EPA-R2-72-038
September 1972 Environmental Protection Technology
Development of the Catalytic
Chamber Process, Final Report
Office of Research and Monitoring
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
EPA-R2-72-038
Development of the
Catalytic Chamber Process,
Final Report
By
A. Gruber and A. Walitt
Tyco Laboratories, Inc.
Bear Hill, Waltham,
Massachusetts 02154
Contract No. 68-02-0008
Program Element 1A2013
Project Officer: Stanley J. Bunas
Control Systems Division
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
OFFICE OF RESEARCH AND MONITORING
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
September 1972
-------
EPA REVIEW NOTICE
This report has been reviewed by the Environmental
Protection Agency and approved for publication.
Approval does not signify that the contents necessarily
reflect the views and policies of the Agency, nor does
mention of trade names or commercial products constitute
endorsement or recommendation for use.
11
-------
ABSTRACT
The goal of this contract was to continue the development of the Tyco Catalytic
Chamber Process for the simultaneous removal of SO0 and NO from power plant flue gas.
& X
As originally conceived, the process involved the gas phase oxidation of SO0 with NO9 followed
Lt Lt
by the high temperature absorption of the nitrogen oxides in sulfuric acid. The dissolved oxides
of nitrogen were recovered and oxidized in a catalytic stripper. The process produced nitric
and sulfuric acids.
The results of the program conducted in a 10-SCFM miniplant were:
1. The SO» gas phase oxidation by NO0 was found to proceed too slowly to permit
the use of practical-sized equipment. However, by contacting the flue gas with 2 wt % nitro-
sylsulfuric acid in 70% sulfuric acid in a packed bed reactor, 90% removal of SCL could be
accomplished at 225 °F in less than 15 sec.
2. Absorption of equimolar quantities of NO and N02 was not found to be
controlled by a conventional absorption mechanism, but is probably dependent on a gas phase
interaction of die two oxides. As a result, the large equipment needed to provide long residence
times makes the process appear relatively expensive for removing SO, and NO together or
NO alone.
x
111
-------
Table of Contents
Section Page No.
I. SUMMARY [[[ j
II. INTRODUCTION [[[ 9
A. Content of This Report .......................................... <-,
B. Initial Process Concept .......................................... ,,
C. Prior Work
D. Contract Goals
< ,
III. ENGINEERING SUBCONTRACTOR - THE BADGER
CORPORATION [[[ 13
A. Engineering Consultation ........................................ ^3
B. Optimization Program ........................................... j3
C. Pilot Plant Design ............................................... j 3
IV. MINIPLANT MODIFICATION AND EQUIPMENT ........................ 15
A. Existing Equipment ............................................. 15
B. SO? Oxidation Reactor ........................................... 15
C. Scrubber Columns .............................................. 15
D. Stripper Columns ............................................... 17
E. Liquid Pumps [[[ 21
F. Gas Pumps [[[ 21
G. Instrumentation ................................................. 22
H. Gas Feed [[[ 22
I. Miniplant Operation ............................................. 24
V. REACTOR EXPERIMENTS .......................................... 25
A. Background [[[ 25
B. Reactor Experiments: Effect of UV Radiation ...................... 25
C. Miniplant Tests: Effect of Temperature on SO,,
Oxidation Rate .................................................. 27
D. Effect of Acid Irrigation on SO» Oxidation Rate ..................... 27
E. Instrumentation Considerations ................................... 35
-------
Table of Contents (Cont)
Section page NO-
C. Initial Experimentation - Parametric Study 64
D. Mathematical Analysis of Experimental Scrubber Data 67
E. Analysis of Scrubber/Stripper Data 74
VII. STRIPPER EVALUATION 81
A. Initial Concept 81
B. Laboratory Scale Stripper Tests 81
C. Miniplant Experimentation 82
D. Analysis of Stripper Mechanism 83
VIII. PROCESS CONTROL 85
A. Goals 85
B. Power Plant Operating Conditions 85
C. Instrumentation Requirements 87
D. Process Control Requirements 91
IX. ECONOMIC EVALUATION 95
A. Prior Evaluations 95
B. New Process Concept 95
C. Cost of NO Removal Only 101
X
X. CONCLUSIONS AND RECOMMENDATIONS 113
A. Conclusions 113
B. Recommendations 114
XI. ACKNOWLEDGMENTS 117
XII. REFERENCES 119
Appendices
I. PRELIMINARY ENGINEERING ANALYSIS OF THE TYCO
CATALYTIC CHAMBER PROCESS BY BADGER CORPORATION I-1
II. MINIPLANT OPERATION II-l
III. DATA FROM MINIPLANT REACTOR EXPERIMENTATION Ill-1
IV. DATA FROM SCRUBBER/STRIPPER EXPERIMENTS IV-1
V. METAL ION INTERFERENCE IN SCRUBBER AND
STRIPPER OPERATION V-l
VI
-------
Table of Contents (Cont.)
Appendices Page No.
VI. CORROSION IN MINIPLANT EQUIPMENT VI-1
VII. MATHEMATICAL ANALYSIS OF SCRUBBER/STRIPPER DATA VII-1
VIII. LABORATORY STRIPPER TESTS VIII-1
IX. INSTRUMENTATION CALIBRATION IX-1
X. OXIDATION OF NO TO NO2 FOR MINIPLANT GAS FEED X-1
vii
-------
List of Illustrations
Figure No. Page No.
1. Catalytic chamber process 3
2. Schematic flowsheet of miniplant 16
3. Sketch of 19-ft scrubber 18
4. Miniplant stripper column and catalyst bed support structure 19
5. Stripper column using glass pipe 20
6. Modified miniplant layout 23
7. Miniplant reactor experiments — SO2 oxidation. Effect of
water vapor in flue gas, residence time, and temperature 29
8. Configuration of miniplant reactor during initial experimentation
involving liquid irrigation 31
9. Miniplant reactor experiments — SC^ oxidation. Effect of reactor
irrigation with 80% H2SO4 32
10. Miniplant reactor experiment — SO2 oxidation. Effect of acid or
water irrigation in packed reactor 33
11. Miniplant reactor experiment — SC>2 oxidation. Comparison of acid
irrigation, wetted wall tests, and no irrigation 34
12. Miniplant reactor experiments - SC>2 oxidation. Comparison of
wetted-packing reactions with and without additional unpacked
residence volume 36
13. Miniplant reactor experiment - SO2 oxidation. Evaluation of low
O2 level in flue gas 37
14. NOg interference calibration curve 39
15. Miniplant reactor experiment - SO2 oxidation. Tracing of
instrumentation recorder input 40
16. Miniplant reactor experiment - S02 oxidation. Tracing of
instrumentation recorder output 41
17. Miniplant reactor experiment - SO£ oxidation. Comparison of
acid and nitrose irrigation 43
18. Miniplant reactor experiment. Effect of HNSO5 concentration
on SO, removal yield 45
19. Miniplant reactor experiment. Effect of HNSO- concentration
on SO2 removal yield 46
20. Miniplant reactor experiment. Effect of HNSO9 concentration
on SO2 removal yield 47
ix
-------
List of Illustrations (Cont.)
Figure No. Page No.
21. Miniplant reactor experiment. Effect of H2SO4 concentration
on SO2 removal yield 48
22. Miniplant reactor experiment. Effect of residence time on
SO2 removal yield 49
23. Interpretation of miniplant reactor experimentation. Tradeoff
between HNS02 concentration and irrigation flow rate in achieving
a specific SO- reaction yield 50
24. Miniplant reactor experiments. Effect of irrigation and
residence time on SO- removal yield 51
25. Miniplant reactor experiments. Effect of countercurrent versus
cocurrent reactor configurations 53
26. Scrubber/stripper parametric study and data sheet 65
27. Series stripper configuration 69
28. First order reaction rate constant KI as a function of NO
inlet concentration. Data from Tyco and EPA * 76
29. Second order reaction rate constant K2 as a function of NOX
inlet concentration. Data from Tyco and EPA 77
30. Standard deviation for first order rate constant values,
85% confidence limits 78
31. Standard deviations for second order rate constant values 79
32. Graph of power plant operating parameters 88
33. Major process control requirements 92
34. Flowsheet showing the Tyco catalytic chamber process used
for NO removal only 96
X
35. Proposed flowsheet for catalytic chamber process utilizing
a wetted reactor for SO, oxidation 100
36. Extrapolation of data from Run 92 to give scrubber column height
for reducing NOx concentration from 6000 ppm to 100 ppm 102
37. Extrapolation of data from Runs 85c, 86c, 92c, 93, and 96c to give
scrubber column height for reducing NOX concentration from
2000 ppm to 100 ppm 107
38. Catalytic chamber process using a low temperature scrubber 115
-------
List of Tables
Table No. Page No.
I. Initial Reactor Studies 26
II. Gas Phase SC>2 Oxidation with NO2 in the
Presence of UV Radiation 28
III. Degree of Conversion of SC>2 at 150 °C in 80% Sulfuric
Acid for Various Residence Times with Diffusion of
SC>2 Being the Rate Limiting Step 60
IV. Degree of Conversion of S02 at 150 °C in 80% Sulfuric
Acid for Various Residence Time with Rate of Conversion
Equal to 1/10 of the Diffusion Limited Rate 61
V. Degree of Conversion of SOr, at 150 °C in 80% Sulfuric
Acid at Various Residence Times Using Modifying
Assumptions Presented in Text 61
VI. Parametric Study and Supplemental Work Plan 68
VII. CC-2 Analysis of Laboratory Stripper Tests 84
VIII. Economics of NOX Scrubbing Plant for 500-MW Power
Plant Based on Tyco SO, Removal Process 97
IX. Economics of NOX Scrubbing Plant for 500-MW Power
Plant Based on Tyco S02 Removal Process 98
X. Economics of NOX Scrubbing Plant for 500-MW Power
Plant Based on Tyco SO2 Removal Process 99
XI. Economics of SO2/NO2 Scrubbing Plant for 800-MW
Power Plant 103
XII. Economics of SO2/NOX Scrubbing Plant for 800-MW
Power Plant 104
XIII. Economics of S02/NOX Scrubbing Plant for 800-MW
Power Plant 105
XIV. Economics of SO2/NOX Scrubbing Plant for 800-MW
Power Plant 106
XV. Economics of NOX Scrubbing Plant for 800-MW Power Plant 108
XVI. Economics of NO Scrubbing Plant for 800-MW Power Plant 109
X
XVII. Economics of NO Scrubbing Plant for 800-MW Power Plant 110
X
XVIII. Economics of NO Scrubbing Plant for 800-MW Power Plant Ill
X
XI
-------
I. SUMMARY
During the four and one-half year period between October 1967 and March 1972, Tyco
Laboratories, Inc., has been involved in the development of the Catalytic Chamber Process for
the removal of sulfur dioxide and nitrogen oxides from power plant flue gas. Under contract to
the Department of Health, Education and Welfare and the Environmental Protection Agency
(Contracts PH 86-68-75, CPA 70-59, and 68-02-0008), Tyco has brought the process from the
conceptual stage through several major modifications to the pre-pilot stage where a reasonable
decision could be made on its economic and technological viability.
The development work on the process was aimed at solving one of the major environ-
mental problems facing our society: air pollution caused by flue gases from power stations
burning fossil fuels. A typical 800 MW, coal-burning power plant produces almost 1.5 million
2
standard ft /min of stack gas containing about 0.3% sulfur dioxide and 0.05% nitrogen oxides.
These pollutants are equivalent to 900 tons and 100 tons of sulfuric and nitric acids (100% basis),
respectively, being dumped into the atmosphere daily. Both species represent a serious
hazard to man and his artifacts as well as to plant and animal life in general. The magnitude
of the sulfur dioxide problem can be demonstrated by the fact that more sulfuric acid equiva-
lent is dumped into the atmosphere from individual sources than is used for all commercial
applications (and it should be remembered that sulfuric acid is the largest tonnage chemical
commodity marketed today).
The need for power plant flue gas cleanup is urgent, but the very magnitude of the
problem which creates this urgency is its biggest impediment. Flue gas cleanup involves the
handling of enormous volumes of gas; ducts are typically 15 to 20 ft in diameter. Any process
which would require reaction residence times beyond a few seconds is extremely unattractive
economically because of the very high cost of large gas handling equipment. Other major ob-
stacles to a viable process are the need for heat transfer with the flue gas and the disposal
problem of the recovered pollutant. As will be seen, the original concept of the Catalytic
Chamber Process avoided most of the major technological problems, while giving a relatively
attractive economic outlook.
The process concept is based on the century-old Lead Chamber Process for sulfuric
acid manufacture. This obsolete process made use of nitrogen dioxide to oxidize sulfur dioxide
(about 13%) to sulfuric acid, which was formed in large lead chambers. The nitric oxide which
-------
was formed during this oxidation was slowly air-oxidized back to nitrogen dioxide to permit
the complete oxidation of sulfur dioxide before the gas left the lead chambers. The gas was
then scrubbed with cold sulfuric acid to remove residual sulfuric acid and oxides of nitrogen.
The gas was dumped to the atmosphere while the nitrated acid (nitrose) was recycled to permit
continued use of the nitrogen dioxide oxidant.
Early experimentation at Tyco indicated that sufficiently rapid oxidation of sulfur
dioxide in power plant flue gas (about 0.3%) could be achieved to convert the gaseous oxide to
sulfuric acid in short enough residence times to permit the use of practical-sized equipment.
The simultaneous reoxidation of nitric oxide to nitrogen dioxide could not be accomplished fast
enough to be economically feasible, making it clear that the Lead Chamber Process could not
be used directly for dilute gas cleanup. In addition, the cold scrubbing of the wet flue gas
(about 7% water) caused excessive dilution of the recycle scrubbing acid, creating a need for
large amounts of heat to reconcentrate the acid. Even more heat was needed to recover the
oxides of nitrogen to permit the recycle of both the gasses and the scrubbing acid.
These problems were conceptually solved during subsequent experimentation by the
evolution of the Catalytic Chamber Process. (See Fig. 1.) In this scheme, the dilute sulfur
dioxide is oxidized to sulfuric acid with an amount of nitrogen dioxide equal to twice the stoi-
chiometric requirement.
SO2 4 2NO2 - S03 4 NO 4 NO2
Thus, half the nitrogen dioxide is converted to nitric oxide, resulting in a one to one mole
ratio of nitrogen dioxide and nitric oxide which is essential for absorption in sulfuric acid:
NO -I- N02 4 2H2S04 - 2HNSOg 4 H20
The nitrogen oxide and sulfuric acid in the gas stream are absorbed in a high temperature
scrubber. The scrubber is maintained at a temperature such that the vapor pressure of water
over the effluent acid is equal to the partial pressure of water in the incoming gas (for 7%
water in the gas the scrubber temperature would be about 250 °F). In this manner, the process
avoids the pickup of water from the flue gas and thus avoids the need for additional heat for
concentrating the acid.
The effluent nitrose from the scrubber is fed to a chamber packed with activated
charcoal. Here the dissolved oxides of nitrogen are oxidized to nitrogen dioxide, which, being
only very slightly soluble in sulfuric acid, bubbled out of the acid ready for recycle.
activated
4HNSO, + 2H0O 4 O0 - 4H9SO, 4 4NO9
a £t & * i & **. ft
charcoal
This stage was shown experimentally to occur in the 250 to 300 °F temperature range, again
avoiding the need for heat.
-------
uas to suit.*.
250 °F
7%
H20 __
150 ppm N0y
300 ppm SO2
Gas
NO
N02
SO0
2
HN(
Air
NO
REACTOR
Gas
N203
1 ^..^VJij
1
\ i
80% H2S04
HIGH
TEMPER ATURE
SCRUBBER
80% H2S04
HNSOp;
1 w
D«
"
ABSORBER
M
Product
HNOg
Flue gas
SOOT
H O
Air .
Product
^
1
1
7
/
/
/
/
///
Y/,
\
J CATALYTIC
, STRIPPHR
/
3000 ppm SOo H so
6000 ppm NO, 2 4
Fig. 1. Catalytic chamber process
-------
This catalytic oxidizer solved two major problems: the reoxidation of the nitric oxide
formed in the sulfur dioxide oxidation stage, and the recovery of the oxides of nitrogen from acid
solution. This stage was the heart of the process at that stage of development. In combination
with the high temperature scrubber, the three shortcomings of the Lead Chamber Process were
avoided, thus creating a process which apparently could recover both nitrogen oxides and sul-
fur dioxide without the need for additional process heat.
The sulfur dioxide was recovered as 80% acid, while the oxides of nitrogen were re-
moved from the off-gas of the catalytic stripper/oxidizer by absorption of nitrogen dioxide in
water to form nitric acid. Preliminary cost estimates showed great promise for economic
feasibility of the process through the sale of commercial grade acids.
As the contract period covered in this report started, the process as outlined above
had been verified in bench scale experimentation as well as preliminary testing of separate
process stages in a 10 SCFM "miniplant" (pre-pilot scale). What remained was the integrated
operation of the total process and the establishment of engineering criteria for scaleup to pilot
level operation at the 2000 to 5000 SCFM level.
The goals of Contract 68-02-0008, therefore, included:
1. Modification of the existing 10 SCFM miniplant to permit integrated opera-
tion.
2. Experimentation in the integrated miniplant to determine probable design
ranges of critical engineering parameters. A parametric study showing the effect of the various
parameters was planned.
3. Preliminary design of a 2000 to 5000 SCFM pilot plant.
4. Evaluation of control levels of SO, and NC- attainable through use of the
£t X
Catalytic Chamber Process and comparison with desired levels. Goals were 90% removal of
S00 and 100 ppm of NO .
£t X
5. Evaluation of the economic feasibility of the overall process and comparison
with EPA standards. For the double removal process (both S02 and NCy, the economic goals
were an installed cost of under $50/KW with an operating cost of less than 2 mils/KWhr.
6. Other goals included preliminary process control evaluations, determina-
tion of a site for the pilot plant, and evaluation of the fly ash problem.
The miniplant was modified to permit integrated operation based on data from pre-
vious experimentation. Improvements in sampling and analytical techniques were incorporated.
The scrubber (4-in. diameter column) was enlarged to a packed height of 27 ft. The reactor
Q
(4-in. dia) was suspended vertically and enlarged to a total of about 15 ft (1.5 ft of
volume) in three columns which could be used in series or in parallel. The stripper was also
set up in three columns, these being 12 in. in diameter with the packing depth in each being
about 35 in. Provisions were again made for a choice of series or parallel operation. Pumps
and heaters were installed which could give the operating range required of a complete para-
metric study.
-------
Initial experimentation involved the determination of operating parameters in the sul-
fur dioxide reactor stage. It became apparent that the purely gas phase interaction between
sulfur dioxide and nitrogen dioxide occurred far too slowly to be feasible. Further testing with
improved sampling and analytical techniques established the fact that the fast sulfur dioxide
oxidation reaction observed previously was in reality a liquid phase interaction. Re-examina-
tion of prior experimentation showed that the liquid phase had always been present in one form
or another when rapid reaction rates had been observed. Once this liquid phase reaction was
verified, it was possible to modify the concept of the Catalytic Chamber Process to include a
stage where the flue gas containing sulfur dioxide contacts a liquid phase consisting of oxides
of nitrogen dissolved in sulfuric acid. Extensive experimentation established a maximum resi-
dence time of 15 sec being required to remove a minimum of 90% of the sulfur dioxide. Solu-
tions containing 2% wt nitrosylsulfuric acid in 70% sulfuric acid were used to contact flue gas
at 225 °F to accomplish this degree of pollutant recovery.
Extensive evaluation of the absorption of equimolar quantities of nitric oxide and nitro-
gen dioxide in sulfuric acid at elevated temperatures (200-275 °F) revealed that the process is
probably a two-step operation and not a simple absorption. Instead of a direct absorption of
nitric oxide and nitrogen dioxide into the sulfuric acid, there is apparently a required gas
phase reaction between the two gases to form nitrogen trioxide which can then be easily
absorbed. This gas phase interaction is relatively slow (compared to the absorption step), and
the equilibrium at elevated temperatures and low partial pressures is strongly shifted in favor
of the unreacted gases. The result is that long residence times in packed towers are required,
thereby greatly enlarging the equipment size.
In estimating the economics of the process, two approaches were taken in sizing the
scrubber system: first, the contact in the scrubber would be the same in the full-sized plant
as in the miniplant experimentation (a gas mass flow equivalent to about 20% of flooding); and
secondly, an assumption that the scrubber could be optimized to approach the industry standard
of about 75% of flooding. The marked difference between the capital and operating costs devel-
oped by using these two assumptions points out the importance of scrubber size in the overall
process.
Using direct scaleup of miniplant data (L/C of 5.0 and a gas mass flow rate of 282 lb/
hr-ft2), an 800-MW power plant producing about 1.5 million standard ft /min of flue gas would
require a scrubber system, including 106 packed columns 25 ft in diameter and 53 ft high to
reduce the NO in the effluent to 100 ppm. The total plant cost would be about $413 million or
X.
about $500/KW installed. Operating costs would be about $ 96 million or 22.2 mils/KWhr,
clearly an impractically high cost. If an assumption of 75% of flooding in the scrubber towers
is made, the costs are $114 million and $26.8 million, respectively, for an installed cost of
$140/KW and an operating cost of 5.5 mils/KWhr. These latter costs are still very high, but
the effect of efficient scrubbing is clearly pointed out. An additional reduction in operating
cost can be achieved by selling the product acids: if sulfuric and nitric acids can be sold for
$10 and $40/ton (100%), respectively, the operating cost can be reduced by about 1 mil/KWhr.
-------
The scrubber operation was evaluated for nitrogen oxide removal only and was found
to be technically feasible, although costly. To reduce the concentration of nitrogen oxides from
2000 ppm to 100 ppm in an 800-MW power-plant, it would take 48 packed towers 25 ft in diam-
eter and 45 ft high if we scaled up directly from miniplant data. The cost of this plant would
be about $175 million, with an annual operating cost of about $41 million (8.5 mils/KWhr).
Again, assuming that optimization would permit operation at 75% of flooding in the scrubbers,
the plant cost could be reduced to $43.5 million and the operating cost to $10.9 million (2.3 mils/
KWhr), showing the premium to oe gained by improving the scrubber operation.
Evaluation of the catalystic stripper/oxidizer was incomplete because of the emphasis
placed on the reactor and scrubber stages. Initial indications were that the oxidation/desorp-
tion mechanism may be something other than a liquid phase oxidation of dissolved oxides of
nitrogen followed by spontaneous generation of nitrogen dioxide. The presence of carbon
dioxide and a large proportion of nitric oxide in the effluent gas at low nitrosylsulfuric acid
concentrations (< 0.3%) suggested that the dissolved oxides of nitrogen were being reduced on
the carbon (and then oxidizing the charcoal). Some nitrogen dioxide was formed by gas phase
oxidation of the nitric oxide, but this reaction is very slow at low nitrose concentrations, and
only about 25 to 40% of the total nitrogen oxide content of the stripper off-gas was in the form
of the desired nitrogen dioxide. Results in this area are contrary to previously-obtained data
at higher nitrose concentrations, and it could well be that operation of the charcoal column
under other conditions may cause the necessary liquid phase oxidation.
Preliminary evaluations of process control requirements were initiated during this
contract. The first step was to ascertain the operating parameters commonly employed in
the power industry. Information obtained included turn-up and turn-down rates, % of excess
air, stability of air flow rates, operating changes during equipment failure, and stack effluent
properties. Two plants were evaluated: the Astoria, New York unit of Consolidated Edison
(primarily oil-fired), and TVA's Chattanooga station (coal-fired). Both showed surprisingly
similar operating parameters. In addition, the Brayton Point station (oil-fired) of the New
England Power Co. was briefly evaluated, with conditions being generally the same.
A general examination of analytical and control instrumentation was made with the
emphasis being placed on response time and operating capabilities. The significant result
of this study was that the current state of the art in control instrumentation and eq uipment is
generally compatible with requirements for a process as complicated as the Catalytic Chamber
Process. Specific procedures were not developed, but it was felt that there were no problems
that could not have been met with current technology.
Discussions were held with Consolidated Edison, TVA, and the New England Power Co.
to ascertain the possibilities of building a pilot plant in conjunction with one of their existing
power stations. All three expressed interest in pursuing the matter further when the process
had achieved a more advanced state of development.
Based on the experimentation and analysis performed during this contract, it would
seem that the Catalytic Chamber Process at its current level of development is technologically
6
-------
feasible, but economically impractical. Control levels of 90% removal of SCL (from 3000 ppm)
and 100 ppm of NO can be achieved, but at $125/KW capital cost and 5.5 mils/KWhr operating
X
cost, both at least twice the economic goal. The only future for the process is to improve the
operation of the three primary process stages or to develop more advantageous techniques.
Recommendations for future work were therefore as follows:
1. Discontinue pre-pilot level experimentation in favor of laboratory bench
scale testing.
2. Continue the evaluation of the liquid phase SO, oxidation to lower the resi-
dence time required for 90% reaction.
3. Re-examine low temperature scrubbing as in the original Lead Chamber
Process.
4. Continue evaluation of activated charcoal in the stripper/oxidizer and
expand evaluation of previously studied candidate catalysts.
-------
II. INTRODUCTION
A. Content of this Report
This document reports the work performed by Tyco Laboratories, Inc., and its sub-
contractor, the Badger Corp., during the one-year period February 1971 to February 1972
concerning the development of the Tyco Catalytic Chamber Process for the removal of sulfur
dioxide and nitrogen oxides from power plant flue gas. The work was performed under Con-
tract No. 68-02-0008 with the Office of Air Programs (OAP) of the Environmental Protection
Agency (EPA). The report will include the experimental and conceptual work accomplished
for the evaluation of the engineering feasibility of the process, as well as consideration of the
economics involved in the full-scale plant construction and operation.
B. Initial Process Concept
At the start of the contract period covered by this report in February 1971, the Tyco
Catalytic Chamber Process had been conceptualized as shown in Fig. 1. The following discus-
sion describes the process as it was pictured at the start of the program and was based on ex-
perimental evidence from previous Contract CPA 70-59, as will be given in Part C of this sec-
tion.
Wat flue gas is mixed with a recycle stream of nitrogen oxides. The mixture under-
goes a gas phase reaction in a reactor stage, with the sulfur dioxide being oxidized by nitrogen
dioxide (NO2) to sulfur trioxide which is immediately hydrolyzed to sulfuric acid:
SO2 4 N02 - NO + SO3 (1)
S03 + H20 - H2S04 (2)
The gases containing sulfuric acid and mixed oxides of nitrogen are then passed into a high
temperature scrubber where both the product acid and the nitrogen oxides are removed by
countercurrent contact with hot sulfuric acid. By operating the scrubber at such a temperature
that the vapor pressure of water over the incoming sulfuric acid scrubbing solution is equal to
the partial pressure of water in the incoming reacted flue gas (taking into account the small
amount of water needed to convert the SOg into 80% sulfuric acid), the sulfuric acid and the
oxides of nitrogen could be scrubbed out without removing any water from the gas. This pre-
vents the dilution of the scrubbing solution.
-------
It has been previously established during Contract Nos. PH-86-68 and CPA 70-59 that
neither nitric oxide (NO) nor nitrogen dioxide are very soluble in sulfuric acid, but can be
readily absorbed as nitrosylsulfuric acid (HNSO,.) in the acid if the two oxides are present in
equimolar proportions:
NO 4 NO2 + 2H2SO4 - 2HNSOg + HgO (3)
Thus, the recycle stream of nitrogen oxides has to be adjusted so that the gas leaving the re-
actor stage will have equimolar quantities of NO and NO0.
&
This technique for the operation of the scrubber stage permits the emission of a gas to
the atmosphere, meeting projected standards for SO« and oxides of nitrogen. The high tempera-
ture operation produces a gas which is buoyant enough to be dispersed in the atmosphere without
further reheat.
The nitrated sulfuric acid leaving the scrubber is passed to a catalytic stripper, which
simultaneously oxidizes the oxides of nitrogen which have been reduced in the SCL oxidation
reactor and also permits the complete denitration of the acid without the addition of heat or the
dilution of the acid. This is accomplished by the countercurrent contact of the nitrated acid
with air on an activated charcoal catalyst:
4HNSOg + 2H2O + O2 - 4NO2 + 4H2SO4 (4)
When the reduced portion of the oxides of nitrogen are oxidized back to NOg, the equimolar
balance of NO and N0« is upset and the nitrogen oxides are spontaneously evolved. The deni-
trated acid can then be recycled to the scrubber after an amount of product sulfuric acid equiv-
alent to the amount of SO2 oxidized in the reactor has been removed.
The gas evolved in the catalytic stripper is then available for recycle back to the ente
ing flue gas to continue the removal of additional SOg. However, if the scrubber /stripper loop
operates at high enough efficiency levels, at this stage there will be excess N0x in the stripper
effluent due to the recovery of additional NO from the flue gas. Therefore, a portion of the
A
stripper effluent is passed through a nitric acid absorber, thereby producing additional by-
product acid. The off-gas from the absorber and the remaining effluent from the stripper are
then recycled to the flue gas stream.
As given here, the primary advantages of the process are:
1. Both SO0 and NO are removed from dilute flue gas while producing sulfuric
A X
and nitric acid as salable byproducts.
2. No additional heat is necessary to operate the process.
3. The process could be added onto existing power stations with only minor
modification of existing plant equipment.
C. Prior Work
Previous to this contract, Tyco had developed the process under two contracts with
the Department of Health, Education and Welfare: No. PH 86-68-75 and No. CPA 70-59. De-
10
-------
tails of the work performed can be found in the final reports for these two contracts, and only
a summary of the findings will be included here.
Experimentation showed that the original Lead Chamber Process was not adequate for
cleaning flue gas because:
1. The required reoxidation of NO simultaneous with SCL oxidation by less than
stoichiometric quantities of NO, in hot, dilute flue gas was too slow to permit reasonable resi-
dence volumes.
2. Scrubbing wet flue gas with cold sulfuric acid caused the dilution of the
acid, thus requiring additional process heat to effect concentration of the acid.
3. Stripping the cold nitrose required additional process heat.
4. The process required approximately 40% more fuel than the power plant
was burning to make up the heat required in (2) and (3) above.
These disadvantages were overcome with the development of the Catalytic Chamber
Process which showed that:
1. Scrubbing of NO could be effected at high temperatures (in the 250 °F
X
range) and thus avoid picking up water and diluting the scrubbing acid. Experiments performed
in a 4-in. tower packed with 3/8-in. Intalox saddles showed 65% removal of NO from 7000 ppm
concentrations in 8.5 ft of column.
2. Nitrose (nitrosylsulfuric acid in sulfuric acid) could be denitrated by contact
with oxygen from air and activated charcoal. In experiments performed at 260 to 290 °F, a
99.7% removal of nitrosylsulfuric acid from nitrose solution (containing 0.7% HNSOg) was
achieved in 22 in. of charcoal packing.
3. Reactor experiments showed that about 95% conversion of S02 to SOg could
be achieved at 300 °F and atmospheric pressure, starting with 3000 ppm of S02 and about 6000
ppm of NO, with a residence time of 15 sec.
4. SO0 oxidation and NO reoxidation could be effected separately without the
£t
need for additional process heat.
D. Contract Goals
The goals of Contract No. 68-02-0008 were to (1) demonstrate process viability by
continued operation in an enlarged and integrated 10 SCFM miniplant, including removal of
SO,, and NO from simulated flue gas to meet 300 ppm SO2 and 150 ppm NOx specifications in
exit gas, (2) establish process and engineering data for scaleup to a 2000 to 5000 SCFM pilot
plant level along with potential site locations, and (3) update process economics.
More specifically, the work was to include the following tasks:
1. Develop engineering data for the SO,, oxidation reactor design. The reactor
was to be run separately from the rest of the miniplant to include optimization of reaction,
temperature, and residence time to achieve 90% conversion yield starting with 3000 ppm SO,, in
the incoming gas.
11
-------
2. Develop design criteria for the NO scrubber and catalytic stripper. The
A
scrubber and stripper were to be operated without the reactor to establish the following operat-
ing parameters in a statistically designed parametric study:
a. Scrubber height necessary to reduce the 7000 ppm NO level to 150 pprn
(also the height of a theoretical transfer unit). [Note: Approximately 600 ppm NO are present
X
in the initial flue gas from the power plant furnace. The remainder results from recycle of
the NO, to the reactor.]
b. Temperature range for scrubber operation
c. L/G for scrubber
d. L/G for stripper
e. Stripper catalyst bed depth
f. Stripper gas and acid space velocities
g. Stripper operating temperature
h. Excess air to stripper.
3. Operation of the integrated miniplant, using a statistically-designed expert^
mental study of major process operating variables. Of particular importance is process con-
trol and process sensitivity near optimum conditions, corrosion tests, and the effect of fly ash
contamination through simulated study.
4. Development of process control criteria. A preliminary evaluation of pro-
cess control considerations was to be made based on the original process, as given in Section
II. Such items as the availability of in-line analytical instrumentation, techniques of fluid flo\v
and concentration adjustment, control equipment, response time requirements, and power plar^
operating procedures were to be determined. A process control system would be recommended
5. Selection of a site for subsequent 2000 to 5000 SCFM pilot plant installation.
Discussions were to be held with utility companies such as Consolidated Edison Company of
New York, the Tennessee Valley Authority, and other interested utilities to examine the pos-
sibility for locating a pilot plant on the site of an existing power station.
6. Prepare a preliminary design for a pilot plant.
In conjunction with a subcontractor, the Badger Corp., a preliminary pilot plant was
to be designed based on the engineering criteria developed in the experimental phases of the
contract. The plant design would hopefully be tied to a specific location and a specific stack
gas.
12
-------
III. ENGINEERING SUBCONTRACTOR — THE BADGER CORPORATION
The Badger Corporation was engaged as subcontractor to Tyco for the purpose of pro-
viding additional expertise in the areas of engineering and design. Badger's role in the contract
was designed to be in three areas: (a) general engineering consultation, (b) design of a com-
puter program for process optimization, and (c) preliminary design of a 2000 SCFM pilot plant.
A. Engineering Consultation
Daring the early stages of the contract, Badger's primary function was to evaluate the
process from a plant design viewpoint and help in the design of experimentation which would
help develop the data needed for the preliminary pilot plant design. Several meetings were
held between Tyco and Badger personnel to exchange information and develop ideas for ex-
perimentation. In this area, Badger made an initial evaluation of the process and made general
suggestions given in Appendix I.
B. Optimization Program
Badger utilized their extensive experience in computer optimization of process vari-
ables in chemical plants to develop a program which was intended to help in the analysis of the
experimental data. Based on the process conceptualization given in Part I of this report, a
program was written around process stoichiometry which would permit optimization of the
operating variables. As new experimental data was reported, the process concept was revised
and the program was appropriately modified. In this manner, the program could be used to
back up the experimental work and keep the laboratory and miniplant work aimed in the proper
direction. As the contract orientation shifted from plant design to reevaluation of process con-
cepts, the optimization work was deemphasized. At the completion of the contract, Badger had
a partially modified program which would require several changes to be useful for further
optimization.
C. Pilot Plant Design
During the early stages of the contract, Badger's major contribution in the pilot plant
design area was to provide assistance in developing the process control system that could be
used with this type of process. This work will be discussed in some detail in Section VIII of
this report. During the latter stages of the contract, it became clear that the process concept
was not sufficiently developed to permit pilot plant design, and this phase of the work was
discontinued.
13
-------
IV. MINIPLANT MODIFICATION AND EQUIPMENT
A. Existing Equipment
At the start of this contract, the all-glass miniplant used in Contract No. CPA 70-59
was operable, but needed extensive modification to permit complete evaluation of the process
variables. The previous contract had established revised operating parameter ranges for im-
portant variables, and it was therefore necessary to change the capacity of certain process
stages to accomplish suitable integrated operation. The SO, oxidation reactor, the scrubber
system, and the stripper stage all had to be modified and enlarged, which created the need for
changes in pumps, flowmeters, piping systems, and instrumentation. Fig. 2 shows the flow-
sheet of the miniplant as conceived at the beginning of this contract in February 1971. As the
contract proceeded, several additional modifications were required, and they will be discussed
in the following sections of the report.
B. SO? Oxidation Reactor
The reactor used in the prior contract was a horizontal glass tube, 4 in. in diameter
and about 5 ft long. Experimentation during the previous contract indicated that the oxidation
of SOp to SO, (with subsequent hydration to FLSOJ might not be as simple as the equation
SO? -f NO« = SO, 4 NO would suggest. The reactor in the miniplant was therefore designed so
that this reaction could be examined in more detail than in the past. The reactor was placed in
3 3
a vertical configuration and its size increased from about 0.5 ft to about 1.5 ft , with the option
existing for using either one-third of the new reactor or the whole volume. The first third of
the reactor was set up so that it could be easily packed, if necessary, with some material for
increasing surface area and possibly to spray the gas with a fine acid mist. Later, an option
was made available for a single 15-ft reactor column to be installed in series with a liquid
phase feed system. The spray capability was added to simulate the conditions in the chambers
of the Lead Chamber Process Plant and thus evaluate the possibility that some phase of the
oxidation involved a liquid phase reaction. The reactor was assembled using a 4-in. glass pipe
and jacketed for heating the simulated flue gas. Previous experience indicated that a maximum
temperature of 375 °F could be achieved in the reactor column.
C. Scrubber Columns
As shown in Fig. 2, the initial plant operation was to utilize a single glass scrubber
15
-------
Bypass to
stack,
T
-O-
tUK
Flow meter
NO
Qxidizer
Filter
Gas
Burnei
Natural
gas
TT
. diluent
gas (M9 or
air) £
ID
Acid feed
To stack
T
O
T
T
O
Corrosion samples
(in packing)
Bypass to stack
Flow meters
NO Air
Corrosion
sample
chamber
1
0
T
O
T
T
O
T
•
}
f
O
T
O
T
O
-X-
I
-
r-,1
1
9
O
T
O
T
¥
«
Corrosion
sample
chamber
(nitrose, gas)
S - Sample port
T - Temperature probe
Corrosion
sample
chamber
(denitrated acid)
Compressed air
Air
Fig. 2. Schematic flowsheet of miniplant (strippers in parallel configuration)
-------
column containing about 7.5 ft of packing (3/8-in. Intalox saddles). This column was similar
to the one used in the prior contract, and it was known that scrubbing efficiencies in excess
of 65% could not be achieved in a column of this height. During the contract period, it was
therefore decided to enlarge the scrubbing system by adding an additional glass column con-
taining 19 ft of packing. A sketch of this column is shown in Fig. 3. All scrubbing experimenta-
tion was performed using the two scrubbers (7.5-ft and 19-ft sections) in series.
D. Stripper Columns
In order to achieve stripper space velocities equal to those used in the laboratory ex-
perimentation performed in Contract CPA 70-59, the stripper stage was redesigned to consist
of three 4 ft high x 12-in. diameter columns operated in parallel (see Fig. 2). The catalyst
was Witco Type 256 activated charcoal 4 x 10 mesh.
As initially designed, the stripper columns consisted of Teflon-coated, stainless steel
tubes as shown in Fig. 4. The tubes had gas feed ports at two levels and four ports for thermo-
well probes on the sides. There were two openings in the top cover (acid feed and gas exit)
and an acid drain in the bottom. Ports which connected to transfer lines were flanged using
standard 150 lb ASA dimensions for easy connection to 1-in. glass pipe. All surfaces which
were to come into contact with acid were coated with Teflon.
Inside the tubes provisions were made to support four beds of activated charcoal.
Three-pronged spiders, resting on half-inch studs welded to the inside wall at four levels,
provided support for the center sag of the fiberglas fabric-wrapped ring which was the actual
support plate for the charcoal. A sketch of this arrangement is also shown in Fig. 4. The three
strippers were each loaded with approximately 25 lb of charcoal to a depth of about 35 in.
During the operation of the system, it became obvious that the coating was not adhering
properly to the surface of the column and the metal was corroding severely (note Appendix VI).
