Off ice of Research and Development Laboratory
Research Triangle Park, North Carolina 27711
EPA-600/7-77-061
luria 1Q77
JUl ?
SLUDGE OXIDATION
IN LIMESTONE FGD SCRUBBERS
Interagency
Energy-Environment
Research and Development
Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into seven series. These seven broad categories
were established to facilitate further development and application of environmental
technology. Elimination of traditional grouping was consciously planned to foster
technology transfer and a maximum interface in related fields. The seven series
are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the effort
funded under the 17-agency Federal Energy/Environment Research and Development
Program. These studies relate to EPA's mission to protect the public health and welfare
from adverse effects of pollutants associated with energy systems. The goal of the
Program is to assure the rapid development of domestic energy supplies in an environ-
mentally-compatible manner by providing the necessary environmental data and
control technology. Investigations include analyses of the transport of energy-related
pollutants and their health and ecological effects; assessments of, and development
of, control technologies for energy systems; and integrated assessments of a wide
range of energy-related environmental issues.
REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect the
views and policies of the Government, nor does mention of trade names or commercial
products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Information
Service, Springfield, Virginia 22161.
-------
EPA-600/7-77-061
June 1977
SLUDGE OXIDATION
IN LIMESTONE FGD SCRUBBERS
by
Robert H. Borgwardt
U.S. Environmental Protection Agency
Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, N.C. 27711
Program Element No. EHE624
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D.C. 20460
-------
CONTENTS
Figures [[[ iv
Tables [[[ VI
Abbreviations and Symbols ........... ...... ........................... vii
Acknowledgements [[[
1. Introduction ................................................ 1
2. Conclusions ................................................. 4
3 . Recommendat ions ............................................. 7
4 . Theoretical Background ............ . .............. . .......... 9
Oxidation Reaction .................................... 9
Oxygen Absorption ..................................... 10
5. Procedure [[[ 23
Equipment ............................................. 23
Approach .............................................. 25
6. Results [[[ 27
Two-Stage Scrubbing ................................... 27
Single-Stage Scrubbing ................................ 38
-------
FIGURES
Number Page
1 Comparison of crystals produced in RTF
limestone scrubber 3
2 Representation of oxygen diffusion from an
air bubble to a liquid, through the liquid film 12
3 Oxygen transfer factor as a function of liquid
depth and air injection rate in air-sparged towers 14
4 Effect of air injection rate on the oxygen transfer
factor at constant pressure, 20°C 15
5 Effect of pressure on the oxygen transfer factor
at constant air injection rate, 20°C 17
6 Oxygen transfer factors for air sparged-towers
at 20°C 19
7 Oxygen transfer efficiency at 50°C as a function
of liquid depth 21
8 Two-stage limestone scrubbing with forced oxidation
in the first stage 28
9 Pilot plant test conditions for estimating oxygen
transfer efficiency at pH 4.5, h = 3.2 m 33
10 Two-stage scrubbing with lime feed, no oxidizer recycle 35
11 Forced oxidation test at high chloride concentration 37
12 Typical operating conditions for single-stage
limestone scrubber at Shawnee Test Facility 39
13 Pilot plant simulation of single-stage scrubbing
with forced oxidation in the scrubbing loop 41
14 Comparison of 0- transfer efficiencies obtained in
the RTF pilot plant with liquid-film diffusion model 44
15 Penberthy air ej ector 46
iv
-------
FIGURES (Continued)
Number Page
16 Forced oxidation in single-stage scrubber using air
ejector to aerate the EHT at pH 6, h = 1.1 m 48
17 Air feed pressure as a function of air injection
rate for air-sparged tower and Penberthy ejector 49
18 Forced oxidation in single-stage scrubber, using
ejector to aerate the EHT: slurry recycle from EHT 51
19 High limestone utilization test 52
20 Slipstream oxidation test 57
-------
TABLES
Number
1 Limestone Utilizations for Data of Figure 19 53
CaSO,-ZtLO Saturation of Scrubber Feed Liquor:
Single-Stage Scrubbing with Forced Oxidation 54
Annual Waste Sludge Production by a 1000 MW
Coal-Fired Power Plant Equipped with Lime-
stone FGD Scrubbers 59
vi
-------
ABBREVIATIONS AND SYMBOLS
2 3
a = specific interfacial area between air bubbles and slurry, m /m
C = 0- concentration in the bulk liquid phase, g mol/1.
C* = 0,, concentration in the liquid phase at the interface when
saturated, g mol/1.
D = bubble diameter, cm
E = 0_ transfer efficiency of oxidizer defined by Equation (5)
EHT = scrubber effluent hold tank
h = liquid depth in oxidizer during aeration, meters
h = static liquid depth in oxidizer, meters
s
H = Henry's law constant, meters of water/mol fraction of 0,.,
IL. = overall mass transfer coefficient based on the liquid, m/hr
k.. = liquid film coefficient for 0^ diffusion
2
m = air injection rate to oxidizer, g mol of air/hr(m )
m = dry mass of gypsum in EHT, g
M = total SO oxidized in oxidizer, g mol
b>U™ Z
P = absolute pressure, meters water
r = relative saturation of gypsum in scrubbing liquor
R = 1.98 cal/g mol °K
t = time, min
T = temperature, °K
V = volume of slurry in oxidizer, liters
y = mole fraction 0 in air feed to oxidizer
y = mole fraction 0_ in bubbles leaving oxidizer
vii
-------
ACKNOWLEDGEMENTS
The pilot scrubber at Research Triangle Park, N. C., is operated for
the EPA by Monsanto Research, Inc. The contributions of this group, and
Mr. James MacQueen in particular, are gratefully acknowledged. The author
also wishes to express his gratitude to Professor Melbourne Jackson of the
University of Idaho for permission to reproduce some of his data in this
report.
viii
-------
SECTION 1
INTRODUCTION
Tests previously reported by EPA showed that calcium sulfite sludge,
which is the normal product of limestone Flue Gas Desulfurization (FGD)
scrubbers, can be oxidized to gypsum under operating conditions anticipated
for scrubbers designed for utilities burning U.S. high-sulfur coals. The
EPA tests showed that the two-stage scrubbing approach, which is used in
(2)
Japan , can be modified to permit the oxidation to be carried out within
the first-stage scrubbing loop. The first stage pH was controlled at 4.5 in
all of those tests, since laboratory investigations of oxidation rate show a
maximum at this pH value. The conversion to gypsum was thus accomplished at
reasonable air stoichiometries, without addition of catalysts, by air-
sparging at atmospheric pressure.
The objectives of forced oxidation are: 1) to improve the settling
properties of the slurry (gypsum settles 10 times faster than calcium sulfite
and yields a smaller volume of settled sludge); 2) to improve the dewatering
characteristics of the sludge (hopefully to the extent that it can be disposed
of directly as landfill without chemical fixation); and 3) to reduce the
quantity of total waste produced (on the basis of the EPA test results,
reductions of 30-40 percent in total waste production should be possible
when forced oxidation is combined with dry fly ash collection). As a
-------
practical matter, separate collection of the fly ash appears to be a
requirement for achieving all three objectives since the settling properties
of the oxidized scrubber slurry can be dominated by fly ash . The improved
settling and dewatering characteristics of the oxidized slurry are attributed
to the difference in size and shape of the CaS04'2H20 and CaSCyl/2 H20
crystals (Figure 1). The larger, thicker gypsum crystals settle faster and
form a more compact mass than the smaller flat plates of calcium sulfite.
Although several scrubber designs now offered in the U.S. are adaptable
to two-stage operation, most existing systems are not. One objective of the
investigations reported here was to determine the feasibility of conducting
forced oxidation in single-stage scrubbers, which operate at higher pH's of
5.6 to 6.4. The advantages favoring the production of gypsum as the throw-
away product would thus be more widely applicable to present systems without
major modification of the scrubber itself. Another purpose of this study
was to further explore the variables that influence oxidation efficiency—
and thus the amount of air required—with respect to the type of aeration
device used, the oxidizer design, and the operating conditions.
-------
A. WITHOUT FORCED OXIDATION: CaSOa Vit^O x 2040
B. WITH FORCED OXIDATION: CaS04-2H20 x 1020
Figure 1. Comparison of crystals produced in RTP limestone scrubber.
(Scanning electron micrographs by Monsanto Research Corp., Dayton,
Ohio)
-------
SECTION 2
CONCLUSIONS
A sludge suitable for direct disposal as landfill can be produced from
limestone scrubber slurries by forced oxidation to gypsum. This can be
accomplished in systems operating with high sulfur coals, with or without
chloride.
The total waste produced by a power plant equipped with limestone FGD
scrubbers can be reduced by about 45 percent, compared to current practice,
if forced oxidation is combined with dry fly ash collection. A final sludge
containing 87 percent solids can be thus produced.
The volume of settled scrubber sludge can be reduced about 30 percent
by forced oxidation.
Efficient oxidation can be accomplished in either two-stage or single-
stage scrubbers. In the former case, the first-stage pH must be controlled
below 5 to avoid gypsum scaling. When sufficient particle retention times
are provided, gypsum crystals filterable to 80 percent solids and settling
at 2 to 3 cm/min can be obtained with either scrubber type. The longer particle
retention times characteristic of the single-stage scrubber, together with
the neutral pH of the oxidized sludge and the avoidance of pH control, favor
the use of single-stage systems.
-------
The oxidation step should be conducted within the scrubbing loop for
most efficient oxidation and for best physical properties of the gypsum.
Lime scrubber slurries can be efficiently oxidized in a two-stage
scrubber.
The oxidation process is controlled by liquid film diffusion of the 0
and thus the performance of air-sparged oxidizers is predictable from theory
developed for waste water treatment. It should be possible to design full
scale FGD oxidizers of this type with confidence.
Oxygen transfer efficiencies of 30 percent were obtained in the pilot
plant with a 5.5 m (18 ft) air-sparged tower at 50°C. The transfer efficiency
was independent of slurry pH from 4.5 to 6 and independent of the sparger
orifice size from 1.6 to 6.4 mm. Air stoichiometries of about 3 will thus
be required for complete oxidation in oxidizers of this depth. Higher
oxidation towers should yield greater transfer efficiencies and require
lower air stoichiometries.
The greatest oxidation efficiency was obtained with an air ejector.
