EPA-600/2-75-057
October 1975
Environmental Protection Technology Series
THE MAGNESIA
PROCESS AS
PLAUT
Iisiareli Laboritffry
Offic® :• jf Hesearcfc and Develipment
U.S. EnviroimiBgrtal Protecfioi Agency
Research Triable Park. N.C. 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development,
U.S. Environmental Protection Agency, have been grouped into
five series. These five broad categories were established to
facilitate further development and application of environmental
technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in
related fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION
TECHNOLOGY series. This series describes research performed
to develop and demonstrate instrumentation, equipment and
methodology to repair or prevent environmental degradation from
point and non-point sources of pollution. This work provides the
new or improved technology required for the control and treatment
of pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. 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 endorse-
ment or recommendation for use.
This document is available to the public through the National
Technical Information Service, Springfield, Virginia 22161.
-------
EPA-600/2-75-057
THE MAGNESIA SCRUBBING PROCESS
AS APPLIED TO
AN OIL-FIRED POWER PLANT
George Koehler (Chemical Construction) and
James A. Burns (Boston Edison)
Chemical Construction Corporation
One Penn Plaza
New York, NY 10001
Contract No. CPA 70-114
ROAPNo. 21ADA-004
Program Element No. 1AB013
EPA Project Officer: C. J. Chatlynne
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
October 1975
-------
ABSTRACT
.A full size demonstration of the magnesia wet-scrubbing
system for Flue Gas Desulfurization was conducted on an
oil fired, 150 MW generating unit. The project involved,
first, design and construction of an S02 removal system
based on firing 2.5% sulfur fuel and an MgO regeneration
facility for 50 ton/day acid production, arid then operation
of both plants over a two year period. Four thousand hours
of operation were logged at the generating station and the
ability of the system to remove 90% of the inlet S02 and
control particulate emissions was shown. Regenerated
magnesia was recycled successfully and over 5,000 tons
of acid were marketed from the regeneration plant demon-
strating the feasibility of the process, correlations
were developed to determine SC>2 removal for varying boiler
loads and fuel sulfur content, and to control regeneration
of acceptable alkali. Several other studies of the process
technology and process chemistry were undertaken as part
of the work.
- 111-
-------
INDEX
Page
i.o
2.0
3.0
4.0
5.0
Summary
1.1
1.2
1.3
1.4
Boston Edison SC>2 Recovery System
Essex Chemical - MgO Regeneration
and Sulfuric Acid Production
MgO Losses
1.3.1 Boston Edison
1.3.2 Essex Chemical
General
Background & Introduction
General Process Description
3.1
3.2
Description of Plants
3.1.1 Power Plant
3.1.2 Acid Plant
Description of Installed Facilities
3.2.1. Magnesia SC-2 Absorption
System
3.2.2 Magnesia Regeneration System
3.2.3 Acid Plant Modifications
Description of Work Performed
4.1
4.2
4.3
4.4
SC-2 Absorption Facility Operations
4.1.1 End of Construction Phase
4.1.2 Pre-Start Up Period
4.1.3 Start-up & Break-in Period
4.1.4 Planned Operational Tests
4.1.5 Operational Testing
Regeneration-Recovery Operations
4.2.1 End of Construction Phase
4.2.2 Pre-Start Up Period
4.2.3 Start-Up & Shake-Down Period
4.2.4 Planned Operations Tests
4.2.5 Operational Testing
Summary of System Modifications
System Availability
1
3
9
11
11
11
12
13
19
24
24
28
30
31
36
42
48
48
48
50
53
58
64
66
66
67
68
71
76
77
82
Equipment Arrangement & Material of
Construction
5.1
Absorber System Areas Subject to
Corrosive Attack
87
96
- IV -
-------
INDEX
5.2 Centrifuge System
5.3 Stack
5.4 Regeneration System
5.5 Corrosion Samples
6.0 Process Chemistry 119
6.1 Effect of MgO Properties on Process 120
6.1.1 Pulverization 121
6.1.2 Hydration of MgO 122
6.2 System pH & S02 Absorption 122
6.2.1 S02 Absorption 127
6.3 Centrifugation & Solids Separation 132
6.4 Regeneration Mechanisms 132
6.4.1 Magnesium Sulfate Control 139
6.4.2 Magnesia Activity Control 141
6.5 MgO Losses & Regeneration Cycles 141
7.0 Description of Test & Development Program. 146
7.1 Test Program 147
7.1.1 Start Up & Break In Tests 147
7.1.2 Variables Screening Tests 147
7.1.3 Long Term Tests 148
7.2 Statistical Methods 148
8.0 Performance Test Results 154
9.0 Data 159
9.1 Operating Information 160
9.2 Range of Significant Variables 169
9.3 Miscellaneous Data 197
10.0 Financial Data 202
10.1 Systems Cost 204
10.2 Operating Costs 206
10.3 Process Economics 206
11.0 References 216
12.0 List of Publications 217
13.0 Conversion From English to Metric Units 219
- v -
-------
INDEX
Page
Appendix A - Magnesium Sulfite Hydrates -
Transition & Formation 220
Appendix B - Physico - Chemical Data 227
Appendix C - Analytical Techniques 236
FIGURES
1 S(>2 Removal Efficiency-Effect of A P 4
2 'Effect of pH on SC>2 Outlet Concentration 5
3 Calciner - Effect of Mid Kiln Temperature
on Bulk Density 10
4 Fossil Fuel Steam Generating Unit Life
.Curve . 15
5 Chemico-Basic Magnesia F.G.D. System -
Schematic Flow Diagram 20
6 Sectional View of Absorber 21
7 Mystic #6 Boiler Cross Section 26
8 Damper Location & Ducts F.G.D. System 27
9 Process Flow Diagram-Absorber & Centrifuge
Systems 32
10 Process Flow Diagram Drying System 38
11 Process Flow Diagram Calcining System 40
12 Sulfuric Acid Plant Schematic 44
13 Process Flow Diagram Tail Gas Scrubber 46
14 Project Schedule 49
15 F.G.Do System Operating Periods & Outages
Jan. 1974 - June 1974 85
16 General Arrangement Plan F0G.D. System 88
17 General Arrangement Plan Regeneration
System 89
18 Venturi Throat Area Wear 98
19 Demister Modules 98
20 Recycle Pump Impeller Wear 99
21 Recycle Pipe-Upper Section 101
22 "T" Section of Recycle Line 102
23 Cross Section of Recycle Pipe 103
24 Microstructure of Recycle Pipe I.D. 105
25 Neutralization Curve 106
- vi -
-------
FIGURE
26 Centrifuge Conveyor Wear 109
27 Centrifuge Conveyor Scale 109
28 Conveyor Wear Plot 110
29 S02 Removal Efficiency Effect of A P 130
30 Dispersed Liquid Surface Area 131
31 Effect of System pH and MgSC>4 Level on
Recycle Solids Content 133
32 Calciner Operation - Effect of Mid-Kiln
Temperatures & % Carbon on Bulk Density 140
33 MgO Consumption by Operating Period 142
34 Mystic #6 Daily Load Swing 149
35 Mystic #6 Seasonal Load Swing 150
36 Process Data Flow Diagram 152
37 ' S02 Removal Efficiency 170
38 S02 Inlet Concentration 171
39 S02 Inlet Concentration 172
40 Boiler Load 173
41 Absorber A P 174
42 Absorber Inlet Temperature 175
43 MgO Slurry Temperature 176
44 Dryer Outlet Temperature 177
45 Dryer Product Temperature 178
46 Dryer Product Solids Content 179
47 Dryer Product % MgS03 180
48 Dryer Product % MgS04 181
49 Dryer Product % MgO 186
50 Centrifuge Feed Rate 183
51 Centrifuge Torque 184
52 Recycle Slurry Flow 185
53 Recycle Slurry Temperature 186
54 Recycle Slurry Solids 187
55 Recycle Slurry MgS04.Content 188
56 Recycle Slurry MgO Content 189
57 Recycle Slurry pH 190
58 Calciner - Mid-Kiln Temperature 191
59 Calciner Feed Rate 192
60 Calciner Product - % MgO 193
61 Calciner Product - % MgS04 194
62 Calciner Product Bulk Density 195
63 Calciner Off-Gas % 02 196
- vn -
-------
TABLES
1 Typical Composition at Major Product
Streams 6
2 Typical Fuel Oil Analyses 28
3 Sulfuric Acid Unit Feed Gas 30
4 Stream Properties & Composition - Absorber
& Centrifuge System . 33
5 Stream Properties & Composition - Drying
System 39
6 Stream Properties & Composition - MgO
Regeneration System 41
7 Stream Properties & Composition - Tail
Gas Scrubber 47
8 Absorber Gas Flow 63
9 F.G.D. System Availability by Test Program
Phase 83
10 F.G.D. System Availability - June 1973 to
June 1974 84
11 Equipment Performance F.G.D. Systems 90
12 Equipment Performance Regeneration System 93
13 Chloride Analyses 97
14 Centrifuge Operating Hours 108
15 Stack Drainings Analyses 112
16 Corrosion Test Data 115
17 Coated Specimen Test Coupons 117
18 Calciner Product Screen Analyses 123
19 Effect of Temperature of Digestion on S02
Removal Capacity 124
20 Rate of Hydration of MgO 125
21 PK Values of Weak Acids 126
22 Potential Sources of MgO Losses 143
23 S02 Removal - Performance Test Results 155
24 Particulate Removal-Performance Test
Results 156
25 Particulate Removal by Particle Size
Performance Test Results 157
26 Oil Analysis During Performance Tests 158
27 F.G.D. System Operating Conditions 161
28 F.G.D. System Stream Analysis 163
29 Regeneration System Operating Conditions 165
30 Regeneration System Stream Analysis 167
31 Mystic #6 Boiler - Fuel Oil % Sulfur 198
- vni -
-------
TABLES
32 High Sulfur Oil Ash Analyses 200
33 Regenerated MgO - Vanadium & Nickel
Content 201
34 Summary of Financial Data 203
35 System Cost Details - F0G0D. Plant ' 204
36 System Cost Details - Regeneration Plant 205
37 Operating Cost - F.G.D. Plant 207
38 Operating Cost - Regeneration Plant
plus Testing & Quality Control
Program 208
39 Project Management Costs 211
40 Summary of Estimated Final Investment -
200 MW Coal-Fired Unit 214
41 Annual Costs - 98% H2S04 From Power
Plant Stack Gas - 200 MW Coal-Fired
Unit 215
- ix -
-------
ACKNOWLEDGEMENTS
The authors wish to acknowledge the contribution of
Boston Edison Co. and Chemico on-site personnel at the
Mystic Station facility, and also the assistance and
contributions of Essex Chemical Co., Basic Chemicals,
and EPA to this program.
The authors are also indebted to Dr. A. Ray and Mr.
E. Dober who assisted in the preparation of this report
and to Mr. M. A. Maxwell of the Environmental Protection
Agency who contributed greatly to the success of the
project.
- x -
-------
1.0 SUMMARY
The control of emissions from fossil fuel fired equipment
has long been a matter of national concern. Of these
pollutants sulfur dioxide (S02) produced in the combustion
of coal and oil at power generating plants represents approxi-
mately 55% of the S02 discharged annually to the natural
environment in the United States.
Acting under the impetus of the Clean Air Act to reduce
S02 emissions and establish air quality standards several
methods of control for this pollutant were implemented.
initially, reduction of S02 levels was accomplished through
the use of low sulfur fuels. However, because of their
limited availability and high costs, the significant impact
on abatement of S02 emissions is expected to come from flue
gas desulfurization (F.G.D0) methods.
The F.G.D. processes are generally divided into two
categories: "throw away" and "regenerable". In the throw
away processes, flue gas SC>2 is absorbed, precipitated as a
sulfur containing solid material and the solid is disposed
of as a waste material. In the regenerable processes, the
S02 is absorbed and subsequently recovered in a usable form
while the absorbent is regenerated and returned to the F.G.D0
system.
One regenerable flue' gas desulfurization process is
based upon the reaction of magnesium oxide (magnesia) with
S02, forming magnesium sulfite. The magnesium sulfite solids
are separated by centrifugation, dried to remove moisture
and then calcined to regenerate magnesium oxide for recycle
and S02 for conversion into sulfuric acid. Once properly
conditioned, the SO2 can be used by existing as well as new
sulfuric acid plants.
Although laboratory and pilot work had been done by
Chemical Construction Corp. on the absorption and integrated
plant operation and by Basic Chemicals on the regeneration steps
by 1970 several questions on full-scale application of this
method remained.
-1-
-------
1) The ability to efficiently remove sulfur
oxides from the flue gases.
2) The ability to continually regenerate a
reactive magnesium oxide.
3) The quality of the product sulfuric acid.
4). Mechanical and materials reliability.
5) Projected construction and operating costs.
6) Transport and storage properties of the
magnesium sulfite and regenerated magnesium
oxide.
In 1970, the U.S. Environmental Protection Agency and
the Boston Edison Company agreed to provide the funds for
the construction and 2-year operation of a large prototype
sulfur dioxide recovery plant based upon magnesia slurry
scrubbing. The Chemico-Basic magnesia process was chosen.
The SC>2 absorption plant was installed at Boston Edison's
Mystic Station in Everett, Massachusetts, and the regenera-
tion facility at Essex Chemical's Rumford, Rhode Island,
sulfuric acid plant. The process was scaled up from a small
pilot plant handling 1,500 cubic feet per minute (CFM) of
gas to a full size unit designed to treat the flue gas
(450,000 ACFM) from a 150 MW generator at the station fired
with high sulfur, No. 6 fuel oil.
A test plan was developed to answer the questions on
full-scale system performance, and to determine the effects
of the numerous control and stream composition variables
on operability and efficiency of the magnesia slurry method
of S02 control. The goals of the program were achieved and
relationships developed to measure the effect of variations
in system operation on both SC^ removal efficiency and
absorbent regenerability.
The program spanned a 21 month period for construction
of the plants and a 27 month period of operations dividied
into a preliminary pre-startup and shakedown period, planned
operational testing period and an end period for continuous
-2-
-------
running at a stable set of conditions, in all, over 4,000
hours of running time of the F.G.D. system were logged.
1.1 BOSTON EDISON S02 RECOVERY
The chemical and mechanical performance of the scrub-
ber itself was excellent. None of the internal plugging
problems which plagued the early operations of the lime and
limestone processes were encountered. The plastic lining of
the scrubber was in sound condition after two years of opera-
tion. Erosion and corrosion were experienced in the carbon
steel recirculating slurry piping. The use of rubber-lined
pumps, valves, and piping in certain areas of this system is
now considered as the most practical solution to this problem.
The S02 removal efficiency was excellent over a wide range
of operation, even during the last six months of the project
when the S02 control system faced frequent boiler outages.
The results of the performance tests which were conducted by
a third party, York Research, gave as an average (four tests) :
S02 Removal 91.7%
Oil Ash Particulate Matter Removal 57%
A correlation of S02 removal efficiency as a function
of inlet S02 concentration and pressure drop across the
scrubber was developed based on actual performance data and
the correlation is shown graphically in Figure 1. in this
same figure, the excellent operability range of the system
is shown by the consistently high removal efficiency curves
for pressure drops above 6 inches of water.
The major control variables used to regulate the scrub-
bing operation were the recycle slurry pH and percent solids.
The effect of pH on S02 concentration is shown in Figure 2.
The effect was controlled by adjusting the rate of MgO addi-
tion. Typical compositions of the recirculating slurry are
shown in Table 1, along with other key streams. The amount
of solids in the slurry was controlled by adjusting the rate
of bleed from the recycle slurry.
The majority of the operating problems experienced at
the power plant installation were closely tied to the solids-
handling system. Many of these problems were due to the
-3-
-------
D_
D.
ar>
LI-
LI-
UJ
_j
era.
z:
LU
cn
CO
DP=t2 IM.
36
DP=6 IM.
DP»2 IN.
=500
nbo
ebo
sbo
1000 1200 1400 1600
INLET SQ2 - PPM
1800
2000
502 REMOVRL....EFFICIENCY
EFFECT
OF DELTfl P flND INLET
502
CONC
Fig. 1
-------
o
o
•
o
C\J.
o
o
2:o
Q_o
•—'0_
en
o
o
•
o.
00
o
o
o
r-,
5.00
5.50
6.00
EFFECT OF PH
-BOSTON EDISON
6.50 7.00 7.50
PH
ON 302 OUTLET
JUNE 1973
8.00
8.50
9.00
9.50
CONC.CPPM)
Fig. 2
-------
TABLE 1
TYPICAL COMPOSITION OF MAJOR PRODUCT STREAMS
Recycle Slurry*
Centrifuge Cake
Calciner Feed
Calciner Product
MgO
0.3
2.2
7.1
87.1
%Of
MgSO3
6.3
37.8
60.3
0
Total Stream
Mgb04
14.4
6.5
10.9
7.7
Water & Inerts
79.0
53.5
21.7
5.1
* Approximately 10% suspended solids in slurry
-6-
-------
nature of the magnesium sulfite crystals obtained from the
scrubber. Although the easily separated magnesium sulfite
hexahydrate (MgS03.6H20) crystals produced in the pilot
operations were expected, any sustained operation of the
scrubber resulted in the production of the finer magnesium
sulfite trihydrate (MgSC>3.3H20) crystals.
This unexpected occurrence resulted in crystal pro-
perties different from those used in the design of the solids-
handling system and the dryer, and was the primary process
problem encountered in the project. Modifications eventually
were required in all of the process areas handling magnesium
sulfite crystals to allow for operation with large amounts
of the trihydrate present.
The rotary dryer, used for removing free and bound
water from the centrifuge slurry solids, was found to be at
the center of these initial operating problems. Since this
dryer was designed for operation with the relatively coarse
hexahydrate crystals, the fine trihydrate crystals caused
excessive dusting, buildup of solids, and loss of drying
ability. The following modifications led to reliable dryings
1) The dryer operation and internal configuration
were modified to allow the dryer to act as a
granulator for the fine crystals.
2) A scalping screen and lump breakers were
installed at the dryer discharge to handle
oversize agglomerated granuales of dried
magnesium sulfite from the dryer.
3) High dust losses from the dryer because of
the fine crystal size were corrected by piping
the dryer off-gas to the venturi scrubber.
4) Hammers were installed on the dryer shell to
prevent the build-up of solids.
5) The dryer conveyor was lengthened to allow
wet centrifuge cake to be added at a point
further into the dryer.
-7-
-------
6) The mixing of the dry material, removed from
the cyclone dust collector operating on the
dryer off-gases, with the wet centrifuge solids.
created a mixture which would "set up" on the
screw conveyor. A pneumatic conveyor system
was installed to transport this material
directly to the MgSOo silo.
As the first regenerated magnesium oxide arrived at
the power plant for reuse, other problems in solids handling
developed. The calcined magnesia returned from the acid plant
contained a fraction of a larger-than-desirable particle size
as well as some overturned MgO particles that would not react
with water to form a reactive slaked slurry. A pulverizer
was installed to grind this oversized MgO to that which was
usable. This pulverizer was eventually transferred to the
acid plant. In addition, magnesium oxide returning from the
acid plant initially showed a lower reactivity than expected.
This was corrected by heating the MgO slurry tank to 180°F
to increase the slaking rate of the MgO.
The combined effects of corrosion and erosion were
experienced in all pumps, valves, and piping that handled
the scrubbing slurry. Because the prototype plant was de-
signed for only a two year period of operation, cast iron
and carbon steel were used in these areas and quickly failed.
Rubber or plastic lining is now considered necessary in
slurry-handling applications.
While centrifuge performance was generally satisfactory,
improved internal washing techniques were required to reduce
wear and improve reliability.
The availability of the S02 recovery system to the
Boston Edison Unit No. 6 boiler ranged from a low of 13
percent to a high of 87 percent during the April 1972 and
June 1974 operating periods,respectively. Many of the lower
value availabilities were caused by the problems discussed
above. During the last four month period of the recovery
system's operation, the monthly availabilities were 87 percent,
81 percent, 57 percent and 80 percent. The lowest reported
value was due to the lack of MgO for S02 removal, caused by
problems in the sulfuric acid plant and an intentional
-8-
-------
emptying of the MgO silos during a controlled system test.
1.2 ESSEX CHEMICAL - MgO REGENERATION AND SULFURIC ACID
PRODUCTION
After solving several initial operating problems, which
are discussed in the following paragraphs, the chemical plant
was consistently able to manufacture high quality 98 percent
sulfuric acid, which was sold in the commercial market.
Over 5,000 tons of H-SO^ were made from captured flue gas
sulfur. During this same period, 3,000 tons of magnesium
oxide were regenerated and shipped to the power plant for
reuse.
The continued ability of the system to operate effi-
ciently with regenerated magnesium oxide was a major variable
investigated during the project. The occasional formation of
a small amount of less reactive MgO was observed. This was
generally accompanied by an increase in density of the magnesia.
Data from this test program were correlated by regression
analysis with operating conditions and the percentage of car-
bon in the feed to predict operating conditions which would
result in a low-bulk-density magnesia. This correlation is
shown in Figure 3 and indicates that the formation of low-
bulk-density (high reactivity) magnesia is favored by low
calciner temperature and an increased amount of carbon in the
calciner feed.
Mechanical problems at the chemical plant were centered
around the calciner operation,. One such problem was air leak-
age into the calciner's firing hood0 The calciner must operate
at very near neutral or reducing conditions in order to allow
for the reduction of magnesium sulfate to magnesia. A con-
siderable amount of effort was expended in tightening up seals
and reducing air leakage. The problem was finally solved by
the installation of a friction seal.
Another problem occurred when oil was fired during the
startup of the cold calciner. Hydrocarbon vapors not removed
by the scrubbing equipment entered the sulfuric acid towers,
which immediately caused blackening of the acid. This problem
was solved by installing a fan and a short stack, which allow-
ed the startup vapors to bypass the sulfuric acid plant.
-9-
-------
V-.
WJ_
n
CO
^
LU
_,
<—'
CD
D.
nj
O
I
O.
=2. 0
700
800
900
1000
II 00
1200
1300
1400
M1DKILN TEMP-DEG F
CRLCINER OPERRTIGN - EFFECT
MIDKILN TEMP. & 7.C. ON BULK DENSITY
1500
OF '
Fig. 3
1600
1700
-------
This system was used only when heating up the calciner and
was never operated during periods of S02 generation.
1.3 MgO LOSSES
During the early period of operation, magnesium oxide
losses were excessive due primarily to spills and required
cleanouts caused, as previously described, by the solids-
handling problems.
1.3.1 Boston Edison
In the final four months of operation, however, the
bulk of these problems were solved and a careful measure-
ment was made of system losses. During these four months,
1,500,000 Pounds of regenerated material were recycled to
the power plant as scrubber slurry makeup. A 13-day test
to identify each specific loss point was also conducted.
The measurements showed a loss over the entire power plant
operation of 0.37 ton per operating day, distributed as
follows :
Loss to stack 0.13 ton per day
Scrubber overflow 0.14 ton per day
Miscellaneous 0.10 ton per day
Total 0.37 ton per day
With an average MgO consumption of 10.61 tons per day
during this period, this total loss amounted to 3.5 percent
of MgO consumption at the power plant. The design loss at
this same location was predicted to be 5 percent.
1.302 Essex Chemical
The greatest losses for this project occurred at the re-
generation plant in Rumford, where 0.5 ton per day was scalped
off the calciner product as large lumps before the pulveriza-
tion process. Future system design will provide for the re-
claiming of these losses. In addition, 1.5 tons per day
were lost from the neutralizer overflow. In subsequent
designs, this large loss can be recovered for recycle by
separation of solids in the neutralizer overflow. Thus,
-11-
-------
almost all of the regeneration plant losses can be reclaimed
by improved design at new regeneration plants.
1.4 GENERAL
Operation of the full size plant yielded much valuable
information for the design of future plants. During the
course of this project:
*
1) Studies of the complex physico-chemistry of the
process were undertaken yielding data not previously re-
ported on:
a) The kinetics of the MgSC>3 hydrate formation
and transformation.
b) The activation phenomena of regenerated MgO
slaking.
c) The formation and influence of MgSC>4 in the
system.
2) Measurements were made of the physical properties
of the chemical components of the system.
3) New analytical methods and techniques were
developed to assure process and quality control at the
plants.
4) Observation and testing of the suitability of
the equipment and materials of construction yielded new
recommendations for the design of improved plants.
Finally, the accumulation of operating experience and
a large data bank for reference allowed initiation of the
second phase of the overall application of the magnesia
slurry system for S02 control at a coal fired generating
station.
* Work in these areas is still ongoing.
-12-
-------
2.0 BACKGROUND AND INTRODUCTION
The control of sulfur oxides emitted to the atmosphere
from power plants has become increasingly important in
recent years. The deleterious effect on human health and
the environment of these substances has been well documen-
ted and needs no review here. The growing demand for energy
has compounded the problems of pollution abatement. For
example, total S02 emissions in 1965 were estimated at 25
million tons per year, and it was anticipated that, if no
controls were imposed, these emissions would rise to 75
million tons per year in the United States by 1975. Of this
amount, two thirds would originate from the products of com-
bustion of fossil fuel fired power stations.
The national concern for this and other problems
related to air pollution was reflected by the Congress ini-
tially in the authorization of a Federal program of research
for air pollution control in 1955, the later adoption of the
Clean Air Act of 1963, and finally the Air Quality Acts of
1967 and 1970. All of these acts focused attention on the
environmental problems created by the conversion of fossil
fuels to energy for heat and power. Also reacting to these
problems, the utilities and chemical industries instituted
numerous programs and studies for environmental protection,
many industries having installed pollution abatement equip-
ment decades before the enactment of the specific legisla-
tion.
Chemico, a leader in the field of the design and con-
struction of air pollution control equipment and in the
incorporation of environmental protection provisions in the
chemical plants which it also designs and constructs, drew
on its 27 years experience in particulate emissions control
and initiated several in-house studies in means of SO2
control from large boilers. By 1966 these Chemico marketing
studies indicated that the New England area would become a
high cost area for fuel oil and that low sulfur content
then proposed as the solution for S0_ emissions, would
itself have the highest cost impact. Engineering studies
had also revealed that alternative schemes for S02 control
which produced waste products were unacceptable in the New
-13-
-------
England area due to the lack of disposal sites. Evaluating
this information, laboratory studies were initiated on
methods of controlling emissions by the use of a regenerable
process. These studies demonstrated that a process built on
the use of magnesia appeared to be the most practical. In
addition, the magnesia system provided a means of incorpora-
ting the concept of a separate, self-contained, central
regeneration plant to serve many power plants. This chemical
plant could then be sized economically for the recovery of
sulfur values, and the design could be based on the longer,
full production capabilities of such plants, rather than the
dimishing power production (Fig.4) experienced in modern
generating stations.
An incentive for rapidly proceeding to full scale plant
operations of the magnesia process, by-passing the smaller
scale semi-works units normally employed in scale-up, was
provided by the Panel on Control of Sulfur Dioxide from
Stationary Combustion Sources. This panel, a committee of
the National Research Council and the National Academy of
Engineering, had recommended funding of prototype facilities
in order to speed the demonstration of the necessary pollu-
tion abatement systems.
The arrangements for funding the magnesia system proto-
type were concluded in a master contract signed in June,1970.
In it the United States Environmental Protection Agency and
Boston Edison Company agreed to provide the funds for a large
prototype system using the Chemico-Basic MgO Sulfur Recovery
Process for control of SC>2 emissions from power plants.
Capital funds for the absorption system to be installed at
the power plant were provided by the Boston Edison Company,
while E.P.A. furnished funds for the magnesium oxide regen-
eration system to be installed at a small sulfuric acid plant
located in Rumford, R.I.,owned by Essex Chemical Company,as
well as the acid plant modifications necessary to accommodate
a "roaster gas" process. Operation and maintenance of the
facilities were the responsibility of Essex Chemical Company
and Boston Edison Company respectively, with reimbursement of
these costs by E.P.A. However, Boston Edison's reimbursement
was limited by a ceiling. Chemico was responsible for
Project Management of the total project, with a part of the
-14-
-------
FIG.4-THE EFFECT OF LOAD FACTOR
UPON S02 RECOVERY REVENUE
100
80%
en
I
60%-
40%
20%-
DECLINING LOAD FACTOR CURVE
OR
PRODUCT RECOVERY
REVENUE CURVE
10
15
20
25
30
35
FOSSIL FUEL STEAM GENERATING UNIT LIFE- YEARS
-------
cost reimbursed by E.P.A.
A history of these and the subsequent events which led
to the full scale demonstration of the magnesia process on a
utility boiler is summarized below:
NEW ENGLAND S02 CONTROL PROJECT HISTORY
1966 - Chemico studies showed New England would
become a high fuel oil cost area.
1966 - Boston Edison was investigating alterna-
tives to the use of low sulfur fuel.
1967 - Laboratory studies of 502 removal/regen-
eration schemes were initiated by Chemica
1967 - Chemico and Boston Edison Co. discussed
testing of these schemes at a Boston
Edison Co. generating station.
1968 - Pilot plant studies of the magnesia system
at generating stations were undertaken by
Chemico.
1969 - Chemico and the Environmental Protection
Agency discussed mutual funding and par-
ticipation in a demonstration project.
1970 - 1)
The master contract, creating the
New England SO- Control Project was
signed by Chemico, Boston Edison Co.,
and E.P.A.
2) Chemico entered into agreements with
Boston Edison Co. for the design and
erection of a magnesia system in their
Mystic #6 boiler.
3) Chemico and Essex Chemical Co. entered
an agreement whereby the regeneration
-16-
-------
section of the magnesia system would
be erected at their Rumford, R.I.
sulfuric acid plant, and the recovered
S02 would be processed into sulfuric
acid.
4) Chemico completed process and design
engineering for both plants.
1971 - Construction of both plants started.
1972 - Construction of both plants completed.
1972 - Initiation of the start-up phase of the
New England SC>2 Control Project Test and
Development Program.
1973 - Initiation of the Planned Operational
testing phase of the program.
1974 - Completion of the program.
This report covers the period from January 1st, 1972
thru the end of the project on June 30th, 1974. During the
first few months of this period, construction at Mystic
Station was completed. This completion had been delayed from
the original forecast date of October 1971 to April 1972 be-
cause of a combination of cold weather, competing maintenance
projects at the same station, and a lower productivity in the
area than had been predicted.
During the following two year period, the operations and
test work was conducted, again with an altered schedule.
The pre start-up period originally scheduled for two
weeks was extended to three months as methods were developed
to cope with the different hydrate and crystal size of mag-
nesium sulfite which was produced in the continuous operation,
as this material had not been encountered in the previous
intermittent pilot plant operation. The Start-up Shakedown
Phase scheduled for a three to six month period was not com-
pleted till the end of April 1973 because of the step-wise
-17-
-------
system necessary to solve the numerous operational, mech-
anical and materials handling problems that arose as the
work progressed.
The planned operational test period was shortened and
concluded by February of 1974. At this time, performance
tests were run on the system, and these demonstrated satis-
factory compliance with guarantees. After the performance
tests were completed, the period from March 1974 to the ter-
mination of the project was devoted to longer and steadier
operational periods at both plants.
Description of the work performed in the previous period
during most of the construction phase, complete engineering
designs of the system, and extensive descriptions of the oper-
ational programs, have been presented in the following
reports:
1) Detailed Engineering Report
Contract CPA 70-114; Chemico Job 1857G
2) Detailed Engineering Report
Contract CPA 70-114; Chemico Job 1858G
3) Report for the period 7/1/70 thru 1/1/72
Contract CPA 70-114
4) Test and Development Program
150 MW Prototype Venturi Scrubbing System
5) Planned Operational Test for New England SC>2
Control Project
-18-
-------
3.0 GENERAL PROCESS DESCRIPTION
Chemico/Basic's Magnesium Oxide System for the recovery
of sulfur dioxide from power plant flue gases is shown sche-
matically in Figure 5. This process, which utilizes the
sulfur dioxide absorption characteristics of an aqueous
slurry of magnesium sulfite, magnesium sulfate, and magnes'-
ium oxide, is composed of five primary steps, these are:
Absorption
Centrifuging
Drying
Materials Handling
Calcination
a) Absorption
The process chemistry which describes the removal of
S02 from the flue gas can be most simply explained as the
diffusion of SO2 through the flue gas to a liquid surface,
absorption of the SO2 gas in the liquid and the subse-
quent reaction of S02 with the hydrated form of MgO, i.e.
M9°(aq) + S02 *~ M9S03 (3-D
The MgSOg produced has a low solubility and can be
separated from the absorbing slurry as a solid.
In the process the flue gas containing sulfur oxides
enters a venturi absorber (Figure 6) of special design,
and contacts the absorbing media which is an aqueous
slurry of magnesium oxide, magnesium sulfite, magnesium
sulfate, and a small percentage'of other components
including fly ash from the fuel oil. The process of S0_
removal that occurs is explaned by conventional mass
transfer principles.
The venturi absorber can be considered as similar to a
co-current, packed vessel. In the venturi the liquid
slurry is introduced and flows downward on surfaces over
which pass an accelerating gas stream. The high velocity
gas stream flowing over the liquid causes wave motion on
-19-
-------
MgO ADDITIVE
SCRUBBER SYSTEM FOR S02 RECOVERY
OIL FIRED BOILER
SCHEMATIC PROCESS FLOW SHEET
MjO FROM ACID PLANT
REGENERATION SYSTEM
MgO RECYCLE PROCESS, FOR PRODUCTION OF 98% SULFURIC ACID
SCHEMATIC PROCESS FLOW SHEET
ACID PLANT
SO2 GAS CLEANING
CONCENTRATED SO2 GAS
SULFURIC ACID PLANT
CONVEYOR
CONVEYOR
POWER PLANTS ~~ ""~. ©MEMICO
Magnesia Flue Gas Desulfurization and Regeneration .system
Schematic Flow Diagram
Fig. 5
-20-
-------
ANNULUS
SPRAYS
TANGENTIAL
WASH
CLEAN GAS
OUTLET
TO STACK
INTERMITTENT
MIST
ELIMINATOR
SPRAYS
CONE
WASH
— NORMAL LIQUOR LEVEL
PUMP
SUCTION
Sectional view of absorber at Boston Edison Company, Boston. Mass.
FIG. 6 -21-
-------
the fluid surface, the waves increase in amplitude and
finally disperse as fine droplets. Thus, the whole mass
of liquid can be distributed in the form of atomized drop-
lets in the gas stream. In the process described in this
report, the dispersed droplets generally have a median
size of 400 micrometers and the surface area available for
mass transfer averages 13 ft^ per ft^ of gas.
There are several advantages in using the venturi as an
absorber: The absorption surface is dispersed into and
flows with the gas stream during the time absorption is
occurring, thus eliminating the problems of plugging asso-
ciated with conventional packed towers (the surface area
per unit volume is approximately equivalent to dumped 3
inch rachig rings). Due to system dynamics, this surface
area relation is relatively invarient over wide turn-down
ratios, and high removal efficiencies can be maintained
over the normal operating range of the power plant's
boiler.
