EPA-650/2-73-013
igust 1973
Environmental Protection Technology Series
«*:
m
K*ss*8s
:•:•:•»:•:•:•:•:•
mm
%vl$S&
S¥i
111
m
WxW<
am
II
Si
-------
c..
EPA-650/2-73-013
EPA ALKALI SCRUBBING
TEST FACILITY:
SODIUM CARBONATE
AND LIMESTONE TEST RESULTS
by
Dr. Michael Epstein,
Louis Sybert, and Irwin A. Raben
Bechtel Corporation
50 Beale Street
San Francisco, California 94119
Contract No. PH 22-68-67
Program Element No. IA20I3
EPA Project Officer: Frank T. Princiotta
Control Systems Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
OFFICE OF.RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, DC 20460
August 1973
-------
This report has been reviewed by the Environmental Protection Agency and
approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the Agency, nor does
mention of trade names or commercial products constitute endorsement
or recommendation for use.
11
-------
ACKNOW LEDGEMENT
The authors wish to acknowledge the contribution of the Bechtel, TVA
and EPA on-site personnel at the Paducah Test Facility. The authors
are also indebted to Mr. C. Leivo, Dr. C. Wang and Mr. C. Rowland
of Bechtel, who aided in the preparation of this report.
111
-------
CONTENTS
Section Page
1 INTRODUCTION AND SUMMARY 1-1
2 TEST FACILITY 2-1
2. 1 Scrubber Selection 2-1
2.2 System Description 2-6
2. 3 EPA Pilot Plant Support 2-10
3 TEST PROGRAM 3-1
3. 1 Test Program Objectives 3-1
3. 2 Test Periods and Test Program Schedule 3-3
3. 3 Test Designs 3-8
3.4 Analytical Program 3-9
3. 5 Data Acquisition and Processing 3-11
4 AIR/WATER AND SODIUM CARBONATE TEST
RESULTS 4-1
4. 1 Pressure Drop Data from Air/Water and
Sodium Carbonate Tests 4-1
4. 2 Sulfur Dioxide Removal Data from Sodium
Carbonate Tests 4-11
5 SHORT-TERM FACTORIAL LIMESTONE
TEST RESULTS 5-1
5. 1 SO2 Removal Results 5-1
5.2 Analytical Results 5-33
5.3 Particulate Removal Results 5-35
v
-------
Section
6 CLOSED LIQUOR LOOP RELIABILITY VERIFI-
CATION TEST RESULTS 6-1
6. 1 Performance Data 6-2
6.2 Material Balances 6-16
7 OPERABILITY AND RELIABILITY OF THE
TEST FACILITY 7-1
7. 1 Closed Liquor Loop Operation 7-1
7.2 Equipment Operating Experience 7-3
7.3 Materials Evaluation 7-15
7.4 Instrument Operating Experience 7-22
7. 5 System Modifications 7-25
7.6 System Reliability 7-31
8 ANALYSIS OF PRESSURE DROP DATA 8-1
8. 1 Venturi Scrubber 8-1
8. 2 TCA Scrubber 8-4
8.3 Hydro-Filter Scrubber 8-11
9 ANALYSIS OF SODIUM CARBONATE SCRUBBING
DATA 9-1
9. 1 High Concentration Sodium Carbonate Data 9-1
9. 2 Low Concentration Sodium Carbonate Data
for Chemico Venturi 9-12
10 ANALYSIS OF SHORT-TERM FACTORIAL LIME-
STONE DATA 10-1
10. 1 Statistical Models for SO2 Removal 10-1
10.2 Closed-Form Correlations for Predicting
SO2 Removal 10-8
10. 3 Computer Models for Predicting SO2
Removal and Slurry Compositions 10-12
11 REFERENCES 11-1
VI
-------
Appendix Page
A CONVERTING UNITS OF MEASURE A-l
B CORRECTION FACTOR FOR SO2 REMOVAL DUE
TO DILUTION EFFECT OF REHEATER GAS AND
WATER VAPOR B-l
C DU PONT CALIBRATION CURVES AND CORREC-
TION FACTORS C-l
D WATER BALANCES FOR SCRUBBER SYSTEMS
DURING CLOSED-LOOP LIMESTONE TESTING D-l
VII
-------
ILLUSTRATIONS
Figure
2-1 Schematic of Venturi Scrubber and After-Absorber
2-2 Schematic of Three-Stage TCA Scrubber Without
Trap-Out Tray 2-4
2-3 Schematic of Hydro-Filter Scrubber 2-5
2-4 Typical Process Flow Diagram for Venturi System 2-7
2-5 Typical Process Flow Diagram for TCA System 2-8
2-6 Typical Process Flow Diagram for Hydro-Filter
System 2-9
2-7 Scrubber Area 2-11
2-8 Operations Building and Thickener Area 2-12
2-9 Control Room 2-13
3-1 Shawnee Test Schedule 3-4
5-1 Preliminary Results for SO^ Removal in the
Chemico Venturi with a Nine-Inch Pressure Drop 5-18
5-2 Effect of Gas and Liquor Flow Rates on SO?
Removal in the Four Header Spray Tower 5-19
5-3 Effect of Liquid-to-Gas Ratio and Gas Velocity on
SO2 Removal in the Four-Header Spray Tower 5-21
5-4 Effect of Inlet Liquor pH on SO2 Removal in the
Four-Header Spray Tower (Limestone Depletion
Run No. 463-1A) 5-23
5-5 Effect of Height of Spheres and Gas Rate on SO2
Removal in the Six-Grid Three-Stage TCA System 5-25
5-6 Effect of Spheres Versus No Spheres and Gas Rate
on SO2 Removal in the Six-Grid TCA System 5-26
IX
-------
Figure Page
5-7 Effect of Liquor and Gas Rate on SC>2 Removal in the
Four-Grid Three-Stage TCA System 5-27
5-8 Effect of Liquid-To-Gas Ratio and Gas Velocity on
SO2 Removal in the Four-Grid Three-Stage TCA
System 5-28
5-9 Effect of Gas and Liquor Flow Rates on SO£ Removal
in the Hydro-Filter with Five Inches of Marbles 5-31
5-10 Effect of Liquid-To-Gas Ratio and Gas Velocity on
SC>2 Removal in the Hydro-Filter with Five Inches
of Marbles 5-32
6-1 Operating Data for Venturi Run 501 -1A 6-3
6-2 Operating Data for TCA Run 501-2A 6-5
6-3 Operating Data for Hydro-Filter Run 501 -3A 6-8
7-1 Schematic of Venturi Scrubber and After -Scrubber
After Modification 7-28
7-2 Schematic of Three -Stage TCA Scrubber Without
Trap-Out Tray After Modification 7-29
7-3 Schematic of Hydro-Filter Scrubber After
Modification 7-30
7-4 Typical Process Flow Diagram for Venturi System
After Modification 7-32
7-5 Typical Process Flow Diagram for TCA System
After Modification 7-33
7-6 Typical Process Flow Diagram for Hydro-Filter
System After Modification 7-34
7-7 Venturi Inspection 7-38
7-8 Venturi After -Scrubber Inspection 7-39
7-9 TCA Inspection 7-44
7-10 Hydro-Filter Inspection 7-48
8-1 Comparison of Experimental Data and Predicted
Values (Equation 8-1) of Pressure Drop for the
Chemlco Venturi 8-3
8-2 Comparison of Experimental Data and Predicted
Values (Equation 8-4) of Pressure Drop for the
Chemico Venturi 8-5
x
-------
Figure Page
8-3 Predicted Pressure Drop for Chemico Venturi:
One Hundred Percent Plug Opening 8-6
8-4 Predicted Pressure Drop for Chemico Venturi:
Fifty Percent Plug Opening 8-7
8-5 Predicted Pressure Drop for Chemico Venturi:
Zero Percent Plug Opening 8-8
8-6 Comparison of Experimental Data and Predicted
Values of Pressure Drop for the TCA System 8-10
8-7 Predicted Pressure Drop for the Four-Grid (No
Spheres) TCA System 8-12
8-8 Predicted Pressure Drop for the Six-Grid (No
Spheres) TCA System 8-13
8-9 Predicted Pressure Drop for the Four-Grid
Three-Stage TCA System: Five Inches of Spheres
Per Stage 8-14
8-10 Predicted Pressure Drop for the Four-Grid
Three-Stage TCA System: Ten Inches of Spheres
Per Stage 8-15
8-11 Comparison of Experimental Data and Predicted Values
(Equation 8-6) of Pressure Drop for the Hydro-Filter 8-17
8-12 Comparison of Experimental Data and Predicted Values
(Equation 8-7) of Pressure Drop for the Hydro-Filter 8-18
8-13 Predicted Pressure Drop for the Hydro-Filter with
Three Inches of Marbles 8-19
8-14 Predicted Pressure Drop for the Hydro-Filter with
Five Inches of Marbles 8-20
9-1 Comparison of Experimental Data and Predicted
Values of SO2 Removal from Venturi Computer
Model 9-4
9-2 Comparison of Experimental Data and Predicted
Values of SO^ Removal in Chemico Venturi from
Equation 9-1 9-5
9-3 Comparison of Experimental Data and Predicted
Values of SO£ Removal in Chemico Venturi from
Equation 9-2 9-6
XI
-------
Figure Page
9-4 Comparison of Experimental Data and Predicted
Values of SO2 Removal for the Low Concentration
Soda-Ash Data with the Chemico Venturi 9-17
10-1 Bechtel Limestone and Lime Wet-Scrubbing Sim-
ulation Program: Venturi Scrubber System 10-15
10-2 Bechtel Limestone and Lime Wet-Scrubbing Sim-
ulation Program: TCA Scrubber System 10-16
B-l Correction of SO2 Removal for Water Vapor and
Reheater Gas Pick-Up B-3
D-l Free Settling Rates of Shawnee Clarifier Feed
Solids D-5
Xll
-------
TABLES
Table Page
3-1 Topical and Final Report Description 3-2
3-2 Field Methods for Batch Chemical Analysis of
Slurry, Coal and Alkali Samples 3-10
3-3 Example of Phase II Report 3-13
3-4 Example of Phase III Report 3-14
3-5 Data Channels 3-15
3-6 Example of Mini-Computer Printout 3-17
3-7 Example of Liquids and Solids Analytical Data
Report 3-18
3-8 Example of Solids Analytical Data Report 3-19
4-1 Pressure Drop Data from Air/Water Runs:
Venturi System 4-2
4-2 Pressure Drop Data from Soda-Ash Runs with Air
and SCU Gas Mixtures: Venturi System 4-3
4-3 Pressure Drop Data from Soda-Ash Runs with
Flue Gas: Venturi System 4-4
4-4 Pressure Drop Data from Air/Water Runs:
TCA System 4-5
4-5 Pressure Drop Data from Soda-Ash Runs with Air
and SO2 Gas Mixtures: TCA System 4-6
4-6 Pressure Drop Data from Soda-Ash Runs with Flue
Gas: TCA System 4-7
4-7 Pressure Drop Data from Air/Water Runs: Hydro-
Filter System 4-8
4-8 Pressure Drop Data from Soda-Ash Runs with Air
and SC>2 Gas Mixtures: Hydro-Filter System 4-9
Xlll
-------
Table Page
4-9 Pressure Drop Data from Soda-Ash Runs with
Flue Gas:: Hydro-Filter System 4-10
4-10 Sodium Carbonate Runs with Air and SO_ Gas
Mixtures: Venturi Scrubber (202-1A to 251 -1 B) 4-12
4-11 Sodium Carbonate Runs with Air and SO? Gas
Mixtures: Venturi Scrubber (259-1A to 260-1C) 4-13
4-12 Sodium Carbonate Runs with Flue Gas: Venturi
Scrubber 4-14
4-13 Sodium Carbonate Runs with Air and SO-, Gas
Mixtures: TCA Scrubber 4-15
4-14 Sodium Carbonate Runs with Flue Gas: TCA
Scrubber 4-16
4-15 Sodium Carbonate Runs with Air and SC^ Gas
Mixtures: Hydro -Filter Scrubber 4-17
4- 16 Sodium Carbonate Runs with Flue Gas: Hydro -
Filter Scrubber 4-18
4-17 Variations in SCL Removal for Sodium Carbonate
Testing 4-19
5-1 Test Results for SCU Removal in Limestone Wet-
Scrubbing Runs: Venturi System 5-2
5-2 Test Results for SC>2 Removal in Limestone Wet-
Scrubbing Runs: TCA System 5-6
5-3 Test Results for SC>2 Removal in Limestone Wet-
Scrubbing Runs: Hydro-Filter System 5-10
5-4 Material Balance Results for Factorial TCA Tests 5-16
5-5 Spray Tower Limestone Depletion Run with Four
Headers (Run No. 463-1A) 5-22
5-6 Effect of Inlet SO£ Concentration on SCs Removal
in a Six-Grid Three-Stage TCA 5-30
5-7 Average Liquor Compositions at the Shawnee Test
Facility During October, 1972 5-34
5-8 Particulate Removal in Venturi and Spray Tower
Scrubber During Factorial Tests 5-36
xiv
-------
Table Page
5-9 Particulate Removal in TCA Scrubber with No
Spheres During Factorial Tests 5-37
5-10 Particulate Removal in Hydro-Filter Scrubber
During Factorial Tests 5-38
6-1 Average Conditions for Initial Reliability Verifica-
tion Runs 6-11
6-2 Average Liquor Compositions for Initial Reliability
Verification Test Runs 6-12
6-3 Material Balances for Venturi Run No. 501-1A 6-17
6-4 Material Balances for TCA Run No. 501-2A 6-20
6-5 Material Balances for Hydro-Filter Run No. 501-3A 6-22
7-1 Test Facility Demister Specifications 7-4
7-2 Centrifuge Test Results 7-13
7-3 Corrosion Test Results 7-18
9-1 Comparison of Measured and Predicted SO? Removal
from the High-Concentration Sodium Carbonate Data
for the TCA Scrubber 9-9
9-2 Comparison of Measured and Predicted SC" Removal
from the High-Concentration Sodium Carbonate Data
for the Hydro-Filter Scrubber 9-11
9-3 Predicted Values of A. for Venturi Model for Low-
pH Soda-Ash Runs 9-14
10-1 Comparison of Measured and Predicted Slurry
Compositions at Scrubber Inlet for TCA Run 412-2A 10-17
B-l Correction Factors for SO? Removal B-2
D-l Water Balances for Closed-Loop Limestone Tests
at 10, 000 ACFM D-2
D-2 Water Balances for Closed-Loop Limestone Tests
at 20, 000 ACFM D-3
D-3 Water Balances for Closed-Loop Limestone Tests
at 30,000 ACFM D-4
xv
-------
Section 1
INTRODUCTION AND SUMMARY
In June 1968, the Environmental Protection Agency (EPA), through its
Office of Research and Development (OR&D), initiated a program to test
a prototype lime and limestone wet-scrubbing system for removing sul-
fur dioxide and particulates from flue gases. The system is integrated
in the flue gas ductwork of a coal-fired boiler at the Tennessee Valley
Authority (TVA) Shawnee Power Station, Paducah, Kentucky.
Bechtel Corporation of San Francisco is the major contractor and test
director, and TVA is the constructor and facility operator.
Three major goals of the test program are: (1) to characterize as com-
pletely as possible the effect of important process variables on sulfur
dioxide and particulate removal; (2) to develop mathematical models to
allow economic scale-up of attractive operating configurations to full-
size scrubber facilities; and, (3) to perform long-term reliability testing.
The test facility consists of three parallel scrubber systems: (1) a
venturi followed by a spray tower; (2) a Turbulent Contact Absorber
(TCA); and, (3) a Marble-Bed Absorber (Hydro-Filter"). Each
The Hydro-Filter scrubber has been recently renamed the "Marble-Bed
Absorber. " It is referred to, however, as "Hydro-Filter" in this report.
1-1
-------
system is capable of treating approximately 10 Mw equivalent (30,000
* o
acfm @ 300 F) of flue gas containing 2300-3300 pprn sulfur dioxide.
The following sequential test blocks were defined for the program:
• Air/water testing
• Sodium carbonate testing
• Limestone wet-scrubbing testing
• Lime wet-scrubbing testing
The air/water and sodium carbonate tests have been completed. As of
early August 1973, short-term (less than one day) limestone wet-scrubbing
factorial tests were 95 percent complete and longer term (over two weeks)
limestone reliability verification tests were nearly complete. Long-term
(4-10 months) limestone testing and short-term factorial lime testing are
scheduled to begin in September 1973.
This report, which is the first of three topical reports to be issued, de-
scribes the test facility and test program, and the results, through June
1973, of air/water, sodium carbonate, and limestone wet-scrubbing testing.
The short-term factorial limestone tests were conducted at high scrub-
ber inlet liquor pH (6. 0-6.2). Series operation of the venturi and spray
tower produced sulfur dioxide removals of up to 80 percent at a total
liquid-to-gas ratio of 80 gal/mcf and a pressure drop of 10 inches H?O.
Although it is the policy of the EPA to use the Metric System for quan-
titative descriptions, the British System is used in this report. Readers
who are more accustomed to metric units are referred to the conversion
table in Appendix A.
1-2
-------
The TCA scrubber obtained up to 96 percent SO^ removal at a liquid-to-
gas ratio of 64 gal/mcf and a pressure drop of seven inches f^O. Re-
movals of 80 percent were achieved with the Hydro-Filter scrubber at
a liquid-to-gas ratio of 40 gal/mcf and a pressure drop of 11 inches HO.
Li
Three initial long-term reliability verification tests have been run at
reduced scrubber inlet liquor pH (5.7-5.9), and consequently, at re-
duced stoichiometric ratio, in order to increase system reliability and
limestone utilization (moles SCU absorbed/moles CaCO, added). For
the TCA system, limestone utilization was 83 percent with a sulfur di-
oxide removal of 80-85 percent and a pressure drop of seven inches H^O.
The operability and reliability of the scrubber systems for the initial re-
liability verification tests have been good. There has been little evidence
of sulfate or sulfite scale after approximately 500 hours of operation of
all three systems, with effluent residence times greater than 20 minutes
and percent solids recirculated greater than 10 percent (40 percent of
solids is flyash). Presently, more severe operating conditions (e.g. ,
lower effluent residence times) are being tested to determine the re-
gions of reliable operation for the three systems.
1-3
-------
Section 2
TEST FACILITY
The test facility consists of three parallel scrubber systems, each with
its own slurry handling system. Scrubbers are of prototype size, each
capable of treating approximately 30,000 acfm of flue gas from the TVA
Shawnee coal boiler No. 10. Therefore, each scrubber system is han-
dling the equivalent of approximately 10 Mw of power plant generation
capacity. The equipment selected was sized for minimum cost consis-
tent with the ability to extrapolate results to commercial scale. The
30, 000 acfm scrubber system was judged to meet these requirements.
Boiler No. 10 burns a high-sulfur bituminous coal which produces SO
LJ
concentrations of 2300-3300 ppm and inlet particulate loadings of about
2 to 5 grains/scf in the flue gas.
The test facility has been designed to provide maximum flexibility and
reliability as discussed in References 1, 2, and 3.
2. 1 SCRUBBER SELECTION
The major criterion for scrubber selection was the potentiality for re-
moving both sulfur dioxide and particulates at high efficiencies (sulfur
2-1
-------
dioxide removal greater than 80 percent and particulate removal greater
than 99 percent). Other criteria considered in the selection of the scrub-
bers were:
• Ability to handle slurries without plugging or excessive
scaling
• Reasonable cost and maintenance
• Ease of control
• Reasonable pressure drop
Based on the information available in the literature, the following scrub-
bers were selected:
(1) Venturi followed by an after-absorber
(2) Turbulent Contact Absorber (TCA)
(3) Marble-Bed Absorber (Hydro-Filter)
The venturi scrubber (manufactured by Chemical Construction Co. ) con-
tains an adjustable throat that permits control of pressure drop under
a wide range of flow conditions. Although a venturi is ordinarily an
effective particulate removal device, gas absorption is limited (in lime-
stone wet-scrubbing systems) by low slurry residence time. For this
reason the after-absorber (spray tower) was included for additional ab-
sorption capability. The TCA scrubber (manufactured by Universal Oil
Products and described in Reference 4) utilizes a fluidized bed of low
density plastic spheres which are free to move between retaining grids.
The Hydro-Filter scrubber (supplied by Combustion Engineering Co.
and described in Reference 5) utilizes a packing of 3/4-inch glass
spheres (marbles). A "turbulent layer" of liquid and gas above the glass
spheres enhances mass transfer and particulate removal. Figures 2-1,
2-2
-------
GAS OUT
CHEVRON DEMISTER
AFTER-SCRUBBER
INLET SLURRY
THROAT
ADJUSTABLE PLUG
VENTURI SCRUBBER
DEMISTER WASH
INLET SLURRY
EFFLUENT SLURRY
5'
H
APPROX. SCALE
EFFLUENT SLURRY
Figure 2-1. Schematic of Venturi Scrubber and After-Absorber
2-3
-------
GAS OUT
CHEVRON DEMISTER
RETAINING GRIDS
GAS IN
DEMISTER WASH
INLET SLURRY
MOBILE PACKING SPHERES
5'
APPROX. SCALE
EFFLUENT SLURRY
Figure 2-2. Schematic of Three-Stage TCA Scrubber
Without Trap-Out Tray
2-4
-------
GAS OUT
CHEVRON DEMISTERS
INLET SLURRY
INLET SLURRY
GAS IN
DEMISTER WASH
TURBULENT LAYER
GLASS SPHERES
EFFLUENT SLURRY
APPROX. SCALE
EFFLUENT SLURRY
Figure 2-3. Schematic of Hydro-Filter Scrubber
2-5
-------
2-2 and 2-3 ( drawn roughly to scale) show the three scrubber systems
along with the demisters selected for de-entraining liquor in the gas
streams.
2. 2 SYSTEM DESCRIPTION
The Shawnee test facility contains five major areas: (1) the scrubber
area (including tanks and pumps); (2) the operations building area (in-
cluding laboratory area, electrical gear, centrifuge and filter); (3) the
thickener area (including pumps and tanks ); (4) the utility area (includ-
ing air compressors, air dryer, limestone storage silos, mix tanks,
gravimetric feeder, and pumps); and, (5) the pond area.
The test facility has been designed so that a varied number of different
scrubber internals and piping configurations can be used with each scrub-
ber system. For example, the TCA scrubber can be operated as a one,
two or three stage unit and solids separation can be achieved •with any
combination of clarifier, filter, centrifuge and pond.
Some typical configurations for limestone testing with the venturi, TCA
and Hydro-Filter scrubber systems are shown schematically in Figures
2-4, 2-5, and 2-6, respectively. Such process details as flue gas sat-
uration sprays and demister wash sprays are not shown.
For all systems, gas is withdrawn from the boiler ahead of the power
plant particulate removal equipment so that the entrained particulate
matter (flyash) can be introduced into the scrubber. The gas flow rate
2-6
-------
CO
>s
CO
o
<4-l
bO
it)
• r-t
Q
^
o
CO
(O
0)
u
o
nJ
o
i
ro
d
tuO
O®®
2-7
-------
5 l
Discharge
'
1
>-
§E*
*<
UJ
(
l/l O
S£
o '^
E
c o
•s t
I5
CO
>>
V)
<
O
H
a
rt
^
&JO
rt
• i-i
Q
to
w
0)
u
o
in
i
(M
0)
IH
d
be
O(8>®
2-i
-------
« 1/1
t/i o
si
o _i
1.1
°B'K
sa
|5
S. trt
a
0)
a
rt
^
&JO
nS
Q
^
o
CO
CD
-------
to each scrubber is measured by venturi flow tubes and controlled by
dampers on the induced-draft fans. The concentration of sulfur dioxide
in the inlet and outlet gas streams is determined continuously by
DuPont photometric analyzers.
The scrubbing systems are controlled from a central graphic panelboard.
An electronic data acquisition system is used to record the operating
data0 The system is hard wired for data output in engineering units
directly on magnetic tape and on-site display of selected information is
available. Also, important process control variables are continuously
recorded and trend recorders are provided for periodic monitoring of
selected data sources.
Figure 2-7 is a view of the scrubber area looking toward the power sta-
tion. Figure 2-8 is a view of the operations building and thickener area.
Figure 2-9 is a view of the scrubber control room.
2. 3 EPA PILOT PLANT SUPPORT
Two smaller scrubbing systems (300 acfm each), which are capable of
operating over a wide range of operating conditions, have been installed
at the EPA facility in Research Triangle Park, North Carolina, in
support of the Shawnee prototype testing activities. The small pilot
scale scrubber systems are capable of simulating the TCA scrubber
system and have generated large quantities of closed liquor loop data
on certain TCA configurations.
2-10
-------
Figure 2-7. Scrubber Area.
2-11
-------
Figure 2-8. Operations Building and Thickener Area.
2-12
-------
Figure 2-9. Control Room.
2-13
-------
Section 3
TEST PROGRAM
This section contains a description of the test program. Other descrip-
tions have been presented in References 2, 3, and 6. In Table 3-1, a
description of the reports which are presently scheduled for general dis-
tribution is presented.
3. 1 TEST PROGRAM OBJECTIVES
The overall objectives of this program are to evaluate the performance,
reliability and economics of closed liquor loop limestone and lime wet-
scrubbing processes. The following are specific goals of the program:
'Investigate and solve operating and design problems
such as scaling, demister plugging, corrosion and
erosion.
Generate test data to characterize scrubber and sys-
tem performances as a function of the important
process variables.
Develop mathematical models to allow economic
scale-up of attractive operating configurations to
full size scrubber facilities and to estimate capital
and operating costs for the scaled-up system designs.
Determine opsrating conditions for optimum SO-,
and particulate removal, consistent with opsrating
cost considerations.
Perform long-term reliability testing.
Study various sludge disposal methods.
3-1
-------
o
hH
H
CX
i— i
0{
U
w
W
Q
O
P.
W
P
O
i— i
C^
o
H
1> - .2
+j rt 4J
rt ^ rt ^
m O D
H p,
Information to be Included
r-<
4->
«H
H
"^
0
ft
-
to
D
H
D
d
o
to
s
tH
^
d
rt
testing with interpretation of data.
CO
r~
^
„
a;
ovemt
2
Summary of operating problems and resolutions
associated with reliability verification testing,
tn i
0} "^
H »
00
|3 tn
^ CU
^ £
U3 ^
•rH J
"rt
r^ >>
to
0)
H
a
c
• rH
2
JH
y^
selection of parameters for limestone long-term
reliability testing.
Tf
f^
^
u
rt
^
0]
Summary of operational problems and resolution
associated with lime reliability verification test-
to
••
<1 -^
-M
to
H
an
•5
-g
3
rH
w
development, interpretation of data and status of
limestone reliability testing.
t^-
(-H
r-H
^
Summary of total test program with particular
emphasis on lime and limestone reliability test
results, mathematical models, scale-up design
in
0)
J.
o
.S ft
l«
JH ' — '
o rt
Co _c
.H in
'rt
r-l r>>
<; ^
< ^
ft rt
W h
^
| and economic studies.
3-2
-------
3. 2 TEST PERIODS AND TEST PROGRAM SCHEDULE
The following sequential test blocks have been defined for the test
program:
(1) Air/water testing
(2) Sodium carbonate testing
(3) Limestone wet-scrubbing testing
(4) Lime wet-scrubbing testing
The test program schedule is presented in Figure 3-1. As can be seen
in the figure, the air/water and sodium carbonate tests have been com-
pleted. As of early August 1973, limestone wet-scrubbing short-
term factorial tests were approximately 95 percent complete and
longer term limestone reliability verification tests were nearly
complete.
3.2.1 Air/Water Testing
These experiments, which use air to simulate flue gas and water to
simulate alkali slurry, are designed to determine pressure drop model
coefficients'1' and observe fluid hydrodynamics (e.g., Hydro-Filter tur-
bulent layer) for all three scrubbers in clean systems.
Mathematical models describing pressure drop, particulate removal
and sulfur dioxide removal for the three scrubber systems have been
presented in Reference 7.
3-3
-------
— )
s
s<
0 „-
•—i 2
u_
->
0
z
o
CO
<
R-1
0s — .
i-H ^
5
<
S.
Lt_
->
O
z
o
CO
CM <
<^ — >
l-H '
->
^
<
s
1 TEST PROGRAM FUNCTIONS
?S
CM
CM
cQ
c5
CM
a
1—4
CM
CM
i— i
a
i«~
i— i
1—1
a
a
2
CM
i— 1
i— I
i-H
O
i-H
OS
oo
f-
>0
in
•<*
rr>
CM
i— 1
1
T_
i
*
u
c
<
\r
I
a
u.
1C
SYSTEM CHECK-OUT
A IR-WATER& SODIUM CARBONATE TESTING
LIMESTONE WET-SCRUBBING TESTING:
Short-Term Factorial Tests
Reliability Verification Tests
Short-Term Factorial Tests
Reliabilitv Tests
'0
^
1 O
ji
>£
to
CO
LIME WET-SCRUBBING TESTING:
Short-Term Factorial Tests
Reliability Verification Tests
PeliahilitvTpttt
ENGINEERING & COST ESTIMATE STUDIES
O
LO
-l->
to
0)
H
(D
fi
bJO
• r-<
h
3-4
-------
3.2.2 Sodium Carbonate Testing
Two series of sodium carbonate tests have been designed. The first,
or high concentration series, utilizes concentrated (~1 wt % sodium
ion) water solutions of sodium carbonate to absorb SO-> from flue gas
and from a synthetic flue gas composed of air and SC>2. These tests
are designed to determine uncertain model coefficients where gas-side
mass transfer is rate controlling. The second, or low concentration
series, uses dilute (0. 1-0. 5 wt % sodium ion) sodium carbonate solu-
tions to absorb SO? from flue gas and synthetic flue gas. For this
series, gas-side mass transfer is not rate controlling and liquid-side
mass transfer uncertain coefficients can be calculated using relation-
ships for gas-side coefficients developed from the high concentration
tests. These runs also help ascertain the absorption capability of
liquors associated with some variations of the Double Alkali scrubbing
process (see Reference 8) over a range of operating conditions.
3. 2. 3 Limestone Wet-Scrubbing Testing
The primary objectives of these test sequences are:
(1) To characterize, as completely as practicable, the
effect of important independent variables on partic-
ulate removal and SO9 removal.
tii
(2) To identify and resolve operating problems, such as
scaling and demister plugging.
(3) To identify areas or regions for reliable operation
of the three scrubber systems, consistent with rea-
sonable SC>2 removal, and to choose economically
attractive operating configurations from within thes-.^
regions.
3-5
-------
(4) To determine long-term operating reliability with
attractive configurations for one or more of the
scrubber systems and to develop more definitive
process economics data and scale-up factors.
To accomplish the first objective, a large number of short-term (over
four hours) factorial tests have been made for each scrubber system.
To accomplish the second and third objectives, a relatively small num-
ber of longer term (over two weeks) reliability verification tests will
be made on each scrubber system. These longer term tests will also
be useful to:
• Obtain more reliable material balances.
• Quantify any variations in SO? and par.ticulate re-
moval and system slurry compositions with time.
The fourth objective will be accomplished by running reliability tests,
lasting from 4 to 10 months, on attractive operating configurations for
one or more of the scrubber systems. During these tests, the systems
will be carefully monitored for potential long-term reliability problems,
such as erosion and corrosion of system components. The ability to
effectively operate such systems under varying gas rate, particulate
loading and SC>2 inlet concentrations will also be studied during this
period.
During the short-term factorial test period (see Figure 3-1), it became
apparent early that it was not feasible to operate the test facility in a
3-6
-------
totally closed liquor loop without facility modifications. A closed-loop
test is a test wherein the raw water input to the system is equal
to the water normally exiting the system in the humidified flue gas and
the waste sludge transferred to the pond. In an open-loop system, raw
water input is greater than the water outflow in the exit gas and
sludge. Therefore, process liquor must be discharged from the
system to maintain an overall water balance. In a commercial sys-
tem such discharge may not be acceptable due to potential water pollu-
tion problems. Also, during open-loop operation reliability may be un-
intentionally enhanced since the additional raw water added tends to
desaturate liquors returning to the scrubber, thereby tending to reduce
scaling and plugging. Open-loop operation was not considered to be a
serious problem during the short-term factorial testing, since, at a
specified scrubber inlet liquor pH, SO removal is not significantly
L-i
affected by liquor composition.
To date, therefore, the order of limestone testing has been (see Figure
3-1):
(1) Open-loop short-term factorial testing
(2) Closed-loop long-term reliability verification testing
>I<
The modifications were completed during a five-week boiler outage in
February and March, 1973 (see Figure 3-1). The major modifications
included: eliminating pump seal water on the Allen-Sherman-Hoff
pumps by changing from Hydroseals to Centriseals; humidifying the
hot inlet flue gases with slurry instead of with raw water; and washing
demisters with process liquor diluted with raw water instead of with
raw water only. Other major modifications to the systems during the
boiler outage, not necessarily affecting the water balance, are dis-
cussed in Section 7 and Reference 1.
3-7
-------
3.2.4 Lime Wet-Scrubbing Testing
This test series, which involves introduction of hydrated lime (calcium
hydroxide) directly in the scrubber circuit, will resemble the limestone
wet-scrubbing test program. The major difference will be the absence
of any open-loop tests. Again, tests will be divided into three general
categories (see Figure 3-1): short-term factorial tests; longer term
reliability verification tests; and, long-term reliability tests.
3. 3 TEST DESIGNS
The test sequences for the air/water, sodium carbonate and limestone
and lime wet-scrubbing experiments are all full or partial factorial de-
signs based upon the chosen independent variables, their levels, and the
restraints of time (outlined in Figure 3-1). The choice of the indepen-
dent variables and their levels was based upon pilot plant test results,
the restraints of the system, and results from mathematical models
which relate the dependent and independent variables.
The air/water and sodium carbonate experiments have been completed
at the test facility and a summary of the test results and the independent
variables and their levels can be found in Section 4 of this report.
A majority of the short-term factorial limestone tests have also been
completed at the facility and a summary of the test results and the in-
dependent variables and their levels can be found in Section 5 of this
report.
3-i
-------
Since each limestone reliability verification test will last about two
weeks and assuming one-third downtime for each system (for inspec-
tions, cleanings, etc. ), only about six tests can be made for each
scrubber system, given the restraints of time outlined in Figure 3-1.
Obviously, not all variables which are assumed to affect system reli-
ability can be comprehensively studied for each scrubber system within
the six-run limitation."'
The variables being investigated are:
• Percent solids recirculated
• Effluent residence times
• Gas rate
• Scrubber inlet liquor pH
• Demister types (e.g. , plastic versus stainless steel
chevron)
Solids separation tests for the clarifier, filter and centrifuge are also
being made on the three systems during the test period. Results from
three initial limestone reliability verification test runs are presented
in Section 6 of this report.
3.4 ANALYTICAL PROGRAM
Samples of slurry, flue gas, limestone and coal are taken periodically
for chemical analyses, particulate mass loading and limestone
These tests will be supplemented with reliability verification tests with
the EPA pilot TCA scrubbers at Research Triangle Park, N. C.
(see Section 2. 3)
3-9
-------
reactivity tests. Locations of slurry and gas sample points are shown
on Figures 2-4, 2-5, and 2-6. A summary of the analytical methods
for determining important species in slurry, coal and alkali is pre-
sented in Table 3-2.
Table 3-2
FIELD METHODS FOR BATCH CHEMICAL ANALYSIS
OF SLURRY, COAL AND ALKALI SAMPLES
Species Desired
Sodium
Potassium
Calcium
Magnesium
Chloride
Total Sulfur
Total Sulfite and Bisulfite
Total Carbonate and Bi-
carbonate
Nitrite
Nitrate
Field Method
Atomic Absorption
X-Ray Fluorescence
Dead Stop lodometric
Infrared Analyzer
Ultraviolet Technique
Six DuPont photometric analyzers are being utilized for continuous SO
L-i
gas analyzing at the inlets and outlets of all three scrubbers. Values of
pH are monitored on a continuous basis using fifteen Universal Interlox
3-10
-------
pH analyzers, and three Universal Interlox electrolytic analyzers are
used to monitor electrical conductivity. A modified EPA particulate
train (manufactured by Aerotherm/Acurex Corporation) is being used
to measure mass loading at the scrubber inlets and outlets.
3. 5 DATA ACQUISITION AND PROCESSING
Operating and analytical data are recorded automatically onto magnetic
tapes at the test facility. These are sent to the Bechtel Corporation
offices in San Francisco for processing. Additional data is recorded
manually in operating logs and graphs by on-site personnel.
3C 5. 1 Operating Data (Scan Data Acquisition)
Over 150 pieces of "scan data"(flow rate, temperature, pH, etc. )
are recorded automatically at fixed time intervals onto magnetic tape
at the test facility. Each piece of scan data has an associated channel
number. The scan data acquisition system was designed and installed
by Electronic Modules Corporation, and the tape recorder was supplied
by Cipher Data Products Corporation. A backup printed record on
paper tape is available if the recorder malfunctions.
The scan data tapes are mailed to Bechtel Corporation in San Francisco
for processing. Preliminary processing consists of reading the tapes,
translating the coded data into comprehensible numbers and producing
a "Phase I" file of the raw data in "time order" (successive channels
at a given time).
3-11
-------
A "Phase II" file is then generated from the Phase I file by sorting
the data in "channel order" (successive times for given channels).
Certain values, recorded as "percent of scale" are converted to engi-
neering units in accordance with calibrations and instrument range
settings supplied by the operators. Certain flow rates are corrected
for temperature. Values outside the specified instrument range are
nagged as high ("H") or low ("L").