The higher 0_ transfer efficiency provided by the ejector will permit the
use of lower air stoichiometries and shallower oxidizers than will be possible
with spargers. They will also permit the substitution of a slurry pump for
an air compressor. The performance of large scale aerators of this type,
however, is less predictable than for a sparged tower.
Forced oxidation can be conducted in a single-stage limestone scrubber
while meeting the operating conditions required for maximum scrubber relia-
-------
bility; i.e., at limestone utilizations of 85 percent and scrubber
feed-liquor supersaturations below 1.3.
-------
SECTION 3
RECOMMENDATIONS
In view of the successful results of the EPA-RTP pilot plant tests with
forced oxidation in a single-stage scrubber, and considering the advantages
of this mode of operation compared to the two-stage approach, further tests
on larger scale equipment are both justified and desirable. The forced
oxidation experiments at EPA's Shawnee Test Facility should, if possible, be
extended to include single-stage tests as well as those now in progress with
the two-stage Venturi/Spray tower. Any such tests should include a larger
version of the Penberthy ejector, which performed exceptionally well in the
pilot plant, but cannot be scaled up with certainty.
The design of air-sparged towers as oxidizers for FGD scrubbers appears
straightforward for systems in which oxidation is conducted within the
scrubbing loop. Liquid-film diffusion models that are already available
should be directly applicable to this situation when corrected for temperature.
The sparger itself should use 6.4 mm orifices to minimize air pressure drop
at the injector. Selection of the most economical tower height requires an
accurate knowledge of the relationship between the 0? transfer efficiency
and slurry depth; tests of an air-sparged tower at depths greater than
5.5 meters are thus desirable to provide further verification of the diffusion
model, at 50°C, for CaSO- slurries. Tests of this type could most easily be
carried out in the RTP pilot plant.
7
-------
The tests with oxidizer recycle have shown a strong effect on the perform-
ance of the sparged tower. It is postulated that countercurrent flow between
the rising air bubbles and downward flowing slurry is responsible for the
observed increase in efficiency at high recycle rates. If this interpretation
is correct, it would permit the effective height of an oxidizer to be
increased without penalties of greater pressure for air injection. Thus, the
scrubber effluent entering the hold tank/oxidizer might be suitable as a
source of downward flow at velocities sufficient to retard the bubbles and
prolong contact time. Some carefully controlled measurements of oxygen
transfer factors in an air-sparged tower during countercurrent flow would be
of potential value.
-------
SECTION 4
THEORETICAL BACKGROUND
In general, the problem of oxidizing the calcium sulfite slurry which
is normally produced in limestone FGD scrubbers involves three steps:
1) the dissolution of the solid CaSCL in order to transfer the S0? to the
liquid phase of the slurry where the oxidation reaction occurs; 2) the
chemical reaction of oxygen and SCL in the liquid to produce calcium sul-
fate; and 3) the absorption of oxygen into the liquid by diffusion from the
air that is injected into the oxidizer. The potential limitations of the
first step can be avoided by conducting the oxidation within the scrubber
loop, which is the procedure used in the tests reported here. In this case,
the slurry is continuously recirculated through the low pH region within the
S0? absorber which accelerates the dissolution of the CaSO_. The other two
steps must be taken into account in any attempt to design an efficient
oxidizer for FGD systems.
OXIDATION REACTION
The uncatalyzed rate of oxidation of calcium sulfite slurries has been
(3)
accurately measured under conditions at which the chemical reaction is
the only resistance. This was accomplished in the laboratory by aerating
the reactor with a large excess of air while stirring at high speed, which
-------
minimized the physical resistance imposed by diffusion upon the absorption
of oxygen into the liquid. At 50°C and pH 6 this rate is:
dMso
1 2 3
chemical oxidation rate = — 7— = 1.4 x 10 g mol/l.(min) (1)
13
The chemical oxidation rate increases to a maximum value of 17 x 10 8
mol/l.(min) at pH 4.5, which is the highest rate that can be expected under
any circumstances at 50°C; i.e., when there is no diffusion resistance to
oxygen absorption.
The first criterion for the design of an oxidizer for FGD systems is
that the volume, V (in liters), of the aerated reactor be sufficiently large
to satisfy Equation (1) for whatever values of pH and S00 feed rate, dM n /dt,
/ 9
one is dealing with. The SO- feed rate includes, of course, both the solid
CaSO_ and the dissolved S0?, in g moles per minute.
OXYGEN ABSORPTION
As a practical matter the overall oxidation rate of SO- in large
equipment is usually controlled, not by the chemical oxidation rate, but by
the rate of transfer of oxygen from the air to the liquid. This is because
of the slow rate of diffusion of the oxygen through the liquid film which
surrounds each bubble and imposes a large resistance to its absorption into
the bulk of the liquid where the oxidation reaction occurs. This resistance
arises mainly from the fact that oxygen is only very slightly soluble (5.5
mg oxygen per liter of pure water at 50°C, and 1 atm air pressure). Thus,
the maximum driving force for diffusion through the liquid film is small.
On the basis of pilot plant tests which varied oxidizer depth, it was
concluded that oxygen transfer was the predominant resistance to the overall
10
-------
oxidation rate of calcium sulfite slurries at pH 4.5 . Overall rates
3
typically observed in the RTF pilot plant at 50°C ranged from 1.1 x 10 to
3 3
2.7 x 10 g mol/1. (min) at pH 4.5, well below the 17 x 10 limit imposed by
chemical oxidation rate. This implies that the resistance due to 02
diffusion is greater than that of the oxidation reaction in solution.
0.0 Transfer Efficiency
The mechanics of oxygen diffusion from an air bubble through the
liquid film that surrounds it is depicted in Figure 2, which considers a
single bubble rising through a volume of liquid, V. The rate of oxygen
transfer from the bubble to the liquid is given by:
|£ = Kja (C* - C) (2)
where C is the concentration of dissolved 09 in the bulk liquid and C* is
the 0« concentration at saturation, or limiting solubility. The term a is
the surface area of the bubble in contact with the liquid, and K is the
Li
liquid film transfer coefficient. At steady-state conditions the rate of
transfer to a sulfite slurry equals the rate of SO., oxidation:
dMso
2 2
V (c* - c> - - I -5T1
When diffusion controls (i.e., the rate of 0« transfer through the film is
slower than the oxidation rate), the 0? concentration in the bulk liquid is
zero. For this situation, the integrated form of Equation (2) for the
(4)
air/water system has been shown by Urza and Jackson to be:
E
, 1 + 0.264E _ 70,000 (K,a) h (P + h /2) (4)
I* j_n _ __ — i_i __ s
3 + 0. 791E 1 - E m H
11
-------
LIQUID
LIQUID FILM
LIQUID
I—VOLUME
V
C* = 02 CONCENTRATION AT AIR/LIQUID INTERFACE
C = 02 CONCENTRATION IN THE BULK LIQUID
Figure 2. Representation of oxygen diffusion from an air
bubble to a liquid, through the liquid film.
12
-------
where
E = oxygen transfer efficiency = ^-o ^ (5)
and y , y are the mole fractions of oxygen in the air feed (= 0.209), and
the bubbles leaving the oxidizer; respectively. The term "oxygen transfer
efficiency" is widely used in the literature on waste water treatment as a
basis for expressing and evaluating the performance of aerated reactors;
it is therefore adopted here for the same purpose.
Equation (4) gives the relationship between the oxygen transfer ef-
ficiency and the transfer factor, K a, the air injection rate, m , and the
Li O
total pressure, P, where H is the Henry's Law constant at the temperature of
the liquor in the oxidizer, h is the expanded liquor depth in the oxidizer
during aeration, and h is the static liquor depth in the oxidizer. The
O
value of K a has been accurately determined experimentally by Jackson for
the transfer of 0? to sulfite solutions at liquid depths to 21.3 meters
(Figure 3). Jackson's data were obtained at 20°C for bubbles formed by 6.4-mm
sparger orifices, which produced bubbles of 5 mm average diameter.
At a given temperature the value of the transfer factor, K a, varies
Lj
directly with the surface area of the bubbles that are in contact with a
given volume of liquid. Thus, at a given pressure this area is proportional
to the amount of air injection per unit time, m , and inversely proportional
to the diameter of the bubbles, D, :
b
m
ex ~ (6)
b
The proportional relationship between K a and m is tested in Figure 4,
Li O
using the data of Figure 3. Logarithmic plots of the K a values measured by
13
-------
O h= 4.3m
D h = 8.2 m
h= 12.2m
0 h = 16.2m
• h = 21.3m
1000
AIR RATE (m0), g mol/hr (m2)
1500
2000
Figure 3. Oxygen transfer factor as a function of liquid depth and air injection rate in air-
sparged towers. Data of Jackson(5) at 20°C.
14
-------
70
60
50
40
30
5 20
10
g
8
7
118
116
109
029
077
200
500 1000
AIR RATE (m0),gmol/hr(m2)
1500
2500
Figure 4. Effect of air injection rate on the oxygen transfer factor at constant pressure.
Data of Figure 3, 20°C.
15
-------
Jackson are shown as a function of the logarithm of the air injection rate
at each of five different constant pressure conditions (tower heights). It
is evident from Figure 4 that Ka increases slightly faster than a linear
function of air rate; the average slope is 1.09 rather than the 1.00 ex-
pected from Equation (6). The difference indicates that some breakage of
the bubbles occurs at the higher injection rates, producing additional
surface area.
Pressure also influences the transfer factor. As the tower height is
increased, the average ambient pressure of the liquid upon the bubbles
increases, thus reducing their size and total area. If the density of the
bubbles at the point of formation remains constant as the pressure is in-
creased, then the total surface area would decrease with the two-thirds
power of the ambient pressure. Figure 5 shows the effect of pressure on
1C. a, again based on Jackson's data of Figure 3, at a constant air injection
Lj
rate. The average IL a decreases in direct proportion to the average pres-
sure as the tower height is increased. The pressure plotted in Figure 5 is:
Average absolute pressure = P + h /2
s
where P is the atmospheric pressure, in meters of water, and h is the
s
static liquid depth in the oxidizer above the sparger, also in meters of
water. Most of this pressure effect is accounted for by the change in
volume of the air, and thus the interfacial surface area. It is also evi-
dent that the density of the bubbles formed at high pressure is significantly
greater than the density of bubbles formed at lower pressure.