A prediction of system efficiency can be obtained with
a fair degree of accuracy by assuming a pure diffusion
phenomena and conventional mass transfer relations. Using
this as an estimate, a maximum efficiency of 96% is poss-
ible for a single stage system. Deviations from that
removal efficiency in the demonstration plant and primarily
caused by equilibrium partial pressures of S02 over the
droplet surfaces greater than zero in the dynamic system.
b) Centrifuging
A stream from the absorption system enters the centri-
fuge where the solids which were formed by the absorption
reaction in the slurry are separated. This bleed stream
is controlled in order to maintain a constant solids con-
tent in the recycle slurry by removal of product magnesium
sulfite and unreacted magnesium oxide and any precipitated
magnesium sulfate. The system is operated so that the
absorbed SC^ is removed as an equivalent amount of the
magnesium sulfur compounds.
-22-
-------
c) Dryer System ,
The wet centrifuge cake containing hydrated magnesium
sulfite, magnesium oxide and magnesium sulfate plus other
solids removed in the venturi absorber-centrifuge system
is passed to a rotary dryer to remove both unbound water
and most of the water of crystallization. The dry pro-
duct is easy to store, and the removal of water reduces
shipping costs. •
d) Materials Handling
The anhydrous magnesium sulfite and magnesium sulfate
produced in the dryer is conveyed to a storage silo for
transportation by truck to the recovery acid plant. The
same transportation equipment is used to return regener-
ated magnesia to the magnesium oxide silo at the power
plant on the return trip. Recycled alkali absorbant from
the magnesium oxide silo and make-up magnesium oxide are
fed to an agitated tank with water where a slurry is pre-
pared for introduction to the recycle system manually on
pH control.
e) Calcination System
Calcination is the process for the regeneration, which
is the reverse of the absorption step. The magnesium
sulfite which has been separated and dried is thermally
decomposed as represented by the following reaction:
MgS03 >- MgO + S02 (3-2)
heat
The recovered SO2 is used in the production of sulfuric
acid and the regenerated MgO returned to the process for
reuse. The dry product transported to the regeneration
acid plant from the power plant is received, weighed and
pneumatically conveyed to a storage silo. It is fed to a
direct-fired rotary kiln at a metered rate, and calcined
to both generate sulfur dioxide gas and regenerate magne-
sium oxide. Coke can be added to provide a reducing
atmosphere, -as necessary, to reduce the residual magne-
sium sulfate to the oxide and sulfur dioxide. The hot
-23-
-------
flue gas containing sulfur dioxide and dust enters a hot
cyclone where essentially all the dust is removed and
returned to the calciner. The flue gas then enters a ven-
turi scrubber for final dust cleaning. At the same time,
the gas is cooled arid adiabatically saturated.
The saturated flue gas is cooled in a direct contact
cooler to meet the requirements of the acid plant water
balance. The cleaned, cooled flue gas then enters .the
drying tower of an existing 50T/D acid plant. At the
Essex Chemicals installation, the resultant product made
from the recovery of the sulfur dioxide is 98% sulfuric
acid. • .
The regenerated magnesia is cooled, conveyed to the mag-
nesia storage silo and recycled back to the power plant.
3.1 DESCRIPTION OF INSTALLATIONS
The magnesia method of flue gas desulfurization removes
S02 from the products of combustion of fossil-fuel fired
eqiupment by absorption in a slurry containing the alkaline
material, magnesium oxide. In subsequent steps the S02 is
driven off by a thermal decomposition of the solid products
of absorption regenerating the magnesia for reuse. Varia-
tions in the process can be used to obtain a range of
strengths of S02 in the gas from the thermal decomposition so
that the products of the recovery operation can vary from
elemental sulfur or sulfur dioxide to sulfuric acid. In the
work described in this report the flue gas to be treated came
from an oil fired power boiler of an electric generating sta-
tion, and the sulfur dioxide was recovered as sulfuric acid at
a remote sulfuric acid plant. The power station and the acid
plant are described in the following sections.
3.1.1. Power Plant
The flue gas desulfurization system was installed at the
Mystic Station of the Boston Edison Company. Mystic Station
is located in Everett, Mass, on a 42 acre site on the north
bank of the Mystic River. The station currently includes six
fossil-fuel steam electric generating units with a total
-24-
-------
generating capacity of 618 MW and a seventh unit, under con-
struction, with a capacity of 600 MW. The existing 618 MW
capacity is provided by three 50 MW units and three 156 MW
units. Flue gas from the three 50 MW units is dispersed
through two stacks 278 feet high and the three 156 MW units
have individual stacks 335 feet high. The units were con-
structed between 1943 and 1961, the last unit (No.6) begin-
ning commercial operation in May 1961. All units were
initially designed for coal firing and were equipped with
electrostatic precipitators for that fuel. The station was
completely converted to oil firing by 1966, and the precip-
itators were deenergized because of the inefficiency of
particulate removal resulting from the change.
The magnesia system was installed on Unit No. 6 at the
station. This generator is powered by a Combustion Engineer-
ing controlled circulation tangentially fired boiler (Fig.
7). This boiler has a rating of 1,000,000 Ib /hr of steam
at 1800 PSIG and 1000°F with reheat to 1000°F.
New ducting was installed at the stack breaching with
dampers in the breaching to divert the flue gas from the
stack to the scrubber system fans; the treated flue gas then
passed back to the stack through a new breaching, also
equipped with a damper (Fig.8). During most of the period
of performance when the scrubber was operated it treated all
of the flue gas (about 40% over design flow at full boiler
loads) which was diverted by fully closing the stack inlet
dampers. During the last three months of operation, the
damper arrangement allowed the excess flue gas to be by-
passed directly to the stack allowing the scrubber to oper-
ate closer to its design capacity and providing reheat to
the combined stack gas.
The fuel oil which was burned at Mystic Station in
boilers 1, 2, 3, 4 and 5 had a maximum of 0.5% sulfur as
required by Massachusetts Law. For the purposes of the Test
and Development program, a special variance was obtained on
October 20, 1970 permitting the use of higher sulfur oil in
the No. 6 unit during the periods of scrubber operation. A
typical analysis of the higher sulfur oil used during scrub-
ber operating is shown in Table. 2.
-25-
-------
FIG. 7
C-E CONTROLLED CIRCULATION BOILER
CAPACItY-.935.000 IB PER MR AI I94O PSI AND IOOO I 1IMP. - RCHEAT IO 1000 F
D,'i,g,.-..d and Buitt By COMBUSTION ENGINHRINC INC.. New »ock
1 M^s:!C STATION-UNITS 445*6
Boston Ediion Coivpnny Evcr»M Mass.
Jft f.Kr.ON A MOP;: A NO •'••,< - .I-..," fr,. npf,.,
-------
TWO EXISTING
ID TANS
E-XISTING
STACK
TWO NEW
F.D FANS
VENTURl
ABSORBER
SECTION 'A-A'
FIG. 8
DAMPER LOCATION & DUCTS
FLUE GAS DESULFURIZATION SYSTEM
MYSTIC 6 STATION OF
BOSTON EDISON CO.
MqO STORAGE
ay.
SILO
JtlgO SLURRY
TANK
TRUCKS
-------
TABLE 2
TYPICAL FUEL OIL ANALYSIS
Sulfur 2.05%
Ash .07%
Carbon 84.39
Hydrogen 11.39
Nitrogen .10
Oxygen 2.00
BTU/lb. 18,350
3.1.2 Acid Plant
The regeneration system was installed at the Rumford
acid plant of Essex Chemical Company located in Rumford,RI.
The Rumford plant is located 55 miles from the installation
at Mystic Station and haulage of the dryer and calciner pro-
ducts between these two sites was done by truck. Both
plants were equipped to load their respective products from
elevated silos and receive their feed by pneumatic unloading.
Freight rates for this haulage were established at 21 cent-.s/
cwt on a round trip basis and 35 cents/cwt for one way
transport.
The plant has been producing sulfuric acid since 1929
when it was build,by Chemical Construction Corporation, and
-28-
-------
for the past several years all acid produced was sold in
the merchant market with no captive use. Markets for the
acid made at this plant are manufacture of detergents, dye-
stuffs, Pharmaceuticals, aluminum sulfate, and tanning
chemicals, steel pickling, boiler water treatment, lead-acid
batteries, galvanization, etc. Sales in this market were at
the full.published price, at the time of this program, ap-
proximately $46 per ton based on 100% H-SO^. No change in
use or pricing was imposed when marketing the acid produced
from MgO regeneration.
The plant is a contact sulfuric acid plant which uses
sulfur as raw material. In the Chemico process the molten
sulfur is injected into the furnace using a spray burner,
with the process air first dried in a tower against 93%
sulfuric acid. The catalyst is vanadium pentoxide.
When first built in 1928 the Rumford plant was an in-
door plant typical of its time and had a capacity of about
20 tons per day of 100% sulfuric acid. Substantial modifi-
cations were made to the plant in 1948 by Chemico to in-
crease its capacity to 50 tons per day. The modifications
included improved converters, a converter heat exchanger,
waste heat boiler and economizer all installed outdoors.
In addition, the cast iron cooling section was enlarged and
moved outside.
This was the plant that was modified in 1971 to accept
the calciner off gas essentially converting it to a metal-
lurgical (roaster) gas plant; however, in the modification
the capability to continue to burn sulfur and augment the
SC>2 from the regeneration plant was retained.
Typical feed gas analyses to the acid unit are shown in
the following table.
-29-
-------
TABLE 3
SULFURIC ACID UNIT FEED GAS
COMPOSITION, MOLE %
Source of Gas N_ C02 P_2 H2p_ S_0_2
Regeneration Section 73 6 5 7 9
Sulfur Burning 79 12 9
3.2 DESCRIPTION OF INSTALLED FACILITIES
The pollution abatement facility, comprising the ab-
sorption, centrifuging and drying section is installed on
the #6 boiler of the Mystic Station, while the regenera-
tion plant (calcining section) is installed at the acid
plant, located approximately 50 miles away. By performing
the drying step at the generating station unbound water and
water of crystallization are removed from the solid prior to
transportation in order to reduce shipping costs.
Transportation of MgSO^ and regenerated MgO between the
two sites was handled by conventional pneumatic discharge
hopper trucks. Originally two 825 ft units were in service;
however, this was eventually reduced to one 1200 ft trailer
hauling MgO from the regeneration facility and MgS03 on the
return trip. The tractor used had a self-contained blower
for unloading material; an auxiliary compressor was avail-
able at the calcining facility to allow the trailer to be
dropped completely if necessary. This compressor was also
used to unload coke brought to the regeneration facility in
a PD (Pneumatic Discharge) hopper rail car which was leased
for this service.
During periods of extended operation of the plants, the
single trailer was unable to transport all the MgSO pro-
duced. In order to move the additional material necessary
for uninterrupted plant operations, a local chemical truck-
ing company provided additional units for the service. In
addition, make-up MgO shipped by rail was transferred to
-30-
-------
this same company's trucks for delivery to the power plant.
All silos, materials handling, and transport systems
were designed on the '.basis of the bulk densities of MgO and
MgSO3 determined in preceding pilot plant work, those were
25 Ib/ft^ and 50 Ib/ft^ respectively. The storage facilities
at both sites were sized to ensure that they were adequate
for continued operation in the event of up to a five day in-
terruption in shipments. This duration was chosen after an
evaluation of information from the Department of Transporta-
tion of both states regarding the frequency and probability
of interruptions due to weather and other circumstances.
Most development work on the chemistry of the process, in-
cluding new analytical techniques, was conducted by Chemico
personnel in a laboratory facility located at the acid plant;
these chemists and technicians also performed the chemical
analysis for stream assay and quality control.
3.2.1 Magnesia SC>2 Absorption System
The SC>2 Absorption System Flow Sheet is shown in Fig.9.
In this system magnesia, both regenerated and make-up mat-
erial, is transferred from pneumatic discharge hopper trucks
to the elevated MgO storage bin through a 4 in pneumatic
unloading system. The MgO bin, 1-101, is 35 ft in diameter
and 42 ft high. Magnesia is fed from the silo, which is
equipped with a vibrating hopper bottom, 0-102, to a 16,000
gallon capicity MgO make-up tank G-102 by an adjustable weigh
feeder, 0-103, with ratio control of make-up water to make
the desired MgO slurry composition. A small pre-mix tank is
interposed between the weighing system and the steam heated
make-up tank to act as a vapor seal. Heated magnesia slurry
is added to the recycle stream by the MgO make-up pump J-101.
The magnesia slurry addition rate is controlled by the oper-
ator to maintain the pH of the absorbing slurry at the de-
sired value.
The recycle stream for the absorber is circulated at a
rate of 14,000 to 15,000 GPM to provide a slurry dispersion
within the vessel sufficient for the desired S02 removal.
The absorber vessel R-201, 31 ft Dia. x 50 ft overall is
supported from a structural steel tower over the recycle
pumps, J201A, B and C, and each of these 350 HP pumps had
-31-
-------
~\ OUST
cov-x.ec.Tora
J3aO
CO
to
CAL.CIMIMO
SYSTEM
-FKOI^V DISCH.
OF EXISTING
I.D.
FIG-. 9
5O2
SYSTEM
-------
TABLE 4
STREAM PROPERTIES AND COMPOSITION
ABSORBER AND CENTRIFUGE SYSTEM
STREAM
NUMBERS
Temp. °F
Flow GPM
Solids Content %•
MgO Ib/Min
MgSO . 6H O Ib/Min
0
11
&. SOLID 5
152
174
13. 15
10. 1
181.7
10.1-
2. 32
0. 273
204.5
1352
1556. 5
5
162
0
411,000
22,840
2,692
6.8
148
.64
0.06
>
12
BREAMS
132
95. 2
10.1
181.7
10.1
2.32
0. 273
204. 5
10.7
213.27
13
161
1342
14
60
64
531
16 17
180
68.4
i
i
i
570
63
-------
half the design flow of the recycle stream.
The recycle stream itself is split into three streams
for distribution before reentering the venturi absorber.
The flow in each stream is adjusted and set by plug valves
to provide a uniform irrigation to a) the converging surfaces
and b) the torroidal section of the throat area. Distri-
bution of absorbing slurry to the torroidal section is made
by 12, six inch nozzles. A similar number of tangential
nozzles supply the outer converging section while the inner
cone is supplied by a single central nozzle.
The slurry enters the upper part of the absorber with
the untreated flue gas, which is diverted from the stack,
using the dampers, through the booster fans K-201 A & B,
each driven by an 800 HP motor,and each designed to handle
283,000 ACFM. at a discharge head of 12 in water. The gas
and slurry mixture passes through the throat area into a
diverging section, then into a central downcomer to exit the
vessel the flow of the cleaned flue gas turns one hundred
eighty degrees upwards. In this step most of the larger
liquid droplets are disengaged.from the flowing gas stream
and fall to the slurry pool in the conical base of the ab-
sorber. The treated flue gas continues upwards through a
slot-type mist eliminator to further remove any entrained
liquid before the gas exits to the stack through the louvre-
type damper.
A stream of slurry, approximately 1.0% of the total
recycle, is taken from the discharge of the recycle pumps
to the centrifuge, R-301. This is a 36 in x 72 in hori-
zontal, solid bowl unit driven by a 200 HP motor,and in
normal operations it removes 50% of the solids in the liquid
stream going to it.
The solids separated in the centrifuge contain an amount
of sulphur (as MgSO3 or MgSC>4) equivalent to the sulfur di-
oxide removal rate in the absorber. The centrate is discharg-
ed directly to an agitated tank, G-301 of 3,000 gallon
capacity which serves as a pump tank for this system,and then
is returned to the basin of the absorber by pumps J-301 A&R.
The additional water, necessary to make up for that lost
-34-
-------
by saturation of the incoming flue gas, is also added to the
basin of the absorber,and in normal operation this reservior
contains approximately 15,000 gallons of slurry. The other
sources of water entering the system are. small amounts added
as pump seal water and an additional quantity used as a mist
eliminator wash; the latter is an intermittent addition.
The flow, sheet for the centrifugeing and drying section
is shown in Fig. 10. The centrifuge cake separated in the
centrifuge is fed to a dryer by screw conveyer 0-402, a 34
ft long unit with a 14 in diameter ribbon flight. The dryer
is a countercurrent, rotary unit 7 ft 6 in in diameter by
60 ft long. Originally of conventional design, this unit
was modified to suit the process as outlined in a later sec-
tion.. Heat is supplied from an integral 18 ft long oil fired
..combustion chamber at a maximum heat release rate of 25 MM
BTU/hr.
The dryer off.gas exits through a multiple cyclone array
and the gas is discharged to the entrance of booster fan K-
201 B. Then, mixed with one-half of the entering flue gas,
it is cleaned of its remaining particulate loading in the
venturi absorber. The cyclones are designed to remove 90%
of the solids entering in the dryer off gas, and the separa-
ted solids pass through a rotary valve to enter a pneumatic
conveying system discharging to the product silo. In addi-
tion, some dryer product entering the spill-back bin of the
dryer is transported in the same system to the product silo.
The dryer product is discharged to MgSO^ conveyer #1,
0-403, a 13 ft long unit with a 9 in diameter, solid flight
screw. The dried solids pass through a finger-type lump
crusher before entering the boot of the MgSC^ elevator de-
signed to handle 5 TPH of anhydrous MgSO.,. The product is
finally conveyed by MgS03 conveyor No.2, 0-405;(similar to
0-403, except only 6 ft long) to the MgSCU storage bin 1-401.
The storage bin, 1-401, was an existing ash storage
silo of tile construction. This silo, 81 ft high and 25 ft
4 in diameter, houses the MgSCU conveyor, 0-403, and the
pneumatic conveyor blower and dust collector in its top sec-
tion. The MgSCU screw conveyor unloading system, and the
control room and motor control center are in its bottom sec-
tion. Approximately a 33 ft high section of the silo, with
-35-
-------
a capacity of 16,000 cu ft is used for the temporary storage
of
. The flat bottom of this silo has four discharge open-
ings connected in pairs to two parallel discharge screw
conveyors (0-407 A&B). Each of the conveyors contain two
short sections of contra -rota ting 14 in helicoid flight
with a center discharge to a similar unit (0-407C), which
cross feeds the MgSO^ to a center discharge located over the
our-loading belt, 0-408, MgS03 Conveyer #3.
This 103 ft long, 18 in wide rubberized belt conveyor
exits the silo into an elevated gallery which carries the
product MgSOo to the truck station where it is gravity load-
ed through a flexible spout to the waiting hopper truck.
Oversize lumps (+3/4 in) of MgS03 are scalped off on an in-
clined screen in the discharge chute and pass to a hammer
mill, MgSO_ recycle crusher, located at grade for size re-
duction. The milled material is conveyed back to the belt
by a continuous flow conveyor.
3.2.2 Magnesia Regeneration System
Magnesium sulfite from the absorption system is received
at the regeneration facility, the .regeneration section
flow sheet is shown in Fig. 11. The salt is unloaded from
hopper trucks to the sulfite silo 1-501 through a 4 in pneu-
matic conveying system similar to that used for MgO at the
power station. The sulfite silo was an existing tile
structure, 20 ft in diameter x 60 ft high, which was water-
proofed before this use. Sulfite is unloaded from this silo
through a pair of inclined screw conveyors, MgSO_ conveyors
1 and 2, 0-509 and O-510, equipped with 12 in diameter heli-
coid flights, which carry the material from the below grade
discharge spout to the boot of the MgSO_ elevator.
This elevator, O-511, a continuous discharge unit 54 ft
high, is designed to handle 9 TPH of MgSO^. It discharges
to the MgS03 feed bin, G-506, a 6 ft diameter by 11 ft high
vessel equipped with a vibrating bottom, which also serves
as a surge bin for the weigh feeder O-514. This weigh feed-
er is continuously variable to a maximum capacity of 3.6 TPH
and discharges to the calciner conveyor 0-512, a 33 ft long
horizontal unit equipped with a 9 in diameter helicoid screw.
-36-
-------
Coke is also discharged to this screw which serves to mix
the two components as calciner feed prior to processing.
The coke is pneumatically conveyed to the coke feed bin
0-505, 9 ft diameter x 19 ft high, which also serves as the
storage bin for this material. It is metered to the calcin-
er conveyor by a weigh feeder which is also continuously
variable up to a maximum capacity of 2.5 pounds per minute.
The two streams (3 & 4 of Fig. 11) enter at points 9 feet
apart and mix in the remaining 22 ft section of the conveyor
before reaching the calciner elevator 0-513, a 43 ft high
centrifugal discharge unit. This elevator feeds directly to
the calciner R-501.
The calciner is a refractory lined, oil fired, rotary
kiln, 7 ft 6 in ID and 120 ft long. Rotational speed is
variable between 1.5 and 2 RPM using a variable diameter
pulley drive, and the kiln has a slope of 3/8 in per foot.
The calciner product empties into four tube coolers attached
directly to the shell of the kiln. The tube coolers are
equipped with internal flights to contact the existing hot
regenerated magnesia with incoming air which serves as
secondary combustion air in the calciner, thus cooling the
calciner product before it empties to the MgO conveyor #1,
0-503.
This is a 9 in diameter helicoid screw conveyor, 54% ft
long, which elevates the product from the calciner tube
cooler discharge seventeen feet to an enclosure housing the
MgO processing equipment.. The regenerated MgO is discharged
from 0-503 to a magnetic pulley for tramp iron separation.
The MgO then passes through a 1 in mesh vibrating screen and
enters a high speed pulverizer which reduces it to the fin-
ished grind shown in Table 18.
The cleaned, cooled, and pulverized regenerated magne"-
sia flows by gravity to the MgO elevator 0-504, a contin-
uous discharge design, 89 ft high, which brings the material
to the top of the MgO storage silo where it is loaded by
means of a horizontal screw conveyor (MgO Conveyor #2) O-505
equipped with a 9 in helicoid screw.
The MgO silo, 1-502, 25 ft in diameter and 45 ft high,
-37-
-------
'8
FIG. 10
DRYING- SYSTEM
-------
TABLE 5
STREAM PROPERTIES AND COMPOSITION
DRYING SYSTEM
STREAM
NUMBERS
Temp. ° F
Pressure IN WC
Gas Flow ACFM
CO
Dry Gas Ib/Min
HO Vapor Ib/Min
SO2 Ib/Min
SO0PPM (Dry Basis)
2
Fly Ash Ib/Min
Mg+S Compounds Ib/Min
Total Dust Ib/Min
1 . 2
GAS STREAMS
400
+ 1
20, 700
771. 0
124
0.8
0
1.06
1.06
|
70
0
10, 300
769
STREAM
NUMBERS
Temp. °F
Flow GPM
MgO Ib/Min
MgSO.,. 6H O Ib/Min
O ^
MgSO4- 7H2O Ib/Min
MgSO, Ib/Min
MgSO Ib/Min
Inerts Ib/Min
Fly. Ash Ib/Min
3.. .!_._ ._.. _:5._..__
LIQUID & SOLID STREAMS
132
10. 1
181. 7
. 10. 1
2.32
0. 273
Total Solids Ib/Min : 204. 5
HO Liquid Ib/Min
i Total Flow Ib/Min
10. 7
215. 2
I % Solids 95. 2
i
, Fuel Oil Ib/Min
i
f i
400
10.0
88.0
4.89
2.32
0. 273
105. 5
0
105. 5
100. 0
i ;
.
250
2. 3
18.8
:
-------
TO SUL.FUR.IC
a-HQT^t*^ VIBRATIN&
l\
SYSTEM
IH^MiCAL CONSTRUCTION CORPORATION
C'.-.'j\.r}>!r, :-?fS'GNiNG 'V'JO CONTRACTING ENGINEERS
U 5.A
-------
TABLE 6
STREAM PROPERTIES AND COMPOSITION
MgO REGENERATION SYSTEM
STREAM
NUMBERS
Temp :°F
Pressure PSIG
Flow GPM
3 4 56 7.8 9 10
11 12 13
LIQUID & SOLID STREAMS
i 300 250 160 160
1.5 91 2
i
Mg(HSO )2 Ib/Min
•
MgO • Ib/Min 8.24 38.35
MgSO0 Ib/Min 72. 50
0 '
MgSO. Ib/Min
4. 12 :
Inerts Ib/Min 1.89 . 1.81
Fly Ash Ib/Min
0.25 0.25
Total Solids Ib/Min 87.0 : 2 40.41 !
HO Ib/Min !
£i •
Total Flow Ib/Min
87.0 ; 2 40.41 760
Fuel Oil Ib/Min 12
I :
STREAM
NUMBERS
Temp. ° F
Pressure In. W. C.
Flow ACFM
90
119
0.88 |
1.5
0. 1
0.1
13.4
15.88
99. 2
150
125. 5
1045
150
Amb. 80
80 1.5 205
• -'1
1
65
I - .
1 2 14 . •
12 1708
GAS STREAMS '
100 70 100 i
-45 0 -45
2,980 , 2,140 640
Flow SCFM 2,460 : 2,100 529
Total Dry Gas Ib/Min
HO Vapor . Ib/Min
Total Wet Gas Ib/Min
SO Ib/Min
O Ib/Min
Dry Gas MW
206.2 161 39.0
9.0 j 0.8
215.2 39.8
44. 7 j
1.8 i
34.4 i 29 i 29
i
-------
is equipped with a vibrating hopper bottom, 0-506, and
elevated on a structural steel support for direct gravity
loading of the returning trucks.
The gas from the calciner, containing SO,, and products
of combustion, as well as a small percentage of excess air,
is first partially cleaned of particulates in the cyclone
dust collectors P-075. This is a dual cyclone array, design-
ed for a 1 in pressure drop. The collected solids are
returned to the calciner with the feed to the unit.
The partially cleaned calciner gas containing 8-10% SO2
is further cleaned in a venturi scrubber of Chemico's spec-
ial design, operated at a pressure drop of approximately 25
in of H20 where it is also adiabatically saturated. Next
the gas enters the separator tower section, which is an in-
tegral part of the venturi equipment. The lower section of
this 4h ft diameter vessel serves as a cyclonic liquid sep-
arator and the upper section, containing eight feet of 3%
in pall rings is irrigated with cooled weak acid to further
reduce the temperature of the gas to 100°F in order to main-
tain the acid plant's water balance. A slip stream of the
cooling liquor is stripped of dissolved S02 in the weak acid
stripping tower P-502, a small (15 in diameter x 14 ft high)
packed contactor.
The stripped SO joins the main gas stream and is duct-
ed to the acid plant in 18 in diameter, FRP pipe.
3.2.3 Acid Plant Modifications
The regeneration of magnesia, described in the previous
section, produces an off-gas from the calciner approximately
100 fold richer in S02 than the power plant combustion gases
treated in the SO2 Absorption System. The calciner gas is
of sufficient strength to be used as a feed for the manufac-
ture of sulfuric acid.
The small, sulfur burning acid plant (Sec. 3.1.2) re-
quired some modifications to enable it to accept the calcin-
er gas. Provisions were also made during these modifica-
tions allowing the plant to burn sulfur as an alternative
source of SO , or to operate on a combined feed from both
-42-
-------
combustion of sulfur and gas from the regeneration plant.
These modifications are shown in the Process Flow Diagram,
Fig. 12.
The principal element replaced in the acid plant was
the Main S0,j Blower, K-901. The original blower handled
only air required for the conversion of SC>2 to 363 and was
not designed to be gas tight (as required when feeding the
acid plant with gas containing 802), or capable of the
required control of suction pressure for operating the
regeneration plant. The replacement blower is an axial flow
compressor of 5,240 CFM capacity driven by a 200 HP motor;
It is capable of a suctiori pressure of up to 44 in of H2O
for the venturi pressure drop, duct and equipment losses,
and calciner draft at the regeneration plarit, and a 75 in
H20 discharge pressure for the acid plant. The unit was
designed to handle either air or a typical feed gas as shown
below:
Acid Plant Feed Gas
S02 6.8%
02 9.4%
N2 74. T/o
C02 8.1%
CO 0.9%
The blower was equipped with an adjustable recycle control
in order to accommodate the variable feed rate of the regen-
eration plant.
Another important element added to the acid plant was
the Cold Heat Exchanger E-901. This piece of equipment
supplies the heat to the incoming cold calciner gas equiva-
lent to that available in the gas when burning sulfur in the
sulfur furnace. This is necessary to ensure that the gas
entering the first mass of the primary converter is hot
enough to sustain the reaction. The exchanger, E-901, is a
vertical shell and tube unit 3 ft diameter x 9 ft high con-
taining 1700 sq ft of surface. Hot gas exiting the secondary
-43-
-------
1
\£2
_D
SFfR
'PS
I I ^
r^^
.r--ji_
1
%J
q
'I
J»A PRODUCT ACID
TO sroBAce:
•-TO »8*4 PUUP TANKS
-C » RETUBM
TO
ATMOSTMEP.C
ABSOBFTtOM
TOWER
FIG. 12
SULFURIC ACID PLANT — PROCESS FLOW DIAGRAM
AT ESSEX CHEMICAL CO. RUMFORD. R.I.
-------
converter enters the tube side on E-901 and heats the cal-
ciner gas passing through the shell side to 540°F.
In the acid plant, calciner gas enters (Stream 1 on
Fig.12) and is diluted with sufficient air through the air
filter for conversion and passes to the acid plant's drying
tower where it is contacted with 93% sulfuric acid to remove
the water from the gas. Entrained liquid is removed.in F-
901A, the Dry Tower. Any S02 absorbed by the 93% acid in
these towers is stripped in the new 93% Acid Stripping Tower
(F-903), a 30 in diameter x 15 ft high ceramic tower con-
taining a 10 ft bed of 1^ in saddles,and returned to the
main gas stream. The small amount of additional air is used
for trim of the oxygen concentration. The gas is pressuriz-
ed by the SC^ blower and enters the Cold Heat Exchanger,
passes to the Converter Heat Exchanger E-902 and finally
enters the converters H-901 A & B. The heat balance for the
conversion step is maintained by using the gas from the
second mass of the Primary Converter, H-901A, to heat the
gas from the Cold Exchanger, and the gas leaving the fourth
mass of the secondary converter to heat the incoming feed.
Next, the stream containing 6.7% SCU is contacted with
98% H2S04 in the Absorption Tower F-902 A & B. The gas
streams from these towers are first demisted in R-902 A & B,
then the remaining SCU is removed in the Tail Gas Scrubbing
section shown in Fig. 13 to reduce the concentration of SO-
leaving the plant to allowable levels.
The absorbant used in the Tail Gas Scrubbing section is
NaOH solution.
In order to allow a rapid change to sulfur burning,
when the plant is operated on 100% calciner gas, the sulfur
furnace was equipped for oil firing. The combustion pro-
ducts from this operation are vented to the atmosphere
through a short stack which could be bypassed. In normal
operation, however, sulfur was burned concurrently because
of a low MgSCU feed rate. This procedure had the advantage
of allowing a rapid change should the calciner gas flow be
interrupted. The alternative was possible because of the
several dampers which had been incorporated into the
ductwork.
-45-
-------
KJOTE.:
* NOt-OH, WATER MAKE-UP TO
^.epFUjeNJT
&CIO T*UANT
IAJATVR
TAMK.THUCK
lin.l. iriMw
MAXIMUM f|,gV>J ^
r=N
COHTi
Y
RCCVC.US
-&?»-s-s
STCAM
ee> IV-
OUMP
TAIL. GAS SCRUBBING SYSTEM
-------
TABLE 7
STREAM PROPERTIES AND COMPOSITION
TAIL GAS SCRUBBER SYSTEM
STREAM NUMBERS
Temperature ° F
Pressure IN WC
Flow ACFM
Dry Gas Ib/Min
H2'O Ib/Min
Wet Gas Ib/Min
SO, (Design) Ib/Min
^ SO2 PPM Dry
-1 By Volume
I
160
0
4, 250
279
0
279
3. 0
5000
2
165
14
4140
279
0
279
3. 0
5000
3
77
0
5600
276. 3
5.6
281. 9
0. 3
500
STREAM NUMBERS
Temperature ° F
Flow GPM
NaOH Ib/Min
Na2SO3 Ib/Min
H2O Liquid Ib/Min
Total Solution Ib/Min
4
77
152
299
1266
1565
5
Amb.
0. 6
3.75
3.75
7. 5
6
77
3.0
5. 9
23. 6
29.5
7
Amb.
3.0
24. 6
24. 6
8
77
76
149. 5
133
782.5
-------
4.0 DESCRIPTION OF WORK PERFORMED
In order to proceed with the development of the process
in the large prototype, the program was organized into a
series of tasks as follows:
a) Construction
b) Pre Start-Up
c) Start-Up and Break-In
d) Planned Operational Tests
e) Operational (Long Term) Tests
A schedule was established for the completion of these
tasks; however, some slippage.from this schedule was exper-
ienced as outlined in the following sections. A comparison
of the actual and predicted schedule showing the dates of
initiation and completion of the important tasks is shown
in Fig. 14.
4.1 BOSTON EDISON
4.1.1. End of Construction Phase
Labor shortages in some crafts and material delivery
delays were the main problems encountered which resulted in
delaying the construction schedule so that tie-in to the
scrubber could not be made before the end of 1971. The two
principal items which had to be completed after the tie-in
was made were the precrete work in the ducts and the scrub-
ber lining application. The precrete work was started to-
wards the end of January 1972.
The cold weather at that time of year made it necessary
to enclose the absorber vessel in canvas and provide heat in
the enclosure so that the FRP liner.could be applied at the
proper temperatures. Application of the liner was started
by the middle of February and was completed by the first
week in March.
The check-out of the dryer, centrifuge and air compres-
sor was completed by the middle of February and installation
of the mist eliminators in the absorber was finished by the
-48-
-------
NEW ENGLAND SO,, CONTROL PROJECT
SCHEDULE DATE
— SIQ-MIUGj-
OF COUTRACT
6-3O-70
START
COM5TRLJCTIQM
—
EMD_jCONJ5J_RyCT 10 U
7-50-71
PRE ^>TA_RT UP
COMPLETION
IO-I5-TI
-START-UP^
BREAKIM COMPLETIOk)
-START
COMSTRUCTIOM
I-II-7O
PRE 5TART-UP
COMPLETiON)
,PLAMMED OPERATIONAL
TESTIW5 COMPLETIOKJ
f-3 0-73
OPERATJOWALO-OKJG TERM)
TESTS COMPLfTJOKJ
i 2-31-73
PREDICTED -
APT! 1 A 1
/*U 1 UAL
OPFRATIOKJA
COMPLFTIOIU
OF COMTRACT
ENJD CONJSTRUCTIOK)
START-UP f
BREAK1M COMPLETION)
PREFORMAMCE
TESTIIpq-
6-5O-7O
OPERATIONAL
TESTS COMPLETIQM
-------
middle of March.
A mechanical check-out of the dampers prior to the
removal of the blanking plates isolating the scrubber show-
ed that different linkage and stronger operators were requi-
red before the dampers could be put into service. This work
was accomplished during the first three weeks in April. The
initial MgO charge was also received during this period. By
the end of April 1972 mechanical check-out had been comple-
ted ending the construction phase.