The three columns were cleaned and recoated with another Teflon formulation recommended by
duPont, but this coating eventually failed also. Rather than continue to experiment with other
coatings, it was decided to redesign the strippers to use an all-glass system with solid Teflon
gaskets and covers. A sketch of this new design is shown in Fig. 5.
This design made use of a glass packing support plate perforated with half-inch holes.
To prevent the activated carbon from falling through the holes, the support plate was covered
with a thin Teflon sheet perforated with quarter-inch holes. A layer of 3/8-in. saddles was
placed over the Teflon plate and the carbon placed on top of the saddles. During several weeks
of operation there was no apparent loss of carbon particles (which were 4 x 10 mesh). This
support design avoided a problem that arose with the fiberglas fabric support configuration. In
the earlier design the fine holes of the fiberglas fabric created such a high pressure drop that
the air pressure required to force air up the column prevented the uniform flow of acid down
the column. The new design had no such high pressure drop and uniform flow was easily
achieved.
17
-------
Flowmerer
li cater
u=
•*—!•
fr-
Thermo-
well
l|J: — Thermowell
Support place and
redistributor
lixit gas
Thermo well
Support plate
and redistributor
19' Scrubber
Support plate and
' redistributor
Support plate and
redistributor
£31
Reservoir
Immersion heater
Fig. 3. Sketch of 19-ft scrubber
18
-------
acid gas
feed exit
J'LJlL
24".
3=__Thermowell ports
gas feed-
ports
Bed Support Structure:
Ring with "Y" spider
covered with fiberglass
fabric
Fig. 4. Miniplant stripper column and catalyst bed support structure
19
-------
Feed acid
Effluent gas
Teflon plate
Thermowells
(equidistant from
the center)
12-in. x 48-in. glass pipe
Packing support plate
(glass)
Feed air
Acid drain
Fig. 5. Stripper column using glass pipe
20
-------
E. Liquid Pumps
During earlier work on the miniplant in Contracts PH 86-68-75 and CPA 70-59, the
acid had been pumped through the system using piston pumps of standard design (Clark-Cooper
Mark II Metering Pump), with all wetted parts being made of Carpenter 20 alloy. These had
required considerable maintenance, and new pumps were ordered for this contract.
All Teflon pumps (Fluorocarbon Co. Saturn Pump) utilizing a nutating disc design
were used during the early stages of the contract, but these too were unsatisfactory. The
major problem with this design was the softening of the Teflon at high temperatures which
caused the internal seal to leak, thus reducing flow to a trickle. These pumps also had a tend-
ency to foul easily, and small particles prevented a good seal between the nutating disk, thus
sharply reducing its flow.
The Saturn Pumps were replaced with ECO gearpumps utilizing Hastelloy B construc-
tion to prevent corrosion. Estimates by both Tyco and ECO indicated that Hastelloy B was a
marginal material of construction for long term use, but should be adequate for the duration
of the contract (about 50 mils/yr corrosion level). In actual use, however, excessive aeration
of the acid combined with the high temperature, high speed operation of the pump caused ero-
sion of the pump body. This resulted in excessive clearance around the gears which reduced
the acid flow to very low levels.
The pumps that eventually proved adequate to the task of pumping 80% sulfuric acid at
temperatures over 200 °F were simple peristaltic pumps manufactured by the Ober Co. for
Cole-Parmer. The Masterflex Pump using Viton tubing could be used for up to about 36 hr
before tubing replacement was required. This meant that complete runs could be made without
interruption for pump maintenance, and even if tubing replacement was required, the change-
over only took a few minutes and thus avoided long periods of down time. This certainly was
not a long term answer to the pumping problem for either the miniplant or the larger pilot
plant, but was certainly adequate for the short term requirements of this contract.
It was interesting to note that on a larger scale, pumps are readily available that can
handle the operating conditions of this system. The Durco Co. manufactures duririon centrif-
ugal pumps in sizes which can handle pumping rates of 5 to 3000 gpm. However, the centrif-
ugal pumps could not easily handle the 0.1 to 0.8 gpm required in the miniplant, and smaller
pumps made of duririon are not available.
F. Gas Pumps
The miniplant had two requirements for pumping of gases: transferring the raw simu-
lated flue gas, and pulling a gas sample from various points in the process gas loop. The first
requirement had been met in the original miniplant design by using a blower manufactured by
MGD Pneumatics Co. This blower easily provided up to 12 SCFM of raw flue gas.
The sample pump was a far more difficult problem because of the combination of high
temperature, high water content, and the corrosive nature of the gas. After trying several
pumps which proved inadequate, a low cost diaphragm pump made by Bellfc Gossett (Fluid
21
-------
Handling Div., I.T.T.) was chosen. This pump required relatively frequent maintenance, but
the low cost of the unit permitted the use of several spares which could be kept ready for rapid
installation while the fouled pump was cleaned. By carefully trapping condensed liquid in the
gas lines, the pump could be operated for several days without problems.
G. Instrumentation
Instrumentation in the miniplant involved three areas: flow rate determination, tem-
perature measurement, and in-line gas analysis. No automatic control equipment was used.
Fig. 6 shows the miniplant after final modifications and sample point installation.
1. Flow rate determination
All liquid and gas flow rates were measured using in-line glass rotameters as
shown in Fig. 6. Liquid levels in reservoirs and columns were controlled by visual inspection
and throttle valve adjustment by the plant operators.
2. Temperature measurement
Over forty temperature measurement points were utilized to help monitor and
control the process. Two Esterline Angus twenty-four point recorders were therefore instaHe-j
to continuously monitor the temperature levels, with the data being extracted from the recorder
printout by the plant operator and recorded permanently on data sheets.
3. Gas analysis
At several points in the system, the process gas lines were monitored for con-
centration of NO0, SO0, and NO. This was accomplished for the first two gases by analysis
a ft
with a duPont Model 400 ultraviolet photometer. Nitric oxide was monitored with a Beckman
Model 315A infrared analyzer. In addition, the water content of the gas was periodically moni-
tored with a EG&G Model 2000 dew point hygrometer.
H. Gas Feed
The flue gas feed was simulated by doping the effluent of a natural gas burner with
fur dioxide as shown in Fig. 6. The exhaust gas from the burner was diluted with air to
the water content of the flue gas fed to the system so that the feed stream contained about 6%
water, 15% oxygen, plus CO2 and nitrogen. Sulfur dioxide was metered into the controlled ex-
haust gas stream to create any desired concentration.
Nitrogen dioxide was fed to the system by oxidizing nitric oxide with air. NO, is a
liquid below 22 °C and is very hard to handle around room temperature. Previous experience
had shown that the gas condensed in flow meters and tubing creating a very difficult cleanup
problem. In order to avoid this problem a large empty reactor chamber was used to oxidize
the NO to NO« before it was fed to the process. Yields of 75 to 85% were obtainable in this
manner. See Appendix X for a discussion of air oxidation of NO for this application.
22
-------
to
co
OQ
0>
•
2
a
M*
K
a
3
H*«
3
3
ET
1
Rescr
Bypass
me
Blowe
Gas
Burner
Air
Gas sample
L
Reactor
sample
Flow
_ meter
&S-ffip
1
pir |^\J ^
1 n Flow
j U meter
fi fH
LJtank
T
HNSO,
Concent
NO C
Oxidizei
JL
rat
1
i
ion
t
^esdrv
r
]Flo
met
V
lir
V
sr
^
Gi
19'
Scr
r
i
is sample
f
ll— •
JJ.
b
lo
Q
R
n
J
er
?r
ah
se
*.A
1
•vcU
1
cid
To
-»(
s
1
E
2)1 m
sam]
LValve
stack
3as
;amp
I'lo\
1CD
Aclc
Sam
t
j
ale
— P
R\
i
le
t
r
Bypas
ito sta
. \ \ - j
T T Valvet
itijippef:^
. LJ 1 1
i i- * *
j
iplcs
test
, cb Filter
1 ^"JF — "^ Stripper
U oypass
I\imp
.ow control
ma ft s
DFlow
meter
Natural
gas
Pump
:ers
Air
-------
I. Miniplant Operation
The approach taken during the contract was to operate the miniplant in separate stages
leading up to integrated operation. First, the reactor stage was fully evaluated and then the
scrubber/stripper loop was investigated. The emphasis in plant design was on simplification
of operation to aid in the rapid accumulation of data. This was particularly pertinent since the
miniplant was operated by two men per shift. Details of the miniplant operation, including
data-taking procedures, gas sampling techniques, and mode of shift operation are presented
in Appendix II.
24
-------
V. REACTOR EXPERIMENTS
A. Background
Experimentation on Contract CPA 70-59 indicated that the gas phase oxidation of SO_
Ct
with NO0 occurred very rapidly at 300 °F such that at a concentration of 3000 ppm SO0, about
ft Ci
95% would be converted to H,SO. in less than 20-sec residence time. These results were
somewhat erratic, but several runs were made that gave consistent yields at this level. The
only reason for questioning the data was the fact that the residence time of the gas in the sam-
ple lines and analyzer was about 15 of the total 20-sec contact time. To verify this earlier
data, a series of experiments was run in the miniplant using a three-reactor column containing
g
about 1.6 ft of volume. The gas sample system was modified so that the gases were in contact
for only about 2 to 4 sec in the sample lines, and the results more closely represented the re-
action occurring in the reactor chamber.
Initial tests at 300 to 350 °F, using gases containing about 3000 ppm of SO, interacting
with NO streams providing about 6000 ppm of NO? and about 2000 ppm of NO, indicated reac-
x ^
tion rates of about 15 to 20%, with reactor residence time of 15 sec and sample line residence
times of about 3 sec. Table I shows some of the data from these initial runs. These low yields
necessitated a more complete evaluation of the reactor system due to the differences between
the experimental system used in the earlier tests and the more recent miniplant and sampling
system design.
Clearly, the most significant of these differences was the shorter sample line residence
time although the total contact time was about the same. In addition, the sample lines were
modified so that heating was more uniform and condensation in the lines was greatly reduced.
Rather than using 1/4-in. glass pipe wrapped with heating tapes, the samples were pumped
through 3/16-in. Teflon tubing which was enclosed in aluminum tubing wrapped with heating
tapes. This approach avoided the flanged sections of pipe which were very hard to wrap with
tapes and permitted continuous lengths of uniformly heated tubing which could be easily cleaned.
To evaluate the effect of these plant modifications as well as examine the operating
parameters of the reactor, a series of miniplant reactor tests were run as discussed below.
B. Reactor Experiments; Effect of UV Radiation
Initially, two experiments were run to determine if UV radiation accelerated the gas
o
phase reaction between SC^ and NC»2. An empty 4-in. glass column containing 0.53 ft of void
25
-------
Table I. Initial Reactor Studies
Reactor temperature (°F)
Gas flow rate (SCFM)
Gas feed: (ppm)
so2
N02
NO
H20 (%)
Effluent gas: (ppm)
so2
N02
NO
Reactor residence time (sec)
Sample line residence time (sec)
Q
Reactor volume (ft )
SO- removal efficiency (%)
Run 4
350
4
3200
6100
1700
-6.0
2750
6050
2200
14.5
~3
1.6
14.1
Run 8
350
4
3050
6000
1400
-6.0
2600
6700
1800
14.5
-3
1.6
14.8
Run 8A
300
4
2950
6000
-6.0
2300
5350
2400
15.5
~3
1.6
18.7
26
-------
volume (6 ft long) was modified so that a UV source emitted radiation down the length of the
tube. Gas containing SO,,, N0_, and NO was fed into one end of the column (the end housing
the UV source), and the gas concentration was monitored before and after passing through.
Two runs were made, as shown in Table II, with the only difference between the two runs being
that in Run 51 the radiation was directed toward the side of the column and was reflected through
the tube by aluminum foil. In Run 51A, the source was aimed along the length of the tube. In
both cases, the radiation passed through a quartz plate which protected the source from the
corrosive gases.
It is clear from the results of the two runs that the UV radiation had no effect on the
oxidation rate of the SO0 by NO0. Although there is a consistent NO material balance error
ft £t X
due to small NO analysis errors, there is no apparent acceleration of the oxidation of S00.
X s*
C. Miniplant Tests: Effect of Temperature on SO, Oxidation Rate
During the early runs, discussed above, there were indications that temperature had a
significant effect on the S02 oxidation rate. To evaluate this effect, the reaction was tested
over a range of temperatures at different flow rates in order to vary the reactor residence
time. At the same time, the effect of water vapor in the gas stream on reaction rate was
evaluated by running the tests with and without the gas burner. It should be noted that the
primary purpose of the gas burner was to introduce water through the oxidation of natural gas.
The gas burner effluent contained about 6% water after dilution. The results of these tests are
shown graphically in Fig. 7.
It had been anticipated that the gas phase oxidation of SO- with NO, proceeded according
to the reaction:
SO2 + N02 - S03 4 NO
and would proceed at a faster rate at higher temperatures. The data in Fig. 7 show that just the
opposite was true: higher temperatures brought about lower oxidation rates. The data also
showed that at lower temperatures the presence of water vapor in the reaction stream was es-
sential to higher reaction rates.
In general, the results were more compatible with a mechanism involving a liquid phase
reaction rather than the expected gas phase oxidation. The increasing reaction rate at lower
temperatures suggested the necessity of dissolving the gas phase reactants in a liquid phase
before the reaction could take place. Comparing the Tyco Process with the original Lead
Chamber Process, it should be noted that one major difference is the absence of liquid addition
in the SO0 reaction stage. Much has been written about the complexity of this reaction, and at
ft
this stage of experimentation it appeared that the liquid phase addition might well be essential
to high reaction rates.
D. Effect of Acid Irrigation on SOg Oxidation Rate
To evaluate this concept, a series of experiments was performed in the miniplant reac-
27
-------
Table II. Gas Phase SO2 Oxidation with N02
in the Presence of UV Radiation
Run 51A
Gas flow rate (SCFM) 2 2
Column temperature (°F) 250 250
Gas feed: (ppm)
S02 2800 3400
NO2 5600 6200
NO 1500 1900
Effluent gas: no UV (ppm)
S02 2000 3100
NO2 8600 9050
NO 1675 3850
Effluent gas: with UV (ppm)
SO2 2050 3100
NO2 8500 9050
NO 1775 3850
SO2 removal without UV (%) 28.6 8.9
SO2 removal with UV (%) 26.8 8.9
Notes: 1. Run 51A was run under identical conditions to
Run 51 except that the UV source was aimed
directly down the reactor column instead of
into the wall.
2. Reaction run in empty glass tube 4 in. x 72 in.
(0.53 ft3).
28
-------
100 -
90 -
80
70
60
50
40
30
20
10
Run 8D
2 SCFM
< 1% H20 in flue gas
Run 8A
4 SCFM
-6% H2O
in flue gas
Reactor Residence Time
at 4 SCFM
°F Sec
150
200
250
300
350
19.3
17.9
16.6
15.5
14.5
Three reactors in series total
volume: 1.6 ft3 (first 0.3 fir
contained 3/8 in. Intalox saddles)
Feed: S02 - 3000 ppm
NO2 - 6000 ppm
NO - 1500 ppm
Run 8B
2 SCFM
~6% H20 in flue gas
100
200
300
400
Temperature, °F
Fig. 7. Miniplant reactor experiments - SC>2 oxidation. Effect of water
vapor in flue gas, residence time, and temperature
29
-------
tor which was modified to permit irrigation of part of the chamber with a liquid phase. Fig. 8
shows the configuration of the reactor during this phase of experimentation. The diagram
2
shows that the gas could be passed through either one reactor chamber (0.64 ft containing
3 3
0.3 ft of 3/8-in. Intalox saddles) or three chambers (1.52 ft total), with only the first reactor
chamber being irrigated. The saddles were placed in this chamber to increase the gas/liquid
contact.
Figs. 9 and 10 show the results of irrigating the first reactor section (containing the
saddles) with 80% sulfuric acid at gas flow rates of 2 and 4 SCFM. It is clear that there is no
apparent advantage to acid irrigation under the conditions shown. Fig. 10 also shows the effect
of irrigating the first reactor section with water rather than acid. Because of the vaporization
of water above 150 °F, it was impossible to continue the experiments above this temperature
but if the effect could be extrapolated to higher temperatures, it would appear that again there
would be little advantage.
During some of the irrigation runs, it had been noticed that, if the acid had been exposed
for a period of time to NO in the gas phase before the SCL was introduced into the flue gas,
initial reaction rate was significantly higher than the eventual steady state reaction level.
led to the concept involving the buildup of dissolved NO in the liquid phase in order to effect
high reaction rates. To test this hypothesis, the column was irrigated with acid for an
period of time, after which the acid flow was stopped. The gas phase reactants were then in-
troduced into the reactor and the effluent concentrations allowed to achieve steady state. Fig<
11 shows the results of this test at 2 SCFM. Clearly, this technique is far superior to either
the dry reaction mode or the continuous acid irrigation.
Examination of Fig. 11 reveals an anomaly which should be considered at this point.
The experiment which involved the continuous irrigation of the reactor with 80% sulfuric acid
exhibited lower reaction rates than the dry interaction, which apparently contradicts the pro-
posed liquid phase reaction mechanism. This can be explained in terms of the actual conditiOtls
in the reactor during most of the experimentation. The tests described up to this point were
usually run for a few hours at the most and one right after the other. The net effect is that the
"dry" condition mentioned above was very rarely the case. Even when running without irrig^.,
tion, it took many hours for the saddles to be completely dried out from either previous runs
or condensation during unirrigated runs. It is therefore quite likely that, during the "2
burner on, no acid" run shown in several of the graphs, there was a significant amount of
present on the saddles which improved the reaction due to buildup of reactant concentration jn
the liquid phase. The extreme of this wetted condition is undoubtedly shown by the top curve j
Fig. 11.
One more series of tests was run to ascertain the advantage of the presence of a liqujd
phase. This involved two runs in which the saddles in the first section of the reactor were
wetted with 80% sulfuric acid before the reaction was initiated. In one of the two tests, the
gases were passed into the other two void sections of the reactor before analysis, while in
the second test the effluent gases from the packed section were analyzed immediately. The
30
-------
Reactor Volumes
1-0.64 ft3 (0.3 ft3 of saddles
85% void volume)
r •*
MO oxidizer
Three-way
valve
^1
' _ 2-0.44 ft (empty)
By pass 3 r J
. . 1 *} f\ A A f+ f ft »VM^*^ T \
to stack J-0.44 ft ^empty)
/
M
V.
j
4-1.05 ft (empty)
Liquid feed
Flue gas
feed (con-
S i taining /
|L- v SCU /
M
t I
NO Air
feed feed
/
3/8 in
/
/
/
/
•
}
Gas
= _». sample
rvrrr/lnlprV
!
^\x\
V\\N
P
L-
porcelain
Intalox
r
| ]
Jk Jl
Gas sample
port (outlet)
X T \
^_^
Three- way
valve
saddles
Liquid
drain
Fig. 8. Configuration of miniplant reactor during initial experimentation
involving liquid irrigation
31
-------
100
90
80
70
60
> 50
o
»*
esj
40
30
20
IO
Three reactors in series total
volume: 1.6 ft3 (first 0.3 ft3
contained 3/8 in. Intalox saddles
and was irrigated)
Gas flow: 2 SCFM
Run 8B
No irrigation
-6% H2O in flue gas
Run 8D
No irrigation
< 1% H20 in
_ flue gas
Run 9B
-6% H2O in flue gas
Irrigation of packed
column with 80% H
H0SO. at -30 cc/min
I
100
200
Temperature, °F
300
400
Fig. 9. Miniplant reactor experiments - SO2 oxidation. Effect of reactor
irrigation with 80%
32
-------
100-
90
80
70
60
1
cT
to
50
40
30
20
Run 10
Irrigated with 37 cc/min
Reactors in series
total volume: 1.6
ft3 (first 0.3 ft3
contained 3/8 in.
Intalox saddles and
was irrigated)
Gas flow: 4 SCFM
Water content: ~6%
Run 9A
Irrigation with 27 cc 'min 80%
H2S04
Run 8A
No irrigation
I
100
200 300
Temperature, °F
400
Fig. 10. Miniplant reactor experiment - SC>2 oxidation. Effect of acid or water
irrigation in packed reactor
33
-------
100
90
80
cf
on
70
_ Three reactors in
60
50
40
30
20
10 -
Run 15
Packed reactor wetted with
80% H£SO4 before reaction
started
series total volume:
1.6 ft3 (first 0.3 ft3
of reactor contained
3/8 in. Intalox saddles)
Water content: ~ 6%
H20
~ Gas flow: 2 SCFM
Run 8B
No irrigation
Run 9B
Irrigation with 30 cc/min
~ H2S04
I
I
100
200 300
Temperature, °F
400
Fig. 11. Miniplant reactor experiment - SOg oxidation. Comparison of acid
irrigation, wetted wall tests, and no irrigation
34
-------
results shown in Fig. 12 make it clear that the 24 sec spent by the gas in the empty reactor
sections contributed little or nothing to the reaction which was achieved in the 5 sec in the
wetted saddles. This key test established that, under the existing conditions and gas concen-
trations of interest, the oxidation of SO, is achieved in the liquid phase.
These tests were followed by a series of experiments which attempted to optimize the
liquid phase oxidation of SOg. The first step was to pack all three reactor sections with 3/8-in.
Intalox saddles and provide irrigation to the top of each column. This created a situation where
reactors 1 and 3 were operated with the liquid and gas contacting each other in a cocurrent
manner, while reactor 2 was in a countercurrent configuration (see Fig. 8). Several tests were
run in which operating parameters were varied, including acid flow rate, acid concentration,
and NO concentration (gas phase), but little improvement in reaction rate was achieved.
A
A typical reaction run is shown in Fig. 13. With elapsed time running from right to left,
it can be seen that with no acid flowing it took about 1-1/2 hr for the system to reach steady
state. During this time, the acid on the saddles was apparently building up in reactant concen-
tration until a steady state condition was obtained. During this run, the effect of high oxygen
concentrations in the flue gas was evaluated. This high oxygen content is the result of diluting
the gas burner effluent gas with air in order to reduce the water content of the gas from about
20% to 6 to 8%. Instead of about 3% Og, which is typical of coal burner flue gas, the simulated
flue gas used in the current experimentation contains about 15% O,. To determine the effect of
this high oxygen level, an apparatus was constructed which permitted the dilution of the gas
burner with nitrogen rather than air. Examination of Fig. 13 shows that the SO, content of the
reactor effluent reached a steady state level of about 1050 ppm after approximately 1-1/2 hr.
At this point, the diluent air was replaced with nitrogen and the experiment continued for an-
other hour under these conditions. The nitrogen was then shut off and air returned to the dilu-
tion line. It is obvious that these changes had no effect at all on the oxidation rate of SO,. It is
therefore safe to say that, in general, the results of the experimentation were not affected by
the presence of the high oxygen levels.
E. Instrumentation Considerations
During these last experiments, it was noticed that there was some apparent interference
by the NO2 on the SO2 reading of the duPont UV Photometer. It was known that NO, absorbed
to some degree at 302 mM where the SO, was measured, but this was supposedly balanced out
through the use of a reference beam which analyzed at a wavelength where SO, and NO, ab-
sorbed equally. In addition, there was a balancing potentiometer which permitted the inter-
fering NO9 to be zeroed out, so there should have been no apparent interference by NO, at the
SO, absorbing wavelength. The problem was that, at the higher N09 levels used in some of
2 ^
the tests (up to about 15,000 ppm), this balance was not accounting for the interference.
Discussions with instrument people at duPont revealed that the balance control was ef-
fective at constant levels of N02 in the SOg-bearing gas stream. If the NO, interference was
balanced out at low NO2 concentrations, there might well be very large levels of interference
35
-------
100
90
80
70
60
§
*
O
oo
so
30
20
10
Run 15
Packed reactor
wetted with 80%
H2SO4 before
reaction started
0.3 ft3 packed,
1.3 ft3 unpacked
Run 8B
No irrigation
Gas flow: 2 SCFM
Water content: ~6%
Run 14
Packed reactor wetted
with 80% H2S04 0.3 ft3
packing 0.2 ft3 unpacked
Run 14
Packed reactor
wetted with 80%
H2S04 0.3 ft3
packing 1.3 ft3
unpacked
I
100
200
Temperature, °F
300
400
Fig. 12. Miniplant reactor experiments — SOo oxidation. Comparison of wetted-
packing reactions with and without additional unpacked residence volume
36
-------
3275 ppm NO
Reactant
concentration
_ 3000 ppm
— 1050 ppm
Flue gas diluted with
Gas feed:
1850 ppm NO
2500 ppm SO2
4800 ppm NO2
Reactor temperature: 360 °F 3
Three reactors, packed with ~ 1.25 ft
of 3/8 in. Intalox saddles; no irrigation
2
Elapsed time, hr
0
Fig. 13. Miniplant reactor experiment — 803 oxidation. Evaluation of low O2 level in
flue gas
-------
at high concentrations. DuPont's recommendation was to balance the instrument at zero NO
concentration and run a calibration curve, giving SCL interference at various levels of NO
This was done, and the resulting interference curve is plotted in Fig. 14. The curve was ob-
tained by reading SCL while feeding known concentrations of NO, to the UV cell. The inter-
ference level is approximately 30:1, which is about what was predicted by duPont. (This
means that 30 ppm of NO, shows up as 1 ppm of SO0.)
i i
The high levels of N02 used in this stage of experimentation caused a difficulty in read-
ing NO2 concentrations above 1.5% (15,000 ppm) because of the saturation of the measuring
photocell's ability to read low light levels. DuPont recommended a shorter UV cell to effect
lower total absorbance in the cell. It was decided to replace the 15-in. cell with a 1.5-in. cell
rather than put the two in series. By using a higher gain mode in the photometer amplifier, «t
was possible to use the shorter cell without loss in sensitivity at the gas concentrations of in~
terest (up to 5000 ppm of SO2 and 10,000 ppm of N02 full scale). The instrument could also be
used to measure very low concentrations (500 ppm SO2 and 1000 ppm NO2 full scale), if de-
sired. The advantage of this system is that it is possible to read NCL at both 10,000 ppm ana
100,000 ppm full scale, thus permitting accurate analysis of the lower concentrations in the
scrubber and reactor effluents and the higher concentrations in the stripper offgas.
The short cell was installed and the interference curve run again with the same levels
of interference being observed. All further analysis indicated in this report was made with
the short path UV cell unless otherwise noted. It was decided that, since the interference l
are quite low at the NO0 concentrations being used throughout most of the experimentation
it *
this phenomenon would be ignored unless the SO« concentrations were in the 0 to 500 ppm ran
Reaction yields in future experiments will not take into account the NO, interference unless
specifically noted.
F. Effect of Irrigation of Reactor with Nitrosylsulfuric Acid
It was apparent from the experimentation described above that the concentration of dls_
solved NO in the liquid phase was of considerable importance in determining the SO, oxidati
rate. It was therefore decided to irrigate the reactor with HNSOg solutions in H«SO4 rather
than H«SO. alone.
Before discussing the runs where the reactor columns were irrigated with nitrosylsul-.
furic acid, it will be of value to examine two runs made with 80% sulfuric acid irrigation in
order to gain some insight into the mechanism of the reaction. The operating data from thea
two 250 °F runs are shown in Figs. 15 and 16 and relate to Runs 27 and 28.
In Run 27, the three reactor columns are irrigated for some time with 80% sulfuric a *j
at the rather slow rate of about 30 cc/min for all three reactors (i.e., about 10 cc/min each)
While the acid continued to drip into the reactors, the SOg input level was set at 3100 ppm
(time zero). The SO2 valve settings were fixed so that we could return to the same feed rate
and the feed bottle was turned off.
For the next 1.5 hr, NO was fed to the columns as attempts were made to fix the de-
sired NO and N02 feed rates. Acid continued to drip during this time. At the 1.75 hr mark Of
38
-------
a
a
*
I
4—>
rt
PH
t->
CM
8
600
500
400
300
200
100
o o
8
10
12
NO2 concentration, 10 ppm
Fig. 14. NC>2 interference calibration curve
14
39
-------
Adjustment of NOX
feed. Acid irrigating
column. No SO2 feed
Reaction temperature: 250 ° F
Acid irrigation: 30 cc/min
- 5400 ppm
N02 feed
2700 ppm
NO feed
1000 ppm
Reaction yield'
53.2%
Add SO2 to
NOX gas
stream
SO.
T
NO
SO, feed
3100
ppm
Reactant
concentration
ELAPSED TIME (hr)
Fig. 15. Miniplant reactor experiment - SO2 oxidation. Tracing of instrumentation
recorder input (Run 27: evaluation of acid irrigation)
40
-------
4100 pp
Reactor temperature: 250°
Irrigation with 20 cc/min
80% HSO. (intermittent)
Reaction yield 52.2%
1700 ppm SO2
Saddles wetted
with acid
3550 ppm
1850 ppm
4750 ppm
3200 ppm NO2
and NO2
feed
R cacti) nt
concentration
ELAPSED TIME (hr)
Fig. 16. Miniplant reactor experiment - S02 oxidation. Tracing of instrumentation
recorder output (Run 28: evaluation of intermittent acid irrigation)
41
-------
this run, the SCL was injected into the gas stream at the previous level and the gas concentra-
tions were monitored for the next hours. The SO2 level quickly leveled out and showed a reac-
tion yield of 53.2%, based on the 3100 ppm input. When the S02 was introduced into the system
the NO level was monitored by cutting the gain in half, which accounts for the apparently low
reading on the chart. The NO,, level in the reactor effluent gas did not change much during the
run.
At the end of the hour of testing, the SOg was turned off and the NO content measured
at the reactor input side. It is highly significant that the NO feed rate was about 3700 ppm
total while the reactor effluent gas during the run exhibited about 10,000 ppm of total NO
X
The reasonable explanation for this behavior is that the acid on the column (which is
not moving very rapidly at 10 cc/min for each column) absorbed a great deal of NO during
the 1-1/2 hr of NO flow rate manipulation, and when the SO« was added, it rapidly reacted
with the NO in solution and quickly reached a "steady state" reaction level. During this reac-
tion period, the oxidation of the SO™ caused the liberation of NO due to the reduction of the
NO0 values in the nitrose. The NO observed in the gas during the reaction far exceeded the
A X
input NO due to the reaction.
X
This hypothesis is further verified by the system shown in Fig. 16, Run 28. After the
saddles were wetted with 80% sulfuric acid prior to time zero, the NO and S00 were intro-
X £i
duced into the system almost simultaneously with no acid running. The SO« saw only gas
phase NO and virtually pure sulfuric acid at the start of the run. With time, however, the
X
NO was gradually absorbed into the acid on the saddles, and the SO, reacted with this NO
at a gradually increasing rate. When the acid was added to the system on an intermittent
(e.g., at the 20-min mark in the test), the reaction rate seemed to level out, but then
to increase when the acid was allowed to sit on the column and absorb NO .
A.
When the gas concentrations leveled out around the 3-hr mark, it could be seen that th
effluent NO concentration in the gas phase was about the same as the input NO level (6600
X X
ppm feed NO compared to 6300 ppm in the reactor effluent at steady state). This indicated
X
that the NO was going into the liquid phase to enter into the reaction at the same rate at whiCK
it was leaving and a real steady state had been achieved. If Run 27 had been continued for
several more hours, it is likely that the same steady state conditions would have been achie\»
It is interesting to note that the reaction yield at steady state in Run 28 (52.5%) was about the
same yield as in the earlier run (27).
Examination of these two runs makes it clear that it is not the NO in the gas phase th
is important to the SO« reaction rate, but the amount of NOx in solution. It seemed obvious at
this point to irrigate with HNSOg solutions to introduce the NOx with the acid rather than
for it to be absorbed in the pure acid from the gas phase.
Two runs (29 and 30) were made to evaluate the effect of interacting flue gas with
solutions rather than sulfuric acid solutions. The nitrose solution used contained 4.4% HNSQ
in 70% sulfuric acid. The data from these runs are shown graphically in Fig. 17 (original
from all reactor experiments are in Appendix III) in comparison to data from acid-irrigated
42
-------
100
_ Run 15'
Wetted column
90
80
70
60
g 50
0)
M
O 40
**»-'
30
20
10
Runs 29 and 30
Continuous irrigation with
4.4% HNSO5 in 70% H2SO4
at 20 cc/min total flow;
three packed reactors
(~ 1.2 ft3)
Run 27
Continuous irrigation
with 80% H2SO4 at
~ 25 cc/min total flow
in three packed reactors
(~ 1.2 ft3)
Run 9B
Continuous irrigation with
80% H2SO4 at ~ 30 cc/min
in one tracked reactor
(0.3 ft3)
100
200
Temperature, °F
300
400
Fig. 17. Miniplant reactor experiment — SO2 oxidation. Comparison of acid
and nitrose irrigation
43
-------
runs. It is clear that there is a definite advantage to irrigating with nitrose as considerably
higher reaction yields are obtainable. Most of the data available with continuous acid irrigation
were taken with the partially packed reactor (Run 9B in Fig. 17) and cannot be directly compared
to the data from Runs 29 and 30, which were made using three completely packed reactors (ton
solid curve in Fig. 17); but the data from Run 27 can be used as a reference point since it
too, was made using all three packed reactor columns (single point in Fig. 17).
These data show a decided advantage to using nitrose solutions rather than dry reactor
or pure acid irrigation, although comparison of the new data with earlier runs shows up an an-
parent inconsistency. If the data from Run 15 (Figs. 12 and 17) is compared with Runs 29 and
30, it would appear that continuous nitrose irrigation is not the only way to achieve high reac-
tion yields. It should be recalled, however, that Run 15 was made under essentially nonstead
state conditions. The single packed column was wetted with acid and then allowed to interact
with the gas feed. The data points reproduced in Fig. 17 were the reaction yields under rela-
tively constant conditions, but there is no doubt that after a considerable period of time the
acid on the column would be removed and the efficiency of the reaction decreased. During
time that the data in Run 15 were taken, there must have been a buildup in the concentration
HNSOg in the acid remaining on the saddles, and this permitted the more rapid oxidation of
Experiments were run to evaluate the effect on reaction rate of HNSO. concentration t
the irrigation acid while maintaining constant HgSO. concentration and constant reactor resi-
dence time. The results are shown in Figs. 18, 19, and 20. It is quite evident from the grar»H
that increasing the HNSOg concentration causes an increase in reaction rate.
The effect of varying the H-SCK strength while maintaining all other conditions consta
is shown in Fig. 21. As can be noted, decreasing the acid concentration causes an increase i
reaction rate within the range of conditions studied.
Fig. 22 shows that the reaction yield increases with increasing reactor residence tinie
certainly a predictable result. In this figure, as in Figs. 18 through 21, the data clearly sho\i.
that increasing the acid flow rate causes an increase in reaction rate. This has a bearing on
the mechanism of the reaction which will be discussed later.
The data from Figs. 18 through 22 can be combined as in Fig. 23 to give some indicati
of the overall direction which must be taken to obtain high reaction yields. In this graph, acin
flow rate was plotted against HNSCX concentration showing lines of constant reaction yield,
constant acid concentration, and constant residence time. This method of presenting the data
shows the trade-off between irrigation rate and HNSOg concentration to achieve the same rea
tion yield. It is highly significant that below a certain HNS(X concentration, the yield cannot
be improved regardless of how much acid is used to irrigate the reactor. Conversely, below
a certain irrigation rate, the yield cannot be improved even though the HNSOg concentration
is greatly increased.
Fig. 24 points out that, although the irrigated column is highly effective in developing
high reaction rates, the wetted columns also are quite effective in causing the oxidation and
44
-------
100
80
3 60
QJ
_o
4-t
u
0)
CJ
CM
O
CO
J
40
20
4.3% HNSO,
1.1% HNSO,
70% H2S04
2 SCFM
250 °F
3 reactors in series
1
IOO 200 300 40O 500
Acid flow, cc/min
Fig. 18. Miniplant reactor experiment. Effect of HNSOs concentration on
SO, removal yield
45
-------
100
80
60
I
^j
O
§
C/3
40
20
I
I
I
1.0% HNSCL
0.53% HNSCL
70% H2S04
4 SCFM
250 °F
3 reactors in series
100 200 300
Acid flow, cc/min
400
500
Fig. 19. Miniplant reactor experiment. Effect of HNSO2 concentration on SO2
removal yield
46
-------
100
80
60
c
"- 40
CM
20
HNSOC
1.26% HNSO,
2 SCFM
250 °F
1
1
1
100 200 300 400
Acid flow, cc/min
500
Fig. 20. Miniplant reactor experiment. Effect of HNSO2 concentration on SC>2
removal yield
47
-------
100
80
3 60
-------
100
cf
C/D
250 °F
3 reactors in series
1
1
100 200 300 400
Acid flow, cc/min
500
Fig. 22. Miniplant reactor experiment. Effect of residence time on SO2
removal yield
-------
4.0
_o
g
c
0)
I
C/3
z
2.0
1.0
70% H2SO4
75% reaction yield
2 SCFM
• 70% H2S04
90% reaction yield
2 SCFM
3 reactors in series
250 °F
(numbers refer to specific
runs - see Appendix III)
38A
70% H2S04
75% reaction yield
4 SCFM
44
100 200 300 400
Acid flow, cc/min
500
Fig. 23. Interpretation of miniplant reactor experimentation. Tradeoff between
HNSO2 concentration and irrigation flow rate in achieving a specific
SC>2 reaction yield
50
-------
100
80
2
& 60
s
(H
40
20
After one irrigated reactor and two
unirrigated, wetted reactors
2 SCFM
Run 46
After one irrigated reactor
2 SCFM
Run 46
After one irrigated reactor and two
unirrigated, wetted reactors
4 SCFM
Run 47
V
After one irrigated reactor
4 SCFM
Run 47
Cocurrent operation:
1 irrigated reactor
2 wetted, unirrigated
reactors
250 °F
I
I
100 200 300 400
Acid flow, cc/min
500
Fig. 24. Miniplant reactor experiments. Effect of irrigation and residence
time on SC>2 removal yield
51
-------
removal of SO0. This verifies earlier data and has a bearing on the eventual operating mode
£i
of the reactor.
Four runs were made in one column only, the first two (Runs 46 and 47) being cocurrent
operation while the other two (Runs 48 and 49) being countercurrent. The data on the cocurrent
runs were taken from analyses made at the end of the first column, although the gases were
passed through all three columns. The saddles in the second and third columns were still
wetted with acid from previous runs and undoubtedly entered into the reaction. The gas effluent
from the third column was also monitored. The countercurrent runs were made with one col-
umn only.
Fig. 25 shows the somewhat surprising result that there is essentially little difference
between countercurrent and cocurrent operation. One might expect the countercurrent mode to
be more effective in generating high reaction yields, but this is not the case. A possible ex-
planation for this is the stripping of the HNSOg from the acid in the countercurrent mode befor
it gets a chance to react with the SCL entering the reactor from the other end. This might be
offset by increasing the NO feed in the gas phase, which would tend to reduce the evolution of
gas from the acid.
A series of liquid phase reactor experiments was run in the 15-ft packed miniplant re-
actor (see the data from Runs 52 through 62 in Appendix III). Runs 52 through 59 were run in
a countercurrent mode while 60 through 62 utilized cocurrent operation. The significant result
of this testing was that reaction yields in excess of 90% could be achieved in residence times
under 15 sec. Realizing that additional testing would be eventually required to optimize the
operation of the reactor at the pilot plant level, experimentation was halted at this point be-
cause it was clear that a reactor of practical size could be designed based on the information
already learned.
Looking back at some of the earlier data (see final report to Contract No. CPA 70-5Q)
it is possible to explain some of the anomalous results in terms of new knowledge. When the
plant was operated such that there were long residence times in the sample lines, the reaction
rate apparently increased with time in much the same manner as shown in Run 28 (Fig. 16).