16
-------
50
40
30
a
< 20
ui
10
m0 = 1290
SLOPE = -1.02
10 20
AVERAGE TOWER PRESSURE, meters water
30
Figure 5. Effect of pressure on the oxygen trans-
fer factor at constant air injection rate. Data of
Figure 3, 20°C.
17
-------
It follows from Figures 4 and 5 that the transfer factor for a sparged
tower (at 20°C) can be expressed empirically as:
1.09
m
., n 0,, o
V = °-214 FT
The accuracy of Equation (7) is correlating Jackson's data is shown in
Figure 6.
From Equations (4) and (7) it is evident that the 0 transfer efficiency
is, for practical purposes, a function only of the liquid depth (or bubble
residence time in the oxidizer) at any given temperature.
Temperature affects the oxygen transfer efficiency in two ways:
1) as temperature increases, the solubility of 0« — and thus the value of H
in Equation (4) — is reduced; and 2) the transfer coefficient for liquid film
diffusion is increased. The magnitude of each of these effects is known.
At a normal FGD scrubber temperature (and therefore, oxidizer temperature)
of 50°C, H = 7 x 10 meters water per mole fraction. The value of 1C a can
be estimated from Equation (7) and the data of Sherwood and Pigford :
50°C = 2.08 Ka @ 20°C
/7QS1- ? I ~\
and a @ 50°C =(|ffJ a @ 20°C
/ \ 1.09
K a = 2.08 ( -77^-] " ,—70 (8)
With Equations (8) and (4) the oxygen transfer efficiency at 50°C can be
represented by:
3.0.791. + * - 0-0*74 . - h (9)
18
-------
10 15 20 25 30
MEASURED K|_a (JACKSON), hr>
35
40
Figure 6. Oxygen transfer factors for air-sparged towers at 20°C. Test of Equation (7) with
data of Figure 3.
19
-------
Equation (9) is plotted in Figure 7 over the range of air injection
rates used in the RTF pilot plant. It shows the relationship between oxi-
dizer height and the oxygen transfer efficiency expected when calcium sul-
fite slurries are air-sparged at 50°C. The correlation includes the effect
of m on the height of the slurry in the oxidation tower, which expands by 1
o
2
and 3 percent, respectively, at mQ = 1000 and mQ = 3000 g mol/hr(m ). The
transfer efficiencies shown in Figure 7 are about 20 percent greater than
those calculated at 20°C. The net effect of higher temperature is thus to
improve the 0 transfer, a result brought about by the relatively large
reduction in film resistance compared to the change in oxygen solubility.
Air Stoichiometry
Equation (9) can be used to estimate the minimum air requirement for
complete oxidation in a given oxidizer configuration when liquid film dif-
fusion controls. The minimum air requirement is defined by an air stoichi-
ometry of 1.0, where
. , . g atoms of 09 injected as air ,,»*
Air Stoichiometry = —-. j—:rr—^— v j "• TC— (10)
J g moles of SO- absorbed in scrubber
(kg/hr air fed) 0.21 (64.1) 2
29 (kg/hr S02 absorbed)
Equations (5) and (10) show that the air Stoichiometry and oxygen transfer
efficiency are related by:
. , . oxidation in oxidizer ,,,v
Air Stoichiometry = —— ^ =-.—; (11)
J 0~ transfer efficiency
where the numerator in Equation (11) is the moles of S09 oxidized in the
oxidizer per mole SO- absorbed in the scrubber. Generally, 15-25 percent of
the S02 absorbed is oxidized in the scrubber, which reduces the amount
20
-------
m0 = IDOOg, mol/hr (m2)
02 4 6 8 10 12 14 16 18 20 22 24
LIQUID DEPTH, meters
Figure 7. Oxygen transfer efficiency at 50°C, as a function of liquid depth. Curves calculated
by Equation (9).
21
-------
of oxygen that must be transferred in the oxidizer. In the experiments
reported here, the oxidation in the oxidizer was determined as:
Oxidation in oxidizer =
Final oxidation of sludge - Oxidation in scrubber
In those situations where the scrubber oxidation could not be separately
determined, such as the experiments with single-stage scrubbers, the tests
were made under conditions which limited the scrubber oxidation to only about
15 percent of the SO absorbed and this value was assumed in using Equa-
tions (10) and (11).
For reasons of operating economy, especially in terms of the power
requirements for air compression, the oxidizer must be operated at an air
stoichiometry as close to 1.0 as practicable. Equations (9)-(11) provide a
basis for oxidizer design where the physical absorption of 0? is the con-
trolling factor. In the following description of the pilot plant tests
maximum effort is made to evaluate and interpret the results in terms of
these relationships.
22
-------
SECTION 5
PROCEDURE
EQUIPMENT
The FGD pilot plant operated by EPA's Industrial Environmental Research
Laboratory at Research Triangle Park (RTP), N.C., consists of two scrubbers,
each 23 cm diam. x 3 m high and each having a flue gas capacity of 8.5 m /min.
They are operated for the purpose of providing in-house experimental support
for EPA's larger prototype-scrubber test facility at TVA's Shawnee power
plant in Paducah, KY. The RTP scrubbers are 1/100 the capacity of the Shawnee
prototypes and 1/1000 full scale.
Two-stage scrubbing experiments were conducted with the two towers
connected in series so that the hot flue gas entered the "first stage," and
then passed into the "second stage," which was the principal S0« absorber.
The limestone feed entered the second stage, from which the partially reacted
slurry was then fed to the first stage so that flue gas and slurry flows were
countercurrent through the two scrubbers. Each stage contained its own
scrubber effluent hold tank (EHT) and slurry recirculation pump. The first-
3
stage SO- absorber was always operated as a spray tower at L/G = 2.7 l./m .
2
The second stage was operated at L/G = 9.3 l./m , either as a spray tower or
as a turbulent contact absorber (TCA). The second-stage EHT was initially
set up with three tanks in series with 9 min total residence time; later
23
-------
tests were made with a single 9-min stirred tank. The first-stage EHT
consisted of an air-sparged tower, containing slurry of variable depth, which
operated in series with a 163-liter stirred tank. The tank served to in-
crease the particle retention time* for growth of the gypsum crystals. The
pH was manually controlled in the first stage by means of the rate of lime-
stone feed to the second stage.
Tests conducted with a single-stage scrubber were made in the TCA con-
figuration, with three 13-cm beds of 3.8-cm diam (5 g) spheres and 2.7 m/sec
gas velocity at L/G = 9 to 11 l./m . The EHT volume was 718 liters, pro-
viding a residence time of 10 minutes. This tank was aerated by either of
two methods: an air-sparged tower or an air ejector. In the latter case, a
single stirred (1725 rpm) tank was used which had a slurry depth of 1.1 meters.
When the sparged tower was used for aeration, it consisted of a 30.5-cm diam
PVC pipe containing 400 liters of slurry at a depth of 5.5 meters; this tower
was followed in series by a 318 liter stirred tank from which the slurry was
pumped back to the scrubber.
The following operating conditions applied to both two-stage and single-
stage testing: inlet SO- concentration in the flue gas = 2700 to 3000 ppm,
inlet 0 concentration in the flue gas = 4 to 6 percent, chloride concen-
tration in the scrubbing liquor = 3000 to 5000 ppm (added as HC1 gas to the
flue gas entering the scrubber), oxidizer temperature = 50°C, and scrubber
slurry = 8 percent solids without fly ash. The sludge was dewatered in a
rotary vacuum filter, 91 cm diam x 61 cm wide, with all filtrate returned to
the scrubber.
*Particle retention time = EHT volume (1.)/slurry purge rate (l./hr)
24
-------
APPROACH
The oxidizer was set up in a manner that permitted the oxidation reaction
to be conducted within the scrubbing loop. With only one exception, the
tests were made without catalyst addition. Since the scrubber towers were
glass-walled and the oxidizer, EHT, and connecting piping were all plastic,
the slurry was not in contact with catalytic surfaces; any catalyst that may
have been present could only have entered the system as an impurity in the
limestone. The limestone used in all tests was obtained from EPA's Shawnee
Test Facility: its composition was 96 percent CaCO.,, 2 percent MgCO_, and
2 percent insoluble inerts.
The goal of the test program was to improve the dewatering properties of
the sludge by means of forced oxidation. The two-stage system, which had
already been successfully tested, was used in the current experiments to
obtain information concerning the effects of pH as an operating variable. On
the basis of experience gained with the two-stage system, the tests were
extended to a single-stage scrubber. Another objective, therefore, was to
establish suitable scrubbing configurations and operating conditions for
applying forced oxidation to all types of scrubbers that will be operating at
power plants in the U.S., and especially those burning high sulfur coal. The
earlier RTP tests showed that complete oxidation of the slurry (at least
90 percent) is required to obtain good sludge properties. The tests reported
here, therefore, had the additional objective of finding an efficient method
of oxidizing the slurry while using the minimum amount of air. Various
oxidation schemes were tested in a manner that permitted estimation of the
minimum air stoichiometry required for complete oxidation.
25
-------
The oxygen transfer efficiency, defined in Section 4, is used as the
basis for estimating the minimum air stoichiometry required for complete
oxidation and for comparing the performance of oxidizers. The oxygen trans-
fer efficiency must be measured at operating conditions which yield incom-
plete oxidation of the calcium sulfite; thus some tests are reported in which
the air injection rate was deliberately limited below the minimum stoich-
iometry. This approach avoids many trial and error tests that would other-
wise be necessary.
26
-------
SECTION 6
RESULTS
TWO-STAGE SCRUBBING
A two-stage scrubber is defined here as a system containing two ab-
sorbers in series, each absorber with its own EHT. Figure 8 illustrates one
type of two-stage system tested at RTF. Although the first stage consisted
of a spray tower in all RTP experiments, a venturi can also be used for the
first stage. Fresh limestone is fed to the second stage where most of the
SO absorption occurs and about 80 percent of the limestone is reacted. The
partially reacted slurry is then fed to the first stage where additional S09
absorption occurs, lowering the pH and further increasing the limestone
utilization. The main advantage of this system, as far as forced oxidation
is concerned, is that it permits the two stages to operate at different pH
values. Thus, the first stage pH can be controlled at low values for maximum
oxidation efficiency, while the second stage operates at the higher pH's
needed for most efficient SO absorption. This arrangement also maximizes
the utilization of the limestone, which averaged 95 percent in the RTP tests.