4.1.2 Pre Start-Up Period :
The trial operational period was started on April 16,
1972 periodically taking gas to the absorber with the boiler
on low sulfur fuel. On April 26, the boiler was switched to
high sulfur fuel and, for the first time, solid MgSCU was
removed from the system. During the following few days sev-
eral new problems were encountered. The solutions to these
problems were to consume considerable time and effort during
the balance of the year.
/
While these problems did cause delays in the demonstra-
tion of the prototype, it had been shown in the first brief
period that the scrubber was capable of performing its pri-.
mary function, removing sulfur dioxide from the power plant
flue gas. SC>2 removal efficiencies of 90% or greater were
measured during some of these early runs.
The first and most frustrating of the problems was the
change of crystal type and size in the scrubber slurry which
made subsequent centrifuging and drying difficult.
Typically, the scrubber would be put on the line when
the solids content built to the concentration for initiation
of centrifuging and the slurry, contained large crystals of
MgSOo. This, in turn, produced a dry, sandy centrifuge
cake. After a short period of continuous operation, usually
on the order of 8-12 hours, the centrifuge cake rapidly
changed from a sandy to a fluid, creamy consistency. It was
found that in the initial operational period of the system,
the crystal distribution in the slurry was either larger
than 200 mesh or smaller than 325 mesh with little material
distributed between these two sizes. (A 200 mesh size was
-50-
-------
adopted as the dividing line between describing larger and
smaller crystals.)
The crystal size found after the change in centrifuge
cake characteristics was generally less than 325 mesh. This
same change from a coarse to a fine crystal could take place
even though the scrubber was> not in operation. If the
scrubber was placed on line and then had to be' taken off the
line before centrifuging, it would be found the following
day that most of the crystals in the slurry would be less
than 200 mesh.
The change also appeared to be associated with temper-
ature, since the crystals appeared to increase in size if
the scrubber slurry were allowed to cool. There was also
evidence that the level of magnesium sulfate in the system
had some influence on the size of the crystal formed.
Early investigations indicated that the coarse crystals
were the hexahydrate form of magnesium sulfite, and the
small crystals were the trihydrate form. Some work was done
in an effort to find a way to keep the crystal mass in the
hexahydrate form, however, the factors that control the
crystal type had not been clearly defined.
Operating policy was still to "shut down" when the
centrifuge cake consistency changed. All attempts to con-
trol the system to produce +200 mesh crystals, by lowering
pH, adjusting slurry solids, etc. were unsuccessful. At
this point conaideration was given to a process change in
which bleed stream to the centrifuge would be cooled and
then retained in a holding tank till a desired crystal size
was attained. An air-slurry cooling tower was designed and
installed and a thickener and associated pumps brought to
the site to pilot this operation on a five gallon per minute
scale. This scheme was never tried, however, principally
because the scrubber would have to operate in some contin-
uous mode to provide a feed for the pilot plant and no more
than a few hours of operation at a time could be attained.
The practical solution to this problem was associated with
the solution to a completely different one which arose in
the first operation.
•51-
-------
When the centrifuge cake was first fed to the dryer it
was found that most of the material could not get through
the dryer, rather, the dried product was entrained and
carried out of the dryer with the off-gas stream. Through-
out May, June, and early July of 1972 a series of modifica-
tions were made to the dryer design to allow a reduction in
dryer gas velocity while still providing a drying capacity.
The most significant of these was the removal, in a series
of changes, of most of the flights in the dryer drum. In
addition, the feed end dam was increased in height to re-
duce spill back and the top of the feed screw housing in
the dryer was cut away to allow cake which was adhering
to this housing to fall into the screw and be reconveyed
to the drum.
As a result of the dryer modifications, all the flights
had been removed from the feed end of the drum,and on some
occasions is was noted that the dryer feed would agglome-
rate in the feed end and pass' through the unit in the form
of lumps. These lumps ranged from balls about one inch in
diameter, to larger agglomerates up to 18 inches in dia-
meter. These large agglomerates caused problems in the
material handling equipment after the dryer, and even the
smaller pieces could not be handled by the pneumatic un-
loading system when the magnesium sulfite was shipped to
Rumford so that the policy of shut down of the F.G.D.
system as soon as troubles developed was continued.
A separate problem which hindered the early runs was a
rapid build-up of solid material on the damper blades and
frames in the fan and outlet ducting. This build up made
the dampers difficult to open when putting the scrubber on •
the line. Several minor modifications were made in an at-
tempt to overcome this difficulty, but it was finally neces-
sary to replace the pneumatic operators with more powerful
ones developing sufficient torque to break the dampers loose
from the accumulated material.
During this first period the dryer burner was still
another problem source as it was subject to frequent flame-
outs and a tendency to coke in the burner block.
All of these difficulties limited the duration of oper-
ating periods to less than 24 hours and resulted in frequent
-52-
-------
"dumping" of the scrubber slurry as attempts were made to
maintain the "sandy" centrifuge cake.
4.1.3 Start-Up and Break-In Period
By early July 1972 it was evident that the crystal type
and size produced in the absorption step could not be con-
trolled by any simple changes in the equipment or operating
conditions. The technique of allowing the scrubber slurry
to cool to form larger crystals had had sporadic success and
was not a satisfactory solution. 'Any significant process
changes would have required extensive pilot testing, and
no facility for this testing was available.
The intermittent operations of the previous month did
indicate that successful operation might be possible if the
centrifuge cake could be made to granulate, if there were no
adherence of the centrifuge cake to the dryer shell, and if
the agglomerates could be handled in the subsequent equip-
ment.
A test was planned in which the dryer temperature pro-
file would be adjusted to give the maximum feed end skin
temperature possible, consistent with good operation, to see
if the build up problem could be eliminated by the formation
of agglomerates.
Feed was to continue to the dryer when the consistency
of the cake changed unless there was an indication of heavy
build-up or other serious problems. This test run was
started on July 18, 1972.
The operation proved to be successful and the system
ran for three days. During this time other unrelated prob-
lems were encountered causing two brief interruptions. The
first was flame-out in the dryer resulting in a 12 hour
interruption. A breakdown of the MgSO3 weigh feeder in the
materials handling train caused the second six hour stoppage.
During this second stoppage, the flue gas was diverted from
the scrubber using the dampers, but all other equipment was
kept operating to facilitate return to service of the system.
After a temporary chute was installed to bypass the MgSO-j
weigh feeder which had caused the trouble, the unit ran for
an additional thirty hours before a tear in the dryer feed
-53-
-------
conveyor trough forced a termination of the test run. This
breakthrough operation demonstrated that no major changes
in equipment or process would be required to continue with
the planned operational program. Additional improvements
and modifications were required, however, before the system
could operate for the long durations necessary to determine
availability and commercial applicability.
Corrosion in areas of the duct was also noted during
this time. These were temporarily patched while plans were
made to protect the parts of the gas flow system which had
not originally been protected by ceilcoat or precrete.
These difficulties limited operations to an intermitt-
ent basis for several weeks while corrective action was
taken.
The scrubber again was put on the line late in August,
and operated continuously for five days, except for a brief
6 hour period when the boiler was out of service for minor
repairs. During this run, all equipment functioned satis-
factorily with the exception of the dryer burner. Although
the burner was no longer subject to flame-outs, now not
enough fuel could be burned to provide enough heat input to
dry the centrifuge cake properly. In order to compensate
for this the draft through the dryer was increased, and in
turn, resulted in excessive amounts of dust being carried
out of the dryer and into the boiler stack. The situation
gradually became worse, and on August 31st, it was decided
to suspend operations until corrective action could be taken.
During the August run, the scrubber system was shown to
be capable of adapting to the changing boiler load which
typically was about 150 MW in the daylight hours and 30 MW
at night. The scrubbing efficiency observed during this run
was also consistently high.
September and part of October were spent in making ca-
pacity checks on the dryer burner, velocity checks on dryer
gas flow, and entrainment checks on the scrubber itself, and
no operations were conducted. System modifications were
also completed during this period including installation of
a lump breaker for the dryer product, a dust collector for
the magnesium sulfite weight totalizer and installation of
-54-
-------
a low sulfur fuel supply to the dryer burner as the burner
capacity checks in early September had shown that the unit
did not burn its rated amount of high sulfur fuel.
This latter problem was attributed to the properties
of the station's fuel. Nominally a #6 oil, it contained a
relatively large amount of volatile material so that when
the oil was heated to the temperature necessary to obtain
the proper viscosity for atomization in the dryer burner,
the volatile component would vaporize with the subsequent
flashing restricting the capacity of the burner. To over-
come this limitation, piping was installed to allow the dry-
er burner to draw upon the station's low sulfur fuel supply.
Later, additional work was done on the burner which allowed
it to operate at a higher pressure, and thus prevent the
flashing of the volatile fraction of the oil to make it
capable of operating at rated capacity on either fuel.
The dryer gas velocity checks showed that high dryer
draft caused the excessive dust carryover experienced on
the run in late August. These tests also showed that even
under normal operating conditions the dust carryover was
higher than desirable as the dryer was originally designed
to process a much coarser material than the fine material
which was formed in the prototype system.
The system was returned to operation in mid-October
following adjustment and debuging the new equipment which
had been installed. A forty-hour run filled the magnesium
sulfite storage silo. Magnesium sulfite was hauled to the.
calciner operation in Rumford and the first 74 tons of re-
cycled magnesium oxide was returned to Boston and added to
the magnesium oxide silo. The scrubber was put back on the
line on October 21st, and within 12 hours, recycled magne-
sium oxide was being fed to the scrubber system. At that
point, severe plugging occurred in the MgO slurry piping,
and the system had to be taken off the line.
It was found that this first recycled magnesium oxide
varied considerably in its properties. Much of the material
was over calcined (hard burned),and the coarse particles did
not slake when the MgO was slurried. After clearing the
piping system and installing flush-out connections, for fut-
ure use the scrubber was again returned to operation,but
-55-
-------
plugging of the MgO slurry system continued to be a problem.
Plugging of the MgO slurry system continued after the
calciner operating conditions were modified to produce a
softer, satisfactory burn material. Indications were that a
grinder was required to reduce the particle•size of the cal-
cine, so that a slurry could be produced free of large part-
icles and grit. Enough data were obtained during this per-
iod to show that the pH of the scrubber slurry could be
maintained at the proper level, and that the pH and free
MgO level was sufficient to remove 90% of the incoming S02-
During the month of November, a scheduled outage of the
#6 boiler was taken for maintenance and repair. During the
same period, maintenance was planned for the S02 abatement
system. Principal items of work done were precreting of the
fan housings to prevent the corrosion caused by condensation
in the duct work. The recycle pumps were inspected and sev-
ere erosion-corrosion was noted on the impellers and wear
plates. The impellers were replaced from spare parts and
orders placed for stainless replacements. All the other
pumps were inspected at this time. Additional steam tracing
was installed where necessary for freeze protection during
the upcoming cold weather. A preliminary attempt to heat
the MgO slurry tank was made. A DuPont S02 analyzer was
added to the system to monitor the inlet and outlet sulfur
dioxide concentrations, as the analyzer installed origin-
ally proved to be unsatisfactory as it plugged up during
brief periods of operation.
An inspection of the scrubber at this time showed no
heavy deposits, but there was a light build-up of solids in
the vicinity of the tangential nozzles. The scrubber lining
was found to be intact and in generally good condition ex-
cept for some wear on the vertex of the torus where over-
lapping spray patterns had caused abrasion of the FRP lining.
A serviceable Micro-Pulverizer was located at the acid
plant and evaluated. These tests indicated it would produce
a ground product suitable for reuse in the system. This
unit was a 2D-H model with a 10 HP drive.
By December, the pulverizer had been installed at the
discharge of the MgO weigh feeder at Mystic Station and an
-56-
-------
attempt was made to return the system to operation using
recycled MgO. The run was aborted after a brief operating
period when the pulverizer plugged.
This plugging problem was caused by tramp material and
clinkers of calcined MgO. Most of the tramp material was
uncalcined MgS03 which had not been completely unloaded from
the trucks prior to loading with MgO.
Vibrating screens were installed to scalp off the tramp
material prior to introduction into the pulverizer feeder,
but because of limited clearances around the MgO feed system
and high maintenance requirements, this was not successful.
During the few short periods of operation in December,
heavy build-up was also experienced in the dryer drum. The
dryer feed caked on the walls to the extent that a shut down
of several days duration was required to clean it.
Also during December, Boston Edison Company asked the
local and state authorities for an extension of the time
limit for the installation of pollution abatement equipment
on their new Mystic #7 unit which was under construction.
The appeal was based on the demonstrated progress of the MgO
process which had been gained in the operations at Mystic
Station since April. The appeal was granted in January 1973.
In the first quarter of 1973, several major modifica-
tions were made based on the previous operating experience
to improve the "on stream" time of the system. In January,
the dryer off-gas was re-routed to the inlet of the scrubber
for particulate matter emissions control after numerous
other schemes had been evaluated and rejected for control.
This selection was based on the conclusion that the 6 inch
pressure drop across the F.G.D. systems venturi scrubber was
sufficient to control the solids carried over from the cyc-
lones of the dryer. The pulverizer was also moved from
Boston to Rumford and installed at the discharge of the mag-
nesium oxide product conveyor.
Over-load of the dryer feed screw which had caused some
shut downs was associated with an excess of magnesium oxide
in the centrifuge cake. Tests showed that slaking of MgO
could be improved by heating the MgO slurry tank. A steam
-57-
-------
sparger was installed in the tank and slurry temperatures up
to 190°F attained.
Other work undertaken during this period included at-
tempts to remove tramp material recycled to Boston by in-
stalling screens at the MgO weigh feeder discharge. MgO
proved difficult to screen as it adheres to surfaces and
wires, blocking even the coarsest screens tried, \ inch
mesh. A pre-mix tank was substituted to allow the grit to
settle while the ground MgO is fed to the slurry tank as
overflow. As a precaution, a stand-pipe was installed on
the discharge nozzle of the MgO slurry tank allowing agglo-
merates to be blown down on a regular basis.
The improved operations during this period proved the
efficacy of the modifications; however, dryer build-up con-
tinued as the most frequent cause of shut down. In April,
a final series' of changes and additions were made to improve
operations. These were: installation of "hammers" on the
dryer, a pneumatic conveying system to take the cyclone
underflow and dryer spill-back directly to storage, and lengthen-
ing of the dryer conveyor in the dryer. A modification
to the pre-mix tank was made to prevent blockages of the
down" comers and new design damper operators were installed
in an attempt to overcome the old sticking damper problem
which had delayed numerous start-ups throughout the program.
In a preliminary run at the end of April most of the old
problems appeared to be corrected,and it was considered pos-
sible to enter into the next phase of the test and develop-
ment program to provide data which would lead to improved
operations and optimization of the system.
4.1.4 Planned Operational Testing
Initial operations in May 1973 indicated a substantial
improvement in the process reliability. Slaking of the MgO
slurry at a temperature of 180°F reduced the consumption of
MgO required for pH control and proper S02 removal. Analy-
sis of the centrifuge product indicated a substantial reduc-
tion in the amount of free MgO indicating a greater utiliza-
tion of the recycled alkali in the absorption step.
A labor problem at the station in May prevented further
operation of the system during that month. Upon resumption
-58-
-------
of work in June, the dust handling system was completed and
the system was started up.
During the month of June nine interruptions occurred.
Six of these outages were caused by boiler related problems
but the other three were failures of the F.G.D. System.
These three were: first, a stalled dryer feed conveyor
caused by neglect to clean the dryer feed screw before the
system was returned to operation; second, a dryer upset
caused by a large agglomerate jamming the dryer product
screw;and finally, a shutdown because of high level in the
dryer product silo. The latter resulted from insufficient
equipment to transport MgSC>3 at the rate it was being pro-
duced.
In all, during June, 376 hours of operation were logged
and scrubber availability to the boiler increased to 68%.
Operations of the system continued into July. The main
difficulty to operations being the constant attention neces-
sary to prevent the MgO pre-mix tank from plugging. On the
4th of July, the #6 boiler developed an economizer tube leak
and was shut down. During this outage leaks in the recycle
piping and valves were patched,and a revised pre-mix tank
was substituted. A manual valve was also substituted for
FCV-5 controlling the flow to the centrifuge as the pinch
valve, which had been part of the original design, proved
to be unsuitable. The system was returned to service on the
6th but deposits on the dryer feed screw and in MgO pre-mix
tank which had to be removed on a regular basis by the oper-
ators and leaks which continued to appear in the recycle
piping caused operating difficulty. In addition, a blockage
(later found to be a plastic bottle) occurred in the centri-
fuge feed line restricting flow. A boiler tube failure on
the 14th terminated this run.
A stack inspection was performed at this time, and it
was found that the upper 10 ft of the inner stack was damag-
ed (this damage was unrelated to the F.G.D. system). The
boiler was returned to service on the 15th, and the F.G.D.
system ran till.the MgO supply was exhausted on July 19th.
When sufficient MgO was received, the plant was returned to
operation and continued in service through the 27th, with
two boiler tube failures and some absorber system related
-59-
-------
problems causing a number of short interruptions during that
period. A scheduled boiler outage was taken on July 27th
and maintenance on the boiler and the absorption system oc-"
cupied the rest of July and most of August.
No. 6 boiler was back on the line on August 28th, and
the absorption system on the 30th; however, difficulty was
again encountered with sticking booster fan dampers which
was not corrected till September 6th. A series of attempts
to restart the system were aborted by:
1) A leaking centrifuge oil cooler
2) Dryer feed screw binding
3) MgO feed problems
On September 15th the system was returned to operation
and during the first three day period several minor problems
were corrected with the F.G.D. system in service switching
from high to low sulfur oil. Smooth operations were obtain-
ed from the 18th through 21st but again a low MgO supply
forced a shut-down after that time.
The inventory was replenished and the system restarted
on the 24th, but after a short period it had to be shut down
because of a slurry leak which developed at the centrifuge
feed port. The cause was fractured centrifuge feed pipe,and
four days were taken to correct this problem as the centri-
fuge was also dismantled for an inspection.
Operations continued intermittently during the first
half of the month of October, with only 78 hours of system
operation achieved by the 15th. The interruptions were a .
series of breakdowns of process equipment interspersed with
a series of boiler related failures. The F.G.D. problems
were•debugging the new cyclone underflow conveying system
and in the MgO pre-mix system.
A 40 hour run started on the 16th was shut down because
of a boiler tube failure. Repairs to the boiler were made
but the units' return to operational capacity was limited by
•a failure of the refractory in the dryer burner block which
prevented firing it at its full rate. The system was
-60-
-------
switched temporarily to low sulfur fuel to control solids
level but another boiler tube failure on the 18th termin-
ated the run,and the system was idle til the 20th. From
the 20th until the end of the month, the system operated
for 106 hours with one interruption, a 17 hour outage caused
by depleting the supply of MgO. After transferring a fresh
supply of MgO the system ran for an additional 87.5 hours,
but was forced to shut down after the MgSO silo at Mystic
filled to capacity with resultant damage to the materials
handling equipment.
While the MgSO was being transferred, the dryer feed
screw gear box and the product bucket elevator were repair-
ed, and operations resumed on November 9 and continued
through the 15th with only a 4.5 hour interruption for lack
of MgO. From the 16th through the end of the month the unit
was down, as the MgO supply had been exhausted, and as a
result of a shortage of fuel oil for the calciner. Over-all
utilization for the month was 24% based on available hours;
utilization for the week ending November 16th was 84%.
During the short operating period in November a stack
testing program was conducted and a second series of scrub-
ber emissions tests was run. A separate particulate matter
emissions test was also conducted by the Environmental Pro-
tection Agency.
Operations started again on December 7th after the fuel
oil situation had been resolved, and during the next two
weeks continued sporadically with numerous interruptions
caused by malfunctions of the cyclone overflow conveying .
system and leaks in the main recycle pumps.
The pump maintenance could not be accomplished because
the plug valves to isolate them were severly eroded, requir-
ing that the system be drained to work on it. A planned •
outage of the absorber system was taken to replace the pumps
and plug valves and to patch several other leaks which had
developed in the top piping. This limited plant operations
to 86 hours during the month of December. The system was
returned to service on January 10, 1974 but again a series
of mechanical problems, including an outage of the boiler,
caused intermittent operation of the absorber and numerous
switches between high and low sulfur fuel oil in the boiler.
-61-
-------
By January 25th, most of the new problems were corrected and
testing was resumed, but after 30 hours a boiler tube fail-
ure required a two day repair outage. When the unit was
started up again difficulty was experienced in controlling
the recycle solids concentration in what appeared to be a
reduction in centrifuge separational efficiency. Operations
were terminated to investigate this,and the centrifuge was
dismantled for inspection and repair.
During January, the system was operated an additional
147 hours to give a total for December and January of 233
hours. During this period inventory control was initiated
and the inventory showed a loss of 80 tons of MgO in the two
months operations.
The absorber system was not returned to service until .
the 19th of February, 1974 while the centrifuge was being
repaired. During this same outage several maintenance items
were completed including the installation of a heavy duty
weigh belt on the MgO weigh feeder, venting the dust collec-
tor in the MgSO silo, replacement of couplings on the re-
cycle pumps, ana the installation of several sample nozzles
for the stack testing to be conducted later in the month.
The system was started up again but ran only two days
before depleting the supply of MgO. The coordination of
shipments of material between the two plants was made the
responsibility of the chief Chemico operator to correct this
recurrent problem. Because of the decision to use only re-
generated alkali while conducting the performance tests
there were no further operations until the 25th in order to
conserve the MgO inventory. During this wait the by-pass
dampers were calibrated and tested to prepare for a new mode
of operation in which a part of the flue gas would be bypas-
sed so that the system could be operated at its design gas
flow. Previous testing had established that the system was
handling as much as 650,000 ACFM when the boiler was being
operated at full load (as compared to the design 440,000
ACFM) because of leakage into the ducts ahead of the boiler
ID Fan (Table 8) .
Also, for the tests, the dryer off-gas which had been
routed to the absorber system in a corrective modification,
was revented to the stack in order to eliminate this
-62-
-------
TABLE 8
FLUE GAS DESULFURIZATION - MYSTIC #6
ABSORBER GAS FLOW
Date
ll-25-70(1)
11-27-70^1)
10-11-73
10-11-73
(2)
10-14-73 (2)
3-5-74(3)
DESIGN GAS FLOW
6-1 Duct
Boiler Total Gas
Load Air Flow
(MW) (%} (ACFM)
142 153
142 163
150
149
149 }
151 335,411
LOW
6-2 Duct Total
Total Gas 6-14
Air Flow 6-2
(%} (ACFM) (ACFM)
153
160
573,000
523,807
645,000
322,796 658,207
440,000
(2)
(3)
Boston Edison Company Communications
Custom Stack Analysis Co. Report
York Research Corporation Report
Y-8419, 4/11/74
-63-
-------
additional solids emission load on the scrubber,as it had
not been part of the original design.
The system was started on February 25th for the first
performance tests of the program, but a boiler tube leak on
March 1st forced a shutdown.
Also during this period, additional tests were conduct-
ed over two days using humidification sprays to determine if
pre-quenching could influence the hydrate formed, and whether
there would be an improvement in the outlet dust loading
The results obtained from the humidification test indicated
that there was no effect from their use.
Repairs to the boiler were completed on March 5th. The
F.G.D. system was put in operation, but again a boiler tube
failure forced termination the following day. The F.G.D.
system was restarted immediately after boiler repairs were
completed and ran satisfactorily for two days before a third
boiler tube failure forced an outage; however, the perform-
ance testing was completed.
4.1.5 Operational Testing
Operations during the test period indicated that reduc-
tion of gas flows nearer to design in the system, coupled
with the extensive repairs and numerous improvements and
modifications, had resulted in a more reliable and control-
lable system. With the substantial experience and operating
information gained in the preceeding two years, it was deci-
ded to start a period of continuous operations 'at a fixed
set of conditions. Prepatory to starting this new phase the
blanking plate, which had been removed from the dryer gas
stack breaching for the test work, was reinstalled so that
the scrubber controlled this particulate discharge source.
The F.G.D. system was returned to operation on March
12th and ran smoothly. On the 15th another boiler tube
failure occurred, repairs were made and operations resumed
on the 18th but were stopped on the 19th by another failure.
Boiler repairs were completed by the 21st,and the F.G.D.sys-
tem ran through the 26th when again a boiler failure forced
an outage. The system came on as soon as boiler repairs
were completed on the 28th, but was forced out because of a
-64-
-------
boiler tube leak the following day.
During this period, there were also four malfunctions
of SC>2 abatement system equipment. The first, on the 18th,
was a broken chain on the cyclone underflow star valve.
These repairs took three hours, and during this time the
system was kept on the line on low sulfur fuel. The second
was a blockage of the dryer screw conveyor on the 22nd (low
dryer temperature allowed wet material to get into the screw
conveyor). This was corrected.in 4h hours, again, without
taking the system off the line. The third incident did
force the unit off the line and was caused by a failure of a
small agitator in the MgO pre-mix tank. A replacement was
on hand but was found to be defective,and an outage of 11
hours had to be taken to repair the agitator. The fourth
interruption was a four hour outage to replace a broken
chain on the MgO weigh feeder drive. ,
When the F.G.D. system was being restarted on the 30th
a bearing failure was noted in the MgS03 elevator,and no
operations were conducted for the first four days of April,
while waiting for repair parts. The system was started on
the evening of the 4th and ran continuously through the llth
A carbon deposit in the dryer burner block caused a hot spot
and the system had to come down for one day for repairs, but
was again in operation on the 13th. It ran continuously
through the 19th, when a boiler tube failure forced it o.ff
line. Boiler related failures on the 22nd, 23rd, 24th, 26th,
and 29th of April resulted in interrupted operations; how-
ever, they also further demonstrated the capability of the
F.G.D. system to cycle with the boiler. Despite the afore-
mentioned problems April operations were smoother, and. one
hundred percent availability was achieved for the two week
period from the 6th through the 19th. In all, during April,
the absorber system operated for 471 hours out of 585 hours
available (because of the numerous boiler failures) for an
81% utilization.
One hundred percent availability of the system was
again achieved during the first two weeks of May. Then the
system was forced down on the 14th when the MgO supply was
exhausted because of acid plant outage. The acid plant
outage necessitated shut down of the regeneration section.
MgO production was resumed in late May after the acid, plant
-65-
-------
had been repaired.
Before starting again three additional days were
taken to clear the system of MgO and MgSCU prior to start-
ing an MgO. Loss Identification and Inventory Control Program,
During the month of May, 1974, boiler tube failures on six
occasions were the major cause of program interruptions.
The boiler and F.G.D,, system were operational from the 4th
through the 5th, the llth through the 14th, and from the
28th through the 30th, a total of 280 hours out of 488
available.
One hundred percent availability of the system was
again achieved during the first two weeks of June. The MgO
Loss Identification Program was concluded on June 18, 1974
(the MgO feed bin was deliberately run empty to finish this
test). S02 addition runs were made on June 13th and June
14th, with the injection of four tons of liquid SO« into
the system to increase inlet concentration to its design
value. But overall availability for the month was still
.limited because of seven boiler tube failures. Two hundred
eighty-eight hours of F.G.D. system operation were achieved
in 359 hours of boiler operation for an 80% availability.
The program contract termination date was June ,30th,
1974 but operations were actually concluded on the 26th
when the dryer feed conveyor failed.
4.2 REGENERATION - RECOVERY OPERATIONS (ESSEX CHEMICAL
PLANT)
4.2.1 End of Construction Phase
The plant construction had been completed on December
21, 1971; however, operations could not be initiated as the
magnesium sulfite needed for feed was the product of the yet
to be completed SO2 abatement plant. In the interim the new
main blower was commissioned for acid plant operations in
January, and instruction of operators and the first firing
of the calciner to "cure" the refractory brick work was done
in March.
During this preliminary work the first operating prob-
lem uncovered was a low draft at the firing hood caused by
-66-
-------
excessive leakage of air through the seals and air intakes
of the calciner. This was a potential process problem, as
the calciner had to operate with a neutral or reducing at-
mosphere in order to reduce MgSO. which was produced in the
scrubbing step.' Considerable work was expended in tighten-
ing up seals and reducing the air intake leakage. (This did
provide some improvement; however, when feed was obtained
for processing excess oxygen content in the calciner atmos.-
phere still remained as a problem).
Another problem wh ich was uncovered in this period
resulted when firing oil in the cold calciner on start-ups.
Hydrocarbon vapors succeeded in passing through' the scrubb-
ing equipment installed for gas clean-up of particulars
matter and entered the acid plant where the hydrocarbon
caused contamination and blackening of the product sulfuric
acid. A 1500 CFM fan was obtained, ducted tc the calciner
and a stub stack discharge erected in order to provide a
separate venting system for the calciner on start-up. (This
vent is not used when MgSO-j is decomposing in the calciner}.
While waiting to receive dryer product from Boston for
actual start up, a high noise level condition from the main
blower air intake was also corrected by relocating the in-
take and erecting a sound baffle around the blower itself.
4.2.2 Pre Start-Up Period
The first dryer product was transported from Boston on
June 20, 1972. The necessary coke had been obtained prior
to this time and was stored in the coke silo.
Pre start-up was limited to testing the conveying sys-
tem with the material on hand. Some work was undertaken in
matching the elevator capacities as the bulk density and
granulation of the. material produced at Boston was. substan-
tially different from that for which the equipment had been
designed based on the original pilot plant work.
The presence of lumps in the feed material caused dif-
ficulty in material handling at the regeneration plant in
both the pneumatic conveying system and the weigh feeder.
(The solution to this problem involved the determination of
proper operating conditions for the dryer and was not
-67-
-------
implemented til'later in the program).
4.2.3 Start-Up and Shake-Down Period
Start up of the system was not initiated until suffic-
ient dryer product was in inventory in insure five days of
continuous operation. Material from the Flue Gas Desulfu-
rization system was first fed to the calciner on August 3,
1972. Problems of low draft in the unit and high oxygen
concentration in the kiln atmosphere were experienced
immediately.
Natural draft occurs in the calciner because of its
length and pitch. This draft resulted in high seal leak-
age at the small gas flows required for proper kiln opera-
tion. This caused "puffing" at the feed hood and excessive
oxygen content in the off gas. The new corrections for
leakage included removing and machining the firing hoodseal
to provide minimum clearance on the hot shell. The draft.,
doors were sealed, all openings closed and the product
discharge ports closed by 50%. This work was done by early
September and resulted in a lessening but not elimination
of the problem.
During this same period operating policies were deve-
loped to overcome the problem of contamination of the pro-
duct acid by hydrocarbon vapors from the regeneration sect-
ion. Principally this was accomplished by shunting the
combustion gases to the vent stack after the mid kiln temp-
erature fell below'1000°F.
Another serious problem was caused by dusting in the
calciner which obscured the flame from the fire, control
monitors and caused frequent shut down of the kiln on a
false "flame failure" signal. Several approaches were tried
to overcome the numerous "flame.outs" which occurred because
of this. These included installation of a second flame
scanner for parallel control,and air purging the sight tube
on the main scanner, and adding magnifying lenses to it.
These improvements did not eliminate the problem as the long
nose burner used extends deeply into the kiln and the high
level of dust still caused shut down.
In reviewing the problem it was found that normal
-68-
-------
industrial practice for operation of dusty kilns was elim-
ination of flame scanning equipment and substitution of
direct operator observation and control. Work was initiated
at this time to obtain the necessary authorization to oper-
ate the calciner this way. ' .
By late September improvement in draft control was acc-
omplished and product of low MgSO^ content was produced;.-
however, the product MgO was still hard burned. It was-
found that the MgO slaking method which had been proposed
to analyze the product was unsatisfactory, and a decision to
enlarge the scope of the quality control laboratory at Rum-
ford was made to develop the.necessary new analytical methods
for the operation.
During the month of October the calciner was operated
for 179 hours. No coke was added during these runs but the
MgS04 content of the product was. held at a low level by ••
increasing the calciner temperature. When this level fell
below 3-4%,however, a drop in the MgO activity was noted,
the result of everburning of the MgO at high kiln tempera-
ture.
The first shipments of regenerated magnesia produced -in
these operations, was made to Boston in November. Problems
in feeding the MgO were immediately encountered there (as
described in the previous section). A factor in the problem
of using the regenerated, recycled MgO was a large percent-
age of grit and coarse particles contained in it.
A micro pulverizer was located and first installed on
the system at the Mystic Station to control the problems
that resulted from the coarseness of the calciner product.
This location was found to be unsatisfactory primarily be^
cause of space and equipment lay-out limitations.
The annual shutdown of the Mystic 6 boiler taken in
early November, coupled with the infrequent operations in
December, as solutions were sought to improve the activity
of the recycled MgO, resulted in a lack of feed for the re-
generation plant. Processing was not resumed until January
13, 1973, after the initial trials.
During this two month outage the pulverizer was removed
-69-
-------
from the Boston Edison plant and re-installed at the regen-
eration facility.. .It was placed at the discharge of the
calciner product screw conveyor after first removing the
MgO weigh belt. (The weigh belt had been previously found
to be very difficult to keep in operation and its removal
did not affect the project.) In addition a magnetic sep-
arator and a vibratory screen were installed at the feed
hopper of the pulverizer to protect the unit from tramp
materials. (This pulverizer, which was-available on short
notice, was used throughout the project despite the severe.
service it was required to perform. Its use has enabled
gathering information on size distribution in order to
specify grinding equipment in the. future.)
After restarting the regeneration plant in mid January
the modifications and improvements resulted in longer oper-
ating periods but a new phenomenon was observed, the occur-
rence of "slides" in 'the calciner. These slides were evidenced
by a sudden surge of material at the discharge of the
kiln which stalled the material handling equipment and the
pulverizer. This phenomenon was partially controllable if
the feed rate to the kiln was limited to 60-80 Ibs/min.,
(substantially less than 106 Ibs/min design). Attempts to
eliminate the "slides" were limited to increasing the rota-
tional speed of the calciner in order to reduce the bed
depth, and control of the feed rate to reduce the throughput.
Neither of these has been completely successful, and
increasing evidence indicates that the dryness and size dis-
tribution of the dryer product are also important.
Operating at a 60-70 Ib/min feed rate and a 1.56 RPM
kiln speed reduced the slide frequency from 30 .per shift to
1 per shift. However, with each slide heavy dusting still
occurred in the calciner causing a flame-out on the false .
signal previously mentioned.
After discussion with the fire underwriters, permission
was obtained to bypass the flame scanner if an operator was
stationed on the firing platform to tend the unit. The
elimination of the frequent interruptions in operation re-
sulting from this change to manual control resulted in a
great improvement in the quality of the calciner product
that could be obtained. ;
-70-
-------
Implementation of the quality control tests for SO
efficiency of the regenerated MgO and operator control Based
on the product bulk density measurements resulted in impro-
vements in both quality and uniformity of the recycled mag-
nesia.