The Phase II file is particularly useful for spotting instrument malfunc-
tion or other erroneous data, for following data trends versus time, for
spotting process variable discontinuities and/or changes, and for estab-
lishing steady-state run conditions.
A portion of a Phase II report is presented in Table 3-3. The columns
represent the Julian day, time, channel number, data value and data
flag. Channel No. 1050 represents the computed percent sulfur removal
for the venturi system, and channel No. 2001 represents the gas inlet
SO concentration in ppm for the TCA system.
LJ
A "Phase III" file is then generated from the Phase II file by "back-
sorting" the data to the original time order and a report is prepared
based upon the resulting Phase III file. The report carries three col-
umns with concurrent data for all operating scrubbers and a fourth
column for data common to all scrubbers. The Phase III report is
maintained as the prime data reference and is used to determine the
values of operating parameters for any scrubber at any given time.
An example of a Phase III report is given in Table 3-4. All three sys-
tems are shown in operation at 1557 hours on October 20 (294), 1972.
The limestone tests in progress at that time were (see Section 5. 1):
3-12
-------
Table 3-3
EXAMPLE OF PHASE II REPORT
SCANIN Tl IO=PFCJJ1 01/29/73 23.21.24
PftGfc 47
Oil 1717 1050
Oil 1717 1050
Oil 1757 1050
Oil 1817 1050
OH 1837 1050
OIL 1857 1050
Oil 1917 1050
Oil 1"?7 1050
Oil 1057 1050
Oil ,7017 1050
Oil 2037 1050
Oil 2057 1050
Oil 2117 1050
OH 2137 1050
OH 2157 1050
Oil 2217 1050
OH 2237 1050
Oil 2257 1050
Oil 2317 1050
Oil ?337 1050
OH 2357 1050
012 0017 1050
01? OC37 1050
012 0057 1050
012 0117 1050
012 0137 1050
012 C157 1050
012 0217 1050
012 0237 1050
012 0257 1050
012 031? 1050
012 0337 1050
012 0357 1050
01? 0417 1050
012 0437 1050
012 0457 1050
012 0517 1050
012 0537 1050
012 055^ 1050
012 0617 1050
012 Of.37 1050
012 0657 1050
012 0717 1050
01? 0737 1050
012 0757 1050
012 0817 1050
Oil 0817 2001
Oil 0837 2001
Oil 0857 2301
Oil 0
-------
Table 3-4
EXAMPLE OF PHASE III REPORT
1972 294 1512
1001
1002
1003
1004
1005
1006
13C7
1008
1009
1010
1012
1013
1014
1015
1016
1017
1018
1020
1021
1022
1023
1024
1025
1026
1027
1028
1029
1031
1032
1033
1034
1035
10^6
1037
1038
1039
1040
1042
10^3
1044
1045
1046
1047
1049
1050
2541.00
1.00
-10.70
0.40
29980.00
330. 10
316.40
14.80
12.10
0.13
0.38
1.08
14.10
50.00
1132.00
-54.88
251.10
1299.00
?6.20
-26.40
135.00
315.60
1.02
6.41
5^8.00
116.10
31. 2J
0.0
7.87
1.10
7.74
32.70
1.04
605.00
6.26
6 . 98
1 1 2 . 40
61.60
116.50
119.70
8.30
80.00
201 .00
55.30
45.43
-
-
-
-
-
-
-
-
-
-
-
-
-
L
-
L
-
-
-
-
-
H
-
I -
-
-
-
I -
I -
L
-
-
-
-
I -
I -
-
-
-
-
-
-
-
-
C
2001
2003
2004
2005
2006
2007
?008
2010
2012
2015
2016
2017
2018
2020
2022
2023
2024
2025
2026
2027
2023
2031
2032
2033
2034
2035
2036
2039
2040
2041
?04?.
2043
2044
2043
?049
2050
2631.00
-9.70
0.40
20450.00
287.60
124.20
10.30
•- 0.86
0.48
116.40
653.00
-22.71
249.20
150.00
-20.80
112.00
271.50
1.05
5.83 I
1249.00
59.90
0.0 I
7.12 I
1.1?
6.86
14.80
1.00
7.07 I
112.00
56.00
77.60
112.00
114.40
33.50
10.10
94.02
-
-
-
-
-
-
-
-
-
-
-
L
-
-
-
-
-
-
-
-
-
-
-
-
-
-
L
-
-
-
-
-
-
-
-
C
3001
3003
3004
3005
3006
3007
3008
3009
3010
3014
3015
3016
3017
3018
3020
3021
3022
3023
3024
3025
3026
3027
3028
3029
3031
3032
3033
3034
3035
3037
3038
3039
3040
3041
3042
3043
3044
3046
3047
3050
2741.00
-9.60
0. 10
19850.00
314.40
50.00 L
9.30
12.90
0.60
12.00
50.00 L
574.00
-87.73 L
247.20
1378.00
19.70
-19.70
94.00
282.80
1.03
6.08 I -
259.00
103.50
21.40
0.0 I -
6.56 I -
1.10 L
6.85
25.40
205.00
6.33 I -
5.39 I -
108.40
101.50
116.40
117.40
113.10
1.50
198.00
46.32 C
4043 666.00
4044 0.0
4046 14.94
4047 38.90
4048 55.40
4050 94.85
4051 -7803.10
I -
3-14
-------
• Venturi Run 411-1A
• TCA Run 404-2A
• Hydro-Filter Run 411-3A
In general, the 1000, 2000, and 3000 channel series represent the ven-
turi, TCA, and Hydro-Filter data, respectively, and the 4000 series
represents the "common" data. A few selected channel numbers are
described in Table 3-5.
A continuous printed record is maintained for the Phase II and Phase III
data files. These files may also be entered into a computer data base
for report generation and plotting.
Table 3-5
DATA CHANNELS
Channel Descriptions
Gas Flow Rate, acfm
Liquor Flor Rate, gpm
Pressure Drop, in. ^O
Inlet SC>2 Concentration, ppm
Outlet SC>2 Concentration, ppm
SO2 Removal, a %
Channel Numbers
Venturi
1005
1037
1009
1001
1020
1050
TCA
2005
2027
2008
2001
2020
2050
Hydro -Filter
3005
3037, 3047
3008
3001
3020
3050
The calculated SC>2 removals have been corrected for the addition of
reheater gas and water vapor (7% by volume) to each system (see
Appendix B).
3-15
-------
3. 5. 2 Analytical Data
The analytical data acquisition system, which records the results of
laboratory analyses on magnetic tape, was designed and (in part) in-
stalled by Radian Corporation. A mini-computer receives inputs,
either directly from laboratory instrumentation or indirectly by read-
ing cards. The mini-computer performs certain calculations and enters
the resultant data on magnetic tape.
The system generates, on-site, a printed summary sheet of analytical
data for each sample. A typical summary sheet showing liquid and
solids analyses for the TCA scrubber inlet slurry (sample point 2816)
is presented in Table 3-6.
In San Francisco, data on tapes received from the test facility is en-
tered into a data base. The data is sorted, further calculations are
made (e. g. , percent ionic imbalance, percent sulfite oxidation, stoichi-
ometric ratio), and reports are prepared which present the data covering
a specified period for a given scrubber. Portions of typical analytical
data reports are presented in Tables 3-7 and 3-8.
3.5.3 Data Packets and Operating Status Sheets
Data packets are assembled daily for each operating system, and a copy
of each packet is sent to Bechtel in San Francisco. Each packet contains:
• The Test Run Instruction Sheet (which indicates the
system flow configuration, test conditions and analy-
tical requirements)
• The daily operating instructions issued'to the test
engineers
3-16
-------
Table 3-6
EXAMPLE OF MINI-COMPUTER PRINTOUT
RESULTS OF SAMPLE ANALYSES
: SAMPLE ID S38H I RUN NUMBER 5022-1
X SAMPLE POINT 2816 I DATE 5-10-73 TIME 1600
:
: TtMPERATURECC)
: CONDUCTIVITY
: PM
FIELU
e.0 :
,0000E 0 :
.0H00E 0 :
LABORATORY
40.0
.9300E 4
.6000E 1
SPECIES SOUGHT
LIQUIDS
CATIONS
CALCIUMCCA)
MAGNESIUM(MG)
SODIUMCNA2U)
ANIONS
SULFITESCS02)
SULFATESCS03)
TOTAL 5ULKURCA5 S03)
CARBON OIOXIDECC02)
CHLORIDES(CL)
SOLIDS
CATIONS
CALCIUM(CAO)
MAGNESIUMCMGO)
ANIONS
SULFITfc5(S02)
SULFATE5CS03)
TOTAL SULFUR(AS S03)
CARBON DIOXIDECC02)
WT x SOLIDS IN SLURRY
METHOD
X-RAY
ATOMIC ABSORPTION
ATOMIC ABSORPTION
AMPEROMETRIC DEAD-STOP
X-RAY
X-KAY
NON-DISPERSIVE IR
X-RAY
X-RAY
ATOMIC ABSORPTION
AMPEROMETRIC DEAD-STOP
X-RAY
X-RAY
SOLID CARBON DIOXIDE
CONCENTRATION
(MOLES/LITER)
0.523319E -1
0.216971E -1
0.59B0B0E -2
0.194959E -2
0.172827E -1
0.192323E -1
0.690654E -»
0.9733301 -1
X SOLIDS
WT
0.3S427BE
0.140106E
0.115313E
0.650655E
0.209177E
0.182640E
0.160390E
2
1
2
1
2
2
2
TCA INLET
S02 IN STACK GAS • 2781. PPM CONTROL ROOM 802 « 2813,
X IONIC IMBALANCE * -0.146707E 1
X OXIDATION • a.31103tft: 2
STOICHIOMtTRIC RATIO • 0.241790E 1
PPM TIME
1800
3-17
-------
o
d<
w
O
i— i
H
Q
i— i
,-i
o
en
Q
l-H
P
a
i— i
^
N
o
w
w
J» ft X J fl! K ^ tf! N J oV 3 2 2 N cJ JJ^*iX^^^^°*^
— O •*• tnmm m <£ h- ** m f* tf*. ~« *tr\if*i~ir-*rt*- 1*4 (^(M*-f
»
I O
j< • *••« "it^^ 'Q"^'^ *^*ffif^"t^pi°0ir^ ™°iaj"aj^wi^r^(^lli*n3'^^*f^°'^1^
(M
Ul
O I
< VI I *SI
a. a: t o
^JO - «*«* ^-00.* -»-- n-N-1--«U,,.0.<--inoifiiriooif*^irooointno^u^ooooo»rimOooomc —•roirooooooo-ootnoo
"• poopopooopopopoooooooopoopopooooooooopooppopcooooo
a. a. t
IE
I ( I I I I 4 I I I t t I I I I I I ( I I I I I I I I I I I I I I I I I I I I I I I I I I I 1 I
-------
Table 3-8
EXAMPLE OF SOLIDS ANALYTICAL DATA REPORT
SOUnS DATA
VENTURI SCRUBBER
PAGE 2
DATE
JAN 11
JAN 12
JAN 13
JAN 14
JAN 24
JAN 25
SAMPLE
POINT
1S18-A
1821-A
1822-A
1815-A
1P16-4
1818-A
If ?1-A
IR22-A
IB15-A
ini6-A
1118-A
1821-A
18?2-A
1R15-A
1316-A
IPlfl-A
1821-A
1P22-A
1815-A
-R
-c
-D
-F
-F
1816-A
-B
-C
-0
-E
-F
1815-A
-B
-C
-0
-E
-f
-G
-H
-I
-J
1816-A
-B
-C
-D
-F
-F
-G
-H
-I
-J
SLURRY
WGT *
SOLIDS
12.81
0.0
3.13
2.08
1.76
1.99
0.0
1.62
2.01
2.33
4.61
0.0
2.23
3.49
3.57
4.22
0.0
3.53
8. 48
8.74
8.01
1.82
5.90
5.98
8.89
8.95
9.03
8.97
5.90
5.99
6.29
6.26
7.67
7.48
9.71
6.21
7.20
6.38
6.09
6.31
6.60
6.61
7.72
8.17
5.90
5.80
7.02
7.08
6.28
6.16
WEIGHT I OF SOLIDS
SO 2
9.10
0.0
8.79
7.43
8.24
5.08
0.0
7.38
5.02
4.52
6.41
0.0
5.62
13.10
10.59
13.04
0.0
12.40
12.95
12.98
12.94
12.54
16.06
15.95
12.06
12.07
12.33
12.74
13.33
13.30
12.13
1?.35
13.07
13.26
10.89
11.15
12.32
12.13
13.66
13.40
12.43
12.76
13.90
13.71
13.07
13.00
11.45
11.60
13.86
13.57
C02
16.44
0.0
16.96
19.58
14.51
16.20
0.0
17.94
19.25
22.22
15.40
0.0
17.85
14.74
17.01
14.32
0.0
16.27
12. P3
13.41
13.62
14.60
10.12
10.12
13.78
13.32
13.30
11.37
10.12
9.24
7.29
6.34
8.36
8.14
5.06
4.18
8.14
7.26
3.74
2.86
8.55
7.67
7.70
7.04
7.48
5.28
4.18
4.40
4.62
4.18
TOT S AS
22.29
0.0
11.31
22.17
23.03
13.49
0.0
19.12
20.33
25.31
15.35
0.0
16.83
26.17
27.59
18.29
0.0
26.27
20.73
21.32
20.12
21.04
32.65
30.19
18.02
18.53
19.26
19.51
19.49
28.43
28.49
28.35
29.04
27.01
26.14
26.11
36.17
36.53
38.00
36.95
23.08
23.97
22.09
22.25
23.99
25.82
27.97
30.79
30.48
34.66
S03 CAO
13.06
0.0
12.53
16.83
15.52
15.59
0.0
15.87
24.48
26.62
22.81
0.0
22.45
29.96
29.83
28.75
0.0
29.04
25.83
27.55
24.41
26.69
25.75
25.28
25.79
27.06
24.70
24.39
0.0
0.0
21.84
23.44
22.20
23.08
19.67
20.50
19.63
20.92
19.60
20.69
22.21
23.34
21.33
22.36
20.76
21.59
20.48
20.98
20.56
21.16
MGO
0.70
0.0
0.73
0.80
0.73
0.73
0.0
0.68
0.88
0.79
0.81
0.0
0.80
0.67
0.69
0.67
0.0
0.74
0.76
0.81
0.79
0.80
0.57
0.56
0.82
0.78
0.78
0.79
0.66
0.67
0.57
0.58
0.71
0.72
0.60
0.52
0.52
0.51
0.43
0.42
0.69
0.71
0.69
0.73
0.61
0.60
0.55
0.55
0.43
0.43
CALCULATED
WEIGHT %
INSOLUBLES
49.79
100.00
60.16
42.47
48.27
55.21
100.00
48.22
36.32
26.19
47.22
100.00
43.47
31.73
27.53
41.23
100.00
30.78
43.09
40.16
44.30
40.00
34.92
37.84
44.61
43.32
45.04
47.13
73.06
64.98
44.85
44.37
42.95
44.37
51.26
51.48
38.41
37.81
41.65
42.43
48. 53
47.51
51.66
51.04
50.43
49.95
49.68
46.18
47.39
42.96
CALCULATED
* ION
IMBALANCE
-160.42
0.0
-120.36
-125.47
-109.19
-80.34
0.0
-115.34
-50.67
-66.01
-26.77
0.0
-46.49
-20.06
-33.03
-4.53
0.0
-30.02
-14.74
-11.53
-23.12
-19.80
-34.64
-30.53
-12.01
-6.30
-17.94
-10.40
0.0
0.0
-29.10
-15.10
-33.56
-21.57
-20.64
-11.19
-73.66
-60.94
-55.29
-38.71
-16.71
-9.08
-13.33
-5.00
-21.81
-10.55
-17.18
-24.94
-28.68
-35.98
3-19
-------
• On-site plots of selected system variables
• The daily logs kept by the Bechtel and TVA
shift engineers
• The mini-computer printout for each laboratory
sample analysis
In addition, to expedite the transmission of scrubber operating status,
a log sheet containing pertinent operating data and the operating status
for all three scrubber systems is transmitted daily by telecopier to
Bechtel Corporation in San Francisco and EPA in Durham.
3-20
-------
Section 4
AIR/WATER AND SODIUM CARBONATE TEST RESULTS
4. 1 PRESSURE DROP DATA FROM AIR/WATER AND SODIUM
CARBONATE TESTS
The pressure drop data for the air/water and sodium carbonate runs for
the three scrubber systems are shown in Tables 4-1 through 4-9. The
"total" pressure drops refer to pressure drops from the point of gas en-
trance to the gas humidification sections to the point of gas exit past the
demister wash headers.
The observed variations in the independent and dependent variables for
the presented air/water and sodium carbonate (soda-ash) runs are:
Gas flow rate ±200 cfm
Liquor flow rate ±10 gpm
Hydro-Filter turbulent layer height ±20%
Total pressure drop ±3%
Demister pressure drop ±8%
As can be seen from Tables 4-1 through 4-9, the replicate runs were all
in agreement to within the estimated experimental variations of pressure
drop.
4-1
-------
W
EH
CO
CJ
D
EH
£
W
>
CO
W
H
O
PH
O
ID
CO
CO
W
go
ao
a .
oj v
sOCO(MCT-(M
^ ^H rf (\
ommo oui
ooooo
ffi
-------
(M
•*
cu
t—i
r£>
rt
CM
O
to
to
w
EH
to
to
K
to >
IB
< w
to ..
si3
O tf
rf t>
fa H
0, O
o
w
to
to
W
Replicate
Runs
sure Drop, in. H^O
Venturi Demister
n
V .-i
M nJ
H
.2
II
E
Temper -
e, °F
3
L/iquor Flow
Rate to Venturi,
gpm
Air Temper-
ature, °F
o u
r-H
o
o
CO
CO
o
o
o
CO
u
CT-
m
CO
^
PJ CT- ^O O
CO CT- CT- CT-
o r- o co
r- co —t -H
^H r- o o
in CT- o O
OOOO
CT- CO 00 00
m tn -u oo
co o o o
oooo
oooo
^H en en en
i-H CO CO CO
< < PQ <
CO PO CO CO
i i i
cccc
o
CO
CO
It
o
m
00
en
o
o
0
CO
u
0
vO
CO
o
PO
4-3
-------
O
fl
oi
0
M
ffi
co
3
w
|
i *H \Q Q** CO
i ,-H ,-H O s£)
ffi
H
i—i
CO
0
'«
ro
rQ
0)
H
*
Q co
8 2
o
w
>
.
o
H
o m o o o in in
ro co in in in m in •
w
g
CO
CO
W
a a
O
o o o o o o
o o o o o o
ft
ooooooooooooo^o^ 0s ^H»-H
0)
o
^1 Tf
^ ^
QO 00
4-4
-------
$«
at
< CQ
rt
EH
w
H
en
f*
in
U
H
CO
H
O
(NJ
Is
O O | ^ '
in oi'iootrttninui
r- r- ^H i
PL.
O
rt
Q
H
r.
CO
CO
W
cJ
PH
4-5
-------
m
i
(U
i— i
,Q
oi
H
CO
o
CO
Q
&
ffi
h-l
^ ^
S
[3 CO
PCJ PH
hr*
" ^t
co
DO
§• c^
0 C
essure
Bottor
Oi _
3
0
«
U
h
JS
CL
CO
« «.
M d
V ft)
Is
Temper-
e, °F
t. «J
I
|E
|3
v to
r.
H §
5*
ȣ
1 8
h .
*l
a .
3 O
s
Q
O U O Q W W fci [*
(VJ — * ^* ^H ••• O N O N ** WON (MN(M OfMfMON ~< O (*J
NOONC4 J
V 4)
n n
0 O
fe h
JS £
U V
tf tf
3S
4-6
-------
S3
en
ni
EH
H O
O
rt
P
W
en
10
W
P5
"ft a!
Co
§J
o- i
r.
3 O
Oj Z
O O
CTN m o o ••"* **">
i-H ^H (M (M r-H PN!
m so « •* * *t
m ^H i-« %
o o o o o o
o o o o o o
o ^* ^ ^ in in
•^ oo so
ro ro N
(VJ (M (M CM (M fl
4-7
-------
i
*S
IE"
* 5
"S ™ e
1 »'S
? S
34.
|| -S
h
loh
V
H w*
M £
A) ^
g ^
1 1
I S
Water Flow
Bottom Sprays
\t
»l
0 nt
<* .-
.S 4
< B!
||
V
5
< O < 0
-D^oo^uiiri^oooo^r-tninjMin*
5
qj
" 2 « -H « ^
in
t*i
o o m o o o o
oo*n-*oooooooooooo
in o in m o o o tn o in m o in
ooocomoi/ir-oovDooor-inr-
1* N -^ •* — « M Tf pq (VJ ^T n tr- o^
r- &
z~
o o o
OOOO
0 0
m o o o
^H ^.
O* O1* 00 CT*
OOOO
OOOO
«"> fo I-- [^
1 1 1 1
(?•
-------
plic
Run
00
I
o1- -H o <
H
+> J3
i oo d
' Z —-"*** -1 ^2?:^^
* -51 .S oo —m*> oo •
_ in —B^- M I* (M ^ i« - " *
rt 5 •
h
0)
gt.
§°
^rt? t-^Oso h-^-HNO^r-r^oo^r- •^•^•^fOfninm^1 in^c^-^fninmin
in in om ininxn in ino Oin
OOO O^OOO^O~or>jpj
-------
QN
1
, n
"
^
r
PR
ffi
H
1— 1
£
CO
55
§B
E £
CO PH
<; co
< *
2 w
S H
S d
VJ »-H
>«H r.T.1
s ,
So
ps as
h p
^ ><
^j |>^
r "T"
EH HH
<;
a
rt
PH
0
rt
Q
»J
P
CO
CO
W
cj
ft
V
n) m
U C
TJ 3
art
u
en
t-i
V
0]
o 1
ff «
.s is
a H «
§• , j
_H C
Q rt c
if v
V TJ JL
£ « -g
3 (Q .5
«
4)
£ 2
' • 0
H
„ ^
« 'v
•3 K c
* t .3
s s.
H S-
»4
n £
».? e
*£'*
V +j
s-a, .
S'S.s
2 *
R
4)
I*
H.£
o 3
3 £
S a
J
1 $
ft &
i &
A H
« ^^
i «
5 J-
P-I ™
<" a
>. *
o e
^ C
o o
J o
w
I-
V fa.
r.
OJ 4,
H I
nl m
O
»l
,2 rt
^ .
- m
o o
-------
An analysis of the pressure drop data for the three scrubber systems
is presented in Section 8 of this report. The analysis includes fitted
equations and plots representing pressure drop for the three scrubbers
as a function of the independent variables.
4. 2 SULFUR DIOXIDE REMOVAL DATA FROM SODIUM
CARBONATE TESTS
A summary of the SO2 removal data from the high concentration and
low concentration sodium carbonate runs is shown in Tables 4-10
through 4-16. The SO2 removals have all been corrected for the dilution
effect of water vapor and reheater gas pickup by the flue gas. The cor-
rection factors are presented in Appendix B.
In most instances, the "nominal" specifications listed in the tables are
within a few percent of the actual variable levels. An exception to this
would be the inlet liquor pH, Na concentration and stoichiometric
(moles Na^COo /moles SO£ in inlet gas) levels for the low concentration
(double-alkali simulation) runs. In a majority of the tests, the Na+ con-
centration was difficult to control within the prescribed limits. Where
stoichiometry was specified (instead of pH) the soda-ash addition rate
was hard to control because of changes in soda-ash composition within
the feed system. Also, the soda-ash addition rate was based upon an
"average" SO2 inlet concentration and, consequently, the true stoichio-
metry varied throughout the flue gas runs as the SO2 inlet concentration
varied. As an example, the stoichiometry of venturi Run 286-1A (see
Table 4-12), which was nominally 0. 75, varied from 0. 85 to 0. 95 dur-
ing the run.
4-11
-------
a>
I—I
J3
flj
H
w
Pi
H
2~
^
w O
«! 7
-------
rt
H
H
rt
D
FH
O ^
w O
Q H
^
W
O g
ffl H
tf >
<
U
Q
O
"fro!
LSI
ill
O c
o S,
ft 0
Ml?
E
h a
§ °
OS 2
r-oor-insooor-in^o
O O O Is-
r^ r-' r^ r-'
O O O O O
PJ ro PJ ro CM
a a a a a
.5 S S S 5
l\ li II II II
§•§•§•§• g-
to to to to to
*O TJ TJ T3 tJ
to to to to to
m a o) CD m
P, 0, PH PH
H . " **H i,
.s1-^ -° - •§
h i ° °
» h fl c o
a) , C R -M
4J rsi o O _,
C O to to P
£ W > O
? C O O O
y _O g g v£)
rt '"£ * v -^
t « ° ° 5
^ 3 C C *^
,O to O O C
*J ft*
Hi •
«
« a
> >
4-13
-------
4
,0
a
H
W
CQ
PQ
U
W
>
ui
<
O
W
ffi
H
W
U
P
O
a,81
2 c °
S • 6
> y v
< SB:
P.
II
5 K
"
O O O O >M <-4
ooooooooo
ooooooooo
ooooooooo oooo
ooooooooo oooo
ooooooooo oooo
o o
o o
o o
<<<<•-<<<<<
ID O
(M m
11
rt ^
4-14
-------
CO
w
PQ
PQ
U
U
CO
W
I
EH
X
O
O
CO
ffi
H
i— i
CO
£
t>
ffj
W
H
O
ffl
BJ
<
u
Q
O
co
t
« *
^c o
m
N
in irt irt in •-* o
o o ooo ^i
m
ininminom m o Is- om
o o ooo o
CTv CT^ O^ O^ (M N
o oooooo o ooo ooooo ooo
O '— ^- OOOOOO O *• OOO OOOOO OOO
ooooo oooooo ooo o ooooo ooo
NOcaooN (Minomoro oino o inoomo rgmo
"^ffjuffij uuuuuu <<;< <<< <;<<<< <<<
rjpjpopjfM pocjporopors] (MPJN CJNCJ (MMMPOM POPJPO
i i i i i i i i i i i iii iii i i i i i iii
OOOOO PJrOPJNPJrO flfim fOrom •^•^•^•^fin xOvDsO
(MNCOPJCJ N(MNfVirJ(M M(NJPJ fM(MPJ MPJPOPOpO (MPJfM
N cn ^
r- CT^ iii
•"•^HCOOOO 1 1 r-I^H(SICOTltminiA
T3
4)
A
U
4)
*4
V
3
n
^,
-o
rt
V
u
1
A
T)
V
4
W *H
% 1
«. «
0 §
C • «
1- §
^ s •<
sasured) gas and
i and middle stag
i stage.
otonically during
le in Table 18.
r||s|
3 t! ^ S a
a £ £ a o
h m m £
^S £ g |
•V £ "£ 2
JS o o 1 I
•S " *" S ^
™ « m v ,H
H h h h ri
«^.Ji ™B
a; D 4) Q y
^ 3^ "o" •9' "4?
4-15
-------
W
w
PQ
U
U3
<
U
o
W
en
2
E>
ffj
W
u
Q
O
CO
r-J
2 C °
S S £
> U K
^i a) 0*
7
m
C
O
4-1
cd
U
u
tS1
I
1
6
o
m
ex
° 0)
V V
•9 v
S U
£ W
•S2™
's'f "°
0)S g
0 >
3
.£f
J C3
^ 0
JS °
S +a
J IS
5 —
u
w 55
ft
1 G
G a
h-
tr it
13 rt
& 1
0 t)
p^ rt
« oT
o|
^ Q
rt
R
rj o --o r- co Is- o
m « _
!S
^, BI ^ ^
o o o
m in
o o
r- r- r- r-
*o -o r-* ^
o o o o o o o
O O O O O O O
o o o o o
o o o o o
O O O O in
o ifi in in tsj
N
"u"
<<
-------
in
i-H
I
nl
H
w
X
co
•<
o
(M
O P
w
PQ
PQ
w
£ '
w
o
cq
OJ
-------
I
«*
0)
4-18
-------
The observed variations in SC>2 removal for the sodium carbonate runs
are shown in Table 4-17. The variations appear to be a function of pH
level''* and the magnitude of SO^ removal.
Table 4-17
VARIATIONS IN SO2 REMOVAL FOR
SODIUM CARBONATE TESTING
Inlet Liquor
PH
8.5-9.5
6. 0 - 7.5
Percent SO a
Removal
85 - 100
50 - 80
90 - 100
50 - 80
Observed
Variation in
SO2 Removal
±1%
±2%
±3%
±7%
See replicate runs in Tables 4-10 through 4-16.
The observed variations are also functions of the scrubber system.
For example, Na concentration was easier to control on the venturi
system than on the TCA and Hydro-Filter systems. Consequently, the
variations in 803 removal were less for the venturi system than for the
other two systems for the low concentration runs.
The high concentration runs came to steady-state within a few minutes,
once the gas or liquor flow rates, plug positions (or pressure drops)
At high concentration/pH the interfacial vapor pressure of SO? is essen-
tially zero and gas-side resistance controls. At low concentration/pH,
the gas and liquid-side resistances are important. Since the liquid-side
resistance is a function of pH (and Na+ concentration), variations in in-
let liquor pH may cause the observed variations of SO2 removal.
4-19
-------
and inlet SOU concentrations were changed. The low concentration runs
had to be run over longer periods of time (greater than six hours), be-
cause of the apparent large variations in SO_ removal (due to variations
in inlet liquor pH, Na concentration and stoichiometry). The average
high concentration run lasted about three hours while the average low
concentration run lasted about 1Z hours.
In almost all instances, for the high concentration experiments, repli-
cate runs were in agreement to within the stated experimental variations
of SO removal.
An analysis of the data from the soda-ash tests for the three scrubber
systems is presented in Section 9 of this report. For the high concen-
tration data, the analysis includes equations and plots representing SO^
removal for the three scrubbers as a function of the independent vari-
ables. The low concentration data has been analyzed rigorously only
for the venturi scrubber.
4-20
-------
Section 5
SHORT-TERM FACTORIAL LIMESTONE TEST RESULTS
5. 1 SO REMOVAL RESULTS
LJ
The results of all short-term factorial limestone test runs, made from
August 21, 1972, to February 2, 1973 (see Figure 3-1), are presented in
Tables 5-1, 5-2, and 5-3 for the venturi, TCA and Hydro-Filter systems,
respectively. Actual (measured) test conditions are shown, including gas
and liquor temperatures and liquor pH's.
An analysis of the data from the limestone short-term factorial tests is
presented in Section 10 of this report. The analysis includes three
types of mathematical models relating SO£ removal to the independent
variables: (1) linear equations produced by a statistical analysis of the
data; (2) closed-form equations •which are compatible with boundary con-
straints; and, (3) complex computer models.
A majority of the short-term factorial tests were made at "high" scrub-
•jf
T
ber stoichiometric ratios (greater than 1.75 moles CaCO3 added/mole
After mid-November, 1972, when a 60 wt % limestone slurry addition
system was installed for the three scrubber systems (a 15 wt % lime-
stone addition system was used previously), some problems developed
with the calibrations of the limestone additive magnetic flowmeters at
the reduced slurry flow rates. Values of stoichiometric ratio after
mid-November, therefore, are uncertain, but are all greater than 1. 75
moles CaCO3 added/mole SO2 absorbed.
-------
n)
H
W
H
CO
H
fc
W
CO
O
§
W
W
ri
u
CO
i
H
W
W
O
H
CO
H
•J
<
O
H
O
CO
O
CO
H
co
W
rf
H
co
W
H
(fl
•si?
rt
*j >
c o
S g
OH (M
O
CO
1
** 2
c c
tiJ U
C
o
u
O 3
* pj
§•
V -H
n S
n C
>
!i
ji
H
1 |
«I
ii
£ 2
•H rt
•6 tf
o ..H
*" t(
o fi
Ji
1 !
• H
rt >.
Liquor F
Venturi
O O Q
[LI O
3|;
ii
V
&
**«,*
0000
o o o o
§000
000
r--or^
i i i i
-H O 0 O
o o in m
7777
nnnn
0
in
N O
00 00 I^ f-
^?^SS
r- m r-- ro
1/1 O
o m o o
o o o o
o o o o
ssss
Ul sO
O^ -H 1 1
^^^^
***so^*m^«**«Nm2««rtNnNrinN^N
, ,, .
OOOOOOOOOOOOOOOOOOOOOOOOOO
oooooooooooooooooooooooooo
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1
oooooooooooooooooooooooooo
oooooooooooooooooooooooooo
MNNNNNMNNNNNNNNNNMN^NNN^
oooooooooooooooooooooooooo
NOom^m-Woommm^mmmm^ooo^^muimn-o
,77, i77,7,7, , ,7, , ,7
»»««>»^M»--»':i«'-»soa--°0-<10sooint-2
S«SS°g°S"§-SSSSS5
(\jr^^N(N](vj—(M(MNMfvj^--.-HM^(*arjiriin-H^ f->-rg m
sisssSSSasssasaSssiisiisi!
oooooooooooooooooooooooooo
oooooooooooooooooooooooooo
«NN«Nn«N«MNMN««M«m«--«««--
<<<<<<<<<
-------
a
o
U
,0
-i rt u
1 ffi^S
h5
&
5
11
On
u
« .s
V -
ID
11
a "
G^O
° «
'3 ."
S i
o E
^ M
1 I
JS ^
(d >-
fn
* ft
o w
S B
§• s
J >
0 tj 0
fc, "o
3 « ©
K
ll
0)
rt
?
m CM PJ (M
-H-H-H-H^-i-t-i-H -H •" ^^^ "" *
OOOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOO
ooooooooooooooooooo
ooooooooooooooooooo
=? •? ?
tf
icross to
2
T)
0)
3
ft
1
•£•
leader.
I
1
J
1
.s
0
C
•2
to
o
ft
V
at
5
"2
01
.C
C
V
nt
C
1
-
" '
E
o
o
Ji
1
CT-
C
00
0>
nt
u
4)
s
Limes
C
rj
V
O
O
O
fc
•3
re all in excea
0)
0
o
4)
a
n
O
V
h
O
Stoich
—
V
nt
C
cd
0)
0)
g
A
C
OB
V
H
a
h
H
1
at
(M
n
d wet S
C
yzers
nt
(d
C
3
Q
V
5
m
1 due to proble
JD
3
41
i
01
01
"id
o
-.
5
0
o
s
ft
* £
0 0)
fr E
c« u
3 S
O °*
0 JH
0 bo
^ *3
•S.
v *
if
"3 o
ro 3
U h
'S o
ft nt
< Q
5-3
-------
w
H
H
2
W
>
en
^
t»
P£{
O
T3
d
.2
"a
O
U
r~ H
in
«\
CQ
PQ
O
i
H
W
£
H
£
0
H
CO
W
J
O
S
w
o
CO
CO
H
J
D
CO
W
H
CO
W
H
3*
a
P-
h
a
tr
;i
j*
*<
1
f4
V
?
s
o
3
[4
^
0)
t,
S
rt
OJ
a
V
H
n
c
•a ^
& »
"* a!
SJ
'^ Tl
O D
K * «
3^3
. -t->
05 O
3
H «
K 3
'** t!
>3
**•* jj
^1
v fl
s r
* a fn
2 g °
il
^ H
2 83
£-HC?
H JS
UJ ^
w o
S 5
5 *i
a> o
'*•* +j
13
>
^ **
^o9
>. J
ll
I »
S3
>
•M m
o
OJ
3
co r- o o r*
^4^4^
^ ^ ^ ^ ^
00 - «, « « -
'-i ' * «i ci J> * ti •oi^ri.mio^ '
^ ^^^^^ ^ J * ^ ^J
o^oo^^-oo-o-a^^-o^o-^
JJ>ociu:,4
ooor^r^ominminrNiONinoiri^-' inoo
mmmmnmmmmm^m^,mmtonM
iii^Iliylisllilli
-asi
^lA^rlcia-o ' 'Mm^^ui*^^^
TTT^i^.irim^jIviij!^
2222222222222---^^
NO
NO
^^^^^
J ^ J ^ J
~*111
i i i
m «^r
2 ^
N — —
1 1 1
2 — -
0 S~
ti ^;-
S MM
2 -S
2 00
M o m
in MM
i i >
M m -i
in o~- in
^trsjt;
eA '•*•*•*
55L,
222^2
. , ,
1 1 1
m-o^S
1 1 1
T
o*
O
in
00
CJ
o
"
-------
TJ
a
o
in
" a.
ifo1*
< H
2 S 3
"•^
rc i
«<^
.-
BS
g^
> w
-2 3
Ss
0) 41
% 73
0 3
>s **
'C *.
sl
V 0)
^5
^
I
rg
•". *. ~ "! ° -; ". i ~. T' N.
r-'
^
. ^. . °. . ^ . 1 .i.|| -I .-II
fw^j^o^DsO^^O vO'J'-i) -O -O-O
•ii
00 CM
\D*O irivDvo-HinuiinsD vein
iri so
1 1 1 1 1 1 1 1 1 1 1 1 1 ^ 1 1 1 1 1
m ^0^^-^.n^
vO
fONiTir'>OOOsO^OO^O"«^OOsl')OOCO'il(OOQO
i i 1 i 1 1 ^ i i 1 i
fOrO'*fOfMMfv](M'~'fVlrOMI«*'TC^'«l(M(M
mvO'*(^rri^'ij'or-r^'*NrJNrgMfa^-^H
^^^o^d^wcA^ao^^w^^rloloi
_rt_rt^F__ -H^^^^^^,
N(M--"(M'-'tM^-CMN"H""|-|NfM^-NMN-H
{"jooocor^r-sON-H — ifioi--'r-Qooroo
QOO(MO-— O^-O«M O^O^COCOQO'^'O'^OCO-^
— 'NMfMO-iONNO'-'-'fMfMrg-Hfg-.-H
>riinr-MOoo^ ou^woo o^or-^int-^Doo
o'-ooo'-t— o^a--->oov^oO'— >o<^ooo
raom<;<<<<;O'-'
^^-Tj.^-^^^.^.^^.^.^^.^.^.^^.-^^
^-m-^^o t~eoo-> 1111 i ii
roi i i i i i (M^ro-^r-r-coajo^o
--tMfO^^OOf-OOO^M^-NrOOJfMMrOfMm
B S
£. £
« nl
J5 "
"rt "2
9 •"
CT _
1) *
S 1
•£ rt
"3 >
o » '
H -a,
ffi 5 .