Forced oxidation testing was begun at RTP using the two-stage—
rather than a single-stage—system because the greatest possibility of
success could be expected at low pH, where the CaSO- solubility and the
27
-------
FIRST
OXIDIZER STAGE
/-^ / v
SECOND
STAGE
^-^
S02 1-94 kg/hr 2770 ppm S02 S02 = 2090 ppm j
Cl - 40 ppm
pH=4.5
SOLIDS
TOTALS AS SOa 51 9 mg/g
S02 17
C02 30
Ca 282
OXIDATION -0.96
UTILIZATION =0.92
10 ^
FILTER
r
^
V-7~
HCIO.O
02 4.3 %
15 liters/min
1
25 liters/min
1
AIR
7.6kg/h
kg/cm 2
(4.4x)
«
£J Ulll
H20
163 1.
9
J
r
f
5.5m
pH
4.5
o
^ .
*•
(24.5%)
A
2.0
m/sec
AP
0.8
cm H20
FOUR
80%
OPEN
GRIDS
pH5.1
\ t
' T~^
VJ7
-t
^
45 liters/hr
i
1 S02 = 470 ppm
(83%)
A
2
.0
m/sec
AP
7.6
cm H20
FOUR
65%
OPEN
GRIDS
pH
I
155
3 min
^
5.3
51 liters/min
Cl - 40 ppm
pH =
5.5
SOLIDS
TOTAL S AS S03 469 mg/g
S02 240
C02 69
Ca
302
OXIDATION =
UTILIZATION =
5.0
1 40% !
1.
102 1.
Cftij O
2 min.
0.36
0.78
kg/hr
1+ FILTRATE
50LIDS
204 1.
4 min.
j
J
'
EHT
Figure 8. Two-stage limestone scrubbing with forced oxidation in the first stage.
28
-------
overall rate of oxidation are greatest. From the experimental standpoint,
the two-stage system is the preferred starting point even though it is
somewhat more complicated, operationally, than a single-stage system. The
initial results with the two-stage system were reported at the New Orleans
FGD symposium (March 1976) and showed that complete oxidation could be ob-
tained at air stoichiometries of 6-7 when a shallow (h = 0.9m) tank was used
as a combination oxidizer/EHT in the first-stage scrubbing loop. By sub-
stituting a tower (h = 5.5m) for the stirred tank, as shown in Figure 8,
complete oxidation was obtained at air stoichiometries as low as 2.3. Aera-
tion was provided in both systems by a sparger containing 22, 1.6 mm ori-
fices. The improvement in oxidation efficiency brought about by increasing
the oxidizer height was attributed to the greater contact time between the
air bubbles and the slurry. This result, which is in accordance with the
effects described by Equation (8), was evidence that oxygen transfer was
controlling the performance at pH 4.5.
Recycle
The importance of air/liquid contact time was further evidenced by
experiments recently completed at RTP, in which the slurry recycle to the
oxidizer was discontinued. This recycle, indicated in Figure 8 by an arrow
entering the top of the oxidizer, was provided for the purpose of mixing the
slurry to prevent settling of solids inside the oxidizer. At normal recycle
rate the slurry moved downward through the 20-cm diameter tower at 340 cm/
min. Tests without recycle showed that it had a significant effect on oxida-
tion efficiency, air stoichiometries of 3 being necessary for complete oxida-
tion at pH 4.5. The observed effect of recycle on oxidation efficiency can
29
-------
be explained in terms of 0 transfer efficiency: the countercurrent flow
imposed by recycle retarded the rising air bubbles and increased the contact
time. The effective height of the tower was increased as a net result of the
recycle.
The 163-1. tank shown in the first stage loop of Figure 8 was added to
provide additional residence time for the growth of gypsum crystals during
the oxidation process. This tank increased the residence time from 7 min to
18 min, based on the slurry recirculation rate, and resulted in a general
improvement in settled density and filterability of the oxidized sludge. The
gypsum thus produced settled at a rate of 3 cm/min, to a final density of
1 gram (dry solids) per ml, and filtered to a sludge containing 75 to 80 per-
cent solids.
Oxidizer pH
The principal objective of the pilot plant tests reported in this
section was to evaluate the importance of pH as a variable influencing the
efficiency of forced oxidation, particularly in terms of the amount of air
required. To put this question in perspective, the two-stage scrubbing
approach was adopted for the initial RTF oxidation experiments on the as-
sumption that low pH is a necessary condition for efficient oxidation.
Numerous laboratory investigations have shown that the oxidation rate of
CaSO., slurries increases as pH is reduced from 7 to 4.5. If the rate of
chemical oxidation in solution is the controlling factor, then two-stage
operation at low oxidizer pH should provide an advantage. If, as the dis-
cussion above implies, the CL transfer from the air to the liquid is control-
ling, then pH would be important only at values greater than that pH at which
chemical reaction rate equals the 0. transfer rate.
30
-------
The first direct indication that low pH may not be a crucial factor for
forced oxidation occurred during experiments aimed at evaluating the effect
of slurry "carry-over." These tests simulated the entrainment of 20 percent
of the first-stage feed liquor into the gas stream. Since the transfer of
this slurry to the second stage must be accompanied by an equal amount of
slurry fed forward from the second stage to the first, the pH is equalized in
the two stages. Despite the higher pH in the oxidizer during the carry-over
tests (pH = 6.0-6.5) complete oxidation was still obtained. This result
prompted further tests intended specifically to determine the effect of the
first stage pH on oxidation efficiency. They were conducted with zero carry-
over, increasing the first stage pH by operating the system at higher lime-
stone feed stoichiometry. The scrubber configuration was that shown in
Figure 8.
The ensuing tests at high limestone stoichiometry showed two important
results: 1) the overall SO- removal efficiency was significantly improved by
increasing the first stage pH—operating with two spray towers in series at
an overall pressure drop of 8 cm water, 85 percent average SO- removal was
obtained with a first stage pH of 6 compared to 76 percent average removal at
pH 4.5; and 2) the first stage tower consistently scaled with gypsum when
operated at pH's greater than 5—controlling the first stage pH below 5 was a
necessary condition for avoiding scaling in the RTF scrubber. The scale
formed on the grids and glass walls, mostly in the upper third of the tower
around the slurry feed nozzle.
It was apparent that the greater S02 removal efficiency that could
potentially be obtained in the two-stage system by operating at high lime-
31
-------
stone stoichiometry could not be realized in practice because of the scaling
problem. As far as the effect of pH on the oxidation efficiency was con-
cerned, the results of these experiments were similar to the carryover tests:
oxidation could still be completed at high limestone stoichiometry and oxi-
dizer pH's as high as 6.5.
The final test with the two-stage scrubber made a direct comparison of
the effect of pH on oxidation efficiency. The recycle to the oxidation tower
was cut off and the air stoichiometry set at 3.0 with an oxidizer height of
3.2 meters. At these conditions, 68 percent oxidation was obtained at pH 4.5,
and 56 percent oxidation was obtained when the pH was increased to 6. The
difference in oxidation rates over this extreme range of pH was insignificant
compared to the difference between the corresponding uncatalyzed chemical
-3 —3
reaction rates: 17 x 10 g mol/1. (min) at pH 4.5 vs_. 1.4 x 10 at pH 6.
Figure 9 summarizes the operating conditions of the test at pH 4.5; the
oxidation rate was far slower than would be expected if chemical reaction
were limiting:
•A • 1900(0. 83) (1 - 0.245)0.68 „ . _ -3 ,/-,/. N
oxidation rate = - ^ — 102(60)64 1 — - = 2.1 x 10 g mol/1. (min).
It is therefore evident that oxygen transfer, rather than the oxidation
reaction, was controlling. Since Equations (9)-(ll) apply in this case:
oxygen transfer efficiency = oxidation ^ oxidizer
(h = 3 2m) air stolcni°metry
The minimum air stoichiometry required for complete oxidation with an oxi-
dizer height of 5.5 meters* would be:
^Transfer efficiency is taken to be. directly proportional to oxidizer height
for h<6 meters, per Figure 7.
32
-------
SO? 1.90 kg/hr 280° PPm S02
HCI 0.0
02 5.7%
15 liters/min
Cl = 230 ppm
pH = 5.0
SOLIDS
TOTAL S AS 803 536 mg/g
S02 133
C02 10
Ca 280
OXIDATION = 68 mol %
UTILIZATION =96 mol %
FILTER CAKE =
64% SOLIDS
|
AIR
4.7 kg/hr
1 0.46^^^^^
kg/cm2
(3.0x)
13 cm
H20
163 1.
1 ,-*.»,
13- PH4.8
I^^BI
«
J
i
•••
^*-
3.2m
51°C
1021.
PH
4.5
O
H^
A
2.0
m/sec
AP
3
cm H20
FOUR
80%
OPEN
GRIDS
pH4.2
H
45 liter
1502=470 ppm
*»
s/hr
d
A
2.0
m/sec
AP
9
cm H20
FOUR
65%
OPEN
GRIDS
pH5.6
1
ll.
155 1.
3
51 liters/min
Cl-90ppm
PH =
6.0
SOLIDS
TOTAL S
AS SO 3 462 mg/g
S
u-
' LI
a
CO? 86
Ca
OXIDATION
299
= 24.5 mol %
UTILIZATION =77 mol %
6
._._ LIMFtTfl
415
102 1.
-Jo -
.2
Ml
'oJ
o
kg/hr
E+ FILTRATE
SOLIDS
204 1.
«
^
-tJ
3 min.
2 min.
4 min.
Figure 9. Pilot plant test conditions for estimating oxygen transfer efficiency at pH 4 5
h = 3.2m.
33
-------
air stoichiometry = (1 - 0.245) 3.2
(h = 5.5m) 0.145 x 5.5
Another test, similar to the one above, was conducted at pH 6 with MnSO,
I |
catalyst added to provide a concentration of 40 ppm Mn in the oxidizer
liquor. If the chemical reaction rate were limiting, the test would be
expected to show an acceleration of the rate of oxidation relative to the
uncatalyzed value of 1.4 x 10 g mol/1.(min). Operating at 5.5 meters
_3
oxidizer height and without recycle, the observed rate was 1.3 x 10 g
mol/1.(min), the same as the uncatalyzed value. The test verified that
oxygen transfer was also limiting at pH 6, since the overall rate was not
influenced by catalysis. The test was conducted at an air stoichiometry of
1.7, at which the final oxidation was 62 percent and the initial (scrubber)
oxidation was 12 percent. Thus, the
oxygen transfer efficiency _ 0.62 - 0.12 _ n oo
/i r r- \ — -i-? — u.^y.