Other changes were made during the first months in 1973
to eliminate recurring causes of operating problems. These
were 1) the redesign and installation of a new firing end
friction seal in order to minimize air infiltration at this
point 2) rebuilding the gas duct at the calciner exit in
order to provide a better transition to reduce pressure drop
in the system, 3) overhaul of the MgO elevator, and 4) an
air intake outside the building provided for the calciner's
forced draft fan. All this work was finished by the end of
April, 1973.
4.2.4 Planned Operational Testing
Operations were resumed early in May and proceeded
smoothly for a four day period. Coke was added at several
rates during this period with good results. During this run
120 tons of feed were processed. The calciner product MgSO^
content averaged 1-3%; S02 utilization factors between 50
and 65 were measured and bulk density of 25-30 Ib/ft were
observed, all indicative of high quality magnesia. These
operations raised questions on the role carbon played in the
process. There was indication that it smoothed operation
because of the different temperature profile obtained with
the addition of coke which can act as a fuel in the feed
stock.
Before this question could be investigated a strike at
Boston Edison cut off the feed supply,and operations were
again interrupted for a time. The system was not restarted
until early June,and at this time it was decided to repro-
cess the dryer product without the addition of coke. During
the June operations the calciner feed rate averaged 36 tons
per day, and 12 tons per day of product were produced and
recycled to the absorption system.
While numerous material problems still remain to be
solved, it was apparent that the calciner system could
-71-
-------
regenerate the dryer product frcm Boston satisfactorily.
Operation of the regeneration system during July 1973 was
almost continuous; 125 hours the first week, 164 hours the
next week, and a total of 65% utilization during the month
with the longest single outage being 24 hours taken to
shorten the chain in the MgO bucket elevator. The dusty
nature of the feed from Boston did cause numerous brief
problems, literally hundreds of floods of the MgSCu weigh
feeders, recycling in the MgSOo conveyor, etc. In addi-
tion the pulverizer motor was found to be too small to
handle high flows and stalled out when slides occurred in
the calciner. Another new problem that became evident with
longer operations was that the C>2 concentration in the cal-
ciner could fall too low as equilibrium was attained. (This
was an illustration of the efficiency of the improved fric-
tion seals). As a result, on several occasions elemental
sulfur formed in the calciner which in turn plugged the
venturi scrubber.
An acid plant shut down which had been scheduled for
the 28th of July was taken and during this time additional
improvements were made at the regeneration facility. A 20
HP drive was obtained and installed in place'of the 10 HP
drive originally furnished on the pulverizer. A variable
speed drive was installed on the calciner in the hope of
controlling slides through speed variation, and the MgSO_
elevator was examined and plans made to convert it to a
continuous rather than a centrifugal unit.
The acid plant was restarted on August 27th and the
regeneration plant on the 30th. Problems resulting from
the dusty feed noted earlier limited feed rates to app-
roximately 50 Ib/min. It was thought that these problems
were caused by a segregation of feed in the Boston silo
caused by the introduction of the cyclone underflow (fine
dust) and more granular dryer product at diametrically
opposite points in the silo. During the July-August outage
this had been corrected by rearranging the dryer product
discharge back to the center of the silo. A program of
gradually increasing the feed rate was undertaken, and a
75 Ib/min rate achieved by the end of September. Since
changes in feed rate resulted in problems in control of
calciner atmosphere, reducing conditions were encountered
and sulfur formation caused blockage in the venturi scrubber
-72-
-------
on two occasions. The oxygen analyzer probe and sampling
system were overhauled to eliminate leaks, and to give better
control.
A period of low pH operations at Boston resulted in a
feed of high iron content (4-5% Fe203), and processing the
resulting calciner product resulted in rapid wear in the
micropulverizer, but did not stop the processing. Opera-
tions were conducted when feed was available with the only
interruption an acid plant leak which terminated a four day
run (September 14 through 18). A planned shut down was
taken on the 26th to coincide with the centrifuge repairs at
Boston.
Operations at this plant continued to improve as a
result of the better mixing of dryer product from the scrub-
ber operation. A feed rate of 80 Ibs per minute was
achieved in October, With 60% operation of the regeneration
system during the first week and 69% operation during the
third week, with most interruptions caused by a shortage of
feed material from Boston.
During the week of October 26th, the unit was shut down
for additional improvements. A new weigh feeder, live bin
bottom, and tapered flight in the calciner product screw
conveyor were installed and these improvements further as-
sisted 'in smoothing the operations so that during the last
week of the month operation at an 84% utilization level was
maintained.
Operations continued into November,with only one major
difficulty, low air pressure to the feeder controller when
the instrument air pressure dropped below 45 psig. To cor-
rect this a new compressor, Sullair Series 10, was ordered
with delivery of the unit scheduled for December.
A bearing failure on the head shaft of the MgS03 eleva-
tor occurred on November 7th, which shut down the plant.
The plant was not restarted after repairs were completed be-
cause the availability of both MgSO3 from Boston and No.4
fuel oil from Essex Chemical could not be assured. In all,
a total 152 hours of operation were logged during the month
of November.
-73-
-------
The fuel supply question was resolved and calcinar.ior.
of magnesium sulfite from Boston started again in early Dec-
ember. During the first and second week of December, the
system was operated for 196 hours at better than 60% avail-
ability, and 140 tons of regenerated magnesium oxide were
produced. Some mechanical difficulty was encountered when
bolts loosened in the bucket elevator and a flood at the
weigh feeder caused a two day outage during the second week.
Operations were briefly continued during the third week and
an additional 21 tons of magnesia produced prior to shutting
down -the system while the Boston unit was being repaired.
During the first three week period, approximately 430 tons
of magnesium sulfite was processed through the kiln . Sulfur
formation in the calciner was experienced once during that
period and again attributed to strong reducing conditions in
the calciner because of a low C>2 concentration. Plans were.
made to further redesign and improve the oxygen analyzer
sampling system in order to provide a better continuous in-
dication of off-gas composition to control the unit.
Magnesium sulfite was processed through the calcining
system during the last two weeks of the month of January,
1974, 72 tons of MgO were regenerated during 115 hours of
operation. (Plans to start the system earlier had to be
postponed when a defective control valve was found in the
main SC>2 blower recycle loop. This was a reoccurrence of
a previous failure of the valve seat in the valve. A new
seat was obtained and installed.)
During the latter part of the month of January, the new
tail gas scrubber system was tied into the plant during an
acid plant outage. After the acid plant was restarted, cal-
cining resumed on the 4th of February, but on the following
day the system was forced down by a trip of the acid plant
blower. During this brief period of operation, the acid
plant reported that some of the acid produced was cloudy, an
indication of the presence of sulfur in the calciner gas.
The shut down period was extended in order to install the
new oxygen analyzer sample system.
The bucket elevator was changed from a centrifugal dis-
charge type to a continuous type during the following week.
In addition, the tail shaft of the MgSO3 screw conveyor was
extented and equipped with a second bearing in order to
-74-
-------
improve its service life.
Operations were not resumed until the 26th, as there '•
had been no operations at Boston to produce feed in the pre-
vious weeks. When the system was restarted, problems were
encountered in the materials handling equipment, principally
flooding and recycling of materials in the bucket elevator.
This recycling was attributed to a dusty feed being receiv-
ed from Boston.
The interrupted operations resulted in only 100 hours
of operation for the month of February, and the production
of 80 tons of regenerated alkali.
The calciner continued to operate through the beginn-
ing of month of March, but the dusty feed continued to cause
materials handling problems. Finally, a bearing failed on
the MgSO-j conveyor,and the system was shut down until the
7th. When the system was restarted, a slide caused an
overload of the MgO conveyor requiring additional time for
maintenance.
A shortage of feed further limited operations, and the
plant was not restarted until the 13th of March.
Operations continued through the 15th, when the run was
terminated again because of a bearing failure on the MgSOo
conveyor. The low feed inventory was replenished during
this time, and the system restarted on the 18th. At this
time it was noted that draft control in the calciner had be-
come difficult. The cause of low draft and inability to
meet'acid plant air requirements were traced to a build-up
of dust in the 18 inch duct from the exit of the calciner
cyclone to the venturi scrubber. An inspection of this line
showed the duct to be almost completely plugged.
The build-up of dust in the duct was an accumulation
over approximately 1 % years. This duct had been inspected
and cleaned at the earlier time of a two to four inch accu-
mulation. Cleanouts are now scheduled on acid plant turn-
around. Additional consideration was given to blanking off
one of the cyclones on the calciner in order to increase
dust removal efficiency; however, there is a serious pro-
blem with draft control, and it is felt that blanking the
-75-
-------
cyclone could cause the situation to worsen. (This latter"' i:
problem results from a higher delta P (35") than expected
across the acid plant's drying and dry towers, which are
presently packed with 3 inch partition rings.) Regener-
ation operation resumed on the 24th and continued through
the end of the month with a series of interruptions from
both acid plant leaks and new problems arising with the
renovated magnesium sulfite bucket elevator. Despite the
sporadic outages of the regeneration facility, enough MgO
was produced so that the SC>2 abatement plant had sufficient
inventory to continue operations all during the month. The
tail gas scrubber was started up on the 25th, and measure-
ments showed that satisfactory S02 removal could be attain-
ed in the unit to reduce the acid plant emissions to acc-
eptable levels.
Repairs were made to the MgSO^ elevator, and, except
for a few brief interruptions caused by acid plant problems
in the early part of April, the system ran until inventory
was exhausted. This was replenished over two days,and re-
generation was resumed on the 7th of April, continuing
through the 12th, when another failure of the tail shaft of
the bucket elevator occurred.
The shaft was replaced with a heavier one, and no fur-
ther breakage was experienced. The previous five day oper-
ation had again reduced feed inventory, and this was built-
up over the next two days.
4.2.5 Operational Testing
Operations were resumed in mid-April for a seven day
period, only stopping after depleting the feed inventory.
The plant was down from the 21st through the 27th awaiting
a supply of feed. During this period, a leaking emergency
water spray in the venturi scrubber which had caused plug-
gage of the upstream duct was repaired.
Shortly after starting up again on the 28th, it became
apparent that some of the refractory brick in the calciner
had loosened. However, MgS03 regeneration continued through
the 3rd of May whe n the plant was shut down when one of the
bricks jammed the product screw conveyor. The problem was
identified as two rings of brick which had loosened near the
-76-
-------
feed end of the calciner. Spare refractory bricks were on
hand so the repair were accomplished quickly,and the system
was returned to operation from the 7th through the 14th. At
this time repair in the acid plant necessitated a shut down
while the waste heat boiler was overhauled and regeneration
did not start again until the 24th. This outage did result
in a shortage of regenerated MgO for Boston, with operations
limited to 152 hours out of an available 280 hours.
Again in June, the final month of the program, inter-
rupted operations at Boston eventually resulted in a short-
age of feed for calcination. Thus, the regeneration plant
operated five days in the first week of June, and another
five days the second week with the interim periods shut down
for lack of feed. This same shortage had prevented attempts
to achieve design capacity of the calcining facility.
Feed was accumulated during the third week of the month
to initiate some rate and capacity checks, but the results
obtained were inconclusive. Operations were ended with the
program termination at Boston; however, the same facility
will be used in a continued program with a coal fired power
plant.
In the period from March 8th through June 14th, 716
tons of MgO were regenerated in 935 hours of operation.
This can be contrasted with 419 tons of MgO regenerated
during 588 hours in the previous four months period.
4.3 SUMMARY OF SYSTEM MODIFICATIONS
During the course of the project numerous modifications
were made to both plants. These ranged from major changes
such as rerouting the dryer duct to the scrubber instead of
the stack,to minor ones such as increasing the size of the
calciner bucket elevator tail shaft by 1/8". While no major
process changes were made or major new equipment required,
all of the many changes and modifications contributed to
improving the operability of the plants in this first proto-
type installation.
The purpose of modifications fell in three catagories:
first to allow the integrated plants to operate continuous-
ly; second, to increase the reliability and raise the
-77-
-------
capacities nearer design values; and third, to achieve the
desired removal efficiencies and demonstrate the regenera-
tion capability of magnesia. In the first catagory num-
erous problems at both sites still had to be corrected
during the start up period. Sound levels from the acid
plant main blower had to be reduced to an acceptable level,
pneumatic operators on the ID booster fan dampers at Boston
needed to be changed to a size capable of opening the
dampers, and capacity matching of the material handling
equipment was required.
1. Other modifications required for operation and
accomplished in the first phases of the program during 1972
were:
A) Removal of dryer lifter flights in the first
and last third of the drum and removal of the
lifter lips in the mid-section of the drum, re-
quired because the centrifuge cake produced in
operation was a much smaller crystal size than
that formed in the pilot plant operation. Other
dryer modifications included increasing the height
of the rear (feed end) dam.
B) The dryer product that was obtained had large
agglomerates and was also very dusty. In order
to correct this problem, a "finger" crusher was
installed at the dryer exit to reduce large agg-
lomerates to 2" maximum. The dust problem took
some time to bring under control, and the methods
are described in a later section.
C) Corrosion of the fan casings at the 6-1, 6-2
booster fans was corrected by installing a pre-
crete lining in the lower half of the case. The
problem was caused by the condensation of vapors
in the flue gas in the ductwork when the absorp-
tion system was idle. The ducts themselves had
been designed and installed with a liner, but the
fans, identical to the Mystic 6 ID fans, had not.
Installation of the lining corrected this problem.
D) The seals on the calciner were not adequate to
prevent in-leakage of air and, thus, a reducing
atmosphere could not be attained. The firing end
-78-
-------
was modified, all necessary openings closed,
. ^2 of the tube cooler air intakes blanked off,
and the seals reraachined to provide the .min-
imum clearance. This work .did reduce air
'infiltration and MgO .was regenerated for reuse
in the absorption step (further.work was done
on improving the firing end seal later in the
program). .
E) The dusty nature of the material in the cal-
ciner created a problem, as the dust continuous-
ly obscured the flame, tripping the-flame scanners
on a false "flame out!l signal. Numerous minor
modifications were tried to resolve the problem
such as:
1) Changing optics on the scanners.
2) Adding a second scanner.
3) Installation of an air blown sight tube on the
scanner.
None of these provided a permanent solution.
F) Calcined magnesia leaving the kiln contained
a significant amount of grit, which would not
slake when it was recycled to the absorption
system. This grit also caused pluggage of the
MgO slurry lines. A small pulverizer was ob-
tained for test purposes in the prototype. .
Initially installed at the Boston facility it
was demonstrated to be capable of producing a
usable grind.
G) The ground regenerated MgO exhibited a low
reactivity, causing difficulty with pH control
and S02 removal. This was corrected by heating
the MgO slurry tank to 180°F to activate the
magnesia. Heating was accomplished with a direct
stream sparge to the tank controlled by an appro-
priate temperature controller.
2. In the second category several modifications were
accomplished in 1973:
A) Re-routing the dryer off gas to the absorber to
control particulate matter emissions. The dryer
•-79-
-------
had originally been equipped only with cyclones
for dust control of the gas passing to the stack.
It had been predicted earlier that these would
not be suitable, however, as there was no exp-
ience with such a system, the cyclone arrangement
was investigated before arriving at the final
design. After testing of particulate matter had
revealed a 10 gr./SCFD inlet and 1 gr./SCFD.outlet
from the cyclone, it was decided to use the absor-
ber for the final dust control. This was accomp-
lished by installing a direct connection to the
6-1 ID Booster Fan inlet or outlet, with a damper
control to select the injection point for absorb-
er in-service or out-of-service application.
B) The pulverizer was relocated to the regenera-
tion facility and installed at the discharge of
the MgO screw conveyor (0-503). Also installed
at the same time iwere a vibrating screen and a
magnetic separator to remove any tramp material
from the calcined MgO before it entered the
pulverizer.
C) A hammer mill was installed at the truck loading
station at Mystic to reduce oversize lumps scalped
off at the discharge of the belt conveyor (0-408).
MgSOo out-loaded to trucks from this belt contain-
ed a significant percentage of +3/4" size lumps
(10%) that had, to this point, been discarded. A
chute was installed from the scalping screen to the
newly installed hammer mill. The pulverized MgSO3
discharging from the mill was then returned to the
belt by means of a conveyor installed at the same
time.
D) External hammers were added to the feed section
of the dryer and the feed screw extended into the
dryer to eliminate adherence of material to the
shell in this area.
E) A friction seal to replace the original laby-
rinth seal was designed and installed at the firing
hood of the calciner to further reduce air leak-
age at this point. This, coupled with the previous
-80-
-------
fixes, allowed the attainment of a reducing atmosphere
in the calciner. Additional control of kiln operation
was provided by the installation of a variable speed
drive for the range 0.7 to 2 R.P.M.
F) A pneumatic dust conveying system to handle the
cyclone underflow and dryer spill-back was designed and
installed. This finally eliminated a problem that
occurred when adding the dry cyclone dust to the wet
centrifuge cake being conveyed to the dryer. These
two materials when admixed often "set-up" and stalled
the conveyor.
G) The MgO slurry system was further improved by the
addition of a pre-mix tank at the discharge of the MgO
weigh belt (0-103) to act as a steam seal between the
now heated MgO slurry tank (G-102), and the enclosure
erected directly over the tank housing the weigh feeder.
This eliminated the problems associated with condensing
water vapors in this dusty area, but created new ones
because of frequent pluggage of the pre-mix tank.
H) The problem of frequent shutdowns of the calciner
on false flame failure signals, because of the high
dust level in the calciner, was solved by by-passing
the flame scanning equipment. Operator control is
substituted for this instrumentation after the kiln
refractory becomes heated to the ignition temperature.
I) Additional improvements were made in the calciner
feed system. A vibrating bin bottom was installed on
G-506, and the MgS03 weight belt 0-514 changed to a
more robust unit with faster response to varying feed
bulk density.
3. in the third category, these other modifications made
were :
A) Conversion of the MgS03 bucket elevator (0-511) at
the regeneration plant from a centrifugal discharge to
a continuous machine allowing it to handle either dusty
or granular material.
B) Installation of a tail gas scrubber at the acid
plant to reduce emissions to the new standards was
completed and the system operated. The acid plant
-81-
-------
itself is of an old design for 95% conversion, and
was in the process of being phased out before the
program started, because of the inability of such
plants to meet the new standards of the Federal
Clean Air Act. In order to continue the program in
compliance with these standards, a caustic (NaOH)
scrubbing system incorporating a venturi absorber
of Chemico's special design was installed to treat
the acid plant's emissions, reducing the SC>2 from
3000 ppm to 300 ppm.
4.4. SYSTEM AVAILABILITY
System availability for the flue gas desulfurization
system installed at Mystic Station is given in Table 9 which
summarizes operating time and availability by program period
and shows the improvement in availability during the 4,127
hours of operation logged with the system treating flue gas
from a high sulfur fired boiler. Table 10 presents the data
on a monthly basis, with comments on the primary causes for
reduced availability during the planned operational testing
and continuous operation phase which comprised the final year
of the program.
The continuous operation phase was interrupted repeatedly
by boiler outages resulting from numerous tube failures un-
related to the FGD system. Because of the problem, the maxi-
mum duration of continuous gas flow to the absorber was limited
to a week at a time, however, during April, May and June, two
periods of 35 and 30 days were obtained with essentially 100%
availability despite the numerous boiler failures. This
demonstrated the ability of the FGD system to cycle with both
load and varying operation of the boiler. This is shown in
Figure 15 which depicts periods of FGD system operation,
average load during those periods and causes of absorber system
outage.
Similar data are not presented for the regeneration plant
as most outages there were caused by interruptions in the Mgo
supply and these, except for two cases, resulted from break-
down in the transportation and trucking operation. In the
other two instances, first the national shortage of petro-
leum products in November 1973 forced the regeneration plant
out of service for several weeks until this situation was
resolved and a fuel oil supply for the calciner secured; and
second, a two week regeneration plant outage was forced by a
failure in a piece of acid plant equipment not related to the
calcination system. Several problems were encountered and
-82-
-------
TABLE 9
FLUE GAS DESULFURIZATION SYSTEM AVAILABILITY
BY TEST PROGRAM PHASE
Description
Pre start-up through
Break-in period
Planned Operational
Testing
Period
April 1972 to
May 1973
June 1973 to
February 1974
Continuous Operation* March 1974 to
June 1974
Operating
Hours Availability
1,127
17%
1,630
1,370
46%
76%
* During the period 3/1/74 to the end of the
program, the Mystic 6 boiler only operated
at total of 1,721 hours because of numerous
boiler tube failures.
-83-
-------
TABLE 10
FLUE GAS DESULFURIZATION SYSTEM AVAILABILITY
Month
Availability
June 1973
July
August
September
October
68%
61%
—
38%
60%
Comment
November
December
26%
13%
January 1974 28%
February
March
April
May
June
25%
87%
81%
57%
80%
Boiler annual overhaul
System availability decreased
during the last part of the year
due to heavy erosion/corrosion
attack in the recycle slurry
piping syst.em.
System availability limited by
boiler related problems which
caused frequent shut downs in
January and February
Low availability in May due to
two week acid plant outage
Program termination.
-84-
-------
-
CD
o:
o
. -
^
21
"
f j
£-)
•-^^ J-T*-,
cr-."
0.
in
]
r
B.
1
~1
*.-»
4
{
t
i
i
*
i
r~
F'-l
—
r
!
I
t
i
|
1
1
I
-
\
\
E
•
r
...._
I
!
A
; /
; \
•
|
"r
il
BA
!
1
|
!
1
\
— »
-i
XI
3(1
ill
ill
ill
ill
III
in
i
J L
B -
1
;
•~1
"
j
1
—
"
!
(
!
j
i
t
i
«M:
|
1
1
.
j
1
i
1
*-
-T
I
'I
j
|
!
|
I
}
i
i
i
r*
CODE SHUTDOWN CAUSE NO.
A FGD System 7
B Power Plant 25
C MgO Supply 1
A.
Jan.
Feb.
March
April
May
June
FIG. 15
BOSTON ED. OPERRT1SNS JRN.-JJNii 1974
FLUE GAS DESULFURIZATION SYSTEM OPERATING PERIODS AND OUTAGES
-------
solved at this plant as detailed in the foregoing section
(Sec. 4.2) on the operation, and in addition the calciner was
never run at its design capacity; however, the supply of
regenerated magnesia was maintained at satisfactory levels
when the plants were operating as an integrated system,.
-86-
-------
5.0 EQUIPMENT ARRANGEMENT AND MATERIALS
OF CONSTRUCTION
Figure 16 shows the general arrangement plan for the
flue gas desulfurization system installed at Boston Edison
Company's Mystic Station and identifies the location of the
major pieces of equipment. In the S02 absorption system,
a wide variety of conditions exist in the equipment as the
process streams range from incoming flue gas to saturated
gas streams, liquid and slurry streams, and varieties of
solid chemical compounds. A range of acidity also exists
throughout the plant. Figure 17 shows the general arrange-
ment of the regeneration plant at Rumford, R.I. and also
shows major equipment location. Tables 11 and 12 list the
equipment by category, and material of construction for
both plants, comments are given if problems were encountered
with a particular item.
-87-
-------
i
00
00
ROTARY DRYER
•It
D
Mos
STORAGE
SILO
r"
-C.-J.,
I
,-J-L-.
_J *- T -,—I
c;; j cbs
MOTHER
FIG. 16
GENERAL ARRANGEMENT-PLAN
SCRUBBING SYSTEM FOR 155 MW CAPACITY AT MYSTIC NO 6 STATION OF
BOSTON EDISON COMPANY
UJ C
_l
u
/ p", BELT CONVEYOR
n
ll
0
I O
L»A
t
TRUCKS
\ \
O
I
-------
VENTUPI
hCRUBBE.R-
COOLEH FUEL OIL TANK
I
00
VD
I
ACID
PLANT
GENERAL ARRANGEMENT-PLAN
CALCINER PLANT AT ESSEX CHEMICAL co. RUMFORD RHODE ISLAND
-------
TABLE 11
1-101
J-101
EQUIPMENT PERFORMANCE
FLUE GAS DESULFURIZATION SYSTEM
Code
No. Equipment
D-401 Combustion Chamber
G-102 MgO Makeup Tank
G-301 Mother Liquor Tank
MgO Makeup Pumps
J-201 Recycle Pumps
J-301 Mother Liquor Pump
K-201 ID Booster Fans
K-401 Dryer ID Fan
K-402 Dryer FD Fan
Material Of
Construction
Refractory Lined
Carbon Steel
Carbon Steel
Carbon Steel
MgO Storage Bin & Dust Collector Carbon Steel
Cast Iron
Cast Iron
Cast Iron
Carbon Steel Wheel
Mild Steel Case
Carbon Steel
Carbon Steel
Comment
Two outages resulted from a failure of the
burner block and resultant failure of the
refractory lining adjacent to it.
Steam sparging of the tank and a pre-mix
unit were added; an 18 in. standpipe was
also added at the outlet nozzle after plugging
of the outlet of the vessel by MgO lumps.
Satisfactory Operation
Satisfactory Operation
Severe erosion of impellers and casing was
experienced.
Severe erosion of pumps was experienced.
Some impeller erosion was experienced.
Some corrosion of case was corrected by
precrete lining the lower half of the case.
Satisfactory Operation
Satisfactory Operation
-------
TABLE 11 CONT'D
Code
No.
M-101
M-301
O-101
Equipment
MgO Tank Agitator
Mother Liquor Tank Agitator
MgO Conveyor
O-102 MgO Vibrating Hopper
O-103 MgO Weighing System
O-401 Dryer
O-402 Dryer Conveyor
O-403 MgSO3 Conveyor #1
O-404 MgSO3 Elevator
Material Of
Construction
304 SS wetted parts
304 SS wetted parts
Aluminum Straight
Tube SS Bend
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Comment
Satisfactory Operation
Satisfactory Operation
A hole developed in the bend after one year
of service.
Satisfactory Operation
Frequent shutdowns caused by jamming of the
weigh control mechanism and belt.
Numerous problems primarily of a process
nature caused by adherance of centrifuge
cake to dryer walls.
Frequent shutdowns caused by hardening of
centrifuge cake with subsequent jamming
of the unit and wear of the unit.
Occasional large lumps from dryer jammed
conveyor.
Satisfactory service except for one outage
because of a bearing failure on the drive
spocket.
-------
TABLE .IICONT'D
Code
No.
O-405
O-406
Equipment
MgSOg Conveyor #2
MgSO3 Weigh Feeder
O-407 MgSO3 Screw Feeder
O-408 MgSO3 Belt Conveyor
O-409 Dust Feeder
R-402 Dust Collector
Material Of
Construction
Carbon Steel
Carbon Steel
Rubberized Belt,
Conventional drive
mechanism
Carbon Steel
Carbon Steel
Comment
Eliminated from material handling train.
Eliminated from material handling train
after numerous failures.
Satisfactory Service
Trouble free operation, however, dusty nature
of feed created secondary clean up problem.
Entire dryer dust handling system was re-
vised as part of dryer modifications.
Trouble free operation except for some in-
stances of packing of collected dust in hopper
after failure of O-409.
R-301 Centrifuge
Carbon Steel
See Sec. 5. 2.
-------
TABLE 12
Code"' •-- ;;
No.
E=501A,B
F-502
G-501
£ G-505
G-507
G-508
1-502
J-050A, B
J-501A.B
J-502A, B
Equipment
Weak Acid Cooler
Weak Acid Stripping Tower
Weak-'Acid Pump Tank
Coke Feed Bin
M;gS'o3:;Feed Bin
•, ••: • -i f , •• !/>_
Bearing Water Pump Tank
' '*•'-.:. - i
Air Receiver
; :.,--)•,'.••. ."!
MgO Silo
R e cy'cle Water Pumps
Calcirier Fuel Oil Pumps
Weak Acid Circulating Pumps
EQUIPMENT PERFORMANCE
REGENERATION SYSTEM
Material Of
Construction
316 SS
Chemical Stoneware
Reinforced Fiberglass Poly-
ester.
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Alloy 20
Cast Iron
Alloy 20
Comments
Satisfactory Service
Satisfactory Service
Satisfactory Service
Satisfactory Service
Plugging problem corrected by in-
stallation of a vibrating bin bottom.
Satisfactory Service. •
Satisfactory Service
Satisfactory Service
Satisfactory Service
Satisfactory Service
Satisfactory Service
-------
TABLE 12CONT'D
Code
No.
J-503
J-504
K-501
O-503
O-504
O-505
O-506
O-508
Equipment
Neutralizing Pump
Bearing Water Pump
Calciner F.D. Fan
MgO Conveyor #1
MgO Elevator
MgO Conveyor #2
MgO Vibrating Hopper
MgO Weigh Feeder
O-509,O-510MgSO3 Conveyor #1
0-511 MgS03 Elevator
Material Of
Construction
Natural Rubber Lined
Cast Iron
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Carbon Steel
Comments
Rubber swelled and bound pump
after two years operation.
Satisfactory Service
Satisfactory Service
Frequent hanger bearing failure.
Wear of sprockets and chain
experienced after 18 months unit
converted to continuous discharge
type.
Satisfactory Service^
Satisfactory Service
Eliminated from material handling
train after numerous failures.
Problems limited to bearing failures.
Frequent break downs required shaft
size increase.
-------
TABLE 12CONT'D
Code
No.
O-512
O-513
O-514
O-515
O-516
O-523
P-075
R-401
R-501
R-020
Equipment
Calciner Conveyor
Calciner Elevator
MgSO3 Weigh Feeder
Coke Weigh Feeder
Cyclone Dust Feeder
MgSOg Airveyor
Cyclone Dust Collector
Fuel Oil Supply & Combustion
Equipment.
Calciner
WA Venturi Scrubber
Materials Of
Construction
Carbon Steel
Carbon Steel
Carbon Steel
316 SS
Carbon Steel, Refractory
lined
316 SS
Comments
Satisfactory Service
Unit converted to continuous discharge
type, bucket fastening bolts failed due
to corrosion.
Numerous failures, unit replaced by
a BIF unit reducing the number of
break downs.
Belt tracking problems.
Satisfactory Service
Frequent pluggage due to tramp
material and lumps in transported
MgS03.
Satisfactory Service
Problem of unsuitable flame failure
control.
A two foot wide section of the re-
fractory lining loosened and failed,
pinion gear failed after two years.
Occasional pluggage due to formation
of elemental sulfur.
-------
5.1 ABSORBER SYSTEM AREAS SUBJECT TO CORROSIVE ATTACK
Carbon steel is the principal material of construction of
the unit but because of S02 absorption, the carbon steel must
be protected, in some cases stainless can be substituted in
the internal structure, but here the corrosion-problem can be
influenced by the amount of chlorides in the recycle stream
(Table 13). Since temperatures in the FGD System are normally
low, (the adiabatic saturation temperature of the gas) as long
as scrubbing liquor is present, the carbon steel structural and
other parts of the venturi absorber were protected by about 60
mils of glass reinforced polyester resins. This material was
chosen over stainless as the concentration of chlorides in the
absorber slurry could increase up to a point where Austenitic
stainless steels are susceptible to stress cracking and sulfuric
attack in the presence of high cloride causes alloys like type
316 to pit excessively.
Some wear was experienced in the absorber during the
twenty-seven months' operation of the plant. The extent of
this wear was minor, limited to the erosion of the top coat
of the two layer application of the FRP liner, and the erosion
of the steel apex of the toroidal ring.
The pattern of the erosion of the torus indicated it
occurred in the areas of over-lap of the adjacent irrigation
sprays. Repair was easily accomplished by welding an angle
section over the damaged areas of the torus. This can be seen
in Figure 18 which shows (from left to right) the inner core,
first slot, torus, second slot, part of the converging wall,
and the tips of two spray nozzles.
The shaded areas on the converging section of inner wall
are the underlayer of the FRP liner. The patched area of the
torus is visible.
The polypropylene chevron modules, used for mist elimina-
tors, have shown no corrosion problems, nor have the plastic
beams used to hold them down. The upper surface of these
modules is shown in Figure 19.
Recycle pumps and piping have been subject to extensive
wear of an erosive-corrosive type. This was particularly
serious in areas of high velocity and turbulence. The recycle
pumps were originally all carbon steel. Figure 20 shows the
-96-
-------
TABLE 13
CHLORIDE ANALYSIS
Scrubber Slurry (11/16/73)
Magnesium, Oxide (New)
Magnesium Oxide (Regenerated)
CHLORIDE %
0. 21%
0.16 - 0.18
0.0
Process Water (Avg. 1972)
Process Water (12/6/73)
Process Water (12/7/73)
CHLORIDE (PPM)
10. 4 PPM
7.8 PPM
3. 4 PPM
-97-
-------
Fig. 18
VENTURI THROAT AREA WEAR
Fig. 19
DE1MISTER MODULES
-98-
-------
Kig. 20
RECYCLE PU.MP IIVIPELI ,ER WEAR
-99-
-------
condition of the carbon steel impellers after one year of
service. The impellers, suction plates, and other parts
of the pumps were replaced in stainless steel, where avail-
able; however, these were also subject to wear when used
in the slurry service.
Similar wastage of metal due to abrasion was experienced
in the MgO slurry pumps. Leaks also developed in sections
of the distribution piping of the recycle lines shown in
Figure 21. A sample was cut from a larger diameter header
pipe which had welded to it, in a horizontal plane a smaller
4-1/2 inch diameter pipe and sent to L. Pitkin, Inc.,
analysts and metallurgists, for evaluation. This pipe had
been weld repaired where it had holed through on a prior
occasion. In the sample submitted there were two holes,
approximately opposite each other near the weld joint, at
an angle of about 45° to the cylindrical axis of the larger
pipe.
The following is a report of the examination.
(Figure 22 is a photograph of the T-section sample as
received, viewed so as to show one of the areas where the
smaller pipe had holed through).
"The outer surface of the."sample showed flaked paint
and rust staining, and except for the holes, appeared to be
in sound condition. The inner surface of the sample was
examined and it was observed that there was extensive, severe
deterioration at the weld .joint. The deterioration diminshed
proceeding away from the joint towards the flame-cut end of
the small pipe. Deterioration was concentrated for about
180 degrees on one side. The other side was, by comparison,
relatively sound.
"Further examination showed the I.D. of the small pipe
to have thinned rather gradually and uniformly on one side
as revealed by a cross-section cut about five inches from
the weld joint. There were no accompanying irregularities
as were present near the weld joint. The cut cross-section
is shown in Figure 23.
"Microspecimens were prepared through the deteriorated
area of side A of the small pipe. The material at the I0D.
merely showed smooth undulations with no corrosion penetra-
tion. The microstructure was comprised of pearlite and
-100-
-------
••'
Fig. 21
RECYCLE PIPE-UPPER SECTION
.-101 -
-------
o
•• '
r" SECTION OF RECYCLE LINE
FIG. 22.
-------
Fig. 23
' CROSS SECTION OF RECYCLE
-------
ferrite in a slightly acicular or needle-like pattern charac-
teristic of normalized medium carbon steel (estimated 0.25%
carbon). Figure 24 is a photomicrograph at 100X showing the
general microstructure and I.D. profile in the deteriorated
area where the wall had been considerably reduced.