What must have been happening was the condensation of sulfuric acid in the sample lines fol-
lowed by the absorption of NOx in the condensate. The SOg-bearing gas came through the sam-
pie lines and reacted with the nitrose on the walls the same way it reacts with nitrose on the
saddles. Once steady state was reached, an apparently high reaction was achieved which \va
really occurring almost completely in the sample lines.
During the testing described above, these experimental artifacts were avoided. The
lines were well heated and not easily contaminated (and were easily cleaned when there wag
condensation). The residence time in the sample lines was down to about 2 sec because of
high sample flow rates and the low volume of the short UV analysis cell. The instrumentatio
behaved very well and data were both reproducible and reliable.
In looking over this data, it is clear that the nitrose irrigation concept represents a
new approach to the problem of obtaining high SOg oxidation rates. There will be a discusaio
52
-------
100
a fin-
>>
o
(-1
8
Cocurrent
Run 46
Countercurrent
Run 48
2 SCFM
73% H2S0
1% HNSO.
5
250 °F
1 reactor
20-
I
I
100 200 300 400
Acid flow rate, cc/min
500
Fig. 25. Miniplant reactor experiments. Effect of countercurrent versus
cocurrent reactor configuration
53
-------
on reaction mechanisms and an analysis of a proposed flowsheet, including this new concept
later in this report. At this stage in the analysis of the reactor, it is safe to say that a more
complete understanding of the reactor has been achieved, and it should be possible to continue
to develop the process to the point where SO. can be removed from dilute gas streams in a
practical manner.
G. SO, Oxidation Reaction Mechanisms
It was originally assumed that the overall chemical reaction occurring in the reactor
was a gas phase oxidation:
NO2 + SO2 - NO + S03
followed by, or simultaneously occurring with, hydration to form H.SO.. Experimentation
during the early phases of this contract showed that this reaction does occur, but at too slow
a rate to be economically feasible. However, subsequent testing showed that irrigation of the
reactor with sulfuric acid or nitrose solutions gave considerably higher yields at equal resi-
dence times, and the most recent testing indicates 90% yields in less than 15 sec residence
time. In order to more fully understand the process and thereby be able to further optimize
the reactor stage, an in-depth analysis of the liquid phase reaction mechanism was initiated
Two earlier studies, which are of direct relevance to the oxidation of SO. in the liquid
phase by nitrogen oxide species in equilibrium with gas phase NO at 25 °C, are discussed
A
here. The relevance of this work to our results at higher temperatures is clarified.
Berl, et al. report the results of a study entitled, "The mechanism of the oxidation
of SO2 by oxygen in the presence of nitrosylsulfuric acid." The experiment involved the vig-
orous shaking together of 20 ml of 0.1 molar nitrosylsulfuric acid in sulfuric acid (20 to 80 wt
%) with approximately 600 ml of a gas mixture consisting of 1/3 oxygen and 2/3 sulfur dioxid
at approximately 25 °C. The rate of uptake of SO, was measured and found to be constant for
any given sulfuric acid concentration. This rate was plotted versus sulfuric acid concentratio
and found to peak sharply at 57.5 wt % HgSO.. Further studies up to 90 °C indicated similar
results. The rates of SO. conversion were much lower in the absence of a liquid phase.
2
Seel reported the results of a study entitled, "The nitrous acid-nitrosyl ion equilib-
rium in aqueous sulfuric acid." A very careful ultraviolet spectroscopic study of nitrose sol
tions at 25 °C indicated that, whereas nitrous acid is the predominant nitrogen species at sul
furic acid concentrations less that 50 wt % HgSO,, nitrosyl ion (ionized nitrosylsulfuric acim
is the predominant nitrogen species at sulfuric acid concentrations greater than 70 wt % H sr»
In the midrange, nitrous acid exists in equilibrium with hydrated nitrosyl ion, FUNO.4, whj .**
is, of course, the same as protonated nitrous acid.
HNSOg ^ NO+ -I- HSO4~
N0+ + HS04~ + H20 * H2N02+ + HSO4~ *
54
-------
The concentrations of HNOg and HgNOg are equal at 55 wt % HgSCh. At 59 wt % H-SCK, the
conversion to H-NCL is 90% complete. As the sulfuric acid concentration is further increased,
the H9NO« becomes less and less hydrated and thus becomes more correctly described at
N0+ (solvated by H2SO4).
Thus, at 57.5 wt % H2S04:
1. Berl observed the highest rates of conversion of SCL to HLSO,
2. Seel observed that H-NCL is the predominant nitrogen specie in nitrose
solutions.
One may reasonably conclude that H-NO, is the nitrogen specie in nitrose solutions which is
responsible for the rapid conversion of SO, in the lead chamber process and in the Tyco SO»
reactor.
In making a statement concerning the optimum nitrose concentrations for the Tyco SO,
reactor, one must consider the effects of increasing the temperature from about 80 °F to the
range of 200 to 300 °F. At 80 °F, 0.1 molar HNSOg, and 64 wt % H2SO4, the partial pressure
of nitrogen oxides is about 0.02 atm whereas at 200 and 300 °F, the corresponding partial pres-
sures are 0.13 and 0.41 atm. Berl's experiments at room temperature were carried out at
constant HNSO.. concentration whereas our experiments at elevated temperatures are more
D
closely approximated by conditions of constant vol % NO in the gas phase. It is entirely pos-
x 4
sible that, at our operating temperatures, we may obtain the highest concentration of HgNOg
(and thus the highest rate of oxidation of SO,,) at some concentration of sulfuric acid higher
than 57.5 wt %.
We assume that the major portion of the SO0 oxidation occurs in the liquid phase by
+ *
reaction of dissolved SO« and nitrosyl ion, NO . The reaction is assumed to be irreversible,
i.e., the reaction products do not inhibit the oxidation of SO-. For our model we postulate the
following processes and possible slow steps:
(SO2) gas = (S02) dissolved (D
diffusion gradients in gas phase (la)
diffusion gradients in liquid phase (Ib)
NO 4 NO2 4 H2SO4 = 2NO"1" 4 2HS04~ 4 H£O (2)
diffusion gradients in gas phase (2a)
diffusion gradients in liquid phase (2b)
S02 4 N0+ 4 H20 = NO 4 H+ 4 HS03 (3)
HSOg 4 NO+ 4 H20 = NO 4 H2S04 4 H+ (4)
To make the situation mathematically tractable and to allow us to develop a physical model
which closely approximates a packed column, we find it useful to introduce reasonable sim-
plifications into the model. We will verify by direct calculation that our simplifications are
reasonable.
55
-------
The first simplification we propose is that we be able to ignore the diffusion gradient
of SO, in the gas phase. A typical case which we may use for the purposes of our calculation
is 3000 ppm SOg at 150 °C over 80 wt % HnSO,. We may obtain an extrapolated value for the
solubility of SO, in 80 wt % H,SO4 at 150 °C using the data of Kuznetsov. Using the data for
79 wt % H2SO4 and doing a least-squares fit of log (CSQ ) versus 1000/T°K, one obtains statis-
tical constants which, in turn, yield the value of 1.00 cc SO, dissolve per cc of sulfuric acid
solution.
4
Jost gives estimates for diffusion coefficients. We find that for gases,
D =-££
P
where f = 1 approximately, rj = viscosity, and p = density. At STP, D values normally fall in
the range of 0.1 to 1.0 cm /sec. We also find that D is proportional to Tn where n is in the
range 1.5 to 2.0. Thus, at 150 °C we may estimate that D is approximately equal to 1
For liquids of the viscosity we are concerned with, D is approximately equal to 10
To calculate SO, flux we use Pick's first law,
£t
J = -D
In the gas phase we assume complete depletion of SO, and a diffusion layer thickness of 0.01
cm, which results in an SO2 flux of 8.6 x 10~ moles/(cm -sec). In the liquid phase we as-
sume complete depletion of the SO2 and the same diffusion layer thickness which, in turn, re
suits in an SO9 flux of 8.6 x 10~ moles/(cm -sec). It is apparent that the limiting flux
-119
in the liquid phase. Thus, even though the maximum flux of 8.6 x 10 moles/(cm -sec)
might be occurring in the liquid phase, this is only 0.001% of the maximum flux in the gas
phase.
The second simplification we propose is that we be able to ignore step 2 as a rate
ing step. The solubility of nitrosylsulfuric acid (0.1 moles/liter for the case given above)
much greater than the solubility of SO, (8.6 x 10 moles/liter for case given above) for the
conditions prevailing in the reactor. There is, thus, a greater driving force for the formati
of nitrosylsulfuric acid than for dissolved SO9. We have confirmed experimentally that NO
X
uptake is very rapid in the reactor. This allows us to disregard step 2 as being rate limitin
The third simplification we propose is that we eliminate reaction 4 as a rate limiting step. if
either reaction 3 or 4 is rate limiting, one would expect that reaction 3 is slower because of
fact that one does not observe high concentrations of any reaction intermediate, such as
We thus dismiss reaction 4 as being rate limiting. *
Our simplified model now consists of the irreversible bimolecular reaction of one
slightly soluble gas phase species with a dissolved, rather higher concentration liquid phase
species. To evaluate this model for our system, a particular configuration was chosen whj .
is compatible with the experimental work that was done. The system chosen was a packed
column with countercurrent gas/liquid flow. The input SO2 level is assumed to be 3000
56
-------
and the input NO, level to be 3000 ppm at a gas flow rate of 4 SCFM. The acid feed is assumed
to be 80 wt % H2SO4 with 0.0623 M HNSOg (at 150°) present and a liquid flux of 0.17 gpm. The
HNSCL concentration of 0.0623 M was chosen since it is the equilibrium concentration for 3000
5
ppm each of NO and NO, in the gas phase at 150 °C in 80 wt % sulfuric acid, as computed from
the data of Berl and Saenger.
The column utilized was 4 in. in inside diameter, packed with 1/4-in. Intalox saddles.
fi 23
Danckwerts gives an effective area of 1.4 cm /cm for superficial flow rates of 0.2 cm/sec
7
on a 1/2-in. packing. From U.S. Stoneware data, we see that a dry area difference of ~ 1.5-
2 3
fold exists, and thus an effective area of 2.2 cm /cm is expected and will be used in our cal-
culations. (Note: In this region, the effective interfacial area is nearly a linear function of the
superficial liquid flow rate (cm/sec so that we may easily refine our effective area estimate
when we experimentally determine the superficial liquid flow rate).
C
The calculation we wish to make involves the solution of Eq. (5-64) of Danckwerts:
where:
D.K.B0
M = A * ; E = l 4 A* ) : R = KA*D
as:
E = enhancement factor,
E = enhancement factor when rate of reaction is controlled by diffusion,
i
M = dimension less time quantity,
2
DA = diffusivity of dissolved gas A (cm /sec),
A 2
Dn = diffusivity of reactant B (cm /sec),
B
k0 = second order rate constant for reaction of A (i /g mole sec),
2
k = liquid film mass transfer coefficient (cm/sec),
A* = concentration of dissolved gas A at interface, in equilibrium
with gas at interface (g moles/cm3),
3
B° = concentration of B in bulk of liquid (g mole/cm ),
Z = number of moles reacting with each mole of A, and
f{ = average rate of absorption over contact time t.
For the solutions of these equations, we need estimates of several parameters. The
state-of-the-art in estimating diffusion coefficients involves extrapolation of room tempera-
ture aqueous data by assuming
57
-------
Dw
-=r- = constant,
2
where D = diffusivity in cm /sec, ^ = viscosity in centipoises, and T = temperature in °K.
_C O
Typical diffusion coefficients for ions and slightly soluble gases are 1.5 ± 10 cm /sec at
4 9
25 °C. We may extrapolate the data of Schwab and Kolb and obtain a value of 16.2 centlpoia
for the viscosity of 80% H2S04 at 150 °C. The viscosity of water is taken as 1 centipoise at
25 °C. This leads to D = 1.3 ± 10~6cm2/sec for NO+ or SO2 in 80% H2SO4 at 150 °C.
As mentioned earlier, B° (concentration of NO ) is taken as 0.0623 M. We assume A*
to be the saturation value or the solubility of SOg. From extrapolation of the data of Kuznets '
to 150 °C and the assumption of the validity of Henry's Law, we obtain;
A* = 1.16 x 10~4m
_2
From Danckwerts, we obtain k, ~0.6 x 10 cm/sec at 25 °C in water on 1/2-in. Intalox sad-
dies. From Danckwerts and also Gehlawat, et al., it is shown that k, is proportional to D
This leads to k, = 1.16 x 10~ cm/sec at 150 °C in 80 wt % H^SO.. For k«, the bimolecula-»-
L, „ £ 4 Z ««»*• re_
action rate constant, we assume a value ^ 10 f /mol-sec. (This is equivalent to saying that k
bimolecular reaction is appreciably faster than the diffusion process.)
Using the parameter estimates obtained above, we calculate that E,, the instantaneou
enhancement factor, is equal to 269.5. Similarly, M ^ 2.675 x 10 . From Danckwerts6 we
confirm that E = E..
Finally, R, the average rate of absorption of SO,,, is found to be 3.62 x 10 mol/crr^
sec. Now that a value for R at the gas inlet to the reactor is determined, we may begin to c i
culate the conversion yield. The SO2 flux is 2.528 x 10 mol/sec and the NO"1" flux is
8.028 x 10~4mol/sec.
We choose to treat the problem by calculating the amount of time necessary to reduc
the SO, level to 99.9% of its input level and then repeat this process until the desired conver-
sion is attained. Before completing this calculation, we find the residence time for a typical
column. For a 16-ft column, the residence time is 10.26 sec (Intalox saddles 75% void
4
void volume 3 x 10 cc; gas flux 2925 cc/sec at 150 °C).
The rate of SOg absorption is:
,-8 mol
3.62 x 10
cc (total) - sec
2 ~8
for 2.2 cm /cc (total) and for 6.482 x 10 mol SO2/cc (total volume) in the gas phase.
that because of the applicability of Henry's Law and the form of the Danckwerts rate equati
the rate of SO2 absorption (mol/sec) is directly proportional to its concentration in the gas
phase (mol/cc). Thus, the rate given above converts to:
fraction SO,
58
-------
absorbed. This is the differential rate law for our column. To obtain the integrated degree of
conversion, we need to sum the results for slight degrees of conversion as shown below.
To remove 0.1% of the SCL requires ^'A* = 8.13 x 10* sec = r sec. At the end of
T sec, 99.9% of the initial SO2 is present. At the end of n • T sec, 100(0.999)"% of the initial
SO is present. We may calculate % conversion for various residence times. Such results are
2
given in Table III.
We may continue our calculations and compute another table corresponding to conver-
sion at 1/10 of the diffusion limited rate. This would be applicable if the bimolecular reaction
C
was slower. From Danckwerts, we see that the corresponding E factor of 26.95 occurs for
M = 1600. We may calculate k, and determine conversions as shown in Table IV.
MKL
= 6.0 x 107
- fi n * 1 n*
cc
g mol sec
S.
Thus, it can be seen that 90% conversion of SO, will occur in the range of 2 to 18 sec
residence time, and presumably much closer to the lower value. A reasonable series of as-
sumptions of:
DD 20% lower
o
B° 20% lower
DA 20% higher
A
A* Same
kL 20% lower
provides the results shown in Table V of 90% conversion in 4.4 sec residence time and 99% con-
version in 8.8 sec.
59
-------
Table III. Degree of Conversion of 803 at 150 °C in 80% Sulfuric
Acid for Various Residence Times with Diffusion of
SC>2 Being the Rate Limiting Step (See Text for Full
Details)
Residence
Time
(sec)
8.13 x 10~4
8.17 x 10~3
0.0857
0.564
1.872
3.75
5.62
7.5
10.00
Column
Length
(ft)
0.0012
0.0128
0.13
0.88
2.93
5.85
8.76
11.7
15.6
% so2
Remaining
99.9
99.0
90.0
50.0
10.0
1.00
0.100
0.00988
0.00045
%so2
Conversion
0.10
1.00
10.00
50.00
90.00
99.00
99.90
99.99012
99.99955
60
-------
Table IV. Degree of Conversion of S02 at 150 °C in 80% Sulfuric
Acid for Various Residence Time with Rate of Conver-
sion Equal to 1/10 of the Diffusion Limited Rate (See
Text for Full Details)
Residence
Time
(sec)
0.00813
0.0817
0.857
5.64
15.39
18.72
Table V.
Residence
Time
(sec)
0.00192
0.0193
0.202
1.35
4.42
8.84
Column
Length
(ft)
0.012
0.13
1.3
8.8
24.0
29.3
%so2
Remaining
99.9
99.0
90.0
50.0
15.07
10.0
% so2
Conversion
0.1
1.0
10.0
50.0
84.9
90.0
Degree of Conversion of SC>2 at 150 °C in 80% Sulfuric
Acid at Various Residence Times Using Modifying
Assumptions Presented in Text (p. 59)
Column
Length
(ft)
0.003
0.03
0.31
2.1
6.9
13.8
% so2
Remaining
99.9
99.0
90.0
50.00
10.00
1.00
%so2
Conversion
0.10
1.00
10.00
50.00
90.00
99.00
61
-------
VI. SCRUBBER EVALUATION
A. Initial Concept
As discussed in the introduction to this report, the purpose of the scrubber is to remove
the oxides of nitrogen from the reacted stack gas, leaving the reactor stage after the oxidation
of the SO,. This is to be accomplished without removing water from the gas stream. The
evaluation tests run during this contract were aimed at establishing two capabilities of the
scrubbing system: absorption of NO at the high concentrations present after SO., oxidation by
NO0, and absorption of low concentrations of NO that would be present in raw flue gas. In this
2 x
manner the Catalytic Chamber Process could be evaluated for combined NO /SO0 removal as
X ft
well as NO removal alone after the SO» was removed in a prior gas treatment step.
X ™
Prior experimentation, as discussed in the final report to Contract CPA 70-59, had
established that a 4-in. column packed with 8.5 ft of 3/8-in. Intalox saddles could absorb about
65% of the NO from an equimolar mixture of 3000 ppm each of NO and N00. Extrapolation of
x &
this data indicated that a column about 30-ft high would absorb about 97% of the NO . Based on
X
this work, the miniplant was used to evaluate the scrubber concept for absorption of NO from
X
dilute and concentrated gas streams.
B. Experimental Procedures
The miniplant was set up as in Fig. 6 with two packed 4-in. diameter scrubber columns
in series containing about 27 ft of 3/8-in. Intalox saddles. The three Teflon-coated stainless
steel strippers were set up in parallel with the scrubber effluent acid being sent to the bank
of strippers for denitration. In this manner the scrubber system could be operated continuously
over a long period of time. It was decided that at the initial stages of scrubber evaluation,
there was no need to run the reactor stage since the reactor work had already been completed
(see Section V of this report). Therefore, the gas that was fed to the bottom of the large scrub-
ber was composed of a simulated reactor off-gas containing about 6% water and approximately
equimolar quantities of NO and NO« totaling 6000 to 7000 ppm NOx for the first series of tests.
The gas was passed through the large scrubber into the small scrubber and then to the stack.
It was analyzed at three points: before and after the large scrubber and after the small scrub-
ber. (At a later stage of experimentation, the gas was also analyzed half way up the large
scrubber.)
63
-------
Nitrose-free sulfuric acid was fed to the top of the small scrubber with the effluent from
this column being fed to the top of the large scrubber. Acid from the bottom of the large column
was fed to the three strippers in parallel for denitration. The acid was analyzed at the bottom
of each of the two scrubbers and as a composite at the bottom of the strippers. Air was fed to
the strippers (at the bottom) and the effluent dumped to the stack after being analyzed for its
NO content.
A.
Temperatures in the system were controlled by externally mounted heating tapes and
column mantles. The goal in temperature control was to feed the acid to each of the columns
at the required temperature of column operation and use column heaters only to maintain the
temperature. It was felt that it would be too difficult and inefficient to heat the acid while in
the columns, especially in the 12-in. diameter strippers.
All temperatures were monitored on the two 24-point recorders while the gas analysis
was performed on the duPont ultraviolet photometer and the Beckman infrared analyzer as in
previous work. The acid was analyzed in batch tests using a hydrometer for H-SO, concentra-
tion and potassium permanganate titration for HNSO, content. (See Appendix II for details of
the KMnO. titration.)
C. Initial Experimentation — Parametric Study
In order to produce a useful amount of data in a short period of time, a parametric
study was developed by Tyco and the Office of Air Programs (GAP) of EPA, which selectively
varied certain key variables so as to obtain sufficient information in a series of twenty runs to
develop a good feel for the scrubber operation. An outline of the study combined with a data
sheet is shown in Fig. 26. This study was oriented around a combined NO^SO, removal pro~
cess, and high levels of NO were used in the feed stream.
The first runs were made under the average conditions of the parametric study (see
column 11 of Fig. 26). The raw data of these runs (nos. 67 and 68) are shown in Appendix IV
as are all other scrubber/stripper experimental data. Run 69 was planned and then omitted.
Runs 70, 70 (repeat), and 71 were run under the same conditions as the tests performed durin«»
the preceding Contract No. CPA 70-59, which showed approximately 65% scrubbing in an 8.5 ft
column. Run 72 was a followup to 71 in an attempt to evaluate a higher L/G.
It is clear from examination of the data that the high scrubbing efficiencies expected bv
extension of earlier tests during the preceding contract were not achieved in any of these initial
experiments. Where the data is most reliable, consistent readings show no better than 75 to
80% scrubbing instead of the anticipated 98%. (See Runs 67-72 in Appendix IV.) Confirmation
of the analysis is obtained from two sources: NOx material balance and consistent readings
over a long period of time. The material balance errors of closure are shown in the table and
are initially quite low (under 10%) in the first two runs, although later runs are more erratic
More will be said about this later. The consistency of the readings is shown by the subruns
given by a letter designation in the table. These subruns show separate data points taken as
the run progressed and, for the most part, imply a continuous operation of the plant at a
set of conditions.
64
-------
(tails Lwvl 1
TNA
CNA
LjG
GHSVHD
'^'Nq,' F.G.
1CSQ[' F.G.
Conflg.
O ~ r.
x F.C.
Acid
op
16/hr-ft2
mole/mole
hr'^STP)
ppro
ppm
250
915
3 1
6.4
2000
0
- Serics(2)
Units
SCFM
GFM
SCFM
1 5 3 O
0104,
0.4,0.6,
Kun Nunhrr
l^evel 2 Average
340
3600
6 1
33.8
10000
500
Parallel(3)
Levels
6 1 12 O (5 6)
(0 25)
2. 1,3.1, (9. 85,1.
275
2290
4 6
2O 1
6000
250
SorP
.25)
275
2290
4 6
2O 1
6000
250
S
5.6
O 25
1 25
250 :*«> 2SO 300 300 251) 300 250 275 275 300 250 300 250
915 3600 915 3600 915 3600 915 3600 2290 2290 3600 915 3600 915
31 61 61 31 6.1 31 31 61 46 46 61 31 31 61
64338 33 8 64 64338338 64201201338 64 64 33 8
2OOO 2000 2000 2000 10OOO BOOO 10000 10000 600O 6000 UOOO UOOO ttOOO UOOO
0 0500500 0 0500500250250500500 0 0
SSSSSSSSSPPPPP
3.0 6l 15120 15120 30 61 56 56 61 3.0 12 0 1.5
01 O4 01 04 01 04 01 04025025 04 01 04 0. 1
04 21 21 04 04 21 21 04125185 31 0.6 06 3.1
25C
3600
3 1
33 8
2000
500
P
12.0
0.4
3.1
300
915
6 1
6 4
2000
500
f
1 S
0 1
0 6
250
3600
6 1
6 4
2000
0
P
6 1
0 4
0 6
31X1
915
1 1
33 8
2000
0
P
3 0
0 1
3 1
275
2290
4 6
20 1
600O
250
P
5 6
0 25
1 85
L/CHD
3
[HNSOS]3'
hr"1 0.14,0.22,0.58,0.86,(0.36.0.54) 0.54 0.22 0.86 0.22 0.86 0.22 0.86 0.22 0.86 0.54 0.36 0.58 0.14 0.58 0.14 0.58 0.14 0.58 0.14 0.36
16/hr-ft2 81,160,325,640, (302) 302 160 325 81 640 81 640 160 325 302 302 325 160 640 81 MO 81 325 160 302
16/hr-ft2 37,110.148.442,(92.5.270) 270 110 442 110 442 110 442 110 442 270 92.5 148 37 198 37 148 37 148 37 92.5
16/hr-ft2 0.27.1.34.1.41,6.97,(4.14.0.84) 0.84 0.27 1.41 1.41 0.27 0.27 1.41 1.41 0.27 0.84 4.14 6.95 1.34 1.34 6.75 95 1.34 1.34 6.95 4.14
mole/mole ^iVw'"^'?^'!)5'23"0' M'2 23-° 17'5 *'* 92-* 23"° 17'5 4'4 92'' M-2 22'9 "•' 1S 4 6'"4 3-° "-9 15'4 M'4 3 ° 22 9
%V 0.14-19.8,(2.99,4.46) 4.46 1.48 0.57 0.14 5.95 3.77 5.65 1.42 15.1 4.46 2.99 1.75 4.96 19.8 0.49 0.77 0.51 2.01 0.19 2.99
SgW 0.03,0.16,0.40,0.79,(0.32) O.32 0.16 0.08 0.08 0.16 0.40 0.79 0.79 0.40 0.32 0.32 0.40 0.79 0.79 0.40 0.16 0.08 0.08 0.16 0.32
Tyco Mini -Flam Data
Date
I-., i °F
'NA r
TEJT> °F
QF.G SCFM
QAir SCFM
2 "*
3
rcu~cn 1 1 91 W
3
1
2
3 36V
|Curu F_G_ |juu
1
2
[CSQA] F.G. pnni
1
2
3 5tv
(Cuo/u H.G. 56 V
1 "
3
[CHjol Air %V
'
1
1
NOTES:
1. KEY: 1- Top of NO,, absorber; 2- Middle
3- Top of HNSO5 decomposer; F.G
2. VCatalyst= 3.72 ft3 (series) ; 5. 57 ft3 (
3. AHD= 0.70 ft3 (series); 2.09 ft3 (paral
4. Calculated, assuming that [Cg^sOsl ~ ((
>
of N(
Flu
>arall
el).
=N01 =
}x absorber
s gas.
el).
= [CN021 =
0
=
Fig. 26. Scrubber/stripper parametric study and data sheet.
-------
Two major problems appeared which contributed to the difficulty of operation and, in
part, to the low scrubbing efficiency. First of all, there was a great deal of difficulty in a-
chieving continuous operation due to frequent pump failure. This was before the more reliable
peristaltic pumps were installed. More pertinent to the scrubbing problem was the presence
of a relatively high concentration of dissolved iron and chromium salts. The metal ions were
clearly obvious in the acid as seen by the bright coloration of the acid while the fluid was hot
and the presence of a light-colored precipitate throughout the system when the acid was cooled.
Appendices V and VI of this report discussed in detail the source and influence of the dissolved
metals on the efficiency of the scrubbing operation, but there did appear to be a connection
between scrubbing efficiency and metal concentration in the acid.
Examination of the data for Runs 68 through 72 shows that as each run progressed,
there was a decrease in scrubbing efficiency. This was due, in part, to the increase in tem-
perature in the scrubbing columns, but even where the temperature was relatively constant
there was a continued decrease in scrubbing. It is hard to evaluate the start of such an effect
due to the way the system was brought up to operating conditions. Fresh acid was put in the
system and the whole plant started up at room temperature. The acid was pumped around the
entire loop through both the scrubbers and the strippers. When the plant was close to operating
temperatures, the NO was introduced into the gas feed and the experimental data taken. By
A
this time the acid had achieved the deep, bright green that was indicative of the relatively high
concentration of chromium ion. Therefore, it was impossible from the data available to deter-
mine the scrubbing efficiency of uncontaminated acid under the desired operating conditions.
The source of the metal ions was not immediately apparent. Due to the failure of the
Hastelloy B pumps, it was at first assumed that the corroding pumps were the source; but after
it was determined that the metal ions were primarily chromium, this source was ruled out
(there is no chromium in Hastelloy B). The only chromium in the system was in the Carpenter
20 of the piston pumps (which were not always in service during these runs and had not shown
a tendency to corrode this severely in previous runs) and in the stainless steel of the several
fittings used to connect the pumps into the glass pipe system. It was not until the end of the
experimentation that the interior of the strippers was examined, and it could be seen that the
Teflon coating had undergone massive failure and the chromium was indeed coming from the
badly corroded stainless steel walls of these columns. (For a more complete discussion on
corrosion of pumps and strippers, see Appendix VI.)
This problem with the coated strippers created a requirement for new stripper columns
New strippers were installed as shown in Fig. 6. The size and shape of the new columns was
the same as before, so there was no change in operation.
After the new strippers were installed, Runs 74 through 81 were made, which indicated
that under certain conditions it was possible to reduce the scrubber effluent to less than 150 Ppm
total NO (see Appendix IV, Run 74).
At this stage in the experimentation, the orientation of the program was shifted to low
concentration removal of NOx> A new parametric study was used to determine the run conditiOn
66
-------
(see Table VI), with NO and NOg being introduced at 250, 500, and 1000 ppm each.
As indicated in Table VI, the stripper configuration was changed from three parallel
columns (1 ft by 4 ft high) to two columns in series (1 ft by 8 ft high). A diagram of series
stripper configuration is given in Fig. 27. An in-depth discussion of stripper performance is
given in Section VIII of this report.
D. Mathematical Analysis of Experimental Scrubber Data
In this section, we present a mathematical treatment for the scrubber /stripper experi-
ment. We also present a short discussion regarding scrubbing mechanisms. Actual data reduc
tion follows at the end of this section and in Appendix VII.
We consider the experimental data to be a function of column height, column cross sec-
tional area, and initial concentrations. Models of first and second order processes are devel-
oped. Equimolar NO and NO™ are assumed for the beginning of each run when scrubbing is
other than pure first order.
For first order scrubbing,
where:
P ,„ = instantaneous concentration of NO in the bulk gas (atm),
NO
t = time (sec), and
K = first order rate constant
For second order scrubbing,
and
(3)
3t K2PNOPNO, '
u
For first and second order processes,
.3
flow rate (^J = -r = G. (5)
volume (ft3) = V = Ah,
time (sec) = t = — = AhG » and (6)
dt= AG^dh (7)
67
-------
Table VI. Parametric Study and Supplemental Work Plan
Run
No.
82P
83p
84?
85p
86P
87P
88
89
90
91
92s
93s
94s
95s
96s
97s
98
99
100
101
102
Acid
Flow
GPM
0.25
0.40
0.40
0.10
0.10
0.25
0.20
0.40
0.10
0.10
0.25
0.10
0.40
0.40
0.10
0.25
0.25
0.25
0.62
0.25
0.25
Main
Gas Flow
SCFM
4.9
11.7*
6.0
1.5
2.9
4.9
2.9
2.9
6.0
11.2
4.9
2.9
6.0
11.7*
1.5
4.9
11.7**
4.9
4.9
4.9
4.9
Stripper
Gas Flow
(Total)
SCFM
1.9
3.1
0.6
3.1
0.6
1.9
3.1
3.1
3.1
3.1
0.9
1.6
1.6
0.3
0.3
0.9
0.9
0.9
0.9
0.9
0.9
L/G
Moles/
Moles
4.6
3.1*
6.1
6.1
3.1
4.6
6.2
12.4
1.5
0.8
4.6
3.1
6.1
3.1*
6.1
4.6
1.9**
4.6
11.4
4.6
4.6
Scrubber temperature: 275 °F Flue gas composition:
Stripper temperature: 290 °F „ n „*
_2_ NO WU2
6% 250 ppm 250 ppm
500 ppm 500 ppm
1000 ppm 1000 ppm
p Parallel strippers parametric study.
s Series strippers parametric study.
* Actual gas flow was 8.8 SCFM, giving an L/G of 4.2 mole/mole.
** Actual gas flow was 8.8 SCFM, giving an L/G of 2.6 mole/mole.
68
-------
Gas exit
Acid feed
4-ft column
sections
Thermocouple lines
Charcoal
Saddles
Redistributor plate
Support plate
Charcoal
Saddles
Support plate
Air feed
Acid exit
Fig. 27. Series stripper configuration.
69
-------
for constants A and G.
Eqs. (1) through (4) become
dpNO AK1PNO
AK1PN02
dh
dpNQ
= - G - ' and (10)
dpNO AK2PNOPN02
= - G - ' (U)
For the second order process, assume that
PNO = PN02 (12)
and Eqs. (10) and (11) become
AK2PNO
- dpNO ' -4^ and (13)
AK2PN02
~ dpN02 = — G - ' (I*)
Integrating the first order NO equation,
P0
NO _x
/P0 PNO dpNO = '
*NO
f
Jp
M-£r) • K.AHtrl • «•>
Integrating the first order N02 equation,
= +KiAHG
70
-------
Integrating the second order NO equation,
fPNO 2 AK fH
J Q PNO dpNO= -— JQ dh and
PNO
-jj- - — = -KgAHG'1 . (20)
PNO
Integrating the second order NO, equation,
. (21)
PN02
Solving for expected P values:
First order - NO
PNO=PNOexp(~KlAHG~1} ' <22)
First order - NO2
PNO = PNO exp (-KiAHcrl) • (23)
2 2
Second order - NO
P°NQ
PNO
Second order -
(24)
PNOn
PN02 = ^p^K.AHG-1 '
Mechanistic tests that seem reasonable are easily devised. First, plot
PNO
} PNO
versus AHG~ Iog1f.e, and in the region where the plot is linear, NO is being removed by a
first order process [cf. Eq. (17)]. The first order rate constant, K., is given by the slope.
A similar plot may be made for NO2< Second, plot
(25)
71
-------
versus (-AHG ) and in the region where the plot is linear NO is being removed by a second
order process [cf. Eq. (20) ]. A similar plot may be made for
Note that in the foregoing tests the intention is to plot all of the data on four plots:
1. First order NO plot.
2. First order NO2 plot.
3. Second order NO plot.
4. Second order NO, plot.
Putting all of the runs on only four plots will have the effect of averaging the data. By visual
inspection one may determine whether or not the data falls into any significant groups. When
this has been ascertained, one may perform meaningful, weighted, least- squares fits of
selected data groups to obtain first- and second-order rate constants applicable to various
scrubbing conditions.
When rate constants have been calculated, one may calculate percentage scrubbing for
various scrubbing conditions and column geometries on the basis of first- or second-order
constants [Eqs. (22) through (25) ]. For a somewhat more precise treatment, one may derive
rate equations for simultaneous first- and second-order processes.
If NO were removed by second-order processes only and NO- is both first- and second-
order processes, we would write;
= —_ , (8)
and
dpN09 AK1PNO()
& ft , et
dh— = G + 0 (26)
We may solve these equations only if we can obtain an expression relating p and jx
Experimental observations tend to indicate that except at very low NO levels, we may validly '
use the approximation:
PNO = PN02 ' (12)
For NO, we obtain:
PNO = PNO exp (~K1AHG ) ' (22)
For NOg,
/»D^,~ NO0 _., /»H
dh
/»n KNO« , /•
f PN02 ^ 2 , = -AC"1 f
I 2 I
J 0 K1PN02 4 K2PN02 -'O
PN02
72
-------
From the general integral solution:
J
dx 1 In a + bx
x
x(a + bx) a
(28)
we obtain
r
dp
N0
In
Then,
0 K,Pnn
pN02 1 N°2
IK + Kn° \
PN02 \jS + K2PN02/|
(29)
O/
/K
PNO ii
2 I
and finally,
PN09 "
— cx.p ^— jv^^vnu ; ,
+ K2PNO? )
2/
0 / -L
K1PN00 exp (~KiAHG )
2
1C -I- K^n" _ F1 - fivn ^-K AHr,~nl
(30)
2NO
2
We may see if Eq. (31) reduces to Eqs. (23) or (25) as Kg or
tively. As Kg approaches zero, Eq. (31) becomes:
PNO
At small values of K^ we may approximate:
e~x = 1 - x ,
and Eq. (31) becomes
(31)
approach zero, respec-
(23)
(32)
(33)
2
And, as K1 approaches zero, this becomes:
.0
'2
PNOO
K2AHcrl '
(25)
73
-------
Thus, the more general equation (Eq. 31) reduces to the simpler equations (Eqs. 23 and 25) in
the limiting cases.
We may now calculate percentage scrubbing for various scrubbing conditions and geom-
etries, using Eqs. (22) and (31) . The only restrictions that apply are that PN(~ should not be
significantly lower than pNQ if Eq. (31) is to hold exactly. The necessary calculations may be
easily performed on a computer.
When we perform the calculations we should be aware that we can make reasonable
guesses at the "chemistry-of-the-scrubbing-process" on the basis of our data treatment. Our
preliminary guess regarding the rate limiting step in the second order process is that it in-
volves gas phase combination of NO and NOg to form gaseous NgOg which then rapidly dissolves
in the acid. If this view is correct, the K, values should be reasonably constant for any given
temperature and reasonably independent of gas flow rate, liquid flow rate, column length, and
column cross-sectional area. Our preliminary guess regarding the rate limiting step in the
first order process is that it involves dissolution and/or liquid phase disproportionation of
2*
If this view is correct, the K- values should be dependent on temperature, gas flow rate, and
liquid flow rate.
E. Analysis of Scrubber /Stripper Data
Both Tyco scrubbing data for 80% sulfuric acid and OAP scrubbing data for aqueous
base solutions were analyzed. Not all of the Tyco data was utilized. Experimental scatter was
the primary criterion for excluding data. Only total column scrubbing data were plotted and
discussed in detail; the 10-ft and 19-ft data were not used.
Before commencing the data reduction, it is possible to obtain an estimate of the degree
of reproducibility obtainable with the Tyco equipment. One may compare the percent removal
for the centerpoint runs which represent the average conditions in the parametric study in
Table V; data for these runs: 82, 87, 92, and 97, are given in Appendix IV). The scrubbing
efficiencies were 63.2, 72.8, 88.1, and 72.6% NO removal. The difference in scrubbing effi-
X
ciencies is 9.6% for Runs 82 and 87 and 5.5% for Runs 92 and 97. The average difference in
scrubbing efficiencies for the series and parallel stripper configurations is 12.3%. Since the
acid from the stripper was quite well denitrated and would not be expected to influence the re-
sults, we must view these differences in scrubbing efficiency as an indication that experimental
scatter is high. To partially overcome these difficulties, we used averaged data for the follo^.
ing analysis presented in this section.
One may also obtain a good estimate of the scrubber NOx removal imbalance, i.e.,
whether or not NO and NO2 were scrubbed in an equimolar ratio. For this purpose we consult
Table VII- 1 in Appendix VII which is a summary sheet of the Tyco scrubbing data. The first
column gives the run number and the second column gives the flow rate in SCFM. The next
columns give inlet and outlet concentrations in ppm. The seventh and eighth columns give the
amount of material scrubbed from the gas stream (in ppm) and the last column gives the dif-
ference of the values in the seventh and eighth columns (in % error). Thus, the NO error
74
-------
value of 20 listed for Run 82AB indicates that 20 ppm more NO than N00 was scrubbed. (The
£t
analogous OAF data are presented in Table VII-4 in Appendix VII.) In an attempt to determine
the source of this NO error, one may determine average values of the NO error for certain
X
groupings of the data. This data is presented in Table VII-2 in Appendix VII. For the six
groupings listed, the values for the average NO error range from -01 ppm to +32 ppm.
Though the standard deviations in these average values are high (between 140 and 300 ppm),
the value for the average NO error is 120 ppm more positive for the 80 series than for the
X
90 series (see data tables in Appendices IV and VII for details). This might be indicative of
an analytical error, perhaps caused by a change in the experimental setup or due to a time
dependent factor, such as "drift," in the analytical equipment. Also, there is definitely a
trend toward more positive values of the average NO errors as the input NO levels to the
X X
scrubber increase. This might reflect concentration dependent analytical errors or it might
reflect the air oxidation of NO to NOg.
In the foregoing paragraphs, we have discussed sources of variation in the NO analyses.