(h = 5.5m) 1. 7
The estimated minimum air stoichiometry required for complete oxidation is:
air stoichiometry = 1.0 - 0.12
(h = 5.5m) ~ 0.29 ~
It is concluded from these tests that the effect of pH (more specifi-
cally, the effect of the chemical oxidation reaction in solution) is insig-
nificant compared to the effect of oxygen transfer on the overall rate of S02
oxidation in the pilot plant at pH levels up to at least 6.
Lime Feed
The two-stage scrubber was operated successfully with lime feed at the
conditions shown in Figure 10. The pH in the first stage was controlled
manually at 4.5, which yielded an average pH of 7.5 in the second-stage
34
-------
S02l.9kg/hr 2800 ppm S02
Cl- 30 ppm
pH = 4.5
SOLIDS
TOTAL S AS S03 556 mg/g
S02 20
C02 4
Ca 285
OXIDATION =0.96
UTILIZATION = 0.97
FILTER CAKE =
84% SO LIDS
I
HCI 0.0
02 3.6%
A
7.6
kg/cm2
(4.6x)
23cm
H20
163 1.
IR
kg/h
••••••
•
«
J
-•«
15 liters/min
mm
h-
5.5m
pH
4.5
O
L1
*•
3-
502 = 2300 ppm ^S02
A
(18%)
2.0
m/sec
AP
2.5
cm H20
FOUR
80%
OPEN
GRIDS
pH4.2
i
i
-*-
45 liters/hr
I
A
2.0
m/sec
AP
16
cm H20
FOUR
65%
OPEN
GRIDS
pH5.5
= 500 ppm
(82%)
51 liters/min
SOLID
TOTALS AS SO
S02
C02
Ca
'OXIDATION =
UTILIZATION =
8.17kg/hr
CaO+FILTRAT
19.6% SO LI OS
460 1.
pH = 7.5
S
3 535 mg/g
321
57
311
0.25
0.86
E
\^s
9 min.
Figure 10. Two-stage scrubbing with lime feed, no oxidizer recycle.
35
-------
scrubber. Automatic control based on a second-stage pH of 7.5 may be
feasible; however, an attempt to operate the RTF scrubber on automatic
control at pH 8 resulted in the same pH in both stages with poor oxidation
and severe first-stage scaling. It is clear that the second stage pH must
be kept below 8; when this is done, good oxidation can be obtained in the
first stage to produce gypsum of excellent settling and dewatering proper-
ties.
Chloride
An example of a run made with no chloride is shown in Figure 8. It was
previously shown that forced oxidation can be conducted at the normal chlo-
ride levels of 3000-5000 ppm expected at Shawnee , and most of the RTP
testing was conducted within this range. A test at very high chloride is
shown in Figure 11. No adverse effect on oxidation efficiency was evident
with 14,000 ppm chloride in the oxidizer liquor.
36
-------
SC»2l.94kg/hr 2890 ppm S02
S0£ = 2320 ppm
S02 = 550 ppm
HCI YhS •
02 5.9%
15 liters/min
Cl = 14,300 ppm
pH = 4.4
SOLIDS
TOTALS AS S03 515mg/g
S02 14
C02 7
Ca 284
OXIDATION =0.97
UTILIZATION = 0.91
FILTER CAKE =
83% SOL IDS
?
>
A
7.6
liuii
v^ »
kg/cr
(3.7
20cm
H20
163
»
1.
o
IR
kg/h
h
n2
x)
I
r
«••
5.5 m
PH
4.2
-O
c
»•
G
(20%)
A
2.0
m/sec
AP
0.8
cm H20
FOUR
80%
OPEN
GRIDS
pH4.4
I
H
^
i
A
2.0
m/sec
AP
7.6
cm H20
FOUR
65%
OPEN
GRIDS
pH
45 liters/hr
5.4
\u i /o;
51 liters/min
6.81 kg/h
LIMESTONE + F
34% SOL IDS
...1
460 1.
r
LTRATE
_J
Cl = 8,040 ppm
pH = 5.6
SOLIDS
TOTALS AS S03 462 mg
S02 235
C02 86
Ca 303
OXIDATION =0.36
UTILIZATION = 0.76
9 min
Figure 11. Forced oxidation test at high chloride concentration.
-------
SINGLE-STAGE SCRUBBING
The importance of the pH question, as addressed in the two-stage
oxidation experiments described in the previous section, has to do with the
prospects for applying forced oxidation to single-stage scrubbers. That
prospect is clearly unfavorable if pH 4.5 is required. The situation is
different, however, if oxidation can be carried out at pH 6 which is within
the range of normal operation of single-stage limestone scrubbers.
As indicated in Section 4, laboratory measurements of the uncatalyzed
_3
oxidation rate of CaSCL slurries are reported to be 1.4 x 10 g mol/1. (min)
at pH 6 and 50°C. The two-stage scrubbing experiments at RTF confirmed that
oxidation rates of at least this magnitude can be sustained in an uncata-
lyzed scrubber under actual operating conditions. If the constraint speci-
fied by Equation (1) can be satisfied with respect to oxidizer volume in a
single-stage scrubber operating at pH 6, then it should be possible to force
the oxidation to completion as long as sufficient oxygen transfer efficiency
is provided. The simplest approach to providing the oxygen transfer would
be to aerate the scrubber effluent hold tank (EHT). In this situation
Equation (1) demands that the EHT have sufficient volume so that the SO
— 3
feed rate does not exceed 1.4 x 10 g mol/1.(min).
Typical operating conditions for a single-stage limestone scrubber are
shown in Figure 12, based on operating experience at the EPA/Shawnee Test
Facility. The SC^ absorbed in the scrubber and fed to the EHT is:
580,000 (1-0.08) 0.90 (0.003) _ = 41.5 g mol/min.
38
-------
S02
ABSORPTION = 90 PERCENT
FLUE GAS
580,000 l/min
(150°C, 8 PERCENT MOISTURE)
S02 = 0. 3 PERCENT DRY BASIS
I
A
SO 2
SCRUBBER
SCRUBBER FEED LIQUOR
4,540 l/min
SCRUBBER EFFLUENT
HOLD TANK
VOLUME =45,400 I
(lOmin)
Figure 12. Typical operating conditions for single-stage lime-
stone scrubber at Shawnee Test Facility.
39
-------
With a 10-min EHT residence time, the maximum oxidation rate required for
complete conversion of this amount of SO^ is:
/ -I r _ o
oxidation (EHT feed) rate = .q,' = 0.91 x 10 g mol/1.(min)
Since this rate is less than the rate observed at pH 6, both in the labora-
tory and in the RTF scrubber, the oxidation efficiency in a single stage
system of this type should be a function only of the 0™ transfer efficiency
in the aerated EHT. As far as the chemical oxidation rate is concerned,
Equation (1) will be satisfied for the system shown in Figure 12 at EHT
residence times as short as 6-1/2 minutes, even if no oxidation occurs in
the scrubber itself and no catalysts are added. The experiments reported in
this section were carried out to determine if oxidation can be forced to
completion in a single-stage system under realistic operating conditions
when good oxygen transfer efficiency is provided.
Air-Sparging Tests
Figure 13 summarizes a single-stage scrubbing test in which the EHT
2
was aerated at a slurry depth of 5.5 meters (oxidizer area = 0.0729 m )
which, as previously shown, has an oxygen transfer efficiency of 0.29.
A second stirred tank was added, as shown in the figure, to provide a total
residence time of 10 minutes. The oxidation tower was sparged with air at
2
a rate of m = 3260 g mol/hr(m ) using 22 orifices of 1.6 mm diameter. As
shown by the results in Figure 13, 97 percent oxidation was obtained at an
air stoichiometry of 2.9 while operating at a limestone stoichiometry of
1.3 (pH = 6.1). The oxidation rate was:
oxidation rate = 97 = 1.4 x lO'3 g mol/1. (min)
40
-------
83%
S02REMOVAL
S02 2.67 kg/hr
HCI 73 g/hr
pH5.1
T
87 l./min
AIR
6.9
kg/hr
18
cm
H20
5.5 m
400 1.
LO
5550 ppm Cl
pH6.1
SO LIDS (8%)
TOTAL S AS S03 470 mg/g
S02 11
C02 46
Ca 275
OXIDATION =97mol%
UTILIZATION = 85 mol%
SETTLING RATE =
2.6 cm/min
SETTLED DENSITY =
0.9 g/ml
FILTER CAKE =
80% SO LIDS
12.8 kg/hr
LIMESTONE + FILTRATE
38% SOLIDS
EFFLUENT
HOLD
TANK
318 1.
OXIDATION
TOWER
Figure 13. Pilot plant simulation of single-stage scrubbing with forced oxidation in the
scrubbing loop; pH = 6.1, air stoichiometry = 2.9.
41
-------
For an accurate estimate of the oxygen transfer efficiency, another run was
made at a lower air stoichiometry of 2.2 at which the oxidation was incom-
2
plete. The air injection rate in this run was 2370 g mol/hr(m ) which
yielded 75 percent final oxidation. Assuming 15 percent oxidation in the
scrubber, the oxygen transfer efficiency is:
f ff- • 0.75 - 0.15 97
oxygen transfer efficiency = -r—^ = U.z/.
(h = 5.5 m)
The air stoichiometry required for complete oxidation, which agrees well
with the values previously obtained with the two-stage scrubber, is:
• i.. 1.0 - 0.15
air stoichiometry = „ = J.l
It is concluded that oxidation can be forced to completion at reason-
able air stoichiometries in a single-stage scrubber operating at normal
limestone stoichiometry. In accordance with Equation (9) one could expect
oxygen transfer efficiencies greater than 0.27 if the EHT were designed with
a slurry depth greater than 5.5 meters. For example, a tank with a slurry
depth of 10 meters (33 ft) would provide an oxygen efficiency of 0.5, re-
quiring an air stoichiometry of only 1.7 for complete oxidation.