"The rust and corrosion product on the I.D0 surface of
the pipe was leached with distilled water and checked for
acidity by acid test paper. The results indicated a pH of
from 6 to 7 - virtually neutral. The leachings were further
checked for sulfites and sulfates which, if present, would
be acidic in nature and, of course, could progressively
.attack the carbon steel. The results were negative - neither
sulfates nor sulfites being found.
"it would appear that any sulfate or sulfite compounds
that were present on the inner surface had been washed away
in subsequent operation or perhaps converted to water soluble
compounds and so not detected. In any event, the inner ;
surface of the T-sample was in a potentially uncorrosive
condition with respect to the slurry handled at the time
it was received.
"The deterioration sustained was only in the small dia-
meter pipe. This was a thinning on one side of the pipe
diminishing towards its. discharge end. The large diameter .
pipe was virtually unaffected. This condition, in the
metallurgist's opinion, was the result of inlet end impinge-
ment erosion. Liquid flow over the sharp 90-degree change ;
in direction into the smaller diameter discharge pipe line
causes eddies and whirls due to the abrupt change in flow
direction and increased velocity. The result is that there
is a mechanical erosion effect accompanied by differential
aeration of the solution with attendant formation of galvanic
oxygen concentration cells which leads to an accelerated
deterioration as observed."
The recycle line wall thickness was gauged using ultra-
sonic testing revealing that there had been general corrosion
throughout the system. A possible cause of the corrosion is
the presence of the bisulfite ion (HS03) which persists up to
a pH of 8.3 (Figure 25) i.e., above the normal system oper-
ating pH. Another contributing factor for corrosion were
short periods of low pH excursion which occurred during the
course of the program because of control upset.
-104-
-------
March 4, 1974
M-3070
fi^M^^fM^^^
V-
FIG. 24
MICROSTRUCTURE AT I.D. SURFACE -
DETERIORATED AREA
100 X
Photomicrograph of longitudinal section through deteriorated
area, side A, of small diameter pipe showing general microstructure of
ferrite and pearlite. The I.D. surface profile (top) shows a smooth
undulating profile with no corrosion penetrations and absence of any
oxide scale build-up.
steel.
The microstructure is normal for a plain medium carbon
-105-
-------
CURVE.
en
i
11
MgO
25 30 35 40 45 So
2 ^
-------
5.2 CENTRIFUGE SYSTEM
Sections of the centrifuge are also subject to wear
and in this case it appeared that abrasion was the cause of
wastage of metal. The parts of the machine which were most
affected were the replaceable items such as bowl plows, and
conveyor flights. The unit, of carbon steel construction,
had the conveyor flights originally edged with Colmonqy #6.
After the first 14 months of intermittent service, the
machine was disassembled and inspected with the following
observations:
Conveyor Blades . •
Conveyor blades were found to be worn an average of
3/32" in the cylindrical section to a maximum of an average
1" in the conical section. Conveyor blades were worn thin -
in the areas adjacent to the feed ports. Badly worn blade
segments were replaced and the blade areas adjacent to the
feed ports were covered with J alloy. Conveyor blades were
built up and re-hardsurfaced with Colmonoy #6, which was
the original hardsurfacing alloy.
Product Build-Up
It was determined upon inspection that the out-of-balance
problems reported were probably caused by build-up of solids
in the drain compartment of the conveyor hub. Virtually all
surfaces of the bowl and conveyor had a significant build-up
of hardened solids resembling concrete, in order to minimize
solids this build-up in the conveyor hub drain compartment,
a tube was installed to blank-off this zone^ All accessible
surfaces of the bowl and conveyor were then covered with
Tropolite plastic coating to provide a smooth corrosion
resistant surface.
Bowl Head, Bowl Head Plows & Wiper Beads
Considerable wear was evident in the solids discharge
area. The hardfaced product wiper beads on the bowl head
periphery were worn off and the bowl head flange had worn
approximately 1/2" radially. The bowl head solids plows
evidenced considerable wear. The bowl head was repaired and
the wiper beads were replaced and an additional eight (8)
beads were applied. Two of the solids plows were replaced
-107-
-------
by the larger diameter case plows and the remainder of the
bowl head plows were re-hardsurfaced arid reinstalled.
The above summary while not complete does highlight
the important steps taken to improve the reliability of the
equipment.
As noted, the conveyor was re-hardsurfaced with the
original (Colmonoy) metal, as were the bowl head plows.
Two of the solid plows were replaced with larger diameter
case plows and the machine returned to service. The hard-
surfacing was touched up in October 1973, when the machine
was disassembled for repair to its feed pipe and the installa-
tion of additional wash nozzles.
The machine was again opened and inspected in January
1974 and wear was still evident in the conveyor (Figure 26).
It was hardsurfaced with Stellite II and remained in service
until the end of the program.
Total time daring which the centrifuge was fed .in the
periods is summarized'in the following table:
TABLE 14
Start Finish Hours Fed Equivalent Days
April 1972 June 1973 1127 48.9
August 29, 1973 Sept. 25, 1973 215.5 9
Sept. 30, 1973 Jan. 28, 1974 818.6 34.1
Feb. 19, 1974 June 26, 1974 1448.1 60.3
A plot of the conveyor wear is given in Figure 28, which
graphically shows the similarity of wear patterns with the
feed to the machine entering between blade 14 and 15. It is
also evident that the Stellite II applied in the latter part
of the program gave improved service (blade wear would have
to exceed two inches before a significant reduction in
efficiency would be experienced). Based on the rate of
wear for the period February through June 1974, it appears
that over one year of continuous operation would be obtained
-108-
-------
I
Centrifuge Conveyor
(Jan., 1974)
Showing Flight Wear
FIGURE 26
Centrifuge Conveyor
(Jan., 1974)
Showing Scale Build-up
on Flights
FIGURE 27
-109-
-------
BIRD CENTRIFUGE
CONVEYOR BLADE WEAR
cL
<
LU
LU
D
<
_J
cD
12.
O
10 15"
BLADE MS.
IS 30 45 fcO TF
IMCHE5 FROM 50LID3 D15CHARG-E
Fig. 28
SCALE -
-110-
-------
before the conveyor would have to be repaired.
5.3 STACK
The existing Mystic #6 stack handled dryer off gas
for an eight month period in addition to the flue gas.
The stack was brick lined, with concrete outer shell, 335'
high with 15' inside diameter at the base of the inner stack.
It is designed to withstand saturated, acidic gases with the
inner stack constructed of solid shale brick laid up in a
potassium silicate mortar (Corlock). The recommended pH
range of the mortar is 1 to 7 where there is no submergence,
and 1 to 5.5 where there is submergence or heavy liquor flow.
The stack was inspected on several occasions. The first
inspection, prior to redirecting the dryer-off gas to the
absorber, revealed several inches of build-up of solid mag-^
nesia salts in the stack. These were washed off in a clean-
down of the stack and no significant deterioration of the
brick work or mortar was observed after the washing. After
the dryer off-gas was treated in the absorber with the flue gas
there was no reoccurrence of the problem.
On a second occasion, the inner stack was observed
to be leaning and touching the outer shell at its top where
a ten foot vertical section of the inner brick stack had
cracked. (During this period construction of Unit 7 was
underway, including pile driving in the area immediately
adjacent to the absorption plant and stack).
After the dryer-off gas was redirected to the absorber,
consideration was given to reheating the treated flue gas.
However, inspection had shown no evidence of deterioration
of the stack due to the condensing liquor (Table 15), and
there was no observation of "raining" from the stack during
operations with the S02 Absorption System. No reheat was
provided for the flue gas except for that resulting from
the mixing of the small portion of untreated gas during the
last four months of operation.
5.4 REGENERATION SYSTEM
With few exceptions, the materials of construction
selected for the calciner installation proved to be entirely
satisfactory., Problems listed below resulted from operating
-111-
-------
TABLE 15
STACK DRAININGS ANALYSIS
6/13/73 7/17/73 7/18/73
Load 145 MW 145 MW 145 MW
pH 4.1 4.3 3.3
SO4(As MgSO4) 13.3% 7.0% 5.3%
Ni 0.01% . 0.004% 0.003%
R2O3* 0.12% 0.18% 0.22%
*R2°3
-112-
-------
conditions which were not anticipated in the original de-
sign.
All material handling equipment for the dry solids,
both feed and product, are of carbon steel. No corrosion
problems were experienced, with the exception of the coke
bin. The coke originally purchased contained from 10% to
17% free moisture. Since coke was seldom used, some scaling
took place on the walls of the coke bin. Even so, the
problem was not one of deterioration of the bin, but rather
that scale would occasionally break off in large pieces and
cause malfunctions of the coke feeder.
No corrosion problems were observed in the calciner
elevator despite the fact that it would frequently become
wet with condensate from combustion gases. This occurred
when the calciner was not being fed but. the burner was
kept on to maintain.temperature.
The gas duct between the calciner and the venturi
scrubber was constructed of Type 310 stainless steel in
anticipation of high exit gas temperatures. The venturi
scrubber itself was constructed of Type 316 stainless steel.
Both these areas were subject to corrosion under certain
conditions. In places where moist solids would accumulate
there was considerable corrosion. This was evidenced by
numerous perforations in the convolutions of the Type 310
expansion joints, in the bottom portion of horizontal or
inclined runs of the Type 310 ducts, and in the Type 316
sampling system for the analyzer monitoring the oxygen
content of the calciner exit gas. However, if the metal
surface was clean and dry, or if it was constantly washed
with a liquid stream as in the scrubber, no corrosion was
experienced.
The gas duct from the venturi scrubber to the acid
plant drying towers was constructed of fiberglass rein-
forced polyester and was satisfactory.
The first 40 feet of the calciner was lined with 9 inch
of high alumina (70%) fire brick, and the remaining 80 feet
with 6 inches of high duty kiln block, backed with 2% inches
of insulating brick. The high alumina brick showed no evi-
dence of degradation. The kiln block suffered only a few
-113-
-------
spalled corners, except in the area located from 25 feet
to 40 feet from the feed end. In this area, considerable
spalling of the face of the block took place. It is pos-
sible that in this area, most of the combined moisture is
released from the feed. In this case, the face of the
lining is subjected to rapid temperature fluctuations as
feed is introduced or stopped, or as the moisture content
of the feed changes.
The insulating brick behind the block was relatively
soft. On two occasions, one or more rings of block fell
out and considerable wear was found on the insulating brick,
This wear allowed the block to move outward toward
the shell.
5.5 CORROSION SAMPLES
Coupons of various metals and coatings were placed in
the system at the initiation of testing in a program of
evaluation of materials of construction. Results for two
of these test racks are given below:
1) A test rack of various metals was immersed in
the venturi absorber basin and removed after
approximately eight months of exposure. Table
16 shows the corrosion rates measured.
2) A test rack of coupons with various coatings
was immersed in the mother liquor tank, an
environment similar to Item 1 above. The
results of the test, after the coated speci-
mens had been immersed for an eight month
period is given in Table 17.
-114-
-------
TABLE 16
CORROSION TEST DATA
Corrosive Media
Location Of Specimens
Type of Test
Temperature
Test Duration
Magnesium oxide slurry scrubbing liquor for SO2
removal from power plant flue gas.
Immersed in venturi scrubber liquid.
Field Aeration. Moderate Agitation. Moderate
F Avg 120 Min 100 Max 140
C Avg 49 Min 38 Max 60 File. 603
From 12/5/72 To 7/31/73 Total Days 238
Material
SS EB 26 1
Incoloy Alloy 825
Inconel Alloy 625
Armco 22-13-5
Allegheny Ludlum 6X
Hastelloy Alloy G
Hastelloy Alloy G
Titanium
SS Type 216
SS Type 317
Carpenter 20 CB-3
Durimet 20
Corrosion
Rate
(MPY)
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.0
Maximum
Pit Depth
(MILS)
5
INC IP
INC IP
INC IP
INCIP
-115-
-------
TABLE 16 (CONT'D)
CORROSION TEST DATA
Material
SS Type 316L
Hastelloy Alloy C-276
SS Cast Grade CF8M
Moriel Alloy 400
Steel HSLA USS COR TEN
Steel AISI Type 1010
Corrosin
Rate
(MPY)
0.0
0.0
0.0
5.0
20.1
20.2
Maximum
Pit Depth
(MILS)
INC IP
13
PERF 40
-116-
-------
TABLE 17 - RESULTS OF COATED SPECIMENS FROM CORROSION SPOOL
IN MOTHER LIQUOR BETWEEN CENTRIFUGE & PUMP
TYPE COATING GMS
CEILCOTE 103 PRIME -.20
+252 T. C.
CEILCOTE 103 0
CEILCOTE 252 -.16
VAL CHEM TAR COAT -.39
HI BUILD VINYL -.02
HI BUILD EPOXY -.19
CARBOGLAS (FIBER
GLASS & POLYESTER) -.46
POLYURETHANE -.35
CARBON STEEL -16. 28
LOCKPRIME - TUFCHEM -.38
MEMBRANE
COMMENTS
EVIDENCE OF ABRASION AND SOME SLIGHT ATTACK, NOT AS
GOOD AS FLAKELINE 103 ALONE.
NO EVIDENCE OF ATTACK
EVIDENCE OF ABRASION AND SOME SLIGHT ATTACK.
COATING PENETRATED ON SIDES OF SPECIMENT BUT NOT ON
END WHERE THEY APPLIED HEAVIER COAT.
VINYL COATING UNATTACKED. EVIDENCE OF CREVICE
CORROSION UNDER TEFLON SPACERS.
SOME STAINING AND OBVIOUS CORROSION ATTACK.
SOME GLASS OBVIOUSLY MISSING. SPECIMEN SMOOTH INSTEAD
OF ROUGH LIKE CEILCOTE COATINGS.
COATING SHOWS HEAVY ATTACK.
BADLY SCALED AND HEAVILY ATTACKED.
SOFT LIKE RUBBER, COATING CRACKED AND WORN AWAY, ALSO
CREVICE CORROSION.
-------
TABLE 17CONTINUED - RESULTS OF COATED SPECIMENS FROM CORROSION SPOOL
IN MOTHER LIQUOR BETWEEN CENTRIFUGE & PUMP
oo
TYPE COATING
PLASITE 4000
(VINYL ESTER)
URETHANE
ELASTOMER
GMS
+ .29
-.09
CYANAPRENE ST-4808 +.04
(ELASTOMERIC URETHANE)
CARBON STEEL
GLASS REINFORCED
VINYL ESTER
-18.33
-.01
COMMENTS
PICKED UP CONSIDERABLE FLY ASH WHICH WOULD NOT
DESTROY COATING. COATING INTACT.
SOME STAINING AND CHIPPING OF COATING AT EDGES,
REST IN GOOD SHAPE.
EDGES OF SPECIMEN WAS NOT COATED. COATING NOT
ATTACHED BUT DOES NOT SHOW STAIN.
HEAVY SCALE AND WASTING AWAY OF SPECIMEN. ALMOST
20% MISSING FROM CORROSION.
ATTACK ON ONE SIDE ONLY. OTHER SIDE STILL SHOWS
MACHINING MARKS.
-------
6.0 PROCESS CHEMISTRY
Pulverized solid magnesium oxide is introduced into a
mixing tank where water is added and conversion to the hy-
droxide commences. The mixture is added as a slurry to the
recycle stream of the venturi absorber to contact flue gas
containing sulfur oxides and products of combustion; The
reactions between magnesium hydroxide and sulfur dioxide
produce principally magnesium sulfite.
The following series of reactions characterizes the
principal ones of the process:
MgO Slaking:
MgO(s)+H20(l) *-Mg(OH)2(s) (6-1)
Mg(OH)2(s) - 3~ Mg(OH)2 (Soln) (6-2)
Mg(OH)2 (Soln) •*- Mg++ + 2 OH"" (6-3)
S02 Absorbtion:
S02(g) - - >- S02 (Soln) (6-4a)
S02 (Soln)+H20 ->• H2S03 (6-4b)
S02 (g} +H20 •>• H2S03 (6-4)
Formation of MgSOs
H2S03 - >-H+ + HS03 (6-5)
H S0~ »- H+ + SOf (6-6)
Mg(HS03)2+2H20 •—'••> Mg++ + 20H~+2H+ + 2HS03 (6~7)
MgS03xH20+2H20 -^ Mg++ + 20H~ + 2H+ + SOf + XH20 (6-8)
X = 3 or 6
Other Reactions:
C02(g)+H20 >- H2C03 (6-9)
H2C03 >- H C03~ + H+ (6-10)
H C03~ ->- COf + H+ (6-11)
Mg C03.5H20 ^Mg++ + C0~3 + 5H20 (6-12)
H20 -- >- H+ + OH~ (6-13)
-119-
-------
Kerr (if has presented the magnesia system equilibrium
composition for a system at 125°F as follows:
Component Molality
HS03~ 5.045.x 10~5
H+ 2.416 x 10~4
S03= 2.182 x 10"3
Mg++ 2.452 x 10~2
HCO ~ 3.876 x 10~2
C03= 2.920 x 10-3
H S03 8.370 x 10~12
H2C03 1.108 x 10~4
OH~ 3.808 x 10~5
-11
Sulfur dioxide partial pressure 1.651 x 10 atm.
Carbon dioxide partial pressure 6.493 x 10 atm.
Thus, from reactions (6-7) and (6-8) the SO,, removal
process is essentially an acid-base reaction, and the re-
action in the liquid phase will be very rapid and almost
quantitative.
While other side reactions take place the most signi-
ficant are those which produce magnesium sulfate:
MgO+S03+7H20 »- Mg S04 •7H20 (10-14)
Mg S03+1/202+7H20 *- Mg S04 •7H20 (10-15)
6.1 EFFECTS OF THE PROPERTIES OF MgO ON THE PROCESS
From equations (6-1) - (6-8), it is seen that the pro-
cess depends on the formation of Mg and OH ions and sub-
sequent neutralization of sulfurous acid. The formation of
Mg++ and OH~ ions, however, is dependent upon the hydration
*See Section 11, References.
-120-
-------
characteristics of solid MgO. A recent paper (2) on the
hydration of MgO has shown that the rate controlling step
is the removal of solid Mg(OH)2 particles from the surface
of the solid MgO. In agreement (3) with the above, both
MgO and Mg(OH)2 of similar surface area have been shown to
react at the same rate with aqueous C02•
The above conclusion is only valid, however, when the
MgO is in a reasonably active form. The other factor which
affects the process is the particle size of MgO. Not only
do the finer particles react faster, but also being smaller,
they remain suspended for longer periods of time and thus
have a better chance to react with S02»
6.1.1 Pulverization
Early attempts to use calcined magnesia in the absorp-
tion system without any intermediate processing were un-
successful. The product MgO, as it exited the kiln coolers,
was gritty and contained a small percentage of agglomerates.
When this material was added as a slurry, it settled rapid-
ly with the result that the MgO slurry pumps and lines
plugged. Finally the lack of a fine dispersion of magnesia
caused further problems with pH control.
A small, high speed hammermill (Micro Pulverizer 2DH)
was installed first at Boston and, after a brief trial
period, was moved to the regeneration facility where it was
installed to pulverize the calciner product.
Several varieties of screens and hammers were tried
in this unit and two of the screen types were moderately
successful in producing a satisfactory pulverized product.
There were herringbone pattern screens and jump gap screens.
Neither type had a satisfactory service life as would be
expected with the abrasive regenerated magnesia.
Herringbone pattern screens had a service life of 4 to
8 hours. The jump gap screens gave better service as did
hard faced hammers and lined head plates. Average replace-
ment times are as follows:
-121-
-------
Jump Gap Screens 185 hours
Hammers 1500 hours
Head Plates 230 hours
Table 18 shows the size distribution before and after
pulverization of the regenerated magnesia as well as the
distribution of the unpulverized calciner product.
6.1.2 Hydration of MgO
Digestion of MgO in water resulted in slaking of the
MgO and formation of Mg(OH)2 which subsequently reacts with
S02 • The conditions necessary for efficient slaking of re-
generated MgO were investigated in the laboratory using the
S02 utilization test (Appendix C ) developed to monitor
calciner operation. The effect of digestion temperature
using this same test is given in Table 19, while the rate
of hydration for various samples treated at 180°F in water
given in Table 20.
6.2 SYSTEM pH AND S02 ABSORPTION
From equations (6-5) and (6-6) it is seen that the for-
mation of solid MgS03 is dependent on the pH of the ab-
sorbing system. It is also evident from the pK values of
Oo,
~2 i
Table 21, that sulfurous acid (considering the
S03~2 ion) is a weaker acid than acetic acid.
Coupled with this is the fact that alkaline-earth
sulfites dissolve readily (4) in moderately strong acids
(even in sulfurous acid) . Thus pH values near 7 were select-
ed for operation.
Because of the weakly acidic character of sulfurous
acid, the salts, especially alkali and alkaline-earths,
are hydrolyzed in solution and consequently the sulfites
are alkaline in nature.
Titrations of S02 in water by MgO and Mg (OH) 2 suspen-
sions support the above conclusion, i.e., for more than 90%
removal the pH of the final suspension should be above 7.
-122-
-------
TABLE 18
DRY SCREEN ANALYSIS
Screen
Size (Tyler)
+50
+100
Unpulverized
Calciner Product
A
30.3
18.1
Pulverized Calciner Product
Herring Bone
Screen
B
39.1
Jump Gap
Screen
C
5.3
9.5
+200
+325
-325
9.4
13.1
29.1
9.4
3.3
48.2
8.8
49.8
26.6
-123-
-------
TABLE 19
EFFECT OF TEMPERATURE OF DIGESTION OF MgO
ON SO2 REMOVAL CAPACITY
Sample
MgO, 97.4%
B/D 46.40#/ft.
MgO, 97. 35%
MgSO4, 0.50
B/D 68#/ft. 3
Treatment
a. Untreated.
b. Digested at 160°F
for 45 mins.
c. Digested at 180°F
for 45 mins.
a. Untreated
b. Digested at 180°F
for 1/2 hr.
c. Digested at 180°F
for 1 hr.
d. Digested at 210°F
for 1/2 hr.
S02
Efficiency
20
50
72
22
42
49
45
-124-
-------
Time in
Minutes
Ul
I
15
30
45
60
90
120
TABLE 20
RATE OF HYDRATION OF MgO at 180°F. in WATER: 10% SLURRY
MgO (AR)
MgO 94.47
SOs, eff. 60.00
% MgO after
Hydration
79.50
74.82
72.32
70.96
69.36
68.65
% of
Hydration
48.20
64.27
71.32
75.70
80.85
83.14
R-1616
B/D 14.00
MgO 88. 36
MgSO4 2. 80
SO 2 eff. 44
% MgO after
Hydration
71.79
67.18
66.36
65.68
64.47
64. 45
o
% of
Hydration
53.35
68.20
70.81
73.03
76.92
76.99
R-1310
B/D 43.00
MgO 90.58
MgSO4 4. 98
SO9 eff. 30.00
% MgO after
Hydration
90.61
90.00
89.34
89.08
88.51
86.55
% of
Hydration
0.00
1.96
4.09
4.93
6.76
13.07
-------
TABLE 21
Ka pKa
f-\
H2S03 *- H+ + H S03~ 1.54 x 10~ 1.82
O "7
H 803" *-H+ + S03~ 1.02 x 10~ 6.91
For comparison the pK values of other acids are:
_ pKa
H2C03 *- H + HC03 6TT7
H C03~ *- H+ + C03~2 10.25
Acetic Acid 4.75
-126-
-------
6.2.1 S02 Absorption
The absorption of gas in a venturi device follows con-
ventional mass transfer principles. In the venturi ab-
sorber used in this work, flue gas containing SC>2 enters
the converging section of the vessel and is accelerated
towards the throat area, passing over surfaces which are
irrigated by the absorbant slurry. The moving gas creates
a wave motion on the liquid surface until, at a critical
velocity, the forces resulting from the frequency and
amplitude of the waves exceeds the cohesive (surface tension,
etc.) forces of the liquid, When this occurs, some portion
of the wave is detached and dispersed into the gas stream.
The liquid, dispersed as droplets into the gas stream,
provides the media for absorption. The equation for point
efficiency,
(1-E) =
™
502 f
= exp
-Kg a
o^
s°
2i
~pe
i_
(6-16)
( E = Efficiency
( P = Partial Pressure of the Absorbed Gas
where ( Kg = Overall Mass Transfer Coefficient
Ib mole/hr. ft.2
( Z = Axial distance, ft.
( G = Mass Velocity, Ib mole/hr.ft.
( a = Area, ft2
was used in deriving relations for removal efficiency.
The surface area can be determined from the liquid to
gas ratio in the absorber and the mean liquid drop size.
S =
where
3.05 x 10° Vi
00 *g
S = Specific Surface Area, ft. /ft.'
V]_ = Volume of liquid, ft.
Vg = Volume of gas, ft.
Do = Mean drop diameter, microns.
(6-17)
-127-
-------
An estimate of Do can be obtained from the Nukiyama
and Tanasawa equation.
/ f s .1 ~\ 0 A.Z, / -, IE:
585 v/CT- +597.1 -^L—=-\ (1000 VI \ (6-18)
" ug nr- \ ^T ) \ v^J
V / ' /
where Ug = Velocity of the gas, meters/sec.
/ = Liquid density, gm/cc
^l = Liquid viscosity, poise
0 = Liquid surface tension, dynes/em
For cases where gas phase resistance is controlling an
analagous form of equation (6-16) based on system pressure
drop has been used (5) to correlate data as A P is a function
of gas flow and thus dispersed area. In this form
j i
I - E = exp ' -A A PB j (6-19)
where AP = differential pressure (in f^Q)
A&B are constants
For other cases other than gas film controlling an
over-all mass transfer co-efficient is defined as:
= 1 + n + n + n
k_a k^a ksa kra
(6-20)
where n = Henry's Law constant (Ib mol/ft. )
(Subscripts refer to the gas film, liquid film,
solubility, and reaction resistances respectively.)
As the chemical reactions involved in the absorption
are acid-base (fast) reactions, the final term in (6-20) is
zero. Recent investigations (1) have also determined that
the solids dissolution resistance is zero, and that the
liquid phase and gas phase resistances are each approximate-
ly one-half of the total. To incorporate the contribution
of liquid film resistance, we have correlated the .data in
-128-
-------
the form - ~]
1 - E = exp i -A APB _| (6-21)
(10PH)C(S02)
D;
where the denominator in (6-21) represents the liquid film
resistance. This correlation is shown graphically in Figure
29, which presents the results obtained using the final form
of the prediction equationI
r -1.014
1 - E = exp ! ' 2.666 (A P ) -3(6-22)
3.75 - 0.271 In S02I . .-6+0.031 pH)
E = Fraction S02 Removal
/N, P = Pressure Drop, Inches H20
S02i = Inlet Concentration S02/ PPM
This prediction equation was developed from both the data
obtained during the operations at Mystic Station in Boston,
and from some subsequent additional data obtained during
intial operations at an installation on a coal fired boiler.
This latter operation provided data for extension of the
equation to higher pressure drops.
The correlation also illustrates the relatively small
change in S02 removal efficiency over wide turn down ratios
of the power generator. The high efficiency of S02 removal
over the ranges of power output of a cycling generating
station is explained by the relative invariance of the
surface area available for mass transfer over that range.
Figure 30 is a plot of surface area, based on uniform drop
size, for a constant liquid rate, which shows only a 33%
reduction is surface area for a 4 to 1 turn down, equivalent
to operations between 40 MW and 150 MW Mystic #6 power
generator.
-129-
-------
CD.
CD
CD-
i
«
U_
u_
LU
l CE
CD
I
LU
CD
UD
CD
CO
CD.
\S)
8
—i—
10
12
DELTfi P - IN-H20
—I—
14
502=1000 PPM
N INLET 502=700 PPM
INLET 502=400 PPM
—i—
16
18
502 REMOVRL EFFICIENCY
EFFECT OF DELTfl P RND INLET SG2 CQNC
Fig. 29
-------
Fl'At
I7.5--
15.0-
195
Surface Area
Absorber Vol.1
/e.o
•7.5
5.0
IOO/S6C. GA5
75/SEC.
/
2O
ON PgOPCOTIES
OF Pug£ WATgO
1
40
i
80
1.000 CFM)
FIG 30
-131-
-------
An estimate of the accuracy of the correlation in pre-
dicting S02 removal efficiency is given below:
ACCURACY OF SO7 REMOVAL CORRELATION
Standard
Source of Data No. of Points Deviation
Boston Edison 600 • 4.0
Coal Fired Boiler 5 2.9
Boston Test Runs
Used to Check Corre-
lation 5 3.1
6.3 CENTRIFUGATION AND SOLIDS SEPARATION
The centrifuge was normally capable of separating 50%
of the incoming solids in the recycle slurry, but some in-
stances were encountered in which the centrifuge did not
perform satisfactorily, i.e., solids were not removed at a
rate sufficient to maintain control of recycle solids con-
centration.
In all, during the entire program, operations were
interrupted three times by inability to control slurry solids
by centrifuging the bleed stream. An analysis of the con-
ditions prevailing showed that during these periods mag-
nesium sulfate level was higher than normal in the recycle
liquor, an effect which arose because of changes in operating
conditions. In addition, the concentration of MgO in the
slurry was also shown to have an effect. These relations
are shown in Figure 31 which relates recycle solids concen-
tration to MgS04 level and pH.
6.4 REGENERATION MECHANISMS
The principal operation in the regeneration plant is
the calcination of the dryer product from the absorption
system. This can be most simply described as a thermal
decomposition:
MgS03 *-. MgO + S02
heat
-132-
-------
o
o
CoJ
O
o
CD J
I — • — •
O
Q_
I
COO
0°
u>
1 UJo
O
UJ
cc
o
o
oJ
o
o
"%.SO
6. 00
6.50 7.00
7.SO
8. 00
M
gS03 . 3H20 Case
8.50
9.00
9-50
10. 00
PH
EFFECT OF SYSTEM PH RND MGS04 LEVEL
ON RECY.SOLIDS "CONTENT -DflTfl TO 3/12/74
Fig. 31
-------
The rate of the decomposition is temperature dependent,
and some decomposition has been observed at temperatures as
low as 300°C. For the production of useful product in this
program, the rotary calciner has been operated at mid kiln
temperatures above 1000°F.
In addition to the simple decomposition of MgS03» the
calcination step also serves the purpose of reducing the
MgS04» a side product in the absorption reaction which is
also present in the feed. This reaction can be represented
by either of the following equations:
2MgS04 + C »- 2MgO + 2S02 + CC>2 (6-24)
,XG iiooop = -3lKCal
MgS04 + C MgO + S02 + CO (6-25)
± G 11000F = ~14 KCal
Other reactions can also take place in the kiln environ-
ment :
1/2 MgS04 =— 02 + 1/2 MgS (6-26)
MgS04 >- Mg02 + S02 (6-27)
2 CO + S02 *~2 C02 +S° (6-28)
1/2 02 + S02 —*-S03 (6-29)
All of these reactions are both time and temperature
dependent and the calciner is equivalent to a multi-zone
linear reactor. In the course of the program, in addition
to the production of thousands of tons of acceptable re-
generated MgO, upset conditions have occasionally resulted
in the formation of high concentrations of 803 and at other
times in the formation of elemental sulfur.
Attempts have been made to study these upset conditions
in the laboratory, and thermal decomposition studies have
been carried out on a laboratory scale using infrared spec-
troscopic techniques to investigate the phenomenon of sulfur
-13.4-
-------
formation by detection of the appearance of its precursors.
Those studies, which are described more fully in Section
6.1, have not as yet yielded any specific information on the
formation of elemental sulfur; however, they have yielded
information indicating that the decomposition process for
MgSC>3 obtained from the trihydrate form and the hexahydrate
form may be different.
Retention time in the calciner can be determined from
the following equation:
6 = 0.19 L
N D S
where 0 = time (min.)
L = kiln length (ft.)
N = Rotational Speed (RPM)
D = Dryer Diameter (ft.)
S = Slope (ft./ft.)
For most operations undertaken during this project re-
tention time was approximately one hour. Because of the use
of tube coolers on the calciner, an additional one hour was
required before final discharge to the product conveyors.
Studies of heat transfer in a rotary kiln (8) have re-
sulted in the formulation of prediction equations for the
solids - gas temperature relations considering the following
1) Gas to solid heat transfer.
2) Gas to calciner wall heat transfer.
3) Calciner wall to solid heat transfer.
4) Conduction thru the kiln wall.
5) Conduction and convection to ambient air
from the kiln.
Such studies have not considered the additional compli-
cations of chemical reaction in the charge or radiation from
the flame at the discharge end.
-135-
-------
The data resulting from the test program phase of this
project were analyzed to provide correlations for both the
percent of MgS04 in the product and the bulk density of the
product. These correlations are given in the following equations
6-31 and 6-32, and provide a means of determing the pro-
cessing condition necessary for the production of an active
magnesia and control of magnesium sulfate concentration.
-136-
-------
CALCINER OPERATIONS
CORRELATION FOR PREDICTION AND CONTROL
OF % MgSO^ IN REGENERATED MAGNESIA
% MgS04= -90.2+(T'*A)+(C*B)-2.46 (%MgSO4 in feed) +0.989 (%MgSOa in Feed)
+ 4.1 (%02 in Acid Gas) + 0.28 (Feed Rate, Lb./Min) (6-31)
I
M
UJ
I
1/2
Where T1 = (1700 - °F Mid Kiln Temperature)
A = -0.870 + 0.185 (% MgSO4 in Feed)
B =15.7-23.4 (Furnace Draft) - 2.19 (% O2 in Acid Gas)
-0.61 (% MgSO4 in Feed)
C = % Carbon in Feed
Statistics: Standard Deviation = 5.4
Multiple Correlation Coefficient =0.77
Confidence Level of F Ratio = 99.9% +
No. of Data Points = 207
-------
CALCINER OPERATIONS
CORRELATION FOR PREDICTION AND CONTROL
OF BULK DENSITY OF REGENERATED MAGNESIA
u>
00
BD=169.2 + (T'*A)+(C*B)+0.741 (% MgSO4 in Feed) -0.744 (% MgSO3 in Feed)
-4.9(% O2 in Acid Gas)-0.4(Feed Rate, LB./MIN.)-104(Furnace Draft) (6-32)
1/2
Where T" = (1700-°F Mid Kiln Temperature)
A = -0.891+0.166 (C)
B = -28.4+3.44(C)+71.2(Furnace Draft)+1.95 (% O2 in Acid Gas)
-1.24(% MgS04 in Feed)
C = % Carbon in Feed
3 '
BD = Product Bulk Density, LB./FT0 (Operations @ 1.82 RPM
Add 5 to BD for 1.56 RPM)
Statistics:
Standard Deviation =7.2
Multiple Correlation Coefficient = 0.79
Confidence Level of F Ratio = 99.9% +
No. of Data Points = 456
-------
6.4.1. Magnesium Sulfate Control
It has been shown (6-14, 6-15) that magnesium sulfate
formation occurs to some extent in the process due to both
the presence of oxygen and 603 in the flue gas entering the
venturi and some additional conversion of MgSC>3 to MgS04 in
the dryer.