X
The deviations noted are of the order of 100 ppm. The averages of the absolute values of the
NO errors are 60, 90, and 140 ppm for the 250, 500, and 1000 ppm each of NO and NOg input
levels, respectively. This amounts to an error which is 20 to 100% of the output NO levels.
X
Since the calculation of rate constant values is sensitive to rather small deviations in the output
levels, one may expect that the rate constant values we obtain will show wide variance. Such is
the case.
The Tyco data analyzed in this section is presented in Table VII-3 in Appendix VII. The
GAP data is presented in Table VII-4 in Appendix VII. Rate constants were calculated with the
aid of the computer programs given in Table VII-5 and Table VII-6 in Appendix VII. The result-
ing rate constants are given in Table VII-7, Table VII-8 in Appendix VII, and Figs. 28 through
31. Only the Tyco results for the total column height of 26-1/2 ft are discussed in this section.
Due to the high variance observed with these calculations, one may not draw firm con-
clusions solely on the basis of the numerical values presented here. Figs. 28 through 31 indi-
cate that the scrubbing reactions are neither pure first order or pure second order reactions.
The upward slope of the Kl plots coupled with the downward slope of the K2 plots is indicative
of rate limiting steps which are intermediate between first and second order. We postulate,
but cannot prove (with the data presented here), that the observed results are a result of a
bimolecular reaction with some mass transfer control. If sufficient data on basic column
parameters were available, we might be able to apply calculations similar to those given in
the section on the reactor mechanism.
The most conclusive evidence that we have regarding the scrubbing mechanism comes
from the wide range of experimental conditions covered by the Tyco and OAP work. Within
experimental error, essentially the same scrubbing rate constant values were observed for
80 wt % sulfuric acid, water, alkaline earth hydroxide slurries, and concentrated sodium
hydroxide solutions. Nitrosyl ion is the stable nitrogen specie expected to be formed in the
scrubber under strongly acid conditions, such as 80 wt % HgSO^ Nitrite ion is the stable
75
-------
-O NO (Tyco)
-• NO2 (Tyco)
NO (Garcia10)
86A
f\C'P2
85
KEY
"A" after run number indicates average
ACP - average of center point runs
Circled numbers indicate Garcia's*" run numbers
(See Table VII-4 in Appendix VII for data)
200
400 600 SOO
Inlet Concentration, ppm
1000
I ZOO
Fig. 28. First order reaction rate constant KI as a function of NOX inlet concentration
Data from Tyco and EPA (Garcia 10). °-
76
-------
16 -
O NO (Tyco)
NO2 (Tyco)
NO (Garcia10)
N02 (Garcia10)
12 •
ACPI
KEY
"A" after run number indicates average
ACP - average of center point runs
Circled numbers indicate Garcia's10 run numbers
(See Table VII-4 in Appendix VII for data)
ACP2
600
too
200
400
1000
uoo
Inlet Concentration, ppm
Fig. 29. Second order reaction rate constant K2 as a function of NOX inlet concentration.
Data from Tyco and EPA (Garcia 10).
77
-------
oo
500
NO (ppm)
A.
1000
Fig. 30. Standard deviation for first order rate constant values, 95% confidence limits.
-------
CO
12
500
N0x (ppm)
1000
Fig. 31. Standard deviations for second order rate constant values.
-------
nitrogen specie expected to be formed in the scrubber under neutral or alkaline conditions,
such as prevailed in the GAP work. One would expect different rates of formation for these
two species because of the wide variation in pH and water activity. If the formation of NO+
or NO,, were rate limiting, we would have expected a difference in scrubbing rates. No dif-
ference was observed. Therefore, the theory that the gas phase reaction between NO and NO
is controlling is substantiated. Another factor is that considering the wide temperature range
one would have expected a wide fluctuation in scrubbing rate if the rate limiting step involved
a liquid phase reaction.
The only rate limiting steps which we can imagine might be responsible for the observed
results would be diffusion and gas phase formation of NgO, followed by rapid dissolution. Bi-
molecular gas phase reactions can have low temperature coefficients (i.e., reaction rates rnav
be a weak function of temperature), which would correspond with the observed results.
The Tyco and GAP scrubbing results may be applied to the design of a scrubber if one
takes care not to extrapolate the derived rate constants too far from the concentration regions
in which they were obtained. It is also necessary to use similar packing materials.
80
-------
VII. STRIPPER EVALUATION
A. Initial Concept
The catalytic stripper stage of the Catalytic Chamber Process had the function of deni-
trating the acid effluent from the high temperature scrubber and simultaneously oxidizing the
nitrogen oxides to NOg. Since the nitrogen oxides were absorbed as equimolar amount of NO
and NO0, this meant that half the NO had to be oxidized.
2 x
Experimentation during Contract CPA 70-59 indicated that exposure of the nitrose solu-
tion to the activated charcoal catalyst in a countercurrent stream of air resulted in a minimum
of 99% denitration, with the off-gas being primarily NO«. The process had not been evaluated
quantitatively and the reaction mechanism had not yet been determined. The earlier experi-
mentation had shown that the acid and air space velocities were critical parameters, and ef-
fective (although not optimum) values had been determined. Work during this contract was to
more completely evaluate the stripper concept in both continuous miniplant tests and in batch
laboratory scale experiments.
B. Laboratory Scale Stripper Tests
Several experiments were carried out with a stripper column made of 1-in. glass pipe
with a total length of 3 ft. The packed portion of the column was 29 in. in length. A thermo-
couple well employing a thin-walled, 5-mm glass tube was situated at the center of the packed
section. In the later experiments of this series, a thermocouple was also used to determine
the feed nitrose temperature. Before this, feed nitrose was at ambient temperature. Nitrogen
or air was used as the stripping gas and this was not preheated. Column packing was either
1/4-in. Intalox saddles or Witco no. 256 activated charcoal. Sulfuric acid concentration was
nominally 80 wt %.
The goals of these batch tests were:
1. To compare the stripping action of inert and catalytic packings
2. To determine the maximum obtainable NO« level in the effluent
gas
3. To obtain an insight into the stripper mechanism
The results of these initial experiments are given in Appendix VIII, Table VIII-1.
Intalox saddles were used as the column packing for Runs 1 through 17. The sample
lines were not properly heated for these runs so that in cases where the N02 level is unexpec-
81
-------
tedly low, one must allow for the probability of nitric acid formation in the sample lines,
thereby creating NO, analysis errors. This would be expected at temperatures where the
partial pressure of water was higher than about 20 torr or, in our case, for column tempera-
tures in the range of 280 to 300 °F. The most significant observation that can be made regard-
ing saddles is that less than 10% stripping of the nitrose occurs at temperatures of 70 to 185 °p
with either nitrogen or air. At 280 to 300 °F, about 50% denitration is accomplished with nitro-
gen and about 90% with air.
In contrast, significant stripping of nitrose solutions is achieved on activated charcoal
even at room temperature, with either nitrogen or air. At 185 °F, the stripping was generally
above 90%. With nitrogen as a stripping gas, large quantities of NO and small quantities of
NO, were observed. With air, the reverse was true.
The maximum steady state levels of NO, which were obtained were approximately 3%
However, higher levels were obtained when: (1) the acid sample was taken (pressure decrease
in column), (2) the column was flooded (Run 33), or (3) the column temperature was increased
In these cases, momentary high levels of up to 10% were observed. Observations of the sort
mentioned above tend to indicate that the NO, absorbs onto the activated charcoal, but that it
can be displaced by oxygen if the nitrose flux is not too great and if the air flux is high enough
Unless the NO- is displaced, there seems to be a tendency for the NO, to be reduced to NO.
Thus, enrichment of the stripper gas with oxygen might very well enable us to obtain higher
steady state levels of NO,.
To follow up the experimentation in the 1-in. column, a series of experiments were run
in the 2-in. laboratory stripper in order to evaluate the capability of the stripper to produce a
gas stream containing higher concentrations of NO,. The data obtained during this limited ex-
perimentation are shown in Table VIII-2 in Appendix VIII.
The data reveal two significant points: almost complete denitration is easily achieved
and the effluent gas stream containing about 3% NO, with fairly poor material balance closure
In view of later developments in the interpretation of the output of the duPont UV photometer
(see Appendix IX), it is quite likely that the N02 concentration is considerably higher than the
figures shown in Table VIII-2. This is due to the fact that the 15-in. cell which was used duri
diese tests cannot give accurate analyses above about 2% NO,. The calibration curve tends to
increase to a peak and then decrease, making it difficult to determine which side of the peak
should be used to interpret the absorption reading. Use of the 1.5-in. cell eliminates this proh
lem, and tests run in the miniplant strippers usually give better material balances. It is signi-
ficant to note that the percentage of NO in the stripper gas effluent is quite low, usually less
than 10% of the
C. Miniplant Experimentation
Data concerning stripper performance is shown in the combined scrubber /stripper d
sheets in Appendix IV. Two primary conclusions can be drawn from the data: high denitratio
yields are generally achieved, but the gaseous effluent contains a very high proportion of
82
-------
and very little NOg. Although several experiments were run in the miniplant to try to reverse
this proportion, there was usually an excess of NO present. The results showing high NO con-
centrations were contrary to laboratory scale test data and necessitated an in-depth examination
of the stripper mechanism.
D. Analysis of Stripper Mechanism
Since more than half of the effluent NO was in the form of NO, there must be some
1\
reductive reaction occurring in the stripper. One possible reaction would be the reduction of
NO or HNSO. to NO by reaction with the activated carbon, e.g.:
A **
4HNSO5 4 C + 2H20 - COg + 4NO 4 4H2SO4 .
If this were the case the off-gas would show COg. The long path (one meter) infrared batch
cell was calibrated for CO« and the gas from several of the laboratory stripper runs (see
Table VIII-1 in Appendix VIII) were analyzed for CO,. The results of this analysis are shown
in Table VII. One may calculate, using average values for Runs 38 through 41 (and using
2.83 ml/min of nitrose, of specific gravity, 1.73 g/ml containing 0.90 wt % HNSO-), that if
all of the HNSOg reacted with C to produce NO and CO2, one would obtain a partial pressure
of CO« corresponding to 1.943 cc/min of CO2 or 0.0096 atm CO,. This compares with the
average measured value of 0.0095 atm CO2-
This analysis leads to the conclusion that the activated charcoal is not acting as a cata-
lyst, but is entering into a reaction with the nitrose to effect denitration. This is contrary to
the experimental evidence developed earlier in the development of this process. Laboratory
batch tests showed definite evidence of catalytic behavior of several materials in addition to
carbon (see the Final Report for Contract No. CPA 70-59). Clearly, the results with carbon
are not encouraging, but considerably more work would have to be done before the catalytic
approach to denitration of nitrose is ruled out.
83
-------
Table VII. CO% Analysis of Laboratory Stripper Tests*
(Feed gas CO, concentration =
Run CQ2
31 0.0344
34 0.0252
35 0.0143
38 0.0060
39 0.0144
40 0.0090
41 0.0085
42 0.0094
43 0.0058
44 0.0147
45 0.0118
46 0.0038
* See Table VIII-1, Appendix VIII, for details of these experiments.
84
-------
VIII. PROCESS CONTROL
A. Goals
One of the goals of this contract was to examine the Catalytic Chamber Process in
terms of the process control that would be necessary to operate a full-scale plant. This task
was taken on early in the development of the process because of two anticipated problems: the
large scale of the gas handling system, and the need to accurately control the NO/NO, ratio to
maintain an equimolar gas mixture in the feed to the high temperature scrubber.
The objectives of the process control studies were: (1) to determine the magnitude of
upset conditions that the process must accommodate, (2) to determine how rapid detection and
response must be to control the process during upset conditions, (3) to survey and assess the
adequacy of existing and potential analytical and control equipment to satisfy the above require-
ments, and (4) to recommend a process control system to the EPA.
B. Power Plant Operating Conditions
To establish the magnitude of power plant upset conditions affecting the Tyco process,
discussions have been held with TVA, Consolidated Edison Co. (N.Y.). and New England Power
Co.
Discussions with William Thompson of TVA (Chattanooga) provided information on
the operating conditions of a typical TVA coal-fired power plant. A summary of these discus-
sions follows:
1. Turndown rate; Under normal controlled conditions, plant output is changed
(turndown or turnup) at the rate of 3 MW/min for 125 to 550 MW power plants.
2. Normal operating schedule; Under normal operating conditions, a 550 MW
power plant will deliver between 400 and 550 MW. Full load is usually delivered between 8 a.m.
and 7 p.m., with the output being reduced after 7 p.m. Demand production and scheduling can
be done on a weekly basis assuming normal operation.
3. Air control; Under normal operating conditions (between 400 and 500 MW),
130% excess air is used: 120% at the combustion zone and 10% at the stack. Air flow is con-
trolled automatically to within 5%. When operating below normal conditions (200 to 400 MW),
140% excess air is used. There is considerable variation in air flow control under catastrophic
turndown conditions.
85
-------
4. Stack effluent properties: At full load, the flue gas analyzed about 7%
water (assuming 10% water in the coal) and 2500 ppm SO- (3% S in the coal). NO is usually not
determined. (The sulfur in the coal at the Shawnee plant varies between 1 and 4%. Thus,
greater stack gas SCL fluctuations are noted on a week-to-week basis than on a day-to-day
basis as sulfur varies with the coal shipment.)
5. Operating conditions; During normal operations, the load is changed by
pulling one pulverizer out which would feed one level of burners in the furnace. The tempera-
ture of the burner is probably not changed, but the length of the burning zone varies, i.e., the
height of the high temperature zone is reduced. When operating below normal power output
(300 MW or lower), supplementary oil torches are needed. When coming up in temperature
the limiting factor is the metal temperature in the turbines. The output is therefore increased
at a maximum of 3 MW/min even if the line voltage or grid voltage must change as a result.
In the case of equipment failure, the following changes occur:
a. Turbine failure: 500 MW drops to 0 MW in seconds
b. Loss of one pulverizer: loss of 20-40 MW in seconds
c. Loss of pump: loss of 1/3 to 1/2 load in seconds
The plant can operate at 200 MW or less as repairs are made. Wide fluctuations of excess air
would be experienced during equipment failure until control would be established manually to
the desired level.
It must be noted that the power demand requirements on the TVA system do not nor-
mally fluctuate to the extent encountered by primarily urban electric utilities, such as the
Consolidated Edison Co. of New York. To assess this factor, discussions were held with
Joseph Cunningham, Plant Engineer, Astoria (N.Y.) Power Station, Consolidated Edison Co.
Throughout these discussions, it was noted that despite the greater fluctuations in demand
profile, the response rate of the equipment to such fluctuations, as well as operating conditions
are virtually identical.
Total generating capacity of Consolidated Edison (N.Y.) as of July 1, 1970, was 8,039
MW. The Astoria station, representing approximately 18% of Consolidated Edison's generating
capacity, contains five units with nominal capacities as follows: unit 1 — 180 MW, unit 2 —.
180 MW, unit 3 — 335 MW, unit 4 — 380 MW, and unit 5 — 380 MW. [Note: Units 1 — 3
were oil fired, unit 4 was being converted from coal to oil, and unit 5 was scheduled for con-
version upon completion of unit 4.]
For the entire Consolidated Edison system, peak loadings occur during summer week-
day afternoons between 3 and 4 p.m. Winter peaks occur from 5 to 6 p.m. as industrial and
lighting loads overlap. Due to the mix of plant efficiencies present in older utilities, these
system peaks are not necessarily reflected at the Astoria station. In practice, the least ef-
ficient generating units are turned on last and turned off first. Thus, the Astoria load profile
being buffered by the rest of the system, requires the nominal maximum output from 9 a.m.
to 10 p.m.
86
-------
Differences in generating efficiencies of units within a given station are also signifi-
cant. The three larger generating units at Astoria operate at a higher efficiency than the two
older units. Thus, the large units are operated at a continuous level throughout the day;
changes in station output are made by thorough controlling of the smaller units. The interplay
within units is shown most clearly in Fig. 32, which represents the operating parameters of
one of the smaller (180 MW) units. Note the continuous daytime and evening operation at maxi-
mum levels. Dips in power output occur at noontime and after 10 p.m. Minimum levels are
reached between 3 to 4 a.m.
1. Turnup and turndown rates; 2 to 4 MW/min is usually practiced, although
the capability exists for controlled rates up to 8 MW/min. Control of process air feed is done
automatically as with TVA.
2. Normal operation; The larger, more efficient units operate at their nom-
inal maximum capacity 24 hr/day. The smaller, less efficient units follow the load profile
described above. Minimum operating levels of 25% of unit capacity are used.
3. Air control; Units operating at the nominal maximum are run at 120 to
125% air. When run at reduced power, up to 150% air is used to maintain steam temperature.
A 5 to 10% higher air excess is normally used in burning coal as compared to oil.
4. Equipment failure; The most common failure appears to be rupture of a
boiler tube. Units are shut down during repairs. The effects of feed pump failures are de-
pendent on the specific unit. The two smaller units have three pumps each. Pump failure thus
results in the abrupt loss of about 40 MW. The three larger units operate with only one large
mp each. Each unit has a 1/2-size backup pump which must be turned on immediately in
the event the main pump fails. Rapid fluctuations in stack gas composition are to be expected
. event of such failures as well as during startup operations. Startup times range from 3 to 6
(depending on unit size) before the unit can generate sufficient power to go on grid (minimum
operating level). The unit continues up on grid at 2 to 4 MW/min (8 MW/min maximum) until
it reaches the demand operating level.
A meeting was held with New England Power Co. concerning plant operating conditions
as well as pilot plant site considerations. Section X discusses this meeting in detail; but, in
eneral, the operating conditions and techniques did not vary significantly from those indicated
above for TVA and Consolidated Edison.
C. Instrumentation Requirements
Regardless of the detailed control system selected, careful analysis of available ana-
lytical instrumentation will be required. Measurements of water, sulfur dioxide, nitrogen
dioxide, nitric oxide concentrations, stream flow rates, and stream temperatures are cur-
ntly expected to be mandatory in the final control scheme. A preliminary analysis of avail-
ble instrumentation for the measurement of SO«, NO,, and NO has been carried out, with
-articular emphasis on automatic continuous analysis and control. Manufacturer's literature
87
-------
Fig. 32. Graph of power plant operating parameters.
88
-------
was first surveyed to establish preliminary applicability. Where required, the appropriate
technical assistance groups were contacted for more detailed information. Only instrumenta-
tion capable of being integrated with process control devices was considered in detail. The
various analytical methods were organized into the following measuring techniques: spectro-
scopic, electrochemical, colorimetric (wet chemical), and gas chromatographic.
1. Spectroscopic instruments
Both Beckman and duPont produce spectroscopic instruments suitable for pro-
cess control use. NO» and SCL are readily analyzed in the UV region using the 436 m/J and
302 mM wavelengths, respectively. Although the instruments are basically similar in approach,
the duPont 400 series photoanalyzer has the advantage that the sample cell can be run at ele-
vated temperatures. The condensation of water and the resulting alteration of sample stream
composition can thus be avoided.
Since NO is practically transparent in the UV region, IR analysis using the 5250 mjz
band is usually used. Water interferes in the IR region; thus, the sample must first be passed
through a cold trap or a desiccant for drying. The analysis is then carried out at ambient
temperatures. Alternatively the NO may first be oxidized to NO_ (using a charcoal catalyst
and a source of oxygen) and the total NOg measured in the UV region. The measured NO levels,
which are determined by differences in the NO- concentration before and after oxidation, would
not be expected to be sufficiently accurate, particularly where low concentrations of NO must
be determined.
The response rate of spectroscopic instruments to changes in concentration at constant
sample stream flow is determined by the volume of the sample cell. The longer sample cells
required for the measurement of lower concentration levels present the most serious limita-
tion. However, worst case response times of 15 sec are quoted for instruments of this type.
Instrument costs are on the order of $3,000 per installation.
2. Electrochemical methods
NO, SO2, and NO,,, when dissolved in an electrolyte, are electrochemically
active; i.e., they can be oxidized or reduced (in the case of N0« and SO,) by the passage of an
electrical current. Devices based on this principal are being sold by Dynasciences (a Division
f Whittaker) and Environmetrics. Conversations with both manufacturers indicate that the
devices are identical.
The sensor in all cases is a simple electrochemical cell consisting of a porous mem-
brane, an electrolyte, and a reference electrode. The potential of the cell is maintained by an
electronics package which also measures the current output (which is proportional to the quanti-
ty of sample reaching the cell). Sample temperatures are limited to 110 °F or less in order to
avoid damage to the electrochemical device.
Both manufacturers claim devices which can measure the presence of SO2 without in-
terference from NO or NOg. Units which measure NO2 in the presence of SO,, appear to require
89
-------
either a scrubber to remove the SO, or a separate SCL detector to null out the SCL current.
Only Environmetrics claims to be able to measure NO independently of NOg. The units are
relatively inexpensive ($250 for the sensor and $2,000 for an electronics unit capable of
handling two separate signals) and compact. Demonstrations of these devices are currently
being arranged. Particular attention must be paid to problems (e.g., water condensation) in-
herent in the reduction of sample temperature. In this area, Dynasciences has a device avail-
able which uses a porous membrane to separate a portion of the gas to be measured from the
hot stream. Cooling of this relatively dilute sample stream would not be expected to cause
condensation.
Although rapid when compared to the process conditions, the response rate of this
type of detector is significantly slower than that of the spectroscopic instruments. Depending
on the details of sensor design which determine the diffusion rate of the dissolving gas, re-
sponse times are reported to range up to several minutes for 90% of full-scale reading.
3. Colorimetric analysis
Many analytical devices based on the automation or semi-automation of the
classical chemical techniques for the analysis of dissolved gases are available. Most of these
units are designed for ambient (low ppm) levels of pollutant gases. Due to relatively long cycle
times and/or limited concentration ranges, their application to source and process concentra-
tion levels is questionable.
4. Gas chromatography
This method may be divided into a measurement or detection stage and a column
or separation stage. Automated gas sampling can be performed at the process stream tempera-
ture. Once a sample is taken, it can be readily transferred through heated tubing to the analyt-
ical column. Ideally, the column would first separate and isolate each gas component and pass
them over a detector. The sensitivity of this detector determines the sensitivity of the method.
The analytical column determines the selectivity. In principle, this method should have con-
siderable versatility.
Commercial process control chromotographs (Bendix Corp., Applied Automation)
typically include automated sample stream switching, multiple columns (to permit relatively
complex separations), and thermal conductivity detectors, as well as the electronics required
to convert what is basically a batch output (e.g., a peak area) into a trend-following signal
which could be used for set-point control. Prices range from $10,000 and up for installations
capable of handling two or three sample points.
Instrument response times would be dependent on the cycle time of the particular
analytical column being used. Thus, from both the process control and analytical standpoint.
column selection presents the most difficult problem. Single column separation of NO0, NO
A *
and SO9 in the presence of the other flue gas components is not likely. Although specific in-
formation was not obtained, Bendix Corp. claims to have experience in such separations using
multiple column technology.
90
-------
Based upon a survey of presently available instrumentation, it appears that instruments
using spectroscopic measuring techniques are the only units sufficiently flexible in terms of
operating range and temperature to allow their further consideration. The main advantages of
the spectroscopic approach lies in the limited amount of sample preparation required as well
as the inherently continuous output signal. The response of this instrument type is adequate for
integration with process control devices.
D. Process Control Requirements
A preliminary evaluation of the major process control requirements (Fig. 33) was
performed during the current contract. Three major control loops are shown: (1) reactor con-
trol, including control of NO/NO2 ratio, (2) control of the L/G ratio in the scrubber, and (3)
control of the NO2 concentration in the stripper offgas. Also indicated is the overall control
of process stream flows required to maintain the Tyco process in an operating mode subser-
vient to that of the power plant.
Three conceptual design changes in the overall system were also made. First, note
that the NO, recycle stream is introduced prior to the main process blower in order to take
advantage of all available mixing capability. (Since the flow velocity in the process blower is
quite high and since the stream temperatures are well above the dew point, corrosion problems
in the blower should not be increased significantly.) A second alteration involves recycling of
some of the acid wash (produced during fly ash concentration) into the scrubber. In addition to
reducing subsequent neutralization costs, this procedure allows the scrubber to be operated at
a higher temperature than that required to produce product acid at the desired concentration.
Excess water is removed by the scrubber, but the concentration is closely controlled by adding
the dilute acid wash back in. In this manner, small fluctuations in scrubber temperature will
not seriously affect product acid concentration. The temperature control requirements involved
in scrubber operation are therefore reduced. In addition, final modification places the power
plant's stack fan after the isothermal scrubber.
1. Reactor control
As diagrammed, control of the oxidation reactor is obtained by adjusting the
flow rate of the recycle NO2 stream in order to maintain a constant 1:1 ratio of NO to NO, in
the reactor output. Reactor response to changes in NO, and SO2 inputs will be extremely fast
(seconds), since the equivalent holdup time (determined by the rate constant of the reaction) is
roximately 14 sec> An alternative to this feedback control approach, which avoids the mea-
urement of NO (the more difficult analysis), is also possible. Feedforward control, based on
measurements of SO2 and NO2 going into the reactor, is feasible if: (1) the ratio of NO, to NO
. the power station flue gas is constant for all operating levels, (2) no reoxidation of NO occurs
rior to the scrubber, and (3) a history of suitable emission control in the exit flue gas has been
demonstrated.
91
-------
CO
to
FLUE GAS
KEY
RFC Recording Flow Controller
HPC Recording Pressure Controller
RTC Recording Temp. Controller
RFM Recording Flow Meter
RTI Recording Temp. Indicator
RPI Recording Pressure Indicator
RC1 Recording Cone. Indicator
LLC Liquid Level Controller
Control Path
D Control Signal Processing
A Signal Subtraction
~i~ Signal Division
RFI Recording Flow Indicator
FRT Recording Flow Transmitter
O Hand Valve
BLOWER Setpoint. .
L/G
FLY ASH CATALYTIC ACID
SEPARATOR STRIPPER COOLER
PRECIPITATOR #3
Fig. 33. Major process control requirements.
-------
2. Scrubber control
By designing the scrubber to handle the maximum possible flue gas flow at the
desired operating efficiency, scrubber control is reduced to a simple maintenance of the L/C
ratio. As long as the L/G ratio is held constant (by controlling the acid feed from the stripper),
it is anticipated that scrubber efficiency will be maintained at all operating levels.
3. Stripper control
Here again, the control system is simplified by ensuring adequate stripper
capacity to handle the maximum nitrose flow. A feedback loop adjusts a variable speed blower
to obtain 10% NO, in the stripper offgas. As designed, that portion of the 10% NCL stream not
required by the SCL oxidation reactor is diverted to the nitric acid plant. The flow rate to the
acid plant is thus subject to variations dependent on the flue gas input. It is possible to main-
tain a constant flow rate in this stream (by bleeding additional air into the stripper off gas),
but constancy of both flow and composition is impossible. The preferred operating mode of
the nitric acid plant will determine whether the constant flow rate or composition control is
preferable.
An evaluation was also made on suitable startup procedures. In general, power station
outages from key facilities are infrequent, let us say 1 to 2 times per year. However, once
they occur (through either equipment failure or scheduled maintenance), downtimes on the
order of days to weeks may be expected. During these intervals, the Tyco process must be
maintained in a mode capable of relatively rapid startup.
The overriding criteria during both startup and standby operation are the maintenance
of operating temperatures (to ensure rapid and efficient startup) and the storage of NOx (to
avoid both further expense and contamination).
Storage of the NO can be accomplished most easily in a low pressure liquid NO?
X "
tank. During normal process operation, the tank is filled by routing stripper offgas through
a compressor and ammonia chilled condenser. Since NO- is a liquid at 70 °F and 1 atm pres-
re elaborate pressure vessels are not required. As is indicated in Fig. 35, the operating
moerature of the scrubber and stripper will be maintained by circulating sulfuric acid (any
"trose will be removed by the stripper) through a steam heat exchanger.
In the event of a power station outage, the flue gases are first bypassed directly to
the stack blower. The acid loop heat exchanger is then activated to maintain operating tem-
rature. HNSO,- present in the acid loop is removed by the hot stripper and sent to either
MO storage or to the HNOg plant. Once the HNSO, is removed, the stripper blower will turn
ff Pure H SOA will continue to circulate until the system is reactivated.
System startup is accomplished by allowing the hot flue gas to pass through the reactor
nd scrubber. NO2 is bled directly from the NOg tank so as to maintain the proper NO/NO-
atio at the reactor outlet. Effective scrubbing and stripping can commence immediately since
the units are maintained at their operating temperature. The NO2 (from storage) can be shut
93
-------
off once the stripper offgas reaches the 10% N02 level. The length of this period will be deter-
mined by the HNSO. holdup in the acid system.
Once the system has stabilized, the NO2 level in the storage tank can be replenished
by diverting the stripper offgas through the condenser. Thus, both startup length and frequency
will determine the NO2 storage requirements. Obviously, initial startup of the Tyco process
can be effected through the use of purchased NO«.
Tank storage of the NO values as a nitrose solution has also been considered. Svs-
X *
tern startup would be accomplished by bleeding the stored nitrose solution directly into the
stripper. If the stripper is bypassed during standby operation, the nitrose can be stored in the
acid loop itself. In this case, the scrubber would have to be started with a hot stream of pure
FLSO. (from storage) in order to avoid stripping of the nitrose solution during process startun
A major advantage of nitrose storage is that it permits the scrubber to be constructed from less
expensive materials which are passive in nitrose solutions. Use of such materials would, of
course, reduce plant investment.
Selection between the alternative methods of NO storage must consider the economics
of each approach. Without further experimental backup at the 2000 SCFM level, it is premature
to select the ultimate NO storage technique.
X
94
-------
IX. ECONOMIC EVALUATION
A. Prior Evaluations
Economic evaluations of the Tyco Catalytic Chamber Process have been made earlier
in the development of the process by Tyco and by the Tennessee Valley Authority (TVA) (under
contract to OAP). Both analyses showed a required investment of about $12 to $16 million for
removal of 90% of the SCL and 60% of the NO from an 800 MW coal burning power station
whose stack gas contained about 2200 ppm of SO, and 600 ppm of NO . Based on the salability
of sulfuric and nitric acid, at $12 and $40 per ton (100%), respectively, the add-on facility
would operate at about the break- even level. (See the final report for Contract No. CPA 70-59
for details of this economic analysis.)
During the contract discussed in this report, an initial economic analysis of the pro-
cess used only for NO removal was presented to OAP. (A suggested flowsheet is presented
in Fig. 34.) Assuming 75% removal of 600 ppm NO and the manufacture and sale of nitric
acid the plant would cost about $12 million and would add $1.48 per ton of coal to the operat-
ing cost of a 500- MW power station. Details of this analysis are shown in Tables VIII, IX, and
X The analysis was based on the experimental data from Contract CPA 70-59.
B. New Process Concept
As a result of the experimentation performed on this contract, there has been a change
. the process concept as well as an overall enlargement of the plant size. The conceptual
hange involves the use of a liquid phase reactor to oxidize the SO,. This reactor would be
larger than the one used in previous process evaluations and would be more costly because
f the need for either packing or spray tower construction. Fig. 35 shows a proposed flowsheet
new reactor concept.
1. Reactor cost
It is difficult to size the new reactor because of the limited amount of data
ilable, ^m assuming the 14-sec residence time required in the miniplant for 90% reaction
ields, we would need about 1000 ft of columns. These could be set up as 20 columns, 25 ft in
diameter and 50 ft high, packed with 3- in. Intalox saddles. The estimated cost of such columns
would be about $6 million. Further optimization of reactor design and contact methods should
reduce the residence time and reactor cost.
95
-------
low NOX gas
to stack
NO
x
Scrubber
Nitrose
NO
Nitrose
decomposer
HNO,
Flue gas after SC>2 removal;
high NO
Air
Fig. 34. Flowsheet showing the Tyco Catalytic Chamber Process used for NOX removal only.
96
-------
Table VIII. Economics of NOX Scrubbing Plant for 500-MW
Power Plant* Based on Tyco SC>2 Removal Pro-
cess
Estimated Major Equipment Cost
1.
2.
3.
4.
5.
6.
7.
8.
Item
Scrubber
Packing
Stripper
HNOg Absorber
Filter
Product Storage
Tank
Exhaust Fan
Pumps
Use
NO scrubbing
X.
For Item 1 - Scrubber
Denitrification
HNOQ Production
o
Remove ash from acid
Hold HNO3
Provide differential pressure
2 stages of pumping to scrub-
ber and to stripper
No.
4
12,000 ft3 ea.
2
1
4
2
4
—
Size
250,000 SCFM $1
20 ft dia
40 ft high
3" Intalox Saddles
20 ft dia
30 ft high
30 ft dia
60 ft high
50,000 gal
15 x 40
or 20 x 22
-12" H20
Total
54,000 GPM
Cost
,500,000
500,000
500,000
400,000
250,000
50,000
100,000
150,000
TOTAL
Use Lange Factor of 3.5
Total Installed Cost = 3.5 x 3.45 x 106 =
$3,450,000
$12,100,000
* 1 x 106 SCFM
600 ppm NO (75% recovery)
A
97
-------
Table IX. Economics of NOX Scrubbing Plant for 500-MW
Power Plant Based on Tyco SO2 Removal Process
Annual Operating Costs
Materials
Charcoal replacement $ 500,000
Direct Labor 110,000
Supervision 30,000
Maintenance 5% of fixed capital 605,000
Supplies 15% of maintenance 90,800
Utilities
Power 250,000
Water 10,000
Heat (startup only) 50,000
TOTAL DIRECT COST $1,645,800
Payroll Burden (20% of direct labor and supervision) 28,000
Plant Overhead (50% of direct labor, supervision, 417,900
maintenance, and supplies)
Packing and 100,000
Shipping
Waste Disposal 50,000
TOTAL INDIRECT COST $ 567,900
Depreciation (10% of fixed capital) non-regulated 1,210,000
(regulated: $403,000)
Taxes (2% of fixed capital) 242,000
Insurance (1% of fixed capital) 121,000
TOTAL FIXED COST non-regulated $1,573,000
Regulated: $766,000)
TOTAL OPERATING COST non-regulated $3,786,700
Regulated: $2,979,700)
98
-------
Table X. Economics of NOX Scrubbing Plant for 500-MW
Power Plant Based on Tyco SOo Removal Process
Net Process Cost
Credit 19,100 tons/yr (100% HNOg) at $40/ton .......... $764,000
Net Cost (regulated) $2,979,700
- 764,000
$2,215,700
............... *1.48/concoa,
Met Cost (non-regulated) $3,786,700
- 764,000
$3,022,700
or yffiffpfo - _ ............... $2.02/ton coal
1,500,000 tons coal/yr v '
99
-------
Reactor
NO
H20 _
Nitric
Acid
Absorber
Raw flue
gas
To stack
Scrubber
1
fc
;ilute
MCO
. f
70% H2SO
.-J
1
Nit
Ab
rose
jorb
I
24 Product
W/UNSOC
NO, NO,
Stripper,'
Oxidizer
70%
(cone. HNSO5)
Air
80% H.SO.
I 4
NO
NOr
Product H-SO.
£i
Product HNO,
Fig. 35. Proposed flowsheet for Catalytic Chamber Process utilizing a wetted reactor for
SO0 oxidation.
100
-------
2. Scrubbing Costs — S02/N0 Removal
If the process were used for removing both SO2 and NO , it would be necessary
to have packed towers about 105 ft high to reduce the concentration of NO from 6000 ppm total
X
NO plus NOg to 100 ppm total NOx. This figure is developed from visual extrapolation of data
produced in Run 92 (see Appendix IV), as shown in Fig. 36. If we assume an L/G of
5.0 moles/mole and a gas flow of 282 Ib/hr-ft as used in the miniplant experimentation, we
would need about 53 columns 25 ft in diameter and 105 ft high, packed with 3-in. Intalox saddles
(although it would be better to use 106 columns half as high to avoid ground loading problems).
These shorter columns would cost about $1 million each for a total purchased cost of $106
million. Assuming a 3.5 factor for installation and auxiliary equipment, the scrubbing facility
alone would cost about $370 million, clearly an unreasonable expense for this process. The
total capital and operating expenditures are estimated in Tables XI and XII.
Typical commercial operation of a packed column, such as the scrubber, is at 75% of
flooding. If it can be assumed that subsequent process optimization would allow operation at
75% of flooding with the same L/G and column height, we would then need 12 columns 25 ft in
diameter and 105 ft high, packed with 3-in. Intalox saddles, or 24 columns half as high. The
purchased cost of these columns would be about $24 million or about $84 million installed. The
entire plant would cost about $114 million, as shown in Table XIII, or over $140/KW installed
for an 800-MW power plant. Estimated operating costs are shown in Table XIV. Comparison
of Tables XI and XII with XIII and XIV show how strongly dependent the capital and operating
costs are on scrubbing efficiency. Economical implementation of the Catalytic Chamber Pro-
cess clearly depends on increasing the scrubber gas loading and decreasing the reactor
residence time.
C. Cost of NO Removal Only
X
In sizing the scrubbing facility for removing NO only, the same two scrubber flow rates
X
were used as with the combined S02/NO system discussed above. The data from Runs 85c,
96c, 92c, and 96c indicate that about 90% of an equimolar NO/N02 gas stream containing 1000
ppm of each can be removed in about 27 ft of packing at an L/G of about 5.0 moles/mole and a
gas mass flow rate of 282 Ib/hr-ft . Extrapolation of the data, as shown in Fig. 37, indicates
that the gas could be cleaned to 100 ppm total NO in about 45 ft of packing. The cost of a plant
to accomplish this would be about $175 million, with an annual operating cost of about $41.0
million (8.5 mils/KWhr). See Tables XV and XVI for details.
If we again assume that optimization of the system would allow us to operate at 75% of
flooding with a similar column height in the scrubber, we would need less scrubbing capacity
and the plant cost would drop to $43.5 million. The operating cost would then be $10.82 million
or 2.3 mils/KWhr. Tables XVII and XVIII present the breakdown of this analysis.
101
-------
1
CJ
OS
40 80
Scrubber height, ft
120
Fig. 36. Extrapolation of data from Run 92 to give scrubber column height for reducing
NOv concentration from 6000 ppm to 100 ppm.
102
-------
Table XI. Economics of SoJNO2 Scrubbing Plant for 800-MW Power Plant
Tyco Catalytic Chamber Process
Estimated Major Equipment Cost
Basis: 282 Ib gas/ft2-hr (Miniplant data)
Item
1. Scrubber
Use
NO scrubber
A
2. Stripper Acid denitrification
3. HNO~
Absorber
4. Filter
5. Product Stor-
age Tanks
6. Exhaust Fan
7. Pumps
HNOo production
Remove ash from
acid
Hold HN03 and H2SO4
Provide differential
pressure
3 stages of pumping
to scrubber, stripper,
and reactor
g Reactor Oxidize S02
9. Nitrose
Absorber
Nitrose production for
reactor feed
No.
106
Size
Cost
16
1
8
20
1
Using Lange Factor of 3 5:
TOTAL INSTALLED COST =
27,000 SCFM $ 106,000, 000
25 ft dia
53 ft high
packed with 3-in.