Id S C
7 3 '
,
i
30
a S
J t
V
ft
g
S
0
Unable
o>
C
.S
g
0
M
0)
.a
n)
O.
c:
*o
0
U 41
41 m
^ a
4) £
nt ^
*• 0
.S c
Magne'
Ventur
"3 2"
5-5
-------
-«'u«i-ii-ii-i'-H{M(M ooor-infOincor-omro-^
-------
fi
.1-1
-1-1
fl
O
I
IT)
Oj
H
w
H
co
CO
N
° O
g to
a,
0,
u
e
o
u
"S o
— 1M
0*
»
Li
O
•5
0)
M
• H .i-l
*o ^
PH
U qj
H
O
bo
M
|h
« 0
|5
i
CM •-*
m in
0 0
m o
f- PO
fM PO
1 1
0 0
in in
CM t-
fM -<
0x0
t 1
O CO
in m
CO CO
*H CO
i ^
. CO
r-
m o
O CO
o m
O ON
CO ^H
o o
in o
CO O
m o
CO PO
i i
in xo
in
i
•^ in
PO CM
xO 5
O 0
m o
~* CM
o o
o m
r- co
-H fM
f- O
O CM
1 t
QN ~*
r- r-
CO 00
-H CO
00 CO
co r*-
o o
-+ CM
O O
0 0
CO O
in m
CM ^
< <
CM CO
1 1
r- co
CO PO
-H PO
CM O
o o
o m
i t
o o
o m
m co
PO PO
.0 r-
CO CO
1 t
PO CO
CO f>
PO I
<* in
CO xD
O O
-H O
CO CO
o o
o m
ro xO
- fM
CO CM
1 1
ON O
-t CM
xO t-
CM CM
If
m
o
in
CO
CM
1
o
o
xO
CM
O
I-
1
-°
xO
O
1
00
o
CO
0
0
PO
o
m
o
o
PO
PO
CO
oo
PO
r-
co
CO CJ
CO CO
0 0
o m
CO 00
CM CM
0 0
m o
m t—
(M CO
O ON
r- xo
i i
xOxO
xO CO
ON O
r- 1
00 *
GO
O O
CO CO
o o
0 0
CM CM
O O
o in
-* o
o o
CO CM
CM PO
fM xO
PO CM
ON
PO
1
CO 00
CO PO
CM r-l
m xo
0 0
in in
PO IT
i i
O 0
in o
PO PO
oo o
i i
m o
m in
i i
o W
t*- cc
o o
o o
CM CO
o o
o o
CO O
o o
CM PO
PO PO
PO PO
o
CO
1
ON ON
PO PO
m cj
CO ON
o o
o m
m •*
PO PJ
i i
0 0
o o
ro co
ro PO
0 CO
xfl CO
1 1
00 00
m CM
co in
i-H CO
i i
XO NO
0 *#
co m
o m
0 0
CO xO
o o
o o
o o
o m
CM —i
PO PO
i i
t^- CO
PO CM
CO
PO i-H
CM ro
•* ON
0 0
ir> o
CO CJ
i i
O 0
m m
CM O
CM CM
-,„
1 1
O PO
r-*
co in
"M CO
r- p-
i i
c- o
xo r-
^t CM
xD CO
O O
O O
PO xO
O O
O O
ON O
^ m
CO CO
CM CM
1 1
ON O
CM PO
^ ^
•- I
r- co .
O O "H
o o o
O O -H
fM xD (Jx
O O O
o o in
o o o
o o in
CJ CO —H
CO CO CM
1 1 1
-• fM ro
PO
1
CO -i
0- CO
0 0
in o
Tf in
PO PO
i i
o o
o m
oo ••$•
•-< PO
oo r-
PO in
i i
r- in
co in
CO CO
CO CO
O -t
1 1
0^
O CO
in m
o o
ON ON
o o
o o
o o
o in
CO CM
CO PO
^ in
CO CO
CO CO
V
rt
were
00
D
Li
0)
•a
00
G
V
1
(0
D
rt
DQ
CO
ber Internal
,0
D
U
in
M
u
H)
o
o
CO
A
bO
3
O
Li
£
00
1
7
S
a
C
o
cu
£
3
C
o1
W
m
ed/molei
T3
T3
rd
CO
o
u
6
m
4)
O
£
Defined as
"«"
ON
C
o
m
Li
1
0
0
xO
(d
1
cd
03
ft)
ftf
CO
Ll
0
rd
Li
O
flj
PO
O
D
£
«
T3
S
Li
e
•fi
CO
Li
U
SOT remove
T)
.1 in excess of 1. 5 moles CaCO3/mole SO? inlet.
rd
te are
1
VI
O
TJ
o
Li
8.
1
Li
O
(4-1
0)
>
a
•o
a
V
4*
n
•0
a)
1
•0
S
0
0)
Li
c-
O
U3
V
bO
.s
Li
•S
£
DO
00
0
13
13
13
rt
"o
03
Limestone
C7
j
.
§
Li
•e
a
£
bo
.9
1
5-7
-------
.3
o
u
(M
in
a
H
W
H
O
H
co
O
W
U
co
i
H
W
H
^
O
H
co
H
O
s
w
o
CO
tf
O
CO
H
J
t>
co
W
EH
CO
ti
o
U
V -t->
"1 V
i 3
j o
J w
M
| »
,JH oocor-^xor-vONOmoor-
)
*, fc
"5 °
v
if
•* ,*
01 1 ~ «A ri N .• 7 2 " 7 " 7 T ™ 7 77 T "? "• • j i A _i i "•
i fi
i ^
I
PH
O C* (O (M
41 ** M -• rg ro
•^ M r^ r-
I O -H nH
fc'
0
v
M
^_ ,-..-_.^.^. I I I I I I i I J
g- ^3 "
4)
H
<4
°. , - m - ^ ^ ru rvi m f^
• r-i o o b^od coobovobooot^-t^cor-bdr-b g-
--- ' ---- ------- - (MoJCV]PJPOCM{MM(MM(NJN(VlPOr»JfM ^
g
NNPOM(V]MPJMrO[v]{VJ OJ
'--'-' ' ' ' ' -•- "i u u o u u u u u u u u u u
I
*3
T)
U
. 1
Ol-HfVJ^vOOOO^^H M>-(
r-(M(NJ(MfMN(MNrO^H ^HM ^ _• Nm
—. i r i i i i i i r m r i 10 —i >-("HEi
Qir-o^foinr-GOO^-Himo^^rofMNro^H i 11 "
O'-'-tfJMr-JpJcocoporo^H (\jr-Ji>j(Nj(M(MMrj i o o o •—"—••—i t\]
o o o
^
00
OC
5-i
-------
W
H
co
>H
CO
<1
O
H
co
O
PQ
T3
(D
r>
.S
-M
s
o
U
^""*
(M
1
m
i— I
ni
H
ffj
O
CO
H
W
^
W
£5
O
H
CO
W
i— i
J
^
I — |
•4
o
w
Nl
o
CO
o
CO
H
CO
w
H
CO
w
H
Clarifier
Underflow
K
CX
§ 5
1 0
j
V
•a
u,
*, O
S .-
1 e
H
h
O -w
JD V
•° 3
o 55
w
1
H
0 -
» i «•
tr •§ t!
3 Is
u
«£<"
CQ go
H
h
4) +J
J2 0)
•51
s3
b, m
O
0)
? "
£ n
2 £ «
a> 3*3
a >•* 5
£ u
H
td
0
II
§ o
«
S
1 1 * 1 1 '.
t*- •£>
•~o in
f\j —> (O ro
* -*** 0
%o -o in -o
njJsJ
oo oo cr- oo oo o
2 « ™ z; - «
O^ O O O O O
^2"r'7^
^00^00^0
CO ft M co (M <*l
(^ C3^ ^H o O PO
O — i M CM -«^— i^-.^— .rorJ
-^^-^.•^^.^.•^.•^^^ -^^.-^.^x.
I- 5
s s
5 £
SI
5-9
-------
CO
OOOOOOOOOOOOOOOOOOO
OOOOOOOOOOOOOOOOOOO
222222S00000
IOOOOOOOOOOO'
o
s
ffl
pq
re
H2O
ONDinoocM'«j*••* O O CM
icoocoooooooocococorMi
u
CO
H
W
W «
o w
r t ^H
HT
ide
nce
min.
CO
I
•8
H
W
A
> O
O ti
5 Q
W ^
(M
O
o
L/G,
al/
oomoomininoinooi
lOinoinomooi
j •« —i en m tM in •
omininooomomini
r-tOOOOOOOOOO'
en
H
ooinoinommomooomooinino
OOOOOOOOOOOOOOOOOO-H
> o o o o
] CM CM CM CM
W
OOOOOOOOOOOOOOOOOOl
•00000
co
W
H
rorocorommfirororoc'ii
oooooooooo^.
~^.OOOOOOOOOOOOOOOO--"^-H
5-10
-------
(U
d
.s
-l->
P!
o
u
ro
I
n)
H
W
O
ffi
CQ
u
CO
"I
o w
H H
> o
O f*J
00
o
CO
ti
O
co
H
CO
W
CO
w
i
g
f.
a
P
V
i
4
i
a
i-
&
i
3
a
4
I
t
c
c
t
a
4
C
(
3
cr
C
•-*
1
3
s
E
1
|
a
C
J
prf
0
3
moval
I (M
0
en
M Q.
0
§
(J
o
X
a
d
o
Q
\n
I
^
« *j
2 „•
n g
0>
« S
Solic
Recircu
a
0
•a
o!
0
£
•8
E
•3
I !r
O Jv
« w
H
w
Ifc,
rt o
rt ©
Crf
0
2
o
o
I
o
0
sC
•*
.
0*
o
in
m
0
o
in
in
o
o
o
o
[V]
o
0
o
o
1
vD
N
_
~4
in ft
o o
o o
o o
0 0
m oo
(M C>J
1 1
-^ 00
1 1
m in
o
in
o ~~~
in in
m o
0 -H
in o
0 0
N IM
0 0
o o
0 0
o o
(M N
< m
t- sD
(M (M
IM CO
-, ^T
*«
r-^or-so^^^^^^r-o
oooooooooooo
oooooooooooo
oooooooooooo
oooooooooooo
t-MOCT'-t-Nooom-Hrg
T T -7 -i ^ lA co o ^ ^ ^ ci
^H O ^
O ff*
ininin-*-*-.M-*-opg-*^
Oinommmoinooinin
oo-^ooooooooo
oooooooooooo
o o o o o
eg r«i pg N rg
OOOOOOOOOOOO
OOOOOOOOO 3OO
ooooooc'-omooo
OOOOOOO^MCMOOO
OCQg
(M 1-1 eg --.
•-o i ' xo pg (\j
-H (M (M 1 I 1 1 1 1 O — I
M rg M--^^d^^^CC
•d
0)
1
s.
4)
0)
n
I
o
o
(M
O
r-
00
0
00
0)
s
u
E
3
ai
.h
at
•3
£
) added
(J
id
U
"o
nt
V
C
ft)
Q
rt
V
Ifl
(it
o
§•
0
h
neasure
s
00
JS
4)
43
O
CO
o
3
0>
's
T?
.3
U
X
01
|
"£
(M
E
6"
u
rt
E
'o
(0
C
nS
0>
0)
13 'S
1 «
• 1
£ I
metrics
e ope rat
3 1
S a
V SH
5-11
-------
H X
c
o
o
*w*
CO
i
m
CO
2
P
OtS
O
5
CQ
CQ
t3
tf
U
U3
i
H
o w
H H
C/D CO
w ><
> o
is
w t^
pj ffi
CO
O
o
h
co
H
J
E>
LO
W
ffi
H
co
U
H
H
Q 8
v a,
3 E f>
g V 0
-------
T3
3
C
0
U
ro
i
tn
rt)
H
w
O
ffl
CQ
O
i
H
H
§2
°S
w ^G
H £
•^ en
> O
O PH
w
ffi
(M
O
en
O
en
H
en
W
en
W
H
V CO
Ifi
So r- r* o o
£ C3
'? « J, fA ' ' J J, I ^r J L* ^! J. I l I
1 O -H
_ e ^^*liiiiiiij.iii
Q S
5; ^ o a o o* *o so -o co
'- <^ j. ^ j « ' „,- ' ' i J. i ' '
0 „
pr
? s.
cQ tn
tn
g-j ^,^^00^,
Ul
.MP.^ ^Co^S^S^MNNN^^^^O
gCO oo^r^ONm^-^vOI^O^lA^Oo
_J|H •**mvmm*(MMfo«-.m ^
S *
Q
I
U
H
ffl tn
1 I ' ' ' o
r- -t -j o o cr- a-
^r
H
h ^
O
V
(U
V
o.
^ .
O
Ort oi^»j>ooomcNO^min'*_. ._
I -^
-------
SC>2 absorbed) and, consequently, at high inlet liquor pH's (6.0-6.2).
The data collected indicate that stoichiometric ratio has an insignificant
effect upon SO removal at values of inlet liquor pH greater than 6. 0
LJ
(see Sections 5. 1. 1 and 10. 1).
As mentioned previously, the SO^ removals, as calculated from the mea-
sured (DuPont analyzers) scrubber inlet and outlet gas SO^ concentrations,
have been "corrected" for the differences in the mass flow of gas between
the two locations (see Appendix B). The SOo removals during factorial
limestone testing have also been corrected for DuPont SC>2 analyzer
calibration errors associated with unstable composition of SC>2 calibra-
tion gas (from September 15 to October 13, 1972) and the deterioration
of the DuPont analyzer optical filters (from June 23 to December 1, 1972).
Prior to system modifications made during the five week boiler outage,
reasonable material balance closures for calcium and sulfur could only
be obtained with the TCA system. During this period of operation, the
venturi and Hydro-Filter systems still had the clarifiers and process
•water hold tanks included in the main slurry loops, while the TCA sys-
tem had been modified so that the main slurry stream circulated be-
tween the hold tank and scrubber, with a bleed stream from the main
slurry stream routed to the solids separation area (see Figures 2-4, 2-5
and 2-6). The poor material balances for the venturi and Hydro-Filter
These correction factors were furnished by the DuPont Company and
have been included in Appendix C. The corrections wer e small for
SO2 removals greater than 70 percent.
During the five-week boiler outage, the flow configurations for the ven-
turi and Hydro-Filter systems were converted to ones similar to that
of the TCA system. These modified configurations are shown in Sec-
tion 7.
5-14
-------
systems were attributable to solids build-ups (or depletions) within the
clarifiers. For the TCA scrubber, the calcium and sulfur leaving the
system could be obtained from the measured flow rate and solids analy-
sis of the "bleed stream" to the solids separation area, and the clari-
fier could be excluded from the material balance enclosure.
Material balances for six TCA open-loop limestone runs are shown in
Table 5-4. The closures are within the limits of the estimated experi-
mental accuracies. The SO? absorbed was computed from the measured
inlet gas rate, the inlet and outlet gas SO^ concentrations and the estimated
gas outlet rate. The calcium added was computed from the measured volu-
metric rate of limestone slurry additive and the solids concentration in
the slurry. The sulfur and calcium discharged were computed from the
measured rate of slurry discharged from the system and the concentra-
tions of sulfur and calcium in the discharge.
Although satisfactory material balance closures were not obtained dur-
ing the open-loop factorial testing for the venturi and Hydro-Filter sys-
tems, confidence in the generated data for commercially reasonable SO
removals (greater than 70 percent) is based on the following:
• "Wet" chemical analyses for SO2 in the inlet and exit
gas streams repeatedly corroborated DuPont SO?
analyzer measurements.
• Sulfur removals in longer term reliability verification
runs, with excellent material balance closures for cal-
cium and sulfur (see Section 6), have been in close
>'<
agreement with factorial replicate''' runs for the ven-
turi, Hydro-Filter and TCA systems.
Replicate runs are made with identical values for all independent
variables.
5-15
-------
un
r-l
-H OO 0s i-H rH LO
i-H i— 1 i— 1 O i-H i-H i-H
41
O O O O O O
co t>- o co r^ vo fe^-
•^ CO ^f O ' — 'CO O
1 — t 1 — 1 1 — 1 1 — 1 1 — 1 1 — 1 1 — 1
• ••••• "I
0 O O O 0 O
0
0s oo co un co m fe~-
o LD oo xo m r~- ur>
I-H O O O O O I-H
O O 0 0 O 0
LO sO ^^ LD sO CO ^~-
00 LO 00 v£) vO 00 O
O 0 O 0 0 O I-H
• ••••• ~H
o o o o o o
jj U
cocococococo S b
i i i i c i .3 S
^Hco^unooo^ -u>o
oooooo "j-J
5-16
-------
In the following sub-sections, the significant SC>2 removal results from
the limestone factorial testing are presented graphically and discussed.
It is recognized that SC>2 removal is affected by SC>2 inlet gas concentra-
tion and scrubber inlet liquor temperature (Reference 9). Care, there-
fore, has been exercised in segregating these non-controlled independent
variables in the presentation of the data.
5. 1. 1 Venturi System
In Figure 5-1, the effect of gas and liquor flow rates on SC>2 removal for
the venturi scrubber with nine inches of pressure drops is shown. All of
the runs are indicated in the figure. The solid data points represent runs
made after December 1, 1972, which do not need the DuPont analyzer
corrections. Obviously, the replicate runs made after December 1, have
consistently lower SO-, removal values. These discrepancies cannot be
explained by differences in either inlet gas SCu concentrations or in
scrubber outlet liquor temperatures. This casts some doubt as to the
accuracy of the DuPont corrections (see Appendix C) at low SO removals.
Figure 5-2 illustrates the effect of gas and liquor flow rates on SO2 re-
moval in the four-header spray tower. The outlet liquor temperatures
(shown in Figure 5-2) varied considerably from run to run. The effect
of outlet liquor temperature on SO2 removal was generally consistent,
with higher SO? removal at lower temperature. A curve representing a
median liquor temperature of approximately 100°F has been drawn for
each liquor flow rate.
5-17
-------
60
50 -•
40 •-
O
CO
I—
z
c
oe.
UJ
20 ••
10 ••
0
DARKENED-IN SYMBOLS REPRESENT DATA WHICH
DO NOT NEED DUPONT SO2 ANALYZER
CORRECTIONS (DATA TAKEN AFTER DEC. 1, 1972).
LIQUOR FLOW RATE = 600 gpm
LIQUOR FLOW RATE = 300 gpm
STOICHIOMETRIC RATIO > 2 moles CaCOg/mole SO2 absorbed
SO INLET CONCENTRATION = 2200-3000 ppm
PERCENT SOLIDS = 5-8%
HOLD TANK RESIDENCE TIME = 33-70 min.
SCRUBBER OUTLET LIQUOR TEMP. = 107-120°F
1
15,000
20,000
25,000
GAS FLOW RATE,acfm @ 330 F
30,000
Figure 5-1. Preliminary Results for SC>2 Removal in the
Chemico Venturi with a Nine Inch Pressure Drop
5-18
-------
70 --
60 --
<50
O
10
I—
§40
30 --
20
I (78-92)
(78-87)
320 gpm
2-3% SOLIDS
(98-111)
(102-114)
(109-114)
(79-116)
LIQUOR RATE - 450 gpm
LIQUOR RATE = 300 gpm
STOICHIOMETRIC RATIO > 2 moles CaCOg/mole
SO- absorbed
PERCENT SOLIDS = 5-9%
RESIDENCE TIME = 40-106 min.
SO2 INLET CONCENTRATION = 2,500-3,400 ppm
LIMESTONE
DEPLETION
RUN 463-1A
HIGHSTOICH.
RATIO
NUMBERS IN PARENTHESES REPRESENT
LIQUOR OUTLET TEMPERATURES ( °F).
(111-118)
1
30,000
10,000
20,000
GAS RATE,acfm@ 330 °F
Figure 5-2. Effect of Gas and Liquor Flow Rates on SO2 Removal
in the Four-Header Spray Tower
5-19
-------
Figure 5-3 is a cross-plot of Figure 5-2, showing the effect of liquid-to-
gas ratio and gas velocity on SG>2 removal at a scrubber outlet liquor
temperature of about 100°F. The SC>2 removals are outside the range
of interest for commercially acceptable gas velocities (greater than 7
ft/sec). ' The results from Figure 5-3 appear to agree reasonably well
with the spray tower data taken by the Hydro-Electric Power Commission
of Ontario (Reference 10), after correcting for the effects of inlet gas
SO concentration (see Section 10. 1. 3).
* *
A spray tower limestone depletion Run 463-1 A was made to determine the
effect of stoichiometric ratio and inlet scrubber liquor pH on SO2 removal.
Results from this run are presented in Table 5-5 and in Figure 5-4. SO2
removals for this run were low because, at that time, the liquid-to-gas
ratio could not be maintained greater than approximately 30 gal/mcf.
In Table 5-5, the stoichiornetric ratio has been calculated from the esti-
mation of the original Ib-moles of CaCO3 in the system and of the
absorbed. A comparison between the stoichiornetric ratios calculated
in this manner with those obtained from the solids analysis could not be
made, unfortunately, because of uncertain solids analytical results dur-
ing this period.
*
A modification to increase the maximum liquor rate from 600 to 1200
gpm for the spray tower is scheduled for completion by the end of May,
1973. Further spray tower limestone factorial data will be obtained
subsequent to that date.
i ^,C
A limestone depletion run is a run in which no limestone make-up is
added during the test period.
The stoichiornetric ratio (moles CaCO3/moles SO2 absorbed) of the
scrubber inlet liquor changes with time during the depletion run as the
SO^ is absorbed (i. e. , one mole of CaSOx is formed and one mole of
CO2 is evolved for every mole of SO2 absorbed).
5-20
-------
80
70 •-
_, 60 ••
O
to
I—
Z
UJ
U
40 •-
30
SO2 INLET CONCENTRATION = 2,500-3,400 ppm
STOICHIOMETRIC RATIO > 2 moles CaCOg/mole SO2 absorbed
PERCENT SOLIDS = 5-9%
HOLD TANK RESIDENCE TIME = 40-106 min.
SCRUBBER OUTLET LIQUOR TEMP, sa 100° F
2.5 ft/sec
sec
20
10
20
30 40 50
LIQUID-TO-GAS RATIO, gal/mcf
60
70
Figure 5-3.
Effect of Liquid-to-Gas Ratio and Gas
Velocity on SO2 Removal in the Four-
Header Spray Tower
5-21
-------
Table 5-5
SPRAY TOWER LIMESTONE DEPLETION RUN WITH FOUR HEADERS
(RUN NO. 463-1A)
DA I'K
1/23/73
1/24/73
IN
ST. »
l.K 1' Sl.UKKY
ms
-.%'
1100 7.70 7.30
1200
1300
1400
1500 6.
1600
1700
1800
1900 6.
2000
2100
2200
2300 6.
2400
0100
0200
0300 7.
0400
0500
0600
0700 8.
0800
0900
1000
1100 9.
1200
1300
1400>
1NI.K 1
SO^ PER(
CONC. . SO RE
ppm
3125 71.
STOICHIOMETRIC RATIO,
:ENT TOTAL MOLES CaCOi
MOVAL MOLE SO2 ABSORBED
0
3094 70.0 38
33 6.50
6.50
6. 25 ' .
6.20
40 6.30
6.20
6.20
6.00
95 6.30
6.30
6.20
6.30
93 6.30
6.30
6.30
6.20
92 6.35
6.30
6.25
6.30
00 6.30
6.30
6.30
3094 60.
3063 58.
3031 53.
0 19
0 13
0 10
3031 49.0 8.6
3031 48.
3000 47.
2969 47.
3000 46.
3125 43.
3250 44.
3187 44.
3063 44.
3063 44.
3063 44.
3000 41.
3063 42.
3063 41.
3125 42.
3125 42.
3156 42.
3156 41.
3031 42.
2875 41.
2813 41.
2813 42.
0 7.4
0 6.5
0 5.8
0 5.3
5 4.8
5 4. 5
0 4. 1
5 3.9
5 3.6
0 3.4
5 3.2
5 3.1
0 2.9
0 2.8
0 2.7
5 2.6
0 2.5
0 2.4
0 2.3
5 2.2
5 2.1
"." 1
• J System down due to high fan vibration.
1/25/73
1/26/73
2100 J
2200
2300 5.
2400
0100
0200
0300 6
0400
0500
0600
0700 7,
0800
0900
1000
1100 6.
1200
1300
1400
1500
1600 5.
1700
1800
1900 7
2000
2100
2200
2300 6
2400
0100
0200
0300 6
0400
0500
0600
0700 6
0800
0900
1000
1100 6
1200
6,15
95 6.20
6.20
6. 10
6.05
60 6. 00
6.10
6.00
6.00
95 6.20
6.20
6.00
6.05
62 6.10
6. 10
6.00
-
5.90
85 5.90
5.95
5.70
50 5.80
5.70
5.65
5.60
22 5.70
5.70
5.60
5.60
78 5. 50
5.40
5.45
5.35
99 5.45
5.40
5.20
5.20
70 5.40
5.30
2812 51.
2812 47.
2750 43.
2812 44.
2562 41.
2812 47.
2750 46.
2812 47.
2750 44.
2750 44.
2813 45.
2813 43.
2813 43.
2500 44.
2438 43.
2375 43.
2375 43.
2375 40.
2312 36.
2000 36.
2312 36.
2188 35.
1812 32.
1750 30.
2000 33.
1938 31.
1875 32.
1688 31.
1781 31.
1906 31.
1906 30.
1813 29.
1906 29.
2000 28.
2344 26.
2344 24
2500 23
5 2, 1
0 2.1
5 2.0
5 2.0
5 1.9
0 1.9
0 1.8
0 1.8
5 1.8
5 1.7
5 1.7
0 1.6
5 1.6
5 1.6
0 1.6
0 1.5
0 1. 5
0 1.5
0 1.5
0 1.4
0 1.4
0 1.4
5 1.4
0 1.4
5 1.4
5 1.4
5 1.3
5 1.3
5 1.3
5 1.3
0 1.3
5 1. 3
0 1.3
0 1.3
0 1. 3
0 1.2
0 1.2
2500 21.0 1.2
2500 21
0 1.2
5-22
-------
60
50 •-
O 40
CM
O
to
h-
z
LJJ
y so
20 ••
10
5.0
GAS RATE = 20,000 acfm @ 330° F
LIQUOR RATE = 450 gpm
SO INLET CONCENTRATION =1,750-3,200 ppm
PERCENT SOLIDS = 6-9%
SCRUBBER OUTLET LIQUOR TEMP. = 97-113° F
HOLD TANK RESIDENCE TIME = 56 min.
-f—i—i—i—i—i—H
5.5 6.0 6.5
SCRUBBER INLET LIQUOR pH
7.0
Figure 5-4. Effect of Inlet Liquor pH on SC>2 Removal in the
Four-Header Spray Tower (Limestone Depletion
Run No. 463-1 A)
5-23
-------
As can be seen in Table 5-5, the SC>2 removal and inlet liquor pH re-
mained at steady values of 40-44% and 6. 0-6. 3, respectively, for a long
period of time (about 35 hours) before the removal and pH began to drop.
The stoichiometric ratios for this period of time were greater than 1.4
moles CaCO3 added/mole SO2 absorbed. The SC>2 removal for this high
stoichiometry region has been included as a data point in Figure 5-2.
Figure 5-4 shows the effect of inlet liquor pH on SC>2 removal for Run
463-1A as the limestone in the system was depleted. Similar effects
of pH upon SC>2 removal have been reported (References 11, 12).
5.1.2 TCA System
^if ^^
The results of the EPA''~ TCA limestone runs are summarized in Figures
5-5 through 5-8.
Figure 5-5 shows the effect of height of spheres (5 and 10 inches/stage)
and gas rate on SCu removal in the TCA system with six grids and three
stages. The effect of spheres versus no spheres in the six-grid TCA
system on SC>2 removal is illustrated in Figure 5-6.
The liquor and gas rate effects on SCu removal in the four-grid three-
stage TCA system are presented in Figure 5-7. Figure 5-8 is a cross-
plot of Figure 5-7, showing the effect of liquid-to-gas ratio and gas
velocity on SC>2 removal.
From November 4, 1972, to January 15, 1973, TVA conducted a spec-
ial series of tests with the TCA scrubber to provide process and equip-
ment scale-up and design information for the 550 Mw coal-fired TVA
Widows Creek Unit 8 retrofit limestone scrubbing system. The re-
sults shown on Figures 5-5 through 5-8 do not include the TVA tests.
5-24
-------
100
95 ••
O 85
UJ
ex.
U
a:
80 -•
75 --
70 -•
65
T 1 1 1 1 r
90 -• J-
(7.1-7.9 in.HO)
1(4.4-5.1 in.H0O)
n 1 1 1 1 1 1 1 r
1(12-15 in.H2O)
(8.2-9.9in.H2O)
1(5.9-7.7 in.H2O)
H 1 1 h
LIQUORRATE= 1,170-1,220 gpm
SO2 INLET CONCENTRATION = 2,400-3,300 ppm
STOICHIOMETRIC RATIO = 1.4-2.3 moles CaC
mole SO. absorbed
PERCENT SOLIDS = 6.5-10.5%
HOLD TANK RESIDENCE TIME = 4.6 min.
SCRUBBER OUTLET LIQUOR TEMP. = 111-125° F
HEIGHT OF SPHERES
5 INCHES/STAGE
10 INCHES/STAGE
NUMBERS IN PARENTHESES REPRESENT TOTAL
PRESSURE DROPS (EXCLUDING DEMISTER).
H 1 \ 1 f
H 1-
15,000
20,000
GAS RATE, acfm
25,000
'280°F
30,000
Figure 5-5. Effect of Height of Spheres and Gas Rate on SO2
Removal in the Six-Grid Three-Stage TCA System
5-25
-------
100
95 --
90 -•
., «s ••
O
I
75 -•
70 ••
65 --
60
102-118 F
(5.5in.H2O)
6 (6.2in.H,
0(9.6 In. HO)
(2.0in.H20)
(2.5in,H20)
)(3.6in.H20)-
LIQUOR RATE = 1,190-1,210 gpm
SO2 INLET CONCENTRATION = 1,700-2,950 PPm
STOICHIOMETRIC RATIO >1.5 moles CaCOg/mole SOj absorbed
PERCENT SOLIDS = 7.5-11.5%
HOLD TANK RESIDENCE TIME = 4.6 min.
SCRUBBER OUTLET LIQUOR TEMP. = 110-120° F (EXCEPT AS NOTED)
HEIGHT OF SPHERES
O 5 INCHES/STAGE (3 STAGES)
D NO SPHERES
NUMBERS IN PARENTHESES REPRESENT TOTAL PRESSURE DROPS
(EXCLUDING DEMISTER & KOCH TRAY).
-I 1 1 1 1 1 1 1 1 1 1
15,000
20,000
GAS FLOW RATE,acfm @ 280 °F
25,000
Figure 5-6. Effect of Spheres Versus No Spheres and Gas Rate on
SC>2 Removal in the Six-Grid TCA System
5-26
-------
100
95 -•
90 --
85 . .LIQUOR RATE=900 gpm
80 -•
z
LU
U
LU
a.
70 -•
65 -•
60 --
55 --
50
T 1 1 1 T
LIQUOR RATE=1200 gpm
T 1 1 T
T 1 T
|(7.0in.H20)
(3.8in.H20)
LIQUOR RATE=600 gpm
(3.5 in. HO)
(3.0m.H20)
((4.4 in.H0O)
~-^ t-
SO2 INLET CONCENTRATION = 1,800-2,500 ppm
STOICHIOMETRIC RATIO > 1.75 moles
SO9 absorbed
PERCENT SOLIDS = 6-11%
HOLD TANK RESIDENCE TIME = 18-35 min.
SCRUBBER OUTLET LIQUOR TEMP. = 111-123° F
HEIGHT OF SPHERES = 5 INCHES/STAGE
NUMBERS IN PARENTHESES REPRESENT TOTAL PRESSURE DROPS
(EXCLUDING DEMISTER & KOCH TRAY).
-I 1 1 1 1 1 1 1 1 1 1 1 1
15,000
20,000
GAS RATE,acfm @ 280 °F
25,000
Figure 5-7. Effect of Liquor and Gas Rate on SO2 Removal in the
Four-Grid Three-Stage TCA System
5-27
-------
100
95 -•
90 -•
85 -
< 80
i
es .,,.
O 75
ff 70
65 -•
60 ••
55 ••
50
7.8 ft/sec
9.8 ft/sec
5.9 ft/sec
SO2 INLET CONCENTRATION = 1,800-2,500 ppm
STOICHIOMETRIC RATIO > 1.75 moles CaCOymole SO2 absorbed
PERCENT SOLIDS = 6-11%
HOLD TANK RESIDENCE TIME = 18-35 min.
SCRUBBER OUTLET LIQUOR TEMP. = 111-123° F
HEIGHT OF SPHERES = 5 INCHES/STAGE
4-
20 30 40 50 60 70
LIQUID-TO-GAS RATIO, gal/mcf
80
90
Figure 5-8. Effect of Liquid-To-Gas Ratio and Gas Velocity on
SO2 Removal in the Four-Grid Three-Stage TCA
System
5-28
-------
The variation in SC^ removal for 5 inches of spheres per stage in the
six-grid, three-stage TCA (shown as open circles in Figures 5-5 and
5-6) is attributed, mainly, to differing average values of SC>2 inlet con-
centrations. In Table 5-6, the SC>2 removals and operating conditions
for these runs have been compared.
R. H. Borgwardt (Reference 12) of EPA has reported that, for his pilot
scale TCA system (see Section 2. 3), the percent SO removal is inversely
£,
proportional to approximately the one-tenth power of inlet SOo concentra-
tion. The difference of five percent in SCU removal between Runs 409-2A,
and 416-2A, for average SC^ concentration differences of 3000 and 2000 ppm,
respectively, is in agreement with the EPA pilot results. The eight percent
difference in the SC>2 removal between Runs 410-2A and 415-2A is attri-
butable both to the differences in inlet SC>2 concentrations and the differ-
ences in the scrubber outlet liquor temperatures.
5. 1. 3 Hydro-Filter System
Figure 5-9 summarizes the effect of gas and liquor flow rates on SO2 re-
moval in the Hydro-Filter system with five inches of marble-bed height.
Figure 5-10 is a cross-plot of Figure 5-9, showing the effect of liquid-to-
gas ratio and gas velocity.
The extended dash-lines showing the ranges of SC>2 removal in Figures
5-9 and 5-10 indicate where the ranges of SC>2 removal would have been
if corrections had not been made (for the deterioration of the DuPont
analyzer optical filters) for these runs. As mentioned previously, there
is some doubt about the accuracy of these corrections at low SO?
removals.
5-29
-------
Table 5-6
EFFECT OF INLET SO2 CONCENTRATION ON SO2
REMOVAL IN A SIX-GRID THREE-STAGE TCA
Run No.
SO2 Removal, %
Inlet SC>2 Cone. , ppm
Gas Rate, acfm @ 280°F
Liquor Rate, gpm
Stoichiometric Ratio
Scrubber Outlet Liquor
Temperature, °F
Percent Solids Recirc.
Hold Tank Res. Time, min.
Ht. of Spheres/Stage, in.
Pressure Drop, in. P^O
409-2A
&
414- 2A
90+3
2800-3250
20, 100
1, 190
1.4-1.6
112-122
7-11
4.6
5
5. 9-7. 7
416-2A
95+1
1750-2200
20, 000
1, 195
>1.5
111-118
8-9
4.6
5
5. 8-6. 6
410-2A
&
411-2A
87+3
2500-3150
15, 100
1, 180
1.2-2. 0
111-120
7-8. 5
4. 6
5
4.4-5. 1
415-2A
95+2*
2250-2750
15, 250
1,200
>1.5
102-118*
7-11
4.6
5
5. 0-6. 0
'":-is:h removal may also be due to lower outlet liquor temperature.
5-30
-------
100 •-
o
o
I—
z
80 •-
60 -•
40 -.
20 --
TOTAL LIQUOR RATE = 400 gpm
TOTAL LIQUOR RATE = 600 gpm
TOTAL LIQUOR RATE = 800 gpm
SO. INLET CONCENTRATION = 2,400-3,200 ppm
STOICHIOMETR1C RATIO > 1.75 moles
mole SO- absorbed
PERCENT SOLIDS = 5-7%
HOLD TANK RESIDENCE TIME = 50 mjn.
SCRUBBER OUTLET LIQUOR TEMP. = 115-125° F
(9-11 m.H20)
'(8-10 Jn.H.O)
1(10-12 in.KLO)
(8-95n.H20)
(9-10in.H20)
NUMBERS IN PARENTHESES REPRESENT HYDRO-FILTER PRESSURE
DROPS (EXCLUDING DEMISTER) IN A SCALE-FREE BED.