It is necessary to reduce and control the magnesium
sulfate concentration in the recycled MgO in order to prevent
both: ultimate conversion of the absorbent to the sulfate
and to prevent centrifugal separational problems associated
with the increasing viscosity of more concentrated MgSC>4
solutions.
Direct reduction of magnesium sulfate in the calciner
by operation at temperatures sufficient to thermally decom-
pose MgS04 (2000°F) would cause "dead burning" of the
magnesia product and render it useless for absorption of
S02- It is evident from 6-24 and 6-25 that the presence of
carbon in the calciner feed can result in the reduction of
the MgS04 contained in it at temperatures which still pro-
duce a useful product.
At the outset of the program carbon, in the form of
coke, was added to the process. However, difficulty in
obtaining a grade of coke sufficiently low in ash prevented
a controlled program of carbon addition. After the initial
operations of the F0G.D. System it was noted that the in-
coming calciner feed contained some carbon which had been
removed in the venturi as uncombusted fuel or "oil smut"
from the flue gas. A number of analyses of the calciner
feed revealed that the feed had an average carbon content
between 0.5% and 1% which was sufficient for control of the
magnesium sulfate in the product. Normally, the calciner
was operated to reduce the MgS04 content from 11% in the
feed to 5% to 8% in the products, i.e., reduction of 50% to
70% of the incoming sulfate.
Equation 6-31 provides the empirical relationship between
the control variables and the MgS04 content in the regenerated
product. This is presented graphically in Figure 32.
•139-
-------
o.
OJ
tO-
Q
O
CO co
C)
^00
900
1000
1100 1200 1300 UOO
MIDKILN TEMP-DEG F
1500
1600
1700
CRLCINER OPERRTION - EFFECT OF
MIDKILN TEMP. & 7.C GN XMG504 IN PRODUCT
32
-------
6.4.2 Magnesia Activity Control
The principal control parameter in the calciner opera-
tion has been the bulk density of the regenerated MgO. Values
of calciner product bulk density between 20 and 25 lb/ft3
have resulted in better performance in the F.G0D. System,
and the property can be readily measured by the calciner
operator using several commercially available devices. In
this work a "paint pigment volumeter" was used.
The effect of carbon content and mid-kiln temperature
on the product bulk density, developed from equation 6-32,
has been shown in Figure 1.
6.5 MgO LOSSES AND REGENERATION CYCLES
A) Losses
In early operation of the system, losses of MgO were
high. The main losses resulted from: frequent clean out of
plugged equipment, discarding both oversize materials and
any spilled solids, entrainment to the stack, overflows, and
cleaning the absorber when the system was shut down.
By June 1973, the installation of lump crushers and re-
routing the dryer off gas to the absorber, reduced the loss
substantially but did not eliminate it. Spills were still
discarded and the venturi absorber was still drained on shut
down. Another continuing contributor to the loss of
alkali was the higher than design gas flow, which resulted
in excessive entrainment of absorber slurry.
In the final period of operation, from March 1974,
controls were imposed to eliminate many of the loss points.
During part of this period a careful measurement of system
losses were made in order to identify the sources which
could not be eliminated in the current program for future
design improvements.
The loss history for the entire operating period is
accounted for on a rate basis in Figure 33.
In the initial operations, startup and shakedown,
system losses averaged 1075 Ib/hr. of operation or a total
-141-
-------
-1
OO
CQ
- I O.
UD
CO
D_
rvj
PRE-STARTUP
AND
SHAKEDOWN
AVERAGE LOSS PER PERIOD
TEST
AND
DEVELOPMENT
OPERATING
U/72
6/73
500
1000 1500 2000 2500 3000
OPERATING HOURS
. 3500
PROCESS MGG CONSUMPTION
BY OPtRRTING PERIODS
MYSTIC STATION
REGENERATION
4000
U500
Fig. 33
-------
of 606 tons as MgO in the 1127 hours logged from initial
startup in April 1972 to the completion of the first phase
in June 1973.
The second period, the test and development program,
had an average MgO loss rate of 415 Ib/hr. of operation.
The total loss in the 1350 hours logged from July 1973 to
February. 1974 for this phase of operations was 151 tons as
MgO.
Similarly, in the final period from March to the end
of the program, losses averaged 234 Ib/hr. as MgO; a total
of 151 tons of MgO lost in 1293 hours logged.
A comparison of the measured losses at Mystic Station
(cross hatched) and the losses from the regeneration plant
for the March through June 1974 period are also shown in
this figure. It should be noted that the regeneration plant
losses were approximately constant during the entire program.
In the final operating period at Mystic Station, gas
flow to the absorber was maintained at the design rate and
the system was run as a closed loop with no vessel drainage,
returning the small spills back to the process. Measurements
were made at the following 14 loss points (Table 22) at
Boston for the continuous operations period and the silos
were emptied and the contents weighed before and after the
run for accuracy of inventory measurements.
Table 22
Potential Sources of MgO Losses at Mystic Station
1) Stack
2) Centrifuge Washing
3) Centrifuge Case Leaks
4) Pump Packing Gland Leaks
5) Absorber Overflow
6) MgO Slurry Tank Blow-Down
7) MgO Slurry Tank Overflow
8) Centrate Tank Overflow
9) Solids Loss at Dryer Feed End
10) Dust Loss at Dryer I.D. Fan
11) Dust Loss at Expansion Joints
-143-
-------
12) Spillage at MgO Feeder
13) Spillage in MgS03 Belt Gallery
14) Spillage at Truck Loading Point
The tests, conducted over 13 days, in which 336,470 Ibs.
of regenerated material were fed (an additional 2,504 Ibs. of
MgO added with the fuel oil was also accounted for), showed
a loss of 0.37 tons/operating day at the absorber system,
distributed as follows:
Ton/Day
Loss to Stack 0.13
Absorber Overflow 0.14
Misc. Measured Loss 0.07
Unmeasured Loss (by difference) 0.03
0.37
With an average MgO consumption in the process of 10.61
tons/day during the period, the 0.37 tons/day loss represents
a 3.5% loss rate at the F.G.D. System.
The greatest losses were found however, to occur at the
regeneration plant. Here 1.5 tons/day of equivalent MgO is
lost from the neutralizer system overflow and another 0.5
tons per day is scalped off, for future reclamation, before
being pulverized.
Both of these losses would be virtually eliminated in a
full size regeneration plant.
B) Regeneration Cycles
The information on total losses has been combined
with the other inventory records including the receiving
records for virgin MgO shipments to evaluate the number of
regenerations. In the start-up and shake-down period, losses
were 606 tons of MgO, with 559 tons of regenerated MgO re-
cycled to the Mystic #6 System. This rate of loss limited
the number of recycles from 2 to 3 before the material was
lost from the system.
-144-
-------
During the test and development period, 1717 tons of
regenerated MgO were returned to the system, while 338 tons
of MgO equivalent were lost. This corresponds to 5.1 cycles
for the magnesia before it is lost from the system.
During the final period, 151 tons of magnesia were
lost, while 875 tons of regenerated alkali was shipped back
to Boston. This would correspond to 5.8 cycles; however,
during this same period there was an inventory build-up
with 80 tons of equivalent MgO inventory at the beginning,
and 214 tons remaining at the close of the program so that
only five regeneration cycles can be accounted for.
-145-
-------
7.0 DESCRIPTION OF THE TEST AND DEVELOPMENT PROGRAM
A test program was developed for the "New England
S02 Control Project" on the basis of two years operation
of the system.
This planned program was divided into three phases,
first a period of start-up and break in operations of the
integrated plants carried out over a three to six months
period. Then a twelve month period in which several
variables, considered to be most important for successful
operation of the system, were to be investigated by operat-
ing the plant setting the variables to evaluate perfor-
mance and determine the best methods of overall system
control. A final period of three to six months of opera-
tion at optimized conditions was planned to demonstrate
the long term capability of the project and provide
further information on the results of continuous "on stream"
running.
The actual operating periods for the Start-Up and
Break-In Tests and the Variables Screening Tests deviated
considerably from the proposed plan to conform to the
situations which are described in the previous sections.
A further complication to controlled operation arose
because of the economic-dispatch method of setting boiler
load which favored maximum load on the Mystic 6 unit when
it was fired with high sulfur fuel. For this reason, the
F.G0D. system treated a higher than average gas flow based
on the load data supplied by Boston Edison Company. Except
for disruption of any test plan, the effect on operation
was not serious as the F.G.D. System had been designed for
continuous full load operation although complications were
introduced by the greater than design gas flow.
It was not till the last six months of the program,
after approval was obtained from local authorities for a
partial gas bypass, that the F.G.D. System could be
operated at its design gas flow while the boiler was fired
at its maximum rating.
-146-
-------
7.1 TEST PROGRAM
While the time schedule of the program was changed,
the goals of the three periods remained the same and the
test plan is summarized to show the scope of the project.
7.1.1 Start-Up and Break-In Tests
An initial period of testing started after mechanical
completion of the plant had been accomplished. A prestart
period was included which was limited to the accumulation
of 2 days production of MgSC>3 at the. Boston Edison site.
This initial period was to demonstrate the operability
of all equipment at Boston Edison as well as the integrity
of the piping and equipment. After pre-start-up, the start-
up and break-in period continued under Chemico supervision
and shipment of MgS03 to the calcining facility was ini-
tiated to begin integrated operation at both sites.
During this period operating, testing and analytical
procedures were tried and evaluated; equipment tested,
and additional training completed. In addition, the first
phase of testing started in order to find the conditions
necessary to achieve even a limited level of system opera-
bility.
This initial period provided wide ranging operating
conditions as flows and equipment settings were adjusted
to and modifications made to obtain a satisfactory level
of semi-continuous performance of the recovery-regenera-
tion facilities. All data collected during this period
were accumulated for use in the statistical analysis pro-
grams used to assess the program.
7.1.2 Variables Screening Tests
Following the Start-up and Shake-down period, a test
and development program, designed to investigate the effects
of primary process variables on the overall operation was
initiated. After determining conditions which made the
plants functional during the first period, certain pro-
cess variables were chosen for study and the effect on
plant operation gauged.
-147-
-------
The program was to establish a model of the principal
process mechanisms from which the effect of changes in
operating conditions could be evaluated. From this model
(and consideration of pilot plant results), preliminary
values for the parameter excursions were chosen, and experi-
mental sets planned so that the data collected could be
used for optimization of the process. After initial tests
to evaluate the effect of the variable excursions, and to
determine if modifications to the scheme are necessary,
further optimization followed an evolutionary plan in order
to minimize system upset. That is, the range allowed to
the parameters under examination were regulated in order
to insure minimum system upset and negligible effect on
power plant operation.
Information was collected at these operating conditions
for a sufficient length of time to insure that the effects
could be determined with sufficient reliability by the use
of statistical methods. Other variables or combinations
of variables were then selected and the process repeated
to provide information for the data file in order to es-
tablish regression equations to determine variable effects
and interactions at levels not specifically tested.
7.1.3 Long Term Tests
During the design of the variables screening portion
of the Operational Tests Program the assumption of careful
control of all variables was assumed. For any test the
boiler load was to be held within specified narrow limits
for the test period in order to evaluate the performance
of the associated flue gas scrubbing equipment under the
controlled conditions. The long term tests were to follow
the normal operating ranges of the power boiler. Figure 34
shows seasonal load variation based on data supplied by
Boston Edison Company and Figure 35 shows average daily
load swings for both spring and summer operation. By
following the load swings the effect on F.G.D. System
operation could be determined.
7.2 STATISTICAL 'METHODS
Throughout the term of operation of the program, large
amounts of process data were generated. In order to effi-
-148-
-------
— 1
1
LOAD [_
IU
ac
<
1 50
1 4O
I 30
1 2.0
1 1 0
1 00
90
SO
70
60
50
40
3.0
20
10
AVERAGE"
SPRING
SUMMER
MYSTIC
DAILY LOAD SWING
FIG . 34
I 2 3 4 5 6 789 10 II 12 \ 2
A 56789 IO II 12
MIDNIGHT
TIME
-149-
-------
MYSTIC 6
SEASONAL LOAD SWING
Q
O
_l
LJ
90
85
80
75
70
65
60
> 55
50
JFMAMJJA50
-150-
-------
ciently assimilate, disseminate, analyze, and store this in-
formation, a computerized, statistical method of handling
the data was employed. As a result, it was possible to
effect a real-time feedback to the process of correlated
and trend results at the time they would be most useful,
along with accumulating a permanent data bank available for
analysis and retrieval. The statistical computer program
package utilized permitted the use of the same format data
file as input to a wide range of sub-programs, including
regression calculations, x-y plotting, file listings, trend
plots, etc. This flexibility reduced the number of files
which needed to be maintained, in addition to the evident
speed advantages.
Figure 36 illustrates the flow of information between
both the scrubber and calciner operating systems and the
computer data bank. Operating conditions and analyses were
entered on punch cards on a daily basis for primary storage
and transferred to computer disk files for further pro-
cessing. At monthly intervals the following statistics
were generated for all variables: average, maximum, mini-
mum, standard deviation, and a percent change relative to
a base period. The final stage in the data flow was the
integration of the operating log and analysis files, plus
the inclusion of appropriate time lags for the scrubber
and regeneration facilities to form single disk files suit-
able for input to correlation and plotter computer programs.
This process data bank is now a source of rapid retrieval
of process information for continued analysis and for com-
parison with results from other installations.
The process correlations presented in previous sections
were developed primarily using the data bank as input to a
stepwise multiple linear regression program. This approach
minimizes the number of computer runs required to determine
equations which are statistically significant, and which
have been evaluated for consistency with observed data.
F ratio tests are employed to establish significance levels
for testing of equation variables, and coefficients are
calculated by least squares techniques. With respect to the
S02 removal efficiency correlation, log transformations
were required by the regression program. Continued investi-
gation of mass transfer coefficients may lead to the use of
non-linear programs.
-151-
-------
DATA. FLOW
01
to
I
SCRUBBED
CALCIM&R.
LOO
FIL&-:
DATA,
FILE-
DISPLAY
FILE-
FIG. 36
-------
As discussed in Section 7.0, Description of the Test
and Development Program, normal excursions in the process
variables were sufficient to satisfy the required scope of
the test program variation in levels of operating data.
-153- :
-------
8.0 PERFORMANCE TEST RESULTS
Performance tests on the Chemico Basic Magnesia System
installed at the Mystic Station were conducted by York Re-
search Corporation for Boston Edison Company. The testing
period extended from February 27, 1974 through March 9, 1974
following a preliminary period in early February for equip-
ment set-up, familiarization, establishing gas velocity
profiles and gas flow measurement.
The performance testing used the methods detailed in
the Federal Register, Vol. 36, No. 247, 1971 for:
a.) Gas analyses by Method 3
b) Particulate removal efficiency by Method 5
c) SC-2 removal efficiency by Method 6
in addition, particulate removal efficiency for a range
of particle sizes was done using the method developed by
J. A. Brink (9). The results are presented for S02 removal
efficiency in Table 23, for overall particulate removal
efficiency in Table 24, and for particle size removal
efficiency in Table 25. Oil analyses data for the test
period are presented in Table 26. These data has been
abstracted from the York Research Corporation's report
No. Y-8419 dated April 11, 1974.
-154-
-------
TABLE 23
S00 REMOVAL - TEST RESULTS
BOILER
TEST LOAD: INLET GAS S02 IN S0'2 OUT % S02 SO? OUT
NO. MW RATE: ACFM PPM - VOL. PPM - VOL. REMOVAL LB/106 BTU
1
M
Ul
(j\
\
1
2
1
146
1
144
446
486
,953
,991
926
1004
-.1
.5
71.
89.
1
0
92
.91
.3.
.1
0
0
.125.
.199
2
151 658,207 983.9 63.3 93.6 0.201
148 1 503,233 833.3 86.6 89.6 0.243
1) Flue gas partially bypassed to stack to attain design gas flow.
2) No flue gas bypass.
-------
TAB Li.' 24
PARTICIPATE REMOVAL - TEST RESULTS
BOILER
TEST LOAD: INLET GAS PARTICIPATES: LB/HR % PARTICULATES
1
U1
1
NO.
1'
2.
3
4
MW
146 -1
1
144
2
151
148 l
RATE :
446
486
658
503
,9
,9
5
ACFM
3
91
,20
,2
3
7
3
3
2
IN . OUT
80 116
*
32 11"5
399 150
1
51 82
REMOVAL
69. 5
50.4
62.4
45.7.
LB/106 BTU
0
0
0
0
.072
.084
.111
.068
1) Flue gas partially bypassed to stack to attain design gas flow.
2) No flue gas bypass.
-------
TABLE 25
PARTICLE SIZE RESULTS
Test Removal Efficiency Removal Efficiency Overall
No. Above 1 Micron >- Below 1 Micron Efficiency
1 89.5 53.6 85.5
2 89.5 57.6 85.8
3 80.75 64.6 74.4
4 94.4 57.2 88.9
-157-
-------
TABLE 26
OIL ANALYSIS
Parameter
BTU/lb
Specific Gravity
% Sulfur
% Ash
% Carbon
% Hydrogen
% Nitrogen
% Oxygen
Test #1
2/27/74
18,284
A
.9431
2.15
0.09
84.77
11.34
0.2
1.45
Test #2
2/28/74
18,348
A
.9433
2.10
0.10
84.79
11.42
0.2
1.39
Test #3
3/5/74
19,083
B
.9316
1.89
0.07
84.54
11.24
0.1
2.16
Test #4
3/9/74
19,145
B
.9366
2.04
0.07
84.39
11.39
0.1
2.01
A - Taken at 126°F, B - Taken at 130°F
-158-
-------
9.0 DATA
9.1 OPERATING RESULTS
Operating data are summarized and presented as monthly
averages for the period from the initiation of Planned
Operational Testing through the end of the program. Includ-
ed in this information are:
Table 27, Operating Conditions for the FGD System
Table 28, Stream Compositions for the FGD System
Table 29, Operating Conditions for the Regeneration
Unit
Table 30, Stream Composition for the Regeneration
Unit
9.2 RANGE OF SIGNIFICANT VARIABLES
Frequency distribution and the range of the important
variables measured in this study are presented in Figures
37 through 63.
9.3 MISCELLANEOUS DATA
Table 31, Sulfur Content of High Sulfur Oil for
Mystic #6 Boiler
Table 32, Oil Ash Analyses
Table 33, Regenerated MgO Vanadium and Nickel
Content
-159-
-------
9.1
OPERATING RESULTS
-16,0-
-------
TABLE 27
FLUE GAS DESULFURIZATION SYSTEM
OPERATING CONDITIONS
VARIARLF
5+6/73
MONTHLY DATA AVERAGES
7/73 8+9/73 10/73
11
/73
12/73
SO? (PPM)
IN
OUT
SRF.MHV
PM
603
78
87
6
• 01
.65
.07
.86
789
106
86
7
.45
.50
.24
.02 7.14
734.
86.
67.
7.
45
53
03
?1
741
107
84
7
.80
'.31
'.24
.13
594
91
83
7
.83
.05
.89
.18
POWER Pt:ANT
R
0
(
ATFO-tvO
IFF. PRESSURES
IN.H20)
TOTAL
MIST
DR/FR
132
q
«
0
.02
.79
.89
.07
127
10
5
0
.26 116.59
.05 9loi
.06 2.63
.04 0.07
123.
9.
?.
0.
6?
0?
49
07
106
7
1
0
. 16
.25
.40
".08
124
9
0
0
.02
.30
.92
.07
TF.MPFRATHrtESCF)
— ARCritJDf-U
INI.FTq
GASOUT
SLURRl
•MGSLUR
-ORYF.K
GASOUT
PROO.
311
113
123
186
483
295
.30
.87
.60
.83
.13
.06
306
116
128
188
469
315
.93 293.46
.73 117.43
.00 123.01
.93 173.16
.51 469.04
.77 350.0?
300.
121 .
124.
177.
449.
339.
69
17
61
08
89
8?
294
124
1.23
184
443
332
.97
.08
.25
.26
.16
.13
288
120
131
172
441
340
.12
.98
.96
.43
.08
.68
FLOWS(GPM)
CTRC. 14
CF.NTFO
MGSLUR
748
118
5?
. 10
.62
.29
14458
110
.67 13778.52
.57 102191
44.94 49.98
13484.
114.
39.
79
1?
??
12688
128
58
'.96
.27
.69
14185
120
45
.61
.60
.59
CFNTRTFllGE
TORO.(G>S)
15
.58
23.22 15.48
-161-
23.
16
24
.71
22
.82
-------
TABLE 27 (CONT'D)
FLUE GAS DESULFURIZATION SYSTEM
OPERATING CONDITIONS
VARIABLE 1/74
s n 2 ( p p M )
IN 634.77
OUT 114.81
*REMOV ai.27
P H . 7.24
POWER PLANT
RATE(yW) 11.3.35
DTEE. PRESSURES
( IN.M20)
TOTAL 8.23
MIST 1.44
DRYER 0.05
TEMPERATllRES(E)
-ARSnR»ER
IMLETn 293.55
GASmiT 124.70
SLURR. 129.30
MGSLUR • 176.27
-ORYER
GASOUT 4P0.83
PRnn. 341.62
FLOWS(GPM)
CIRC. 14971.15
CEMTEn 120.55
MGSLUP ' 23.55
CENTRIFUGE
TORR.'CG>S) 22.81
MONTHLY DATA AVERAGES
2/74 3/74 '4/74 5/74 6/74
8.30.12 763.36 667.60 634.10 563.
107.61 125.
82.58 76.
7.20 7.13 7.01 6.94 6.
136.05 138.33 130.10 136.36 127.
7.02 6.47 5.90 5.53 6.
0.50 0.61 0.85 Ol77 1 .
0.05 0.06 0.07 Ol07 0.
299.63 287.88 2fM.96 287.04 287.
129.40 129.25 127.1/4 129.59 130.
128.26 128.87 130.88 130.32 130.
173.79 177177 179. ?4 176.74 177.
426.30 426.02 408.77 400.37 394.
349.96 334.36 120. «7 321.86 3l«.
14274.07 15492.71 15421.59 15245.29 1fl725.
123.68 130.12 129.42 130.01 130.
17.67 14.65 13.96 20.19 16.
16.90 13.36 12.97 10.87 7.
-162-
31
19
98
98
59
25
14
08
23
39
16
10
82
27
40
90
98
15
-------
TABLE 28
FLUE GAS DESULFURIZATION SYSTEM
STREAM ANALYSIS
MHMTHLY DATA AVERAGES
VARTARI.F 5*6/73
HRYE*
* SPL. 81.79
XM'ISm 60. ?5
*MGMI4 11.56
*MGp 3.58
CFNTPTFnGE
? S;H. . . 75.68
.3 M G S (1 3 5 3 . 0 7
SMGsn/4 7.47
*MGO ?.01
RECYCLE
? StlL. 10.49
PH 6.95
-FILTRATE
SMGS03 1.05
' KMGSD4 16.56
-CAKF
%MGf,03 5?. 67
*MGsn
-------
TABLE 28 (CONT'D)
FLUE GAS DESULFURIZATION SYSTEM
STREAM ANALYSIS
VARIABLE
MONTHLY DATA AVERAGES
1/74 ?/74 3/74 4/7/1
1.73
MOTHER LTOUnP
% SOL. 8.37
-CAKF
1 .82
3.53
3.03
4.6?
3.0ft
6.57
5/74
2.97
3.13
6/74
ORYF.R
* sni.
SMGSD3
XMGP.
C F N T R T F 1 ! ft E
* SHL.
'*Mr,5un
fcMGsna
*MGH
RECYCLE
* SOI.'.
PH
-FILTRATE
•XMGS04
-CAKE
* M G S 0 1
88.57 87.61 85.41 87. si 09^3 86.78
65.00 65.89 64.05 60.70 63^98 59.93
16.36 11.61 10.32 11.10 11.68 14.77
2.78 3.47 3.39 5.75 4.09 3.44
86.01 81.27 82.20 79.^0 79.17 78.57
46.61 48.64 . 46.93 4ft. ?n 47'.56 45.84
lf>. 13 6.75 6.86 /.33 8.07 10.47
'•21 1.47 1.70 3.61. 2.43 2.06
11.90 7.54 8.39 10.37 8.24 7.86
7.11 7.18 7.19 7.?1 7.n 7.17
15.83 12.41 13.73 15.08 16.69 18.58
• „ n
2.26
3.12
XMGH
2.10
2.25
3.75
4.00
4.93
4.03
-164-
-------
TABLE 29
REGENERATION SYSTEM
OPERATING CONDITIONS
VARTAHLF 54.4/73
c a L c T M E R
-
•
TEMPERATIIPES(F)
M 0 K T L M 122J./2
GASnUT 650.27
P R n D F Y 159. n 1
SOLID FLnwS(PPM)
M G S n 3 * 3 . 8 7
COKE 0.84
NEUT. PH Llo. 5.64
AC
FE
in PLANT
r.n GAsrS)
Cn n.io
MONTHLY DATA AVERAGES
7/73 84-9/73 10/73
12?8.00 1189.51
665.56 588.73
1.'8.80 H7;i?
59.89 66.35
0.97 0.25
5.25 5.51
0.17 n. a.
1091 .74
569. *1
161.33
7?. 05
0.0
5.94
n. a.
11/73
1001 .
555.
202.
76.
0.
5.
n. ;
24
66
57
81
0
75
l.
12/73
983
563
149
65
0.
5
n
.28
.23
.07
.47
0
.53
.a.
r'2 *.96 5.20 10.?5 4.43 4.?9 3.69
sn;? 3.5fl n. a. n. a. n. a. n. a. n. a.
-165-
-------
TABLE 29(C6NT'D)
REGENERATION SYSTEM
OPERATING CONDITIONS
MONTHLY DATA AVERAGES
VARIABLE
C A L C T N E R
-TFMPFRATUPES
MCKTLN
GASH u T
p R n n F v
-SOLin FLOWSf
MGSH3
COKE
NPUT. PM LIO.
ACID PLAMT
FEED GASTX)
en
n?
sn?
1/74 2/74
(F)
958.45 1045.89
53*. 55 571.68
154.7? 19?.? 3
PPM)
62.60 76.81
0.0 0.0.
5.72 5.11
n. a. n. a.
5.01 5.36
n. a. n. a.
3/74
1005.60
546.98
168.95
76.45
0.0
5.19
n. a.
5.31
n. a.
4/74 5/74
1019. *5 1072.56
550.53 560.16
165.4? 179.00
77.80 77.00
0.0 0.0
5.A8 5.7?
n. a. n. a.
4 . n 9 4.97
n. a. n. a.
6/74
1052.75
538.03
172.61
78.94
0.0
5.66
n. a.
4.66
n. a.
-166-
-------
TABLE 30
REGENERATION SYSTEM
STREAM ANALYSIS
VART ARLF
FFFO
* H?f]
t M G S H 3
* M G s n a
*Mr,n
?! C A * R M
CALCTMF.P
PRODUCT
% M G S fl 3
x M G s n a
* M G (1
S02FFF
RLKDFN
5+6/73 7/73
16.47 1.7.79
58.94 5B.92
10.31, 10.93
7.56 7.S2
n. a, n. a.
0,06 0.02
5.38 6.05
91.24 89.59
52.27 47.33
28.30 31.33
fl-t-9/73 10/73
16.86 14.88
57.30 59. tn
1 1 .59 12.63
8.59 6 . *9
n.a. 1 .n5
0.0 O.na
8.71 10.47
93.38 85^6
23.43
40.18 34.08
11/73 12/73
15.56 13.20
59.09 61.53
12.54 12.38
7.28 7.78
0 ". fl 1 0.45
Oi.0 0.01
10.91 B.2fl
84".04 86.11
34.04 49.46
-167-
-------
V ft « I A R L F
TABLE 30(CONT'D)
REGENERATION SYSTEM
STREAM ANALYSIS
MONTHLY DATA AVERAGES
?/74
5/74
6/74
C A L f T :\l F p
FFF.D
% H?H
XMGS0.3
*Mr,Sn« '
x M G :i
9; CAROM
CALCTNJFR
PRODUCT
XMGsm
X M G S PI 4
*MGn
S02F.FF
B 1, K n E M
1?
' 65
10
s
1
0
17
84
,7
.MR
.31
.95
.36
.65
.00
.5?.
. 1 3
.65
Q
6/1
1 0
6
1
0
16
R6
3/4
?7
.86
.64
.13
.88
.02
.02
.75
.93
.00
.76
13.33 15.06
64.13 5 R . 3 R
6.9? 9. QP
6.79 7.RO
2.05 1 .06
0.37 0.0
4.37 4.0?
8 R . ? 1 90.10
24.0? 3?.SR
1?
61
11
4
1
0
1?
83
20
.70
.97
.95
.7ft
.04
.0
.4?
.80
.85
15.
58.
1 1.
5.
1.
0.
7.
86.
25.
20
66
87
75
1 3
12
60
69
80
-168-
-------
9.2
RANGE OF SIGNIFICANT VARIABLES
-169-
-------
CO
I—
°-<
I
.'. ED"
o
i
"o
0_
o_
1/7
°60
65
70
—r~
75
—I f—
80 85
PCT.
90
—i—
95
100
105
BOSTON ED.OPERRTI0N - 502
FREQUENCY DISTRIBUTION
RFMOVR
EFF.
Fig. 37
-------
CJ_
CO
^oo
moo
soa
1000
120O 14 GO
1600
1800
PPM
BOSTON ED.QPERRTIQN
FREQUENCY DI SIR I BUT I ON
S02 INLET CONC
Fig. 38
-------
o
o_
\r>
o
0_
CO
50
100
150
200
PPM
BOSTON ED.OPERRTI3N -
FREQUENCY DISTRIBUTION
250
S02
300
350
400
450
OUTLET CONC
Fig. 39
-------
f—
OJ
I
^
C.J
un
V
/w
LOO
I SO
2OO
25O
3OG
BOSTON ED.OPERflTION
FREQUENCY DiSTRiBUTiON
L_ c ri U
Fig. 40
-------
°0
IN. H2G
BOSTON ED-OPERRTION - TOTflL DELTfl
FREQUENCY DISTRIBUTION
P
Fig. 41
-------
o
o
oT
O
O
O
OJ
CD
LT
I
O
o.
o
0_
\n
240
260
280
300 320
DEG. F
340
360
380
400
BOSTON ED.OPERRTION
FREQUENCY DISTRIBUTION
RBSOR6.INLET TEMP
Fig.42
-------
CD
in.
CD
CD
LT>
(SD
en
I
CD
0
I MO
20O .
220
R n s T n M F n n p F p^ p T T n IM
LJ L v J 1 l_y I 'J • L..' a ^__; | I ) II ! i >_/ 1 VJ
FREQUENCY iJ I SIR 1 BUT I ON
Fig. 43,
-------
o
o
CD.
o.
oo
CO
o.
U_co
CD.
CD
CD.
0_
C\)
300
350
400
450
F
500
550
600
650
700
BOSTON ED-OPERRTION
FREQUENCY DISTRIBUTION
DRYER GRS OUT TEMP
Fig. 44
-------
600
650
F
BOSTON ED-OPERRTION
FREQUENCY DISTRIBUTION
DRYER PRODUCT TFMP
Fig. 45
-------
CD.
BO
CO
I—
Q_
VD
I
CD.
zr
70
LOO
I OS
U O
-"; ."n I i n ^;
..'vj L. j U --
BOSTON ED.OPERRTION
FREQUENCY DISTRIBUTION
- DRYER PRODUCT
Fig. 46
-------
Q_,
(-• ED
00 .
O
I
SO 7O SO
PCT- MG-503
100
l 10
180
BOSTON EDaOPERRTION - DRYER
FREQUENCY DISTRIBUTION
PRODUCT
Fig. 47
-------
Q__ i
00
M
I
LT7 I
LQ
20
PCT-
25
- —r' ~
30
BOSTON ED-0PERRTI3N
FREQUENCY DISTRIBUT ION
nR Y PR p Rnr;1
!_• I > i ' 1 i 1 i i < J ! ' l_-
r
Fig. 48
-------
cr> i
LO 4
in
00
to
I
—i
45
10
15 ^o
PCT- MGO
BOSTON ED.OPERflTION
FREQUENCY DISTRIBUTION
DRYER PRODUC1
Fig. 4;9
-------
CO
oo
30O
4OO
BOSTON EDoOPERHTION
FREQUENCY DISTFU BUT ION
- CENTRIFUGE PEED
Fig. 50
-------
en
CD-
rr
10
•-40 50
Ft.-Lb. /32
60
BO
STON ED.OPERflTlON
CENTRIFUGE
TORQUE
FREQUENCY DISTRIBUT13M
Fig. 51
-------
00
Ul
i
ST.)
O i
CD
r\i
ISOOO
GTM
BOSTON ED.OPERflTI Oh
FREQiJENCY D I SIR I BUT I ON
l 60OO
p p" r- Y
V
I 70OO
i Boaa
'DOOO
Fig. 52
-------
j 1
'
CX)
CTi
I
ISO
•i 55
ED.OPERflTION
BOSTON
FREQUENCY DI3TRIBUT I ON
- RECY- SLUR
T r M P
! L_ I I i
Fig. 53
-------
00
^J
I
u_ —
ED
en
CD-
/
lO
t 5
25
PCT.
40
BOSTON, ED.OPERflTION - RECYCLE SOLID!
FREQUENCY DISTRIBUTION
Fig. 54
-------
ii
oo
I
ca-
ui
I
I
\
\
\
10
PCT-
BOSTON ED.OPERflTION
FREQUENCY DISTRIBUTION
- RECYCLE MGS04
Fig. 55
-------
I-.
00
kO
I
Q_
C3
' i CO
—i—
12
fen
10
lG
PCT.
BOSTON ED.OPERRtlON
FRtQUENCY DI3TRIBUT 1 ON
f
\ I ,- i !
Fig. 56
-------
en
en.
CD
C3.
OO
Q-.
i
U_i
O ED
5.50
6-00
e. so
7.OQ
7.50
TH
s.oo
S.50
-OO
3-50
to.oo
BOSTON ED-OPERRTION
FREQUENCY DISTRIBUTION
- RECY. PH
Fig. 57
-------
CD
CD
CD
CD.
00
CD.
I U_CO
UD °
I
.CD
CD.
CD
CM
600
80O
LOOO
1200
DEG.
1400
ISOO
1800
2000
2200
F
CflLClNER OPERflTION
FREQUENCY DISTRIBUTION
- MIDKILN TEMP.
Fig. 58
-------
10
N)
I
80 90
LB./MIN.
100
110
120
130
CRLCINER OPERRTI0N -
FREQUENCY DISTRIBUTION
MG503 FEED RflTE
Fig. 59
-------
»u
SO 90
PCT. MG-Q
LOO
130
CRLCINER OPERRTION
FREQUENCY DISTRIBUTION
PRODUCT flNflLYSI
Fig. 60
-------
40 50
PCT. MGS04
60
70
80
90
CRLCINER 0PERRTION - PRODUCT
FREQUENCY DISTRIBUTION
RNRLYSIS
Fig. 61
-------
Ul
I
X
\
\
yo so
B./CU-P"T.