Intalox saddles
20 ft dia
30 ft high
30 ft dia
60 ft high
50, 000 gal
4,000,000
400, 000
350,000
200,000
12-in.H2O 300,000
Total 85,000 GPM 400,000
50 ft high
25 ft dia
15 ft dia
40 ft high
6,000, 000
250,000
$ 118,900,000
$ 413,000,000
103
-------
Table XII. Economics of S09/NOV Scrubbing Plant for 800 MW Power Plant*
£t X
Tyco Catalytic Chamber Process
Annual Operating Costs
2
Basis: 282 Ib/hr-ft scrubber gas flow (Miniplant data)
Catalyst replacement (2 replacements per year) $ 2, 000,000
Direct labor 250,000
Supervision 75,000
Maintenance 5% of fixed capital 20,600,000
Supplies 15% of maintenance 3,100,000
Utilities
Power 4,000,000
Water 100,000
Heat (startup only) 100,000
TOTAL DIRECT COST $ 30, 325,000
Payroll burden (20% of direct labor and supervision) 65.000
Plant overhead (50% of direct labor, supervision, 12,013,000
maintenance, and supplies)
Packing and shipping 100,000
Waste disposal 50, OOP
TOTAL INDIRECT COST $ 12,228,000
Depreciation (10% of fixed capital; non-regulated) 41, 300,000
Taxes (2% of fixed capital) 8,260,000
Insurance (1% of fixed capital) 4, 130,000
$ 53,690,000
TOTAL OPERATING COST (non-regulated) $ 96,243,000
or 19.8 mll/KWhr
* 330 days of operation per year at 75% capacity
104
-------
Table XIII. Economics of SO2/NOx Scrubbing Plant for 800-MW Power Plant
Tyco Catalytic Chamber Process
Estimated Major Equipment Cost
Basis: 75% of flooding in scrubber
Item
I Scrubber
Use
No. Size
2. Stripper
3. HNO«
Absdrber
4. Filter
NO scrubbing
Acid denitrication
HNO, production
Remove ash from
acid
5 Product Stor- Hold HNOo and H0SO,
age Tanks J ^
6. Exhaust Fan Provide differential
pressure
7. Pumps
8. Reactor
9 Nitrose
Absorber
3 stages of pumping
to scrubber, stripper,
and reactor
Nitrose production for
reactor feed
24 120,000
25 ft dia
53 ft high
packed with
3-in. Intalox
saddles
4 20 ft dia
30 ft high
1 30 ft dia
60 ft high
50,000 gal
12-in.H20
Total 85, 000 GPM
20 50 ft high
25 ft dia
1 40 ft high
15 ft dia
Cost
$ 24,000,000
Using Lange Factor of 3.5:
TOTAL INSTALLED COST =
1,000,000
400,000
350,000
200,000
150,000
250,000
6,000, 000
250,000
$ 32,700,000
$ 114,000,000
105
-------
Table XIV. Economics of SO9/N
-------
20 40
Scrubber height, ft
60
Fig. 37. Extrapolation of data from Runs 85c, 86c, 92c, 93c, and 96c to give scrubber
column height for reducing NOX concentration from 2000 ppm to 100 ppm.
107
-------
Table XV. Economics of NOX Scrubbing Plant for 800-MW Power Plant
Tyco Catalytic Chamber Process
Estimated Major Equipment Cost
Basis: 282 Ib gas/hr-ft2 (Miniplant data)
Item Use No. Size
Cost
1. Scrubber NOV scrubbing
X
27,000 SCFM $ 40,000,000
25 ftdia
45 ft high
3.
4.
5.
6.
Stripper
HNO,
Absorber
Filter
Product
Storage
Tank
Acid denitrification
HN03 production
Remove ash from acid
Hold HNO3
3
1
4
2
7. Exhaust fan Provide differential
pressure
8. Pumps 2 stages of pumping
to scrubber and stripper
Using Lange Factor of 3.5:
TOTAL INSTALLED COST =
10,000
t3ea
3
1
4
2
4
-
3-in.Intalox
saddles
20 ft dia
30 ft high
30 ft dia
60 ft high
-
50, 000 gal
- 12-in H20
Total 50.000GPM
8,000,000
750, 000
400,000
200,000
50,000
200,000
300,000
$ 49,900,000
$ 175,000,000
108
-------
Table XVI. Economics of NO Scrubbing Plant for 800-MW Power Plant'
A
Tyco Catalytic Chamber Process
Basis: 282 Ib gas/hr- ft2 (Miniplant data)
Annual Operating Costs
Catalyst replacement (twice per year) $ 375,000
Direct labor 20°,dou
Supervision 60,000
Maintenance 5% of fixed capital 8,750,000
Supplies 15% of maintenance 1,310,000
Utilities
Power 2,000,000
Water 100,000
Heat (startup only) 150,000
TOTAL DIRECT COST 12, 925, 000
Payroll burden (20% of direct labor supervision) 52,000
Plant overhead (50% of direct labor, supervision, 5,160,000
maintenance,and supplies)
Packing and shipping 20,000
Waste disposal 50,000
TOTAL INDIRECT COST 5, 282, 000
Depreciation (10% of fixed capital) 17.500, 000
Taxes (2% of fixed capital) 3,500,000
Insurance (1% of fixed capital) 1,750,000
TOTAL FIXED COST $22,750,000
TOTAL OPERATING COST (non-regulated) $ 4Q 957 Q00
or 8.5 mil/KWhr
*330 days operation per year at 75% capacity
109
-------
Table XVII. Economics of NOV Scrubbing Plant for 800-MW Power Plant
A
Tyco Catalytic Chamber Process
Estimated Major Equipment Cost
Basis: 75% of flooding in scrubber
Item
Use
Scrubber NOV scrubbing
A
No.
12
2. Packing For item 1 - scrubber 20,000
3. Stripper Acid denitrification
ft3ea
4. HNOo HNO« production
Absofber °
5. Filter
6. Product
Storage
Tank
Remove ash from acid
Hold HNO0
2
1
4
2
7. Exhaust Fan Provide differential
pressure
8. Pumps 2 stages of pumping
to scrubber and stripper
Using Lange Factor of 3.5:
INSTALLED COST =
Size
120,OOOSCFM
25 ft dia
45 ft high
3-in. Intalox
saddles
20 ft dia
30 ft high
30 ft dia
60 ft high
50, 000 gal
H2O
Total 50,000 GPM
Cost
$ 10,000,000
2,000,000
500,000
400,000
200,000
50,000
100,000
150,000
$ 12,400,000
$ 43, 500, 000
110
-------
Table XVIII. Economics of NOV Scrubbing Plant for 800-MW Power Plant*
X
Tyco Catalytic Chamber Process
Annual Operating Costs
Basis: 75% of flooding in scrubber
Catalyst replacement (twice per year) $ 250, 000
Direct labor 150,000
Supervision 50,000
Maintenance 5% of fixed capital 2,170,000
Supplies 15% of maintenance 325,000
Utilities
Power 750,000
Water 25,000
Heat (startup only) 50, 000
TOTAL DIRECT COST 3,760, 000
Payroll burden (20% of direct labor supervision) 40,000
Plant overhead (50% of direct labor, supervision, 1, 350, 000
maintenance, and supplies)
Packing and shipping 20,000
Waste disposal 50,000
TOTAL INDIRECT COST 1, 460, 000
Depreciation (10% of fixed capital) 4,350,000
Taxes (2% of fixed capital) 870,000
Insurance (1% of fixed capital) 435,000
TOTAL FIXED COST $5,655,000
TOTAL OPERATING COST (non-regulated) $10, 875, 000
or 2.3 mil/KWhr
* 330 days operation per year at 75% capacity
111
-------
X. CONCLUSIONS AND RECOMMENDATIONS
A. Conclusions
1. The gas phase oxidation of SO, by NO, is too slow to be practical with feed
SO levels of 3000 ppm. Residence times of several minutes at 250 to 300 °F would be neces-
sary to achieve 90% oxidation. This would necessitate reactor volumes of several million cubic
feet which would be extraordinarily expensive as well as creating enormous technological and
logistical problems (e.g., a 5 million cubic foot reactor could be envisioned as 3 miles of 20-ft.
diameter ducting).
2. Liquid phase oxidation of SO, using 2 wt%HNSO,. in 70% H,SO. can be ac-
complished in packed towers in residence times of about 15 sec for 90% SO« oxidation. The
data indicates that the reaction is probably not diffusion controlled, with the limiting step being
the liquid phase reaction itself. Techniques for providing a more intimate gas/liquid contact
should be able to reduce the residence time down to a workable 10-sec time.
3, The rate of absorption of equimolar quantities of NO and NO, in 80% sulfuric
acid at 250 °F is not controlled by the individual absorption of the two gases. It appears to be
dependent on a gas phase interaction between the two species to form N,Og, which is readily
absorbed at these conditions. The gas phase reaction is probably slower than the absorption
step, in addition to which the equilibrium at these elevated temperatures is shifted towards
the unreacted species. These conditions make the absorption from dilute gases very slow com-
pared to absorption in cool acid under the more concentrated conditions of the standard Lead
Chamber Process.
4. At 250 to 300 °F, in the presence of less than 0.3% nitrosylsulfuric acid and
air the activated charcoal does not behave as an oxidizing catalyst. Thus, in the stripper/
oxidizer state the charcoal probably enters into a reaction with the nitrated sulfuric acid to
form NO and CO,. This is contrary to results obtained at higher nitrosylsulfuric acid con-
centration and should be verified by further experimentation.
5. The cost of an add-on Catalytic Chamber Plant to remove both SO, and NO?
from power plant stack gas would cost about $413 million for an 800 MW plant if its size were
based on miniplant data directly. Its operating cost would be almost 20 mil/KWhr. If the plant
were sized using an industry practice level of 75% of flooding in the scrubber, the plant cost
ould be reduced to $114 million with an annual operating cost of about $27 million or $125/KW
and 5.5 mil/KWhr respectively.
113
-------
6. If the process were used to remove NO only from the stack, the plant cost
A
would be about $175 million and the annual operating cost about $41 million if miniplant data
were used directly for scale-up. If 75% of flooding were used here in the scrubbers, the costs
would drop to $43.5 million and $10.8 million, respectively. This would be equivalent to $B5/KW
installed cost and 2.3 mil/KWhr operating cost. Both of these sets of cost estimates bring out
the importance of efficient scrubbing operation.
7. The overall conclusion is that the Catalytic Chamber Process at its current
level of development is technologically feasible but economically impractical. Major changes
must be accomplished in the three primary process stages — SO, oxidation reactor, NO
adsorber, and NO stripper/oxidizer — before we have a process of sufficient practicability
to justify pilot plant evaluation.
B. Recommendations
In order to bring the Catalytic Chamber Process to a level of development which would
justify larger scale evaluation, basic changes must be made in the process. Experimentation
to develop these changes should be conducted on each stage separately at the bench scale,
rather than the integrated miniplant level, to avoid having equipment problems overshadow
basic technological problems. The three primary process stages should be examined:
1. The liquid phase reaction for the oxidation of SOg with nitrose should be
examined further to see if spray towers or other gas/liquid contractors can be used to reduce
the residence time from 15 sec to the order of 5 sec. The current achievement level of 90%
reaction in somewhat under 15 sec is not out of the question, but it does require a packed tower
having a volume of over 300,000 ft , which would be quite costly. Considering that no effort
had been made to optimize the contact method in the current contract, it is hard to imagine that
considerable improvement in residence time cannot be accomplished.
2. Since the NO and N02 do not interact in the gas phase fast enough to permit
the use of a high temperature scrubber concept, it may be advantageous to return to the original
Lead Chamber Process and use a low temperature scrubber (at about 100 ° F). The actual NO/
NO, reaction rate would be slower, but the equilibrium would be shifted to the NgOg side which
should make the overall process much faster. The process economics looked poor when the low
temperature scrubber was originally examined, but it may be possible to avoid the large heat
loss needed to concentrate all the recycle acid by modifying the technique somewhat. One ap-
proach would be to withdraw only part of the acid and concentrate it to about 95% acid, with-
draw product acid, and men return the remainder to the diluted recycle acid, thus concentrating
the recycle stream back to the desired level. In this case there would be no need to recycle the
most concentrated acid possible (since this is not the market acid), and it would be advanta-
geous to work at the lowest concentrations, compatible with the process concept. Recent work
shows this is about 67% sulfuric acid. A low temperature process which might be practical is
shown in Fig. 38.
114
-------
To stack
so2
Oxidizer
Flue gas
LJ
Nitric
Acid
Absorbe:
Product
95% HS
Fig. 38. Catalytic Chamber Process using a low temperature scrubber.
115
-------
3. More detailed studies should be made of the interaction between dissolved
NO and activated charcoal to see if there are conditions under which the system can oxidize
X
NO rather than reduce it. The data obtained in the contract are not in complete agreement
X.
with earlier data obtained at higher nitrosylsulfuric acid concentrations. Prior work also
showed that other candidate catalysts denitrated the nitrose, although not as well as the acti-
vated charcoal. Further tests should be made with these, including silicon nitride, tungsten
carbide, boron carbide, and others (see final report to Contract CPA 70-59).
If work in any of these three areas is not successful, the development of the Catalytic
Chamber Process should be discontinued.
116
-------
XI. ACKNOWLEDGMENTS
The work summarized in this report was sponsored by the National Environmental
Research Center, Environmental Protection Agency. The guidance and assistance of the
project Officer, Mr. Stanley J. Bunas is acknowledged with sincere thanks.
The report was prepared by Arnold H. Gruber and Arthur Walitt of the Corporate
Technology Center, Tyco Laboratories. The authors are grateful to the following Tyco
personnel for their significant contributions: Dr. David Cogley, Mr. Lewis Gaines and
Or. S. Barry Brummer, along with the pilot plant crew — Edward Broderick, Arthur Chernosky,
Irving Frutkoff, Mourad Ghobrial, Gary Hanson, Kamal Ladha, and Vincent Sljaka.
Acknowledgment is also gratefully given to the efforts and contributions of our sub-
contractor, The Badger Company, and particularly Dr. T.C. Dauphine, Mr. George A.
Randall and Mr. Stephen T. Williams.
In addition, we would also like to acknowledge the helpful discussions with
k/lr. William Thompson of the Tennessee Valley Authority, Mr. Roger Ramsdell and Mr.
Joseph Cunningham of the Consolidated Edison Company of New York, and Mr. Sashil Batra
of the New England Power Service Company.
117
-------
XII. REFERENCES
!. E. Berl, H. Hillebrandt, and K. Winnacker, Z. Anorg. Allg. Chem., 214, 369 (1933),
2. F. Seel and R.Z. Winkler, Physik. Chem. N.F., 25, 217 (1960).
3. D.A. Kuznetsov, J. Chem. Ind., 16, 3 (1941).
4 W. Jost, Diffusion in Solids, Liquids, and Gases, Chapters X and XII.
5> E. Berl and H.H. Saenger, Anorg. u. Allg. Chem., 202, 113 (1931).
6< P.V. Danckwerts, "Gas-Liquid Reactions, " McGraw-Hill Book Co. (1971).
n U.S. Stoneware, Bulletin S-31-R.
I •
8. J.K. Gehlawot and M.M. Sharma, Chem. Eng. Sci., 23, 1173 (1968).
o G. Schwab and E.Z. Kolb, Physik. Chem., 3, 52 (1955).
y* —
IQ L. Garcia, "Adsorption Studies of Equimolar Concentrations of NO and N02
in Alkaline Solutions, " Second International Limestone-Wet Scrubbing Symposium,
New Orleans, November 1971.
119
-------
APPENDIX I. PRELIMINARY ENGINEERING ANALYSIS OF THE
TYCO CATALYTIC CHAMBER PROCESS BY BADGER
CORPORATION
T. C. Dauphine' of Badger Corp. evaluated the Tyco/Catalytic Chamber Process
during the first month of the contract and made recommendations for continued experimenta-
tion based on Badger's experience in the operation of pilot-plant facilities as well as the design
and construction of full-scale plants. Experiment equipment was modified after consultation
with Dr. Dauphine'and George Randall of Badger. The following are excerpts from Dr. Dauphine''s
report.
A. Introduction
Badger has been asked to review the data and correlations made available to Tyco and
to advise on additional information that may be desirable or necessary for proper definition of
the process.
After reviewing the available information and discussing the process conceptions with
representatives of both Tyco and Badger, I have the following comments and suggestions.
B. General Background
An essential requirement of the Tyco process is that it must function at high effective-
ness over the maximum range of operating rates and conditions of the power plant itself. These
eqUirements thus differ from those of most process plants which are set up for relatively nar-
row design conditions concerning feed rates and specification, our system must be designed to
maintain the effectiveness.of the equipment and controls in spite of relatively large variations
in feed rate and composition during the daily cycle.
For future experimentation, we have agreed to consider the problems involved with a
coal fired plant.
The process problems caused by variation in water concentration in the flue gas must
be studied and understood.
A unique feature of the Tyco catalytic process is the high temperature scrubbing of
stack gas with 80% sulfuric acid to remove sulfur oxides and nitrogen oxides without absorption
of net water into the acid system. Water input into stack gas is subject to change, even at
constant firing rate.
1-1
-------
C. Reactor (SO2 Oxidizer)
The reactor mixes a relatively small stream containing N02 (from the catalytic strip-
per) with a large volume of incoming flue gas to oxidize contained SO0 to S00. The reaction
£ O
involved is:
2NO2 + SO2 = S03 + NO2 H- NO .
Past laboratory work has indicated that this reaction in small-scale equipment proceeds
very rapidly almost to completion. However, as the scale increases, the reaction does not
appear to proceed as rapidly.
For commercial scale operation, the problems inherent to very dilute gas phase reac-
tions and problems of mixing large gas streams must be considered and solved.
There also exists a question whether the reaction proceeds rapidly under dry conditions
(above the dew point) or depends upon the presence of liquid acid film.
Another question concerns the effect of light in catalyzing the reaction (in miniplant,
this reaction has been studied in Pyrex glass equipment).
Experiments now being contemplated will determine the effect on the reaction of addi-
tional surface to volume, both wetted and dry.
Other experiments should be made to determine the effect of larger scale on reaction
rate.
Temperature (between practical limits anticipated in a plant) should be explored.
Whereas it is desirable to have the input ratio of NO2 to SO2 equal to 2:1 so that a
mixture of NgOg results, it would be well to explore the effect of ratios different from this
to learn what may happen if operating conditions are upset.
For some of these experiments, it may be useful to set up bench equipment separate
from the miniplant so that it could run independently from the miniplant to explore a broad
range of conditions. Such a setup need not be glass and preferably might be of suitable metal
for short time use, and can be useful to get corrosion information.
It would be desirable to know if there is unconverted S02 in the exit stream from the
reactor, how much SOg is present as sulfuric acid mist, the amount of NO2 and NO in the exit
gas, the effect of fly ash upon the reaction, the effect of soot upon the reaction as well as the
effect of various amounts of water in the flue gas feed upon the dew point, and the resultant
acid condensate which might be formed at high water content levels.
A device which may be useful both for contacting the gases and for dust and mist re-
moval is a "Sirocco" centrifical device which has been used commercially to remove dust and
mist from SO2 gas streams from Pyrites burners.
A check in the literature for materials which might catalyze or poison the reaction is
also suggested.
D. High Temperature Scrubber
The high temperature scrubber uses 80% sulfuric acid at about 250 °F to scrub out SO
3*
1-2
-------
H2SO4, and NOX ft"0"1 me oxidized flue gas. By controlling the acid inlet temperature, the
water vapor leaving with exit flue gas can be controlled to equal that entering, thereby eliminat-
ing dilution of acid. Since the amount of water entering per mole of flue gas will change with
time, a suitable means for control must be developed.
One alternate would be to shift the temperature of the scrubbing acid to keep the partial
pressure of the water vapor in the exit gas equal to that in the entering gas. This might be
accomplished with relatively little lag by use of a suitable heater (and cooler).
Another alternate would be to operate the scrubber at a temperature high enough to
eliminate the maximum amount of water which might enter the system and to inject water (as
required) into the inlet gas to maintain a constant and predetermined partial pressure.
Calculations based on probable changes in water partial pressure will help determine
which system might be preferable and what range of variables should be studied now.
This scrubbing equipment must have the ability to operate efficiently at rates well below
the maximum design throughout; packed scrubbers are not very satisfactory in this regard and
operate efficiently only over a narrow range of throughput, fairly close to the flooding velocity.
Channelling of liquid and vapor at lower rates results in poor contact and diminished efficiency.
The problem becomes much more severe as the diameter of the equipment becomes larger and
the devices to distribute liquid and vapor within the system become more awkward and ineffec-
tive.
This may be overcome by using a number of parallel scrubbers, each using a suitable
high ratio of neiSnt to diameter, and planning to shut off one or more scrubbers from the sys-
tem as throughput is diminished in order to maintain optimum loadings in the remaining operat-
ing columns. In consideration of this, the miniplant work should avoid studying scrubber packed
sections which are abnormally short in relation to their diameter; and such ratio of dimensions
should not be considered for commercial equipment.
It is important to know accurately whether serious mist removal problems may exist,
narticularly regarding the flue gas stream from the scrubber. Mist formation in the reactor
and its presence in the feed to the scrubber is probably of no concern, since the scrubber itself
should be an effective mist remover.
Two approaches appear interesting for eliminating mist. The first is to design the
scrubber for low exit gas velocities so mist formation will be minimized and can be removed
by a simple "demister" in the top of the column. The second is to provide additional mist
eliminators, assuming that it is cheaper to provide this equipment and design the scrubbers
for higher exit velocities.
The Sirocco unit for dust and mist removal should also be considered for scrubber exit
trSiS in case mist is a problem; it also provides additional pressure for moving gas.
It is doubtful that operation in laboratory equipment can give any realistic results con-
cerning mist formation. A review of conventional sulfuric acid technology may provide prac-
tical information on this point.
1-3
-------
The high temperature scrubber has been designed to obtain a given efficiency of NO
X
removal, with the NO,/NO mol ratio being maintained at unity in the entering gas. The effect
of different NO2/NO ratios in the entering gas, and a range of liquid/gas ratios, on the effi-
ciency of the NO absorption must be explored over the maximum probable range of variation.
A
It is very important to know how sensitive the efficiency of scrubbing is to variations in these
ratios. If both SO, and NO removal are relatively unaffected by changes in these ratios, it
will make control of the commercial unit easy and indicate a high probability of success for
this essential part of the Tyco process.
The effect of fly ash upon the scrubber and its operation should be established soon.
Fly ash contains a number of metallic impurities, some soluble in the acid, which could affect
the chemical reaction in the scrubber. It must also be prevented from physically plugging
lines and equipment. A simple test would be to suspend fly ash (or leach it), with 80% sulfuric
acid used in scrubbing, and use this in the miniplant to check results against those obtained
when using pure 80% sulfuric acid.
Fly ash may be scrubbed out in the acid and at some point would have to be removed
from the acid and discarded. The handling properties of typical fly ash in appropriate concen-
trations in sulfuric acid can be checked on the bench, along with characteristics for drainage
filtration, and centrifuging. Simple experiments could give indications whether fly ash cake
has "setting up" or "blinding" properties which would make removal in the plant difficult and
expensive. Such experiments should check properties of filtered, neutralized (or alkaline)
fly ash as well as for abnormal handling problems.
For a given plant, the gas rate through the scrubber is fixed between certain practical
limits. Work in the miniplant should determine the minimum acid scrubbing rate (L/G) needed
to attain specified removal of SO- and NO and performance over a range of higher L/G ratios
These figures are needed for optimizing design of the scrubbing unit and minimizing acid cir-
culation rate.
Another variable planned for miniplant study is the effect of operating temperature upon
efficiency over a practical temperature range.
Past runs have been explored at maximum concentrations of NO^Q in tne inlet gas, and
the scrubber has not been able to reduce exit concentration to the desired limit. For assurance
in design, it is desirable to know what happens at very dilute concentrations of N_O, approach-
ing design concentration in exit gas. Additional runs in the miniplant are suggested with very
low concentrations of NgOg in the gas inlet, other conditions of operation being similar to pre-
vious runs.
E. Catalytic Oxidizing Stripper
The heart of the Tyco catalytic process is the catalytic oxidizing stripper, which in-
volves most of the unique and developmental features of the process. The stripper accom-
plishes two things: first, it contacts the sulfuric acid-nitrose solution from the high tempera-
ture scrubber with oxygen (air) over an activated charcoal catalyst, so that HNSOg is oxidized
1-4
-------
and hydrated to NO2 and H2SO4; second, it strips NOg from the sulfuric acid to produce sale-
able acid. To date, the most successful operation has been with use of 4 x 10 mesh pellets of
No. 256 activated charcoal obtained from Witco.
For the small equipment used to date, this size charcoal packing is appropriate, but
would have disadvantages for industrial scale use. First, the mesh size is very small for use
in large equipment for countercurrent gas-liquid contacting and may result in high pressure
drop and inefficient vapor-liquid contact. Second, the material is mechanically weak and fri-
able, and it may break up and dust in beds deeper than 2 ft. (The manufacturers claim that
it has been run in deep beds, and this is usual for decolorization of liquids; but conditions are
quite different for its use as a tower packing for countercurrent contact of gas and liquid.)
A search of other available activated charcoals is suggested to try to obtain some of
larger size, better mechanical strength, minimum friability and dusting tendencies, and long
catalyst life.
Such exploration of their available activated charcoals can use the bench scale single
tube screening test already used to select activated carbon from other potential catalysts.
Catalyst life requires continued research for selected catalysts of good initial character-
istics. Current studies show negligible loss of activity after passing 78 volume acid/volume
catalyst.
It is recommended that future experimentation include determination of the effect of fly
and soot on catalyst life and activity. In particular, a few simple tests to delineate whether
impurities will impose any new problems or restraints are desirable at an early date.
Past experiments on Witco activated charcoal have shown that scrubber operating tem-
perature (at least 290 °F) and air rate must be maintained at suitable levels to prevent rapid
loss of catalyst activity. Obviously, if new alternate catalysts seem worth study, this type of
•nforrration must be obtained for them. In addition, further experiments at higher temperatures
e suggested, with the catalyst used for forthcoming miniplant experiments.
Tests on the oxidizing stripper, using fixed bed charcoal packing, should determine
efficiency of contact over this packing, the effect of changing liquid/gas ratios, and the limiting
apacity of the packing for acid and air flow (flooding characteristics).
It would be desirable in this equipment, as well as in the scrubber, to determine heat
ffects; however, the nature and type of equipment available would appear to render this im-
practical.
Corrosion tests on various metals should also be run in the inlet and outlet from this
Vstem. Any effect of dissolved metal upon catalyst reactivity, acid purity, etc., should be
noted, as well as corrosion rate.
It may also be interesting to test the stripping of nitrose from sulfuric acid solutions
ising bench tests and simple equipment, removing the carbon catalyst and having present,
•nstead, contaminants from fly ash and soot. This test could discover if any useful catalytic
activity came from their presence.
1-5
-------
There may be hazards at some temperature levels in having an oxidizable bed of char-
coal in contact with air and an oxidizing acid. Small-scale tests are proposed to determine
upper temperature limits which would be safe for operation and to eliminate the possibility of
explosive ignition of the carbon bed.
1-6
-------
APPENDIX II: MINIPLANT OPERATION
A. Overall Approach
During the scrubber/stripper studies, the operators had to record 27 temperatures,
7 fluid flows, 27 heater control settings (controlling 52 individual heaters), and 11 gas sample
concentrations plus 3 complete acid analyses, in addition to keeping track of a variety of pro-
cess equipment indicators (reservoir levels, column drainage levels, cooling water flows,
pump and blower performance, and sample system operation and overall system integrity).
Reactor experiments required 8 temperature points, 6 flow rates, and 10 heater control set-
tings as other checks and analyses. (Data sheets showing the information required during the
experiments are shown in Tables II-l through II-4.) Since it was necessary to record a complete
set of data on an hourly schedule, a large task for 2 men, it was important that the plant be set
_ as conveniently as possible. This was accomplished by installing all controls on the upper
level of the two-level miniplant. The only tasks that had to be performed on the lower level
were the sampling of the acid effluent from the 19-ft scrubber and the enlarged vertical reactor
(these, too, could have been taken on the upper level, but there would have been a concentration
lag due to the volume in the reservoirs). To further simplify operation, a closed circuit tele-
ision system was installed to monitor the reservoir levels on the lower plant level, with the
level controls being located on the upper level.
B. Gas Sampling System
Gas samples were taken at two points in the reactor system (as shown in Fig. 6) at the
"nlet and at the outlet. Samples were required at five locations in the scrubber/stripper studies:
(1) inlet to 19-ft scrubber, (2) halfway up the 19-ft scrubber, (3) exit from the 19-ft scrubber
(feed to 7.5-ft scrubber), (4) exit from the 7.5-ft scrubber, and (5) exit from the strippers. In
order to take these samples and analyze them in one system of analytical instrumentation, a
manifold system had to be arranged as shown in Fig. II-l.
The system utilized two UV cells for analysis of NO2 to give accurate analysis at two
oncentration levels which differed by an order of magnitude, 1% full scale and 10% full scale.
These ranges were expected from the scrubber system and stripper system, respectively.
II-l
-------
Table II-1. Reactor Data Sheet
DATE:
RUN:
TIME
FLOWS
Infrared Sample
Total Sample
O2: psig
SCFM
NO: psig
SCFM
Low Air: psig
SCFM
SO2: psig
SCFM
High Air: SCFM
H2S04
HNSOg Concentration
TEMPERATURES
Blower Inlet
Inlet Reactor 1
Ultraviolet Analyzer
Acid Feed
Middle Reactor 1
Bottom Reactor 1
Middle Reactor 2
Middle Reactor 3
VOLTAGES
Middle Reactor 1
Bottom Reactor 1
Middle Reactor 2
Bottom Reactor 3
Top Reactor 3
Acid Feed
Top Reactor 2
Inlet to Reactor 1
Bottom Reactor 2
Top Reactor 1
CHECKLIST
Gas Burner
Cooling Water
Drains
Acid Level
OPERATOR
II-2
-------
TEMPERATURE
Scrubber 2 Reservoir
Top of Stripper No. 1
" No. 2
" No. 3
Middle of Stripper No. 1
" No. 2
" No. 3
Bottom of Stripper No. 1
11 No. 2
» No. 3
Acid Feed: Lower
Middle
Top
Stripper Reservoir
VARIACS
Acid Feed Line
Acid Feed (Res)
Acid Feed Tapes
Stripper Heaters Top
Middle
Bottom
Acid Feed
Top Stripper Tapes
FLOWS
Acid
Air to Stripper
CHECKLIST
Res. Level
Exit Level
Flows
OPERATOR
Table II-2. Stripper Data Sheet
Run No. Date
Time
Sheet No.
II-3
-------
Table II-3. Scrubber No. 2 Data Sheet
Run No. Date Sheet No.
Time
TEMPERATURES
Main Air Feed
Inlet Air
Acid Feed: Top
Middle
Bottom
Feed Res (Scr 4 Res)
Sample Line
CONTROL VARIACS
Main Air
Gas Feed to Scr Bottom
Gas Feed tc Scr Bottom
Top (1 ft Mantle)
Top (4 ft Mantle)
Top/middle (4 ft Mantle)
Middle (1 ft Mantle)
Middle/Bottom (4 ft Mantle)
Bottom (4 ft Mantle)
Bottom (1 ft Mantle)
Acid Feed: Immersion Heater
Acid Feed: Tapes
Reservoir
FLOWS
Main Air Meter Reading
Main Air Flow
Acid
NO to Oxidizer Meter Reading
Air to Oxidizer Meter Reading
SO2 Feed Flow Meter Reading
CHECKLIST
Reservoir Level (TV)
Gas Burner
OPERATOR
II-4
-------
Table II-4. Scrubber No. 4 Data Sheet
Run No. Date Sheet No.
Time
TEMPERATURES
Acid Feed
Gas Feed: Top
Middle
Bottom
Feed Res (Stripper Res)
CONTROL VARIACS
Feed Res (Stripper Res)
Bottom Mantles
Middle Mantles
Top Mantles
Acid Feed: Top Immersion
Acid Feed: Tapes
Gas Feed
Reservoir
_FLOWS_
Acid Feed
CHECKLIST
Reservoir
Heater
Column Exit
Leaks
Sample Box
Sample Line
OPERATOR
II-5
-------
From top of
strippers
From
before
reactor
DuPont \JV Photometer
f ~ ~1?" ceH ~ 11/7" -c<
n/7" ~ein-
Manometer
n
Air
pump
Flow meter
Flowmeter
T6
Calibration gases
O \»— From top
Vp of scrubber
From after
reactor
Recorder
Dew-
point
Hygron
?
Beckman
IR Analv:
t
eter
Flowmeter
Condenser
To before reacto
To stack
T
To after
reactor
Fig. II-1. Flowsheet of gas analysis system.
-------
The gas was pumped out of the miniplant with a small diaphragm pump (about 0.5 SCFM)
and fed to the system shown in Fig. II-1. The gas analysis system was set up so that the gas
could be returned to the miniplant after it had been analyzed to avoid intermittant fluctuations
in process flow levels.
The gas sample was controlled at constant temperature and pressure in the analyzers.
This was more important than constant flow rate since the duPont and Beckman instruments
both based their analysis on the number of molecules in the path of the radiation beam. A
manometer and temperature readout were installed in the sample system near the UV photo-
meter.
The most important feature of the gas sampling system was the control of the condition
of the sample. It was necessary to avoid condensation of liquid in the sample lines, a condition
which had been known to cause erroneous results in sample analysis. The sample lines were
Teflon tubes encased (two or three at a time) in aluminum tubes which were wrapped with heat-
ing tapes. In this manner, the gas sample could be kept at about 275 °F while it was outside
the miniplant.
If the temperature dropped much below this level, condensation was formed with cor-
roded the stainless steel fittings used in the sample system. The corrosion products tended to
block the sample line, resulting in either a high pressure drop (making it difficult to maintain
the proper pressure in the system) or a complete blockage of the lines. The system was set
up so that an individual line could be isolated and washed with water without removing it from
the aluminum tubes. Although the system was workable as used, it was concluded that a more
desirable system would utilize individually traced lines. These are more expensive than the
system used, but they involve far less maintenance.
Nitric oxide was analyzed with the Beckman infrared analyzer, as shown in Fig. II-1.
The primary consideration in this analysis was the removal of water from the gas sample
before it reached the analyzer because water interferes with the quantitative analysis of NO
in the infrared region of the spectrum. To remove the water, the sample was passed through
a cold trap. Experience showed that a large U-tube filled with saddles and immersed in alcohol
at -35 °C would drop out most of the water without removing significant amounts of nitrogen
oxides. The trap had a tendency to plug up with ice several times a day, but a quick-disconnect
system allowed the replacement of the trap in a few minutes and very little time was lost in
this type of maintenance. An operator learned to recognize the sudden erratic behavior of the
Beckman analyzer output as the symptom of a plugged cold trap.
C. Data-Taking Procedures
As noted above, a complete set of data was taken every hour during a run. A run was
not considered to be complete until the essential data were constant for two consecutive sets
of readings. The temperatures were taken off the printout of the 24-point recorders and logged
onto the data sheets shown earlier. The recorder charts themselves were not retained. Based
on the temperature readings, heating rates were adjusted whenever necessary.
H-7
-------
Gas sampling was performed continuously, with SOg, NO,, and NO being analyzed at
each sample point. A gas concentration was considered to be meaningful if the output of the
analyzer instrument was constant for 5 to 10 min. Since SOg and NO2 were analyzed on the
same instrument, it took several minutes to get a complete analysis at any one sample point
Acid samples were taken from all required points within a few minutes of each other
and analyzed later for acid strength (percent sulfuric acid) and nitrosylsulfuric acid content
In this manner it was hoped that all the data would come close to giving the process condition
at one particular time. The acid strength was obtained with a hydrometer (correcting for tem-
perature) and the nitrosylsulfuric acid content determined by titration with potassium perman-
ganate in acid solution:
5 HNSO. + H MnO, 4 2H.O -» 5H.,SO4 -f HNO, + 2Mn(NO0)0 .
0 1 £i 61 O 0 ^
The acid analysis is quite rapid, except in very dilute solution where it takes as much
as 15 min for the color change to become permanent. With HNSO- concentrations above 0.1%
the titration takes less than 5 min and is quite accurate and reproducible. At lower concentra-
tions the variation between replications can be as much as 5%.
The NOx content of the acid can be determined by an alternate procedure: the nitrometer
In this case the reducible nitrogen oxides (NO,"", NO,") are reduced with mercury to NO gas
which is evolved from the solution. The volume of the NO is measured and the original amount
of HNSO- can be calculated if it is assumed that all the NO was present as N000. This is a
o x 2 3
very accurate method but is very time consuming because it takes a long time for the mercury to
react with all the NO in the acid solution. In addition, it is not too accurate for very dilute
solutions since a small error in reading the colume of NO can make a sizable error in the
HNSOg concentration. All analyses performed during the normal course of miniplant experi-
mentation was done by permanganate titration with occasional checks with the nitrometer on
the more concentrated samples.
Fluid flows were recorded once an hour, but were monitored frequently by the miniplant
operators. This was accomplished by observing rotameter settings as well as liquid levels in
the reservoirs.
D. Shift Operation
Experimentation in the miniplant was performed in continuous "batch" operations. The
equipment was run in a continuous fashion for several hours at a time to accomplish the goals
of a particular experiment. Then the conditions were changed and another "batch" experiment
was performed. This type of testing might have lent itself to an 8-hr day program, except for
the fact that it took several hours to bring the plant up to temperature. Because of limitations
in the heating equipment and the necessity for slow heating rates on the all-glass piping and
columns, it took 4 to 8 hr to achieve thermal steady state.
Due to these requirements and the desire to accomplish as much experimentation as
possible during the contract period, the plant operation was conducted on a two or three shift,
II-8
-------
5-day week basis. Two operators were present at all times during the three shift work. For
the last several weeks of the contract period, the plant was operated for two shifts per day to
permit more time during the experimentation for data analysis. The plant was maintained at
about 200 °F during the 7-hr period that it was unattended, which cut down the heating time to
under 4 hr. This left about 13 hr for experimentation and shutdown. The two shift operation
had another advantage in that it provided at least 4 hr every day for equipment maintenance and
sample system cleanout. This permitted smoother plant operation during the experimental
period.
II-9
-------
APPENDIX III. DATA FROM MINIPLANT REACTOR EXPERIMENTATION
III-l
-------
Table III-l. Miniplant Reactor Experiment— Irrigation
with Nitrose [Three Reactor Columns (1.6
Packed with 3/8-in. Intalox Saddles, Acid Feed
4.4% HNSOg in 70% H2SC>4 at 20 cc/min]
Run No.
Reactor temperature, °F
Gas flow rate (SCFM)
Feed SO, (ppm)
Feed NO, (ppm)
Feed NO (ppm)
Feed H00 (%)
£i
Outlet SO, (ppm)
Outlet N02 (ppm)
Outlet NO (ppm)
Reactor residence
time (sec*)
SOp removal efficiency, $
30A 308 30C
225 225 225
222
3450 3450 3450
4000 10,400 4000
2325 6000 2325
~6 ~6 ~6
650 700 450
3700 16,400 -5500
4700 6800 4850
34.2 34.2 34.2
70.5 87.4 84.2 94.0 91.3
29A
250
2
3550
4000
2325
~6
1200
3900
3850
33.2
29B
225
2
3550
4000
2325
~6
600
4200
5000
34.2
SOD
200
2
3450
4000
2325
~6
150
5300
5300
35.8
30E
200
2
3450
10,400
6000
~6
400
11,800
6600
35.8
100
100
* Sample line residence time was less than 1 sec using the short LJV cell.
** NO, interference was taken into account.
III-2
-------
I
CO
Table 111-2. Miniplant Experiments — Irrigation with
Nitrose [Three Reactor Columns (1.6 ft3)
Packed with 3/8-in. Intalox Saddles*]
Run No.
Reactor temperature (°F)
Gas flow rate (SCFM)
Gas feed:
SO. (ppm)
Ndn (ppm)
NO (ppm)
Acid flow rate (cc/min)t
Acid feed:
HNSOg (%)
H2SO4 (%)
Gas effluent:
SO2 (ppm)
NOo (ppm)
NO (ppm)
Acid effluent:
HNSOjj (%)
H2S04 (%)
Reactor residence
time (sec)
Reaction yield (%)
NOx material balance
error flb) t
* Reactors 1 and 3
f Total flow to all
(Material balance
31
250
2
3450
0
0
15
4.3
70.0
1900
2900
4250
0.33
66.5
33.2
45.0
-134
250
2
3450
6400
2725
15
4.3
70.0
1100
6100
4700
0.27
64.4
33.2
68.0
10
250
2
3450
6400
2725
60
4.3
70.0
760
16,900
6200
0.78
69.0
33.2
78.0
-16
were in cocurrent flow rate;
3 reactors.