Orifice Size—
(4)
Urza and Jackson found that the oxygen transfer efficiency of an
air-sparged tower was independent of the sparger orifice diameter within the
range of 6.4 to 3.2 mm. The sparger used in the tests conducted at RTF
consisted of 2.5-cm diam. PVC pipe containing 22, 1.6 mm orifices formed by
drilling through the pipe wall. It was expected that the smaller holes
might give higher I^a values, and thus better transfer efficiency, by reducing
42
-------
the size of the bubbles formed. It is evident that the transfer efficiencies
obtained at RTP (0.29 and 0.27 with a slurry depth of 5.5 meters) are not
significantly different from the values predicted on the basis of Jackson's
data for 6.4 mm orifices, shown in Figure 7. It may be concluded, there-
fore, that the transfer efficiency is not affected by the size of the sparger
orifices even when the range is extended to 1.6 mm.
The above conclusion was tested by experiments with a sparger contain-
ing 22, 6.4 mm orifices. The results of this test showed 90.9 percent
oxidation in the 5.5 m tower at an air stoichiometry of 2.3 and m = 2660 g
2
mol/hr(m ). Assuming 15 percent oxidation in the scrubber, the oxygen
transfer efficiency was 0.33, confirming that no loss of oxidation efficiency
was incurred by increasing the size of the sparger holes. The scrubbing
configuration used in this test was the same as that shown in Figure 13,
with a pH of 6.1 in the EHT.
Figure 14 compares the oxygen transfer efficiencies obtained in the RTP
pilot plant, using various sparger orifices and tower heights, with the
values predicted by Equation (9) . The data agree well with the model based
on liquid-film diffusion control, within the range of oxidizer heights
tested. Since the transfer efficiency is independent of orifice diameter
within the range of 1.6 to 6.4 mm, the larger size should be used because of
the lower air pressure required for any given air injection rate. Thus, the
overall power input for air compression will be minimized.
Chloride--
It was shown above that slurries containing from zero to 14,000 ppm
chloride could be oxidized in the two-stage scrubber. Additional tests
43
-------
0.9
0.8
0.7
0.6
0.5
UJ
o
£ 0.4
CO
CD
0.3
0.2
0.1
EQUATION (9)
= 3000gmol/hr(ni2)
ORIFICE
O 1.6 mm W/CI
O 6.4 mm W/0 Cl
A 6.4 mm W/CI
• 1.6 mm W/0 Cl
8 10 12 14
LIQUID DEPTH, meters
16
18
20
22
24
Figure 14. Comparison of 02 transfer efficiencies obtained in the RTP pilot plant with the.
liquid-film diffusion model at 50°C.
-------
were made with the single-stage scrubber, measuring the 0_ transfer ef-
ficiency in the absence of chloride. Using a 5.5-m slurry depth, 6.4 mm
orifices, and an air stoichiometry of 2.28, the total oxidation was 87 per-
cent. By Equation (11) the 02 transfer efficiency was 0.31 which is not
significantly different from the efficiencies obtained when chloride was
present. Contrary to the results reported by other investigators, that
bubble size is influenced by ionic strength, it is concluded from the IERL-
RTP tests that oxidation efficiency will not be affected by chloride (which
increases the ionic strength).
Air Ejector Tests
In addition to the air-sparged tower, an air ejector was also tested as
a means of aerating the EHT in a single-stage scrubber. The ejector, shown
in Figure 15, was a Penberthy model 164A manufactured by Houdaille Indus-
tries, Inc. It was made of bronze and was 22 cm long. The ejector was
mounted in the side of the EHT near the bottom of the tank. Slurry was
pumped through it at a rate of 98 liters/min at an inlet pressure of about
9
1.4 kg/cm . The high velocity of the slurry in the 10.3 mm nozzle aspirated
air into the liquid stream. The high shear thus developed in the throat of
the ejector broke the air into minute bubbles which were ejected horizon-
tally across the bottom of the EHT and dispersed upward, aerating the slurry
in the tank. The main feature of the ejector was its ability to create
smaller bubbles than could be obtained with a sparger, resulting in higher
transfer factors in accordance with Equation (6). The greater transfer
efficiencies thus produced should permit a reduction of tower height and/or
a reduction of air stoichiometry.
45
-------
AIR IN
SLURRY
IN
TO
HOLD TANK
Figure 15. Penberthy air ejector.
46
-------
The ejector tests were conducted in the manner shown in Figure 16,
aerating the EHT (h = 1.1 m) so that oxidation was forced in the slurry
recirculating within the scrubbing loop of a single-stage TCA scrubber. As
shown by the results summarized in Figure 16, 99% oxidation was obtained at
an air stoichiometry of 1.8. As pointed out in the description of the two-
stage tests, this same system required an air stoichiometry of 6 to 7 for
good oxidation when aerating by air-sparging. The ejector clearly produced
a marked improvement in oxygen transfer. Tests of the ejector made with the
EHT stirrer turned off (normal stirring speed = 1725 rpm) required an air
stoichiometry of 2.7 for complete oxidation, indicating that the stirring
contributed to the overall performance of the ejector. These tests also
showed that the solids could be maintained in suspension without any mixing
other than that provided by the ejector.
Air Feed Pressure—
In addition to the high oxygen transfer factors associated with ejector
aeration, which permit efficient oxidation to be achieved in shallow tanks,
this approach also has the advantage of a lower air pressure requirement.
This derives from the reduced pressure developed by the high velocity of the
slurry in the ejector throat, which aspirates air into it. Figure 17 com-
pares the air pressure required for the various methods of injection tested
at RTF. It shows that the pressure required by the ejector is significantly
lower than that needed for air-sparging at any given air injection rate.
One objective of the ejector tests was to reduce the air feed pressure to
the range that can be provided by air blowers. Successful tests were made
2
at air feed pressures as low as 0.07 kg/cm (1 psig), which is well within
47
-------
S02
2.47 kg/hr
4.7% 02
2870ppmS02
t
525 ppm S02
(82%)
A"
2.7
m/sec
AP
15
pH 5.2
5290 ppm Cl
pH6.0
SOLIDS (9.4%)
TOTAL S AS S03 443 mg/g
S02 0
C02 56
Ca 282
OXIDATION = 100mol%
UTILIZATION = 79mol%
SETTLING RATE =
3.2 cm/min
SETTLED DENSITY =
0.90 g/ml
FILTER CAKE =
82% SO LI OS
12.8 kg/hr LIMESTONE + FILTRATE
37% SOLIDS
AP
12cm
H20
kg/cm2
AIR 3.93 kg/hr
Figure 16. Forced oxidation in single-stage scrubber using air ejector to aerate the EHT
at pH 6, h = 1.1 m.
48
-------
5.5m TOWER
1.6 mm ORIFICES
5.5 m TOWER
6.4 mm ORIFICES
AIR EJECTOR
1.1 mTANK
300
AIR FEED RATE, gmol/hr
Figure 17. Air feed pressure as a function of air injection rate for air-sparged tower
and Penberthy ejector.
49
-------
that range. Thus, the substitution of an air blower—and a slurry pump—for
an air compressor should be feasible in systems employing ejectors.
Most of the tests were conducted with the ejector feed slurry taken
from a separate shallow tank located at the bottom of the tower, as shown in
Figure 16. This arrangement allowed for the oxidation of some of the SO^
while it was still concentrated in the liquor coming out of the absorber.
Successful tests were also conducted without that tank, with the ejector fed
with slurry recycled from the EHT, as shown in Figure 18. The latter con-
figuration required somewhat higher air feed pressures as indicated by the
data of Figure 17 (open circles).
Limestone Utilization—
Fouling of the mist eliminators has traditionally been the chief ob-
stacle to good operating reliability of limestone scrubbers. Experience at
the EPA/Shawnee Test Facility has shown that the fouling problem is
closely related to the level of limestone utilization at which the scrubber
is operated; fouling of the mist eliminator does not occur when the utili-
zation is maintained above 85 percent (or the limestone stoichiometry is
maintained below 1.18). It is therefore important to determine whether a
limestone scrubber, in which oxidation is forced to completion within the
scrubber loop, can operate at a limestone utilization of 85 percent or more.
Successful tests were made at this level of utilization while operating the
RTF scrubber at 98 percent oxidation. One of these tests is summarized in
Figure 19; the limestone utilizations calculated from these data are given
in Table 1 for each of the three possible methods of measurement.
50
-------
S02
2.38 kg/hr
5.1% Oz
2680 ppm S02
450 ppm S02
(83%)
A
2.7
m/sec
AP
13
cm H20
QQOOQ
IAAJUUL:
pH5.3
72 liters/min
FOUR
80% OPEN
GRIDS +
THREE
20-cm
BEDS OF
3.8 cm
(5g.)
SPHERES
12.7 kg/hr LIMESTONE + FILTRATE
10.46 kg/cm2
AIR
7.9 kg/hr.
(3.7x)
14
E.H.T.
718 1.
(lOmin)
—€
pH6.1
5020 ppm Cl
SOLIDS 0.2%)
TOTAL S AS S03 453 mg/g
S02
C02
Ca
5
57
276
OXIDATION =99mol%
UTILIZATION = 82mol%
SETTLING RATE =
3.7 cm/min
SETTLED DENSITY =
l.lg/ml
98 l./min
Figure 18. Forced oxidation in single-stage scrubber, using ejector to aerate
the EHT: slurry recycle from EHT,
51
-------
S02
2.59 kg/hr
HCI
73 g/hr
2860 ppm S02
280 ppm S02
(90%)
2.7
m/sec
AP
40
J3QOOCX
pHS.1
87 1./min
12.3 kg/hr LIMESTONE + FILTRATE
35% SOLIDS
Y-|
u.
1
I
^l
EJECTOR
X_
\
EHT
M81.
C
^*f
o
0.41
kg/cm
AIR 6.63 kg/hr
<2.6x)
Figure 19. High limestone utilization test.
5540 ppm Cl
pH6.1
SOLIDS (8.3%)
TOTAL S AS SOa 498 mg/g
S02
C02
Ca
35
265
OXIDATION = 98 mol%
SETTLING RATE =
3.1 cm/min
SETTLED DENSITY =
0.97 g/ml
FILTER CAKE =
77% SO LIDS
52
-------
TABLE 1. LIMESTONE UTILIZATIONS FOR DATA OF FIGURE 19
Basis of Measurement
Utilization, mol percent
Solids S/Ca mol ratio
Solids C02/Ca mol ratio
Scrubber material balance
93
88
90
It was concluded that a limestone scrubber can be operated at the utiliza-
tion levels required for the prevention of mist eliminator problems while
forcing oxidation in a single-stage system.