60
,x
f -
70
80
90
CflLCINER OPERflTION
FREQUENCY DISTRIBUTION
- PRODUCT BULK DENS
Fig. -62
-------
o
10
20
PCT.
CPLCINER QPERflTIQN
FREQUENCY DISTRIBUT ION
flCID
FEED GPS 02
Fig. 63.
-------
9.3
MISCELLANEOUS DATA
-197-
-------
TABLE 31
MYSTIC STATION
UNIT #6
SULFUR - #6 FUEL OIL
7. S % S
April 72 1.72 Jan. 73 2.04
2.08
May 72 1.86 V2.11
1.86 2.05
1.93
1.88 Feb. 73 2.08
2.07
June 72 1.93 2.17
1.91 2.17
1.91
2.05 March 73 2.11
2.13
July 72 2.05 2.11
2.03 2.12
2.08
2.05 April 73 2.08
2.10
Aug. 72 2.00 2.05
2.14
2^09 May 73 2.10
2.09 2.06
1.99 2.09
Sept. 72 2.03 June 73 2.10
2.10 2.06
2.04 2.05
2.01 1.73
Oct. 72 2.01 July 73 2.04
2.01 2.08
1.81 1.95
1.97 2.13
Nov. 72 2.00 Aug. 73 2.00
2.04 2.08
1.96 2.10
2.04 2.11
2.02 2.08
Dec. 72 1.96 Sept. 73 2.03
2.01 2.05
2.10 2.07
2.09 2.02
.. -198-
-------
TABLE 31 (CONT'D)
STATION #200 - UNIT .#6
Oct. 73
Nov. 73
Dec. 73
Jan. 74
Feb. 74
March 74
April 74
May 74
June 74
2.10
2.07
1.95
2.15
2.25
2.13
2.19
2.11
2.13
2.16
2.25
2.19
2.18
2.23
2.13
2.15
2.03
2.03
1.92
2.02
2.02
2.11
2.02
2.05
1.97
1.88
1.90
96
86
60
81
98
85
65
1.67
1.60
1.58
1.88
1.91
-199-.
-------
TABLE 32
HIGH SULFUR OIL ASH ANALYSIS
METHOD OF TEST: Emission Spectroscopy
SAMPLE IDENTIFICATION: Sample #1235
Ash, % 0.046
Semi-Quantitative Spectrographic Analysis of Ash:
Vanadium A
Nickel 3B
Sodium B
Tin 3C
Silicon 3C
Calcium C
Iron C
Magnesium C
Aluminum 3D
Boron 3D
Barium 3D
Cobalt 3D
Cadmium 3D
Copper D
Chromium D
Molybdenum D
Lead D
Strontium D
Titanium D
Manganese 3E
Silver E
Key: A = Greater Than 10% E = .001 - .01%
B = 1 - 10% F = .0001 - .001%
C = . 1 - 1. 0% G = Less Than . 0001%
D = . 01 - . 1% 3B = Three Times Letter Value
-200-
-------
TABLE 33
REGENERATED MgO VANADIUM AND NICKEL CONTENT
Date % Vanadium % Nickel
6-12-73
7-3-73
7-19-73
7-27-73
9-12-73
11-5-73
11-6-73
12-4-73
12-5-73
0.43
0.55
0.69
1.55
0.45
1.00
0.80
0.83
0.83
0.085
0.100
0.093
0.374
0.136
0.134
0.170
0.08
0.08
-201-
-------
10.0 FINANCIAL DATA
Project costs are presented in this section for the plants
construction and for their operation, this data is summarized in Table
34. A description of some guide lines and recommendations for the
operating requirements of future plants is also presented and used in a
comparison with cost data developed in a separate study to provide a
basis for economic evaluation of the process.
-202-
-------
10.1 Systems Cost
Chemico - Basic Magnesia
F. G.D. System at Mystic Station
Regeneration Plant and Acid
Plant Modification
Tail Gas Scrubber
Total System Cost
10. 2 ' System Operation
l
NJ
O
U)
F. G.D. System Operation &
Maintenance
Regeneration Plant Operation
and Maintenance
Project Management
TOTAL OPERATIONS COST
TABLE 34
SUMMARY OF FI NANC1AL DATA
S3,635,000
2,294,220
238,110
$6,167,330
Pre Start Up
Through
Break In Period
$ 573,591
747,717
488,872
$1, 810, 180
pi "
PERIOD
--, eel Operational Continuous
Testing Operations
$ 619,418
578,716
260,800
$1,458,934
$326,116
391,564
94,500
$812,260
Project
Termination
Total
$ 89,843 $1,608,968
1, 717, 997
34,582 878,834
$124,425 $4,205,799
-------
10.1 SYSTEMS COST
The costs of the prototype plants are presented for the
F. G. D. System in Table 35 and for the regeneration system plus
the cost of modifications to the acid plant in Table 36.
These are broken down to show total costs by specific
catagories identified by Chemico standard codes. (The equipment
included in each of these sections has been listed in Tables 11 and
12 of Section 5).
10.1.1 System Cost Details
Table 35
MAGNESIA F. G. D. SYSTEM
Chemico
Cost Code
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
Description
Foundations
Buildings
Structural Steel
Ductwork
Tanks
Storage Bins
Pumps
Fans
Machinery
Material Handling Equipment
Special Equipment
Insulation
Piping
Instruments
Electrical
Painting
Sitework
Sub Total
Cost *
$ 229,456
17,351
303,338
522,751
29,798
22,138
86,938
151,973
1,826
261,053
462,102
52,578
271,960
166,838
153,581
28,942
205,646
$2,979,415
Site clearance and preparation)
Initial MgO charge, etc. ) 655,585
Total System Cost $3, 635, 000
* Includes Engineering Cost
-204-
-------
TABLE 36
10.1.2 System Cost Details
Regeneration Plant and Acid Plant
Modifications
Chemico
Cost Code
A
B
C
D
E
F
G
I
J
K
0
P
R
S
T
U
V
W
X
Y
Z
Description
Foundations $
Buildings
Structural Steel
Ductwork
Heat Exchangers
Tower
Tanks
Storage Bins
Pumps
Fans
Material Handling Equipment
Dust Collection
Calciner
Insulation
Piping
Instruments
Electrical
Catalyst
Painting
Sitework
Leasing
Total $2
Cost *
176,942
57,342
166,342
24,774
40,803
6,611
16,888
52,487
13,579
38,096
86,743
9,970
584,811
44,447
424,791
72,994
179,433
1,926
23,018
55,357
216,866
,294,220
* Includes Engineering Costs
- 205 -
-------
10.2 OPERATING COSTS
Operating costs for the F. G.D. installation at Mystic
Station are shown in Table 37. These costs are separated to show ex-
penses for operating labor, utilities and maintenance as well as costs
that can be identified as specific to the test program. A further
separation is provided to show the costs incurred during the three
principal periods of the test program.
Similar information is shown in Table 36 for the regenera-
tion section. Included in this table is the expense for maintaining the
laboratory and testing facilities at both sites and field administration
costs. Because modifications and maintenance of the calciner and
material handling equipment constitutes the largest percentage of the
maintenance costs these areas have been identified separately in this
section.
Table 37 presents the project management cost details as
a separate item. Almost all of these costs are specific to operation
of a prototype system, administration of the test and development
effort, data handling and report preparation.
10. 3 PEOCESS ECONOMICS
The program described in this report combined elements
of process development and pilot plant work with normal start-up
procedures for the prototype system. Similarly, maintenance work
on the unit was often in the form of modifications to pieces of its
equipment. Because of these aspects, it is not possible to determine
operating costs directly from the data. Instead, the information
contained in this section has been combined with the observations and
findings of the program to provide guidelines for establishing speci-
fications for future plants and required quantities of materials for
plant operations.
A source of data on process economics for the MgO slurry
F. G. D. System is the Environmental Protection Agency Report,
"Conceptual Design and Cost Study-Sulfur Oxide Removal from Power
Plant Stack Gas-Magnesia Scrubbing-Regeneration" (10) which pre-
sents economics for a variety of plant sizes and configurations.
-206-
-------
10. 2. 1 Operating & Maintenance Cost Details
F. G.D. System - Boston Edison Co.
TABLE 37
NJ
O
Operations
Labor* (Incl. Supv. )
Electricity
Fuel Oil (for dryer)
Water
Steam
Maintenance
Labor*
Supplies*
Spare Parts*
Subtotal O&M
Test Program Costs
Training Program*
Cleanup*
Subtotal Test Program
Termination & Mothballing Costs*
TOTAL F. G. D. SYSTEM COSTS
Pre-Startup
Thru Break-In
Period
$232, 324
26,421
13, 806
3,476
6, 858
134,090
67, 842
- 0 -
484,817
6,401
82, 373
88,774
$573,591
Planned
Operational
Testing
$212, 169
33, 195
15, 980
3, 958
2, 599
180, 955
55, 213
50, 947
555,016
64, 402
64,402
$619,418
Continuous Project
Operations Termination
SI 15, 408
45,743
25,024
4,740
2, 756
59,881
52,742
- 0 -
306, 294
19,822
19,822
$89,843
$326,116 $89,843
Total
$ 559,901
105,359
54,810
12,174
12, 213
374, 926
175,797
' 50, 947
1, 346,127
6,401
166,597
172, 998
89,843
$1, 608, 968
*Incl. 50% General Overhead or Mark-Up
-------
10. 2. 2 Regeneration Plant Operations Cost Details
Including Maintenance & Modification Plus
Testing and Quality Control Program
TABLE 38
O
03
I
Administrative Costs
Field Office (1)
Field Transportation (Boston & Rumford)
Home Office
Salaries (Incl. Overhead)
Miscellaneous (Incl. Communication &
Reproduction )
Payroll Taxes & Insurance
Travel & Living Expense
Subtotal Administration
Operations Costs
Essex Chemical Leasing Fees
Operating Costs
Operating Labor (Incl. Overhead)
Downtime
Pre-Startup Planned
Thru Break- In Operational
Period Period
$ 52,
5,
41,
2,
4,
3,
$110,
262
400
649
712
970
928
924
head for Field
152
98
,372(2)
,297
468
$33,
3,
33,
6,
3,
3,
$83,
466
200
893
016
890
435
900
Continuous
Operations
$16,512
1,000
13,
2,
1,
2,
$37,
019
756
992
486
761
Total
$102,
9,
88,
11,
240
600
561
480
10,852
9,
$232,
849
585
Accountant
72
60
,398
, 300
383
55
37
,600
,704
144
280
196
,370
,301
995
-------
TABLE 38
(Continued)
10. 2. 2
(Cont'd)
l
10
o
kD
I
Utility Costs
Fuel Oil
Cooling Water
Electricity
(3)
Operating Chemicals
MgO (makeup)
Coke
Other Chemicals
• Transportation (MgO &
Subtotal Operations
Pre-Startup
Thru Break-In
Period"
22, 081
2, 538
21, 508
47,629
5,430
1,474
13, 807
$365, 604
Planned
Operational
Period
23,914
2, 983
15, 386
64,152
5, 827
904
25,964
$272,211
Continuous
Operations
30,814
2,046
10,951
41,073
2, 192
3, 294(4)
23,017
$206,835
Total
76,809
7,568
47, 845
152,854
13,449
5, 672
62,788
$844,651
(2)
(3)
(4)
Includes leasing costs from 10/71
Includes freight
Includes $2, 928 caustic for tail gas scrubber system
Maintenance Costs
Spare Parts
Plant Maintenance Labor
Essex Supplied Material
Outside Maintenance' '
Routine Maintenance & Supplies
Code O Bins/Mat'l Handling/Pulverizer
Code R Calciner Modifications
21,354
21,354
30, 631
17, 133
21,557
13,259
13,271
9,238
2,483
23,066
17,676
11, 292
3, 654
3,731
8,061
13,369
6,404
43, 523
23, 347
52, 684
44, 304
30, 967
-------
10.2.2
NJ
!-•
O
TABLE 38
(Continued)
(Cont'd)
Pre-Startup Planned
Thru Break-In Operational
Period Period
Miscellaneous Tools & Supplies 8, 504
Subtotal Maintenance $1 25, 709
(5)
Includes Subcontract Labor
Quality Control & Testing Costs
Salaries (Incl. Supervision & Overhead)
F. G. D. Plant at Mystic Station 983
Regeneration Plant 131,807
Material and Supplies 11,660
Laboratory Trailer (F. G. D. Plant) 1,030
Special Test Work
Stack Analyses (F. G. D. Plant)
Spectrometric Characterization
Metallurgical Testing
Subtotal $145,480
TOTAL REGENERATION PLANT COSTS $747, 717
6, 913
$ 70,668
& Shipping Costs
55, 218
69, 799
5, 531
1., 030
20,359
$151, 937
$578,716
Continuous
Operations
4, 631
$ 39,850
41, 298
39,709
10, 974
734
11,682
2,720
$107, 117
$391,564
Total
20,
$ 236,
-
97,
241,
28,
2,
20,
11,
2,
$ 404
$1,717
048
227
499
315
315
794
359
682
720
, 534
, 997
-------
TABLE 39
10. 2. 3
Project Management Cost Details
i
ro
Direct Salaries ,, >
Field Costs, (Incl. Operating Supvr. )
Home Office, (Incl. Engineering)^ )
Outside Engineering & Inspection
Miscellaneous (Incl. communication &
Reproduction )
Payroll Taxes & Insurance
Travel & Living Expense
TOTAL PROJECT MANAGEMENT
PrerStartup
Thru Break-In
Period
$106, 212
275,732
33,360
13,239
10,297
50,032
$488,872
Planned
Operational
Testing
$ 84, 265
117,190
11, 156
15, 121
5,436
27,632
$260,800
Continuous Project
Operation Termination
$43,520 $ -0-
25,978 34,582
799
9,895
2,925
11,463
$94,580 $34,582
Total
$233, 997
453,482
45,315
38,255
18, 658
89,127
$878,834
(1) Includes Overhead
-------
In order to apply the findings of this study to improving
economic predictions for the magnesia slurry F. G. D. process, a
comparison of costs is made for one case presented in the "Conceptual
Design and Cost Study" using costs developed from the guidelines re-
sulting from this work.
The case chosen is a Magnesia Slurry Absorption-Regenera-
tion System for a 200 MW, coal fired unit, 3. 5% S in fuel and a 150
TPD sulfuric acid plant. (11) An advantage of this case is that it in-
tegrates all of the facilities from the boiler to the acid plant at a sin-
gle site which eliminates the uncertainty associated with estimating
transportation and storage costs.
Following are some of the significant aspects used in de-
veloping the manufacturing costs:
A. Delivered Raw Material - Measurement of system
losses during the controlled run indicate an MgO makeup rate
equal to 5% of the feed is reasonable.
B. Conversion Costs - Labor- Successful operation of
the F. G. D. System was attained with the assignment of one control
operator and one auxiliary operator per shift. It should be noted
that this staff could handle two trains of equipment.
Regeneration plant and acid plant operation is accomplished with
two operators plus one roving operator per shift.
Utilities - Fuel Oil - The higher fuel consumption of the
dryer in the study was related to the reduced efficiency caused by
the modification noted. Future installations would employ properly
designed dryers, therefore fuel consumption can oe estimated by
conventional methods. No provision is made for reheat as no
separate reheat was used in the prototype operation.
Steam - MgO slaking at 180 F is provided by direct
steam sparging.
-212-
-------
C. Maintenance - Despite the high costs normally associated
with maintenance and modifications of a prototype system, actual
maintenance costs incurred were only 5|% of erected value for the
F. G. D. unit and 4-3/4% for the regeneration plant. Using this basis,
a 5% maintenance charge appears conservative.
D. Analysis - Reliance on wet chemical methods for criti-
cal control characteristics of the F. G. D. and regeneration system
plus the normal requirements of the acid plant mandate close analyti-
cal support. One chemist per shift plus supervision and supplies are
included.
Table 40 summarizes the fixed investment costs using the
format developed in the E. P. A. study. In both cases the design
parameters are as follows:
200 MW coal fired unit
3|% sulfur in the coal
SO2 removal 90%
Plant on stream time 80%
Coal consumption 554, 200 T/Yr.
Heat rate 9, 500 BTU/kwh
The E. P. A. study noted the inclusion of stack reheat by steam
to 175 F though no costs were included for operation. No reheat is
included in our comparison as the prototype plant did not employ reheat.
Table 41 summarizes the manufacturing costs, again using
the E. P. A. format. It should be noted that only the direct costs are
listed, no allowance is made for revenue from the sale of sulfuric
acid or for depreciation and other indirect costs as these will vary
from the viewpoint of either the utility or chemical manufacturer.
Other arrangements for processing the products from this
F. G. D. system have been proposed. In one of these a tariff is paid
by the power plant to an acid manufacturer to regenerate the absorbent.
A study* of the plan, based on a $15/ton differential cost between high
and low sulfur coal, has shown a saving to the utility of 1. 6 mills/kwh
while paying a processing fee sufficient to provide a 25% before-tax
return on investment to the chemical company. Similar economics
obtain for a fuel oil fired power plant at a $3/bbl differential between
high and low sulfur fuel.
* Additional information is presented in the E. P. A. Capsule
Report on Magnesia Scrubbing (EPA 625/2-75-007).
-213-
-------
TABLE 40
SUMMARY OF ESTIMATED FIXED INVESTMENT
MAGNESIA SLURRY SCRUBBING-REGENERATION PROCESS^
(200-MW New Coal-Fired Power Unit, 3. 5%S in Fuef;T50~TPD~H~2SO4)
From EPA Cost Study From This Study
Investment, $ ^' Investment, $'2'
Land, site clearance, excavation, landscaping, roads, railways, walkways 200, 000 -0-
Particulate scrubbers (2 scrubbers with surge tanks, agitators, pumps, and fly
ash neutralization and disposal facilities) 1, 445, 000 3, 790, 000
Sulfur dioxide scrubbers.(2 scrubbers with surge tanks, agitators, pumps,
mist eliminators and fans) 1, 602, 000 4, 360, 000
Optional bypass duct around scrubbers • 209,000 -0-
Slurry processing (screens, tanks, pumps, agitators and heating coils,
purification facilities, centrifuges, and conveyors) 416,000 (2,510,000
Drying (drying system, dust collectors, conveyors and MgSOs storage hopper) 470, 000 (
Calcining (calcining system, fans, MgO and coke storage hoppers,
feeders, conveyors, elevators, waste heat boiler, dust collectors) 635,000 1,830,000
Magnesium oxide slurrying (MgO unloading and storage facilities, feeders,
conveyors, elevators, slurry tank, agitator, and pumps) 140,000 410,000
Sulfuric acid plant (complete contact unit for sulfuric acid production,
dry gas purification system) 1, 495,000 4, 500, 000'4^
Sulfuric acid storage (storage and shipping facilities for 30 days
production of H2SO4) 108,000 -0-
Fuel oil storage (fuel oil storage and distribution system including storage
tank, hold tanks, heat exchanger, transfer and feed pumps) 94, 000 -0-
Control room building, including motor controls, laboratory, and lockers 150,000 Included Above
Service facilities and buildings allocation for maintenance, shops, and offices 410, 000 -0-
Subtotal direct investment 7, 374, 000 17,400, 000
Engineering design and supervision 664,000 (
Construction expense 811,000 (Included Above
Contractor fees 442,000 (
Contingency 959, 000 (
Subtotal fixed capital investment 10, 250, 000 17,400, 000
Allowance for startup and modifications 1, 025,000 Included Above
Interest during construction (8%/annum rate) 410, OOP -0-
Total fixed capital investment 11, 685, 000 17,400, 000
(1) 1972 Costs
(2) 1975 Costs
(3) "Dry" Sulfuric Acid Plant & Tail-Gas Scrubber
(4) Double Absorption Sulfuric Acid Plant
-------
TABLE 41
ANNUAL MANUFACTURING COSTS FOR 98% H2SO4 FROM SCRUBBED POWER PLANT STACK GAS
MAGNESIA SLURRY SCRUBBING-REGENERATION
"T200-MW Existing Coal-Fired"~Power Unit, 3. 5%TrTn Fuel)
From EPA Cost Study
(10)
Annual
Quantity
Delivered raw material
Lime
Magnesium oxide (98%)
Coke
Catalyst
Sub-Total raw material
Conversion Costs
Operating labor and supervision
Utilities
Fuel Oil
Steam
Heat Credit
Process Water
Electricity
Maintenance
Labor and Material, . 07 x $13, 083, 000
Analyses
Sub-Total conversion costs
Total direct costs
(1)1972 Costs
(2)1975 Costs
56. 6 tons
463 tons
322 tons
760 liters
30, 440 man-hr
3, 166, 000 gal
-- M Ib
8, 600 MM Btu
931 MM gal
28. 2 MM kwh
Unit Cost
16. 00/ton
102. 40/ton
23.50/ton
1. 51/liter
6. 00/mh
0. 09/gal
0. 60/M Ib
0. 40/MM Btu
0. 05/M gal
0. 007/kwh
5. 5%S in Fuel)
Total*1 >
Annual
Cost
$
900
47,400
7,600
1, 100
~57,000
182, 600
284, 900
(3,400)
46,600
197,300
915,800
45,000
1, 668, 800
1, 725,800
From
Annual Quantity .
-0-
875 tons
490 tons
2, 700 liters
50, 000 man-hr.
2, 890, 000 gal.
4MM Ib
-0-
400 MM gal
34 MM kwh
. 05x$17, 400, 000
This Study .
Unit Cost(c)
$
150/ton
80/ton
1. 25 'liter
8. 00/mh
0. 298/gal
2/M Ib
0. 25/M gal
0. 01 /kwh
Total(2)
Annual
Cost
S
-0-
131, 250
39, 200
3, 375
173.T25
400, 000
861. 220
8,000
-0-
100, 000
340, 000
870, 000
100, 000
2, 679, 220
2, 853, 045
Note: Direct cost 2 mills/kwh (80% rating^
8
-------
H.O REFERENCES
1) Kerr, C. P. Sulfur Dioxide Removal in Venturi Scrubbers. I & EC,
Process Design 1_3, No. 3, 222 (1974). :
2) Smithson, G. L. and Bakhshi. Kinetics and Mechanism of the Hydration
of Magnesium Oxide in a Batch Reactor. Canadian Journal of Chemical
Engineering, 47, 10,508 (1969).
3) Evans, R.L. and St. Clair, H.W. Carbonation of Aqueous Suspensions
Containing Magnesium Oxides and Hydroxides. I & EC, 41, 2814 (1969).
4) Yost, D. M. and Russel, H. Systematic Inorganic Chemistry. New York:
Prentice Hall, Inc. (1946^
5) Epstein, M. E.P. A. Alkali Scrubbing Test Facility: Sodium Carbonate
and Limestone Test Results. .EPA - 650/2-73-013, (August 1973).
6) Kovache, T., Bakalov, V. and Trendafelov, D. Attempts to Obtain Lower
Hydrates of MgSOq and the Anhydrous Salt. Khimiya i Industriya
Sofia, 4_2, 5, 209 (1970).
7) Hagesawa, H. On Magnesium Sulfite. Bull. Inst. Phy. Chem. Research
(Tokyo) 12, 976 (1933).
8) Sass, Allan. Simulation of the Heat Transfer Phenomena in a Rotary Kiln.
I & EC Process Design and Development 6, No. 4, 532 (1967).
9) Brink, J. A. Cascade Impactor for Adeabatic Measurements. I & EC,
5£, No. 4, 645 (1958).
10) McGlamery, G. G. et al, Conceptual Design and Cost Study - Sulfur
Oxide Removal from Power Plant Stack Gas - Magnesia Scrubbing.
Regeneration. EPA-R2-73-244 (May 1973)
11) Op cit, pg 206.
-216-
-------
12.0 LIST OF PUBLICATIONS
1) Sulfuric Acid from the Stack. Chemical Week, 197 (3) (1970).
2) Shah, I. S., Wechselblatt, .P.M., Radway, J.E. SOp Recovery from
Smelters with Magnesium Base SO? Recovery Process. AIME
Environmental Quality Conference, Washington, D.C. (June 1971)
3) Shah, I. S. Removing SO9 and Acid Mist with Venturi Scrubbers. CEP,
(May 1971) Vol. 67, No. 5.
4) Kleiman, G. and Willett, H. Relative Economics of Stack Gas Scrubbing
US Residual Oil Desulfurization. API Session on Desulfurization and
SO2 Recovery San Francisco, Calif., (May 12, 1971).
5) Shah, I. S., Quigley, C.P. Magnesium Base SO? Recovery Process, A
Prototype Installation. 70th AIChE National Meeting, (August 1971).
6) Wechselblatt, P.M., Quig, R.H. Magnesium Base SO? Recovery Scrubbing
Systems. 71st AIChE National Meeting, (February 1972).
7) Maxwell, M. A., Koehler, G. R. The Magnesia Slurry SO? Recovery
Process with a Large Prototype System. 65th AIChE Annual Meeting,
(November 1972)
8) Quigley, C.P. Progress Report - Magnesium Oxide System at Boston
Edison Company's Mystic Station. Electrical Worlds Technical
Conference, Chicago, (October 1972).
9) Houston, P. and Koehler, G. Application of Magnesia SOo Control System
to a 150 MW Power Plant. International Conference on SO2 Control,
Manchester, England (April 1973).
10) Koehler, G. R. (Part I) and Quigley, C.P. (Part II) Operational Performance
of the Chemico Basic Magnesium Oxide System at the Boston Edison
Company Flue Gas Desulfurization Symposium. New Orleans,
(May 1973). (EPA-650/2-73-038, December 1973).
11) Radway, J.E. and Rohrbach, R.R. Progress Report on the Chemico-Basic
Magnesium Oxide Flue Gas Desulfurization System at the Boston Edison
Company. Presented at the 30th Annual Meeting, East Central Section
APCA, Cleveland, Ohio (September 26, 1973).
-217-
-------
LIST OF PUBLICATIONS (CONT'D)
12) Maxwell, M.A. Application of the Magnesia Slurry SO? Recovery Process
to Stack Gas Desulfurization. Paper presented at 25th Southeastern
Regional Meeting, American Chemical Society, Charleston, S. C.
(November 8, 1973).
13) Koehler, G. R. New England SO2 Recovery Project - System Performance.
66th AIChE Annual Meeting, Philadelphia (November 1973).
14) Koehler, G. R. Alkaline Scrubbing Removes Sulfur Dioxide. Chemical
Engineering Progress, 70, No. 6, 74.
15) Koehler, G.R. and Dober, E. Magnesia SO2 Absorption, Process Develop-
ment . Flue Gas Desulfurization Symposium, Atlanta, Ga.
(November 1974) (EPA-650/2-74-126).
16) Quigley, C.P. and Burns, J. A. Assessment of Prototype Operation and
Future Expansion Study - Magnesia Scrubbing Mystic Generating Station
Boston, Massachusetts (Ibid).
17) Erdman, D.A. Mag-Ox Scrubbing Experience at the Coal-Fired Dickerson
Station. Ptomac Electric Power Company. Washington, D.C. (Ibid).
18) Zonis, I. S., Olmsted, F. , Hoist, K. A. and Cunningham, D.M.
The Production and Marketing of Sulfuric Acid From the Magnesium
Oxide Flue Gas Desulfurization Process.(Ibid).
19) Koehler, G. Report of Operation of a Magnesia FGD System on and Oil
Fired Boiler. AIChE Symposium Series, No. 148, (July 1975).
-218-
-------
13. CONVERSION FROM ENGLISH TO METRIC UNITS
To Convert from
atmosphere (normal)
atmosphere (normal)
barrel (42 US gallons)
British thermal unit (Btu)
Btu/hour
Btu/pound mass
Btu/pound mass - °F
foot
foot2
foot3
foot /minute
foot-pound force
gallon (US)
gallon (US)/minute
grain
horsepower
inch
inch H2O (60°F)
pound force
pound mass av
pound mass av
pound force/inch
pound mass/foot^
°Rankine
ton mass (US short)
ton mass (US long)
To
.bar
pascal
meter^
joule
watt
joules/gram
joules/gram - °K
meter
O
meter
meter
meter ^/minute
joule
meter^
meter'3/hour
milligram
kilowatt
centimeter
kilopascal
newton
kilogram
metric ton (tonne)
kilopascal
kilograms /meter
°Kelvin
kilogram
kilogram
Multiply by
1.01325
101,325
0.15899
1055.1
0.29307
2.32600
4.18680
0.30480
0.09290
0.02832
0.02832
1.35582
0.00379
0.22712
64.7989 '
0.74570
2.5400
0.24884
4.44822
0.45359
0.0004536
6.89476
16.0185
0.55556
907.185
1016.05
-219-
-------
APPENDIX "A"
MAGNESIUM SULFITE HYDRATES
TRANSITION & FORMATION
Process considerations are given to the existence of
the two hydrates of MgS03 which have differing physico-
chemical properties as the absorption processes are normally
operated near the transition temperature region for these
hydrates.
MgS03*3H?0 - MqS03'6H20 Transition
It has been observed (6) that the transition of
MgS03-6H20 to MgS03-3H2 at 40°C (104°F) is negligible. A
measurable, although still slow transition can be observed
at 65° (149°F) and at 70°C (158°F) the transition is completed
in 10 hours.
Further studies (7) indicated that at 15°C (59°F) and
at 35°C (95 °F) no transformation of MgS03«6H2O to MgS03.3H20
occurred. Also in this work, the concentration of MgS04 in
solution was varied from 0 to 30%. With high MgS04 concentra-
tion in the solution the sulfite crystals were highly contami-
nated (to the extend of 40%) MgS04- At 55°C (141°F) and at
75°C (167°F) rapid transformation of the hexahydrate to tri-
hydrate occurred. These authors were of the opinion that
MgS04 solutions aid in the transformation of hexahydrate to
trihydrate.
Since the temperature of the absorber at Boston is 135°F
and no such rapid transformation of the hydrates was observed,
a study of the transition of the two hydrates was undertaken
using the following procedure :
220
-------
Slurries were prepared with a composition similar to
that obtained in normal operation and were placed in glass-
stoppered bottles. The bottles were placed in a constant
temperature bath. After a predetermined period of time
the bottles were taken out and the contents filtered
immediately, washed with methanol and dried in air and at
40°C. The crystals were analyzed for sulfite, sulfate,
and MgO content.
The ratio of the two hydrates in the crystals was
analyzed using a chemical method developed at Chemico's
laboratory. This method is discussed in Appendix C,
Analytical Techniques.
RESULTS
At room temperature (in contact with water) the tri-
hydrate is converted to the hexahydrate. This transforma-
tion also takes place in 8% MgSO^ solution. But when the
concentration of MgSC>4 in solution reaches 15% or above,
no further transition to hexahydrate occurs. In this
transition the crystal size of the resultant hexahydrate
becomes 8 to 10 fold larger than the original trihydrate
crystals tending to support the hypothesis that the crystals
first dissolve in the solvent and then crystallize out in
the other form. In contrast to the trihydrate, the hexa-
hydrate appeared to change over to the trihydrate with the
rise in concentration of MgS04 in solution. However, con-
flicting results were obtained in this study and, consequent-
ly, no firm conclusions could be drawn. Table A-l summarizes
the room temperature data.
221
-------
In our study, no transition of the hexahydrate to
trihydrate was observed at 135°F (57°C), even after 48
hours, nor was the catalytic effect of MgS04 claimed by
Kbvachev and others observed. The transition temperature
of MgS03-6H20 to MgS03«3H20 appears to be a few degrees
above 135°F and most probably at 138°F. The pH of the
medium seems to have some effect on the rate of transition
of the hydrates. It appears (Table A-2) that alkaline pH
retards while acid pH favors the rate of transition. In
summary, the major factor in the transition of the hexa-
hydrate to trihydrate appears to be temperature. The
transition temperature is 138°F (59°c). MgSC>4 does not
appear to play a significant role in the transition. On
the other hand, pH and presence of trihydrate do have some
effect on the rate and temperature of transition.
222
-------
Table A-l
MAGNESIUM SULFITE
HYDRATE TRANSITION AT ROOM TEMPERATURE
Conditions for Digestion Trihydrate
1. in 8% MgSC>4 solution All
for 24 hrs. Hexahydrate
Hexahydrate
No change
2. In 15% MgSO4
a. 20 Hrs.
b. 64 Hrs.
c. 97 Hrs.
No Change
No Change
Mostly hexa with
some trihydrate
40% 6H20
60% 3H20
No change
3. In 20% MgS04
a. 20 Hrs.
b. 64 Hrs.
c. 72 Hrs.
d. 97 Hrs.
No Change
No Change
80% 6H20
20% 3H20
40% 6H20
60% 3H20
No change
No change
223
-------
TABLE A-2
TRANSITION OF MgS03.6H20
pH ADJUSTED BY OIL. H2S04
Conditions for Digestion
1. pH 7.00: 138°F.
45 mins. in water
2. pH 7.00: 135°F.
45 mins. in water
3. pH 7.10: 140-146°F.
45 mins. in water
4. pH 7.00: 134-135°F.
45 mins.; 15% MgS04
5. pH 7.00: 132-134°F.
45 mins. in water
6. pH 7.15: 134-136°F.
45 mins. in water
7. pH 7.40: 138-141°F.
45 mins. in water
Results
All trihydrate
Hexahydrate with few
percent Trihydrate
All trihydrate
All hexahydrate
All hexahydrate
All trihydrate
A little trihydrate
10%: 90% Hexahydrate
224
-------
Magnesium Sulfite Hydrate Formation. .
Magnesium sulfite trihydrate (MgS03.3H20) is formed
in the prototype FGD system after equilibrium is attained.
Laboratory studies were undertaken in an effort to determine
the conditions which led to the formation of each of the
hydrates and the influence of the systems components on
crystal type. These experiments including the effort to
simulate the reaction occuring in the venturi absorber, were
unsuccessful in duplicating the hydrate formed in the pro-
totype system.
Two experimental conditions were used: a) SQ2 at
room temperature was reacted with MgO slurry in (i) water
and in (ii) MgSQ4 solution at 135°F and at a pH of 6.8 to
7.20; b) SC>2 at 300°F was reacted with MgO slurry in water
at 135°F; and a pH of 6.80 to 7.20. The reaction period
in both cases was of 75 minutes and 862 gas (pure) at .a
rate of 200 ml./min. and air at a rate of 3 liters/minute
were added through a glass T-tube connected to a Vycortube
wrapped with heating tape.
The gas was bubbled through 400 ml. of water containing
1-2 gm. of MgO maintained at a temperature of 135°F in a 600
ml. beaker. The temperatue of the reaction-beaker was con-
trolled at 135°F by a water-bath.
The pH of the reaction medium was kept at 6.8-7.2 by
controlled addition of MgO slurry (10% in water). The whole
reaction lasted 75 minutes.
After reaction the slurry was filtered and the coke
and filter analyzed. In all cases the exclusive product
was MgS03.6H20 as shown in Table A-3.