NO
r-rwT
(in) —
N0x
N0v (out)
X
(in)
225
2
3450
6400
2725
60
4.3
70.0
550
18,400
7000
—
34.2
84.0
reactor 2
- x 100.
32A
250
2
3100
5800
2375
25
1.12
70.8
1150
4000
4400
—
33.2
62.9
—
250
2
3100
5800
2375
50
1.12
70.8
850
4400
4100
—
33.2
72.6
—
250
2
3100
5800
2375
100
1.12
70.8
750
4900
1500
0.64
71.3
33.2
76.5
30
was countercurrent.
32B
250
2
135
0.63
68.7
33.2
74.1
250
2
2900 2900
6600 6600
2650 2650
240
0.63
68.7
750 600
8500 13,400
5000 6000
0.37
64.7
33.2
79.3
-42
33A
250
2
3425
0
0
32
1.23
78.9
2900
250
1050
—
33.2
15.5
—
250
2
3425
5500
1700
32
1.23
78.9
2150
3900
2550
1.54
76.3
33.2
37.2
2
250
2
3425
5500
1700
165
—
1700
3000
1800
1.39
76.3
33.2
50.1
6
250
2
3425
0
0
165
—
2650
2900
—
33.2
22.9
—
-------
Table III-2. (Continued)
Run No.
Reactor temperature (°F)
Gas flow rate (SCFM)
Gas feed:
SO2 (ppm)
NO2 (ppm)
NO (ppm)
Acid flow rate, cc/mint
Acid feed:
HNSOg (%)
H2S04 (%)
Gas effluent:
S02 (ppm)
NO2 (ppm)
NO (ppm)
Acid effluent:
HNSOg (%)
H2S04 (%)
Reactor residence
time (sec)
Reaction yield (%)
NOX material balance
error (%) t
* Reactors 1 and 3
t Total flow to all
1 Material balance
33B
300
2
3050
6000
2050
60
1.27
79.6
2550
4600
3100
—
29.9
16.4
were in
300
2
3050
6000
2050
160
1.27
79.6
2200
4100
3075
—
29.9
27.9
cocurrent
300
2
3050
6000
2050
160
1.27
79.6
2150
4000
3350
1.03
77.3
29.9
29.5
14
flow rate,
34
260
2
3500
7000
2050
15
1.46
60.0
2100
3400
2900
—
32.9
40.0
reactor 2
260
2
3500
0
0
15
1.46
60.0
2900
—
32.9
17.1
260
2
3500
7000
2050
57
1.46
60.0
1250
4100
4700
—
32.9
64.4
260
2
3500
7000
2050
150
1.46
60.0
550
6100
6000
—
32.9
84.3
260
2
3500
7000
2050
185
1.46
60.0
425
6100
6600
—
32.9
87.8
35
250
2
3200
6500
2400
60
1.13
53.0
200
—
33.2
93.8
250 250
2 2
3200 3200
6500 0
2400 0
60 60
1.13 1.13
53.0 53.0
50 600
— —
33.2 33.2
98.4 81.3
36
250
2
3200
5050
1775
15
1.13
71.6§
1200
—
33.2
62.0
37
250 250 250
2 22
3200 3150 3150
5050 0 0
1775 0 0
50 25 80
1.13 0 0
71.6§ 79.8** 79.8**
2200 3150 3150
— — —
33.2 33.2 33.2
31.7 0 0
was countercurrent.
3 reactors.
error =
NOx (in) -
- NO (out)
- x mn
(in)
§2.24%HNO3 added.
**2.95% HNO, added.
o
-------
Table III-2. (Continued)
en
Run No.
Reactor temperature (°F)
Gas flow rate (SCFM)
Gas feed:
SOn (ppm)
NCJ2 (ppm)
NO (ppm)
Acid flow rate (cc/min)
Acid feed:
HNSOg(%)
H2SO4 (%)
Gas effluent:
SO2 (ppm)
NO™ (ppm)
NO (ppm)
Acid effluent:
HNSO5 (%)
H2SO4 (%)
Reactor residence
time (sec)
Reaction yield (%)
NOX material balance
error ft)
38
250
2
3100
4700
1225
20
1.17
71.25
1400
3900
2600
1.58
74.9
33.2
54.8
-14
38A
250
2
2950
5500
1775
60
1.04
70.7
1150
4200
2900
—
—
33.2
62.7
—
* Reactors 1 and 3 were in cocurrent flow
t Total flow to all 3 reactors.
j Material balance
NO
250
2
2950
5500
1775
140
1.04
70.7
400
6100
4400
—
—
33.2
86.4
—
250
2
2950
5500
1775
260
1.04
70.7
150
6800
4625
—
—
33.2
94.9
—
rate, reactor 2
250
2
2950
0
0
260
1.04
70.7
500
3600
4400
—
—
33.2
83.1
—
39
250 250 250
2 22
3200 3200 3200
5500 5500 5500
1900 1900 1900
33 50 181
1.26 1.26 1.26
80.4 80.4 80.4
2000 2250 1850
— — —
— — —
— — —
33.2 33.2 33.2
37.5 29.5 42.2
— — —
40
250
2
3200
6100
2100
70
2.06
79.3
1100
3100
2100
—
—
33.2
65.6
—
250
2
3200
6100
2100
140
2.06
79.3
950
4000
3100
—
—
33.2
70.3
—
250
2
3200
6100
2100
260
2.06
79.3
500
2500
2325
—
—
33.2
84.4
—
250
2
3200
0
0
260
2.06
79.3
1150
300
4250
—
—
33.2
64.0
—
was countercurrent.
(in) — NO., (out)
NO (in)
A
x 100.
50.45% HNO, present.
o
-------
Table III-2. (Continued)
Run No.
Reactor temperature (°F)
Gas flow rate (SCFM)
Gas feed:
SO2 (ppm)
N02 (ppm)
NO (ppm)
Acid flow rate (cc/min) t
Acid feed:
HNSO5 (%)
H2S04 (%)
Gas effluent:
S02 (ppm)
N02 (ppm)
NO (ppm)
Acid effluent:
HNSO5 (%)
H2SO4 (%)
Reactor residence
time (sec)
Reaction yield (%)
NOX material balance
41
250
4
3850
7500
2900
55
1.0
71.1
3000
900
3625
—
16.6
22.1
—
250
4
3850
7500
2900
55
1.0
71.1
2275
5800
4400
0.59
72.9
16.6
40.7
6
250
4
3850
7500
2900
120
1.0
71.1
1175
7500
5400
0.96
72.6
16.6
69.6
-18
250
4
3850
7500
2900
280
1.0
71.1
550
8500
6000
0.70
71.2
16.6
85.5
-11
250
4
3850
7500
2900
500
1.0
71.1
0
7500
5600
0.86
69.6
16.6
100
-4
250
4
3850
0
0
500
1.0
71.1
375
5500
5900
16.6
90.3
—
42
250
4
3150
5600
1950
120
0.91
75.8
2150
—
16.6
31.8
—
250
4
3150
5600
1950
255
0.91
75.8
1850
16.6
41.5
—
43
250
8
3050
6100
2325
110
1.06
69.7
2050
z
8.3
32.8
—
250
8
3050
6100
2325
280
1.06
69.7
1400
z
8.3
54.1
—
44
250
4
3200
6500
2175
64
0.53
70.0
1775
4350
3225
0.45
70.0
16.6
44.6
13
250
4
3200
0
0
64
0.53
70.0
2825
200
950
z
16.6
11.7
—
250
4
3200
6500
2175
122
0.53
70.0
1400
4600
3500
0.45
71.0
16.6
56.3
9
250
4
3200
6500
2175
257
0.53
70.0
1225
5000
3850
0.42
70.0
16.6
62.0
5
250
4
3200
6500
2175
484
0.53
70.0
1075
4725
3750
0.39
70.6
16.6
66.6
13
error (%) t
* Reactors 1 and 3 were in cocurrent flow rate, reactor 2 was countercurrent.
t Total flow to all 3 reactors.
NO (in) —NO (out)
t Material balance error = - x 100.
N0x (in)
-------
Table III-2. (Continued)
Run No. 45
Reactor temperature (° F) 250
Gas flow rate (SCFM) 6
Gas feed:
S02 (ppm) —
NOjj (ppm) —
NO (ppm) —
Acid flow rate (cc/min)f —
Acid feed:
HNSO5 (%) —
H2SO4 (%) —
Gas effluent:
SO2 (ppm) —
N02 (ppm) —
NO (ppm) —
Acid effluent:
HNSOg (%) —
Reactor residence —
time (sec)
Reaction yield (%) —
NOX material balance —
error (%) t
46$
250
2
3300
6400
1800
15
1.12
71.5
2550
6100
3850
1.64
72.8
24.9
22.7 **
-23
250
2
3300
6400
1800
42
1.12
71.5
2300
6700
4025
1.85
80.4
33.2
30.3
SeTT
-38
250
2
3300
6400
1800
57
1.12
71.5
2400
6700
3975
2.1
74.9
33.2
27.7
TO
-47
250
2
3300
6400
1800
89
1.12
71.5
1500
7700
4625
1.36
77.9
33.2
54.6
TO
-39
250
2
3300
6400
1800
134
1.12
71.5
1400
8600
5000
0.81
67.7
33.2
57.6
-22
250
2
3300
6400
1800
248
1.12
71.5
900
8125
4025
0.65
72.9
33.2
72.9
T5o
9
250
2
3300
0
0
248
1.12
71.5
975
6200
4850
33.2
70.6
SO
* Reactors 1 and 3 were in cocurrent flow rate; reactor 2 was countercurrent.
f Total flow to all 3 reactors.
Material balance error
N0x (in) — N0x (out)
N0x (in)
x 100.
475
250
4
3250
6200
1950
29
0.96
73.4
2925
4950
3500
—
16.6
10.0 **
2O
—
250
4
3250
6200
1950
82
0.96
73.4
2600
5000
3550
0.62
16.6
20.0
2O
3
250
4
3250
6200
1950
179
0.96
73.4
2200
4600
4000
0.67
74.4
16.6
32.4
Z3TO"
7
250
4
3250
6200
1950
220
0.96
73.4
1950
4900
4025
0.63
73.7
16.6
40.0
5O
8
§ One reactor: cocurrent operation — effluent data for one column only.
** The yield values represent SO2 removal rates after one irrigated column (above the slash line) and after the third reactor; i.e., one
irrigated reactor and unirrigated, but wetted packed columns (below the slash line).
-------
Table III-2. (Continued)
Run No. 485 49 §
Reactor temperature (°F) 250 250 250 250 250 250 250
Gas flow rate (SCFM) 2222 44 4
Gas feed:
SO, (ppm) 3050 3050 3050 3050 3450 3450 3450
NCT2 (ppm) 5850 5850 5850 5850 5900 5900 5900
NO (ppm) 1950 1950 1950 1950 2200 2200 2200
Acid flow rate (cc/min) t 29 59 136 204 30 180 350
Acid feed:
HNSO5 (%) 0.95 0.95 0.95 0.95 0.95 0.95 0.95
H2SO4 (%) 72.7 72.7 72.7 72.7 72.7 72.7 72.7
Gas effluent:
SO, (ppm) 2175 1550 1275 1200 — — —
NOJj (ppm) 4600 5200 5300 5600 5500 4500 4400
NO (ppm) 3750 4700 5050 5100 3350 3900 2650
Acid effluent:
HNSOg (%) 0.54 0.56 0.79 0.89 0.71 0.67 0.69
H2S04(%) 75.5 74.7 73.9 73.9 71.0 70.9 71.2
Reactor residence 33.2 33.2 33.2 33.2 16.6 16.6 16.8
time (sec)
Reaction yield (%) 28.7 49.0 58.5 60.0 9.9 27.3 31.7
NOX material balance 1 -8-10-13 -6 8 20
error (%) $
* Reactors 1 and 3 were In cocurrem flow rate; reactor 2 was countercurrent.
t Total flow to all 3 reactors.
NO (in) — NO (out)
t Material balance error = —u/. .. ,x x 100.
x 'ln'
§One reactor was countercurrent.
m-8
-------
Table III-3. Miniplant Reactor Experiments — Irrigation with Nitrose (One 15-ft Reactor in Counter current
Operation Reactor Packed with 3/8-in. Intalox Saddles)
i
to
Run No.
Reactor temperature (°F)
Gas flow rate (SCFM)
Gas feed:
SO2 (ppm)
NO, (ppm)
NO (ppm)
Acid flow rate (cc/min)
Acid feed:
HNSOg (%)
Gas effluent:
SO2 (ppm)
NO2 (ppm)
NO (ppm)
Acid effluent:
HNSOg (%)
Reactor residence
time (sec)
Reaction yield (%)
NOX material balance
error (%)
dc\ f n«-n.-iril It'll 0 «*-•<=»
52
210
2
3500
5800
1900
330
1.9
76.0
850
7200
4700
1.2
76.3
29
76.0
+ 21
208
2
3500
5800
1900
330
1.8
72.0
750
5100
3850
1.0
72.3
29
78.5
+ 33
NOx(in) -
52A
218
2
3100
6250
1775
320
1.10
69.4
750
10,800
5010
0.77
69.5
28
75.9
-7
NOX (out)
222
2
3100
6250
1775
320
0.78
72.1
550
4350
4100
0.47
72.4
28
79.0
+ 23
100
0.8
71.3
1350
4300
4200
0.39
71.3
14
51.7
-1
20
1.29
73.1
100
4000
3750
0.81
72.3
28
98.5
+ 12
25
1.26
73.1
300
4650
4200
0.89
72.9
28
89.7
+ 1
35
1.53
73.1
100
5600
4250
0.93
73.0
28
96.4
-2
35
1.53
73.1
325
5700
4025
0.93
73.0
28
88.8
-1
230
0.06
68.1
-1700
4100
3550
0.14
68.4
27
42.5
-10
180
2.13
69.8
1850
18,000
3100
0.20
70.4
27
43.9
+ 18
NOX (in)
fUnstahle system.
-------
Table III-3. (Continued)
I
H»
o
Run No.
Reactor temperature (°F)
Gas flow rate (SCFM)
Gas feed:
SO, (ppm)
Ntt (ppm)
NOlppm)
Acid flow rate (cc/min)
Acid feed:
HNS05 f%)
H,SO. (%)
2 4
Gas effluent:
S02 (ppm)
NO2 (ppm)
NO (ppm)
Acid effluent:
HNS05 (%)
Reactor residence
time (sec)
Reaction yield (%)
NOx material balance
error {%)
'Material balance
56B
217
2
3300
5900
1050
200
1.66
69.8
1000
15,000
2900
0.26
69.8
28
66.6
+ 18
N0x
56C
201
2
3300
5900
1050
210
1.86
69.4
700
12,000
2900
0.23
69.8
29
78.8
+ 38
(in) - N0x (out)
N0x (in)
57A
201
2
3000
5875
2200
45
1.65
67.8
I100J
15,000
~6000
0.16
68.9
29
63.2
-76
57A'
200
2
3250
0
0
35
2.16
70.1
2300
5000
0.25
67.3
29
29.3
200
2
3250
6000
1800
35
2.16
70.1
500
4500
5000
0.25
67.3
29
84.6
+ 16
57B
225
2
3250
6000
1800
50
2.69
69.5
800
6000
5050
0.55
73.6
28
75.5
+ 18
226
2
3250
6000
1800
35
—
875
8000
5050
—
28
74.4
57B'
225
2
3300
6100
2600
35
2.55
71.6
250
2400
6000
0.21
70.4
28
92.5
+ 35
57C
250
2
2675f
6200
2400
40
2.40
70.0
1450f
8600
7000
0.82
76.6
27
45.9
-25
58
220
2
2575f
7500
3175
30
1.01
72.4
150f
7500
2550
0.62
72.7
28
94.3
+ 10
220
2
2625f
7500
3175
30
1.19
72.6
250f
7500
2550
0.59
72.7
28
90.5
+ 13
59
225
4
3200f
6400
2150
35
0.92
72.0
450f
3800
2800
0.52
72.3
14
79.6
+ 25
tCorrected for zero error.
^Instruments in error.
-------
Table III-4. Mlniplant Reactor Experiments — Irrigation with Nitrose (One 15-ft
Reactor in Cocurrent Operation Reactor Packed with 3/8-m. Intalox
Saddles)
Run No.
Reactor temperature (°F)
Gas flow rate (SCFM)
Gas feed:
SO2 (ppm)
N02 (ppm)
NO (ppm)
Acid flow rate (cc/min)
Acid feed:
HNSO5 (%)
H2S04 (%)
Gas effluent:
SO2 (ppm)
N02 (ppm)
NO (ppm)
Acid effluent:
HNS05 (%)
H.SO. (%)
2 4
Reactor residence
time (sec)
Reaction yield (%)
NOV material balance
60
228
4
2450t
6000
2850
205
1.3
71.0
400t
3550
0.8
72.1
14
83.8
—
228
4
2450f
6000
2850
200
1.5
71.5
400t
4100
3425
1.08
71.0
14
83.7
421
213
4
3200f
6400
1800
200
1.93
72.2
250f
7100
~5500
0.61
71.2
14
92.4
+ 15
61
221
4
3300f
6300
3025
30
1.83
70.3
soot
4900
5000
0.31
70.4
14
75.7
+ 6
62A
222 J
4
2875 1
6000
1000
305
1.29
70.4
350f
6600
2375
0.78
70.7
14
87.7
+ 22
62B
222 {
4
2575f
6160
2250
380
2.65
71.3
150f
12,500
11,400
1.27
72.3
14
94.3
-3
erVor (%)
NOx (in) - NOX (out)
*Material balance error =
N0x (in)
tCorrected for zero error.
JReactor temperature: top - 195 °F, middle - 234 "F, bottom - 230 °F.
-------
APPENDIX IV. DATA FROM SCRUBBER/STRIPPER EXPERIMENTS
IV-1
-------
Run No.
NA
HD
Gas
Acid
SCFM
GPM
SCFM
67
Table IV-1. Scrubber/Stripper Data
68 (Repeat of 67)
240/ 230/ 230/ 235/ 265/
300 300 300 300 300
240/ 240/ 250/ 250/ 240/
290 300 310 315 310
5.6
[C 1
5
[C ]
24
[CNQ ]
2
NO Scrubbing
rSoJ
NO Mail. Bal.
X
[c 0]
2
1
2
3
1
2
3**
Flue Gas
1
2
3
Efficiency
Flue Gas
1
2
3
Error
Flue Gas ***
Air
wt%
"
11
wt%
1'
"
ppm
"
"
ii
ppm
It
if
ir
%
—
3900
—
3400
—
—
1300
950
erratic
—
~6
~1
5.6 5.6 5.6 5.6
0.22 0.22 0.22 0.22 0.22
0.08 0.11 0.22 —
0.34 0.37 0.76 —
0.08 0.09 0.18 —
— 76.6 — —
— 78.0 — —
— 77.4 — —
4150 3450 3600 3050
875 400 500 0
1000 400 0 350
71.1 73.2 68.3 66.4
1300 2600 2375 2735 3025
1075 1150 1500 2050
1225 1400 1900 1850
— 8.9 -1.9 -51.8 —
~6 ~6 ~6 ~6
A
145/
205
ISO/
200
5.6
0.24
—
B
140/
285
160/
190
5.6
0.24
0.30
74.7
C*
1807
285
170/
220
5.6
0.22
—
—
D
2007
285
2007
220
5.6
0.22
0.17
0.25
0.08
76.0
76.0
76.0
E
2457
280
2007
220
5.6
0.22
0.15
0.23
0.06
76.0
76.0
76.0
F
2457
280
2007
220
5.6
0.22
—
—
G
250/
278
230
5.6
0.2
0.23
0.29
0.14
76.0
76.0
76.0
H
2507
278
230
5.6
0.2
—
—
I
2457
275
235
5.6
0.2
0.16
0.27
0.06
76.3
76.3
76.3
J
2457
275
235
erratic
0.2
0.59
0.35
0.18
75.5
82.1
75.7
3300 3900 3200 3150 3300 2300 3100 3300 3450 3450
800 375 3175 2100 2000 1800 2325 2050 1450 —
2500 — 5050 2900 3175 — 3025 — 2000 5050
79.0 90.0 21.0 39.6 58.8 51.8 40.6 36.0 52.T 61.0
2900 2950 4100 3975
500 300 2600 2200
500 — 3500 2400
4700 4325 4100 3500 3625 3350
1300 1400 1900 2300 1900 —
1950 — 2200 — 2500 5100
— — — 2.3 22.9 —
""O "-6 —6 -^G -^6 ~6
9.3 — 9.1 —
~6 ~6 ~6 ~6
•f
Key to Effluent Streams: 1 - N0x scrubber 4; 2 - NO scrubber 2; 3 - stripper.
*Plant shut down after 68E and started up again. Sample line problems (this is where
the Teflon sample lines fused and gave us difficulty).
**No SO2 was fed to the scrubbers during Runs 68-72.
***No H2O analyses were made on the scrubber and stripper effluent gases.
-------
Table IV-1. (Continued)
<
CO
Run No.
NA
HD
-'Flue Gas
[CHNSOC
"N0n
1
2
3
1
2
3
Flue Gas
1
2
3
NO Scrubbing Efficiency
[CNO] Flue Gas
1
2
3
NO Mail. Bal. Error
fCH n] Flue Gas
H2U Air
op
°F
SCFM
GPM
SCFM
wt%
II
wt%
M
ppm
ppm
70(2
A
230/
260
250/
270
5
0.19
0.03
0.51
78.2
78.0
2800
400
700
27.5
3650
1050
950
-43.7
C
~*> 1
Strippers)
B
2357
265
255/
275
5
0.16
0.03
0.51
78.2
78.0
2900
550
900
76.4
3350
925
875
— fi
-1
C
2357
265
255/
275
5
0.16
0.08
1.05
78.4
78.6
—
2950
700
1000
77.4
3425
750
875
__
-6
«
D
240/
270
2407
295
5
0.17
0.42
1.13
0.21
77.9
78.7
78.2
3500
1200
1350
—
71.3
3175
750
850
—
-82.8
~6
E
255/
305
245/
295
5
0.14
—
—
3800
1125
1200
—
72.4
3425
875
950
—
—
— fi
— 1
F
2557
305
245/
295
5
0.14
0.29
0.80
77.3
78.0
78.7
3350
1300
1400
—
65.8
3350
1000
1100
—
—
-6
•f
G
2607
280
245/
295
5
0.16
2.25
0.31
0.81
0.31
77.5
78.0
78.9
3650
1300
1300
—
67.9
3350
950
1075
—
-15.4
-6
«
70 Repeat (2
H
2457
260
-260
5
0.16
—
—
—
—
3100
300
1600
—
74.5
3750
1400
2250
—
—
~6
<•
J
170/
265
330/
280
5
0.18
0.02
0.45*
0.01
—
—
3050
400
600
—
79.0
3350
950
925
—
-20.5
~6
4
Strippers)
K
175/
260
230/
280
5
—
—
—
—
—
—
—
3300
400
1000
1400
82.0
3350
800
1100
1800
—
~6
-1
L**
240/
260
230/
285
5
—
—
—
—
80.4
79.3
78.9
3600
1200
1700
—
58.0
3300
1400
1725
—
—
a
«
71
AT
230/
~250
-240
5
0.18
—
—
—
—
—
—
2900
1900
2700
—
51.4
2550
750
950
—
—
-1
— 1
BT
1707
255
1907
240
5
0.18
0.07*
0.25*
0.02*
77.0
77.0
77.0
3000
2300
2900
—
36.0
2150
1000
1025
—
-27.0
-1
<•
CJ
170/
255
225/
260
5
0.16
0.07*
0.38*
0.05*
77.2
77.2
77.2
3000
2000
2700
—
50.6
2850
925
1050
—
-20.0
~6
~ 1
Key to Effluent Streams:
*L'sing acid blank for analysis,
"Instrumentation problems.
tDry gas feed.
JWet gas feed.
1 - NOV scrubber 4; 2 - NOX scrubber 2; 3 - stripper.
-------
Table IV-1. (Continued)
Run No.
Relative Time hr.
TNA °F
T2 - Avg. °F
T4 - Avg. °F
THD °F
THD - Avg. °F
QFlueGas* SCFM
QAcid CPM
QAir** SCFM
[CHNS051 1 wr%
2
3>t
[CH2SO41 L wt,^
2,
3
[CfjO2l Flue Gas ppm
1
2
3
4
[Cfjol Flue Gas ppm
1
2
3
4
NO Scrubbing Efficiency
10.0ft. %
19.0ft.
26.5ft.
NOx Matl.Bal. Error %
74
A
0
1907
290
252
210
230/
255
247
5.6
0.28
2.55
-
-
~
-
~
~
500
-
50
_
1100
-
550
_
_
_
-
-
B
0.5
190/
290
252
217
2307
255
247
5.6
0.28
2.55
-
-
-
-
-
-
500
-
55
_
1000
-
325
_
_
_
-
-
C
1.0
195/
280
252
217
240/
260
250
5.6
0.27
2.55
-
-
-
-
-
-
450
-
75
_
725
-
375
_
_
_
-
-
D
2.25
195/
280
252
217
240/
260
250
5.6
0.28
3.6
0.02
0.06
0.03
69.8
69.8
69.8
575
110
170
20
450
175
160
150
_
.
71.0
10.0
E
3.75
1957
280
252
207
2407
260
250
5.6
0.28
3.6
-
-
-
-
-
-
580
70
120
:
300
150
100
~
75.0
-
F
5.25
1907
250
227
207
2457
255
250
5.6
0.28
3.6
.008
.015
.019
78 0
77.7
77.3
470
10
70
~
300
110
100
-
84.5
-
G
7.25
1957
250
227
213
245 /
250
250
5.6
0.26
3.6
-
-
-
_
-
-
_
30
-
0
_
100
110
83.2
-
H
8.25
1957
250
231
213
2457
250
250
5.6
0.31
3.6
.007
.008
.005
78.0
77.3
77.9
645
-
40
-
450
_
75
-
89.0
-
75
A
0
2107
290
272
253
245 /
260
255
5.6
0.25
3.4
0.05
256
.010
79 2
77.2
79.2
3100
1000
1300
1350
3725
650
1000
5800
76.0
6.3
B
0.5
2107
290
272
255
2457
260
254
5.6
0.25
3.4
.049
.258
.015
78.2
78.2
78.2
3100
1000
1300
1300
3725
675
1050
6000
75.5
-6
C
1.5
210/
290
272
255
245'
260
254
5.6
0.25
3.4
049
.258
.015
78 2
78 2
78.2
3100
1000
1300
1450
3760
575
950
5900
77.0
-6
D
3.0
2157
275
252
232
255 /
265
258
5.6
0.25
1.9
070
320
.008
77 9
77 '9
78.2
3100
1000
1250
5100
3850
500
775
8200
78.4
-2.2
Key to Effluent Streams:
,tThe flue gas contains ^ 6% H2O.
The air fed to the strippers contains ^1
1 - NO scrubber 4 (7.5 ft.); 2 - NO scrubber 2 (19 ft.)
3 - stripper; 4 - NOX scrubber 2 (halfway up the 19 ft. scrubber).
,H20.
-------
Table IV-1. (Continued)
Run No.
NA
'HD
Flue Gas
Acid
Air
[C
NCL
1
2
3
1
2
3
Flue Gas
1
2
3
NO Scrubbing Efficiency
[CNfJ Flue Gas
1
2
3
NO Matl. Bal. Error
[CH „] Flue Gas
H2° Air
SCFM
GPM
SCFM
wt%
rt
wt%
If
II
ppm
ppm
72
A
ISO/
260
250/
270
4
0.17
0.30
0.64
0.21
77.5
71.2
77.7
3050
2250
2300
45.4
3050
1075
725
-47.3
~6
~1
B
ISO/
265
250/
270
4
0.16
0.30
0.60
0.30
76.6
77.7
77.5
3100
2150
2250
48.4
3100 -
1050
825
-16.4
~6
4
C
ISO/
260
255/
270
4
0.18
0.70
0.76
0.30
77.3
77.3
77.3
3100
2300
2400
44.8
-3400
1150
775
-44.4
~6
~1
Key to Effluent Streams:
1 - NO scrubber 4; 2 - NO scrubber 2; 3 - stripper.
X X
-------
Table 1V-1. (Continued)
Run No.
Relative Time
TNA
T2 - Avg.
^4 - Avg.
TIID
THD - Avg.
Q Flue Gas
Q Acid
QAir
1
2
3
1
2
3
hr.
°F
"F
op
SCFM
GPM
SCFM
wt%
wt%
ppm
[CNO!
pp
Flue Gas
1
2
3
4
Flue Gas
1
2
3
4
NO Scrubbing Efficicnc
x 10.0ft.
19 0 ft
26.5 ft.
N0x Matl.Bal. Error
76
A
0
220/
275
264
242
240/
260
251
5.6
0.26
1.9
;
-
540
150
B
1.0
2307
280
264
242
250/
260
251
5.6
0.27
1.9
.009
.038
.006
78.7
78.7
78.6
550
235
350
330
C
4.75
215'
260
231
252
2507
260
253
5.6
0.27
1.9
.015
.032
.002
79.1
78.7
78.9
620
360
320
D
5.25
225/
265
234
231
2507
260
252
5.6
0.28
1.9
.015
.032
.002
79.1
78.7
78.9
600
330
350
380
825 700 450 475
350 275 150 150
300 150 200
2050 - 3025
77'
1257
260
244
243
240/
255
251
5.6
0.25
1.9
0.19
.023
.015
78.6
78.2
78.9
330
300
150
150
78**
A
0
220'
270
264
235
2407
260
251
5.6
0.25
1.9
.011
.037
.007
79.1
79.7
79.4
530,
100X
400,
300^
B
3.0
220/
270
259
237
240/
260
251
5.6
0.25
1.9
015
.022
.007
78.2
77.9
78.3
850
300
370
150
1075
475
400
2000
63.5
59.3
8.5
52.3
-2.3
55.4
57.2
60.0
43.0
Key to Effluent Streams:
1 - NO scrubber 4 (7.5 ft.); 2 - NO* scrubber 2 (19 ft.)
3 - stripper; 4 - NOX scrubber 2 (halfway up the 19 ft. scrubber).
,tNoisy run - levels hard to determine.
New NO oxidizer - 0.05 ft3.
''Readings not steady.
IV-6
-------
Table IV-1. (Continued)
<
Run No.
Relative Time
TNA
T2 - Avg.
T4 - Avg.
THD
THD - Avg
Q- Flue Gas
Q Acid
QAir
[CHNS05]
[CH so 1
24
[CNO2I
[CMQ]
1
2
3
1
2
3
Flue Gas
1
2
3
4
Flue Gas
1
2
3
4
hr.
°F
op
°F
op
op
SCFM
GPM
SCFM
wt%
wt%
"
"
ppm
11
"
M
ppm
11
"
79
A*
0
275 /
270
262
242
240 /
260
252
2.8
0.12
1.9
.009
018
.013
78.7
78 6
78.6
500
0
190
-
-
450
225
150
-
B
16.5
2257
275
259
245
2557
270
260
2.8
0.12
1.9
-
-
_
-
-
470
180
_
0
-
450
275
_
650
-
C
19.25
2257
275
259
257
255 /
270
260
2.8
0.12
1.9
015
.026
.007
80.2
80.3
80.0
490
120
_
-
-
450
175
_
-
-
D
21.0
225/
275
259
257
255/
270
260
2.8
0.12
1.9
010
040
.007
79.6
79.6
79.6
840
180
-
0
-
825
400
-
2500
-
NO Scrubbine Efficiency
X
10.0ft.
19.0ft.
26 5 ft.
%
%
y
70
-
_
70.0
-
_
50.5
-
-
69.0
-
-
67.5
NOx Matl.Bal. Error
80
A
0
250/
280
269
262
180/
265
227
2.8
0.25
1.9
022
.029
.022
80.0
80 0
80.0
570
130
550
175
67.0
29.4
B
3.5
255 /
280
265
267
250/
265
259
2.8
0.26
1.9
015
019
.001
79.8
79.8
79.8
460
85
500
300
60.0
27.1
C
6.5
260/
285
267
272
260/
275
267
2.8
0.26
1.9
-
-
490
50
500
260
68.7
-
D
7.0
260/
285
267
272
260/
275
267
2.8
0.26
**
.015
022
.011
80.7
80.7
80.7
515
100
185
150
425
200
110
1675
68.2
16.9
E
7.5
260/
285
267
272
260/
275
267
2.8
0.26
1.9
-
;
85
1100
;
-
F
7.75
260/
290
267
272
260/
280
267
2.8
0.26
1.9
-
;
550
90
190
65
425
260
125
1065
62.1
-
Key to Effluent Streams:
1 - NOX scrubber 4 (7.5 ft.); 2 - N
-------
00
Run No.
Relative Time
TNA
T2 - Avg.
T4 - Avg.
THD
THD - Avg.
Q Flue Ga$
Q Acid
QAir
[CHNSCkl
[Cf^soJ
1
2
3
1
2
3
[CN02] Flue Gas
[CNO] Flue
2
3
4
Gas
1
2
3
4
hr.
°F
°F
OF
°F
°F
SCFM
GPM
SCFM
wt%
"
"
wt%
11
"
ppm
"
"
"
ppm
"
"
"
"
81
A
0
2607
280
271
271
260/
280
267
2.8
0.26
1.9
015
.049
.011
80 9
80.9
80.9
1100
150
310
-
-
1080
325
160
-
-
A'
0.25
260/
280
271
271
260 /
280
267
2.8
0.26
1.9
_
-
-
_
-
-
145
-
250
-
.
260
-
2900
-
B
1.5
2557
285
268
272
265/
280
270
2.8
0.26
1.9
.008
.013
.011
80 6
80 7
80.2
1000
190
300
270
-
1050
260
125
2150
-
C
2.5
255/
285
268
272
265/
280
270
2.8
0.26
1.9
.004
.007
.006
80.6
80.6
80.9
1000
-
-
_
850
_
-
-
-
D
4.0
255/
285
266
274
265 /
280
271
2.8
0.26
1.9
.003
014
.006
80.9
80.9
80.7
1050
150
240
200
1050
250
160
2385
NO Scrubbing Efficiency
x 10
19
26.
NOX Matl.Bal
Oft.
Oft
5 ft.
. Error
V
/D
(r
"
_
_
78.3
9.8
_
.
81.5
-
_
_
78.0
61.8
_
_
_
-
_
_
81.0
59.8
Table IV-1. (Continued)
82(a)
A
0
220/
280
207
235
245/
270
258
4.9
0.26
1.9
.026
.019
.015
82 0
82 0
82.0
275
170
200
85
D
2.0 4.75 5.5 6.0 6.75 7.5
245/
280
271
269
250 /
270
261
4.9
0.26
1.9
.017
.026
.011
80.5
81.1
81.5
265
150
180
20
250 235
170 100
175 100
200 165
260/
295
279
271
265/
290
279
4.9
0.23
1.9
28.5
34.3
13.7
44.0
50.0
-10.9
275
140
200
210
450
250
225
275
33.2
41 4
45 0
265/
295
280
271
280/
300
289
4.9
0.23
1.9
031
061
.013
81.9
81.9
80.7
230
120
165
15
185
275
225
150
280
235
7.8
37.7
31.8
-103
2657
295
280
271
280,'
300
289
4.9
0.23
1.9
230
105
150
0
190
235
140
135
200
160
24 7
38.7
47.4
270/
275
285
273
2857
305
292
4.9
0.23
1.9
.035
068
.010
81 5
81.3
81.1
230
95
135
0
170
235
155
135
195
160
270/
295
285
273
285/
305
292
4.9
0.23
1.9
230
95
135
0
160
235
250
135
195
160
29.0 31 2
41 9 41 9
47.3 47.3
<-100
Key to Effluent Streams:
1 - NOX scrubber 4 (7.5 ft.); 2 - NO., scrubber 2 (19 ft )
3 - stripper; 4 - NOX scrubber 2 (halfway up the 19 ft. scrubber).
-------
Table 1V-1. (Continued)
Run No.
Relative Time
TMA
NA
T2 - Avg.
T4 - Avg.
T
HD
THD - Avg.
Q Flue Gas
Q Acid
Q Air
lcHNSOsl
fcH2SO4'
[CN02] Klue
[CNQ] Flue
2
3
2
3
Gas
1
2
3
4
Gas
1
2
3
4
hr.
01-
'"
OF
"p
°F
°F
SCFM
GPM
SCFM
wt%
wt%
"
ppm
"
"
"
ppm
n
"
"
"
NO- Scrubbing Efficiency
x 10
19
26.
NOX Matl.Bal
Stripper Matl
Oft.
Oft
5 ft.
%
%
%
82(b)
A
0
265/
300
284
271
285/
310
295
4.9
0.23
1.9
-
-
-
-
485
265
320
0
365
500
275
255
175
325
Not
Steady
State
45.3
B
0.75
27 5 /
300
284
284
285/
310
293
4.9
0.23
1.9
-
-
-
-
-
250
305
200
350
-
250
240
2600
250
39.0
44.7
49.0
.Error %
.Bal.
Error %
Key to Effluent Streams:
-
1 - NOX scrubber
3 - stripper; 4 -
C
2.0
265/
295
281
279
280/
300
292
4.9
0.23
1.9
.033
.059
.014
80 6
80.9
81.1
490
210
255
190
315
500
140
175
2600
275
40.3
56.6
64.5
-50
-
4(7.5
D
3.0
270/
300
283
277
280/
305
295
4.9
0.23
1.9
037
.059
.011
80 7
80 7
81.1
490
150
230
250
240
550
150
175
2900
275
49.7
58 4
70.7
-49.5
-
ft.); 2
NOX scrubber
82(c)
A
0
270/
300
283
277
265/
305
292
4.9
0.24
1.9
.048
066
.018
81 3
81 5
81.3
1050
570
680
390
770
950
300
400
4200
550
35.0
46.0
56.5
6.5
-66.8
B
1.5
270 /
300
283
277
265 /
305
292
4.9
0.24
1.9
041
066
.022
81.3
81.3
81.1
1050
490
600
80
680
925
250
400
1550
550
37.6
49.3
63.2
14.3
19.4
- NOX scrubber 2
2 (halfway
up the
C*
3.5
270/
300
283
277
265/
305
292
4.9
0.24
3.8
037
066
.019
81.5
81.5
81.3
1000
420
-
10
-
950
225
1025
-
-
_
67.2
6.5
7.2
(19ft.)
19 ft. scrubber).
High stripper air.
IV-9
-------
I
O
83a
83b
Table IV-1. (Continued)
83c 83a' 83b'
Run No.
Relative Time
hr.
83b"
TNA
T2 - Avg.
T4 - Avg.