The pH of the aerated slurry in the EHT was higher than the pH's ob-
served without aeration. For example, at 85 percent limestone utilization
and with 5000 ppm Cl in the scrubbing liquor, a pH of 5.6 would be expected
when operating without forced oxidation, based on experience at both the
Shawnee Test Facility and the RTF pilot plant. As indicated in Figure 19
pH's of 6 were consistently obtained when forcing oxidation at utilizations
up to 90 percent. The main factor influencing the scrubber feed pH during
these tests was not limestone utilization, but air stoichiometry. The pH
increased with -the air injection rate, to values as high as 6.6 at an air
stoichiometry of 5.
No scaling of the tower occurred during any of the forced oxidation
tests with the single-stage scrubber, which were conducted during 4 months
of operation. Normal test conditions for the scrubber were L/G =9.8 li-
o
ters/m , EHT residence time = 10 min, and slurry solids = 8 percent without
fly ash. Average CaSO,-2H?0 saturations of the scrubber feed liquor, as
determined by chemical analysis, are shown in Table 2.
53
-------
TABLE 2. CaSO • 2H00 SATURATION OF SCRUBBER FEED LIQUOR:
4 2.
SINGLE-STAGE SCRUBBING WITH FORCED OXIDATION
Constituent
Ca
Mg
so3
SO.
4
co3
Cl
PH
Ion balance*, % error
CaSO -2H 0
Saturation*, %
Concentration,
Nov
1-5
2480
474
13
1504
89
4790
6.0
-2.9
103.3
Nov
8-12
2630
545
169
1404
106
5200
5.8
-2.2
95.2
Nov
15-19
2510
662
0
1520
106
5440
6.2
-3.6
96.0
ppm
Nov
22-24
2180
741
39
1651
93
4750
6.1
-0.2
95.2
Dec
6-10
2330
806
153
1626
119
5550
6.1
-6.7
92.6
Within the accuracy of the sampling and analytical methods used, the
scrubber feed liquor had no significant supersaturation with respect to
dissolved gypsum at 50°C. Saturation tests at 25°C showed relative satura-
tions in the range of 1.1 to 1.2, which agrees with the analytical data when
corrected for temperature; i.e., about 1.0 saturation. The absence of
supersaturation is attributed to the high concentration of pure gypsum seed
crystals in the EHT slurry when oxidation is forced to completion. It is
interesting to compare these values with the saturation level that would be
expected on the basis of measurements of the crystallization rate of pure
/Q\
gypsum, which is :
CaS04'2H20 crystallization rate (g mol/min) =
m (r-1) 2.1 x 109 e "
(12)
^Calculated with Bechtel-Modified Radian Equilibrium Program at 50°C.
54
-------
where m = the total grams of CaSO^I^O (dry solid) present in the EHT,
r = the relative saturation of dissolved gypsum in the liquor, R = 1.98 cal/g
mol °K, and T = temperature, °K. At the operating conditions used in the RTF
tests, the crystallization rate (= SC>2 absorption rate x oxidation) is:
2600 (0.85) 0.97 ... .
(64.1) 60 = °'555 8 mol/mm
and the total gypsum crystals present in the EHT (at 8 percent solids and
718 liters volume) is:
80 (718) = 57,400 grams CaS04«2H20
which, with Equation (12), gives:
relative saturation = 1.04 at 50°C
Thus the observed saturation levels are in agreement with the values expected
on the basis of laboratory measurements of the crystallization rate of gyp-
sum. It was concluded from these tests that forced oxidation can be con-
ducted within the scrubber loop of the single-stage scrubber at pH 6 without
scaling. The difference between this result and the less favorable result
obtained with the two-stage system at pH 6 is attributed primarily to the
higher liquid-to-gas ratio at which the single-stage scrubber operates.
PROPERTIES OF OXIDIZED SLURRY
The physical properties of the oxidized slurries obtained with the
single-stage scrubber operating at pH 6 were similar to those obtained with
the two-stage scrubber operating at pH 4.5. These properties, which were
reported at the 1975 FGD symposium in New Orleans , were 2 to 3 cm/min
settling rate and 0.9 to 1.0 g/ml settled (dry bulk) density. Compared to
sulfite slurries this represents an improvement by a factor of 10 in settling
rate and a factor of 2 in density. As a result of the greater sludge density,
55
-------
the volume is also reduced by about 30 percent. Most importantly, however,
the oxidized slurry was filterable to 80 percent solids compared to 62 per-
cent solids normally obtained with sulfite slurries at RTF; the sulfite
slurries produced at Shawnee have generally been even less filterable,
averaging only 50 percent solids. The filterability of the solids obtained
with the single-stage system were generally better than that obtained with
two stages—a result that may be associated with the longer particle re-
tention times in the larger EHT of the single-stage system.
Attempts to conduct the oxidation outside the scrubber loop (slipstream
aeration) showed that higher air stoichiometries were necessary to complete
the oxidation. Using the Penberthy ejector to aerate a tank of the same
volume (718 liters) as that used for the EHT in the single stage oxidation
tests yielded only 70 percent oxidation at 3.5 air stoichiometry when tested
in the slipstream mode. Figure 20 shows the scrubber configuration for this
_3
test. The oxidation rate in this case was only 0.35 x 10 g mol/l.(min),
which suggests that the dissolution of the solid CaSO- may be the limiting
factor. In this situation all of the SO- absorbed in the scrubber must be
redissolved at high pH, which is not the case when oxidation is conducted
within the scrubbing loop. By increasing the air stoichiometry to 5.4,
96 percent oxidation was obtained in the slipstream. The properties of the
gypsum thus produced, however, were inferior to those obtained by oxidation
within the scrubbing loop. It was concluded from these results that forced
oxidation must be conducted within the scrubbing loop in order to obtain the
best-settling, most filterable sludge with the least amount of air injected.
56
-------
S02
2.5 kg/hr
PRODUCT SOLIDS
TOTALS AS SOa 407 mg/g
S02 99
C02 83
Ca 292
OXIDATION =70mol%
UTILIZATION = 70 mol%
5.4% 02
OXIDIZER
718 1.
50°C
2430 ppm S02
440ppmS02
(82%)
2.7
m/sec
S9SS9P
AP
14
cm H20
JJUULXJJ
pH5.2
72 liters/mi n
FOUR
80% OPEN
GRIDS +
THREE
20-cm
BEDS OF
3.8 cm
(5g.)
SPHERES
pH5.6
3950 ppm Cl
SOLIDS (8%)
TOTALS AS 803 412 nig/!
S02 259
C02 94
Ca 294
OXIDATION =21mol%
UTILIZATION = 70 mol%
12.8 kg/hr
LIMESTONE + FILTRATE
38% SO LI OS
1 liter/min 72 liters/min
E.H.T.
430 1.
6 min
TO FILTER
AIR
7.8 kg/hr (3.5x)
Figure 20. Slip stream oxidation test.
57
-------
SECTION 7
DISCUSSION OF RESULTS
The most important result of these experiments is the attainment of
80 percent solids in the filter cake when oxidation is forced from calcium
sulfite to gypsum. This degree of dewatering has been demonstrated in the
pilot plant with both single-stage and two-stage scrubbing configurations.
A final sludge dryness of 80 percent solids is important because it is ex-
pected to meet the minimum requirement for direct disposal as landfill. It
therefore opens the possibility that chemical fixation of the sludge may be
avoided. The disposal of dry fly ash at the same site will further increase
the dryness of the sludge to 87 percent solids—considerably higher than can
be achieved by any other method, and well within the requirements for stable
landfill of optimum compaction characteristics.
On the basis of IERL-RTP results with single-stage and two-stage oxida-
tion, Table 3 projects the effect of the improved dewatering properties of
the sludge upon the total waste production in a 1000-MW power plant (3.5 per-
cent sulfur and 12 percent ash). It shows that a potential reduction of
47 percent is possible in the total waste produced by the power plant when
forced oxidation is employed. The base case for this comparison represents
(9)
current practice as defined in the SOTSEP report . Another factor that is
sometimes overlooked is that forced oxidation also reduces the volume of the
sludge, in addition to reducing the total tons produced per hour. In our
experience this reduction in volume amounts to about 30 percent.
58
-------
TABLE 3. ANNUAL WASTE SLUDGE PRODUCTION BY A 1000-MW COAL-FIRED3
POWER PLANT EQUIPPED WITH LIMESTONE FGD SCRUBBERS
(SHORT TONS)
Fly Ash Collection
Dewatering Procedure
Limestone Utilization,
mol %
Oxidation, mol %
Coal Ash, Dry
CaS03'l/2 H20
CaC03
Solids Moisture, %
Total Wet Sludge
Total Waste
Reduction of Waste
Compared to Base Case, %
Final Sludge°
Density, % Solids
Wet
Settling
60
10
338,000
322,000
48,000
185,000
50
1,790,000
1,790,000
Base Case
50
60
10
338,000
322,000
48,000
185,000
50
1,110,000
1,448,000
19
63
Dry
Oxidation/Filtration
80
97
338,000
7,200
463,000
69,000
20
673,800
1,011,800
44
87
95
97
338,000
7,200
463,000
14,600
20
606,000
944,000
47
87
Coal = 12% ash, 3.5% sulfur
bSOTSEP Report^ , p. 60.
CDry fly ash blended with wet sludge
59
-------
The importance of pH as a variable influencing forced oxidation is
evidently much less than previously supposed. The results of IERL-RTP tests
with the two-stage scrubber at oxidizer pH's of 4.5 and 6 did not show a
significant difference within the accuracy of the measurements of oxidation
rate. Nor was the performance of the two-stage system at low pH superior to
the oxidation efficiency obtained in the single-stage system at high pH. The
argument that low pH might provide better performance if the tests were made
with more efficient aerators having a greater transfer factor is refuted by
the single-stage test results. The single-stage tests show conclusively that
the oxidation rates measured in the laboratory—and confirmed in the RTF
pilot plant—are great enough even at pH 6 to accomplish complete oxidation
in systems of current design without catalyst addition. The only apparent
constraint is that the physical resistance to oxygen absorption must be
overcome, either by using large amounts of excess air or by providing for
efficient oxygen transfer in the oxidizer. In the latter case, the RTF tests
have shown that the transfer efficiencies needed for realistic air stoichi-
ometries can be obtained with air-sparged towers of 5.5 meters depth or
greater, or by the use of air ejectors. Once adequate 0_ transfer efficiency
has been designed into the system, pH's lower than 6 are simply unnecessary
for good oxidation.