H 1 1 1-
20,000 25,000
GAS FLOW RATE,acfm @ 330 °F
1
30,000
Figure 5-9. Effect of Gas and Liquor Flow Rates on SOo Removal
in the Hydro-Filter with Five Inches of Marbles
5-31
-------
100
80 --
60 --
oe.
cs
O
40 ••
04
LU
O.
20 ••
10
SO2 INLET CONCENTRATION - 2,400-3,200 ppm
STOICHIOMETRIC RATIO> 1.75 moles CaCOg/
mole SO« absorbed
PERCENT SOLIDS = 5-7%
HOLD TANK RESIDENCE TIME = 50 min.
SCRUBBER OUTLET LIQUOR TEMP. = 115-125° F
20
30 40 50
LIQUID-TO-GAS RATIO, gal/mcf
60
70
Figure 5-10, Effect of Liquid-To-Gas Ratio and Gas Velocity on
SO2 Removal in the Hydro-Filter with Five Inches
of Marbles
5-32
-------
5. 2 ANALYTICAL RESULTS
A comparison between measured and predicted liquid and solids analy-
tical data for the venturi and TCA systems during open liquor loop short-
term factorial testing is presented in Section 10. 3. 2. Analytical data
for closed liquor loop reliability verification testing is presented in Sec-
tion 6. 1.
5. 2. 1 Liquid Data
Table 5-7 shows the average scrubber inlet liquor compositions for the
open-loop factorial test runs. During the period of factorial testing,
there did not appear to be a continual build-up of magnesium, sodium
or chloride ions within the liquor. The large concentrations of chloride
ions are attributable to chlorides present in the coal which were con-
verted to HC1 and absorbed from the flue gas in the scrubber. A.
Saleem (Reference 14) of Ontario Hydro has reported similar chloride con-
centrations during limestone wet-scrubbing tests with flue gas from a coal-
fired boiler.
Table 5-7 indicates that the venturi and Hydro-Filter systems had lower
overall dissolved solids than the TCA system. This was expected, since
the quantity of input raw water for these systems was greater than for the
TCA system (the TCA system liquor is close to the predicted "saturation"
level for CaSO4).
The liquid analytical data are tested for consistency by inputting the
measured compositions and pH's into a modified Radian Equilibrium
Computer Program (Reference 13), which then predicts the ionic im-
balance. For the data shown in Table 5-7, the ionic imbalances were
all less than 10 percent.
5-33
-------
Table 5-7
AVERAGE LIQUOR COMPOSITIONS AT THE SHAWNEE TEST
FACILITY DURING OCTOBER, 1972
System
Species Concentration, mg/1 (ppm)
_
so3
Venturi 200
TCA 300
Hydro-
Filter 250
_
co3
140
60
100
_
so4
600
1700
1200
++
Ca
1000
1300
800
Mg
60
100
40
+
Na
50
50
30
Cl
1400
1300
800
Total
3500
4800
3200
5. 2. 2 Solids Data
Analyses of the Fredonia Valley limestone used at the Shawnee facility
showed an average composition of 90% CaCO_, 5% MgCO and 5% inerts.
J J
Dry sieve analyses showed approximately 90 percent of the ground lime-
'!*
stone passing through 325 mesh. A MikroPul' particle size analysis
showed approximately 7% of the ground limestone less than 3 microns,
30% less than 6 microns and 85% less than 20 microns.
The coal burned in boiler No. 10 during these limestone tests is Old Ben
24 and contains approximately 18% ash, 10% total moisture, 3.2% sulfur
and 0. 3% chloride. The analyses of ash from boiler No. 10 showed about
50% Si02, 18% A12O3, !6%Fe2O3, 7% CaO, 1.3%MgO, 1. 3% SOj, 2.3%
K2O, 1% Na2O and 3. 2% ignition loss.
A division of United States Filter Corporation.
5-34
-------
The composition of solids in the slurry is fixed by the moles CaCO.,
added per mole SC>2 absorbed (stoichiometric ratio), the overall percent
oxidation of sulfite to sulfate within the system and the percent of fly ash.
The mole percent oxidations averaged approximately 30 percent during
the open-loop factorial testing and the flyash comprised from 30 to 50
wt % of the solids for the three scrubber systems.
5. 3 PARTICIPATE REMOVAL RESULTS
Particulate removal results for the three scrubbers are presented in
Tables 5-8, 5-9, and 5-10. Only those data which were taken at close-
to-isokinetic sampling conditions have been included in the tables. All
of the outlet p articulate data have been corrected for soot-contamination
from the gas reheaters. The soot amounted to less than 30 percent of
the total mass of the outlet particulates.
During the open-loop factorial testing there were solids accumulations
(depositions) in the demisters for much of the test period. These solids
accumulations may explain some of the very low measured outlet grain
loadings in Tables 5-8, 5-9, and 5-10, especially for the TCA (multi-
grid tower) at 1. 5 inches H^O of pressure drop.
During open-loop factorial testing, the demisters were all washed from
above with raw water (see Figures 2-1, 2-2, and 2-3). During the
boiler outage, provisions were made for washing the demisters from
below, with a mixture of clarified liquor and raw water, and for the
installation of a Koch tray in the TCA scrubber (Figures 7-1, 7-2, and
7-3).
5-35
-------
00
I
ID
rt)
H
ti
W
PQ
rf
U
CO
W
O
H
<5 CO
tf H
co W
PH
H-l -J
PH- ^
D ^
H H
**
W <
> fa
fa1
ff!
W
H
O
P
'c 'fl
4) ;
u c
l«
M
C
•H
T}
3 j°
•l-<
rd
(H
O
PL
O
JH
qj h-j
fH
05 C
05 "-1
Q)
>H
DH
Q)
rt
^ PL
2 M
p
3
-
0) «
<« rt
0 rt
u
•M
rt
Q
6
c
tf
H
3
>
)
H
;
^
-i
•M
0)
tj
0
^
Q)
£
O
O
0s
..
~
fM
t^-
1
r-
tM
tM
U
I
00
*
00
•
(7^
0s
O
O
o
^o
.
fM
m
fM
in
.
rM
o
~*o
"t
fM
1— t
O
o
o^
„
1 — 1
ro
r--
i
r-H
CO
i
fM
_*;,
rn
i
rO
^t
m
.
00
Is-
0
o
(M
^
.
^
in
tM
m
r-
.
0
o
m
r^
tM
l-H
O
O
O^
*,
l-H
ro
t —
I
•^
O
1
1— t
*«
1
5
ro
00
vD
in
o
0
00
PO
.
PO
m
rM
o
r-
,
o
o
in
^
fM
i-H
O
O
a*
*
i — i
ro
f^
1
in
O
i
i— i
%
1
sO
5
-Q
flj
>. I
c
o
!H
(L)
I
rS
t-,
CO
O
P-
5-36
-------
d
H
W
W
ffi
O
ffi
H
^^
M ^
g^
C ) •-<
«g
^
>
O
w
H
H
U
i—i
H
CU
"G
0)
u
(H
tn
W)
-
G
TJ
rt
O
^
fl
•i-H
rt
O
c
c
tn
CO *
CP
^
c
0 4,"
cr1 rt
3 tf
*
to
-M
rt
Q
6
^
c
•H
>
ouiavr
-4-1
i — 1
O
4->
»-H
a
i — i
D
M
q
—i
i
^
•H
3
1
OJD
O
o
o-i
©
o
rt
oo co ND in -Nf
QN QN 00 ON ON
ON ON ON ON ON
O CM 1-1 I-H CM
O O O O O
O O O O O
o NO CM ON m
f~ i— i ro CM NO
i — t ^ i-H CO CO
m in in un o
i — 1 t— 1 i-H f-H i-H
o o o in m
01 oo ro ^f t~-
r^ r^ r^ t~- r<^
0 O 0 O O
o o o o o
CM CO CO ^ CO
ON QN CJN ON ON
i — 1 i-H i-H i — li — t
CM CO CO CO CO
1 1 1 1 1
I-H s^3 ON CM "Nf
CM O O i-H i— i
1 1 1 1 1
CM i — I I-H i — i — i
<; <; _ CM
m in m i— i I-H
i i i i i
U U U U U
en
to
Q)
rt
.!->
O
" «
in
(U
U
n M
-*->
QJ QJ
in
0)
O
!H
en
tn
OJ
a) o
D C
0)
h
ro ra
0) G
S
H
5-37
-------
I
un
n)
H
W
O
O
o
OJ
ID
Q
trJ
W
W
CQ
U
w
oJ
W
H
fc
i
O
Q
K
o
2
W
p{
W
H
{3
O
hH
H
tf
<^
PM
!H !P
a) n
" 6
1> 0)
PH P4
<+-{ -M
O 0
-•U. 'M
00 Q
BO
G
n)
o
( i -*->
J5
•^ M
cti
O
O
4> ffi3
IH l1*
r-t •
» .3
4) -
£ °
" ^
O
^ .
§ JB g
CJ4 nJ CL
'^J ^ tuO
fa
0
CO
m *i CO
en cu
Cti ~tj fr)")
,K nj N2/
05 g
U
rt
K
-*-1
cti
P
6
^
c
?
rt
00 CJN CO CM
00 00 ON 0s
ON ON ON ON
o in CM co
CO CO CO CO
o o o o
o o o o
CM CO ^
NO CO ^ CM
^ ^ ^ ^
o o o m
CM CM O CM
o o o o
(— I I — 1 r— * O
00 CO 00 00
O O 0 0
o o o o
O 0 0 O
o o o o
CM CM CM CM
CM CM CM CM
CN- r — r — r —
i i i i
CO xO 00 CM
— i I-H CM O
1 1 1 1
i-H i-H i-l PO
< < PQ U
co co co co
till
CM CM CM CM
^P •* ^P -^
CO
00
r-
CM
o
o
ON
-*
CM
O
r-l
m
o
00
o
o
o
o
CM
CM
f^
1
^
CM
1
CM
m
co
i
r^
CM
t
CO
ON
ON
in
CM
O
o
00
^
CO
m
i-H
O
r-H
CO
o
o
o
0
CM
CM
[ —
1
oo
CM
1
CM
<
CO
1
00
CM
^P
NO
ON
O
i— 1
O
o
CM
i-H
,-JH
m
i-H
O
r— I
00
0
o
o
0
CM
CM
f —
1
ON
CM
1
CM
CO
1
00
CM
^
O
ON
m
CO
0
o
CO
NO
CO
m
•-<
o
T— 1
00
0
o
o
o
(M
Is-
1
O
co
1
CM
co
i
00
CM
TP
IT)
ON
o
CM
O
0
0
CM
^f
0
[N-
o
o
o
o
ON
i — 1
CO
f^
1
f^
0
1
-^
-------
Section 6
CLOSED LIQUOR LOOP RELIABILITY VERIFICATION
TEST RESULTS
The objects of the closed liquor loop reliability verification tests are to:
(1) identify areas or regions for reliable operation consistent with rea-
sonable SO removal; (2) choose attractive operating configurations from
within these regions; (3) obtain more reliable material balances; and,
(4) quantify any variations in SO_ and particulate removal and system
slurry compositions with time.
The initial tests are to be run at reduced scrubber inlet liquor pH
(5. 7-5. 9) to improve system reliability and increase limestone utiliza-
tion. System reliability can be improved by higher oxidation rates at
the reduced pH's, resulting in a larger percentage of "seed" CaSO^
crystals within the process slurry and in improved solids separation.
A modest reduction in SO0 removal (from high-pH performance) is the
LJ
price of the increased system reliability and limestone utilization (see
Table 5-5 and Figure 5-4).
The performance data for the initial reliability verification test runs
on the three scrubber systems are presented in Section 6. 1 and the re-
sults of material balances for sulfur and calcium (which are satisfac-
tory) are presented in Section 6. 2.
6-1
-------
6. 1 PERFORMANCE DATA
Data for the first 400-500 hours of operation on the initial reliability
verification test runs are summarized in Figures 6-1, 6-2, and 6-3 for
the venturi, TCA, and Hydro-Filter systems, respectively. The upper
plot of each figure shows the operating periods (blank space indicates
shut-down), and such critical variables as SO_ removal, liquor pH and
LJ
stoichiometric ratio. The middle plot of each figure gives some analy-
ses of the solids in the scrubber inlet liquor and the lower plot gives
concentrations of some dissolved species in the scrubber inlet liquor.
Also shown in Figures 6-1, 6-2, and 6-3 are the depletion (line-out)
periods for the tests. Fresh limestone slurries (no CaSO "seeding")
were introduced in the effluent hold tanks, and SCU removal was used to
reduce the slurry pH until the desired level of SO-> removal was attained.
This level was approximately 10 percent below that attainable in high-pH,
open-loop, operation (see Figure 5-4).
Before addition of the limestone, the systems were inspected to make
certain they were free of any scaling or erosion that might have occurred
during the high-pH period of the line-out. Inspection shut-downs are
scheduled periodically (approximately weekly) to monitor scaling and
erosion in sensitive areas of the systems.
Operability and reliability of the three scrubber systems during the ini-
tial runs are discussed in Section 7.
An overall summary of the initial run data appears in Table 6-1, which
presents average values for some significant parameters (from Figures
6-1, 6-2, and 6-3).
6-2
-------
• UGtN UN JOI-1A
Gu Ritt = 20.000 icfm @ 330 °F
Liquor Riu to Vantun = BOO gpm
Liquor Riu to Spriy Toxtr - 600 gpm
Spray Towtr L/G = 40 gil/mcf
Spriy Town- Gu Vtlocity = 50 It/me
E.H T Rnidinct Timt = 20 mm
No of Spriy Huteri = 2
G« Into! S02 Cone. • 2.600-3.300 ppm
Scrubbtt InM Liquor Timp = 120-125 °F
Liquid Conductmtv - 7.000-15.000 M mhoi/cm
DiBhirgi (Clwrfitr) Solidi Cone. • 23-27 wt %
TEST TIME, h
VIS I VIA
CALENDAR DAY
14,000
11,000
10,000
1 1,000
«• 6,000
s
o
^ 4,000
I
5 2,000
J f — • _
• TOTAL DISSOLVED SOLIDS A) MAGNESIUM (Mg ** )
O CALCIUM (Co ** ) A SODIUM (No * )
Q SULFATE (SO4 = ) • SULFITi (SOj1 )
A CMLOIIDE (Cl -) O CAMONATI (COj= )
-
"
-
_ • A
\ * * :
u,ooo
12,000
10,000
8,000
'
4,000
2,000
0
iiiir"
TtST TIME, hoim
VI i t 4/)6
CALENDAR DAV
FIGURE 6-1.
OPERATING DATA FOR VENTURI
RUN 501-1A
6-3
-------
•JN 81-IA CONTINUED
l A SHAY TOWER , - VERTURI & SPRAY TOW» r— VtNTUHI ft S VAY TOWER
(Ul) / OUTLET (LAI) / INLET (IN-Lir4£ METEk
V2I I 4/22 I 4/23 I 4/24 I 4/25 t 4/24 I 4/27 I 4/28 I
Gd« Ran - 20.000 ecfm sj> 330 °F
Liquor Rut to Vtntun = 600 gpm
Liquor Rit< to Spray To«Mr * 600 gpm
Sixty TOOK L/G = 40 gil/mcf
Spray Tower G« Velocity • 5.0 ft/nc
E.H.T. Rendence Time - 20 mm
No. of Spray Heederi' 2
Ga Intel SO; Cone. ' 2,400-3.200 ppm
Scrubber Inltt Liquor Temp - 120-125 °f
Liquid Conductivity - B.MO-16.500 jimboi/cm
Dacherje ICterrlier) Solids Cone • 20-27 wt %
TEST TIME, h
4/33 I 4/U |
CALENDAt DAV
14,000
1 2,000
10.000
^ 8,000
O 4,000
§
£ 4,000
Z
| 2,000
1
• TOTAL DISSOLVED SOLIDS + MAGNEStUM (Mg ** )
O CALCIUM (Co ** 1 A SODIUM (No * )
D SULFATE (SO4 ) • SULFITE (SO," )
A CMLC*IDE (Cl - 1 O CAHONATE (CO3 )
-
* •
-
A
A *
o o o
._ a D a
14,000
12,000
10,000
8,000
6,000
4,000
2,000
0
0 «0
s
£ soo
I «o
300
300
100
• o
* * •
OB "
A o a
i i i i i i i i i i i i n • i i i i i i i i i
2fl0300320MOU03K400 4204«,4«04B030052I
TEST TIME, Ham
4/11 1 4/22 1 V33 1 4/« t 4/35 1 4/2* 1 4/17 1 4/38 1 4/39 1 4/W 1
CALENDAI DAY
FIGURE 6-1 (CONTINUED).
OPERATING DATA FOR VENTURI
RUN 501-1A
6-4
-------
Gat Rm = 20.000 »cfm 0 300 °F
Liquor Riu - 1,200 9pm
L/G ' 80 gil/mcf
Gn Vtlocity • 7.8 ft/in
E.H.T. RnMfnct Tm» = 20 mm
Thru St«OB. 5 in s
GB Inltt S02 Cone. - 2,200-3.200 ppm
Soubbtr Inltt Liquor Ttmp. = 116-125 "f
Liquid Conducltvily = 4,MO-tO,000 M mhoi/cm
Duchtroi (ClHrfnO Solidi Cone - 25-39 wl %
100 120
TEST TIME, hour*
3/23 I 3/24 1 V25 I 3/24 t 3/27 I 3/28 I 3/29
CALENDAR DAY
160 1BO 200
3/30 I
§ 3*
o 2 •
1
s a ,-
I Sa
INSOLUBLES (ASK)
TOTAL SULFUR (SO.)
TEST TIME, houn
VM I 3/5* I 3/M I 3/27
CALENDAR DAY
3/2$ I 3/29 |
230 240
3/31 I
14,000
12,000
10,000
1
•v 8,000
§ 4,000
2
5 4,000
Z
* 2,000
Z 0
• TOTAL DISSOLVED SOLIDS • MAGNESIUM (Mg ** )
O CALCIUM (Co ** ) & SODIUM (N« + )
O SUtFATE (S04 - ) • SULFITE
-------
•UN SOI-2A CONTINUCD
m an
4/1 < VI 14/3
310 340 340 3M «0
TEST TIME, howi
4/4 1 4/S 1 */fc I 4/7 t
2;* "
£ 06 »
u SI <"
• gg.
U.OOO
12,000
10,000
_ 8,000
8" 6,000
o
£ «,000
z
| 3,000
s
1/1 0
• TOTAL DISSOLVED SOLIDS + MAGNESIUM (Mg ** )
<> OLCIUM (Ca ** 1 A SODIUM (NO * )
D SULFATE (SO4 ) • SULFITE (SO. '
A CMLOWDE (Cl - ) O CAtlONATE (COj )
-
• ;
•
A
OD ~
14,000
12,000
10,000
8,000
6,000
4,000
2,000
0
Gu Ran - 20.000 ufffl * 300 "F
Liquor Ritt - 1,200 gpm
L/G-80j«l/md
G«Vilocit>.7.8ft/«c
E.H.T. RMtfina THM = 20 mm
Thru Sups, s m iplwrn/ntgt
G« Ink! S02 Cone. * 2.300-3.300 ppm
Scrublw Inltt Liquor Timp. * 117-127 °F
Liquid Conductinty • 5,400-19.500 M mhoi/cm
Oatlwin (CUtitiw) Solnta Cone. • 30-M wt %
TEST TIME. Ko»n
CALENOAt DAY
6-6
FIGURE 6-2 (CONTINUED).
OPERATING DATA FOR TCA
RUN 501-2A
-------
•UN S01-2A CONTMUCO
II
N
SHU! DOWN FO* CLEANING
| Of INLET DUCT
VINT I HOC UMOVfO
• TOTAL, EXCLUDING DfMISTER 1 KOCH TtAY
--{^MISTER & KOCH TRAY
V- OUTLET (I
00
Gn Ratt = 20,000 actm (3 300 "F
Liquor Rat* = 1,200 gpm
L/6 = 80 gil/mcf
Git Wtocity = 78 ft/»t
E.H.T Rradtnct Tim* = 20 mm
ThrnSttgti, 5 in phn«/>ugi
Gil Inlet S02 Cone = 2,300-3,200 ppm
Soubbtr Inltt Liquor Ttmp = 117 128 °F
Liquid Conductivity = 12,200-17,600 u.mhoi/cm
Onch>rg< (Clwrfnr) Solids Cone • 25 33 »t s
__!_
-J
480S»5Mi40i60S80600«06406«)680
TEST TIME, hoon
4/11 I 4/12 I 4/13 I 4/14 I 4/15 I 4/14 I 4/1,' I 4/IB 1 4/19
CALENDAR DAY
s
85*
3 I i
| IS
| sg
3 S "
INSOLUM.ES (ASH)
i 2*.
Hi
0—
o-
- TOTAL SULFUR [SCU
MO
4/13
TEST TIME, houn
CALENDAR DAY
4/17 I 4/IB I 4/JS
u.ooo
12,000
10,000
1
1
- 8,000
f
O d,000
g
£ 4,000
z
a 2,000
5
z °
• TOTAL DISSOLVED SOLIDS * MAGNESIUM (Mg ** )
O CALCIUM (C« ** ) A SODIUM (No * )
O SUIUTE (504 = ) • SULFtJt (SO. " )
4t CMLQtIOE (Ci ' ) O CAR»ONATE (CO3 - )
-
-
.
A
.
14,000
12,000
10,000
8,000
6.000
4,000
2,000
0
4/12 I */13 I 4/14 t
TEST TIME, hou«
4/15 I 4/14
CALENDAR DAY
4/18 I 4/19 I 4/20 I
6-7
FIGURE 6-2 (CONTINUED).
OPERATING DATA FOR TCA
RUN 501-2A
-------
f -°^8
/ .pf-^
TIST TIME, Iwun
VI* I V»
CALENOAXOAY
14.000
12,000
10.000
{M.
g »,ooo
! ""
| 2,000
™ 0
• TOTAL DISSOLVED SOtlDS 4> MAGNESIUM (Ml **)
O CALCIUM (C, « ) a SODIUM (N. * 1
D SULFATE (SO4 • ) • SULFITI (SOj - 1
A CHIOWOUCI -1 O CAHONATE KO, • )
.
*
- o 8I
" "
14,000
12,000
10,000
8,000
4,000
4,000
2,000
0
G« Rill = 20.000 idm* 330 °F
Liquor Ritt * 100 gpm (tonl)
L/G = 53j«l/mcl
GiiV«kKitv-5.1tt/t«
E.H.T. Rwdtnct Timt- 30 mm
Mvbli B«l Htuht - 3.5 n
Gil Inltl S02 Cone. • 2,700-3,300 ppm
Senibbif Intat Liquor Ttfflp. ' 111-125 °F
Liquid Conductmty • 6.500-17.000 JL mhot/cm
Dtwhirgi (Ctarilwr) Soldi Cone. - 22-25 M S
I Vis 1 Vi*
TEST TIMC, hawn
VI* I V20
CALENDAR DAY
I 3/22 1 V23
FIGURE 6-3.
OPERATING DATA FOR HYDRO-FILTER
RUN 501-3A
6-8
-------
IND UN »I-U I
! MOW *JN JBI-M
8 }J M
' II.
SVSTU4DMMDTO
•uovionus
UM-OOt IWJKTION
°00 ~
INUT (IN-UM Mtm)
OOWNCOMEI OUTLET AM)
1 1 i 1 1 1 1 1 1 1 1- — 1 1 1 1 1 1— 1 1 1 J 1 1
TEST TIMC, hMM
tV»'V»*IV»7'W»IV»IV*>l 3/31 1 4/1 14/2 | 4/S
CALENOUDAY
1.6
1.4
1.2
1.0
Gi> RIM - 20,000 Kim ffl 330 °F
Liquor Rtrt = 800 gpm (Ion!)
L/C • 53 gal/mcf
GilVllocitv'5.1 ft/tic
E.H.T RnidinctTin»*30mm
Mwblt Bid Hnght = 3.5 in
Gil Into S02 Cone. * 2,500-3,200 ppm
Scrubbtf Into! Liquor Tfmp -117-125 °F
Liquid Conductivity « 7,700-14,000 )L mhot/cm
Ductwgt (Clintltr) Solids Cone. • 19-29 wt %
li •
y~ TOTAl
—sC .
-^r"r
MSOIUIUIIUMI
TOTAL SULHM
(SOj)
—o^p
^-
-.s
—6
iOi j—tuumeaf
a
TBTTIMI, taun
I V17 I V»t 1 VW * Vlo t 3/31 t 4/t
OUfMDAIOAY
U.100
12,00
•.on
1
V"0
§ 4.000
JM.
1 2,000
n
z •
0 400
| OT
« 40B
300
200
100
0
24
• TOTAi MttOLVEO SOUK • MAONBHJM **»**)
O CAlOUMfC*++> ^ SOCHUWpH,*)
O sutMTf do4') • suvnipo,- )
A CHlOoWiri*) O CAMONAT1 (CO, - )
_
•
.-
• *
o *"
D 0"
a-
•
•
• •
*
0 *
'
03IO]K3003I13<0340300400410t«4404l
14,000
t2,000
10,000
1,000
4,000
4,000
2,000
0
400
300
400
300
200
100
0
0
FIGURE 6-3 (CONTINUED).
100 OPERATING DATA FOR HYDRO-FILTER
RUN 501-3A &3B
TEST TIME, town
V» I V* I 2/57 t VM I V*f 1 3/» I VSl I
CALoMDWMY
I4/2IV9
6-9
-------
RUN JOI-M CONTINUfO
I 1 U—I L 1 1 1 ' 1 1.1 i . 1
Hi"
s *
if!
j ?§
gl:
4/1 i 4/r
TEST TIMt, hew*
4/1 i 4/» i 4/w
CALENDAR DAY
vn i «/i;
Gu Rate * 20,000 iclm '" 330 °F
Liquor Rltt = 800 gum (lotil)
l/G-SSjll/mcf
Gu Vtlocitv « i 1 ft/we
E.HT rtwdinc«Timi = 3l>min
Micbll Btd Htight = 3.5 in
Gil Inltl SO; Cone. = 2.500-3,300 ppm
Scrubbw Ink! Liquor Timp. = 117-126 °F
Liquid Conductivity * 6,300-13,700 u, mhos/cm
Ditchirgt ICIvifw) Solids Cone. > 20-29 wt S
s s
a s •
.
18} ,
S O O
TOTAL SULFUI (SO.)
TEST TIME, haun
14,000
12,000
10,000
•* 0,000
5 *'°°°
JMCR INLET UQU
i i
y o
• TOTAL DISSOLVED SOLIDS
O CALCIUM (Co ** >
D SW.FATE (S04 - )
A CHLOtIK (0 ' )
•
A
.
• MAGNESIUM (Mg ** )
A SODIUM (No * )
• SULFITE fiO}" )
O CAtiONATE
-------
Table 6-1
AVERAGE CONDITIONS FOR INITIAL
RELIABILITY VERIFICATION RUNS
Parameters
Operating time, hrs
Gas velocity, ft/sec
L/G, gal/mcf
Pressure drop, in. H^O
Percent solids recirculated
Percent SO removal
Stoichiometric ratio
Limestone utilization
Inlet liquor pH
Percent oxidation
Dissolved solids, ppm
Hydro-Filter
Run 501-3A
520
5
53
9
11
65-70
1.25
80%
5.8
30
8000
TCA
Run 501-2A
550
7. 8
80
6
15
80-85
1.20
83%
5. 8
20-30
7500
Venturi
Run 501-1A
410
5a
40b
10. 5°
15
70-75
1. 5
67%
5. 8-5. 9
15
7000
a) Spray tower
b) L/G's of 40 for spray tower and 40 for venturi.
c) Nine inches across venturi and 1.5 inches across spray tower.
A summary of the liquid analytical data is presented in Table 6-2. Most
dissolved species appear to have approached steady state concentrations
during the period of operation. However, magnesium ion (Mg ) con-
centration exhibited a steady increase in the venturi and TCA systems
(see Figures 6-1 and 6-2).
It is of interest to compare the liquid analytical data for the open-loop
factorial and closed-loop reliability verification test runs (see Tables 5-7
6-11
-------
and 6-2). For the TCA system, sulfate concentrations for both periods
of operation were close to the "saturation" levels. For the venturi and
Hydro-Filter systems, the sulfate concentrations during the initial closed-
loop runs were well above that measured in the open-loop test runs. As
expected, in all three systems, the level of chlorides and total dissolved
solids during the closed-loop runs was far greater than that obtained dur-
ing open-loop testing.
Lack of confidence in the long-term reliability of the in-line pH meters
led to a decision to control SO- removal in the initial tests, rather than
to directly control pH within the desired 5. 7 to 5. 9 region. Results of
open-loop limestone depletion runs were used to estimate SO removals
consistent with the desired pH (see Figure 5-4). In general, these SO9
removal levels were selected at about 10 percent below that attainable
in open-loop operations at a pH above 6. 0 (high-pH). Control of SO~ re
i-4
moval was established by varying the rate of limestone addition.
Table 6-2
AVERAGE LIQUOR COMPOSITIONS FOR INITIAL
RELIABILITY VERIFICATION TEST RUNS
Scrubber
System
Liquor
so3= co3=
Venturi 200 200
TCA 150 150
Hydro-
Filter 300 150
Species
so/
1500
1800
1800
Concentrations, mg/1 (ppm)
Ca++
2000
2000
2000
Mg++
250a
300a
200
Na+
50
50
50
Cl"
3000
3000
3500
Total
7200
7400
8000
This species increased gradually throughout the time period. The
values shown on this table are the maximum values, at the end of
the plotted periods.
6-12
-------
6. 1. 1 Venturi Run 501-1A (see Figure 6-1 and Table 6-1)
Open-loop factorial testing at high pH indicated SO removals of approxi-
mately 42 percent in the venturi and 57 percent in the spray tower,
which is equivalent to an overall removal, for the combined system, of
75 percent. Thus, to achieve the desired low-pH operation, a target of
65 percent removal was indicated.
From April 14 to April 21, 1973, SO9 removal was controlled at about
C-i
74 percent, instead of the targeted 65 percent. During this period the
average stoichiometric ratio was 1. 5 (moles CaCOo/mole SO2 absorbed)
and the average oxidation was 15 percent.
From April 22 to April 27, 1973, SO removal was controlled at about
70 percent, and the stoichiometric ratio went from 1.3 to 1.8. This in-
crease in stoichiometric ratio, while maintaining the same SO removal,
was indicative of some "degradation" in the system (e.g. , drop in lime-
stone reactivity, erosion of spray nozzles). Oxidation remained at 15
percent and inlet liquor pH remained at 5. 7-6. 0.
Toward the end of the operating period depicted (see Figure 6-1), the
SO_ removal dropped below 70 percent and was restored to a value
slightly above 70 percent. On April 28 (low removal), the stoichiometric
This removal was estimated from Equation 10-9 and "corrected" for
the change in inlet SO concentrations (see Equation 10-10 and Sec-
tion 5. 1. 2)
The high pH removal at the selected venturi run conditions was originally
estimated at 85 percent. The early part of Run 501-1A is thus at a pH
and SO removal somewhat higher than desired.
6-13
-------
ratio was about 1. 5 and increased to above 2. 0 by April 30. The scrub-
ber inlet liquor pH remained at 5. 9-6. 0 during this period.
6.1.2 TCA Run 501-2A (see Figure 6-2 and Table 6-1)
The predicted SO removal value (at high pH) for the TCA system, ope-
rating at the test conditions of Run 501-2A, is about 95 percent (see
Equations 10-5 and 10-10). Therefore, the controlled SO7 removal
for Run 501-2A was chosen at 85 percent. After an initial operating pe-
riod (from March 24 to March 31, 1973) in which there were relatively
large fluctuations in SO removal, a relatively steady period of about
£>
five days ensued (from April 1 to April 6), where the SO removal va-
ried from 80 to 85 percent and the scrubber inlet liquor pH varied from
about 5. 7 to 5. 9. The stoichiometric ratio during this period was about
1.15 (which corresponds to a limestone utilization of about 87 percent),
and the oxidation was about 30 percent.
Toward the latter part of the plotted operating periods (from April 12
to April 15 and from April 18 to April 20), •while the SO removal was
£j
still controlled between 80 and 85 percent, there appeared to be an in-
crease in t e stoichiometric ratio to an average value of 1. 4 (limestone
utilization of 71 percent), which, again, is indicative of "degradation"
in the system (e. g. , drop in limestone reactivity, pluggage of spray
nozzles). The oxidation dropped slightly during these periods to an ave-
rage of about 20 percent and the inlet pH ranged between 5. 7 and 6. 0.
A removal of about 96 percent was obtained in the line-out (depletion)
period for Run 501-2A. This confirms the estimate from the open-loop
factorial data.
6-14
-------
The system was shut down a number of times because solids plugged the
inlet duct in the vicinity of the humidification section and increased the
total pressure drop of the system (see Reference 1 and Section 7).
6.1.3 Hydro-Filter Run 501-3A and 3B (see Figure 6-3 and
Table 6-1)
The predicted SO~ removal value (at high pH) for the Hydro-Filter sys-
L*
tern, operating under the test conditions of Run 501-3A, is about 80 per-
•3*
cent (see Equations 10-7 and 10-11). Therefore, the controlled SO-
Ci
removal target for Run 501-3A was 70 percent. During most of the ope-
rating period for Run 501-3A, the SO? removal was controlled between
65 and 70 percent, the average stoichiometric ratio was about 1.3, and
the average percent oxidation and inlet liquor pH were about 30 percent
and 5. 8, respectively.
After the system was drained to remove debris (marbles) on March 28
and March 29, another depletion (or line-out) period was conducted for
Run 501-3B. From April 3 to April 13, the SO removal was held be-
LJ
tween 65 and 70 percent, and the average stoichiometric ratio was about
1.4. The stoichiometry, during this period of the run, appeared to
gradually increase, from an initial average ratio of about 1. 3 (from
April 3 to April 6) to a final average of about 1. 5. The percent oxida-
tion remained relatively steady during this period (at about 30 percent)
During a brief period of high stoichiometric ratio (about 1. 5) and inlet
liquor pH (about 6. 1) at about 110 hours in Figure 6-3, the SO2 removal
increased to about 80 percent. This substantiates the predicted high-
pH removal.
6-15
-------
as well as the inlet liquor pH (at about 5. 8). The increase in stoichio-
metry, for the same SO removal, could again indicate some "degrada-
LJ
tion" within the system.
6. 2 MATERIAL BALANCES
As mentioned previously (see Section 5. 1), during open-loop testing,
good material balances for calcium and sulfur could be obtained only
with the TCA system. The poor material balances for the venturi and
Hydro-Filter systems were attributable to solids build-ups (or deple-
tions) in the clarifiers, which could not be excluded from the material
•t*
ff"
balance enclosures. During the five-week boiler outage, the venturi
and Hydro-Filter flow configurations were modified to ones similar to
that of the TCA system, (see Figures 7-4, 7-5, and 7-6). It was expected,
therefore, that good material balances for calcium and sulfur would be
obtained on all three scrubber systems, based on the measured flow rate
and solids compositions of the bleed streams to the solids separation
area, the measured limestone addition rates, and the SO_ removals.
c*
6.2.1 Venturi Run No. 501-1A
Table 6-3 gives the results of a material balance for calcium and sulfur
for venturi Run 501-1 A, during a continuous 142 hour operating period
from April 14 to April 19, 1973 (see Figure 6-1).
For the TCA, the main slurry stream circulated between the hold tank
and scrubber, with a "bleed stream" from the main slurry stream
routed to the solids separation area (see Figure 2-5).
6-16
-------
o
in
O
ro
i
rt
H
D
H
^
W
>
tf
O
W
U
FQ
w
H
a A
* rS « G
O o -r; i
« W Q ja
i E
a ^ ^ "
u ° ,?
K r-t
S°
W
w n
t! N u C
388,1
OOOOOOOO
O(M(MOO(MfM(M
OOOOOOOO
OOOOOOOO
rocooooocooo
OOOOOOOO
43 £ £
•si
M X " ^
SO rt rt
ra U U
6-17
-------
The results of the balance showed that the measured sulfur discharged
(4. 38 Ib-moles/hr) is 3. 1 percent less than the measured SO absorbed
LJ
(4. 52 Ib-moles/hr) and that the measured calcium added (6. 10 lb-moles/
hr) is 6.0 percent less than the measured calcium discharged (6.49 Ib-
moles/hr). Both closures are satisfactory, in spite of some difficulties
experienced in measuring the limestone feed rate during the initial re-
O;
liability tests.
The ionic balances for the solids analyses, from which the calcium and
sulfur discharge rates were calculated, averaged less than +3 percent
(more cations than anions) for the bleed stream shown in Table 6-3.
Note that for both sulfur and calcium the measured inlet and outlet rates
do not necessarily balance during each individual computational period
in Table 6-3. This is due to the unsteady conditions which prevail (e. g. ,
changing percent solids) and the resultant accumulation (or depletion) of
the species in the system. However, over a long period of time (e. g. ,
~150 hours) the accumulation term becomes negligible as compared to
the total input or output for the entire computational period.
The average stoichiometric ratio (see Table 6-3) of 1.48 moles Ca/mole
SO~ absorbed, based on solids analysis, is probably more accurate than
the value of 1. 35 based on the measured limestone addition rate and SO
tii
absorption, because of uncertainties in the limestone slurry feed rate.