THD
T
1 HD - Avg.
QFlue Gas
QAcid
QAir (St)
fCHNS05l 1
3
fcH2SO4l 1
3
fCNO2l F'116 Gas
1
2
3
4
[CNO] Flue Gas
1
2
3
4
°F
"F
"F
"F
"F
SCFM
GPM
SCFM
wt*
wt%
pRm
ppm
275/
310
279
295
240/
270
250
8.6
.40
3.1
.015
.022
.015
80.9
80.9
80.9
320
0
10
50
30
250
250
300
550
325
255/
295
284
265
235/
270
254
8.6
.40
3.1
.018
026
.018
80.7
80.7
80.7
480
0
0
10
10
575
470
475
375
475
260/
285
273
272
230/
270
253
8.6
.40
3.1
.018
.037
.018
80.9
80.9
80.9
1100
0
0
5
0
1000
700
700
500
650
250/
275
253
265
250/
275
-
8.8
.39
3.1
.022
.033
.014
80.9
80.7
80.9
240
65
95
15
140
240
120
125
650
150
260/
290
287
275
260/
275
-
8.8
.38
3.1
.018
.029
.015
80.8
80.6
81.3
520
320
375
130
435
475
225
275
1000
350
280/
285
282
280
270/
280
-
8.8
.38
3.1
.029
066
.018
80 4
80 4
80.6
975
490
500
340
615
1050
385
500
2000
635
280/
295
283
287
270/
290
_
8.8
.39
3.1
.026
.070
.018
80.9
80 7
82.3
925
455
515
300
625
1000
375
460
1975
575
260/
280
269
269
270/
285
_
8.8
.39
3.1
.018
030
.015
82.6
81 2
82.0
530
250
310
100
410
490
240
275
1225
360
275/
285
_
280
265/
285
275
6.0
.40
.6
018
025
!014
79.4
79 1
79! 1
240
60
90
50
110
300
140
160
1675
240
2757
280
279
27 O/
285
281
6.0
.40
.6
.015
022
[on
79 4
79 4
79^4
260
80
120
0
120
275
125
125
1400
175
NO Scrubbing Efficiency
x 10.0ft
19.0ft.
26.5 ft.
NOx Matl.Bal. Error
w
ft
%
37.7
45.6
56.1
21.6
54.0
55.0
55.0
31.1
69 0
66.7
66.7
42.6
39.6
54.3
61.7
6.7
21.1
34.7
45.2
16.2
36.7
49.4
55.7
12.7
37 7
49 4
56.9
7.8
24.5
42 6
52 .'0
20.2
35 2
53 7
63^0
-18.0
44.9
54 2
61 ! 7
4.3
Key to Effluent Streams:
1 - NOx scrubber 4 (7.5 ft.); 2 - NOX scrubber 2 (19 ft.)
3 - stripper; 4 - NOX scrubber 2 (halfway up the 19 ft. scrubber).
-------
Table IV-1. (Continued)
84c
Run No.
Relative Time
TNA
T2 - Avg.
T, - Avg.
4 6
THD
THD - Avg.
Qplue Gas
QAcid
Q Air (St)
2
3
[CH2S04] 1
3
[CNQ2] Flue Gas
2
3
4
[Cjsiol F^ue <~'SLS
1
2
3
4
hr.
op
°F
°F
op
SCFM
GPM
SCFM
wt_%
wt%
ppm
"
ppm
••
NO Scrubbing Efficiency
x 10.0ft. %
19.0ft.
26.5ft.
NO Matl. Bal. En
ror%
0
280/
280
_
280
275/
290
281
6.0
.40
.6
.015
037
.011
80.0
80.0
80.0
530
190
270
280
260
500
150
175
3025
275
48.1
56.1
67.0
-1.8
— — -
1.0
27 5/
275
_
274
275/
290
281
6.0
.40
.6
018
.037
.011
80 0
80.0
80.0
500
200
;
475
110
-
68.2
-3.8
—
0
270/
280
-
274
275/
290
281
6.0
.40
.6
.019
041
.008
80.2
80.2
80.2
1050
415
515
1100
530
1050
250
325
6000
550
48.7
60.0
68.3
25.2
B
0.5
260/
280
271
270
27 5/
295
285
6.0
.40
.6
.023
.038
.011
80 4
80.4
80.4
1000
420
-
1100
250
:
68.1
28.9
A B* C** D**
0 5.0 5.25 6.0
265/
290
283 -
276 -
270/
300
289
1.4 -
0.1 -
3.1 -
.019
022 -
.011
83.6
83.6
83.6
1000 HW) fsQ/l 840
60 35 40 10
95 - - 15
80 100
150 50
1250 IW W 900
250 375 400 400
225 - - 375
500 865
350 350
77.7
QC Q _
OO .0 7 R S /
86-2 8i;s 79-5
63.6
A
0
270/
315
280
292
235/
290
269
1.4
.11
3.1
_
_
200
0
0
50
250
150
150
250
125
;
-
B
1.25
245/
310
278
278
235/
290
269
1.4
.13
3.1
019
022
.011
80.7
80.7
80.7
210
5
30
70
250
150
175
175
46.7
56 5
66.3
-14.7
C
1.75
2457
310
278
278
235/
290
269
1.4
.13
3.1
022
034
.015
80.7
80.7
80.7
230
30
-
250
125
-
67.7
-36.6
"Two measurements of feed level.
"'Difficulty in achieving steady readings
- sample lines probably plugged.
-------
Table IV-1. (Continued)
t-»
to
85b
Run No.
Relative Time hr.
TNA °F
T2 - Avg. »F
T4 - Avg. op
THD °F
T
1 HD - Avg. "F
Q Flue Gas SCFM
Q Acid CPM
Q Air (St) SCFM
[CuMsoJ 1 wt %
a 2
3
[CH2S04] 1 wt,%
3
[CNO2J Flue Gas ppm
2
311
4
[CNQ] Flue Gas ppm
1 "
2 "
3
4
NO- Scrubbing Efficiency
x 10.T) ft. %
19.0ft. 'y
26.5ft.
NOxMatl.Bal. Error %
A
0
260/
290
276
274
240/
285
265
1.4
0.12
3.1
.023
.038
.011
80.9
80.9
80.9
550
80
140
20
210
475
100
125
525
200
60.0
74.1
82.4
-7.3
B
1.0
255/
275
260
268
240/
290
267
1.4
0.12
3.1
.030
.041
.014
80.9
80.9
80.9
540
70
110
0
150
500
100
125
700
160
70 2
77 4
83.7
-5.0
A
0
250/
280
263
268
240/
290
268
1.4
0.12
3.1
.041
.114
.023
81.3
81.3
81.3
1100
90
130
10
180
1000
75
90
625
150
84.3
89.5
92.1
-42.1
B»
1.0
260/
280
270
267
240/
290
267
1.4
0.12
3.1
\
-
1100
240
:
975
100
-
83.4
-
A
0
250/
290
269
290
280/
330
312
2.9
0.1
0.6
.023
.023
.016
84.5
84.5
84.3
230
140
150
10
130
225
90
100
100
150
45.5
49.5
18.2
B
1.25
250/
285
272
278
-
-
2.9
0.08
0.6
.018
.026
.0145
85.2
83.3
84.0
250
170
190
1
250
100
100
42.0
46.0
-13.2
A
0
260/
295
279
271
285/
320
306
2.9
0.1
0.6
.018
013
.015
82.7
81.3
81.9
550
160
400
385
-0
3900
-
68.7
B
1.0
260/
295
281
267
285/
320
299
2.9
0.11
0.6
.024
.009
.013
83.1
83.1
84.5
570
210
25
400
305
325
0
10
3900
125
54.0
70.5
76.5
<50
A
0
270/
285
278
274
285/
310
296
2.9
0.12
0.6
.011
.011
.013
84.5
82.7
83.6
510
110
140
1140
200
400
25
85
4050
125
64.3
85.2
<50
B
0.5
270/
285
278
274
285/
310
296
2.9
0.12
0.6
.011
.013
.011
84.7
83.1
84.5
510
110
155
1400
125
400
15
100
3975
315
64 5
72.0
86.4
68.5
C**
5.0
220/
310
268
260
270/
305
285
2.9
0.11
0.6
032
.101
.052
81.5
82.3
82.2
540
10
90
-1200
625
100
135
-5000
80.5
91.4
13.2
D
7.5
265 /
295
281
276
290 /
310
303
2.9
0.10
0.6
.013
.090
.013
83.3
80. 5/
82.5
525
55
95
650
110
700
140
150
4400
230
74 0
79.9
84.2
-32.7
"Heater failed; copper powder exposed to acid in Scrubber 2
"Acid flow variations - wide temperature fluctuations.
/System was cooled down between 5.0 and 7.5 hours and may have picked up some water
-------
Table IV-1. (Continued)
I
l-t
CO
86c
86d*
87 a
87b
87c
88a
88b
Run No.
Relative Time hr.
TNA
T2 - Avg.
T4 - Avg.
THD
OF
°F
°F
°F
THD - Avg. UF
Q Flue Gas SCFM
Q Acid
Q Air
[QHNSOS]
[CH2S041
[CN021
[CNO!
GPM
SCFM
1 wt%
2
3
1 wt%
2
3
Flue Gas ppm
1
2
3
4
Flue Gas ppm
1
2
3
4
NO Scrubbing Efficiency
x 10 () ft. %
19.0ft. (t
26.5ft.
NOV Matl.
. Bal. Error %
0
260/
295
283
274
300/
320
305
2.9
0.1
0.6
026
.106
.015
83.1
82.5
82.5
930
15
60
790
110
1075
125
125
4850
225
83.3
90.8
93.0
25.7
1.5
270/
290
284
277
290/
320
306
2.9
0.1
0.6
030
.114
.019
83.1
83.1
82.5
1000
15
50
780
100
1050
125
125
4700
175
86 6
91.5
93.6
24.0
250/
290
270
277
300/
320
312
2.9
0.1
0.6
329**
.038
82.3
82.7
1000
20
490
60
1100
150
175
88.9
91.9
-79.0
260/
290
273
273
270/
290
279
4.9
.25
1.9
.015
.022
.011
82 0
81.6
84.6
260
120
150
90
185
225
110
125
475
140
33.0
43.3
52.6
4.1
270 /
285
278
278
280/
290
285
4.9
.25
1.9
.018
.029
.015
82.3
81.9
81.9
520
205
265
430
350
550
150
165
2025
300
39.3
50 5
67.5
35.8
.
270/
290
275
276
280/
295
287
4.9
.25
1.9
.029
.081
.018
81.4
81 2
81.6
1000
365
445
1280
580
1000
180
250
3390
390
51 5
65.3
72.8
5.9
_
255 /
290
284
262
270 /
285
274
2.9
0.2
3.1
.019
030
.011
81.3
81.3
81.3
230
70
80
70
90
280
160
200
250
340
33.3
45.1
54.9
-
_
260;
280
273
269
270/
285
275
2.9
0.2
3.1
.027
038
.017
81.5
81.5
81.5
510
130
160
90
230
490
150
160
640
225
54.5
68 0
72.0
~
A
0
265/
290
280
272
270/
280'
274
2.9
0.2
3.1
030
.068
.022
81.7
81.7
81.7
1000
150
220
200
340
950
350
375
1225
475
58.2
69.5
74.4
"
B
1.25
27 O/
290
278
273
270/
280
272
2.9
0.2
3.1
.034
061
.019
81.7
81.7
81.7
950
130
_
1050
350
76.0
*Run made with Scrubber 2 only (19-ft.of column).
* 'Concentration double-checked.
Fresh acid from stripper fed to Scrubber 2 directly.
-------
Table IV-1. (Continued)
I
H>
it*
Run No.
Relative Time
TNA
T2 - Avg.
T4 - Avg.
T HD
T HD - Avg.
Q Flue Gas
Q Acid
Q Air
[CHNSO 1
HN 5
[CH2S041
ICN02] Fli
1
2
3
1
2
3
je Gas
1
2
3
4
[CNQ] Flue Gas
NO Scrubbir
x 10
19
26,
2
3
4
hr.
°F
"F
°F
op
°F
SCFM
GPM
SCFM
wt%
"
"
wt%
"
ppm
rt
*•
rt
ppm
"
*'
"
92a
A»
0
260/
295
273
280
270/
295
283
4.9
0.25
0.9
.019
026
.008
81.5
81.5
81.5
270
125
150
20
160
250
125
140
1150
165
B
1.0
265/
300
288
270
285/
300
292
4.9
0.25
0.9
.019
026
.011
81.5
81.5
81.5
270
140
-
-
-
240
125
-
_
-
ig Efficiency
.lift.
.Oft.
.5ft.
7?
fr
"
37.6
44.3
52.0
-
-
48.3
NOX Mail. Bal. Error
-14.9 -7.8
92b
A
0
255/
280
269
270
290/
300
296
4.9
0.25
0.9
-
~
480
40
80
**
160
550
160
175
**
240
61.3
75.5
80.6
-
B
1.25
260/
290
275
273
280/
300
294
4.9
0.25
0.9
0.02
0.08
.007
82.1
81.9
81.9
560
40
80
325
130
470
150
160
2375
225
65.5
76.7
81.6
-57.2
C
1.75
-
-
-
-
-
-
-
-
\
-
450
575
-
-
92c
250/
285
279
269
285/
295
290
4.9
0.25
0.9
015
0.16
.006
80.9
80.8
80.9
1125
145
260
1100
420
1100
120
175
4700
320
66.7
80.4
88.1
-50.8
-1125
1150
93a
260/
290
273
271
280/
290
278
2.6
0.1
1.6
.011
.049
.009
82.1
80.8
81.2
280
0
10
0
25
225 225
85 100
100
375
130
69.3
78.2
83.2 78.2
-32.0
275
10
93b
260/
305
280
271
280/
300
286
2.6
0.1
1.6
.007
.051
.006
81.8
81.3
81.1
500
0
0
0
10
475
150
135
450
200
71.3
86 1
84.6
11.6
470
0
500
150
•All runs after 92 were made with 2 strippers in series.
"Pump broke - reading unsteady.
-------
Table IV-1. (Continued)
93c
en
Run No.
Relative Time hr.
T NA °F
T2 - Avg. °f
T4 - Avg. -F
THD °F
T HD - Avg. "F
Q Flue Gas SCFM
Q Acid GPM
Q Air SCFM
[QHNS051 1 wt,*
3
[CH2S041 1 wt*
3
[CNOol F\ue Gas ppm
2
3
4
[CNO] Flue Gas ppm
2
3
4
NO Scrubbing Efficiency
x 10.0ft. %
19 0 ft
26.5ft.
NO Matl.Bal. Error %
A B
0
260/
295
274
280
270/
305
286
2.6
0.1
1.6
.016
096
'.009
81.1
81 4
80.4
1200 900
10 0
40
_* -
190
1070 1025
200 175
200
_ *
300
78.4
89.4
90.7 91.9
29.3
A
0
250/
270
264
257
270/
275
272
6.0
0.4
1.6
.023
.037
.011
81.3
81.3
81.3
240
50
70
10
130
225
200
200
700
200
29 1
40 8
46.2
-65.5
B
250/
270
264
257
270/
275
272
6,0
0.4
1.6
.018
.030
.011
81 3
81.3
81.3
230
30
70
_
120
241
225
225
225
26.6
37.2
45.7
-39.3
A
250/
270
264
257
270/
275
273
6.0
0.4
1.6
026
041
.015
81.3
81.3
81.3
540
90
150
20
280
550
250
280
noo
350
42.2
60.6
68.8
2.4
B
2707
290
273
279
2757
280
279
6.0
0.4
1.6
.022
.034
.011
81.3
81.3
81.3
530
80
_
-
525
280
_
-
-
-
-
65.9
4.9
A
2707
290
273
279
2757
280
279
6.0
0.4
1.6
030
053
.015
81.3
81.3
81.3
1100
300
470
230
630
1050
350
375
1450
550
45.1
60.7
69.8
23.4
B
2607
290
268
275
2807
280
279
6.0
0.4
1.6
037
056
.015
81.3
81.3
81.3
1100
290
-
-
-
950
375
-
-
-
-
-
67.6
10.0
A
250'
285
258
268
2707
270
272
8.8
0.4
0.3
;
-
_
-
265
100
160
-
-
250
180
175
-
-
-
35 0
45.6
-
B
250'
285
258
268
2707
270
272
8.8
0.4
0.3
019
.030
.011
80 7
80 7
80.7
300
135
185
730
220
180
135
140
3250
150
22.9
32.3
43.8
-20.2
C
255
280
271
264
270,'
270
272
8.8
0.4
0.3
-
-
-
-
270
-
-
-
175
-
-
-
-
-
-
-
D
255 '
280
271
264
270'
270
272
8.8
0.4
0.3
-
-
_
-
230
85
-
-
-
250
175
-
-
-
-
-
-
-
'Possible instrumentation problem - plugged sample lines.
-------
Table IV-1. (Continued)
Run No.
Relative Time
TNA
T2 - Avg.
T4 - Avg.
THD
THD - Avg.
Q Flue Gas
Q Acid
Q Air
[CHNSO51 1
3
1
2
3
hr.
°F
SCFM
GPM
SCFM
wt%
wt%
Flue Gas ppm
2
3
4
Flue Gas ppm
2
3
4
NC> Scrubbing Efflclenc
x 10.0 ft.
19.0ft.
26.5ft.
NOX Matl. Bal. Error
95b
A
0
255/
280
271
264
270/
270
272
8.8
0.4
0.3
.026
.037
.011
80.9
80.9
80.9
500
190
280
2900
360
550
280
300
6000
400
27.6
44.8
55.2
7.8
B
0.75
250/
280
270
263
275/
280
280
8.8
0.4
0.3
.023
.033
.015
80.9
80.9
80.9
615
320
575
275
50.0
17.0
95c
A
0
245/
280
269
264
270/
290
276
8.8
0.4
0.3
.060
.079
.015
80.7
80.7
80.7
910
310
420
6500
570
950
400
475
8000
640
34.9
51.9
61.8
-1.1
B
1.75
260/
285
276
267
280/
280
281
8.8
0.4
0.6
~
™
5750
6000
-
-
C
3.25
260/
285
275
277
280/
285
283
8.8
0.4
1.2
~
-
1700'
4250'
-
-
*Sharpdrop - then slow decrease to these levels.
IV-16
-------
Table IV-1. (Continued)
Run No.
<
i
Relative Time
TNA
T2 - Avg.
T4 - Avg.
THD
THD - AvS.
Qplue Gas
Q Acid
Q Air (St)
[CHNSOS] '
[CH2S04I
hr.
op
°F
°F
SCFM
GPM
SCFM
wt^c
wt%
m
[CN021 Flue Cas PP
2
3
4
Flue Gas ppm
1
2
3
4
NO Scrubbing Efficiency
x 10 0 ft %
19 Oft.
26.5ft.
NOX Matl Bal. Error %
90a
A
0
260'
280
271
273
270
290
281
6.0
.12
3.1
.026
037
.011
80.4
80.4
80.4
270
130
150
0
190
250
125
150
260
175
29.8
42.4
51.0
8.8
B
1.25
260,'
290
275
Til
270,'
300
281
6.0
.12
3.1
-
-
-
_
_
-
250
100
-
-
-
275
150
-
_
-
.
.
52.5
-
90b
A
0
2607
290
275
277
2707
300
281
6.0
.12
3.1
034
105
.015
80.7
80.7
80.7
490
250
280
10
350
550
225
240
640
260
41 3
50 0
54.3
-14.4
B
1.25
2607
285
277
273
270'
310
287
6.0
.12
3.1
-
-
-
-
-
-
470
220
-
-
-
525
175
-
-
-
_
-
60.5
-
90c
A
0
2707
290
278
283
2707
310
288
6.0
.12
3.1
023
265
.011
81.3
81 3
81.3
1050
450
500
50
600
1100
275
350
1775
500
48 8
60 5
66^3
-30.7
B
1.50
255 /
300
283
275
2707
310
289
6.0
.12
3.1
_
_
-
_
_
-
1100
450
_
_
-
1100
290
_
-
.
_
66.4
-
90d
-
255 /
300
283
275
275
210
289
6.0
.12
3.1
037
634
.019
81.3
81.3
81.3
2000
900
940
130
1100
2050
375
475
2600
700
55.5
65.1
68.5
-53.5
91a
A
0
2357
275
261
257
2807
310
297
11.2
. l
3.1
Oil
022
.007
81.7
81.7
81.7
200
140
145
400
160
225
175
200
60
225
9.4
18.8
25.8
13.4
B
1.0
-
-
-
-
-
-
-
-
-
-
-
-
200
160
-
-
-
225
175
-
-
~
-
-
21.2
-
91b
245 /
285
262
270
290'
315
298
11.2
.1
3.1
019
030
.011
82 3
82 3
82.3
480
240
260
80
300
550
350
460
1100
450
0.75
510
250
500
300
27 2
35.9
42.7 45.5
34.7
Kev to Effluent Streams:
1 - NOX scrubber 4 (7.5ft.); 2 - NOy scrubber 2 (19 ft.)
3 - stripper; 4 - NO scrubber 2 (halfway up the 19 ft. scrubber).
-------
Table 1V-1. (Continued)
i-*
oo
Run No.
Relative Time
TNA
T2 - Avg.
T4 - Avg.
THD
THD - Ave-
QFlue Gas
QAcid
Q.Mr (St)
[CHNS051 1
3
[CH2S04] 1
3
[CNO21 Flue Gas
2
3
4
[CNQ] Flue Gas
2
3
4
hr.
°F
°F
°F
°F
op
SCFM
GPM
SCFM
wt%
11
wt%
"
ppm
"
"
"
ppm
"
"
"
NO Scrubbing Efficiency
x 10.0ft. %
19.0ft.
26.5ft.
ir
"
91c
A
0
250/
295
262
278
295/
310
303
11.2
.1
3.1
034
.205
.015
82.7
82.7
82.7
1100
400
490
170
660
1050
450
525
2400
650
39.1
52.8
60.5
B
1.50
-
-
-
-
-
-
-
-
:
-
-
-
1150
475
-
-
-
1075
480
-
-
-
_
-
57.1
NOX Mart. Bal. Error
27.2
91d
A B
0
250/
305
263
278
290/
305
300
11.2
.1
3.1
.046
.589
.023
82.9
82.9
82.9
2000 2000
280 300
420
660
710
2100 2050
650 550
750
4100
1000
58.3
71.5
77.3 79.0
25.9
96a
A
0
270/
280
277
275
280/
295
288
1.5
.12
.3
.007
.018
.007
80.1
80.3
80.3
240
0
50
0
120
290
150
110
1000
140
50.9
69.8
71.7
B
1.0
265/
290
275
276
380/
310
291
1.5
.12
.3
-
;
260
10
300
125
75.9
96b
A
0
240/
285
275
267
280/
320
302
1.5
.13
.3
.006
048
.008
81.0
80.2
80.4
520
35
20*
565
380*
550
175
225*
2375
225*
43.5
77.1
80.4
B
1.75
240/
290
274
269
240/
310
302
1.5
.13
.3
":
;
500
0
550
300*
73.2
96c
A B
0 1.0
265/
290
272
278
295/
305
302
1.5
.13
.3
Oil
.075
.007
80.2
80.2
80.6
900 1000
10 10*
90
490
210
1000 1000
200 175*
225
3250
350
70.5
83.7
88.9 90.8
Key to Effluent Streams: 1 - NOX scrubber 4 (7.5 ft.); 2 - N
-------
Table IV-1. (Continued)
i
H*
CO
Run No.
Relative Time
TNA
T2 - Avg.
T4 - Avg.
THD
THD - Avg.
Qplue Gas
QAcid
QAir (St)
[CHNS051 1
3 2
3
£CH2S04] 1
3
[CN021 Flue Gas
2
3
4
[CN01 Flue Gas
2
3
4
hr.
°F
°F
°F
°F
"F
SCFM
GPM
SCFM
wt%
wt%
pprn
ppm
NO Scrubbing Efficiency
x 10.0ft. %
19 0 ft.
26 5 ft.
97 a
A
0
240/
280
258
259
270/
280
274
4.9
.25
.9
.019
023
.011
80 4
80.4
80.4
280
175
195
50
220
225
100
100
950
125
31 7
41 6
45 5
B
1.25
250,'
280
261
269
2707
285
279
4.9
.25
.9
-
-
270
170
225
100
45.5
97b
A
0
2507
290
263
270
275/
280
279
4.9
.25
.9
006
047
.006
80.3
80 2
80.3
500
155
225
160
295
500
220
250
1950
350
35.5
52 5
62 5
B
1.25
250/
280
265
263
2707
285
279
4.9
.25
.9
-
-
480
170
550
180
66.0
97 c
250'
285
268
268
2807
290
285
4.9
.25
005
094
.007
80 2
79.9
80.0
1025
360
500
560
590
1090
220
285
4600
500
1.0
250 /
285
268
272
2807
290
287
4.9
.25
.9
1000
360
1125
235
48 5
62 9
72 6 72.0
98a
A
0
250/
280
264
267
280/
290
288
8.6
.25
.9
007
031
.009
80 4
80 0
80.0
265
160
200
30
230
265
170
175
1650
225
14 2
29.2
32.7
B
1.0
245'
280
261
263
280 /
290
287
8.6
.25
.9
_
_
-
_
-
-
270
155
_
-
-
275
200
-
_
-
.
-
34.9
98b
A
0
2407
275
260
271
2807
290
285
8.6
.25
.9
007
046
!006
79.8
79.5
79.8
475
240
305
165
355
525
285
300
1410
375
27.0
39 5
47.5
B
1.25
2407
275
260
271
2807
290
285
8.6
.25
.9
-
-
-
-
-
-
480
245
-
-
-
550
285
-
-
-
_
-
48.5
Kev to Effluent Streams:
1 - NOX scrubber 4 (7.5ft.); 2 - NOX scrubber 2 (19ft.)
3 - stripper 4 - NOX scrubber 2 (halfway up the 19 ft. scrubber).
-------
Table IV-1. (Continued)
to
o
Run No.
Relative Time hr.
TNA -F
T2 - Avg. op
T4 - Avg. »F
THD °F
THD • AVK- °F
QFlue Gas SCFM
QAcid GPM
QAlr (St) SCFM
1 wt%
2
3
[CH2S041 1
3
[CN02] Flue^Cas
2
3
4
[Cj^fj] Flue Gas
1
2
3
4
NO Scrubbing Efficii
x 10.0ft.
19.0ft.
26.5ft.
ppm
99a
A
0
-
280
180
230
45
270
275
125
125
1300
170
20.7
36.0
45.0
B
0.50
~
385
215
250
125
46.5
C
1.50
185/
210
193
197
275/
380
278
4.9
.25
.9
.009
.020
.008
72.9
72.5
72.5
255
110
130
65
175
300
175
180
1125
250
23.4
44.1
48.6
99b
A
0
190/
210
199
198
270/
280
277
4.9
.25
.9
.009
.043
.007
73.0
72.6
73.0
530
200
260
135
335
555
245
255
1550
350
36.9
52.5
59.0
B
0.75
~
520
205
560
260
56.9
99c
190/
210
200
201
275/
280
279
4.9
.25
.9
.017
.100
.009
73 0
72.7
72.9
1050*
360
460
515
575
1060*
225
275
4290
475
50.2
65.2
72.3
99d
A
0
190/
215
199
201
275/
280
278
4.9
.25
.9
021
.197
.011
73.1
72.8
73.0
2050
670
830
1550
955
2025
300
400
•-5800
750
58.2
69.8
76.2
B
1.00
-
2050
670
2100
300
76.6
lOOa
ISO/
210
199
185
250/
260
255
4.9
.62
.9
.006
.027
.006
74.7
74.0
74.2
270
10
15
280
270
125
135
2325
72.2
75.0
lOOb
185/
210
195
199
200/
230
223
4.9
.62
.9
.006
.056
.004
74 4
74.0
74.2
535'
40
110
1000
555'
300
300
3225
67.4
68.8
Key to Effluent Streams:
1 - NO, scrubber 4 (7.5ft.); 2 - NO,, scrubber 2 (19ft.)
3 - stripper: 4 - NOX scrubber 2 (halfway up the 19 ft. scrubber).
Erratic response - probably UV cell contamination.
-------
Table IV-1. (Continued)
i
to
Run No.
Relative Time
TNA
T2 - Avg.
T4 - Avg.
THD
THD - Avs-
QFlue Gas
QAcid
QAir (St)
[CHNS05] 1
3
[CH2S04] 1
3
[CNO?] Flue Gas
L 1
2
3
4
[CNOI Flue Gas
2
3
4
hr.
op
op
op
°F
°F
SCFM
GPM
SCFM
wt%
"
wt%
"
ppm
"
"
11
ppm
"
"
**
lOOc
A B
0
185/
210
192
192
220/
230
226
4.9
.62
.9
010
.093
.006
74.3
73.6
74.3
1100 1050
195
295 275
210
-
1025 1060
160
175 225
4350
-
C
-
-
-
-
-
-
-
-
-
-
:
-
1100
-
-
-
975
-
-
-
NO Scrubbing Efficiency
x 100ft.
19.0ft.
26 5ft
CP
/o
»r
"
-
78 0 76 3
83.3
-
-
-
IQlb
A
0
4.9
.25
.9
.015
022
.011
79.1
79.1
79.1
120
30
30
30
60
650 650
500 460
475
900
475
65
20
30 5
34.4
31.2
32.9
lOlc
240/
300
283
270
2607
265
262
4.9
.25
.9
018
.022
.015
79.1
79.1
79.1
240
40
60
120
95
1150
850
850
1725
875
30.2
34.5
36 0
220
1150
lOld
240/
290
281
257
260/
270
265
4.9
.25
.018
022
.011
79.1
79.1
79.1
320
40
60
370
110
2200
1675
1700
3025
1800
24 2
30.2
31.9
330
2325
102a
210/
290
276
251
265/
275
269
4.9
.25
.9
.018
026
.011
79.1
79.1
79.1
350
240
260
75
270
75
60
60
175
60
350
250
75
40
22.4
24.7
29 4 29.4
Key to Effluent Streams:
1 - NOX scrubber 4 (7.5 ft.); 2 - NO scrubber 2 (19 ft.)
3 - stripper; 4 - NOX scrubber 2 (halfway up the 19 ft. scrubber).
-------
102b
to
to
Run No.
Relative Time
TNA
"^2 " Avg.
T~4 - Avg.
THD
THD - Av8-
QFlue Gas
QAcid
QAir (St)
[CHNSO5] 1
3
[CjbsoJ 1
2
3
[CNO21 Flue Gas
2
3
4
[CNO1 Flue Gas
1
2
3
4
hr.
°F
°F
"F
°F
°F
SCFM
GPM
SCFM
wt%
"
wt*
11
"
ppm
tf
"
"
ppm
11
"
"
0
210/
290
275
247
270/
275
271
4.9
.25
.9
.018
.033
.015
79.1
79.1
79.1
490
330
360
110
380
100
60
75
500
75
_
_
_
_
.
.
-
.
_
-
500
330
_
.
100
60
_
-
NO Scrubbing Efficiency
10.0 ft.
19.0ft
26.5ft.
V
7o
tr
**
22.9
26 3
33.9
_
_
35.0
Table IV-1. (Continued)
102c 102d
220/
285
277
251
270/
270
270
4.9
.25
.9
.022
.041
.018
79.1
79.1
79.1
1000*
620
700
220
705
225*
100
110
725
130
31 8
33.9
41 2
230/
295
278
260
265/
270
268
4.9
.25
.9
.026
.059
.018
79 1
79.1
79.1
2000
1300
1350
490
1350
350
175
200
925
225
33.0
34 0
37.2
Key to Effluent Streams:
1 - NO, scrubber 4 (7.5 ft.); 2 - NOX scrubber 2 (19 ft )
3 - stripper; 4 - NOX scrubber 2 (halfway up the 19 ft. s'crutter).
Erratic response - probably UV cell contamination.
-------
APPENDIX V. METAL ION INTERFERENCE IN SCRUBBER
AND STRIPPER OPERATION
A. HNSOC Determination — Interference by Corrosion Products
o
During the course of miniplant experimentation it was necessary to analyze the acid
streams for HNSOg content. The standard procedure, whereby acid samples are titrated to
a pale pink endpoint with aqueous 0.1N KMnO. solution, was found to indicate inordinately high
concentrations of oxidizable material in the acid streams. Calculations of available material
and our past experience with similar experiments led us to believe that oxidizable materials
other than HNSOg must be present in the acid. The source of these interfering materials was
believed to be corrosion products of metal alloys exposed to the acid streams. Visible spectro-
photometric analysis, laboratory corrosion tests, and replacement of some of the parts of the
miniplant equipment suggested that the major source of corrosion products was probably stain-
less steel. Of the possible stainless steel locations in the acid feed stream, the strippers of-
fered the highest surface area if the Teflon lining had failed. Dismantling and inspection of the
interior of the strippers revealed that the Teflon lining had failed and severe corrosion had
occurred.
The principle components of SS316 are iron, chromium, nickel, manganese, and molyb-
denum. If the reaction of any of these species with carbon is rapid or if corrosion is rapid,
then one or more of them could be in a reduced state while present in the acid. Titration with
permanganate would then result in an indication of oxidizable material present which could be
mistaken for HNSOg. Iron in sulfuric acid could be present as Fe or Fe ; chromium as
Cr or Cr (oxidation to chromate is slow and occurs only with very strong oxidants; nickel
as Ni , manganese as Mn4 , Mn4 , Mn (oxidation to permanganate would be unlikely; and
molybdenum as Mo , Mo, Mo+ , Mo, and perhaps Mo . Contact with the air for a
sufficient period of time would probably be sufficient to oxidize any of them somewhat. We
have found that samples which give high errors by KMnO, titration (relative to nitrometer re-
sults) when recently removed from the miniplant give much smaller errors by KMnO4 titration
after a week's storage in a closed container. A small error, equivalent to perhaps 0.04 wt %
HNSOK, is still observed (relative to the nitrometer results). Experiments with synthetically
prepared solutions using chromic and nitrosylsulfuric acid solutions lead us to believe that
this error is due to the slow oxidation of chromic ion to chromate by KMnO,. Nitrosylsulfuric
acid does not oxidize chromic ion to chromate.
V-l
-------
Interference by corrosion products can be avoided by using the nitrometer to analyze
for HNSO- instead of the KMn04 titration. Other titrimetric methods can probably be devel-
oped; however, the time required for such an undertaking would be excessive, considering the
availability of the nitrometer procedure.
B. Effect of the Presence of Corrosion Products on HNSO,.
Solubility and Scrubber Performance
The presence of corrosion products could lessen scrubber performance by decreasing
the solubility of HNSO- at any given partial pressures of NO and NOg. This effect would be
observed on the scrubber column as a decrease in overall scrubbing efficiency. However, all
sections of a large scrubber would have the same efficiency per unit length. Thus, if the bot-
tom section of the scrubber was scrubbing NO from the gas phase, the top section would fur-
X
ther reduce the gas phase NO concentration as long as the acid entering the column was free
of dissolved HNSO..
5
One may treat the problem of decreased HNSO.. solubility in the following semiquantita-
tive manner. The dissolution reaction is:
NO + NO2 4 2H2SO4 = 2NO4 4 2HSO4~ 4 HgO .
One may write an equilibrium constant expression for this reaction:
[N04]2[HS04"]2[H20]
If one sets PNQ = pN~ = p and if one disregards small changes in solvent concentrations (i.e.,
H-O and HgSOJ, this expression reduces to:
K =
and it may be rearranged to solve for HNSO- solubility:
HNSO, = [NO4] = P^K .
5 [HS04~]
HNSO-, which dissociates in sulfuric acid to form NO and HSO4 , decreases in solubility as
the HSO. concentration increases. If any increase occurs in HS04 concentration, the solu-
bility of HNSO. decreases proportionately.
We know that corrosion products are present in the hot scrubber acid at a concentration
close to the saturation value. Precipitation of metal salts is observed during cooling of the sys-
tem. We can estimate the concentration of dissolved metal salts to be several moles per liter
from a casual observation of the volume of solids which precipitate out when scrubber acid
samples are cooled. The solids content is approximately 5 vol %. The corrosion products are
V-2
-------
undoubtedly present as bisulfates, since there are no other anions present in appreciable con-
centration. Since the metal ions which we would expect to be present are polyvalent, there will
be at least two bisulfate ions present for each metal ion which is present. We may conserva-
tively estimate that the increase in bisulfate ion due to the presence of corrosion products in
the scrubber is 5 M.
Now that we have estimated the increase in bisulfate concentration, we must estimate
the original bisulfate concentration if we are to obtain a proper notion of the expected decrease
in HNSO- solubility. lonization Constants of Acids and Bases, by Albert and Serjeant, lists for
the first dissociation of H-SO. a pK of -3.0. The Hammet acidity function is -3.0 at a sul-
furic acid concentration of 6 M, according to the same authors. Using these two values and
i _1UL
simple mass balance expressions for H2O, FLO , HSO. , and H-SO,, one may estimate that
at 80 wt % H2SO. (14M HgSOA the bisulfate concentration is approximately 4M at room tem-
perature. At the higher temperatures encountered in the scrubber, the extent of dissociation
Is probably greater and the bisulfate concentration might be 5 or 6 M.
Thus, the concentration of bisulfate is approximately 5 M in the absence of corrosion
products and might be as high as 10 M in the presence of corrosion products. This doubling of
bisulfate concentration would result in halving of the HNSO- solubility according to the equations
we developed above. This decrease in solubility could halve the efficiency of the scrubber.
Ilius, the scrubbing efficiency, which is about 70%, might well be expected to provide higher
recoveries of NO in the absence of excessive amounts of corrosion products.
The number of assumptions involved in our treatment of HNSOg solubility (and its de-
pendence on corrosion products) makes it advisable that we consider the results of our calcula-
tions to be semiquantitative. It would be possible to measure HNSCX solubility in the presence
and absence of corrosion products.
C. Effect of the Presence of Corrosion Products on Stripper Performance
The performance of the stripper directly affects the extent to which the scrubber may
-remove NO from the gas phase. If the feed acid to the scrubber contains residual HNSOg that
was not removed by the stripper, then there will be a residual gas phase N0x level equal to or
greater than the equilibrium partial pressure of NO corresponding to the equilibrium value for
the feed acid. The behavior of the scrubber will be such that the lower portions of the scrubber
will operate fairly well (because of the high gas phase NOx input levels), but the upper sections
Of the scrubber will remove very little more NO (due to the residual liquid phase HNSO&).
Corrosion products present in the acid loop precipitate out of solution when the system
is cooled. This problem is probably very serious in the stripper, where precipitation can
result in the blockage of pores in the activated charcoal and a resultant drastic decrease in
available surface area. Such blocking could drastically reduce stripper performance. This is
a difficult hypothesis to check directly, but the scrubber behavior supports this view.
V-3
-------
APPENDIX VI. CORROSION IN MINIPLANT EQUIPMENT
At the beginning of the testing we employed three Eco gear pumps made of Hastelloy B
bodies with carbon wear plates and bearings. This construction was recommended for the ap-
plication by the manufacturer, whose experience indicated that these materials would withstand
the corrosive nature of hot sulfuric acid and, in fact, claimed that other similar pumps were
sold for this purpose. Examination of standard corrosion tables indicated that the Hastelloy B
material was marginal for long-term application, but should be satisfactory for a period of
several months.
One pump was purchased early in the program, tested under the desired operating con-
ditions, and showed the capability of pumping over one gallon per minute of hot sulfuric acid
(only 0.4 GPM was required for the test series). Upon this successful evaluation, two more
identical pumps were purchased and installed in the system.
When the testing was begun, it was noticed that the pump capacity seemed to drop some-
what as the temperature increased. Extended operation of the system caused all three pumps
to eventually lose their capability to pump the minimum required quantity of acid. The pumps
failed in order of their installation into the system. Examination of the interior of the pumps
showed that there was extensive corrosion and erosion of the pump bodies as well as the drive
shaft and the drive gears. The pump manufacturer was shown the results of only a few days of
operation, and he recommended replacement of the Hastelloy B drive gear with a carbon drive
gear, but this did not prove to be adequate. The corrosion of the pump bodies created a large
clearance between the pumping gears and the bodies, and the resulting back slip of fluid pre-
vented adequate pumping.
The pump manufacturer cooperated with us to the extent that he supplied us with sam-
ples of his pump parts to evaluate in the desired application, but it is clear that the highly cor-
jrosive conditions aggravated by the highly turbulent conditions present in a gear pump plus the
presence of air bubbles causing aeration of the corrosive fluid all combine to rule out the use
of available materials in our system. As shown in Fig. VI-1, samples of Hastelloy B, Hastelloy
C Carpenter 20, and stainless steel 316L all failed quite badly under the plant's operating con-
ditions. (Note; An obvious question would be how our Carpenter 20 piston pumps have survived
almost 2 yr of intermittant use without apparent failure. The answer appears to be in the na-
VI-1
-------
ALLOY 20
HASTELLOY C
cc
0 •
HASTELLOY B
316 STAINLESS
Fig. VI-1. Photographs of corroded pump parts showing comparison with untreated parts;
the Hastelloy B sample came from one of the Eco pumps in miniplant service.'