The advantages of conducting forced oxidation at pH 6 instead of pH 4.5
are very clear. First, it will permit the conversion to be carried out in
the simplest possible scrubbing configuration—a single stage system, without
the addition of extra tanks, absorbers, or chemicals. Oxidation can thus be
conducted in scrubbers of current design that are already in operation.
Secondly, it means great simplification of the control requirements, since
60
-------
the feed rates do not have to be controlled in two independent reactors one
at low pH for oxidation and the other at high pH for good SCL removal.
Thirdly, and probably the most important, is the neutral pH of the sludge
produced when oxidation is conducted in a single-stage scrubber. Sludge
produced at pH 4.5 probably cannot be used as landfill without additional
treatment to raise the pH. Finally, RTF results seem to indicate that somewhat
better physical properties are obtained when the sludge is oxidized in a
single-stage scrubber. This is not unreasonable, considering the difference
in particle retention times characteristic of single stage and two stage
systems: gypsum crystals are retained about 16 hours in the EHT of a single-
stage scrubber operating at 3000 ppm inlet SO,.,, but only 5 hours in the first
stage of a two-stage system operating with the same 10-min EHT residence time
in the first stage. Thus, the gypsum crystals are in contact with the super-
saturated liquor for a longer period of time in a single-stage system and
should grow to a larger average size.
The technique of forced oxidation used in Japan, which is based on low
pH—even to the extent of adding H SO, in large amounts—is somewhat baffling
in view of the RTF results. It is especially so considering the high efficiency
aeration that is obtained with the JECCO spinning-cup air atomizer. A possible
explanation of the poor performance reported for those systems at pH's greater
than 5 may be that the CaSO~ solid dissolution plays the dominant role in the
performance of the oxidizer when the aeration is conducted outside the scrub-
bing loop. The observed loss in "oxidation rate" may actually have been
caused by low dissolution rate of the CaS03 at high pH. The tests with
slipstream oxidation tend to support such an interpretation in that the
61
-------
apparent rate of oxidation tended to be lower than those observed when oxida-
tion was conducted in the EHT. The potential limitation of the solid dissolu-
tion step is avoided when oxidation is forced within the scrubbing loop,
because the CaSO~ is exposed to low pH when it is recirculated through the
scrubber.
The conclusion that oxygen transfer is the controlling factor when
oxidation is carried out in the scrubbing loop appears to be well established
by the IERL-RTP results. The conclusion was tentatively drawn earlier on
the basis of the observed effect of oxidizer height on the air stoichiometry
required for complete oxidation in the two-stage scrubber. The conclusion is
verified by three additional observations reported here: 1) oxidizer recycle
influences the bubble residence time, and thus the oxygen transfer efficiency,
by prolonging the contact between air and liquid; 2) the absence of any
strong effect of pH on oxidation efficiency indicates that the oxidation
reaction is fast once the oxygen gets into the solution (i.e.; the chemical
reaction rate is not an important resistance); and 3) the very large effect
of interfacial area seen with the ejector tests at high pH confirms the
importance of oxygen diffusion as the controlling factor in the overall
oxidation process.
The conclusion that oxygen transfer is the controlling resistance is
especially important because the underlying mechanism of that process—liquid
film diffusion—is the same as that which governs the process of waste water
aeration. The technology for waste water treatment has been developed over a
long period of time and is well understood in terms of design criteria for
large scale equipment. Thus, those same criteria should be directly applicable
62
-------
to the design of FGD oxidation systems. The work of Jackson in this
is probably the most comprehensive and the most useful for this purpose, and
has been used in this report as a basis for evaluating the RTF results. It
is clear that the equation for oxygen transfer derived by Jackson (Equation 4)
predicts the performance of the RTF oxidizer within the accuracy of the
measured transfer efficiencies. It can therefore be assumed valid for CaSO-
slurries as well as for the clear sodium sulfite solutions with which Jackson
worked. The significance of this conclusion is that it should be possible to
design air-sparged towers for FGD systems with considerable confidence. It
is interesting that one of Jackson's conclusions is that the amount of oxygen
that can be transferred per unit of energy expended can be increased by the
use of deeper aeration tanks. Although energy efficiencies of 0.75 to 3.5 kg
of 0,,/kwh are characteristic of current commercial aerators (h v. 4 m), Jackson
(4)
estimates that an efficiency of 8.7 kg of O^/kwh should be possible at
h - 16 m. For this reason large scale scrubber tests of a combination EHT/oxi-
dizer of this depth should be undertaken.
The results also indicate that the use of an air ejector(s) in a shallow
tank would be an alternative approach, especially applicable to systems that
are already in operation. It would require only the installation of the
ejector(s) in the hold tank and the installation of a second slurry pump to
recirculate slurry through the ejector. Although a second pump is required,
IERL-RTP tests indicate that it can probably replace the air compressor that
would be needed for a tower. (An oxidation tower would not necessarily need
two pumps, since the hydrostatic head in the EHT/oxidizer would be sufficient
63
-------
to feed the slurry to the scrubber.) The main question regarding the use of
ejectors is the effect of scale up: whereas the performance of a sparged
tower is predictable for larger units, the performance of larger ejectors is
not. The shear developed in the throat, and other factors that affect the
size of the bubbles produced are not constant as the throat size is increased
and can result in poorer performance of larger units . Unless small
ejectors are used in multiple units, this approach should not be attempted on
full scale FGD systems until a larger ejector has been tested in the Shawnee
scrubber. Such tests will be made in 1977.
64
-------
REFERENCES
1. Borgwardt, R. H., "IERL-RTP Scrubber Studies Related to Forced Oxidation,"
in Proceedings: Symposium on Flue Gas Desulfurization—New Orleans
1976, Volume I, EPA-600/2-76-136a (NTIS No. PB 255-317/AS) pp. 117-143,
March 1976.
2. Uno, T. et al., "The Pilot Scale R&D and Prototype Plant of MHI Lime-
Gypsum Process," presented at the Second International Lime/Limestone
Wet Scrubbing Symposium, New Orleans, LA, November 1971.
3. Gladkii, A. V. et al., "State Scientific Research Institute of Industrial
Gas Cleaning (Moscow)," report for Protocol Point A-l, Development of
Lime/Limestone Scrubbing for Stack Gas Desulfurization, US/USSR Sulfur
Oxides Technology Sub Group, 1974.
4. Urza, I. J. and Jackson, M. L., "Pressure Aeration in a 55-ft Bubble
Column," Ind. Eng. Chem. Process Des. Dev., L5 pp. 106-113, April 1975.
5. Jackson, M. L. et al., "Oxygen Transfer in a 23-Meter Bubble Column,"
AIChE Symposium Series, "Water-1975" No. 151, Vol. 71, 1976.
6. Sherwood, T. K., and R. L. Pigford, "Absorption and Extraction," pp. 288-
9, McGraw Hill, New York, 1952.
7. Epstein, M. et al., "Results of Mist Eliminator and Alkali Utilization
Testing at the EPA Alkali Scrubbing Test Facility," in Proceedings:
Symposium on Flue Gas Desulfurization—New Orleans 1976, Volume I, EPA-
600/2-76-136a (NTIS No. PB 255-317/AS) pp. 145-204, March 1976.
8. Ottmers, D. et al., "A Theoretical and Experimental Study of the Lime/Lime-
stone Wet Scrubbing Process," EPA-650/2-75-006 (NTIS No. 243-399/AS)
p. 48, December 1974.
9. Princiotta, F. T., "Sulfur Oxide Throwaway Sludge Evaluation Panel
(SOTSEP): Final Report, Volume II," EPA-650/2-75-010-b (NTIS No. PB 242-
619/AS), April 1975.
10. Jackson, M. L. and W. D. Collins, "Scale-Up of a Venturi Aerator,"
Ind. Eng. Chem. Process Des. Dev., 3^ pp. 386-93, October 1964.
65
-------
TECHNICAL REPORT DATA
(Please readlnurucrions on the reverse before completing)
1. REPORT NO.
EPA-600/7-77-061
3. RECIPIENT'S ACCESSION-NO.
4. TITLE AND SUBTITLE
Sludge Oxidation in Limestone FGD Scrubbers
5. REPORT DATE
June 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
Robert H. Borgwardt
Emissions/Effluent Technology Branch
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO.
E HE 624
See Block 12, below.
11. CONTRACT/GRANT NO.
NA (Inhouse)
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 3/76-1/77
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES Author Borgwardt's mail drop is 65, and his phone is 919/549-8411,
Ext 2234.
16. ABSTRACT
The report gives results of an experimental study of techniques suitable for
forcing the oxidation of calcium sulfite (a throwaway product of flue gas desulfurization
scrubbers now operating in the U.S.) to gypsum, over a range of scrubber operating
conditions applicable to the use of high-sulfur coals. Potential advantages of conver-
ting this product to gypsum include: a) reduction of the total waste production, b)
improved settling and dewatering properties of the sludge, and c) avoidance of chem-
ical fixation of the sludge. It is shown that appropriate oxidizer designs can achieve
complete conversion with reasonable air stoichiometries at atmospheric pressure
without catalysts. Gypsum filterable to 80% solids can be obtained in either two- or
single-stage scrubbers. In either case, efficiency is determined primarily by the
physical absorption of oxygen from the air injected into the oxidizer, when the oxida-
tion step is conducted within the scrubbing loop. The study shows that the required
oxygen transfer efficiencies can be obtained either with an air-sparged tower or an
with an air ejector, when used to aerate the slurry in the scrubber effluent hold tank.
Merits o.f the various approaches are compared and discussed in terms of the quality
of gypsum produced, operability of the system, air-feed pressure requirements, and
predictability of performance when scaled-up to larger systems.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
Air Pollution
Calcium Inorganic
Compounds
Gypsum
Oxidation
Flue Gases
Desulfurization
Limestone
Scrubbers
Wastes
Sludge
Oxidizers
Air Pollution Control
Stationary Sources
Calcium Sulfite
Waste Reduction
Chemical Fixation
13B 07A,07D
07B
08G
2 IB 11G
13. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
74
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
66
------- |