This measurement problem will be alleviated after May 4, when replace-
ment flowmeter elements (for small flow ranges) are installed in all
three limestone feed system magnetic flowmeters.
6-18
-------
6.2.2 TCA Run No. 501-2A
Table 6-4 gives the results of material balance calculations for TCA Run
501-2A, covering a period of 150 hours of uninterrupted operation from
March 30 to April 6, 1973 (see Figure 6-2).
The results of the balance showed that the sulfur discharged (4. 34 Ib-
moles/hr) is 7 percent less than the SO absorbed (4. 67 Ib-moles/hr),
while the calcium added (4.45 Ib-moles/hr) is 11 percent less than that
discharged (4. 99 Ib-moles/hr). The closures are considered to be quite
acceptable.
In Table 6-4, the sulfur input in each individual computational period is
generally greater than the sulfur output, and the reverse is true for cal-
cium. The ionic imbalances for the solids analyses during these periods
were mostly positive (more cations than anions) and averaged about
+ 5 percent. In other words, the reported sulfur content in the bleed
solids might have been too low, or the calcium content too high, or both.
If this factor is taken into account, either or both of the sulfur and cal-
cium balances would be better than those reported.
Again, due to uncertainties in limestone addition measurement, the ave-
rage stoichiometric ratio of 1. 15 (moles Ca/mole SO- absorbed), based
on the solids analysis, is a more reliable number than the value of 0. 95,
based on the measured limestone addition rate and SO absorption.
6-19
-------
.3.5.8 .S
S'is 6
IS
^3 *"
8
*,$$*
O o » ft
M W « ?
81'
n — -T H-I «
«E^SS
0* -^
v
Is.
« "2
II
•s-s
a n
(d A
ffl ffl
00 tf) 00 00 CM sO I
(M (M N N « «^
00 00 CO 00 CO 00 00
O O O O O O O
o o o o o o o
(M O O O O O O
O O O O O O O
O O O O O O O
CO ^ ^ •*•*•* U-i
O M M CM M N O
O O O O O O O
O O O O O O O
CO 00 O O O O O
N O O O O O O
6-20
-------
6.2.3 Hydro-Filter Run No. 501-3A
Table 6-5 gives the results of material balance calculations for Hydro-
Filter Run No. 501-3A, covering a period of 150 operating hours from
March 16 to March 22, 1973 (see Figure 6-3).
For sulfur, the average discharge rate (4. 11 Ib-moles/hr) is only 3 per-
cent less than the SO^ absorption rate (4.24 Ib-moles/hr). For cal-
cium, the rate of addition (4.49 Ib-moles/hr) is 13 percent less than
the discharge rate (5. 16 Ib-moles/hr). The balance is satisfactory, con-
sidering the uncertainties in the limestone slurry addition rate during
the period.
The ionic imbalances for the solids analyses, from which the calcium
and sulfur discharge rates were calculated, averaged less than +2 per-
cent (more cations than anions) for the bleed stream shown in Table 6-5.
Again, the average stoichiometric ratio of 1.26 (moles Ca/mole SO ab-
sorbed), based on solids analysis, in Table 6-5, is probably more ac-
curate than the value of 1. 06 based on the measured limestone addition
rate and SO_ absorption.
£~t
6-21
-------
in
sD
i
i—t
O
Balanc
ilcium
O
«
u
c
«
fl
IH
3
£»
t)
V
e
0
imest
a
V
h
$
"O
u
a
•
|
:
.5 5 ^ I
rt 3 o |
*" Q J>
w
.S« 'S 1
3Jb! j
c « * *
o"3 1 1
T3 o>
.-S |
J3 JD
5=" *
u
|JS E
CS »
aflS'*
5l^i
.S.S'V
SfllS?
•S ieS
3 -S J t
0
» »• 6
caa
i
o1^
"1
38 § a
IK
3ES^S
hll
ff
i
Q
_^^^^0_00~r-0--0COf-*^^-
OOOn.M-^^OCO»m-f-ON-^0
SSS5SSSKS5SKSSSSSS5
vOC-^>t--o*o^ooonNJJ-oa-M-'n
nnn^^rommrommmmmnrommn
•fr
0 , _--- 00000-
NNNNNNNNNNNNNNNNNNN
«««««««- — —««««"«
^^o^oCo;dddddd-;-:-odoo'd
SSSSSSSSSSS^S^SSS
ooooooooooooooooooo
OOOOOOOOOOOOOOOOOOO
ooooooooi/>oinoomiftininmo
•^'OMtnfOMfMt-fOCOfO'OMM-H-Hm'***!
OOOOOOOOOOOOOOOOOOO
(MNNrsl(vJN(Mr4N(M(MN(M(MNNrOr
O
in
3
H
I "-
fe •* r-
•3 I
sS
U (, TJ
S §1
4) 3 «
5—I O^ vO
rt •« ^
Iff
•° (Q
6-22
-------
Section 7
OPERABILITY AND RELIABILITY OF THE TEST FACILITY
In this section, the operating experience of the test facility during the
open-loop, short-term factorial testing and the initial closed-loop lime-
stone reliability verification testing are discussed. Also discussed
are system modifications made, primarily, during the five-week boiler
outage in February and March, 1973, and results of a material
*
evaluation program. Scaling and plugging tendencies of the systems
are discussed, primarily, in Section 7. 5.
7. 1 CLOSED LIQUOR LOOP OPERATION
The original test facility design included slurry pumps with water seals
(Hydroseals) for bearing protection, water quench sprays for gas cool-
ing, water sprays for mist eliminator washing, a water wash for the
Koch tray, and a dilute limestone slurry feed (10 - 20 wt % limestone).
The water input under these conditions exceeded the makeup require-
ment for closed liquor loop operation. The systems operated, there-
fore, for about six months with partially open liquor loops during
limestone short-term factorial tests. This was not considered to be
In this report "scale" refers only to crystalline hard
solids, and "solids" or "soft-solids" refer to mud-like
slurry solids.
Closed versus open liquor loop testing is discussed in
Section 3. 2. 3.
7-1
-------
a serious problem for, at a specified scrubber inlet liquor pH, SO
C*
removal is not significantly affected by liquor composition. However,
little information was gained about the effect of scaling potential on
reliability during this period.
The absorbent feed systems were changed in November, 1972, to pro-
vide slurry feeds with up to 60 wt % limestone concentration. During
the five-week boiler outage in February and March, 1973, the Hydro-
seal slurry pumps were converted to a Centriseal type (mechanical
seal supplemented with air purge); quench spray systems using circu-
lating slurry were provided for the TCA and Hydro-Filter scrubbers;
and, the Koch tray wash system on the TCA scrubber and the mist
eliminator wash systems on the spray tower and the Hydro-Filter
scrubber were converted to use clarified liquor plus raw water make-
up. Required revisions to bleed control, flow measurements and con-
trol instrumentation were also made during this period.
As a result of the modifications to the test facility, closed liquor loop
operation (i. e. , raw water input to the system equals the water nor-
mally discharged with the humidified gas and waste sludge) has been
attainable during the limestone reliability verification tests. Water
balances for the three scrubber systems, using clarifiers for solids
separation, are presented in Appendix D. As seen in Tables D-l,
D-2 and D-3, the weight percent solids in the TCA system clarifier
underflow approaches the •weight percent solids that would reside with
typical settled pond sludge (~40 wt % solids). The settled slurry dis-
charged from the smaller venturi and Hydro-Filter clarifiers, however,
7-2
-------
is typically higher in moisture content. A reduction in moisture con-
tent in the discharged solids from these systems can be achieved either
by adding coagulent to the process liquor or by routing the clarifier
bottoms to the centrifuge or filter.
7. 2 EQUIPMENT OPERATING EXPERIENCE
7. 2. 1 Demisters
The specifications for the demisters tested during the limestone
factorial tests and the three initial reliability verification tests are
given in Table 7-1.
The original facility design provided only for top (downstream) wash
sprays for the spray tower, TCA and Hydro-Filter demisters. The
earliest open-loop limestone short-term factorial tests were performed
with no demister washing, and demister pluggage developed within two
or three days of operation. Subsequent open-loop factorial tests em-
ployed intermittent demister top washing at 3-5 gpm/ft for 1/2 —
1 minute per hour. Although the open-loop limestone factorial test
period was relatively short and the test conditions varied considerably
with time, the following observations can be made:
Top washing of the demisters with raw water has sub-
stantially improved demister performance for all three
scrubbers over dry operation. Tests could only be
performed with periodic demister cleaning; therefore,
this type of operation is not acceptable for a full-scale
facility.
The quantity of dissolved solids within the process liquor is
proportional to the percent solids discharged from the system.
7-3
-------
The TCA scrubber has had the most serious demister
pluggage problems, the Hydro-Filter scrubber has had
less serious problems, and the spray tower has had the
least pluggage. Many of the venturi system tests have
not utilized the spray tower, which substantially re-
duces the entrained liquor that impacts upon the de-
mister. The reasons for the severity of the TCA plug-
gage are attributed to the high gas velocity through
the scrubber, and the relative tortuosity of the six-
pass chevron demister compared to the three pass de-
mister configurations of the Hydro-Filter and spray
tower scrubbers.
Table 7-1
TEST FACILITY DEMISTER SPECIFICATIONS
Material of Construction
Design
Number of Vanes (Passes)
Total Depth of Demister
Center-to-Center Distance
Between Vanes
Angle Between Vanes
Spray Tower
Stainless Steel
Chevron, open
3
7-11/16-in.
3-9/16-in.
100°
TCA
Stainless Steel
Chevron, closed
6
14-in.
1-1/8-in.
120°
Hydro-Filter
Stainless Steel
Chevron, closed
3
7-1/8-in.
3-in.
80°
Open-vanes not joined, closed - vanes joined.
In order to remedy the demister solids accumulation problems, the
following modifications were made to the systems:
(1) In November, 1972, a Koch Flexitray wash tray was
installed in the TCA scrubber between the inlet liquor
7-4
-------
spray header and the chevron demister, and a
steam sparger was provided for washing (cleaning)
the underside of the wash tray. At first, irrigation
was obtained with raw water. A subsequent modifi-
cation in February, 1973, allowed for irrigation
with process liquor, diluted with the available raw
water makeup.
(2) During the boiler outage in early 1973, the spray
tower and Hydro-Filter demister systems were
modified to allow for washing from both the up-
stream (underside) and downstream directions with
process liquor, diluted with the available raw water
makeup.
The Koch Flexitray wash tray has been successful, to date, in signif-
icantly reducing the solids accumulation on the TCA demister blades.
However, heavy solids buildup occurred below the Koch tray with
intermittent steam sparging for 1 minute per eight hour shift. Sub-
sequent to the five-week boiler outage, the steam sparging was in-
creased to 1 minute per hour, which has substantially reduced the
solids accumulation below the tray.
During the three initial limestone reliability verification tests, the
liquor wash to the demisters (and to the Koch tray) has varied from a
ratio of about one part fresh water and three parts clarified liquor to
o-
*T*
half and half mixtures. The undersides of the Hydro-Filter and
venturi demisters have been washed intermittently, on a cycle that
has averaged about one minute "on" and three minutes "off, " at an av-
erage rate of about 1 gpm/
buildup has been detected.
2
erage rate of about 1 gpm/ft and no significant scaling or solids
The mixture ratio is dependent upon the percent solids
discharged, the percent solids recirculated and the gas
flow rate.
7-5
-------
Based on the results to date and the experiences at other facilities, it
appears that the following design provisions may alleviate demister
plugging problems:
Washing the demister from both the upstream and
downstream directions with a mixture of clarified
liquor and the required makeup water.
Utilizing a wash-tray between the uppermost stage
and the demister.
Using a relatively open demister (low number of
stages and large blade angle).
Maximizing the distance between the uppermost
scrubber stage and the demister.
7. 2. 2 Reheaters
Flue gas is reheated after evolving from the scrubber to prevent con-
densation and corrosion in the exhaust system, to facilitate isokinetic
and analytical sampling, to protect the induced draft fans from solid
deposits and droplet erosion, and to increase plume bouyancy. The
reheaters employed are fuel oil fired units with a separate combustion
air supply and with combustion occurring in the flue gas stream. The
reheaters had been difficult to start and operate during the short-term
factorial testing and combustion had been incomplete, which led to a
visible plume containing significant quantities of soot. This made it
difficult to interpret outlet particulate data and affected gas sampling
by the DuPont SC>2 photometric analyzers. The difficulty appeared to
result from quenching of the flame by the cold (128°F) flue gas before
complete combustion could occur, and from operating with the same
fuel atomizing nozzles over a wide range of flow rates.
7-6
-------
The reheater systems were modified during the scheduled boiler outage
in early 1973. Internal stainless steel sleeves (40 inches in diameter by
4 feet high) were installed to provide approximately 50 cubic feet of
isolated combustion zone for each reheater. Also, the turbulent mix-
ing type nozzles supplied originally were replaced with mechanical
atomizing nozzles. These new nozzles are designed for a narrow
range of oil flow rate and have to be changed when the reheat require-
ments change. Nozzle replacement, however, is a simple job.
To date, the above modifications appear to have been effective. Essen-
tially no soot is visible in the stack gas, and the outlet particulate sam-
ples have shown no appreciable quantities of carbon from the reheaters.
Therefore, plans for installation of an external combustion system on
one of the reheaters have been deferred.
7.2.3 Nozzles
Nozzle reliability at the test facility has been greatly reduced by the
frequent plugging of spray nozzles with foreign material (plastic
spheres, marbles, debris, etc.), and the erosion of some spray
nozzles by the abrasive solids in the circulating slurries. It has be-
come apparent that nozzle pluggage could be reduced substantially by
placing screens over open vessels in the scrubber systems and/or
within the circulating slurry lines.
Spray Tower. Limestone factorial testing in the spray tower started
with the use of spiral tip, 316 SS, full cone, Bete No. ST-24 FCN
nozzles (capacity: 12 gpm @ 12 psig) manufactured by Bete Fog
Nozzles, Inc. Because of frequent plugging with slurry and/or debris,
7-7
-------
these nozzles were replaced in September, 1972, with Bete No.
ST-32 FCN nozzles (capacity: 21 gpm @ 10 psig). Plugging of the
larger Bete nozzles became less frequent. Neither type of nozzle
showed any significant sign of erosion.
To allow for increased liquor flow to the four-header spray tower,
Bete No. ST-48 FCN stainless steel nozzles (capacity: 47 gpm @
10 psig) were installed during the February, 1973, shutdown. During
the first limestone reliability verification test (No. 501-1A), five of
the 28 nozzles became totally plugged with debris and four nozzles
became partially plugged. Although erosion of these stainless steel
nozzles has not been observed to date, they will be replaced with
identical stellite-tipped ST-48 FCN nozzles in the near future.
TCA. The large Spraco 1969, full cone, 316 SS, open-type slurry
feed nozzles have performed satisfactorily and without significant
erosion since the original startup of the unit. Occasional partial
pluggage by large debris did not necessitate premature termination
of any test run.
The four nozzles (Type 7LB, Carpenter 20, manufactured by Spray
Engineering Company) used for gas humidification and located in the
flue gas duct close to the TCA scrubber entrance, became severely
eroded after being in circulating slurry service for approximately
1500 hours during the short-term limestone factorial testing. These
nozzles were replaced by a Ventri-Rod presaturator, which had
operated successfully at the TVA Colbert pilot plant. The Ventri- •
Rod did not perform satisfactorily and there was continual rapid
7-8
-------
buildup of solids both on and downstream of the rods. The Ventri-Rod
presaturator was replaced by four spiral tip, 316 SS, Bete No. ST-Z4
FCN nozzles, which displayed plugging tendencies and were replaced,
in turn, with four Bete No. ST-32 FCN nozzles, which performed with-
out plugging during the final phase (4-1/2 days) of the first limestone
reliability verification test (Run No. 501-2A).
Hydro-Filter. The 22 original slurry feed spray nozzles lined with
Solathane 291 and equipped with internal Adiprine LD 315 swirl vanes
failed in various ways during short-term factorial testing. The swirl
vanes in all 22 nozzles eroded, the liners of four bottom nozzles
collapsed, and two bottom nozzles disintegrated. The nozzles fre-
quently became plugged with slurry and debris.
The original slurry feed nozzles were replaced (during the February,
1973, shutdown) with improved nozzles supplied by Combustion Engi-
neering (stronger, Adiprine LD 3056 lining with improved bonding
using Thixon 1244 between the liner and the body of the nozzle and a
locking groove to hold the vanes in place). The diffusion vanes of 13
(of the 16) bottom spray nozzles failed during the initial limestone reli-
ability verification test run (No. 501-3A), after 764 hours of operation.
The Schutte and Koerting No. 661-S saturation spray nozzles operated
satisfactorily during the short-term factorial limestone test period
with no sign of corrosion or erosion. The only problem encountered
was the buildup of solids at the wet/dry interface in the vertical sec-
tion of the duct near the scrubber entrance. To alleviate this interface
problem, a cooling spray system using four Bete No. ST-20 FCN
nozzles (capacity: 8 gpm @ 10 psig) was installed in February, 1973,
7-9
-------
during the shutdown. During the first reliability verification test run,
the nozzles plugged several times with slurry solids and debris. In
addition, two nozzles eroded seriously.
7.2.4 Waste Solids Handling
The test facility is equipped to study alternate methods of waste solids
dewatering and disposal where separate clarifiers are provided for
each scrubber, and a rotary drum vacuum filter, a horizontal solid
bowl centrifuge and a slurry settling pond are common to the three
systems. Solids separation can be achieved with any combination of
clarifier, filter, centrifuge and pond.
Clarifiers. The clarifiers are conventional solids contact units with
a heavy duty rake and scraper mechanism supported from a bridge.
The vessels are flake-glass lined with a stainless steel rotating
mechanism. The venturi and Hydro-Filter systems have 20-foot
diameter units while the TCA clarifier is 30 feet in diameter.
The performance of the clarifiers during the short-term factorial
test period was unsatisfactory. Solids carryover in the overflow of
the two smaller units was a problem and the solids concentration in
the underflow streams of all three units could not be controlled.
These problems could be attributed to the following:
• Excessive bleed (clarifier feed) rates (up to
100 gpm).
• Erratic bleed rate control, due to oversized
piping and inadequate flow control systems
7-10
-------
(the rubber lining of the pinch valves seriously
eroded during the factorial testing).
• The solids loading for the given area of the two
smaller units was excessive.
• The underflow rates could not be reduced and
controlled below approximately 20 gpm, and the
height of the liquid in the clarifiers caused
siphoning through the pumps.
To improve clarifier operation, the following modifications were made
during the boiler outage in February, 1973 (some of these modifica-
tions were part of an overall revision for improved density and level
control of the scrubber systems):
• Separate 100 gpm capacity, rubber lined, variable
speed pumps were installed in the venturi and TCA
scrubber systems for slurry bleed handling.
• Three parallel Clarkson pinch valves of different
sizes were provided to control the bleed from the
Hydro-Filter system.
• Magnetic flowmeters were installed in the bleed
lines to each clarifier.
• The six and eight inch clarifier feed lines were
replaced with two inch rubber hoses to maintain
required velocities at the reduced slurry bleed
rates required for closed-loop operation.
• High elevation discharge piping was provided in
the clarifier underflow systems to eliminate
siphoning through the pumps.
• The suction piping of the clarifier underflow pump
was changed from one and a half to three inches
to minimize line plugging.
• Low liquid level alarms were installed in each of
the three clarifiers.
7-11
-------
The above modifications have considerably improved the performance
of the clarifiers. The concentration of solids in the underflow of the
large TCA unit approaches the expected final settled density of the
sludge (approximately 40 percent by weight). However, the poor
settling characteristics of certain solids components, particularly
calcium sulfite and fine flyash, and the high solids loading in the bleed
continued to result in periodic solids carryover in the overflow of the
20 foot diameter units (venturi and Hydro-Filter systems). In addition
the adjustable V-notch metal plate at the top of the clarifiers does not
provide a tight seal, resulting in turbid overflow at high solids/liquid
interface level, particularly in the Hydro-Filter system.
Filter. Initial tests with the rotary vacuum filter during the February,
1973, boiler outage were not successful. The filter cake was thixo-
tropic, and, although it appeared dry and firm under vacuum, the cake
became fluid as the vacuum was reduced and the internal water was
released. The wet, sticky cake would not separate from the filter
cloth. Dewatering was restricted by formation of cracks in the cake
which prevented operation at maximum vacuum. Preliminary tests
indicate that approximately 55 percent solids in the filter cake can
be ultimately obtained.
Centrifuge. Short-term, exploratory tests were carried out in late
April, 1973, to establish the optimum solids dewatering capability of
the centrifuge. The test results are presented in Table 7-2.
It appears that the centrifuge is effective in reducing the moisture
content well below the level attained by settling. The centrate clarity
7-12
-------
(M
t
•9
H
w
H
W
W
H
W
O
t)
H
U
Solids
in Centrate,
wt. %
M CU _p
-"-1 ni
.2 u j
,g U «s
O W 4J
• *-(
T3 C) C
0> -)j C
— 1 •— '
d o o d o
t^ so oo o •— '
i i i i i
co "^ GX vD r^
•^ -^ CO oO 00
X "M- ~*r "*r
(\J ^^1 ^N [^- ^J^
oo ro ro oo .-I
i i i i i
m xD 0s o^ o
i — 1 r—t t — 1 i— i , — I
oo oo ro ro in
00, OO OO CO CO
i i i i i
r— 1 O O 0s ' — I
i — 1 i — 1 i — 1 i — 1
cu CD CD cu ^ ;_, ;~-
ia 3 3 S 3 S a ? -° -2 "S
j^C ^C ^C ^C M1^10
ni^ rtS rt° rtS ^'^d
O O O O r-Q i — i J~^
I-U HH HH HH ^
W
O O O O O
o o o o o
o o o m m
oo oo oo oo oo
— < CM co -^ in
0)
-j->
rt
^
i — t
n)
a>
in
ft
E
0
(X
JH
O
U
0)
o
o
o
,£!
rt
0)
tn
n!
0
o
0)
u
a
o
U
7-13
-------
is satisfactory, containing 0. 6 wt % solids or less. The break point, i. e. ,
the deterioration of solids recovery and centrate clarity, was found to be
between 32 to 35 gpm feed rate.
Pond. A three-section settling pond was constructed in an area previous-
ly used as an ash storage pond. The dikes were made from flyash and the
entire pond was covered with local clay (about 6 inches thick). The small
starter pond was used during the early limestone factorial tests and be-
came filled with waste material and taken out of service by the end of
November, 1972. For the remainder of the program, slurry will be dis-
charged into the large settling area and the required supernate will be re-
turned through a smaller "polishing" pond. The slurry to the pond can
come directly from the scrubber system, from the thickener underflow,
or from the filter and centrifuge as reslurried cake.
Laboratory tests are being made on various clays to determine their po-
tential for lining sludge ponds. In addition, a cost study is underway to
determine the feasibility of building three test ponds for sludge disposal
studies.
7. 2. 5 Fans
Initially, considerable difficulty was experienced with the induced draft
fans. Some of the problems included high fan vibration, fan motor fail-
ure, fan damper control failure and fan blade deformation. All of the
problems, except for blade deformation, necessitated repeated shutdowns
of the affected scrubber systems.
The unacceptable high vibration problem of all three fans was greatly re-
duced in June, 1972, by insulating the fan housing, adding additional
7-14
-------
bracing to the outboard pedestals, and welding balance weights on the fan
shrouds. However, occasional high fan vibration continued to hinder
scrubber operation, particularly on the venturi system, and required
either addition of shims to the bearings or replacement of the bearings.
The motors of the venturi and TCA fans had to be returned once to the
supplier for repair and correction of serious manufacturing problems.
Stable flue gas flow control was achieved by increasing the "fully open"
to "fully closed" fan damper response time from 10 to 100 seconds
with new actuators. Three scrubber system shutdowns were caused
by inoperable linkage and a broken shear pin.
Distortions of several blades (arc shapes as contrasted to the original
straight line configuration) of the venturi and TCA fans were observed in
March, 1973. The maximum deformation was 0. 55 inch on blade No. 5
of the Hydro-Filter fan. The manufacturer indicated that the deforma-
tion was probably caused by stress relieving during fan operation and
that the warping of the blades did not interfere with efficient, safe opera-
tion. No significant continuing deformation of the blades has been ob-
served to date.
7. 3 MATERIALS EVALUATION
7. 3. 1 System Components
A thorough inspection of all system components was conducted during
the extended February and March, 1973, boiler outage. Each of the
three scrubber systems had been operated for about 1800 hours during
the factorial limestone scrubbing tests (see Figure 3-1).
7-15
-------
Localized deposits of loose flyash accumulated in the mild steel gas
ducts between the boiler and scrubber structure. The surfaces were
coated with a thin iron oxide scale and moderate pitting had occurred at
the uninsulated connections. The flanges and access doors have been
insulated.
The rubber lining in the scrubbers was in excellent condition; no erosion
or deterioration was noted. The rubber linings in pumps, piping and
process water tanks were also in excellent condition. Slight wear was
noted on some of the rubber-coated agitator blades.
Several hairline cracks were noted in the Flakeline lining in the effluent
hold tanks and clarifier tanks. The cracks did not appear to penetrate the
entire thickness of the lining, they were most prevalent at the junctions
between the baffles and tank walls.
The most severe corrosion was found on Type 316 stainless steel surfaces,
particularly on the mist eliminator blades in the TCA system. In gener-
al, the corrosion was in the form of pitting with some pits as large 1/16
inch diameter and 30 to 35 mils deep.
Significant erosion was noted on the pump sleeves, at the intersections of
the wire of support grids in the TCA scrubber, and on the impeller and
casing of the 316 SS Gould limestone slurry pump.
Weight loss of the mobile bed packing material has been detected. The
polypropylene and polyethylene spheres in the TCA scrubber have worn
noticeably and some are so thin that they have collapsed. Random sam-
ples of these collapsed spheres showed about 60 percent weight loss. Most
7-16
-------
of the spheres were still intact but had an average weight loss of 20 per-
cent. The glass marbles in the Hydro-Filter scrubber have lost about 6
percent of their initial weight.
The performance of the various spray nozzles has been discussed in
Section 7. 2. 3.
7. 3. 2 Test Coupons
Test coupons of several different materials of construction, together
with stressed and welded specimens, were exposed for periods of 1680
hours or longer to various slurry and gas environments. The corrosion
rates observed are presented in Table 7-3.
Corrosion of Hastelloy C-276 was negligible to 5 mils per year. This
alloy showed no evidence of localized attack in any test location. Next
in resistance to corrosion were Inconel 625, Incoloy 825, Carpenter
20Cb-3, and Type 316L stainless steel alloys. The corrosion rates
for each material ranged from negligible to 5, 7, 14 and 15 mils per
year, respectively. These alloys had few minute corrosion pits and/
or crevice corrosion. Type 316L, the fifth alloy in corrosion resistance,
is the least expensive of this group of materials.
Three nonferrous alloys, Cupro-Nickel 70-30, Monel 400, and Hastelloy
B, each had minimum corrosion rates of less than 1 mil. Maximum cor-
rosion rates were 49, 57 and 100 mils per year, respectively. Only
one or two specimens pitted. In three tests of Monel and in one test of
Cupro-Nickel 70-30, the welds were inferior to the parent metal.
7-17
-------
i
Is-
CO
H
H
H
CO
W
H
u
•8
to
0)
&
H
V
0
"8
Number
7
O,
1
n
^
«l
**
<;
—
£
J
I
^
3
c
Crevice A
i
s
c
is
3
m
J a
11
'A
H
|
0
4
5
I
TJ
V
=r^ 3 2 ~ — 2
i it • 2, » « « ' ' ' JJ. ' JJ, ' ' ' oj
2 i£ ?£ ^
O
-1
, , -< , N , -H ,^<^^nr^-M^
** *» ** ** +*
iS SiS -----M-i-g
1 1 1 1 1 1 N , 1
is is is is i is N ' is
i« iMto — — NO«^.O_ — _cri\jin
o o
OOOOQOOOO
^•inor-oMrt'^PJPJrJinrt-'^
in in I- -i -< •* in -i / ,' V \ V V ' V
| | | | | - - X | | | | | - 7 X 7
o
<
^-i ft)
111
t
T) » „
oj C «
I 1 ''
... „ a
OJ (0 ^_
cj "2 o
•2 S *«
rt "* 4.
ce.
of each met
penetration
« « ° fl
ill!
M w Q C
,fl u T:
0s U") "-•
• 2 •
rJ CM tn
O
O
°§
O
lis
» -3 o
!3§
CQ h a
Ul
0
en
n)
CU
c
C -
n
T) I
(0 1
!
« i
S ]
qj 1
** i
S^
2 J
>.J^1
r- nl i
C ., ^
? SL
g s L
S j» ;
p, rt ^
"3 k "
^ *O 0
y c a
ocalized attc
of Bondstra
-i CD i
QJ 0) fl
* M *Q. '
> 6 *~
CO CO r^
V H-,
i i i
i i i
S
«
in ^
in ^H
t^ o qj
en . C
r-H rt 0)
- * ft
^ 3 £•
^ 4-1 0
S n) a)
m z z
CO
(U
r£)
3
'
'
J
'
)
•
}
1
)
t
'
*
}
.
4 to
C
; |J3
) nl
J u
0
M>
in
PJ
1-1
-------
The corrosion rates of Type 446 stainless steel, E-Brite 26-1, Incoloy
800, USS 18-18-2, and Type 304 stainless steel, ranged from negligible
to values which indicated that the alloy specimen was completely des-
troyed at one or more test locations. The values for the specimen
failures ranged from greater than 140 mils per year for Type 446 to
greater than 200 mils for both USS 18-18-2 and Type 304 stainless steels.
These five alloys were highly susceptible to localized corrosion.
Another group of alloys, Type 410 stainless steel, 3003 aluminum, A-283
mild steel, and Cor-Ten B, had minimum corrosion rates of less
than 1 mil per year and maximum corrosion rates of greater than 250
mils for Type 410 to greater than 1400 mils for A-283 and Cor-Ten B.
Pitting and crevice corrison occurred on the four alloys.
In general, the stressed specimens (five alloys only) were not corroded
at rates higher than their counterpart disk-type specimens.
Specimens of Bondstrand 4000, Flakeline 200, and Transite materials
were tested at 21 locations. Bondstrand 4000 showed good corrosion
resistance in 12 tests and poor resistance in nine tests. Only six speci-
mens each of the following materials were tested: Qua-Corr plastic,
butyl natural rubber and neoprene rubber. The results were: five good
specimens and one poor specimen for Qua-Corr plastic, and six good
specimens for each type rubber.
With few exceptions, mainly in the TCA system, the greatest loss of
weight from metal specimens occurred in areas where the velocity of the
unscrubbed, partially humidified, flue gas was comparatively high. Im-
pingement on the specimens of the slurry caused erosion and corrosion.
Pitting and crevice corrosion were not important factors where erosion
7-19
-------
and corrosion kept the specimens clean. In other areas of the three scrub-
ber systems where solids accumulated, the frequency of localized corro-
sion was high. However, each of the 17 alloys tested showed good cor-
rosion resistance at one or more test locations in each scrubber system.
Venturi Scrubber. The venturi scrubber coupons were more severely
affected by both corrosion and erosion than coupons in any other section
of the three systems. Nine alloy and four non-metallic specimens failed.
The following alloys showed the best resistance to corrosion: Hastelloy
C-276, Inconel 625, Incoloy 825, Carpenter 20Cb-3, and 316 L SS.
The butyl and natural rubbers and neoprene remained in good condition,
while Teflon spacers on the spools and the Qua-Corr specimens were
damaged by erosion.
Scrubber Towers. Corrosion rates were less than one mil per year for
several alloys in both the TCA and Hydro-Filter. These alloys include
Carpenter 20Cb-3, Hastelloy C-276, Incoloy 825, Inconel 625, and Type
316L, stainless steel. Minute pitting of Type 316L, stainless steel oc-
curred on the test coupon located between the Koch tray and demister
in the TCA. Pitting and crevice corrosion were common on the other
alloys tested. Pits to a depth of 25 mils were noted on Type 304L
stainless steel. Test specimens were not installed in the venturi
after-absorber.
In the TCA scrubber, A-283 mild steel was corroded at rates of 23
to 250 mils per year, Cor-Ten B at 13 to 268 mils per year, and
3003 aluminum at 4 to 26 mils per year. The highest values were found
7-20
-------
above the Koch tray, and the lowest values were found in the middle bed
area. This might be due to the protection provided by solids accumula-
tion on the specimens at the lower elevations.
In the Hydro-Filter scrubber, the corrosion rates for mild steel and
Cor-Ten B were 14 to 37 and 13 to 50 mils per year, respectively.
The higher corrosion rate values were found below the bed.
Butyl, natural and neoprene rubbers and Qua-Corr plastic showed good
corrosion resistance in both scrubbers. The corrosion resistance of
the other non-metallic materials (Bondstrand, Flakeline and Transite)
varied from "fair" to "poor. "
Scrubber Outlet Ducts. In general, the corrosion rates of coupons were
the least in theTCA outlet duct and greatest in the venturi outlet duct. The
corrosion rates were no greater than 5 mils per year in all tests, ex-
cept for Cor-Ten B (18 mils per year) and mild steel (16 mils per
year) in the venturi duct.
The corrosion resistance of Flakeline material was poor in the venturi
outlet duct. Bondstrand material had poor corrosion resistance in all
three ducts (apparently the temperature exceeded the service limit for
this material).
Effluent Hold Tanks. In general, coupon corrosion rates in all three
tanks were negligible for alloys, but pitting and crevice corrosion were
common. Aluminum 3003, Cor-Ten B, and mild steel had corrosion
rates of 20, 70 and 210 mils per year, respectively.
7-21
-------
Recirculation Tanks. Corrosion rates were comparable to those in the
effluent hold tanks.
Clarifiers. Corrosion rates in clarifiers were negligible. The highest
corrosion rates rangedfrom five to nine mils per year for Cor-Ten B
and mild steel.
7. 4 INSTRUMENT OPERATING EXPERIENCE
7. 4. 1 Sulfur Dioxide Analyzers
Essentially trouble-free operation was experienced with the DuPont
Model 400 UV sulfur dioxide analyzers following the modification of the
sampling system and the replacement of interference filters in Novem-
ber, 1972. Initially, the sampling system was particularly vulnerable
to condensation, solid particulates, oil, soot, corrosion, or the com-
binations of these factors which led to leakage or plugging of the sam-
pling lines, plugging of the filters, or coating of the optical lens. All
of these effects caused erroneous sulfur dioxide analyzer readings.
To eliminate the problem areas, the sampling handling system was
modified as follows:
• All heat sinks and sharp bends in the sample lines were
eliminated. A new 3/8 inch diameter Dekeron sample
line was installed to replace the original 1/4 inch stain-
less steel line. Heat tracing was installed along the full
length of the sample line.
• Stainless steel shields furnished by DuPont were installed
around the probe filters. The original ceramic probe fil-
ters were replaced by probe filters made from 316 stain-
less steel and recently developed by DuPont.
7-22
-------
• An automatic zero and air blow-back system was in-
stalled on the SO? analyzers in the inlet gas ducts,
similar to those provided originally in the scrubbed
gas ducts.
• Stainless steel lines and fittings were replaced with
Dekeron or Teflon plastic wherever possible.
• Calibration methods were changed to use a stainless
steel wire mesh reference filter rather than bottled
standard reference gas.
One additional problem associated with all six analyzers was the deteriora-
tion of the interference filter in the optic section. All of these filters,
which filter out all except the desired light wave lengths, were replaced.
The failure and subsequent deterioration of the filter was attributed by
DuPont to the exposure of the analyzers to freezing conditions prior to
their installation. It was theorized by DuPont that the freezing caused
minute cracks which then deteriorated with time (see Appendix C).
7. 4. 2 Magnetic Flowmeters
Operating experience with the Foxboro magnetic flowmeters has generally
been good. The main problem has been in obtaining accurate flow mea-
surements at very low flow rates with meters designed to measure flow
over a wide range. To assure accuracy, Foxboro recommended a mini-
mum linear velocity of 3 ft/sec through the flow element. Periodic clean-
ing of the electrodes and calibration checks are also required for correct
flow measurements. Routine cleaning and maintenance of flowmeters
are made during extended shutdown periods.
7.4.3 Control Valves
Operating experience with control valves has generally been good when the
control valves were used within reasonable design flow ranges. However,
7-23
-------
when excessive throtling of the valves was required, the increased
velocity caused severe erosion in a short time. This has been observed
in both stainless steel plug valves and rubber pinch valves.
7.4.4 pH Meters
Operating experience with the Uniloc Model 1000 pH meters has thus far
been limited to in-line flow-type meters. No significant scaling of the
electrodes has been noted to date. However, frequent calibration checks
with buffer solution are required to maintain the desired meter accuracy.
Calibration checks are made twice a week on a routine basis (or more if
required). Because of the desirability to control pH to within ± 0. 1 pH unit,
future test program plans include evaluation of another type of pH meter.
7.4.5 Density Meters
Operating experience has been gained with both the Ohmart radiation-type
density meter and the bubble-type (differential pressure) density meter.
Both meters require further study and modification to achieve adequate
reliability in their respective control service.
7. 4. 6 Scan Data Acquisition System
Early in the test program, there was considerable difficulty in recover-
ing the scan data from the tapes recorded on-site. Changes •were made
to reduce the effect of industrial noise on the system. A special com-
puter program was also written to eliminate defective records. Since
the tape recorder was neither enclosed nor located within a pressurized
area, periodic cleaning was initiated on a weekly basis. Subsequent to
these changes, operation of the data acquisition system improved.