225
-------
TABLE A-3
REACTION OF SOo with .' MgO
Slurry
Slurry
Temp. (°F) pH
MgS04(%)
S02
Reaction Products
Temp. (°F) Filtrate
Cake
%MgSO4 MgSO3 Hydrate
130-136
130-136
135
135
135
6.8
6.8
6.8
6.8
6.8
-7.2
-7.2
-7.2
-7.2
- 7.2
0
15%
0
0.5
1.4
Amb.
Amb.
300°
300°
300°
0.43
12.1
0.5
1.4
1.3
Hexahydrate
Hexahydrate
Hexahydrate
Hexahydrate
Hexahydrate
Composition %
MgSO3
48.6
46.4
47.9
48.8
48.3
MgSQ4
0.2
1.9
0.3
1.0
0.6
MgO
0.6
1.0
0.7
0.7
1.2
to
-------
APPENDIX B
PHYSICO-CHEMICAL DATA
It. is well known that magnesium sulfite exists in two forms: MgSOg
and MgSO3. 3H2O. A third form (MgSO3. H2O) has been reported but no
evidence of its existence has been found in this study.
X-RAY CRYSTALLOGRAPHIC DATA
X-ray crystallographic study of samples submitted for analyses showed
the hexahydrate is rhombohedral with space group R3, with three molecules
per unit cell with the following dimensions:
a = 8.820 A
o
c = 9.052 A
The calculated density is 1. 724 gm. /ml. while the measured value is 1. 730
gm. /ml.
The trihydrate is orthorhombic with space group Pbn^. There are four mole-
o
cules per unit cell: a, 9.36; b, 9.45; and c, 5.51 A. The calculated and
measured densities are 2.138 gm. /ml. and 2.180, respectively.
The major X-ray spectra for the hydrates are given in Table B-l.
IR & RAMAN SPECTRAL DATA
-1
Laser Raman Spectrum. (Solid) in CM
MgSO3.6H2O MgSO3.3H2O
962 (s) ) 967.7 (s) triplet
) doublet
944 (s) ) 477 (m) with shoulder
482 (m)
227
-------
O -1
The .laser used was Argon excitation at 4880 A with scanning from 200 cm
_ i ' c - • • '
to 1400 cm . (Because of low solubility in water no sharp spectrum could
be obtained in aqueous medium).. . .
The infrared study of the two hydrates of MgSOg was conducted in: KBr
pellet, nujol mull, as well as MIR with KRS-5 plate. IR spectra of the two
hydrates vs. temperature were also studied and an attempt was .made to cor-
relate the high temperature IR spectra with a DTA study of the-two hydrates.
ROOM TEMPERATURE SPECTRA OF THE HYDRATES
-1
IN THE SOLID STATE (KBr PELLET) IN cm
Assignment
'- 1
i 2
'< 3
"' 4
' 5
' 6
Na9SO3
1010
633
941
496
MgSOQ.6H0O
950 (s)
640 (m)
925 (s)
490 (w)
MgSO3. 3H0O
950 (s)
900 (s)
640 (m, sh
400 (m)
450 (w)
850 (ms)
320 (ms)
700 (m) 700 (s)
260 (ms) 260 (w)
The 700 cm'1 band in MgSOg. 6H2O and in MgSOg. 3H2O is assigned to
coordinated water.
228
-------
From the IR spectrum it appears that the sulfite ion in the hexahydrate has
C3 symmetry while in the trihydrate this symmetry is lowered to Cg. The
correlation between the two are as follows:
V; N)
€3 - 1 ' 2 * 3 Y 4
i ..: v / o y/ ,\
Cs ' 1 ' 3 2 * 5 4 > X6
From the IR data, as also from chemical reactions and DTA work, two hy-
drates can be represented structurally as follows:
MgSO3.6H2Q Mg (H20)6 SO3
MgS03- 3H20 Mg (H20)3S03
The two hydrates can be easily distinguished from their IR s.pectrum by the
320 cm" , 450 cm"1, and 850 cm'1, bands. Also, in the hexahydrate the 700
and 640 cm bands appear as a doublet, but in the trihydrate only the 700
cm band is pronounced. The presence of more bands in the trihydrate spec-
trum than expected is probably due to the coordination and crystal field effects.
The previous representation of the two hydrates is also supported by their
reactions with AgNO3 solutions: (The difference in reaction of the two hy-
drates was discovered during this work and used as a method of analysis for
them)
2 Mg(H20)6 S03 + AgNOg •*- Mg^ g Ag(SO3)2 + 1/2 Mg(NO3) +12H2O
cf . 2Na2SQ3 + AgN03 *~Na3 Ag(SO3)2
-2
showing that the SO3 ion is present in the hexahydrate as a salt-like ion.
With the trihydrate the reaction is:
Mg(H20)3 S03 + 2AgN03 *-Ag2SO3 + Mg(NO3>2 + 3H2O
229
-------
— p
Because of the presence of SO inside the coordination sphere and the low
o
-2
ionisation constant of such complexes, the concentration of SOg ion in solu-
+
tion is too low to form complexes with Ag and only the insoluble AggSOo salt
is formed.
DTA of the hexahydrate shows two prominent endotherms - one at
130°C, and the other at 240 c, while the trihydrate gives only one
endotherm at 240 c.
The two hydrates were heated at 130°c and 160°C, for extended periods of
time (24 hours or more) in an oven. The results are as follows:
Temperature °c Tri- _ Hexa- _
130°C for 7 days Wt. loss 24% Wt. loss 44%
Composition Composition
MgSO3.H2o
160°C for 94 hrs. Wt. loss 29.5% Wt. loss 47%
Composition Composition
MgSO3.0.5H2O MgSO3.0.5H2O
IR spectrum of the two hydrates (Table B-2)were taken while being heated at
130°C and 160°c in a heated cell. After 24 hours of heating the two hydrates
gave essentially identical spectrum for the same heating temperature. The
160°C spectrum contained bands due to sulfate ion present in a complex en-
vironment. This suggests that at least up to 160°C the SOg ion in the tri-
hydrate is still present inside the coordination sphere. As for the hexahydrate,
O 9
at about 130 C the SOo ion enters the coordination sphere displacing water
and remains there at least up to 160°C, at which temperature partial oxidation
of sulfite ion takes place.
230
-------
PHYSICAL PROPERTIES
The specific gravities of the two hydrates were measured with a pycnometer,
in 2-methyl butanol, in which neither of the hydrates dissolve.
Hydrate Sp. Gravity
Magnesium Sulfite Hexahydrate 1. 730 gm. /ml.
Magnesium Sulfite Trihydrate 2. 18 gm. /ml.
The solubility of the two hydrates in water was also found to be different:
Hydrate %MgSOq In Water g. /liter
Magnesium Sulfite Hexahydrate 0.42 8.56
Magnesium Sulfite Trihydrate 0.76 11.55
Additional data is given in Table B-3.
The solubility of MgSC>3.6H2O increases with the rise in concentration of
MgSC>4 in solution as well as with temperature.
SOLUBILITY OF MgSO3. 6H2O in MgSO4 SOLUTION
Cone, of MgSO4 in MgSO3
Soln. (as %) (in gm. /liter)
5 7.5
10 8.5
15 9.0
20 9.20
25 8.90
NOTE: At 25% MgSO4 solution the solubility starts to decrease.
231
-------
SOLUBILITY OF THE HEXAHYDRATE IN WATER AND
IN 25% MgSQ4 SOLUTION vs. TEMPERATURE
% MgSOg in
Temperature °F Water 25% MgSO4
70 0. 75 0. 94
110 0.99 1.43
135 1.42 2.63
180 1.75 5.99
(Since at 180 F the hexahydrate changes over to trihydrate, the solubi-
lities at 180°F are probably those of the trihydrate).
SOLUBILITY OF THE TWO HYDRATES IN MgSO4 SOLUTION
in Solution
%MgSO4 Tri- Hexa-
15 1.22 1.02
25 1.14 0.94
232
-------
TABLE B-l
X-RAY CRYSTALLOGRAPHIC DATA
MgSO3.3H2O d() 3.35 6.71 4.27 6.71
I/I0 100 80 60 80
MgSO3.6H2O d 3.87 2.74 . 4.40 5.7
I/I0 100 100 80 4
233
-------
TABLE B-2
IR SPECTRA OF THE HYDRATES AT HIGH TEMPERATURE
Temperature (°c)
130°
160°
Ionic
-2
1104 (vs)
613 (s)
Band Intensity
s - strong
m - medium
w - weak
vw - very weak
Hexahydrate
990 )
) (s)
960 ) triplet
)
900 )
690 (m)
400 (w)
320 (s)
260 (w)
1050 (m)
900 (s)
700 (m)
670 (m)
390 (m)
Trihydrate ,
950 (s) (broad)
690 (m)
400 (w)
350 (sh)
320 (m)
260 (vw)
1050 (m)
900 (s)
700 (m)
640 (m)
390 (m)
300 (w)
Complex
1032 - 1044 (s)
1117 - 1143 (s)
970 (m)
645 (s)
604 (s)
438 (m)
234 -
-------
TABLE B-3
SOLUBILITY IN HgO
MgSO3.6H2Q MgSO3.3H2o
Temp. °C % MgSO3 Temp. °C % MgSO3
(320F) o 0.338 (100.4 °F) 38 1.034
(59°F) 15 0.497 (107.6°F) 42 0.937
(77°F) 25 0.646 (122 °F) 50 0.844
(95°F) 35 0.846 (131 °F) 55 0.817
(113°F) 45 1.116 (144 °F) 62.5 0.748
(131°F) 55 1.456 . (167 °F) 75 0.664
(144°F) 62.5 1.95 (185 °F) 85 0.623
(203 °F) 95 0.615
235
-------
APPENDIX C
ANALYTICAL TECHNIQUES FOR MAGNESIA MATERIAL
A. MgO is analyzed by acid-base titration.
1. Reagents
(a) Standard sulfuric acid 0. 5N. Reagent grade acid is diluted as: 1 ml.
of acid is added to 71 ml. of distilled water. Concentration checked
by titrating against standard alkali.
(b) Standard NaOH 0. 5N solution. Dissolve 20 gin. of reagent grade alkali
in a liter of distilled water. Concentration is checked by titrating against
standard potassium hydrogen phthalate.
(c) Potassium hydrogen phthalate (0. IN solution). Dissolve 20.4230 gm.
of reagent grade salt in 1 liter of distilled water.
(d) Indicator (mixed). Dissolve 1.25 gm. of methyl red and 0.825 gm. of
methylene blue in 1 liter of 95% alcohol. The indicator is pink in acid
and green in alkaline solution. It is colorless or gray in neutral media.
Use 10 drops for each titration.
2. Apparatus
(a) Two 50 ml. automatic burettes attached to plastic'bottles of 15 to 20
liters capacity.
(b) Beakers, 250 and 400 ml.
(c) Pipettes, 25 and 50 ml.
236
-------
(d) Teflon coated stirring bars.
(e) Analytical balance.
(f) Magnetic stirrer.
(g) Hot plate.
(h) Weighing paper.
3. Standardization of Acid and Alkali
Transfer 50 ml. of standard potassium hydrogen phthalate solution into a
250 ml. beaker. Add 100 - 125 ml. distilled water and 10 drops of indica-
tor. Put a magnetic stirring bar inside the beaker and place it on the stir-
rer and titrate with NaOH. End point is reached when the indicator changes
from pink to green. Repeat the process.
Determine the normality of H2SO4 using NaOH as the standard. Take 25
ml. of H2SO4 and titrate against NaOH using the above procedure.
Calculation
Normality of KHCsH4O4
Sample Weight = N
20.4230
Normality of NaOH
= 50 x normality of KH4C8H4O4
ml. of NaOH
= N (alkali)
Normality of H2SO4
= 25 = N (acid)
ml. of NaOH x N alkali
237
-------
4. Percentage of MgO in the sample
Weigh accurately 1 gm. of sample for centrifuge cakes, dryer products
and calciner feeds, but 0. 25 gm. for calciner products, in a 400 ml.
beaker. Add 150 - 200 ml. of distilled water and 5-10 ml. of 30% H2O2.
Heat and boil for one-half hour. To the hot solution (not boiling) add 50
ml. of standard 0. 5N I^SC^, heat for another one-half hour. Cool to
room temperature and titrate with 0. 5N NaOH using the above indicator.
Calculation
% MgO = 2.02 (50 x normality of H2SC>4 - ml. of alkali x normality of)
sample weight alkali )
B. MgSOs in Samples
This is determined idiometrically.
Dissolve 25 gm. of KI in as little water as possible. Weigh about 12. 70 gm. of
iodine (AE) and dissolve completely in the KI solution. Any insoluble can be
filtered through a glass Gooch crucible. Dilute the clear solution to a liter by
distilled water. Store in a dark bottle.
Standardize the iodine solution as follows:
Weight accurately, as close as possible, 0. 2 gm. of AR grade As2Og in
a beaker. Dissolve in the smallest amount of 40% NaOH solution. Dilute
to 200 ml., add a drop of phenolphthalein indicator. Add, dropwise, 1:1
H2SOA until the pink color disappears. Add about 5 gm. of NaHCOs to the
solution, stir with a magnetic stirrer and titrate with 0. IN iodine solution
with starch indicator.
238
-------
Calculation
Wt. of AS^OS taken = A
0.04948
Normality of iodine - _ A
ml. of 12 solution
1. 0. IN Na2S2O3 solution:
Dissolve 25 gm. of ^28203. SIH^O per liter. Add 0. 1 gm. of Borax per
liter to stabilize the solution.
Sodium, thiosulfate is standardized titrating against I2 solution with starch
as indicator. Take 25 ml. of Na2S2Oq solution in a 250 ml. beaker, add
100 - 125 ml. of water and a few drops (10) of starch indicator. Titrate
until the blue color persists for 1 minute.
Normality of ^28203 Solution
= Normality of Iodine x ml. of 1 2
ml. of Na2S2Og Solution
2. Starch Indicator
Make a thin paste of about 9 gms. of soluble starch-in cold water, then
add it to one liter of boiling water with constant stirring. Boil for 2 to
3 minutes after all the starch has been added. Cool, add 2 ml. of
Chloroform to preserve the solution.
239
-------
3. Phenolphthalein Indicator
Dissolve 1 gm. of phenolphthalein in 50 ml. of alcohol and add 50 ml. of
water. Filter off any insolubles.
Apparatus
1. A brown bottle (1 liter) fitted with a brown automatic burette
(50ml.).
2. An automatic burette attached to a ^28203 reservoir (10 - 15
liters).
3. Pipettes: 10, 25 and 50 ml.
4. Magnetic stirring bars.
5. Magnetic stirrer.
Procedure
Weigh out accurately about 0. 25 gm. of the material. Add to a beaker con-
taining 50 ml. of standardized iodine solution. Stir with a magnetic stirrer,
add 10 ml. of 1:1 HC1 and titrate immediately with standard sodium thiosul-
fate solution. Add 5 ml. of starch indicator to the beaker as soon as the
color of the mixture becomes pale yellow. Continue titrating until the blue
color disappears and the solution remains colorless for 30 seconds.
Calculation
% SC<2 = 3. 22 x (ml. of 12 solution x Normality - ml. of Na2S2C>3
sample weight solution x Normality)
% MgSO3 = % SO2 x 1. 625
240 -
-------
C. Determination of MgSQ4 ;
This is determined nephelometrically. A Hach colorimeter (DRA) was used for
this determination.
Apparatus
1. Hach Colorimeter (DRA)
2. Pipettes: 1, 2, 5 and 10 ml.
3. Volumetric flasks.
4. Beakers.
Procedure
Weigh about 2. 50 gm. of sample accurately in a beaker (250 ml.). Add 100 ml. of
water and 25 ml. of 1:1 HC1. Boil for one-half hour. Filter off any insolubles.
Dilute to 250 ml. in a volumetric flask.
Take appropriate volume (1, 2, or 5 ml.) in the bottle provided with the Colori-
meter. Add the reagent supplied by Hach, dilute to mark and then read p. p. m.
from, the card in the Colorimeter. (The solution concentration is manipulated in
such a way that p. p.m. reading is between 0 - 75).
Calculation •
% 804 = p.p.m x *' = A
sample weight
% MgSO4 = 1. 253 x A
/ = 0.05 for 5 ml. of the original solution
= 0.10 for 2. 5 ml. of the original solution
= 0.125 for 2.0 ml. of the original solution
= 0. 25 for 1 ml. of the original solution
- 241:
-------
D. Carbon
Carbon in the sample is determined in Leco semiautomatic combustion fur-
nace (#572-100). Follow the procedure described in the Leco instruction
booklet.
E. Moisture Analysis
Two varieties of moisture are recognized in the centrifuge cakes:
(1) free or absorbed water
(2) combined or chemically bound water
1. Free or absorbed water:
Weigh out a beaker (250 ml.) and record weight. Add about 50 gm. of
sample and record weight. Weigh a filter-paper and a clean watch-glass
(large enough to accommodate the filter paper). Add approximately 70
ml. of acetone to the sample in the beaker and mix well. Transfer all the
slurry (add more acetone if necessary) to the filter paper in a Buchner
funnel, and filter under suction. Wash twice with acetone. Air-dry the
sample. Transfer the paper to the watch-glass and dry in an oven at 40°c
for one-half hour. Re-weigh the sample and watch-glass and filter paper
to get the weight of the oven dried sample.
Calculation
% Solid = Wt. of oven dried sample x 100
Wt. of original sample
% Free water = 100 - % Solid
242
-------
2. Combined Water
Ground the oven dried sample in a mortar. Place sample (10 g.) on the
aluminum pan of the Ohaus balance (#610). Adjust the wattage and the
height of the lamp so that a temperature of 160 - 165 C is attained at
the center of the pan. Heat for 50 minutes.
Calculation
% combined water = scale reading.
F. Other Analyses
1. Iron
Determined colorimetrically using Hatch. Follow the procedure given in
water analysis Handbook by Hatch Chemical Company. For calculation and
sample preparation: Weigh accurately about 1 g. of the sample. Transfer
to a 250 ml. beaker, add 50 ml. water and 5 ml. of concentrated HC1.
Boil for 15 minutes. Filter off insolubles and dilute to 1000 ml. Take
suitable aliquots to give a reading of 2 p. p. m. or lower.
Calculation
Reading in (p. p. m.) x 25 = mgs. /1000 ml. = A
aliquot
0.1 A = % Fe
sample wt.
Note: The pH of the iron solution should be between 3 and 6.
If too acidic, adjust pH by adding alkali.
243
-------
2. Nickel
Weigh accurately about 1 gm. of sample in a 250 ml. beaker. Add concen-
trated HC1 (10 ml.) and 100 ml. of water. Boil until solution is complete.
Add 5 ml. of 30% H2O2 and boil for 15 minutes. Add NH4C1 (2 gm.) and
then add NH^OH until it just smells of NHg. Filter hot, wash the precipi-
tate with hot water. Add a few drops of concentrated HC1 until acid. To
the hot solution add 25 to 50 ml. of dimethyl glyoxime solution and then 3
gms. of solid sodium acetate. Keep on low heat for at least 1 hour. Filter
through a weighed No. 4 glass Gooch crucible - wash with hot water - dry
at 110°c. for 1 hour - cool in a desiccator. Weigh. The difference in the
two weights is the weight of Nickel dimethyl glyoxime complex.
% Ni = Wt. of Ni-complex x 0. 2031 x 100
Wt. of sample taken
Dimethyl glyoxime solution
Dissolve 0. 6 gm. of dimethyl glyoxime (solid) in 100 ml. of 95% ethyl
alcohol. Filter if necessary.
3. Vanadium
Weigh accurately about 1 gm. of the sample in a 100 ml. beaker. Add
5 to 10 ml. of concentrated HC1, heat to dryness, cool, add 3-5 ml.
of concentrated H2SC>4 and heat again to dryness (fumes of 803). Cool.
Add 50 ml. of water, boil for one-half hour. 'Filter, wash and dilute to
100 ml.
244
-------
Take 10 ml. of the stock solution in a 100 ml. flask, add 5 ml. of 3%
H2O2 and 0. 25 g. of NaF. Make up to the mark with 2N H2SO4.
Transfer the solution to the cell and read absorption at 450 m/w. Then
determine the p.p.m. from the graph.
Calculation
% V = p. p.m. x 0.1
sample wt.
NOTE: For 20 ml. of stock solution
% V = p.p.m. x 0.05
sample wt.
Construction of graph
Weigh accurately analytical grade ammonium vanadate to make a stock
solution of 1 mg. of V/ml. (1000 p. p.m.) in 5N Ir^SO^.. Make up solutions
of 10, 20, 30, 40, 50, 60, 70, 80, 90 and 100 p. p. m. vanadium content by
proper dilution. Proceed as above and plot p. p.m. vs. absorbance curve.
4. Chloride
The chloride is determined by AgNC>3 titration.
Weigh accurately about 1 gm. of the sample in a 250 ml. beaker. Add 100
ml. of distilled water and 5 ml. of concentrated HNC>3. Boil until the solu-
tion is reduced to half its original volume. Filter off any insolubles. Dilute
to 200 ml. in a beaker or Erlynmeyer flask, add 5 ml. of 1:1 HNO3, 25 ml.
of standard AgNC>3, 1-2 ml. of ferric alum indicator and titrate with stand-
ard NH^CNS solution until the reddish-brown color persists on stirring (use
a magnetic stirring-bar during titration).
245
-------
Calculation
ml. of AgNO3 x N of AgNO3 -ml. of NH4CNS x N of NH4CNS = A
% Cl = 0.03546A x 100
wt. of sample
For liquid samples use the above procedure. Take 25 to 100 ml. of
liquid sample depending upon chloride content.
mg. of Cl/liter = 35.46 A x 1000
vol. of sample taken
Standard AgNOcj Solution
Dissolve 16. 9870 gm. of reagent grade AgNOo per liter. Store in a
dark bottle away from direct light.
Normality of AgNO3 = Wt. of AgNO3 taken
16.9870
Standard NH4CNS Solution
Dissolve 7. 6 gm. of NH^CNS per liter. Standardize against AgNO3 solu-
tion. Take 25 ml. of standard AgNO3 solution in a 250 ml. beaker. Add
100 ml. of distilled water and 5 ml. of 1:1 HNO3. Add 1 ml. of ferric
alum indicator and titrate to reddish-brown endpoint with NH4CNS.
Calculation
Normality of NH4 CNS = 25 x Normality of AgNO3
ml. of NH4CNS
Ferric alum indicator
Dissolve 140 gm. of ferric ammonium sulfate (pure) in 400 ml. of water.
Add 10 ml. of 6N HNO3 (390 ml. of concentrated HNO3 diluted to 1 liter).
Heat if necessary. Filter off any insoluble. Cool. Dilute to 500 ml. with
6N HNO3.
246
-------
Magnesium and Calcium by EDTA
Preparation of EDTA Solution
Dissolve 37. 2240 gm. of reagent grade disodium dihydrate salt of EDTA
in 1 liter of water - 0.1M EDTA. Standardize against AR CaCOg or Mg
salt. Store in a plastic bottle.
1 ml. of 0.1M EDTA = 2. 4320 mg. of Mg+2
= 4. 0080 mg. of Ca+2
Buffer
Dissolve 67. 50 gm. of NH4C1 in distilled water. Add 570 ml. of NH4OH
and dilute to 1 liter.
Indicator
\,.
Dissolve 0. 5 gm. of Eriochrome Black T in 100 ml. of ethyl alcohol.
Add 1 gm. of Murexide to 100 gm. of pure NaCl, mix well in a mortar.
Store in a bottle.
Ammonium Sulfide Solution
Saturate 200 ml. of concentrated NH4OH with E^S (keep the solution cold
while passing H2S). Add 200 ml. of concentrated NH4OH and dilute to 1
liter.
Standardization of EDTA using Ca+2
Weigh any Ca compound (pure) accurately so that no more than 0. 5 gm.
of Ca*2 is present per liter.
+2
Take 100 ml. of standard Ca solution, add NaOH solution (4 gm. of NaOH
per liter) until the solution has a pH 12 (check with pH paper). At this point
precipitation might take place. Add a small amount of solid Murexide - NaCl
indicator and titrate with EDTA to a color change of orange to violet.
247
-------
Calculation
Molality of EDTA = ml. of EDTA x gm. of Ca+2 in 100 ml.
0.04008
Standardization of EDTA by Mg
+2
Weigh any AR grade Mg salt accurately so that no more than 1 gm. of
Mg+2 is present in 1000 ml.
To 100 ml. of nearly neutral solution add 10 ml. of buffer and 1 ml. of
the Black T indicator and titrate from red to clear blue color.
1 ml. of 0.1M EDTA = 0.002432 gm. Mg+2
i p -|_2
Determination of Ca * and Mg in centrifuge cake, dryer product, etc.
Weigh accurately 1.5 to 2 gm. of the sample in a 250 ml. beaker. Add
10 ml. of concentrated HC1 and 10 ml. of water. Cover the beaker with
a watch glass and heat (on low heat) almost to dryness. Add 100 - 150 ml.
distilled water, boil until solution (a few ml. of concentrated HC1 may be
added). Add 5 gm. of NH4.C1 and NH4OH dropwise until ammoniacal. Add
10 ml. of (NH^S solution. Filter off any precipitate, wash with hot water
and dilute to 1 liter.
Total Ca+2 and Mg+2
For Mg+2
To 25 ml. of the solution (almost neutral) add 5 ml. of buffer and 10 drops
of Eriochrome Black T indicator. Titrate (with stirring) to blue endpoint.
Note the ml. of EDTA used = 'A1.
For Ca+2
To 25 ml. of stock solution add enough IN NaOH until the pH of the solution
reaches at least 12 (check with pH paper). Add 0.1 gm. of Murexide - NaCl
248
-------
indicator. Titrate to violet end point with EDTA. A mount of EDTA
consumed = 'B1 ml.
Calculation
For Mg+2
A - B = 'C' ml.
Amount of Mg+2 in 25 ml.
= C x molality of EDTA x 0. 02432
= C gm.
In 1000 ml. = C x 40 = C gm.
% Mg+2 = C x 100
wt. of sample
For Ca+2
+2
Amount of Ca in 25 ml.
= B x molality of EDTA x 0. 04008 = b gm.
+2 f
Amount of Ca in 1000 ml. = b x 40 = b gm.
% Ca+2 = b x 100
wt. of sample
6. Degree of Hydration of MgO
Determine the percentage of MgO in the sample using the procedure A.
Place sufficient amount of the sample in a beaker, add water to make a
10-15% slurry. Place the beaker inside an oil bath (set at the desired tem-
perature level). Stir the contents of the beaker with a stirrer. Keep the
beaker in the bath for the desired period of time. Filter under suction,
wash the sample with acetone, dry in an oven at 45 - 50°C for 1 hour.
Analyze again for MgO content by procedure A.
249 ~
-------
Calculation
% MgO in the sample before hydration = A
% MgO after hydration = B .
% Hydration = (A - B) x 3. 22
7. Determination of the Ratio of MgSO3. 6H2O and MgSO3. 3H9Q in the
Centrifuge Cake
The two hydrates of magnesium sulfite behave differently towards silver
nitrate solutions:
1 gm. of MgSO3. 3H2O = 2.1450 gm. of AgNO3
1 gm. of MgSO3.6H2O = 0.4000 gm. of AgNO3
Reagents
1. 0. 5N AgNOs solution
2. 0. IN NH4CNS solution
3. Ferric alum indicator
Procedure
Analyze the cake for H2O, MgO, MgSO4 and MgSO3 contents in the usual
way.
(a) amount of MgSO4 less than 5%
and
(b) amount of MgSO4 more than 5%.
For (a) no treatment is necessary.
For (b) add about 3 gm. of powdered centrifuge cake
250
-------
to 200 ml. of absolute alcohol while stirring. Stir for one-half hour,
filter under suction, wash with acetone and dry in air. Repeat the process
until the amount of MgSO4 is below 5%. (Solubility of MgSO4, 0. 2 gm. /100
ml. of alcohol).
To 25 ml. of 0. 5N AgNO3 solution in a clean 100 ml. beaker add exactly
1 gm. of sample. Stir the mixture with a Teflon coated magnetic stirrer
for 5 minutes. Filter through a clean sintered glass Gooch crucible (fine).
Wash the precipitate twice with 5 ml. (each) of 50/50 methanol-water mix-
ture.
'Dilute the filtrate to 100 ml. in a volumetric flask. The total amount of
AgNQ3 is determined by titration with standard NH^CNS using ferric alum
as indicator,.
Calculation
100 - (%H20 + % MgSO3 + %MgSO4 + %MgO) = X
% MgO + (%MgSO4 x 2. 05) = Y
Amount of AgNO3 consumed per 1 gm. of sample = C
Corrected amount of AgNC>3 consumed = C + C (X + Y) = D
"Too
Determine the ratio of MgSO3. 3H2O: MgSO3. 6H2O from the graph.
Construction of graph
Draw a graph from the data in the following Table. Plot along X-axis the
ratio of the two hydrates and the amount of AgNO3 consumed on the Y-axis.
•251
-------
Ratio of 3HpQ
6H2O gms. of AgNQ3 consumed
100/0 2.1450
80/20 1.7960
60/40 1.4470
50/50 1.2725
40/60 1.0980
20/80 0.7490
0/100 0.4000
Note: The following, elements interfere with this method:
Sulfide, all Halogens (except Fluoride), Cyanides,
Thiosulfate.
- 252
-------
REACTIVITY TESTS
None of the procedures practiced in the Magnesia industry e. g. surface
area measurements (by iodine method) or degree of hydration is applicable
to this project. Hence, the following method was developed in order to
measure the quality of regenerated MgO.
Magnesium oxide, in aqueous suspension, reacts with sulfur dioxide as:
MgO + XH20 + S02 *- MgS03.XH2Q
Magnesium sulfite, being insoluble, precipitates out of the solution.
From the above equation:
40. 31 gm. of MgO reacts with 64. 06 gms. of SO2.
Thus, the theoretical SO2 removal capacity of MgO is
64.06 X-f - 100
40.31
The value of ' T ' being 62. 93. In this procedure it is assumed that
MgSOo is completely insoluble in water and no side reactions occur.
APPARATUS
Pure SO2 gas-cylinder with single stage regulator (Model No. 71) from
Matheson.
Flow meter (Roger Gilmore Industries, Inc.) No. 12 (Cat. No. F-2260A)
253
-------
Two 3 - necked glass flasks (500 ml.)
Two gas purgers (sintered glass extra coarse porosity)
Magnetic stirrer (bar and plate)
500 ml. Volumetric flasks
Burettes (50 ml. )
Pipettes (5 ml. and 2 ml.)
REAGENTS
Caustic soda
Starch indicator
Standard iodine solution
Standard sodium thiosulfate solution
PROCEDURE
gas flow is adjusted in such a manner that about 24 gms. of SC>2 pass
through the purger in 20 minutes. The equipment arrangement for this
test is shown in Fig. 1.
The amount of SC>2 passing through the system for any particular setting
of the rotometer, is measured by absorption in 50% NaOH and subsequent
i
titration of the sulfite by iodine solution.
From the amount of SO2 passing through the system the theoretical amount
of MgO necessary is calculated as follows:
254
-------
Amount of SO2 in grams X 0. 6292 = amount of MgO (in gm. ) required.
The required amount of MgO, calculated as above, is placed in flask No. 1.
The 300 ml. of standard iodine solution, in the second flask, absorbs any
SO2 that escaped from flask No. 1.
After the passage of SO2 through the system, the whole system is purged
with air or nitrogen to remove any SO2 in the line and transfer it into the
two absorbers. The pH of the MgSOg-MgO slurry is measured.
The magnesium sulfite slurry is filtered under suction, washed twice with
water. The filtrate is diluted to 500 ml. and the SO2 content determined
idiometrically. The cake is washed with acetone, air and oven (40°c. ) dried
and analyzed (if desired) for MgO, MgSO4, MgSO3 and H2O content. The con-
tents of iodine flask is also diluted to 500 ml. and the SO2 content is determined.
CALCULATIONS
Amount of SO2 passed 'a1 gm.
Amount of MgO taken 'b1 gm.
(for 'a' gm. of SO2)
Amount of SO2 in MgSO3 filtrate 'c' gm.
Amount of SO2 in iodine flask 'd' gm.
255
-------
SO9 removal capacity of MgO = a - (c + d) x 62.93
I b
N.B. For routine purposes the iodine flask can be discarded, as
only a small amount of SO2 escapes from flask No. 1.
REMARKS
The SO2 removal capacity is a pure and arbitrary number.
Because of the nature of assumptions made, e. g. insolubility of MgSC>3
in the slurry, no side reactions, disregarding bisulfite formation, etc.,
the results obtained are low. Thus, a value of 40 - 50 is considered to be
good. Even virgin MgO, under these circumstances, gives a value of 60 - 65.
256
-------
TECHNICAL REPORT DATA . .
(Please read Instructions on the reverse before completing) ': .'"V, ',.. ,-• ,
. REPORT NO. 2.
EPA-600/2-75-057
I. TITLE AND SUBTITLE
The i Magnesia Scrubbing Process as Applied to an
Oil- Fired Power Plant
AUTHORS George Koehler (Chemical Construction) and
James A. Burns (Boston Edison)
3. PERFORMING OR3ANIZATION NAME AND ADDRESS
Chemical Construction Corp.
One Penn Plaza
New York, NY 10001
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
October 1975
6. PERFORMING ORGANIZATION CODE
8. PERFORMING ORGANIZATION REPORT NO.
10. PROGRAM ELEMENT NO.
1AB013; ROAP 21ADA-004
11. CONTRACT/GRANT NO.
CPA 70-114
13. TYPE OF REPORT AND PERIOD COVERED
Final; 6/70-6/74
14. SPONSORING AGENCY CODE
'15. SUPPLEMENTARY NOTES
is. ABSTRACT The rep0r£ gjves results of B. full size demonstration of the magnesia wet-
scrubbing system for flue gas desulfurization on an oil fired, 150 MW generating
unit. The project involved: design and construction of both an SO2 removal system
(based on firing 2. 5% sulfur fuel) and an MgO regeneration facility for 50 ton/day
I acid production; and operation of both plants over a 2 year period. The report shows
that the system removed 90% of the inlet SO2 and controlled particulate emissions
from the generating station, where 4000 hours of operation were logged. Regener-
ated magnesia was recycled successfully and over 5000 tons of acid was marketed
from the regeneration plant, demonstrating the feasibility of the process. Corre-
lations were developed both to determine SO2 removal for varying boiler loads and
fuel sulfur content, and to control regeneration of acceptable alkali. Several other
studies of the process technology and chemistry were undertaken as part of the work.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Magnesium Oxides
Regeneration
(Engineering)
Fuel Oil
Scrubbers
Combustion
Products
Flue Gases
Desulfurization
Sulfur Dioxide
Sulfuric Acid
Air Pollution Control
Stationary Sources
Magnesia Scrubbing
Particulate
13B
07B
2 ID
07A
2 IB
07D
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (Tliis Report/
Unclassified
21. NO. OF PAGES
267
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
£PA Form 2220-1 (9-73)
257
------- |