7-24
-------
7. 4. 7 Analytical Data Acquisition System
Operation of the x-ray unit has been satisfactory. Since both the x-ray
unit and the computer are enclosed in a pressurized air conditioned
room, the problems of recovering the analytical data from the magnetic
tapes have been minimal. Some minor problems •were initally experienced
with the interface between the x-ray unit and the computer and with the
peripheral hardware equipment. These problems have been solved.
7. 5 SYSTEM MODIFICATIONS
Operating experience during the sodium carbonate and limestone short-
term factorial testing revealed the need for extensive system modifica-
tions to facilitate:
• Closed liquor loop operation
• Improved operability of the systems
• Long-term system reliability
• Additional operating flexibility
These modifications were made, generally, during the five week shut-
down of Unit No. 10 in February and March, 1973. Since the purpose
of the modifications might overlap to some extent, no attempt has been
made to list the individual changes in the four categories shown above.
The major modifications as of the end of the first closed-loop test run
on each of the three scrubbers are listed below:
7-25
-------
Installation of four Moyno pumps and revision of the
limestone slurry makeup piping for the addition of
60 wt % slurry to the effluent hold tanks only (all
three systems).
Conversion of nine slurry circulating pumps from
water seals (Hydroseals) to mechanical seals (Centri-
seals) supplemented with air purge (all three sys-
tems - a total of nine pumps).
Provision to use pond return liquor to sluice fresh
limestone.
Provision to use clarified liquor for demister flush
(all three systems).
Rerouting of slurry discharges from conductivity
and pH cells from sewers to circulating slurry loop
(all three systems).
Modification of all three clarifier underflow pumps
(sheaves to increase pump speed) to reduce slurry
discharge rates to disposal (all three systems).
Modification to the clarifier underflow piping to
the re slurry tank to eliminate siphoning at low flow
rates (all three systems).
Installation of stainless steel sleeves in the reheat-
ers, mechanical atomizing nozzles for improved
combustion, and additional flame detectors (all
three systems).
Purchase of an external combustion chamber for re-
heating the venturi exhaust gas (to be installed at
a future date).
Installation of a Koch Flexitray (with a bottom steam
sparger) in the TCA scrubber to prevent the bulk of
the entrained slurry from impinging onto the demis-
ter (November, 1972). The effluent irrigation liquor
was re-routed from the sewer to the effluent hold
tank in February, 1973.
Installation of a bubble-type density element in the
limestone slurry makeup tank.
Revision of the piping and associated instrumenta-
tion to provide automatic density and level control for
the three scrubber systems, including new variable
7-26
-------
speed Centriseal type pumps (venturi and TCA systems)
and Clarkson pinch valves (Hydro-Filter system).
• Installation of a Ventri-Rod presaturator in the TCA
inlet duct for gas cooling, complete with slurry feed
pump and flow measuring device.
• Modification of the makeup water control valves for con-
trol at low rates required for closed liquor loop
operation.
• Enlargement of the clarifier underflow pump suction
piping from 1-1/2 inches to 3 inches to minimize
line pluggage (all three systems).
• Provision of low liquid level alarms in the clarifiers
(all three systems).
• Modification of the DuPont SC>2 analyzer sampling sys-
tem on the three inlet flue gas ducts in November,
1972 (see Section 7.4. 1 for details).
• Provision of panel-mounted control of limestone slur-
ry addition and connection of the new instrumentation
read-outs to the EMC data acquisition system.
• Modification of the demister flush piping to pro-
vide underspray in the spray tower and Hydro-
Filter. Combined with the revision for den-
sity control, both fresh water and clarified liquor
can be routed to the spray tower and Hydro-
Filter demisters and to the Koch tray in the TCA
scrubber.
• Provision of strainer baskets in five slurry circula-
ting tanks to reduce spray nozzle pluggage (to be in-
stalled at a later date).
• Revision of the spray tower slurry feed system
for increased liquid circulation (to be installed
at a later date).
Figures 7-1, 7-2 and 7-3 (drawn roughly to scale) show the three scrub-
bers with the modifications to the demister systems and with the inclu-
sion of the Koch Flexitray in the TCA system. These figures can be
compared with Figures 2-1, 2-2, and 2-3, which show the three scrub-
bers before the modifications.
7-27
-------
GAS OUT
CHEVRON DEMISTER
AFTER-SCRUBBER
INLET SLURRY
THROAT
ADJUSTABLE PLUG
YENTURI SCRUBBER
DEMISTER WASH
DEMISTER WASH
INLET SLURRY
EFFLUENT SLURRY
5'
APPROX.SCALE
EFFLUENT SLURRY
Figure 7-1 Schematic of Venturi Scrubber and
After-Scrubber After Modification
7-28
-------
GAS OUT
CHEVRON DEMISTER
INLET KOCH TRAY-
WASH LIQUOR
*^rir-^n n n r-M«fj-i
.. J W V W It]
KOCH TRAY
w
EFFLUENT KOCH
TRAY WASH LIQUOR
1 1
STEAM SPARGE
RETAINING GRIDS /
/»•* IU
A A A
0"°"°" ~
00°o0o
00 0
P*P_P-.o_Q.
on ° o
o o
s° O O ° O
»°0*f»
a°o_ O_°QJ£
\ /
i '
|
INLET SLURRY
/MOBILE PACKING SPHERES
5'
APPROX. SCALE
EFFLUENT SLURRY
Figure 7-2 Schematic of Three-Stage TCA Scrubber Without
Trap-out Tray After Modification
7-29
-------
GAS OUT
DEMISTER WASH
INLET SLURRY
INLET SLURRY
GAS IN
DEMISTER WASH
CHEVRON DEMISTERS
TURBULENT LAYER
GLASS SPHERES
EFFLUENT SLURRY
5'
APPROX. SCALE
EFFLUENT SLURRY
Figure 7-3 Schematic of Hydro-Filter Scrubber
After Modification
7-30
-------
Figures 7-4, 7-5 and 7-6 show typical modified system configurations for
the venturi, TCA and Hydro-Filter systems, respectively. These figures
can be compared with Figures 2-4, 2-5 and 2-6, which show the three
scrubber system configurations before the modifications in February and
March, 1973.
7. 6 SYSTEM RELIABILITY
In this section, the reliability (e. g. , scaling and plugging potential, long-
term equipment operability) for the three scrubber systems will be dis-
cussed. Due to the nature of the factorial tests (i. e. , open liquor loop,
steady run conditions for short periods), no formal attempt was made to
evaluate scaling potential for each test. The equipment operability dur-
ing the factorial testing has already been covered in detail in the prev-
ious sections.
7. 6. 1 Short-Term Factorial Testing
Throughout the limestone factorial test period, no significant sulfate-
based scaling occurred in the scrubber systems. Except for a single
series of tests on the TCA system (discussed below), no significant
sulfite scaling occurred in the systems. The demisters, however,
did require periodic cleaning of soft solids, until modifications were
made to the demister wash systems (see Section 7. 2. 1).
During a special closed liquor loop test sequence with the TCA system
(see WC run series in Table 5-2) to simulate TVA's Colbert pilot plant
(TCA scrubber with five grids and no spheres), the feed stoichiometry
was, inadvertently, in excess of two moles CaCOo/mole SC>2 inlet.''" The
See Section 5. 1 for discussion of problems with limestone additive flow
control after mid-November, 1972.
7-31
-------
il
rt
o
• ~H
-M
R)
O
0)
<3
CO
o
a
Q
o
en
en
CD
O
O
ri
o
-------
i
1 u.
' 1
ET
re
a
_? .,_/*-
1
>-
DC
ae^£
§1
UJ
i
un o
V 3.
£
I I
5 3 >-
o .a t
» r 3
5 re —
a. co
O®®
a
o
•tH
4J
(fl
U
'
O
0)
4-1
4-1
to
U
H
SP
•H
Q
0)
o
o
U
• rH
&
H
LH
I
7-33
-------
i/l o
s^
O ^J
c
o
1
t/>
*fe
Pi
o
o
5
(D
4J
CO
f-t
CD
I
bJO
P
O
w
CO
0)
o
o
I
r-
0)
^
S
bD
O®0
7-34
-------
scrubber walls and bottom grids became heavily coated with sulfite-based
scale during the 21 days of testing. It has since been determined that the
sulfite scale formation was caused by operation of the system with a scrub-
ber outlet liquor pH in excess of 6. 3. During a simulation of the TV A
tests at the EPA pilot plant in Durham, N. C. (Reference 16), sulfite
2
scale formed rapidly (60 mg/m in less than 40 hours) with a stoichio-
metric ratio of 2. 5 and a scrubber effluent pH of 6.4, while no scale
formed during 400 hours of operation with a stoichiometric ratio of 1. 25
and a scrubber effluent pH of 6. 0. TVA has also confirmed, at their
Colbert pilot plant, that the excess stoichiometry (and corresponding
high pH) caused the scale formation.
As discussed previously, the operability (e.g. , solids control) of the
three scrubber systems was poor during the open-loop factorial testing.
Following the February-March, 1973, modifications (see Section 7. 5),
the operability and reliability of the three scrubber systems was con-
siderably improved. It is still necessary, however, to make the following
improvements:
Elimination of solids build-up and saturation spray
nozzle pluggage in the inlet ducts of the TCA and
Hydro-Filter scrubbers.
Selection of proper slurry spray nozzles to sustain
efficient operation and minimize the plugging of the
marble-bed in the Hydro-Filter.
Addition of limestone slurry makeup in automatic
operation for proper stoichiometry control.
Automatic control for circulating slurry density
and for tank liquid levels on all three systems.
Automatic density control of the clarifier under-
flow streams.
7-35
-------
7. 6. 2 Reliability Verification Testing
Venturi System. Test Run No. 501- 1A (see Figure 6-1) commenced at
1400 hours on April 9, 1973, and continued intermittently for a total of
645 hours until its termination at 1700 hours on May 9, 1973.
There were three shutdowns during the run. The first occurred at 0530
hours on April 10, for seven hours at the end of the limestone depletion
period. The second shutdown was at 0530 hours on April 11 due to high
induced draft fan vibration and lasted 67 hours. The third shutdown was
at 1250 hours on April 27 for five hours to replace a damaged shear
pin connecting the induced draft fan damper to the damper actuator.
Inspection during the April 10 shutdown (after a 16 hour limestone de-
pletion period) indicated the system to be in very good condition. There
was light scattered sulfate base scaling (5-10 mils thick) on the bottom
of the demister, on the bottom of the trap-out tray, and on the adjacent
scrubber wall areas. The top of the demister was clean and the flue
gas outlet duct was free of soot, oil, and moisture accumulation. The
system was restarted without cleaning.
The system was inspected again on May 10, after an additional 629
hours of operation, at the completion of Run 501-1 A.
The venturi scrubber was coated with a thin sulfate base scale (less than
10 mils). Approximately one-third of the two annealed 316 stainless
steel, bolt-nut assemblies of the guide vanes were eroded. The flooded
elbow was covered with a 35 mil sulfate base scale. The bottom of the
trap-out tray and isolated sections of the top slurry and demister bot-
tom flush headers were covered with a heavy mud type deposit. The
7-36
-------
spray tower and the top and third (from the top) spray headers were
covered with scale of non-uniform thickness (up to 15-25 mils).
The reheater sleeve was severely warped (on the north side) and the re-
fractory was cracked. The exhaust gas duct above the reheater was
covered with about 1/8 inch of soot-limestone deposit with no evi-
dence of moisture or oil.
The preceding inspection summary is presented in Figures 7-7 and 7-8.
Inspection forms shown in the figures have been developed for rapid
dissemination of inspection results and are not intended to replace the
corresponding detailed inspection reports.
The overall control of the slurry density and tank levels was much im-
proved as compared with operation during the factorial tests. However,
the density and level controls could not be operated successfully on auto-
matic control as had been intended. The performance of the bubbler type
density element was erratic, and the demister flush liquor rate (con-
taining the raw water makeup) had to be controlled manually.
The unreliable performance of the effluent hold tank level element (dia-
phragm type) made it necessary to control the waste slurry bleed on flow
rather than on level control.
The variable speed Centriseal pumps performed satisfactorily. How-
ever, vapor lock occurred in the new Centriseal pump in slurry bleed
service at flows below 40 gpm. It was necessary to install a pump dis-
charge-to-suction recirculating line and a pinch valve in the bleed line
to enable satisfactory pump operation at the required low flow rates.
7-37
-------
O «/»
Z 13
> O
„ 9
S
to
< o
z o
O 2
P Q
U j
a.
i^
s
a
ce
fl
o
•H
•U
U
CD
(^
X
>s
D^
^
r — ~^
\
^ X
V
Vw
S,
^N^
S
s.
— x,^—
>» -5?
V
X
^>>
fe * N
^
CD
*i^*m
A\\\\\\\\\\\\
>
^1
h*-< |
Sx
M
bo
•rH
3
7-38
-------
Z Q
Z
<
rc/~
CO
CO
ZD
DC.
O
CO
eg
UJ
a:
•o
m
o
• M
4->
o
I)
u
i
-------
The fresh limestone slurry flowmeter was inaccurate at the required low
additive flow rates (about 1. 5 gpm or less).
The performance of the clarifier was generally satisfactory except for
occasional periods of turbid overflow. The density indicator control
loop on the clarifier underflow could not be operated on automatic. The
density element (Ohmart gamma radiation type) on the discharge of the
variable speed pump measures the density of the diluted underflow (the
pump is provided with water seal). The seal water adversely affects
the density of the combined stream as the underflow rate drops. This,
in turn, causes the pump to idle and, with no underflow, density control
is lost (constant seal water rate cannot be maintained with the existing
system). Density control was established indirectly, i.e. , by adjusting
the underflow rate manually. In spite of the poor accuracy of the 0 to 80
gpm underflow meter at the required flow rates, the solids concentra-
tion was maintained within the 17 to 27 wt % range.
System level control was maintained manually by adjusting the cycle of
the demister wash timer.
The ratio of clarified liquor to raw water was maintained between approxi-
mately 1:1 to 3:1. The demister was flushed intermittently (at a timer
cycle of approximately 70 percent "on" and 30 percenf'off") on the under-
2
side. The average flush rate (during the "on" cycle) was about 0. 4 gpm/ft .
The overall scrubber pressure drop and the pressure drop across the
demister remained constant throughout the test run.
7-40
-------
TCA System. In general, the control and operability of the three scrubber
systems was similar following the February-March, 1973, modifications.
Therefore, only brief references will be made to similar problem areas
for the TCA and Hydro-Filter systems.
Test Run 501-2A (see Figure 6-2) started at 0315 hours on March 22, 1973,
and continued intermittently for a total of 579 hours until its completion
at 1145 hours on April 23, 1973. Ten shutdowns of the system during the
test run totalled 197 hours.
The first shutdown was at 0045 hours on March 23, at the end of the lime-
stone depletion period. Solids deposition •was limited to the Ventri-Rod
presaturator (approximately one-fifth of the flow area was plugged) and
to the gas duct area located downstream. A thin five mil scale covered
the scrubber walls below the bottom bed. The bottom of the Koch tray was
sparsely covered with slurry solids. The rest of the scrubber, the de-
mister and the exhaust gas duct were virtually free of deposits.
The unit was shut down on March 25, 27, 30, 31, and April 11 and 16,
for a total of 126 hours because of the high pressure drop (between 5. 5
and 13.0 inches water) across the Ventri-Rod presaturator. The solids
deposition rate did not diminish when the ceramic nozzle (for introduc-
ing the presaturator slurry feed) was replaced with a stainless steel
open nipple.
Because of maintenance work on the AEC water system on April 6, the
unit was srm^ down and inspected. Once again the Ventri-Rod presaturator
was found to be seriously plugged (about 70 percent). With the exception
of a thin (5 mils) scale on parts of the scrubber walls beneath the Koch
7-41
-------
tray and on the demister bottom, the system was very clean, including
the exhaust gas duct. The absence of solids accumulation on the bottom
of the Koch tray is attributed to frequent steam sparging (once an hour
for one minute as contrasted with once a shift for one minute during
factorial testing). The stainless steel sleeve in the reheater was not
deformed. However, the refractory was severely cracked above the
burner ports and had to be patched. Warping of the fan blades was not
significant (0. 041 to 0. 167 inch from the horizontal).
During the April 16 shutdown, the Ventri-Rod presaturator assembly was
replaced with a humidification section consisting of four Bete ST-24 FCN
nozzles. These nozzles were, in turn, replaced with larger ones (Bete
ST-32 FCN nozzles) during the April 18 shutdown, to reduce the frequent
plugging.
The system was inspected on April 23, at the completion of Run 501-2A.
Approximately four cubic feet of solids deposit accumulated upstream
>'<
of the cooling spray nozzles. ' One nozzle was found plugged with for-
eign material. There was no solids deposition between the cooling noz-
zles and the scrubber.
The wires of the bottom grid of the middle bed eroded at the perpendi-
cular junctions over a 14 square inch area (two apertures) through which
all the spheres of the middle bed had dropped to the bottom bed. Sever-
al wires of the bottom grids of the top and bottom beds were also found
loose with the cross-wire junctions significantly eroded.
More recent reliability verification tests have shown that soot-blowing
and spray nozzle configuration changes will alleviate this pluggage
problem.
7-42
-------
Slight solids accumulation occurred on the east and -west scrubber walls
immediately beneath the Koch tray. The bottom of the Koch tray was
covered with lightly scattered solids while the top remained clean. Ap-
proximately one sixteenth inch scale accumulated on the bottom vanes
of the demister. The reheater and the exhaust duct were free of soot
and solids deposit.
In general, the scrubber was in a very clean condition as shown in
Figure 7-9.
About 36 of the polyethylene and polypropylene spheres were eroded
and collapsed.
The ratio of the clarified liquor to fresh water in the irrigation liquid
to the Koch tray ranged between about 1:1 and 3:1. The irrigation rate
varied between 17 and 29 gpm. The solids concentration of the Koch tray
effluent remained below 1. 0 percent by weight. This effluent was routed
to the circulating slurry system for closed-loop operation.
During steam sparging periods, a portion of the steam was discharged
through the exhaust system.
The overall scrubber pressure drop remained between 5 to 7 inches of
water, and the combined pressure drop across the Koch tray-demister
system remained at two inches of water throughout the test run.
The 30 foot diameter clarifier operated with a clear overflow and at an
underflow solids concentration of about 40 percent by weight — the
expected final settled density of the sludge. This solids concentration
was maintained by flow control.
7-43
-------
R
o\
>»
«
1
M
z^
-^
5^
CL
2 8
h- VI
^ 2
?
CS|
i
I
^
>s
•S
*iS
I
«
^
3
•§
^
It
$S
^
1
1
1
1
__
--
rn
SJ1!
B
^
ill
Oco
UJ
§>N
gSg
8-z
UJ
*SI
n
PO^
,Sn t-
^*>
^
• 'w .Jv
i^OV.
!^l| '
Q k- —
• x3- 1
i$aii
ST.
^i,x
^i>
^h>
Sk1 ^
N^^
k I^K>
^ < < fC
"^
j
•>
J
^
^o:
UJ
Q
<;S Q
21 £
x>
^2
O.
l/>
^ T*
.00
^
vS
•^
43
1
«M
A
s
A
K
AVDdS
ONHOOD
to
^
/
V
*?.'
•M
s
CO
CO
ID
CG
o
CO
<
O
fl
o
•i-*
-1-1
u
V
ft
CO
fl
HH
«;
o
H
o
i
r-
-------
System reliability and equipment performance for the TCA system were
similar to those listed for the venturi system in the following areas:
• Slurry density and tank level control
• Operation of the Centriseal bleed pump at low flow
rates
• Automatic control of clarifier underflow density
• Routing of clarified liquor to the Koch tray
Hydro-Filter System. Test Run No. 501-3A (see Figure 6-3) started at
1955 hours on March 14, 1973, and continued intermittently for a total
of 266 hours until 0530 hours on March 28. Run 501-3B started at 1340
hours on March 30, and was completed at 0515 hours on April 23, fol-
lowing intermittent operation for a total of 498 hours.
The system was shutdown at 0845 hours on March 16, at the end of the
depletion period. Solids deposits in the scrubber were practically non-
existent. Only a thin scale was observed on the scrubber walls, slurry
headers, spray piping and on the bottom of the demister. A very small
solids deposit was found on the north side of the inlet gas duct at the
cooling spray header and did not require removal.
Since this was the first closed-loop run of the entire test facility, the
unit was shutdown for inspection at 0600 hours on March 23, after 158
hours of operation. Three of the cooling spray nozzles and two of the
bottom spray nozzles were plugged with debris and slurry. The inlet
duct between the cooling spray header and the scrubber was plugged
with about one cubic foot of solids. A small section of the bed support
grid was plugged.
7-45
-------
Additional solids deposit was not noticed on the demister. There was
no evidence of soot, oil or moisture carryover in the reheater outlet
duct, but some solids deposit was noticed in the fan inlet ducts. One
of the eight fan blades was significantly warped and another only slightly
warped.
The system -was cleaned (gas inlet duct and the bed support grid) before
the continuation of Run 501-3A at 1740 hours on March 23.
The unit was shut down on March 24, 25, 26, 27 and 28 for a total of 91
hours to clean the plugged bottom spray nozzles. The nozzle pluggage
was caused partly by the loss of marbles to the effluent hold tank through
a loose section of the support grid and partly by accumulation of other
foreign material in the tank.
Run 501-3A was terminated at 0530 hours on March 28, to clean both
the scrubber internals and the effluent hold tank. Replicate Run 501-3B
was started at 0340 hours on March 30, and the depletion part was com-
pleted at 1545 hours on March 31. Very small deposits were noticed in
the inlet duct on the north and south sides, downstream of the cooling
sprays. A high sulfate base scale of about 10 mils was deposited on
the bottom spray headers and nozzles. The bed support grid was free
of solids. The bottom of the demister was coated with one-sixteenth
inch of solids and the top was coated with a scattered flaky deposit.
There was no evidence of oil or soot accumulation in the duct above the
reheater. The north side of the reheater sleeve was slightly warped.
The slurry discharge line from the bed overflow weirs into the effluent
hold tank was extended below the liquid level for vacuum seal. The sys-
tem was not cleaned during the shutdown.
7-46
-------
Run 501-3B was interrupted on April 4, 6, and 11 — twice for cleaning
the plugged cooling spray nozzles and once for maintenance work on the
AEG water supply. Six of the bottom sjaray nozzles were also found
plugged during the April 11 shutdown. To minimize plugging of these
nozzles, the dead end of the spray header was eliminated by using a
recirculating line to the effluent hold tank.
The system was inspected on April 23, after the end of Run 501-3B at
0515 hours. The accumulation of solids and the extent of bed pluggage
is shown in Figure 7-10.
About two cubic feet of solids buildup was found on the north and south
sides of the inlet duct between the cooling spray header and the scrubber.
Up to one inch of solids deposit and scale was found on the bottom spray
headers and scrubber walls below the bed. About 29 percent of the bed
was either plugged or had stratified layers of marbles (i. e. , in the ini-
tial stage of plugging). The bottom of the demister was covered with
one-eighth inch of slurry solids, the top with light, non-uniform, one-
sixteenth inch thick scale and solids. The diffuser vanes in 13 of the
16 improved type bottom spray nozzles disappeared.
The reheater sleeve continued to warp on the north side, and additional
cracking of the refractory occurred. The reheater outlet duct collec-
ted a one-sixteenth inch deposit of dry soot and solids. The deforma-
tion of the induced draft fan blades continued and their deformation from
the straight line pattern varied from 0. 3 to 0. 55 inch.
The apparent weight loss of the randomly sampled bed marbles averaged
about six percent.
7-47
-------
fl
o
• H
4->
o
OJ
ffi
o
r—I
I
bo
7-48
-------
The ratio of clarified liquor to fresh water in the demister flush liquid
ranged between 1:1 and 3:1. The demister flush rate varied between 0.4
and 0. 6 gpm/ft^ during flushing periods. The demister was flushed in-
termittently on the underside at a timer cycle of approximately 70 per-
cent "on" and 30 percent "off". Liquid carryover through the demister
during flush periods was excessive (as viewed through the observation
windows) and was noticeable as steam discharged to the atmosphere.
The overall pressure drop varied between 7.8 and 10.0 inches of water,
depending upon the solids buildup in the duct (at the cooling spray header)
and across the marble bed. The pressure drop across the demister re-
mained constant at about 0. 2 inch of water
The 20-foot diameter clarifier was unable to handle the solids loading,
resulting in turbid overflow operation throughout the run. The solids/
liquid interface was practically at the top of the clarifier and the poor
seal between the top V-notch metal plate and the clarifier wall resulted
in considerable solids carryover in the overflow. The overflow solids
content at times was as high as four percent by weight. The recycle
of solids to the scrubber loop made it practially impossible to main-
tain the solids concentration of the circulating slurry at the required
level.
System reliability and equipment performance for the Hydro-Filter sys-
tem were similar to those listed for the venturi system in the following
areas:
• Slurry density and tank level control
• Automatic control of clarifier underflow density
• Routing of the clarified liquor - fresh water mixture
to demister flush
7-49
-------
Section 8
ANALYSIS OF PRESSURE DROP DATA
8. 1 VENTURI SCRUBBER
In Reference 7, a proposed correlation was presented for fitting venturi
pressure drop. ' A further analysis of the differential equations which
describe pressure drop for this system (see Reference 16) indicated
that the ratio of throat length to plug diameter,-* ID, should be included
in the expression. The inclusion results in a much improved fit to the
i, especially at high gas velocities and high liquid-to-gas ratios.
The following equation is a fit to the venturi air/water and soda-ash
pressure drop data (see Tables 4-1, 4-2, and 4-3):
*
The proposed correlation (Equation 3 of Reference 7) should have
shown the coefficient A- multiplied by L/^ .
Volgin (Reference 17) has included a "throat length" term in his venturi
pressure drop correlation, but gives no theoretical grounds for the
inclusion.
8-1
-------
witK
(8-2)
= 1.8
where
Ap = pressure drop across venturi, in. HO
gas velocity at throat, ft/sec
liquid-to-gas ratio through scrubber, gal/mcf
throat length, ft
Q = venturi plug diameter = 3. 2 ft
o
n*- = venturi throat area (used to calculate gas velocity), ft
The preconstant in the first term on the right-hand side of Equation 8-1
was fit to the "air only" (L = 0) data. The four coefficients in the second
term on the right-hand side fit to the remaining data.
Equation 8-1 accounts for 9V percent of the variation of the air/water
and soda-ash data, with a standard error estimate of 0. 7 inches H2O.
Measured and predicted (Equation 8-1) values of venturi pressure drop
are compared in Figure 8-1. Also shown in Figure 8-1, but not in-
cluded in the fit of Equation 8-1, is the pressure drop data for the fac-
torial limestone wet-scrubbing runs presented in Table 5-1.
8-2
-------
MEASURED PRESSURE DROP, in. HO
Figure 8-1. Comparison of Experimental Data and Predicted Values
(Equation 8-1) of Pressure Drop for the Chemico Venturi
8-3
-------
The four coefficients in the second term on the righthand side of Equa
tion 8-1 were also fit to the factorial limestone data. The resultant
equation is:
Ap - * ~''
The equation accounts for 92 percent of the variation of the data. Mea-
sured and predicted values of pressure drop are compared in Figure
8-2.
Parametric plots of venturi pressure drop from Equation 8-4, as a
function of Ljfy and (/£.for three different plug positions, are shown in
Figures 8-3, 8-4, and 8-5.
8. 2 TCA SCRUBBER
The following equation was fit to the TCA pressure drop data for the air/
water and soda-ash runs (see Tables 4-4, 4-5, and 4-6) and for the lime-
*i'
stone wet-scrubbing runs (Table 5-2):
Ap = 1.2. -1- &.5-IO L/ hp/dp + ^S (8-5)
*
Thirteen runs were eliminated from this analysis. These included nine
TVA grid-tower runs for which the inlet gas duct was partially plugged
and four runs operated under "flooding" conditions.
8-4
-------
15
. 10 -•
LLJ
Of.
(/)
CO
a.
a
o 5
oe.
a.
O WET-LIMESTONE SCRUBBING DATA
MEASURED PRESSURE DROP, in. HO
Figure 8-2. Comparison of Experimental Data, and Predicted Values
(Equation 8-4) of Pressure Drop for the Chemico Venturi
-------
25
20 ••
CN
15 ••
10 ••
LU
OC
Q.
5 ••
THROAT LENGTH = 1.46 ft
THROAT AREA = 6.6 ft2
10
20
30
40
50
LIQUID-TO-GAS RATIO, gal/mcf
Figure 8-3. Predicted Pressure Drop for Chemico Ven-
turi: One Hundred Percent Plug Opening
8-6
-------
25
20 • •
15 •
Q.
O
LU
UJ
Of
a.
10 •
5 •
10
THROAT LENGTH- 1.01 ft
THROAT AREA-4.2 ft2
20
30
40
50
LIQUID-TO-GAS RATIO, gal/mcf
Figure 8-4. Predicted Pressure Drop for Chemico Venturi:
Fifty Percent Plug Opening
1-7
-------
25
20 •-
CN
15 -•
Q-
o
to
LU
10 ••
5 ••
THROAT LENGTH = 0.56 ft
THROAT AREA = 1.8 ft2
-4-
•4-
10
20
30
40
50
LIQUID-TO-GAS RATIO, gal/mcf
Figure 8-5. Predicted Pressure Drop for Chemico Ventxiri:
Zero Percent Plug Opening
-------
where:
Ap = pressure drop across TCA (excluding demister), in. HO
w = gas velocity through scrubber, ft/sec
|_/Cl = liquid-to-gas ratio through scrubber, gal/mcf
Up = total height of packing, in.
u.p = diameter of packing =1.5 in.
MS = number of grids (screens)
Equation 8-5 gives the best least-squares fit not only for the combined
air/water, soda-ash and limestone data, but also for the limestone data
alone. The equation accounts for 91 percent of the variation of the com-
bined data and for 94 percent of the variation of the limestone data. Stan-
dard errors of estimate are 0. 70 inches H_O overall and 0.62 inches H_O
for the limestone data. Measured and predicted values of pressure drop
are compared in Figure 8-6.
The pre-constant of 1. 2 inches HO on the righthand side of Equation
8-5 represents the average pressure drop for the TCA system when
operated as a "spray tower" (no grids or spheres). The second term on
the righthand side represents the pressure drop across the TCA bed of
spheres. The form for this expression was obtained from the work of
Happel (Reference 18) and Leva (Reference 19), for pressure drop
*•'<
through a two-phase fluidized bed at high Reynolds numbers. ""
•-_ , , , (plastic sphere diameter) (gas velocity) (gas density)
Reynolds number =-^ c ; 7-^ :—: r ° *~L
(gas viscosity)
8-9
-------
12
10 •-
. 8 -•
0.
O
6 •-
Q
LLJ
2 •-
O LIMESTONE WET-SCRUBBING DATA
O AIR/WATER & SODA-ASH DATA
4-
+
+
-4-
-h
468
MEASURED PRESSURE DROP, in. I
10
12
Figure 8-6. Comparison of Experimental Data and
Predicted Values of Pressure Drop for
the TCA System
8-10
-------
Parametric plots of TCA pressure drop from Equation 8-5, as a func-
tion of l/£j and \f for various internal configurations (number of grids
and height of spheres), are shown in Figures 8-7 through 8-10.
8. 3 HYDRO-FILTER SCRUBBER
The following equation was fit to the Hydro-Filter pressure drop data
for the air/water and soda-ash runs (see Tables 4-7, 4-8, and 4-9):
Ap a O.OSb IT2"* 0.0 1 1 (<-/<*) l^'"' (
where
= pressure drop across Hydro-Filter (excluding demlster),
in. H20
\f - gas velocity through scrubber, ft/ sec
L|6i - liquid-to-gas ratio through scrubber, gal/mcf
H^ = height of marbles, in.
^H\ = marble diameter = 0.75 in.
Equation 8-6 accounts for 86 percent of the total variation of the data,
with a standard error of estimate of 1.2 inches HO. Further attempts
L-I
to improve the fit of the equation by including other variables, such as
height of turbulent layer, were unsuccessful.
The pre-constant in the first term on the righthand side of Equation
8-6 was fit to the "air only" (L = 0) data. The three coefficients in
the second term on the righthand side were fit by least squares to the
remaining data.
8-11
-------
5 •-
c
fe
OL
LLJ
a:
to
co
4 - .
2 - -
10
GAS VELOCITY, ft/sec.
Figure 8-7. Predicted Pressure Drop for the
Four-Grid (No Spheres) TCA System
12
8-12
-------
o
CM
0.
O
g
to
GAS VELOCITY, ft/sec.
Figure 8-8. Predicted Pressure Drop for the Six-
Grid (No Spheres) TCA System
8-13
-------
15
o
o
^c
V
Q.
CX.
ULJ
D
oo
UJ
§f
5 •-
GAS VELOCITY, ft/sec.
10
12
Figure 8-9.
Predicted Pressure Drop for the Four-Grid
Three -Stage TCA System: Five Inches of
Spheres Per Stage
8-14
-------
15
o
c»
c
10 •-
oc
«/>
l/>
UJ
Q£
a.
5 -•
10
12
GAS VELOCITY, ft/sec.
Figure 8-10. Predicted Pressure Drop for the Four-Grid
Three-Stage TCA System: Ten Inches of
Spheres Per Stage
1-15
-------
Measured and predicted values of Hydro-Filter pressure drop for Equa-
tion 8-6 are compared in Figure 8-11.
Also shown in Figure 8-11, but not included in the fit of Equation 8-6, is
the pressure drop data for the open-loop limestone wet-scrubbing runs
presented in Table 5-3. The equation is not a good representation of the
limestone data, perhaps due to an effect of percent solids in the slurry
and also because there was gradual pluggage of the marble-bed during
many of the test runs (several runs whose pressure drops clearly indi-
cated pluggage were omitted from the analysis).
The following equation, based on the form of Equation 8-6, was fit to
the data from the limestone wet-scrubbing runs:
A6*) , ,4 / L / j \
=• 6.\ t o.oo6Z(L|6j) u* " tHm/am) (8-7)
Equation 8-7 accounts for 75 percent of the total variation of the data,
with a. standard error of estimate of 0.8 inches HO. The equation does
not hold for pressure drops less than six inches H_O.
LJ
Measured and predicted values of Hydro-Filter pressure drop for Equa-
tion 8-7 are compared in Figure 8-12.
Parametric plots of Hydro-Filter pressure drop from Equation 8-7, as
a function of L \(% and (f for three and five inches of marbles, are shown
in Figures 8-13 and 8-14.
8-16
-------
12 --
10 ••
Q_
§ •
LLJ
Of
00
OO
6 •-
y
Q
V 4
2 •-
O LIMESTONE WET-SCRUBBING DATA
O AIR/WATER & SODA-ASH DATA
O
4 6 8 10
MEASURED PRESSURE DROP, in. HJ
12
Figure 8-11. Comparison of. Experimental Data and Predicted
Values (Equation 8-6) of Pressure Drop for the Hydro-Filter
1-17
-------
12
11 ••
o^
a.
§ '
UJ
to
o.
O
£
y
5
8 ••
7 •-
6 •-
O LIMESTONE WET-SCRUBBING DATA
MEASURED PRESSURE DROP, in.
Figure 8-12. Comparison of Experimental Data and Predicted
Values (Equation 8-7) of Pressure Drop for the Hydro-Filter
8-18
-------
15
13 ••
12 ••
CM
c 11
a.
oc
D
t/i
10 ••
9 ••
8 •-
6 •-
GAS VELOCITY, ft/sec.
Figure 8-13. Predicted Pressure Drop for the Hydro-Filter
with Three Inches of Marbles
8-19
-------
15
14 ••
13 ••
12 ••
c' 11 ••
O-
o
10 ••
O£
5>
1C 9
on
a.
8 ••
7 ••
6 ••
10
GAS VELOCITY, ft/sec.
Figure 8-14. Predicted Pressure Drop for the Hydro-Filter
With Five Inches of Marbles
8-ZO
-------
Section 9
ANALYSIS OF SODIUM CARBONATE SCRUBBING DATA
In Section 9. 1, the high concentration sodium carbonate (soda-ash) data
is analyzed and models for predicting SO removal efficiencies for the
three scrubber systems are presented. In Section 9.2, the low concen-
tration soda-ash data for the Chemico venturi is analyzed and a model
for predicting SO removal efficiency and the liquid-side mass transfer
resistance is presented.
9. 1 HIGH CONCENTRATION SODIUM CARBONATE DATA
As mentioned previously, the high concentration soda-ash tests were
designed, primarily, to determine uncertain model coefficients where
the gas-side mass transfer is rate controlling.
The initial venturi sequence (Table 4-10, Runs 202-1A through 202-1D)
showed, for an inlet liquor pH of 9. 5 and Na concentration greater than
0. 5 wt %, that the vapor pressure of SO at the gas-liquid interface
e^-rywhere within the scrubber was essentially zero, i.e. , gas-side
resistance is rate controlling everywhere. The data also indicated
For gas-side controlling mass transfer, the gas absorption efficiency
is independent of liquor composition (see following Section). Note, that
Runs 202-1A and 202-1D in Table 4-10 give the same absorption efficiency.
9-1
-------
that gas-side resistance is controlling above an inlet liquor pH of 8. 5,
for a Na concentration of 1 . 0 wt %. These results are, generally, in
agreement with predictions made with the use of the Radian equilibrium
computer program (Reference 13).
The high concentration venturi sequence (Runs 250-1A through 250-IB
in Table 4-10) showed that SO inlet gas concentration has no effect on
t-i
SO removal, which is in agreement with theory when gas-side resis-
tance is controlling (see following Section).
9. 1. 1 Venturi Scrubber
The venturi scrubber model (Reference 7) for gas-phase limited SO
LJ
removal has been fitted to the Shawnee high concentration soda-ash data
(pH = 9. 5, Na = 1 wt %) for the scrubbing of air and SO gas mixtures
and the scrubbing of flue gas. As was indicated in Reference 7, the
predicted SO removal is sensitive to both the mean diameter of the
I—I
liquid droplets formed at the venturi throat and the initial velocity of
these droplets down the throat.
Using the Nukiyama/Tanasawa (Reference 20) equation (based upon the
throat cross-sectional area) to predict mean droplet diameter, the best
overall fit of the computer model to the data was obtained for an assumed
initial droplet velocity at the throat entrance of 10 ft/sec (the maximum
free-fall velocity of the liquid from the nozzle is approximately 16 ft/
sec). The comparison between predicted and measured SO removals
t-i
The results obtained for SO^ removal, under identical conditions of
gas and liquor mass flow rates and plug position, were identical be-
tween the air-SO and flue-gas runs.
LJ
9-2
-------
is shown in Figure 9-1. Not all of the high pH experiments have been
plotted in Figure 9-1, since replicated runs had been averaged to rep-
resent single data points.
The following equations for predicting SO removal in the venturi were
"fitted" to the high concentration soda-ash data from the Shawnee facil-
J
J
(9-D
•]
5fe
(9-2)
A comparison between measured and predicted SO removals from
Equations 9-1 and 9-2 is shown in Figures 9-2 and 9-3. Equations 9-1
and 9-2 account for 99 percent and 96 percent of the variation in the
data, respectively. The standard errors of estimate are 1.2 percent
SO removal for Equation 9-1 and 1. 9 percent for Equation 9-2.
t-i
The forms of Equations 9-1 and 9-2 were obtained from the following ex-
pression which represents SO2 absorption for the condition of gas-side
resistance being rate-controlling (Reference 7):
x
»
(9-3)
;« n
See Section 8. 1 for nomenclature and for equations relating J( and /•-) ,
to percent opening of venturi plug.
9-3
-------
50
Figure 9-1.
MEASURED SO REMOVAL, %
Comparison of Experimental Data and Predicted Values
of SO Removal fromVenturi Computer Model
9-4
-------
MEASURED SO REMOVAL, %
Figure 9-2. Comparison of Experimental Data and Predicted Values
of SO Removal in Chemico Venturi from Equation 9-1
9-5
-------
MEASURED SO2 REMOVAL, %
Figure 9-3. Comparison of Experimental Data and Predicted Values
of SO Removal in Chemico Venturi from Equation 9-2
Li
9-6
-------
where
P 2
Jfyfa ~ gas-side mass transfer coefficient, Ib-mole/hr ft
-^ = gas-liquid interfacial area per scrubber volume, 1/ft
T- = axial distance, ft
^ = gas rate per cross-sectional area, Ib-mole/hr ft
9.1.2 TCA Scrubber
The pre-constant in the following equation for predicting the gas-side
mass transfer coefficient for the TCA scrubber, operated as a spray-
tower (no screens or spheres), has been fitted to the high concentration
soda-ash data from Runs 225-2C through 230-2C in Table 4-13:
, , ,^4 o.B
2. - OA<* L 6 (9-4)*
where
^ = effective height of spray tower, ft
1_ = liquor flow rate per scrubber area, Ib-mole/hr ft
The coefficients for liquid and gas rates in the above correlation have
been obtained from the equation developed in Reference 21 for spray
towers, which was based upon the work of Fair (Reference 22).
As the appropriate value for 2 in the TCA scrubber at Shawnee is
unknown at the present time, Equation 9-4 is expressed in terms of
-£"t rather than JO
9-7
-------
Measured and predicted (Equation 9-4) values of SO- removal and jft,- t-
LJ ^
A
are compared in Table 9-1. The $&&.• i fit explains 93 percent of
the variation in the data and has a standard error of estimate of 1 5 Ib-
mole/hr ft . The SO removal fit explains 65 percent of the variation
in the data with a standard error of 1. 3 percent removal.
9.1.3 Hydro-Filter Scrubber
The two pre-constants in the following equation were fit to the high con-
centration soda-ash data for the Hydro-Filter scrubber:
^4- 2T = h.O rV'^m + O.SI f'V8 (9-5)
where
~£-T ~ distance between top and bottom spray nozzles = Z. 13 ft
L- = total liquor flow rate per scrubber area (sum of top and
bottom sprays), Ib-mole/hr ft
<-m = height of marble layer, ft
The coefficients for liquor and gas rates in the first term on the right-
hand side of Equation 9-5 were obtained from the equation presented in
Reference 23 for the Vf^tfin the glass sphere region of the Hydro-Filter
system.
9-8
-------
I
0s
rt
H
Pi
H
PQ
PQ
t>
51 Pi
0 °
y co
&<
hu
JH
> H
O E
S H
H Pi
Pi O
[T,
C0«
0 <;
w P
P <
H P
r. _i
aS
P <
H £
Pi O
AND P
I GARB
0 3
wB
g p
1=3 O
w co
^
1 o
r=n n
hS
°S
i— _ r
£ H
0 £
£ w
^
ft 0
5 u
"•^i 1
R E
U o
t—4
E
W
E
H
CO
±i
SH
rH
43
?
0
I— 1
h
rH
0
3
cr
•|H
rH
fe
0
r-H
r
PH
rH
•H
ni
T3
(U
H-J
0
H
ro
ni
P
S
^
PH
be
«T
•4->
rt
Pi
g
0)
tf
r~- ui o
f-H en m
CO r-l CO
oo r*- co
r-H CO Ifl
CO rH CO
^ ^ CO
N r-H [>-
^ 0 00
°' 0 0
o o o
on o o
QN G^ (JN
o o o
r-- o r~-
co co o
O sO vO
O O O
o o o
r^ o ^
•, •.
0)
i— i
d
•H
rH
0)
43
f\
4-1
0
rH
U
CO
ni
ni
V
'O
rt
g
(U
rH
(1)
L of these runs w
^-t
i— t
<
00
CO
0)
£
E
PH
H->
9)
r-H
1
rH
0)
43
"§
(4
0
CO
0)
43
E-i
d
PH
PH
0
O
0s
<*H
0
(H
O
• H
^_}
ni
rH
H->
d
once
y
CO
O
CO
-»->
(U
r-H
d
•i-(
T3
ight percent, an
a;
1*
*
(U
CO
ni
o
higher in every
rH
O
.
h
0
o
0
i-H
r^
D
-4->
ni
a
• rH
Vj
r*l
O
rH
PH
PH
ni
V
rH
n!
CO
d
pi
rH
cu
CO
0)
£
•*— •
JH
O
«4H
CQ
0)
rH
e gas temperatu
43
H
•
CO
£
sD
•
on
CO
•H
n)
0)
rH
n)
I-H
n)
d
O
• H
-«->
U
(U
CO
i
CO
e scrubber cros
43
H
9-9
-------
The effects of the turbulent layer and of the upper and lower spray zones
on jfy(A cannot be easily separated, as these effects show a similar
variation with liquor and gas rates. Hence, the spray and turbulent
layer effects have been combined into a single term (the second term
on the right-hand side of Equation 9-5), with the same coefficients of gas
'/- F
and liquor rates as those developed in Reference 21 for a spray tower.
Equation 9-5 can also be written as:
V' *r - !>* f'c-** +3&.0 6, (9-6)
HTl = i.|0-4lf-7V'D4 (9-7)
where nrt is the height of the turbulent layer in feet.
Equation 9-7 was obtained by empirically fitting the turbulent layer
height data from the air /water and soda-ash runs to the gas and liquid
flow rates (see Tables 4-7, 4-8 and 4-9).
The coefficient of HTL. in Equation 9-6 was assumed to be the same as that
theoretically obtained in Reference 23 for the turbulent layer region.
Measured and predicted values of i^H, (Equation 9-5 or 9-6) and SC>2 re-
moval (Equation 9-3) are compared in Table 9-2. The JK(^ fit explains
88 percent of the variation in the data and has a standard error of estimate
3
of 7. 7-lb mole/hr ft . The SO removal fit explains 79 percent of the
£
variation in the data with a standard error of 1. 1 percent removal.
9-10
-------
pj
H
CP
PQ
(4
W U
ffi w
H Pi
^ H
o
/^> *~i
r f""
i
. ~1 y-N
^ o
< f£J
> Q
O £
M
S w
H
CO EH
°0
H.-4
7 i<^
§°
^ PQ
Q PH
W <
Pi O
< g
S Q
S o
ti
o ^;
•7 O
£-< t-H
O H
CO
Pi
s
"S
O
H
o
h
,j<
(
2
!
P
o
•^H ^< ^^ oo oo Is* co co o^ PO o o I-O co iTi co r*~ r^~ so **& ^* o^
inu^inOLOirt oooomirtoo LoOLOioinininin
OO ^^ •^3 r— t O^ r^ O C> 0s ^^ ^3 ^^ ^3 ^5 ^3 ^3 CT^ CT^ ^3 ^5 CT^ CT^
invOsDvor-oo rxixOM^^^ro^-^ rxj-rhcoiriro^foin
OOOOOO OOOOOOOO OOOOOOOO
OOOOOO OOOOOOOO OOOOOOOO
o^c^roiniTiro romoocoiTiinicro rofococococororo
no (M Mrvjfv]rsjcMro(N]fvJro
J(M(MC\3
9-11
-------
9. 2 LOW CONCENTRATION SODIUM CARBONATE DATA FOR
CHEMICO VENTURI
The low concentration (Na < 0. 6 wt %, pH < 8. 5) soda-ash test runs
with the Chemico venturi (see Tables 4-10, 4-11 and 4-12) were made
under conditions of significant gas-side and liquid-side resistance and
of negligible equilibrium vapor pressure, ^ , of SO over the bulk
liquid. The venturi scrubber model, including both gas and liquid-side
resistances, has been presented in Reference 7.
9.2.1 Liquid-Side Mass Transfer Coefficient
The liquid-side mass transfer coefficient can be expressed as (see Ref-
erence 7, Equations 25 through 28):
[0.
Oj3(T-7/>'; (9-8)
where
= liquid-side mass transfer coefficient
Q
•= liquid-side mass transfer coefficient in absence of chemical
reaction
Ag = function of concentration of dissolved reagent B (Ca) in bulk
liquor (see Reference 7, Equation 28)
A^ = concentration of dissolved reactant A (SO ) at gas-liquid
interface
V = stoichiometric coefficient relating the number of moles of
B reacting with one mole of A
T = liquor temperature, °F
9-12
-------
The correlation of Handles and Baron (Reference 24) is used in the ven-
turi scrubber model to calculate liquid-side mass transfer coefficient
for a "circulating" liquid droplet in the absence of chemical reaction:
A 1 A-
-------
Table 9-3
PREDICTED VALUES OF X FOR VENTURI
MODEL FOR LOW-pH SO DA-ASH RUNS
Run No.
259-1A
259-1B
241-1A
243-1A
243-1B
244-1 A
281-1A
280-1A
286-1A
Measured SO?
Removal, %
74
63
54
56
59
68
42
43
62
Fitted
Constant,
0.7
0. 6
1. 3
1.7
1.7
1. 5
1.0
0.4
1.0
Although this approximate value of 1.0 should apply for soda-ash sys-
tems, it may not be valid for limestone systems. For example, pres-
ence of solid particles in the droplets could increase the "circulation"
within the droplets, increasing the value of /\ .
9.2.2 Sulfur Dioxide Removal
For conditions of significant gas and liquid-side mass transfer resis-
tance and for negligible equilibrium vapor pressure of SO over the
bulk liquid, the following expression represents SO absorption:
9-14
-------
(9-10)
where T\ is Henry's law constant.
For the venturi scrubber, the
Equations 9-2 and 9-3.
can be obtained from
Ap°-
(9-11)
The following equation for predicting SO removal in the Chemico ven-
turi has been obtained by combining Equation 9-11 with Equation 9-10
and by fitting the term M $,£./.$,&, to the low concentration soda-ash
data:
Rew\ox>&) r
Ap
o.
(9-12)
where:
= liquid-to-gas ratio, mole/mole
= mole fraction SO in inlet gas
= mole fraction Na in inlet liquor
pH = liquor pH at scrubber inlet
9-15
-------
The proportionality between fyJ&f A Jh Q_ and M- /X»i shown in
Equation 9-12 can be predicted theoretically. Also, Tt\-K^A./-T?ifi is
expected to be a function of liquid-to-gas ratio, stoichiometry, etc.
The statistical significance of ^^ //., and of the interaction group
I £) » / ( -7 } was verified by the linear regression model.
Measured and predicted values of SO removal from Equation 9-12 are
compared in Figure 9-4. Equation 9-12 explains 85 percent of the vari-
ation in the data with a standard error of estimate of 4. 1 percent SO
removal..
The term ^s^\/X(sj^ is analogous to the term Xfl/Xg. i*1 Equation 9-8,
i.e. ty<£, 1S proportional to /(n and XKI is proportional to )^_
For many of the low concentration soda-ash runs, accurate stoichio-
metries were not available. Maximum utility of Equation 9-12 would
require knowledge of stoichiometry as a function of the other indepen-
dent variables. Either stoichiometry or inlet pH was controlled for
the low concentration soda-ash runs.
9-16
-------
90
80
70
O
«» 60
y
Q
LU
QC
Q-
50
40
30
O
30
40
50
60
70
80
90
Figure 9-4.
MEASURED SC>2 REMOVAL, %
Comparison of Experimental Data and Predicted
Values of SO? Removal for the Lo\v Concentration
Soda-Ash Data with the Chemico Venturi
9-17
-------
Section 10
ANALYSIS OF SHORT-TERM FACTORIAL LIMESTONE DATA
In Section 19. 1, "linear" equations are presented which relate SO
removal to measured parameters for the three scrubber systems.
These linear equations, which were produced by a statistical analysis
of the data, are the simplest equations available for adequately depicting
data within the range of variables measured.
In Section 10.2, a theoretical approach is employed to relate SO re-
Lt
moval to the measured parameters for the spray tower, TCA and Hydro-
Filter. A general closed-form equation is developed, which is not in-
compatible with boundary constraints, and which should permit reasonable
extrapolations. Those variables which were found to be significant in
Section 10. 1 were introduced into the general closed-form equations.
In Section 10. 3, complex mathematical models are discussed for computing
SO removal and slurry compositions for the scrubber systems. These
models are, generally, systems of partial differential equations, which
are solved with numerical computer methods.
10. 1 STATISTICAL MODELS FOR SO REMOVAL
LJ
Results of a statistical analysis of the data from the short-term lime-
stone factorial runs are presented in this section as "linear" equations
10-1
-------
relating SO removal to the independent variables. The linear equations
L*
identify the important independent variables affecting SO removal.
Variables may not appear in the linear equations for a number of
reasons. Some of the reasons are:
• The variable did not significantly affect SO2 removal
over the range tested, i. e., the variable was not
statistically significant in improving the fit of the
equations over the range tested.
• The effect of the variable was masked by a simul-
taneous variation of more significant variables.
• The variable was substantially constant for the data
set being analyzed.
• The variable was "non-controlled" (e.g., inlet gas
SO2 concentration, liquor temperature) and may
not have varied in a manner conducive to determina-
tion of its effect on SO2 removal.
10. 1. 1 Venturi Scrubber
A stepwise regression analysis of the 10 venturi runs (with no slurry
in after-scrubber) made in December, 1972, resulted in the folio-wing
equation:
*
% Removal = 9.4 + 0.04 L + 0. 9 Ap (10-1)
The effect of pressure drop on SO2 removal was only observed
below nine inches H^O. Changes in pressure drop above this
value (e. g. , 9-12 inches H?O) had little effect on SO? removal.
10-2
-------
The range of variables covered by these runs, and therefore the re-
gion of application, is:
Gas Flow Rate: 15, 000-30, 000 acfm
Liquor Flow Rate (L): 300-600 gpm
Pressure Drop (Ap): 6-12 in. t^O
Inlet SO7 Concentration: 2,400-2,800 ppm
L* _i..
Stoichiometric Ratio: 1.0-3.0''" moles CaCC>3/mole SO2 inlet
Percent Solids Recirculated: 6-7%
Hold Tank Residence Time: 33-70 min.
Scrubber Cutlet Liquor Temperature: 112-117°F
Equation 10-1 accounts for 89 percent of the total variation of the data.
Gas rate, Stoichiometric ratio, hold tank residence time, percent solids,
inlet SO concentration and liquor temperature were not statistically
LJ
significant (over the above ranges) in improving the fit of the equation.
10.1.2 Spray Tower
The following equation was fit to data from the 15 test runs made with
the four-header spray tower with no liquid to the venturi (see Figures
5-2 and 5-3):
% Removal = 1 6 + 0. 9 L/G (10-2)
The range of variables covered by these runs, and therefore the region
of application, is:
High stoichiometries (greater than 1.75 moles CaCC>3/mole
inlet), after mid-November 1972, were the result of calibration
problems with the limestone additive flowmeters (see Section 5. 1).
10-3
-------
Gas Flow Rate: 10, 000-30, 000 acfm
Gas Velocity: 2. 5-7. 5 ft/sec
Liquid-to-gas Ratio (L/G): 13-61 gal/mcf
Inlet SO2 Concentration: 2, 700-3, 300 ppm
Stoichiometric Ratio: 1.0-3.0 moles CaCO /mole SO inlet
Percent Solids Recirculated: 2-8%
Hold Tank Residence Time: 40-106 min.
Scrubber Outlet Liquor Temperature: 98-128°F
Equation 10-2 accounts for 95 percent of the total variation of the data.
Inlet SO concentration, Stoichiometric ratio, percent solids recirculated,
hold tank residence time and liquor temperature were not statistically
significant (over the above ranges) in improving the fit of the equation.
At constant L/G, a gas velocity of 7. 5 ft/sec gave slightly more removal
( ~ 3 percent) than 5 ft/sec. This velocity effect was not observed
below about 5 ft/sec. Percent solids were actually between 6 to 8 percent
for most runs, with only one run at 2 percent.
10.1.3 TCA Scrubber
The following equation was fit to the 31 EPA TCA runs (see Figures 5-5
through 5-8):
% Removal =47 + 0. 034 L + 1. 4 PSR + 0. 5 Hp - 0. 006 ppm (10-3)
The equation accounts for 85 percent of the total variation of the data.
The range of variables covered by these runs, and therefore the region
of application, is:
10-4
-------
Gas Flow Rate: 15, 000-27, 500 acfm
Gas Velocity (V): 6-11 ft/sec
Liquor Flow Rate (L): 600-1,200 gpm
Percent Solids Recirculated (PSR): 6-14%
Inlet SO2 Concentration: 1, 800-3, 200 ppm
Stoichiometric Ratio: 1.0-3.0 moles CaCOo/mole 50=2 inlet
Hold Tank Residence Time: 3.8-35 min.
Scrubber Outlet Liquor Temperature: 105-122°F
Number of Grids: 4, 6
Total Height of Packing (Hp): 0-30 in.
The percent removal decreases with increasing inlet SG^ concentration
(~6 percent per 1000 ppm). Gas velocity, Stoichiometric ratio, hold tank
residence time, scrubber outlet liquor temperature, and the number
of grids were not statistically significant (over the above ranges) in
improving the fit of the equation. Although not an independent variable,
the pressure drop in the scrubber was also examined and was found not
to correlate well with SO removal. For example, runs giving 92 percent
LJ
removal have been made at pressure drops of 4, 6, and 9 inches HO.
<_<
The 17 EPA six-grid TCA runs were analyzed as a group. The following
equation was fit to these runs (see Figures 5-5 and 5-6):
% Removal = 67 + 0. 02 L + 1. 0 V + 0. 44 Hp - 0.006 ppm (10-4)
The equation accounts for 78 percent of the total variation of the data.
The analysis is restricted to the previously mentioned range of variables,
with the exception of percent solids recirculated, which only varied from
7-10 percent. Note the gas velocity term for this group, which was not
10-5
-------
statistically significant for the entire set of runs (see Equation 10-3).
Again, stoichiometric ratio, hold tank residence time, and scrubber
outlet liquor temperature did not significantly affect SO removal over
LJ
the above ranges.
The 14 EPA, four-grid, three-stage TCA runs were also analyzed as
group. The following equation was fit to these runs (see Figures 5-7
and 5-8):
% Removal = 53 + 0. 04 L + 1.4 PSR - 0. 007 ppm (10-5)
The equation accounts for 96 percent of the total variation of the data.
These test runs were made with five inches of spheres per stage, for a
total of 1 5 inches. The range of variables is otherwise the same as
that for the 31 test run group. Gas velocity, stoichiometric ratio, hold
tank residence time, and scrubber outlet liquor temperature did not
significantly affect SO removal over the above ranges.
The following equation was fit to the 16 runs made without spheres in
the five-grid TCA tower (TVA tests):
% Removal = 90 + 0.034 L - 0. 46 T (10-6)
J_j
The range of variables covered by these runs, and therefore the region
of application, is:
10-6
-------
Gas Flow Rate: 19, 000-30, 000 acfm
Gas Velocity: 7. 5-12 ft/sec
Liquor Flow Rate (L): 375-1,070 gpm
Inlet SO2 Concentration: 2, 200-3, 200 ppm
Stoichiometric Ratio: 1. 0-3. 0 moles CaCO /mole SO- inlet
Percent Solids Recirculated: 14%
Hold Tank Residence Time: 5-15 min.
Scrubber Inlet Liquor Temperature (TL): 63-110 °F
Scrubber Outlet Liquor Temperature: 89-115 °F
Pressure Drop: 1-7 in.
Equation 10-6 accounts for 65 percent of the total variation of the data.
Stoichiometric ratio, inlet SO concentration and hold tank residence
time -were not statistically significant (over the above ranges) in im-
proving the fit of the equation.
10.1.4 Hydro-Filter Scrubber
A stepwise regression analysis of 27 Hydro-Filter runs'"" resulted in
the following equation (see Figures 5-9 and 5-10):
% Removal = 17. 9 + 0. 1 L - 2. 0 V (10-7)
Eleven runs made during October, 1972, were excluded from the
analysis due to doubtful low values of SO2 removal obtained during
this period. Recent closed loop data has validated this exclusion.
10-7
-------
The range of variables covered by these runs, and therefore the region
of application, is:
Gas Flow Rate: 12, 000-30, 000 acfm
Gas Velocity (V): 3-8 ft/sec
Liquor Flow Rate (L): 200-800 gpm
Inlet SO Concentration: 2, 000-3, 500 ppm
Stoichiometric Ratio: 1. 5-3.0 moles CaCG"3/mole SO2 inlet
Percent Solids Recirculated: 6-12%
Hold Tank Residence Time: 50 min.
Scrubber Outlet Liquor Temperature: 85-125 °F
Height of Marbles: 3-5 in.
Equation 10-7 accounts for 94 percent of the total variation of the data.
Inlet SO concentration, stoichiometric ratio, percent solids, liquor
C*
temperature, and height of marbles were not statistically significant
(over the above ranges) in improving the fit of the equation.
10.2 CLOSED-FORM CORRELATIONS FOR PREDICTING SO2
REMOVAL
Analysis of the factorial limestone data, using the Radian Equilibrium
Computer Program (Reference 13), has shown that the equilibrium mole
fraction of SO2 over the bulk liquid, ^ , is negligible with respect to
the SO, mole fraction within the gas for the spray tower, TCA and
*
Hydro-Filter scrubbers. For this condition, Equation 9-10 represents
SO removal.
Due to low liquor residence times, the amount of limestone dissolved
jfc.
within the venturi scrubber is relatively small. Hence, U , can be
significant.
10-8
-------
Also, scrubber computer models using previously fitted gas-side mass
transfer coefficients (see Reference 2) have shown that liquid-side
resistance controls (i. e. , J?L/rr\ « J?§ ) for the spray tower, TCA and
Hydro-Filter scrubbers, and for a majority of the limestone data. For
this condition, Equation 9-10 can be written as:
= \ - exo -\*
The liquid- side coefficient (see Equation 9-8) is expected to be a function
of gas and liquor flow rates, scrubber configuration, amount of dissolved
reactant, interfacial concentration of dissolved SO (H SO )''" and
Li C* O
temperature.
The form of Equation 10-8 has been fitted (by multiple regression) to the
factorial limestone data for the spray tower, TCA and Hydro-Filter
scrubbers. The significance of the independent variables in the fitted
equations was demonstrated by the statistical analysis (see Section 10. 1).
All the factorial limestone data was obtained at relatively high stoichio-
metries (greater than 1.5 moles CaCO /moles SO absorbed), and,
.J £i
consequently, at high scrubber inlet liquor pH's (from 6. 0 to 6. 3). Within
this regime of operation, stoichiometric ratio showed no apparent effect
upon SO removal. The effect of stoichiometric ratio (and scrubber inlet
LJ
liquor pH) for the scrubber systems will, hopefully, be obtained during
As the concentration of 803 in the gas increases, Aft increases and,
consequently, *$L decreases (see Equation 9-8). This is an explana-
tion for the empirical fact that as the SO? inlet gas concentration
increases, for limestone wet-scrubbing systems, the SO? removal
decreases.
10-0
-------
the reliability verification testing now in progress. Other variables which
showed negligible effects upon SO removal during the factorial testing,
LJ
such as percent solids recirculated, may also have more significant ef-
fects at reduced stoichiometries during reliability verification testing.
The effect of inlet gas SO concentration (a non-controlled independent
£.
variable) upon SO? removal has been included only in the fitted equation
for the four-grid, three-stage TCA scrubber, although it is presumed
that a similar effect exists for the other scrubbers. Also, the effect of
inlet scrubber liquor temperature (a non-controlled independent variable),
which was determined to be significant from the TVA TCA runs (see
Equation 10-6), has not presently been included in the closed-form equa-
tions. The effects of inlet gas concentration and temperature will be
included in the final forms of all the correlations when the analyses of
other pilot data and the Shawnee reliability verification data have been
completed.
10.2.1 Spray Tower
The following equation was fit to 1 5 factorial limestone test runs made
with the four-header spray tower (see Figures 5-2 and 5-3 and Equation
10-2):
- 0-031 U/c-- do-9)
where L/^q is in gal/mcf
Equation 10-9 accounts for 94 percent of the variation in the data (cor-
relation coefficient of 0. 97) with a standard error of estimate of 3. 7 per
cent SO_ removal.
10-10
-------
10.2.2 TCA Scrubber
The following equation was fit to 11 factorial limestone test runs made
with the four-grid, three-stage TCA scrubber (see Figures 5-7 and 5-i
and Equation 10-5):"
F ra c i i o A _ \ _ .,. v ,.,
10-10)
where
L, = liquor flow rate per cross-sectional area, gpm/ft^
,^ = SO2 concentration in inlet gas, mole fraction
Equation 10-10 accounts for 99 percent of the variation in the data with a
standard error to estimate of 1. 3 percent SO removal. As previously
L*
mentioned, it is assumed that the measured effect-of VL^ for the TCA
scrubber will be similar for the other systems.
10.2.3 Hydro-Filter Scrubber
The following equation was fit to 27 factorial limestone test runs made
with the Hydro-Filter scrubber (see Figures 5-9 and 5-10 and Equation
~
* Two runs at relatively high weight percent solids and one "limestone
depletion" run were eliminated from this analysis.
4- Eleven runs made during October, 1972, were excluded from the
analysis due to anomalously low values of SO^ removal obtained
during this period. Recent closed-loop data has affirmed this exclusion.
10-11
-------
Equation 10-11 accounts for 95 percent of the variation in the data with
a standard error of estimate of 4. 1 percent SO removal.
L*
10. 3 COMPUTER MODELS FOR PREDICTING SO2 REMOVAL
AND SLURRY COMPOSITIONS
10.3.1 Scrubber Models
In Reference 7, mathematical models were presented for predicting SO
removal in the venturi, TCA and Hydro-Filter scrubbers. The models
are, generally, sets of partial differential equations which describe SO£
absorption into the process liquor (in accordance with the two-film theory),
reaction between the absorbed SO (H SO ) and the species in the liquor
& L, 3
and the dissolution of solids (e.g., CaCO ) within the process liquor.
The assumption has been made, for these systems, that the liquor is at
all times in equilibrium with an interfacial vapor pressure of 0. 1
atmosphere of CO?, i. e. , the rate of absorption of CO from the flue
gas is large. The thermodynamic equilibria for the models are obtained
from the Radian Computer Program (Reference 13).
Computer models have been written for the three scrubber systems,
which numerically solve the systems of differential equations. It has
been planned to fit the gas and liquid-side mass transfer coefficients
to the high and low-concentration soda-ash data (see Section 4) and then
fit the solids dissolution rate constants to the limestone data. The fitting
of the gas-side coefficients for all three scrubbers has been presented
in Section 9. To date, only the liquid-side coefficient for the venturi
scrubber has been fit to the low concentration soda-ash data.
10-12
-------
As discussed in the previous section, for the open-loop (high-stoichio-
metry) data, the equilibrium mole fraction of SO over the bulk liquid
L-i
is essentially zero'1" for the spray tower, TCA and Hydro-Filter scrub-
bers. For this regime, therefore, the models describing SO absorption
for the three scrubbers can be greatly simplified (see Equation 9-10).
For the venturi (spray tower) scrubber, however, the residence time of
the liquor is low (less than 0. 1 second), the dissolution of limestone within
the scrubber is small, and, consequently, the equilibrium mole fraction
of SO over the bulk liquid is not zero everywhere, for the ranges of
variables tested.
Results from the venturi computer model, using the previously fitted
gas and liquid-side mass transfer coefficients, have shown that an as-
sumption of zero dissolution of solids will give a reasonable fit to the
open-loop limestone data.
10. 3. 2 Simulation Model
The simulation model is a computer model which determines the slurry
compositions of the -waste streams and the scrubber inlet and outlet
streams for the three scrubber systems. The major assumption in
the model is: equilibrium occurs between the liquid and solids in the slurry
leaving the effluent hold tank at a specified equilibrium partial pressure of
CO . The equilibrium relationships between the liquid and solid species
LJ
are obtained from the Radian Equilibrium Program.
* This implies that the kinetic rate of dissolution of limestone
within the scrubbers is high.
+ The specified CO2 partial pressure was chosen to match the
measured EHT outlet liquor pH's and compositons. Predictions
with the Radian program indicate relatively constant CO->
equilibrium partial pressures from 0. 05 to 0. 1 atmosphere.
10-13
-------
The simulation model takes as input all of the independent variables
(e. g. , stoichiometric ratio and gas flow rate), the percent sulfite
oxidation, the percent ash in the solids, and the concentration of chloride
and magnesium in the process liquor. If a scrubber model (either
simplified closed-form or computer model) is used, the simulation model
will (iteratively) predict SO removal, as well as the slurry compositions.
LJ
If no scrubber model is used, then SO? removal must be put into the
simulation program along with the independent variables.
Two results from the simulation model are presented here. The first
simulation, for venturi Runs 419-1A and 421-1 A, is shown pictorially
in Figure 10-1 and incorporates the venturi scrubber computer model
(which assumes zero dissolution of solids). The predicted SO_ removal
LJ
of 45% is close to the average measured values of 42 - 5% (see Figure 5-1
and Equation 10-1). The second simulation, for TCA Run 412-2A, is
shown pictorially in Figure 10-2, and does not incorporate a scrubber
model (the measured removal of 96% was input to the program). The
agreement between the predicted and measured scrubber inlet slurry
composition for this TCA simulation is shown in Table 10-1.
* Ultimately, models will be developed for predicting sulfite
oxidation, ash in the solids, and the concentrations of chloride
and magnesium in the liquor.
10-14
-------
rh
ly
s)
JOT Rote = 600 gpm (16,900 Ib-mole/hr)
acfmf 1, 600 Ib-mole/hr) @ 320°F
ratlon= 2,500 ppm MAJOR ASSUMPTIONS:
0=1.5 • Effluent Hold Tank Slurry in equilibrium
f ' with 0.05 arm CO. .
. e CaSOj & CaSOl concentrations in liquo
rculated= 6% .rrea™ leaving E.H.T. are 1 & 1 tir«, t
;harged = 14% equilibrium saturation levels, respective
olids = 40% • Scrubber Slurry (in contact with flue-gal
Idation = 20 % In •qullibrlum with 0. 1 arm COj, i.e.
ui -j .. i /,,™ ,. rate of transfer of CO? is large.
blondes^ 14 gm-mole/1000 liters
lrf - ) ]7°p * 2.*"* dissolution of solids in scrubber
(all dissolution of solids in E.H.T.).
f!«iij.;if H
IliMIIJJjJ
z
^ CM — 1
£ S 2 '
g a- j;
in 5 O S
•* > < O
J O^-"
S S 3 5 j
I-1
jio „„
| i « 0 •»' —
n: ^
O O <*>
/ § o-fs'sso i
|i
UJ^
r
S jjj j||9, g,
f
> -o -; 0 0* CM
_O Ij *" "~
U.
w CM m
G °° ^
s iifjm
M3
i
P_.II
i
CM CM
S8
SI
.1.4lb-moleArCO2
UJ
1
Make-up Water
14 gpm
me Wet-Scrubbing Simulation Pr<
X ^ S
n™g « to
s h
88
ii
0 O
MJ T '
CO
X
a.
J
\
\
2.2 Ib-mole/hr SO 2 1
160 rb-mole/hr CO2
1
^^
fl> ^
>2*?0-CM g^
'o'ofcOCOrv-^O .r-4 fl
g J5 O K CM — W
J3
•" ff "^ "^
Q jf 8 ° °"o ffl >
— O 5 U »o 10 U
10 ooo'o'o(o^4 •
1
o
i— H
^
GO
10-15
-------
.1 Mlf-
* -C-HI-:
1 j=hfi
•£ i»1{ii
t- I o S £ ° ~
^. § g _•-£-.£
SSo-sH **§
E J8 £«J ~ i °
1 sii-I8 hi
l^-'jpl
oj^siiin
<£l«3lf^.£8
* • e e
KM- Rate =1, 180 gpm (32,700 Ib-mole/hr)
acfm( 3,000 Ib-mole/hr) @ 290 °F
ration = 3,000 ppm
o = 1.25
;f
rculated = 8.5%
harged = 18%
*lids = 30%
idation = 30 %
hlorides = 42.5 gm -mole/1 000 liten
re = 120°F
i§§HJi-:^n
«£S«M:S-fis&
tio-r3?!5!1;
ii * 1 ,; n i § 1 1
Sc
Ga
>02 REMOVAL = 96 % ,n|
Comments: c»,
SIMULATION OF TCA SCRUBBER SYSTEM RUN l/<
412-2A ON OCTOBER 29-31 , 1972. pel
REAL SCRUBBER MODEL NOT USED IN SIMULATION pe
(SO REMOVAL INPUT TO PROGRAM.) pe(
Pel
Co
Scr
h
Is
j
,/
_i
i5
56;
s^;
1
3£™. '°. ° "°. "*.&
*"| ^0^' -OtNO™
n: a
X g X n —
M
JTJ"* -Ots«og
jla«--=aj
§ JfSSg'D
u; s s S S o £0
CO
^O
II
r
o.
s
^ s
33
"S
p— -
*
Ib-moleAr CO2
Not Computed
L2
3
r*" "
o o i
SJ
ii i
? p 1
j) j E
no (3
= 1 '
u
i
0.
/ 1
•«— n
I
C-l' CN W CN
o o o o
-------
Table 10-1
COMPARISON OF MEASURED AND PREDICTED SLURRY
COMPOSITIONS AT SCRUBBER INLET FOR TCA RUN 412-2A
Gas Rate = 27, 500 acfm
Liquor Rate = 1, 170 gpm
L/G = 5? gal/mcf
Pressure Drop = 14 in. t^O
Three stages, 5 inches/stage
Species
PH
SO/
CO/
so/
Ca++
Mg+ +
Cl~
Species Concentrations, gm mole/1
Liquid
Measured I
5.9
1.8
1.2
24
35
5.5
43
Predicted
6.
2.
2.
1
3
1
7
9
43
0*
000 liters
Solid
Measured
—
210
220
86
510
21
1 Predicted
—
300
120
120
550
0
43* - -
Input to computer model.
10-17
-------
Section 11
REFERENCES
1. H. W. Elder, et al, "Operability and Reliability of the EPA Lime/
Limestone Scrubbing Test Facility, " presented at Flue Gas Desul-
furization Symposium, New Orleans, Louisiana, May 14-17, 1973.
2. F. T. Princiotta and M. Epstein, "Operating Experience with a
Prototype Lime-Lime stone Scrubbing Test Facility," presented at
the Sixty-Fifth Annual Meeting of the A. I. Ch. E. , New York City,
November 26-30, 1972.
3. M. Epstein, et al. , "Test Program for the EPA Alkali Scrubbing
Test Facility at the Shawnee Power Plant, " presented at Second In-
ternational Lime/Limestone Wet Scrubbing Symposium, New Orleans,
Louisiana, November 8-12, 1971.
4. Universal Oil Products, Air Correction Division, Bulletin No. 608,
"UOP Wet Scrubbers, " 1967.
5. National Dust Collector Corporation, General Catalog, December 23,
1968.
6. M. Epstein, et al. , "Test Results from the EPA Lime/Lime stone
Scrubbing Test Facility, " presented at the Flue Gas Desulfurization
Symposium, New Orleans, Louisiana, May 14-17, 1973.
7. M. Epstein, et al. , "Mathematical Models for Pressure Drop, Par-
ticulate Removal and SO_ Removal in Venturi, TCA and Hydro-Filter
Scrubbers, " presented at Second International Lime/Limestone Wet
Scrubbing Symposium, New Orleans, Louisiana, November 8-12,
1971.
11-1
-------
8. F. T. Princiotta and N. Kaplan, "Control of Sulfur Oxide Pollution
from Power Plants, " presented at EASCON, Washington, D. C. ,
October 18, 1972.
9. J. M. Potts, et al. , "Removal of Sulfur Dioxide from Stack Gases by
Scrubbing with Limestone Slurry: Small Scale Studies at TVA, " pre-
sented at Second International Lime/Lime stone Wet Scrubbing Sym-
posium, New Orleans, Louisiana, November 8-12, 1971.
10. A. Saleem, et al. , "Sulfur Dioxide Removal by Limestone Slurry in
a Spray Tower, " ibid.
11. A. D. Little, Inc., Evaluation of Problems Related to Scaling in
Limestone Wet Scrubbing, EPA Report, April 1973.
12. R. H. Borgwardt, Limestone Scrubbing at EPA Pilot Plant, Prog-
ress Report No. 6, EPA Report, January 1973.
13. Radian Corporation, A Theoretical Description of the Limestone-
Injection Wet Scrubbing Process^ NAPCA Report, June 9, 1970.
14. A. Saleem, J. Air Pollution Control Assoc. , Vol. 22, No. 3,
March 1972.
15. R. H. Borgwardt, Limestone Scrubbing at EPA Pilot Plant, Prog-
ress Report No. 8, EPA Report, March 1973.
16. R. H. Boll and C. A. Leeman, "Particle Collection and Pressure
Drop in Venturi Scrubbers, " AIChE 69th National Meeting, Cincin-
nati, Ohio, May 16-19, 1971.
17. B. P. Volgin, et al. , Int'l Chem. Eng. , Vol. 8, No. 1, p. 113,
1968.
18. J. Happel, Ind. Eng. Chem., Vol. 41, 1949, p. 1161.
11-2
-------
19. M. Leva, Fluidization, McGraw Hill, New York, 1959.
20. S. Nukiyama and Y. Tanasawa, Trans. Soc. Mech. Engrs., (Japan),
Vol. 4, No. 14, p. 86, 1938.
21. Bechtel Corporation, Alkali Scrubbing Test Facility - Progress Re-
port: Mathematical Models for Venturi Scrubber and After-Scrubbers,
APCO Report, February 1971.
22. J. R. Fair, Petro. /Chem. Eng. , Vol. 33, No. 9, p. 57, August
1961.
23. Bechtel Corporation, Alkali Scrubbing Test Facility - Progress Re-
port: Mathematical Models for Hydro-Filter Scrubber, APCO Re-
port, June 1971.
24. A. E. Handles and T. Baron, AIChE Journal, Vol. 3, No. 1, March
1957.
11-3
-------
Appendix A
CONVERTING UNITS OF MEASURE
Environmental Protection Agency policy is to express all measurements
in Agency documents in metric units. When implementing this practice
will result in undue costs or lack of clarity, conversion factors are
provided for the non-metric units used in the report. Generally, this
report uses British units of measure. For conversion to the metric
system, use the following conversions:
To Convert From
acfm
cfm
°F
ft
ft/hr
sec
To
f\
ft /tons per day
gal/mcf
gpm
gpm/ft^
g r / s cf
in.
in. H2O
Ib
Ib- moles
Ib-moles/hr
Ib-moles/hr ft2
Ib -moles /min
nm /hr
m3/hr
m
m/hr
m/sec
m2
m /metric tons
per day
//m3
^€/min
-£/min/m2
cm
mm Hg
gm
gm-moles
gm- moles /min
gm-moles /min/m2
gm-moles/sec
Multiply By
1.70
1.70
subtract 32 then
-7-1.8
0.305
0.305
0.305
0.0929
0. 102
0. 134
3.79
40.8
2.29
2.54
1.87
454
454
7.56
81.4
7.56
A-l
-------
Appendix B
CORRECTION FACTOR FOR SO2 REMOVAL DUE TO
DILUTION EFFECT OF REHEATER GAS AND WATER VAPOR
Since the flue gas picks up water vapor and reheater gas before reaching
the outlet DuPont SO analyzers, a correction factor should be used to
Lj
account for this dilution effect. It has been found that the correction
factor is a sensitive function of the "scrubber" temperature (and hence
the amount of water vapor pickup). The amount of reheater gas, of
course, has a direct effect on the values of the correction factor.
The following equation relates the "corrected" SO removal to the
LJ
"uncorrected" SO,, removal:
(B-l)
(B-2)
where
- - measured SO2 concentrations at the
scrubber inlet and reheater outlet,
respectively
-r - (?£./&, = correction factor
&, t &&, - molar flow rates per unit time of
total flue gas at the scrubber inlet and
reheater outlet, respectively
Table B-l gives the correction factors,/, to be applied for the SO? re-
moval at different scrubber pressures and temperatures. The assumptions
B-l
-------
used to arrive at the table values are listed in the notes for the table.
Figure B-l gives a convenient chart for correction of SC>2 removal based
on Equations B-l and B-2, once the correction factor has been deter-
mined. Any amount of the air leakage into the scrubber system can be
simply added to the correction factor as a part of the reheater gas.
Table B-l
CORRECTION FACTORS FOR SO REMOVAL
L-i
Scrubber
pressure,
inches W. G.
-10
-20
-30
Correction factor, f (±0.015)
Scrubber temperature, °F
120
1.088
1.092
1.095
122
1.096
1. 100
1. 104
124
1. 105
1. 109
1. 114
126
1. 114
1. 119
1. 124
128
1. 124
1. 129
1. 134
130
1. 134
1. 139
1. 144
Assumptions:
Flue gas at scrubber inlet contains 8 percent by
volume of moisture.
Flue gas at scrubber outlet is saturated with water.
Reheater gas molar flow rate is 4. 5 percent (± 1.5)
of flue gas at scrubber inlet (= 0. 045 G^).
Flue gas mole change due to absorption (or desorption)
of SO2 and other inert gases in the scrubber is neglected.
Note that the correction is insignificant when the SO removal is high
L*
(above approximately 85 percent). At low SO? removal, however, the
correction can be substantial. The SO-, removal efficiencies reported
for flue gas in this topical report have been based upon a correction
factor of 1.11.
B-2
-------
100
90
80
70
3 60
Ul
ex
CS|
o
in
& 50
u
114
O
u 40
30
20
1.11
1.15
07
30 40 50 60 70 80 90 100
UNCORRECTED SC>2 REMOVAL, %
Figure B-l. Correction of SC>2 Removal for Water Vapor and Reheater
Gas Pick-up
B-3
-------
Appendix C
DU PONT CALIBRATION CURVES AND CORRECTION FACTORS
C-l
-------
DV-1011 REV. t-70
E. 1. DU PONT DE NEMOURS & COMPANY
INCORPORATED
WILMINGTON, DELAWARE 19898
INSTRUMENT PRODUCTS DIVISION
February 5, 1973
Mr. Louis Sybert
TVA Shawnee Steam Plant
P. 0. Box 2000
Paducah, Kentucky 42001
Subject: Calibration of Du Pont 460 Analyzer Systems at
Shawnee Steam Plant, Wet Scrubber Project, Tag
No's AE1001, AE2001, AE3001, AE1Q20, AE2020 and
AE3020
Dear Mr. Sybert:
This is an interim letter to provide you the calibration
curves and daily correction factors for the S02 analyses by
the Du Pont 460 analyzer systems (from late June to December,
1972). A more complete report of my investigation of the
analysis problems will follow shortly.
The attached graph contains the calibration curves for
each analyzer with the interference filter used during the
June-December period.
Tables I through VI provide the daily correction factors
based on my review of the instrument log sheets kept by your
personnel on these analyzer systems. These correction factors
assume the analyzers were operating at 220°F and 3" Hg vacuum.
The recorded analysis data may be corrected as follows:
1) From the attached calibration curve for the analyzer
providing the data being reviewed, determine the
"base" SO2 concentration.
2) Multiply the "base" concentration by the "x" factor
for the analyzer on the given day.
For example, for Analyzer AE2020, on October 17 (10/17),
4:30 p.m., let us say the analyzer logged 1000 ppm S02- From
the calibration curve, the base concentration would be 1160 ppm
SO2. From Table V, the x-factor would be 0.875 from 3:30 until 9:00
p.m. on 10/17. Therefore, the corrected concentration would be
0.875 x 1160 = 1015 ppm.
C-3
BETTER THINGS FOR BETTER LIVING . . . THROUGH CHEMISTRY
-------
Louis Sybert - 2 - February 5, 1973
As may be seen from the above example, the x-factor at
a given time, is the number in the table recorded at the
nearest previous time.
I hope that the data provided in this letter will be of
help and that I will be able to send the complete report
within a week.
If you have any questions, please call me at (302) 453-2740,
I will be away from the office the week of February 12.
Very truly yours,
Robert S. Saltzman
Senior Applications Engineer
RSS/psm
Attachments
cc: John Reese
Phil Stone
John Williams
\Charles Levio
C-4
-------
CALIBRATION CURVES FOR 460 S02 ANALYZERS
3200
3000 ••
VENTURI INLET
TCA INLET
HYDRO-FILTER INLET
VENTURI OUTLET
TCA OUTLET
HYDRO-FILTER OUTLET
[Redrawn by Bechtel on 2/28/73]
200 400 600
1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 3000
"BASE" SO2 CONCENTRATION (PPM)
C-5
-------
DU
TABLE I
CALIBRATION FACTORS FOR
PONT S02 ANALYZER NO. 2757
TAG AE 1001
Date
7/7
7/8
7/9
7/10
7/11
7/13
7/14
7/15
7/16
7/17
7/18
7/19
7/21
7/22
7/23
7/24
7/25
7/26
7/27
7/28
7/29
7/30
7/31
X
Factor
1.23
1.23
1.23
1.23
1.23
1.22
1.21
1.22
1.22
-
1.21
1.22
1.22
1.21
1.22
1.21
1.19
1.18
1.20
1.19
1.19
1.18
1.17
Date
9/14
9/16
9/17
9/21
9/22
9/23
9/24
10/1
10/2
10/3
10/4
10/5
10/6
10/7
10/8
10/9
10/10
10/11
10/12
10/13
10/14
10/15
10/16
X
Factor Date
1.16
1.04
1.02
1.13
1.13
1.12
1.12
1.14
1.04
1.00
1.05
1.01
1.02
1.03
1.11
1.05
1.11
1.07
1.15
1.09
1.13
1.13
1.14
1.14
10/18
10/19
,1.06 (3:30p) 10/20
10/21
10/22
10/23
10/24
,1.00 (3p) 10/25
10/26
10/27
,1.02 (2p) 10/28
10/29
10/30
(3p)1.06 (9p) 10/31
,1.06 (3:30p) 11/8
11/9
,1.05 (4p) 11/10
,1.15 'C4:30p) 11/11
,1.10 (lOp) 11/12
(5:30p) 11/14
(9;30p)
11/15
11/16
,1.09 (3:30p)
V /
NT
X
Factor
1.13
1.13
1.09
1.08,1.06
1.13
1.11
1.15,1.09
1.12,1.07
1.10
1.11,1.13
1.11,1.09
1.07
1.07,1.02
1.05,1.03
1.04,1.07
1.06
1.05
0.896
0.87
0.891
0.88
0.896
'i
Jy
V
(3:40p)
(9:30p)
(2p)
(9p)
(6p)
(9p)
(3p)
(10p)
C-6
-------
TABLE II
CALIBRATION FACTORS FOR
DU PONT SO2 ANALYZER NO. 2758
TAG AE 2001
Date
6/23
6/24
6/25
6/30
7/1
7/2
7/3
7/7
7/9
7/10
7/11
7/13
7/14
7/20
7/21
7/22
7/23
7/24
7/25
7/26
7/29
7/30
7/31
8/17
8/18
10/8
10/9
X
Factor
1.23
1.22 (3:30p)1.19 (9:35p)
1.24,1.19 (l:30p) .99(5p)
1.17
1.16,1.15 (9:30p)
1.15
1.15
1.14
1.14
1.12 (2:30p)
1.08
1.1
1.09
1.03
1.04
1.04
1.04
1.03
1.02
1.0
.98,1.0
1.03
1.04
.78 (3p)
.78
1.10
1.08 (8:30p)
Date
10/10
10/11
10/13
10/14
10/15
10/16
10/17
10/18
10/19
10/20
10/21
10/22
10/23
10/25
10/26
10/27
11/16
12/1
X
Factor
1.09,1.06 (4p)
1.09
1.08
1.08
1.08
1.06 (5:45p)1.07 (9:30p)
1.07,1,08 (3:30p)
1.1 (3:30p)
1.13
1.02
1.08
1.07,1.08 (3:30p)
1.08,1.07 (ll:15p)
1.10 (9:30p)
1.19 (9:30p)
1.08
1.08
1.11
1.19
C-7
-------
TABLE III
CALIBRATION FACTORS FOR
DU PONT SO2 ANALYZER NO. 2759
TAG AE 3001
Date
6/30
7/1
7/2
7/3
7/15
7/19
7/20
7/21
7/23
7/24
7/25
7/26
7/27
7/29
7/30
7/31-
8/18
8/23
8/24
X
Factor
1
1
1
1
1
1
1
1
1
1
1
• 1
1
1
1
1
1
•
1
.23
.2,1.19
(4:30p)
.24
.23
.19
.11
.11
.14
.13
.13
.14
.13
.12
.12
.10
.11
.12
99
.09
Date
8/28
9/7
9/8
9/9
9/10
9/22
9/23
9/24
9/26
9/28
9/29
9/30
10/1
10/2
10/3
10/4
10/5
10/6
10/7
X
Factor
.99
.07
.98
1.13
1.12
1.11
.99,1
0.995
.99
1.09
1.09,
.96,.
.99,.
.95
.975,
1.09,
1.09,
0.97
.965,
.97,.
.98
.98,.
.09
,.99
1.11
955
(3:30p)
(3p)
(9P)
(9:30p)
Date
10/8
10/9
10/10
10/11
10/14
10/15
10/16
10/17
10/18
10/19
10/20
10/21
98 (3p)
X
Factor
1.15,
1.10
1.12,
.995,
.94,.
.98
.99,1
1.09,
1.14,
1.13,
1.12
1.15,
1.11,
1.10
.935
(3
:30p)
(9p)
.99
.98
965
.10
.98
1.11
1.14
1.11
1.09
(4p)
(4:
(9:
(5:
(3
(2
(9
(3
30p)
30p)
45p)
:30p)
:15p)
:30p)
:40p)
(9p)
(9:15p)
.95
.98
1.11
(9p)
(9p)
'(2p)
(9p)
.96
995
(9p)
975
(9p)
(3p)
(3p)
10/22
10/23
10/24
10/25
10/26
11/10
.!
V
1.10,
1.11,
1.13
1.14,
1.11
1.14
I
J,
Y
1.09
1.12
1.11
(3
(9
:30p)
:30p)
(4p)
.94 (9p)
-------
TABLE II
CALIBRATION FACTORS FOR
DU PONT SO2 ANALYZER NO. 2758
TAG AE 2001
Date
6/23
6/24
6/25
6/30
7/1
7/2
7/3
7/7
7/9
7/10
7/11
7/13
7/14
7/20
7/21
7/22
7/23
7/24
7/25
7/26
7/29
7/30
7/31
8/17
8/18
10/8
10/9
X
Factor
1.23
1.22 (3:30p)1.19 (9:35p)
1.24,1. 19 (l:30p).99(5p)
1.17
1.16,1.15 (9:30p)
1.15
1.15
1.14
1.14
1.12 (2:30p)
1.08
1.1
1.09
1.03
1.04
1.04
1.04
1.03
1.02
1.0
.98,1.0
1.03
1.04
.78 (3p)
.78
1.10
1.08 (8:30p)
Date
10/10
10/11
10/13
10/14
10/15
10/16
10/17
10/18
10/19
10/20
10/21
10/22
10/23
10/25
10/26
10/27
11/16
12/1
X
Factor
1.09,1.06 (4p)
1.09
1.08
1.08
1.08
1.06 (5:45p)1.07 (9:30p)
1.07,1,08 (3:30p)
1.1 (3:30p)
1.13
1.02
1.08
1.07,1.08 (3:30p)
1.08,1.07 (ll:15p)
1.10 (9:30p)
1.19 (9:30p)
1.08
1.08
1.11
1.19
C-7
-------
TABLE III
CALIBRATION FACTORS FOR
DU PONT SO2 ANALYZER NO. 2759
TAG AE 3001
Date
6/30
7/1
7/2
7/3
7/15
7/19
7/20
7/21
7/23
7/24
7/25
7/26
7/27
7/29
7/30
7/31-
8/18
8/23
8/24
X
Factor
1.23
1.2,1.19
(4:30p)
1.24
1.23
1.19
1.11
1.11
1.14
1.13
1.13
1.14
' 1.13
•
1.12
1.12
1.10
1.11
1.12
.99
1.09
Date
8/28
9/7
9/8
9/9
9/10
9/22
9/23
9/24
9/26
9/28
9/29
9/30
10/1
10/2
10/3
10/4
10/5
10/6
X
Factor
.99
.07
.98
1.13
1.12
1.11
.99,1.09 (3:30p)
0.995, .99 (3p)
.99
1.09
1.09,1.11 (9p)
.96, .955 (9:30p)
.99, .98 (3p)
.95 (9:15p)
.975, .95 (9p)
1.09, .98 (9p)
1.09,1.11 (2p)
0.97 (9p)
.965, .96 (9p)
.97, .995 (3p)
.98 (9p)
Date
10/8
10/9
10/10
10/11
10/14
10/15
10/16
10/17
10/18
10/19
10/20
10/21
10/22
10/23
10/24
10/25
10/26
11/10
X
Factor
1.15, .935 (3:30p)
1.10 (9p)
1.12, .99
.995,. 98 (4p)
.94, .965 (4:30p)
.98
.99,1.10 (9:30p)
1.09, .98 (5:45p)
1.14,1.11 (3:30p)
1.13,1.14 (2:15p)
1.12
1.15,1.11 (9:30p)
1.11,1.09 (3:40p)
1.10 (9p)
1.10,1.09 (3:30p)
1.11,1.12 (9:30p)
1.13
1.14,1.11 (4p)
1.11
1.14
1
10/7 .98,.975 (3p)
.94 (9p)
Co
- O
-------
TABLE IV
CALIBRATION FACTORS FOR
DU PONT SO2 ANALYZER NO. 2760
TAG AE 1020
x
Date Factor
9/17 .975
9/21 .99
9/22 1.04
9/23 1.01
9/24 0.975
9/26 .965
10/1 1.04
10/3 0.965
10/4 0.965
10/5 0.98
10/6 0.91,0.995 (9p)
JO/7 0.98,1.02 (9p)
10/8 1.07
10/9 0.98,0.973 (8:30p)
10/10 1.0
10/11 1.0,0.98 (4:30p)
10/12 0.98,1.01 (7:30p)
10/13 0.675 (5p)
10/14 0.664/0.658 (9p)
Date
10/15
10/16
10/17
10/18
10/19
10/20
10/21
10/22
10/23
10/25
10/26
10/27
10/28
10/29
10/30
10/31
11/8
11/10
x
Factor
0.685/0.67 (3:30p)
0.654
0.696 (9p)
0.67(l:40p) 0.83(6.20p)
0.72 (9p)
0.73 (4:30p)
0.73,0.70 (9:30p)
0.72,0.707 (9:30p)
0.688,0.674 (3:30p)
0.68 (9:30p)
0.673,0.695 (4p)
0.696 (9:30p)
0.692,0.715 (4:30p)
0.688 (9:30p)
0.77,0.66 (4:05p)
.695 (9:30p)
0.633 (4p).079 (6p)
0.73,0.81 (9p)
0.70
0.778
0.72
0.735
0.73
0.715
C-9
-------
TABLE V
CALIBRATION FACTORS FOR
DU PONT SO2 ANALYZER NO. 2761
TAG AE 2020
Date
6/23
6/24
6/25
6/30
7/1
7/2
7/3
7/7
7/8
7/9
7/10
7/11
7/13
7/14
7/15
7/20
7/21
7/22
7/23
7/24
7/25
7/26
7/29
7/30
7/31
8/17
8/18
X
Factor
1.15
1.07,1.02 (Ip)
1.00 (3:30p)
1.18,1. 17 (4:45p)
1.01
1.01
1.01
1.20 (3p)
1.02
1.01
1.00
1.01
1.02,1.40(6:15p)
1.00
1.19
1.20
1.01 (5p)
1.20
1.19
1.16
1.12
1.12
1.13
1.15,1.21
1.00
1.16
1.12
1.15
Date
8/19
10/8
10/9
10/10
10/11
10/13
10/13
10/14
10/15
10/16
10/17
10/18
10/19
10/20
10/21
10/22
10/23
10/25
10/26
10/27
10/28
10/29
10/30
10/31
11/1
11/3
X
Factor
1.12
1.25
1.25
1.30,1.21 (4p)
1.27,1.21 (7:30p)
0.83 (5p)
0.83
0.86,0.85 (4:30p)
0.88 (9:15p)
,
0.875,0.87 (3:30p)
0.875 (9:30p)
0.83,0.825 (9:30p)
0.86,0.875 (3:30p)
0.86 (9p)
0.87 (6:20p)
0.875 (4:30p)0.87 (9:30p)
0.875,0.825 (9:30p)
0.825,0.79 (3:40p)
.79,0.765 (9:30p)
0.77
0.80 (9:30p)
0.77,0.78
0.85,0.78 (2p)
0.79
0.79
0.79
0.76,0.795 (3p)
0.74,0.76 (2:45p)
0.87
C-10
-------
TABLE VI
CALIBRATION FACTORS FOR
DU PONT S02 ANALYZER NO. 2762
TAG AE 3020
Date
6/30
7/1
7/2
7/3
7/19
7/20
7/21
7/22
7/23
7/24
7/25
7/26
7/27
7/29
7/30
7/31
8/18
8/24
8/28
9/7
9/8
9/9
9/23
9/24
9/26
9/27
X
Factor
1.15
1.13,1.15
1.18
1.20
1.15
1.15
1.12
1.12
1.15
1.14
1.13
1.11
1.13
1.13
1.14
1.14
1.10
1.05
1.12
1.12
1.12
1.12
1.46,1.49
1.42,1.36
1.45
1.44
Date
9/28
(4:30p) 9/29
10/1
10/2
10/3
10/4
10/5
10/6
10/7
10/8
10/9
10/10
10/11
10/13
10/14
10/15
10/16
10/17
10/18
10/19
10/20
10/21
(3:30p)
10/24
(3:30p)1.46(9:30p)
10/23
10/24
10/25
X
Factor
1.46
1.43,1.47 (3p)
1.49,1.46 (3p)
1.37,1.64 (4p)
1.49,1.52 (9p)
1.42,1.44 (2p)1.51
1.49,1.46 (3p)1.50
1.43,1.46(3p) 1.45
1.46,1.45 (3p)
1.44,1.45 (3p)
1.52 (8:30p)
1.52,1.71 (4p)
1.58,1.51 (4:30p)
1.03
1. 01(11. 30p) 1.03
1.04 (9:15p)
1.08,1.12 (3:30p)
1.08,1.10 (9:30p)
1.06,1.07(3:30p) 1
1.00,1.05 (9:15p)
1.05
1.07,1.03 (9:30p)
1.05,1.08(3:40p) 1
1.08,1.06(3:30p) 1
1.07,1.03 (9:30p)
1.05
1.04,1.05 (4p)
(9p)
(9p)
(9p)
(4:30p)
1.09(9:30pJ
.09 (9p)
.05(9:30p)
.03(9:30p)
C-ll
-------
TABLE VI
(Cont'd)
x
Date Factor
10/26 1.03(4p) 1.05(6p)
10/27 1.06,0.97(2:10p) 0.95 (9p)
10/28 . 1.06,0.96(5:45p) 1.03(9:30p)
10/29 1.01,1.02(4:30p) 1.04(9:30p)
10/30 1.05
10/31 1.04,1.05 (3p)
11/1 1.03(3:15p) l.OO(lOp)
11/3 1.03
C-12
-------
Appendix D
WATER BALANCES FOR SCRUBBER SYSTEMS DURING
CLOSED-LOOP LIMESTONE TESTING
Water balances for closed-loop limestone tests have been made for the
three scrubber systems utilizing clarifiers for solids separation.
Tables D-l, D-2, and D-3 give the results of the water balance calcula-
tions for inlet flue gas flow rates of 10,000, 20,000, and 30,000 acfm,
respectively, with 6, 10, and 15 wt % solids in the slurry bleed streams
(clarifier feeds). The solids concentration in the clarifier underflows
are calculated from the following equation:
UA =
where,
UA = clarifier unit area, ft /tons per day of solids (TPD)
F, D = clarifier feed and discharge (underflow) dilutions,
respectively, Ibs water/lb solids
R = solids free settling rate (lime settling) at feed
dilution F, ft/hr
The solids free settling rates, R, at different initial (feed) weight per-
cent solids are shown in Figure D-l. The curve in Figure D-l is ob-
tained from batch settling tests in a graduated cylinder performed by
D-l
-------
P
O
O ^H ^H Q)
«
• H
1
0
T3
ffi
O
H
'h
rj
4J
>
Scrubber System
O rO
f^ 00 i— < O^
CO ro
o . r-
0s 0s O PJ
Is- t"
C- CO
o • *^ •
r- r- ^H o
rO Tf
Q
&
V.O ^^-
ro
Assumed SC>2 Removal,
Waste Solids, TPD
Clarifier Area, ft. ^
Unit Area Available, ft.
in
m o oo
- •* (M
«n
^H T(* CM f^j
m m
vO O . .
N N
in o o °
„
o
o o m
o
•* CM ^
Tf
in o oo
-i •* ;M
O O CM "*
-< •* ^ pj
m
\o o .
"^ ° fM
(M
•0
^^
* 6 M
Wt. % Solids in Clarifie
Wt. % Solids in Underfl
Feed Flow Rate, gpm
Underflow Slurry, gpm
O "H
CM CM
O -1
CM CM
O -H
. +1
N CM
t+7
CM CM
. +i
CM CM
. + 1
CM CM
cr- rt
. +1
-H CM
. +1
-H CM
cr I-H
. +i
—i CM
ro
CO
nj
CM" O
Water Output, gpm:
Clarifier Underflow'
Evaporation to Flue
0 -H
I .
0 CM CM
Tf O -H
O CM CM
^* O »— i
. •+!
O CM (M
. o+7
0 Tf
. o+T
O ^*
• O-H*
O ^
. o+7
O "^
. 0-H
O Tf
M.
. o+7
O ^
at
^,
H
n ^
*n o
Water Input, gpm:
Limestone Feed
Pump Seal Water
Total Makeup Water
Demister Wash, Kc
etc.
CO
•
rt
£
'S
o
a
OH
nl
at
cu
00
at
CJ
CO
•3
1
a
H-»
rd
CO
cu
at
.
?
O
T*
r^J
d
0
u
3
CO
OH
s
OH
[J
_0
cu
1
-p
ro
h
o
00
CM
1-1
m
t-t
<«H
0
n
a>
fH
2
nt
h
V
a
g
5
>.
^1
h
to
1
•§
h
U
CQ
fH
O
1
U
at
0
o
o
0
d
a.
bo
ro
-
e
o
fH
4-1
ca
CU
C
nt
CO
rt
BC
cu
3
v!
Q
+1
. C
c .2
O 4J
£ rt
at at
^ (U
cu
rd cu
* z
ro
J>,
0)
• rH
U
a
CO
a>
D-2
-------
O
M
i
Q
O
ft) 41 f<
33^
A
44
C
0
T3
K
g
h
._|
»H
1
ft)
>
Scrubber System
o \o
rt C^ ^H
S "^ t t
(f.
^
*G O CO .
„
* ^ -*
mO..
*"" ' ^ 0s vO
0 0 0 *7
"•< "* CO „£>
NO O . .
^ .-) x£>
in
in oo "^ ^
«-* ro in t-
O vQ ^
rt CM CM OT
00
vO O i-H
** M1 ^
•d
4> —
(U _,
h ~fc
H§ CM
Wt. % Solids in Clarific
Wt. % Solids in Underfl
Feed Flow Rate, gpm
Underflow Slurry, gpm
m CM
t-llj
CO CM
00 4
O CM
.+1
O" CM
. +1
O CM
•*<*
0s CM
.-H
sD CM
.+1
Tf CM
i |
t- *f
00 CM
.+1
ro Tf
m
m
_ nl
CM O
Water Output, gpm:
Clarifier Underflow'
Evaporation to Flue
o-o N
d CM ef»
°~* ° CM
• • ~H
O CM CJ.
CJ* O (\j
• • "tl
O CM ID
rt CM
. O -H
rt 00
rt CM
rt 00
rt CM
. 0 -H
rt CO
CM
. Q ~H
0 rt
CO Jjj
0 ~
CO CM
• o +1
o T-
S •*
•8 8
Water Input, gpm:
Limestone Feed
Pump Seal Water
Total Makeup Water
Demister Wash, K<
Tray, etc.
«
o>
o
5
^~
ll
0
rH
rt
'x
0
a
rt
CD
.H
C
o
u
-u
S
U
a
o
U
•S
f— 1
O
01
O
1— (
tn
a>
-o
S
p
4)
nl
a
'rt
^
o
fl
H
S
*w
2
^
,_,
rt
0)
a
"S
a>
a
rt
0
o
S
Ji
2
g
a
00
CM
^
roximate:
a
8-
^
rt
rt
S 4)
•u rt
— •
(*
0
00
CM
i-i
I
in
rt
<«4
O
a
4)
h
4^
rt
h
ft)
a
fi
5
!>.
IH
_3
{0
ic rubber
w
»H
ti
»M
U
nJ
0
0
O
*.
0
r— 4
fi
bo
CO
i
rH
S
o
to
ft)
00
rt
to
rt
M
Q)
3
d
O
4J
S
o
4J
rt
O
P<
rt
U
-~
J^
rt
o
.{!H
0
ft)
a
D)
ft)
D-3
-------
CO
I
Q
•8
H
M
4)
+1
P— I
• H
1
Q
h
•o
£
O
H
'n
5
g
o
1 Scrubber System
O -H
\T) • ^ .
0 . ^ .
O o O -^i
co ^ ro
rf sO
O . "^ .
t^ CO ^ to
N m — i
Q
rH
Assumed SO2 Removal, %
Waste Solids, TPD
Clarifier Area, ft. 2
Unit Area Available, ft. 2/T
*
»-l (M Cd ^O
1— t
if) in
O "^O • •
,H -H f- N
CO f\J
vO O Tj*
•^ -£• t^t
in oo
t— i CO 0s i— 1
CJ ~H
in co
o t>-
i—t to in o
Tf -1
in ^
vO O . .
•* r- o
in o
m 1-1
*-H oo to m
ro -H
CO
•-I —I CO O
CO
vO 0 "? ™
SO
•o
. ^
^ ~^~'
1—1
^
* ^
00 "^
i-H
l> ro
. +i
if) xO
CO
Water Output, gpm:
Clarifier Underflow^2)
Evaporation to Flue Gas^
. 0
« ft
* 2-
6 rt
• r3 <0
O n)
&ff
"* rt
co J3
™* o
r< CO
0 ^
rt y*
In g
1 g.
(J .jj
C tS
° %
cc -*-*
r^ flj
° &
» 0
S. *"*
o
*w .
4) Q
TJ -^
§ g
4J CO
3 P.
S §
'3 p.
"tj I
o g
S "**
2 3
.5 c
M 3
h
O
00
CM
i— i
1
in
1-1
t**
0
m
O
It
S
a
In
IU
a
fi
V
-4-»
rr
3
CO
h
^
3
^
CO
V,
S
U
m
o
o
o
0
"a
a
bfi
m
^_)
S
o
in
CO
IU
C
n)
t-i
ta
at
bn
IU
3
H-t
O
c 2
O -4J
S 2
•o o
nl a
tu "
•
j?
>
4J
0
a)
Pu
a
.*-»
o
Z
rt cy
-------
3.0
1
2.0
o
z
1.0
TERMINAL VELOCITY
(VELOCITY AT INFINITE
DILUTION) FOR 74 JU
(200 MESH) PARTICLE £
60FT/HR
BATCH SETTLING IN GRADUATED CYLINDERS WITHOUT.
COAGULANT AT ROOM TEMPERATURE.
O
D
»
DORR-OLIVER DATA (1/22/73)
TVA DATA (11/8/72)
SOLIDS FROM HYDRO-FILTER CLARIFIER
UNDERFLOW (RUN 501-3A, 3/18/73)
13
14
56 78 9 10 11 12
INITIAL (FEED) WEIGHT PERCENT SOLIDS
Figure D-l. Free Settling Rates of Shawnee Clarifier Feed Solids
15
D-5
-------
TVA and Dorr-Oliver using Shawnee clarifier solids. A data point for
solids from Hydro-Filter clarifier underflow during Run 501-3A (approx-
imately 10% sulfite oxidation in solids) is also included in the figure.
The data shown in Figure D-l is obtained from tests conducted at room
temperatures. At high clarifier operating temperatures (about 100°F),
the settling rates would be slightly higher than the figure indicates be-
cause of lower liquor viscosity. Also, because of the lower slurry pH
(about 5. 7 to 5. 9) operation intended for the future limestone test runs,
the degree of sulfite oxidation in the bleed solids could be expected to be
higher with resultant high CaSO4 content in the solids. This would give
better solids settling characteristics in the clarifiers. Therefore, it is
reasonable to consider the free settling rates given in Figure D-l as
conservative.
The last lines in Tables D-l, D-2, and D-3 show the predicted makeup
water requirements under the indicated operating conditions for each
scrubber system. The net water evaporation to flue gas varies with the
scrubber slurry temperature and is estimated to range from 1 to 3 gpm
per 10,000 acfm of inlet gas for slurry temperatures of 115 to 128 °F,
respectively.
The maximum obtainable clarifier underflow solids concentration is
about 40 wt %. When the scrubber is operated in conjunction with the
centrifuge or vacuum filter, instead of the clarifier, the makeup water
requirements as shown in the tables would be less because of the expected
lower water content in solids discharged from the centrifuge or vacuum
filter. For example, Table D-2 shows a water makeup of 8 gpm for TCA
systems at 20, 000 acfm and 40 wt % solids and 4. 9 gpm of water in clari-
fier underflow. If instead, the vacuum filter is used and the filter cake
D-6
-------
contains 70 wt % solids, the water discharged with the cake (19.4 TPD
of dry solids) would be only 1. 4 gpm. The required makeup water would,
therefore, be less by 3. 5 gpm.
D-7
-------
rnggwcM 'ERi>T-6°50/2-73-013
4. Title and Subtitle
EPA Alkali Scrubbing Test Facility: Sodium Carbonate and
Limestone Test Results
'. Auchor(s)
M ET)Stein L Svbert and I Raben
>. Performing Organization Name and Address
Bechtel Corporation
50 Beale Street
San Francisco, California 94119
12. Sponsoring Organization Name and Address
EPA, Office of Research and Development
NERC-RTP, Control Systems Laboratory
Research Triangle Park, North Carolina 27711
3' Recipient's Accession No.
5' Report Date
August 1973
6.
8. Performing Organization Rept.
No.
10. Project/Task/Work Unit No.
11. Contract /Grant No.
PH 22-68-67
13. Type of Report & Period
Covered
Final
14.
15. Supplementary Notes
16. Abstractsrpne repOrt describes and presents initial results of testing a prototype wet-
lime/limestone scrubbing facility for removing SO2 and particulates from flue gases.
The facility consists of three parallel scrubbers—a venturi/spray tower, a Turbulen
Contact Absorber (TCA), and a marble-bed scrubber--each able to treat a 10-Mw
equivalent (30,000 acfm) of flue gas from a coal-fired boiler at TVA's Shawnee
Station. Na2CO3 tests were completed in 7/72. As of 6/73, short-term (less than 1
day) limestone factorial tests were essentially complete, and longer term (2+ week)
reliability verification tests were 50% complete. Long-term (4-10 month) limestone
tests and initial lime tests are scheduled to begin 9/73. The short-term limestone
tests, conducted at high scrubber inlet liquor pH (6.0-6. 2), saw SO2 removals of 80
(venturi/spray tower and marble-bed scrubber) to 96% (TCA). Initial longer term
tests were run at reduced stoichiometries to increase system reliability and lime-
stone utilization. For the TCA, limeston
utilization was 83% with SO2 removal of
80-85%. Operability and reliability of th<
scrubbers for these tests were good.
17. Key Words and Document Analysis. 17a. Descriptors
Test Facilities
Prototypes
Air Pollution
Calcium Oxides
Limestone
Washing
Sulfur Dioxide
Flue Gases
Spray Tanks
Coal
Boilers
17b. Identifiers/Open-Ended Terms
Air Pollution Control
Stationary Sources
Particulates
Venturi/Spray Tower Scrubber
Turbulent Contact Absorber (TCA)
Marble-Bed Scrubber
17c. COSATI Field/Group
J4J)
18. Availability Statement
Unlimited
19. Security Class (This
Report)
UNCLASSIFIED
20. Security Class (This
Page
UNCLASSIFIED
21. No. of Pages
247
22. Price
FORM NTIS-35 (REV. 3-72)
USCOMM-DC M852-P
------- |