VI-2
-------
turc of die piston pump, which is essentially a low-speed mechanism with no tight tolerance
requirements. Both the piston and the pump head cavity could corrode slowly under the non-
turbulent conditions of this pump's operation and there would be little or no danger of pump
failure. The only problem might involve the ball checks and, indeed, these did show signs of
excessive corrosion and reduced operating efficiency.)
This experience with the pumps did provide a qualitative input to the program in terms
of the utility of various materials of construction. Clearly, Hastelloy B and C arc incapable of
standing up to the corrosive nature of the liquid system. Carpenter 20 is marginal and may be
of value in certain applications.
B. Strippers
As mentioned in Appendix II, examination of the strippers after 2 to 3 weeks of opera-
tion showed that the Teflon coating had undergone catastrophic failure. When the cover of one
of the strippers was removed, die appearance of the coating was as shown in Fig. VI-2. The
acid had obviously penetrated the Teflon at a great many points, apparently through pinholes
in the coating. The acid then corroded the stainless steel of the stripper walls and the coating
was peeled back as the corrosion occurred in a circular manner. The result was that the coat-
ing failure took the form of a great many small circles being cut out of the Teflon. Many of
these circles interconnected with each other and some of the coating came off in lace-like
films. In other areas the coating was completely removed, and there was a thick (up to 1/4 in.)
buildup of precipitated metal salts on the walls. It should be noted that the coating used was
recommended by duPont for this application, with the actual coating applied by a duPont ap-
proved vendor.
A light-colored precipitate was obtained by cooling the acid circulated through the sys-
tem. X-ray diffraction analysis of this precipitate disclosed that the material was Fc-^O.).,.
This experience indicates that Teflon coating is virtually worthless in this type of appli-
cation. It is also apparent that Type 316L stainless steel is unusable for hot nitrosylsulfuric
acid solutions. It should be pointed out that Teflon lining of equipment has not been ruled out
by this work and remains to be tested. (Teflon lining has the advantage of being a relatively
thick, continuous layer applied by extrusion techniques as opposed to the thin spray-and-bake
application of die Teflon coating which is prone to pinholes.)
VI-3
-------
Fig. VI-2. Photograph of the top of a stripper column after 3 weeks service showing
damaged coating.
VI-4
-------
APPENDIX VII. MATHEMATICAL ANALYSIS OF SCRUBBER/STRIPPER DATA
Table VII-1. Tyco Scrubbing Data Summary Sheet
Table VII-2. Average Values of NO Error
X.
Table VII.3. Tyco Scrubbing Data
Table VII-4. OAF Scrubbing Data
Table VII-5. Computer Program for Tyco Data
Table VII-6. Computer Program for GAP Data
Table VII-7. Rate Constants from the Tyco Data
Table VII-8. Rate Constants from the GAP Data
VII-1
-------
Table VII-1. Tyco Scrubbing Data Summary Sheet
RUN
82AB
82BD
B2CB
83AP
83BP
83CPB
84AB
84BA
84CA
B5AB
05DT
85CPA
86A ,
86BPD
86CB
87A
878
87C
92 A A
92B. i
92CA
93AA
93BA
93CA
94 AB
94BA
94CA
95AB
95BA
95CA
96A.,
96 B A
96CA
97AA
978A
97CA
4.9
4.9
4.9
8.8
a. a
8.0
6.0
6.0
6.0
1.4
1.4
1.4
2.9
2.9
2.9
4.9
4.9
4.9
4.9
4.9
4.9
2.6
2.6
2.6
6.0
6.0
6.0
8.8
8.8
8.8
1.5
1.5
1.5
4.9
4.9
4.9
INLET
235
500
925
240
475
10'i.i
275
50:
1050
250
50.')
10.J.
2'5
70.!
1050
21:5
500
10 :>
250
470
llOd
2?5
475
1070
240
500
1050
180
500
950
290
500
100.)
225
500
1090
NO
27 FT
10i)
150
250
120
225
375
125
150
250
150
ion
75
90
140
125
110
150
180
125
150
120
85
150
200
225
250
350
135
280
400
150
175
200
100
220
220
N02
INLET
265
470
1050
240
520
925
260
530
1050
210
540
110.)
230
525
100 )
260
520
1000
270
560
1 25
280
50;)
1200
230
540
1 00
300
500
910
240
520
900
280
50
1025
27 FT
150
150
490
65
320
45
80
190
415
5
70
90
140
50
15
120
205
365
125
40
145
0
0
10
30
90
300
135
190
310
0
35
10
175
15b
360
DNO ON02
135
40 i
675
120
250
625
150
350
80;,
400
925
135
560
925
H5
40;)
820
125
320
980
140
325
870
15
300
70!)
45
270
500
140
375
80
125
280
870
1!.5
320
560
175
20;;
470
180
340
635
205
470
1010
90
470
985
140
315
635
145
520
980
280
50'i
1 90
200
450
80;)
165
310
600
240
485
890
105
345
6 5
NOX ER
20
80
li.5
-5.,
15
-30
10
165
-105
-70
-85
M r\
9fl
-60
-25
-140
-320
-185
-150
-lO.i
-120
-40
-50
-110
-90
20
-65
205
VII-2
-------
Table VII-2. Average Values of NO Error (at 95% Confidence Level)
A
For 80's NO error = 32 ± 180
As
For 90's NO error = - 91 ± 220
X
For80's490's NO error = - 30 ± 230
A
For 250 ppm NO error = - 58 ± 140
X
For 500 ppm NO error = - 41 ± 200
X
For 1000 ppm NO error = 10 ± 300
X
VII-3
-------
Table VH-3. Tyco Scrubbing Data
NO
RUN
82AB
82BO
82CB
83AP
83BP
83CPU
04AU
84L5A
U4CA
(35 AO
8bD :
flbCPA
86 A •
86BPU
86CO
87A
878
87C
92AA
92Bi;
92CA
93 A,;
93BA
93CA
94 AB
94BA
94CA
95AB
9b8A
95CA
96AA
968A
96CA
97AA
9713A
97CA
Q
4.9
4.9
4.9
a.a
a.a
a.a
6.0
6.0
6.0
1 .4
1.4
1.4
2.9
2.9
2.9
4.9
4.9
4.9
4.9
4.9
4.9
2.6
2.6
2.6
6.0
6.0
6.0
a.a
a.a
a.a
1.5
1.5
1.5
4.9
4.9
4.9
INLET
235
550
92b
240
475
10:. 0
275
bOu
lUbil
2bU
50.'
lOUU
22b
70n
lObO
22b
bbO
10U ^
2bO
47u
IICK)
22b
47b
1070
240
5bO
1050
180
bbu
950
290
5:50
1000
2^b
50U
1090
9.bFT
200
275
b')0
150
350
b7b
175
27b
b- 0
175
160
IbU
IbU
230
17b
140
30!)
390
165
2,Jb
320
130
20:)
30U
2^5
3bO
5bO
150
40 1
640
140
225
350
125
350
50;)
19 FT
10:)
175
40D
125
275
460
125
17b
32b
17b
12b
90
ion
Ibi)
12b
12b
Ibb
2bO
140
160
175
100
13b
20
22b
280
375
140
30 0
475
1:0
225
225
100
250
285
27 FT
10!)
IbO
2bU
120
225
375
125
150
2bO
150
10.'
7b
90
141)
12b
i ; u
IbO
lao
125
ibn
120
ab
IbO
200
22b
2bO
350
13b
280
40;)
150
175
200
100
220
220
INLET
265
470
1050
240
52 (J
92b
26U
b3!)
10b()
210
bin
110; >
23u
525
10 ' • J
T(j(J
b?:)
1 0 : i . i
27!)
bbt)
1125
28!)
bnn
12!).'
23U
b 4 D
HUiJ
300
50 (J
910
240
52U
90u
280
50 J
1025
N02
9.5FT
2UO
240
680
14!)
435
62b
120
2b()
b3(J
70
IbiJ
IfllJ
13;)
1 n
, .^
3;>;i
L>» U
1 f > 1 1
13.)
42!)
2b
0
190
12!)
230
630
220
3ta!)
b70
120
380
210
220
29 b
b90
19 FT
180
230
95
37b
bib
12!)
271)
bib
2b
1 10
130
Ibi)
9b
bii
' b'i
<:I,D
4- -b
Ibi)
2o;j
10
4il
7:j
IbO
470
18b
420
bO
20
90
19b
225
50,
27 KT
1 bO
IbO
4 ') II
ijb
32!)
4V)
fifl
1 90
4 ; ^
!-,
7i)
'•»!)
1 •' !j
* u}
1 a. i J
20L,
3o b
12b
;» ' i
145
(J
0
1 i'
.*) : 1
1 )i '•
3i.
1 s^-
i - • . j
1 -if'
3 1 n
j
3b
1 1 1
1 7b
Ibb
3di)
VII-4
-------
Table VII-4. GAP Scrubbing Data
10
RUN
NO
INLET OUTLET
N02
INLET OUTLET
UNO
NOX ER>
14-1
14-2
14-3
14-4
14-5
14-6
15-1
15-2
15-3
15-4
15-5
15-6
17-1
17-2
17-3
17-4
17-5
17-6
18-1
18-2
18-3
18-4
18-5
18-6
273
370
475
575
695
013
256
343
451
555
664
769
253
366
462
598
681
758
253
345
450
552
651
769
215
268
330
368
434
553
204
250
331
393
463
521
187
26i,
331
405
4
-------
Table VII-5. Computer Program for Tyco Data
/RATES/
11.02 JSJM/ CALCULATION OF RATE CONSTANTS FOR C-141 SCRUBBING EXPERIMENTS
ON NO AND N02
11.03 A=PI/36.0. CJ=0
11.04 H1=26.5.H2=19.0.H4=9.5
11.05 DEMAND INPfOUP
11.06 OPEN INP FOR INPUT AS FILE 3
11.07 OPEN OUP FOR OUTPUT AS FILE 4
11.08 TYPE "
RUN GAS HEIGHT AHG-1 PO/PX LNIPO/PX) 1/PO-l/PX Kl «2
fl
11.21 READ FROM 3: X .Q.N20.N21 .N22.N24 »N10»N11»N12.N14
11.211 DO PART 13
11.22 00 PART 12
11.225 CJ=CJ-H
11.226 DO PART 14 IF FP CJ/5=0 IPAGING ROUTINE
11.23 TO STEP 11.21
12.21 R11=(N10/N11)»R12=(N10/N12}.R14=(N10/N14)
12.22 L11=LOG(R11) .L12=LOG (R12 ) .L14=LOG(R14)
12.23 AG1=(A*H1)/Q.AG2=(A*H2)/Q.AG4=(A*H4)/Q
12.2t F11=L11/AG1»F12=L12/AG2.F14=L14/AG4
12.311 R21=(N20/N21)»R22=(N20/N22).R24=(N20/N24)
12.312 L21=LOG (R21 > » L22=LOG ( R22) »L24=LOG (R24 )
12.313 F21=L21/AG1»F22=L22/AG2.F24=L24/AG4
12.321 D11=(1/N10)-(1/N11).D12=<1/N10)-(1/N12)»D14=<1/N10>-(1/N14)
12.322 S11=-D11/AG1.S12=-D12/AG2»S14=-014/AG4
12 .331 D21= (1/N20) -(1/N21) .022= (1/N20 ) -(1/N22 ) .024= (1/N20 ) - (1/N24)
12.332 S21=-D21/AG1»S22=-022/AG2.S24=-D24/AG4
12.101 TYPE IN FORM 69: X.HI,AG1»R11.Lll.Dll.Fll,S11
12.4011 WRITE ON 4 IN FORM 69! X»H1»AG1»R11^L11^D11»F11rSll
12.402 TYPE IN FORM 69: X»H2»AG2,R12»L12,D12tF12»S12
12.4021 WRITE ON 4 IN FORM 69! X»H2»AG2»R12»L12»D12»F12»S12
12.403 TYPE IN FORM 69: X »H4 »AG1 »R11»L11 »D14 »F14. »S11
12.4031 WRITE ON 4 IN FORM 69: X .HI »AG4 .Rl^.Ll^Ol^Flt, S14
12.10H TYPE " "
12.411 TYPE IN FORM 70: X»H1»AG1»R21»L21»D21»F21»S21
12.4111 WRITE ON «* IN FORM 70: X »H1» AG1 >R21 >L21.021 .F21» S21
12.412 TYPE IN FORM 70: XrH2rAG2.R22.L22.D22.F22.S22
12.4121 WRITE ON 4 IN FORM 70: X.H2.AG2.R22.L22.D22.F22.S22
12.413 TYPE IN FORM 70: X »H4 . AG4 .R24 >L24^ .024 »F24> .524
12.4131 WRITE ON 4 IN FORM 70: X.H4.AG4.R24.L24.D24.F24.S24
12.414 TYPE "
tf
13.11 tSHORT PART TO IGNORE ANY INCOMPLETE DATA SET
13.12 CM=0.000001
13.21 TO STEP 11.21 IF Q=CM
13.22 TO STEP 11.21 IF N20=CM
13.23 TO STEP 11.21 IF N21=CM
13.24 TO STEP 11.21 IF N22=CM
13.25 TO STEP 11.21 IF N24=CM
13.26 TO STEP 11.21 IF N10=CM
13.27 TO STEP 11.21 IF N11=CM
13.28 TO STEP 11.21 IF N12=CM
13.29 TO STEP 11.21 IF N14=CM
14.11 TYPE "SL
n
14.12 DO STEP 11.08
VII-6
-------
Table VII-6. Computer Program for OAP Data10
/GRATES/
j.1. 02 ISJM/ CALCULATION OF RATE CONSTANTS FOR C-141 SCRUBBING EXPERIMENTS
ON NO AND N02
11.05 A=(PI/<*>*(2. 25/12)*(2. 25/12) , CJ=0
1 1.04 Hl=1.5
1 i .05 DEMAND OUP
11.06 OPEN /L6D2/ FOR INPUT AS FILE 3
11.07 OPEN OUP FOR OUTPUT AS FILE 4
11.08 TYPE "
RUN GAS HEIGHT AHG-1 PO/PX LNIPO/PX) 1/PO-l/PX Kl K2
11.21 READ FROM 3! X >N20 »N21 rNIO »N11
1 .22 DO PART 12
1 1.225 CJ=CJ-H
1 .226 DO PART 1«V IF FP CJ/30=0 » PAGING ROUTINE
11.23 TO STEP 11.21
12.21 R11=(N10/N11)
12.22 L11=LOG(R11)
12.23 AG1=(A*H1)/Q
12. 2M- F11=L11/AG1
12.311 R21=(N20/N21)
12.312 L21=LOG(R21)
12.313 F21=L21/AG1
,2.321 D11=(1/N10)-(1/NU>
12.32.: S11=-D11/AG1
12.331 D21=(1/N20)-(1/N21)
12.332 S21=-D21/AG1
12.401 TYPE IN FORM 69: X rHl t AG1»R11 .LI 1»011 »F 11 »S11
12. H011 WRITE ON 4 IN FORM 69: X .HI ,AG1 .Rll »L1 1 »D11 ,F1 1 »S1 1
12.411 TYPE IN FORM 70: X >H1 »AG1 .R21 »L21 »D21 >F21 »S21
12.41H WRITE ON 4 IN FORM 70: X »H1 » AG1 »R21 »L21 »D21 »F21 »S21
1<+.11 TYPE "8L
1<+.12 00 STEP 11.08
lb.ll »SJM-DRC» CONVERSION OF LOU GARCIA GRAPHICAL DATA , FEB 1972
.,5.12 Q=20/60» NCT=2^» CNT=0
15.21 OPEN /LGD/ FOR INPUT AS FILE b
!5.22 OPEN /LGD2/ FOR OUTPUT AS FILE 6
15.31 READ FROM 5: X»X1C» YlCr X2C • Y2C
15.32 CNT=CNT + 1
l5.4l N10=100*10-(0.0752*X1C )
ib.42 R1A=(2.5E-5)*10-(0.07265*Y1C)
1«5.£*3 Nll = N10-( (R1A*(9.91E6)/60)/Q)
t>3.46 N20=100*10- (0.07525*X2C )
•.5.17 R2A=(2.5E-5)*10»(0.07265*Y2C)
15.18 N21=N20-( (R2A*(9.91E6)/60)/Q)
!5.bl WRITE ON 6 IN FORM 6: X> N20r N21 • N10» Nil
lb.6l TO STEP 15.81 IF CNT=NCT
15.62 TO STEP 15.31
!5.81 CLOSE 5
15.82 CLOSE 6
15.83 DO PART 11
61
VII-7
-------
Table VII-7. Rate Constants from the Tyco Data
RUN
GAS HEIGHT AHG-1
PO/PX
LN4E-oi
421E-02
628E-01
8H4E-01
421E-02
2.628E-01
1.
9.
2.
1.
9.
8' 4E-01
421E-02
628E-01
884E-01
421E-02
2.
2.
1.
1.
1.
1.
3.
3.
2.
3.
2.
1.
3.
2.
1.
2.
1.
1.
2.
1.
1.
3.
2.
1.
2.
1.
1.
1.
1.
1.
350E+00
350E+00
175E+00
767E+00
472E+00
325E+00
6 -7E+00
143E+00
OOOE+00
133E+00
043E+00
958E+00
700E+00
312E+00
682E-t-00
143E-I-00
750E-I-O.J
544E+00
OOOE+00
920E-»-00
600E+00
692E+OU
526E+00
714E+00
111E+00
727E+00
357E+00
625E+0!.
387E+00
195E+00
8.544E-01
8.544E-01
1.613E-01
5.691E-01
3.868E-01
2.814E-01
1.299E+00
1.145E+OU
6.931E-01
1.142E+on
7.147E-01
6.721E-01
1.308E+00
8.383E-01
5.199E-01
7.621E-01
5.596E-01
4.345E-01
6.931E-01
6.523E-01
4.700E-01
1.306E+00
9.268E-01
5.390E-01
7.U72E-01
5.465E-01
3.051E-01
4.855E-01
3.269E-01
1.785E-01
-5.745E-03
-5.7U5E-03
-7.447E-04
-2.893E-03
-1.782E-03
-1.226E-03
-4.848E-03
-3.896E-03
-1.818E-03
-U.539E-03
-2.220E-03
-2.039E-03
-2.919E-03
-1.419E-03
-7.371E-04
-1.08nE-03
-7.143E-01
-5.182E-04
-4.167E-03
-3.833E-03
-2.50 E-03
-1.122E-02
-6.360E-03
-2.976E-03
-2.339E-03
-1.531E-03
-7.519E-04
-1.202E-03
-7.136E-04
-3.758E-04
1
2
9
1
1
1
2
3
4
2
2
3
2
2
3
1
1
2
2
3
4
n
M
5
2
2
3
1
1
1
.eioE-i-of)
.52bE+0
.532E-01
.206E+0'
. 143E+00
.6f,3E*0
.753E+0
.384E-I-0.)
.097E+0:i
.420E1-0
. 112E+0
.972EtO -
.772E+0 )
.477E+0 !
.073E+0!i
.615E*0
.65aEfQ
.568EfO
.638E+0,-
.462E-t-0'o
.989E<-0 i
.971E + 0::
.919E+0 .
.721E+0
,843Eton
.901E+00
.242E+00
,848E-t-Oii
.735E>OU
.895E+0(J
1.217E-02
1.698E-02
4.401E-03
6.130E-D3
5.26i. E-03
7.249E-03
1.027E-02
1.1CJ1E-!J2
1.075E-02
9.GKJE-03
0.501E-D3
1.20'iE-fi:.'
6.105E-03
4.193E-11 1
4.357£-,-)3
2.306E-n^
2.1. £-1)3
3.063E-..K.
1.506E-02
2.035E-02
2.G54E-0?
4.2f>9E-02
3..',75E-02
3.159E-02
8.901E-03
8.126E-03
7.981E-03
4.574E-03
3.947E-03
3.989E-03
VII-8
-------
RUN
Table VII-7. (Com.)
GAS HEIGHT AHG-1
PO/PX
LN(POXPX) 1/PO-l/PX
Kl
03CPB
03CPB
83CPB
03CPB
O3CPO
83CPB
-4AO
O4AO
O^AO
flUAO
fli+AU
flUAB
8HOA
04BA
fl4BA
Q4BA
8<+BA
Q4DA
fjfC A
rj4C A
84C A
04CA
84CA
84CA
fl'.jAO
of) AB
05AH
flrjAB
"NO"
"NO"
"NO"
"NC2"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
26.50
19.00
9.50
26.50
19.0
9.50
26.50
19.0.)
9.50
26.50
19.0 >
9.50
26.50
19.00
9.50
26.50
19. On
9.50
26.50
19.0D
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.0.)
9.50
2
1
9
2
1
9
3
2
1
3
2
1
3
2
1
3
2
1
3
2
1
3
2
1
1
1
5
1
1
5
.628E-01
.0 V4E-01
.421E-02
.62RE-01
.R.-4E-01
.421E-02
.854E-01
.763E-01
.3R2E-01
.B54E-01
.703E-01
.302E-01
.854E-01
.763E-01
.3R2E-01
.854E-01
.763E-01
.3O2E-01
.8-54E-01
.763E-01
.382E-01
.854E-01
.763E-01
.3R2E-01
.652E+00
.184E+0
.92?E-01
.652E+0'
. 1R4E+0<|
.92 E-01
2
2
1
2
1
1
2
2
1
3
2
2
3
2
1
2
1
2
^
3
1
2
2
1
1
1
1
4
8
3
.6-,7E*0
.174E + O.I
.739E-I-0;)
.03^E + 0,i
.796E*00
.480E>no
.200E+0
,20f)E-»-Or)
.571E+OM
.250E+0;)
.167E+00
.167E+OG
.3 E+0
.857E+01
.810E+00
.7Q9E-t-Oi'i
.963E+OH
.030E-fOr,
.200E+00
.231E40.)
.909E400
.530E+00
.039E+0
.981E+00
.6f.7E+0,;
.429E+00
.429E+OU
.200E+01
.40HE+00
.OOi'E + 00
9
7
5
7
5
3
7
7
4
1
7
7
1
1
5
1
6
7
1
1
6
9
7
6
5
3
3
3
2
1
.800E-01
.765E-01
.b34E-01
.095E-01
.856E-01
.920E-01
.Rn5E-01
.8''5E-01
.520E-f)l
. 179E + OI
,732E-ni
.732E-01
.204E+O.S
.OSOE + O'i
.978E-01
.026E+0
.745E-01
,12,'E-Ol
.435E + 0.'
.173E + 0 -'
.46.. E-01
.283E-01
.124E-01
.837E-01
.108E-01
.567E-01
.567E-01
.738E+O.I
.128E+0'.!
. Q9C>E+QC,
-1.
-1 .
-7.
— 1 i
-0.
"* ^J <
-4 ,
-2,
-a.
— li
^h
-4
-3
-1
-3
™ 1
~ i
-3
-2
-0
-1
-9
-9
-2
-1
-1
-1
-3
-9
,0. .7E-D3
, 174E-H3
, 391E-04
. I 7E-03
.607E-04
, in')E-04
. v>4F-m
. 304E-D3
.078E-03
.654E-03
."a7F.-03
.4R7F-03
.6. .7E-03
.714E-03
.63--.E-03
.376E-03
.017E-03
,95'.C-03
.048E-03
.125E-03
.658E-04
.457E-03
.894E-04
.34 C-04
.6»7E-03
.714E-03
.714E-03
.952E-01
.524E-02
.524E-03
j 75°Ef
4.12 LEfO .'-..^.Vltr-i!^
G.H74F.-.-0 7.n4uf-;..',
2.70 E4-.I . 4.?.-:'-)F -,,.',
3.1 ORE +i") ! 4.'.")r',(>.C;-(l?)
4.16'F+n 5.:.!inr-fi -,
2.046E+0 1. ' .*,^t.-;u'
2,R53F+n- I .',7 if- )-
.^.271E + 0 : i .'ji'U,-- ,•
3.nr,nH.tn ?. <-v - ,.
2.79t-5Efii ' ..,.'->-- i/
r>/-9(iF+.) .•,.,'•.'•.!.-!./
3.124E + 0 ^ .:-V. R- •
3 .7Q:'F rf) 1 . "j4 f-.. '
4.327F + H ' 1 . .:.WL-:..
2.6 2E+0 . f'..7,.:ir- .-,
2.4- IF. + O • 6.S7 >'.-•• •
5.154E+0 l.-».''.l>.i''
3.723E + 0 7.9.iVP- :,
4.24-E + O 7..',n F-, 7,
4.6flnEt : f,.?.. F-,. ;,
2.40REr,l '. T-\ It-,,.',
2.57RE+0 J.:-,anr- •,
4.9-Rr+n .L?-..'.^-,::
3.092E-01 l.(jl4E-iU
3.012E-01 1.4 7E-0.1,
G.023E-ni r.f^SE-n-,
2.;.'63E + U. l..;i2E-iii
1.797Ef'l 2.97r>E-l)r
1.85 Erii 1 .f.DOE-n,1
VII-9
-------
Table VU-7. (Cent.)
RUN
GAS HEIGHT AHG-1
PO/PX
LN(PO/PX) 1/PO-l/PX
Kl
K2
85BI,
8561)
85BO
85BI5
85BI1
858(3
85CPA
SSCPA
85CPA
85CPA
85CPA
85CPA
86AA
86AA
S6AA
86 A A
86AA
86AA
86BPD
86BPD
86BPD
86BPD
86BPD
flfiRPD
86CB
86CB
86CB
86CB
flACR
86CB
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
1.652E+00
1.184E+0
5.922E-01
1.652E+00
1.184E+00
5.922E-01
1.652E+00
1.18UE+00
5.922E-01
1.652E+00
1.184E+00
5.922E-01
7.971E-01
5.717E-01
2.859E-01
7.974E-01
5.717E-01
2.859E-01
7.974E-01
5.717E-01
2.85QE-01
7.97UE-01
S.717E-01
2.859E-01
7.974E-01
S.717E-01
2.859E-01
7.97«*E-01
5.717E-01
2.859E-01
5.00UE+00
4.000E+00
3.125E+00
7.714E+00
4.909E+00
3.600E+00
1.333E+01
1.111E+01
6.6G7E+00
1.222E+01
8.462E+00
6.111E+00
2.500E+00
2.250E+00
1.500E+On
1.6«»3E+0
1.533E+00
1.769E+00
5.000E+00
1.667E+00
3.043E+00
9.5f5E+00
5.526E+00
4.773E+00
8.^00E+00
8.^ooE+on
6.000E+00
6.667E+01
2.000E+01
i.oonE+oi
2.590E+00 -1.233E-02
2.108E+00 -l.OUE-02
1.897E-1-00 -5.6o7E-03
2.503E-I-00 -1.020E-02
2.136E+00 -6.783E-03
1.810E+00 -
9.163E-01 -6.667E-03
8.109E-01 -5.556E-03
4.055E-01 -2.22rE-03
^.964-E-Ol -2.795E-03
4.27*tE-01 -2.319E-03
5.705E-01 -3.34t.E-03
1.609E+00 -5.714E-03
1.540E+00 -5.238E-03
1.113E+00 -2.919E-03
2.256E+00 -1.628E-02
1.710E+00 -8.622E-03
1.563E+00 -7.186E-03
2.128E+00 -7.0^8E-03
2.128E+00 -7.0«t8E-03
1.792E+00 -4.762E-03
<+.200E+0 -6.567E-02
2.996E+OU -1.900E-02
2.303E+00 -9.000E-03
1.568E+00
1.515E-»-nn
1.803E+00
3.057E+0.)
1.149E-1-0
l.«+18E+On
l.«H8E+Of)
6.2?5E-01
7.176E-01
2.018E+00
3.893E+00
2.829E+0;
2.990E+.1
2.669E+00
3.722E+0,'
6.268E+00
5.267E+0.1
5.210E+0 t
8.055E+00
7.047E-n,'J
8.360E-03
9.717E-03
7.773E-03
U.056E-03
1.17nE-,)2
7.tfv..E-(13
9.162E-;! j
1.021E-0?
2.041E-02
l.riOBE-02
8.83-)E-03
1.237.E-02
8.235E-02
3. '23E-02
3.1U8E-0?
VII-10
-------
Table VH-7. (Cont.)
RUN GAS HEIGHT AHG-1
PO/PX
LN(POXPX) 1/PO-l/PX
Kl
K2
87A
#7A
87A
87A
87A
87A
87B
87B
87B
87D
878
87B
87C
87C
87C
87C
87C
87C
92AA
g2A'
c52 A A
92AA
92 A V
£?2B''
p2B> •
c?2B~
2BD
ej2B^
o2BH
"NO"
"NO"
" NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.0
9.50
26.50
19.00
9.50
26.50
19.0
9.50
26.50
19.0
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.5fl
U.720E-01
3.3R4E-01
1.692E-01
4.720E-01
3.3RUE-01
1.A92E-01
4.720E-01
3.3R4E-01
1.692E-01
4.720E-01
3..1R4E-01
t.ft92E-01
4.720E-01
3.384E-01
1.692E-01
4.720E-01
3 TO liC~ f\ 1
• »J Cl *T t U A
1.692E-01
4.720E-01
3.3R4E-01
1.692E-01
U.720E-01
3.38HE-01
1.A92E-01
4.720E-01
3,38tE-01
1.A92E-01
U.720E-01
3.38fE-01
1.692E-01
2(
1,
!'
2,
1,
1,
3
3
1
2
1
1
5
4
2
2
2
1
2
1
1
2
1
1
3
2
2
1
7
4
.045E+0,'
.800E+00
.607E+0
,167E-»-00
.733E+0
.105E+00
.667E+00
,333E>00
.833E+00
.537E+00
.962E+00
.1+86E-I-00
.5b6E+00
.00 E+00
.564E+00
.740E>00
.2f7E+00
,72tE+00
.OOOE+00
.786E+00
.515E+00
.160E-I-00
.800E+00
.6B7E+00
.133E+00
.937E+00
.089E+00
.^OOE+01
.OOOE+00
.308E+00
7,
5,
It
7,
5,
3,
1
1
6
9
6
3
1
1
9
1
8
5
6
5
^
7
5
5
1
1
7
2
1
1
•156E-01
,87fiE-01
.745E-01
.732E-01
.500E-01
.403E-01
.299E-t-On
.204E+OiJ
.061E-01
.308E-01
.711E-01
.959E-01
.715E+00
.386E+OI1
.416E-01
.008E+0;)
.097E-01
.t^7E-01
.931E-01
.798E-01
.15r)E-01
.701E-01
.878E-01
.232E-01
.142E+00
,078E+0:i
.360E-01
.639E+00
.946E+00
.460E+00
-4.
-3.
-2.
-4.
-2.
-1.
-4 .
-t.
-1.
-2,
-1,
-9,
^
-3
-1
-1
-1
-7
-I*
-3
-2
-4
-2
-2
-1
-4
-2
-2
-1
-5
646E-D3
5:>6E-D3
698E-03
487E-03
821E-03
•559E-03
,R4->E-03
>?f 2E-03
,51r.E-03
,95f.E-03
.851E-03
.341E-OU
.5h6E-03
.orn)E-n3
.564E-03
.710E-03
.217E-03
.2U1E-04
.0" E-03
.lt3E-03
,061E-n3
.296E-03
.963E-03
.516E-03
.539E-03
.122E-03
.317E-03
.321E-02
.071E-02
.907E-03
1.516E+0
1.737E+0-
2.804E + 0!!
1.63RE + 0':
1.626E-1-0.
2.011E-I-0:'
2.753E+t)n
3.558E+00
:s.583E+o;i
1.972E+0..
1.9 i?F. + 0^;
2.3^0E + n.;
3.63^+n.)
4.097E+0'i
S.OfiHE+n.i
2.136E+OH
2.393E + 0 •!
3.2,.'OE + i)
1.469E+OH
1 .714E+0''
2.456E-I-01.
1.632E + 0 ••
1.737E+On
3.093E+0'i
2.420E+00
3.184E+0"
^.354E+On
5.:.-92E+Oi)
5.751E+00
8.632E+0;^
9.R4r>E-03
1 .051E-Oi;
t .rj9SE-02
9. SORE-OS
M.3-.5F-0 ;
9.2UjF-i)3
1 .027F.-H,':
1 .?'j4F-i-)P
8. 95 E-O.'i
A.?.,lF-,n
S.-^nPK-.'i '
r. . ,?^ -,!.,
9.653F-M.'
fi.'>f> E-0.'
Q.2'*riE-fn
3.f.R..F-:i'>
,J«OO • r ^ ' ' . J
^ ? f\ il F ^ i ' "*t
*.*7V.-9E-n2
^.91PE-0?
3.16..E-02
3.1+91E-02
VII-11
-------
RUN
Table V1I-7. (Cont.)
GAS HEIGHT AHG-1
PO/PX
LN(POXPX) 1/PO-l/PX
Kl
K2
92CA
92CA
92CA
92CA
92CA
92CA
93AA
93AA
93AA
93A«
93AA
93BA
93BA
93BA
93BA
93BA
93BA
93CA
93CA
93CA
93CA
93CA
93CA
94 AR
94AB
94AB
9UAB
9UAB
9UAR
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
3
1
3
1
.720E-01
.384E-01
.692E-01
.720E-01
.384E-01
.692E-01
8.89«*E-01
6.377E-01
3.189E-01
8
6
3
8
6
3
8
6
3
8
6
3
a
6
3
3
2
1
3
2
1
.894E-01
.377E-01
.189E-01
.891+E-01
.377E-01
.189E-01
.377E-01
.189E-01
.894E-01
.377E-01
.189E-01
.894E-01
.377E-01
.189E-01
,85«4-E-01
.763E-01
.382E-01
.854E-01
.763E-01
.382E-01
9
6
3
7
2
2
2
1
2
2
1
3
3
2
5
5
5
5
5
3
1
3
6
1
1
1
.167E+00
.286E+00
.137E+00
.759E+00
.327E+00
.679E+00
.6U7E+00
.250E+00
.731E+00
.800E+07
.800E+01
.120E+01
.167E+00
.519E-I-00
.375E+00
.OOOE+07
.OnOE+07
.OOOE-t-07
.350E-fOn
.350E+00
.567E+00
.200E+02
.OOOE+01
.316E+00
.067E+00
.067E+00
.067E+00
7.667E+00
3.286E+00
1.917E+00
2.216E+00
1.838E+00
1.235E+00
2.049E+00
9.853E-01
9.731E-01
8.109E-01
5.486E-01
1
3
2
.715E+01
.332E+00
.iH6E-(-0()
1.153E+00
1.258E+0
8.650E-01
1
1
1
1
1
1
3
1
J
.773E+01
.773E+01
.773E+01
.677E-I-00
.677E+00
.272E+00
.101E+00
.843E+00
M c ii F~»n ?
n E^IL F •• 02
tl^flF-r-f)?
2.037E+00
1.190E+00
6.506E-01
-4.805E-03
-2.216E-03
-6.008E-03
-2.957E-03
-7.320E-03
-5.5b6E-03
-3.21flE-03
-l.OOOE+05
-9.643E-02
-3.643E-02
-4.561E-03
-5.302E-03
-2.895E-03
-l.OfViE+05
-1.0DfiE+05
-t.065E-03
-2.39 E-03
-9.917E-02
-2.417E-02
-1.430E-03
-2.778E-04
-2.778E-04
-2.778E-04
-2.899E-02
-9.938E-03
-3.986E-03
7.298E+0
4.341E+00
5.821E+00
1.272E+00
1.720E-K) ,
1.928E+01
7. 577^+0
1.296E+0 I
1.973E+0
2.713E + 0 i
2.780E+01
3.98r,E+00
5.383E+00
5.780E+0
1.674E-01
2.3 5E-01
H.671E-01
5.285E+OU
1.573E-02
1.420E-02
1.310E-02
1.273E-02
8.739E-03
0.r,19E-03
8.230E-0.5
8.712F-03
1 .019E-0?
I .r>i2F;-n i.
9.07RE-03
i!S,Efn5
7.523E-0 'i
3.790E-02
1.3R9E-02
l!o05E-03
2.01i'E-03
7.520E-02
3.f)9f,E-02
2.R .^E-0?
VII-12
-------
Table VH-7. (Cont.)
RUN GAS HEIGHT AHG-1
PO/PX
94BA
94BA
910A
gUBA
g^CA
guCA
gtCA
g5AB
g5AB
95AB
95AB
95AB
g5AB
gSBA
95BA
gSBA
95BA
gSBA
95BA
95CA
gSCA
gSCA
95CA
P5CA
95CA
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.0"
9.50
26.50
19.00
9.50
3.854E-01
2.763E-01
1.382E-01
3.B54E-01
2.763E-01
1.382E-01
3.854E-01
2.763E-01
1.382E-01
3.854E-01
2.763E-01
1.382E-01
2.628E-01
1.8R4E-01
9.421E-02
2.628E-01
9.421E-02
2.628E-01
1.8R^E-01
P.628E-01
1.88*fE-01
Q.421E-02
2.628E-01
g!421E-02
2.628E-01
9!l21E-02
2.200E+00
1.571E+00
6.000E+00
3.600E-I-00
1.929E+00
3.onoE+on
2.800E+00
1.909E+00
3.667E+00
2.340E+00
1.7^6E+00
1.333E+OU
1.286E+00
1.200E+00
2.222E+00
1.622E+00
1.364E+00
1.964E+00
1.833E+00
1.375E+00
2.632E+00
1.786E+00
1.389E+00
2.375E-1-00
2.000E+00
2.935E+00
2.167E+00
1.596E+00
LN(POXPX) 1/PO-l/PX
7.805E-01 -2.182E-03
6.751E-01 -1.753E-03
4.520E-01 -1.039E-03
1.792E+00 -9.259E-03
1.281E+00 -M..815E-03
6.568E-01 -1.720E-03
1.09QE+00 -1.905E-03
1.030E+00 -1.71^E-03
6.166E-01 -8.650E-04
1.29f'E+00 -2.424-E-03
8.503E-01 -1.219E-03
5.573E-01 -6.782E-04
2.877E-01 -1.852E-03
2.513E-01 -1.587E-03
1.823E-01 -1.1HE-03
7.985E-01 -1.074E-03
4.834E-01 -2.072E-03
3.102E-01 -1.212E-03
6.751E-01 -1.753E-03
6.061E-01 -1.515E-03
3.185E-01 -£
Kl
2.0U6E+0
9.676E-01 -3.263E-03
5.798E-01 -1.571E-03
3.285E-01 -7.778E-OU
8.650E-01 -1.I+47E-03
6.931E-01 -1.053E-03
3.950E-01 -5.099E-04
1.077E+00 -2.127E-03
7.732E-01 -1.282E-03
IU678E-01 -6.555E-04
K2
5.6 1E-0')
7.519E-03
1.
2.850E+Ou
3.726E-1-0
4..680E+0 .
3.371E+0
3.077E + 0 ••••
U.034E+0-
6.203E-03
6 . 2b' E — 0/>
«*.900E-.i3
1.095E + 0 i
1.935E + 0!)
3.039EtO';
2.5f» E+n ;
3.292E+0-!
2.569E + 0 i
3.217E+00
3.682E+0
3.077E+0 )
3.292E+00
3.679E+00
4.193E+00
7.n^7r-.).i
R.^p-E-n^
1. :.7')E-02
1.5 OE-02
1. ' 0'.F-n2
1.207E-0,?
fi.O'*2E-f).''j
7.237E-03
4.966E+00
5.50RE-03
5.587E-03
5.»tl2E-03
8.094E-03
fi!958E-03
VII-13
-------
RUN
Table VII-7. (Cont.)
GAS HEIGHT AHG-1
PO/PX
LN(POXPX) 1/PO-l/PX
Kl
K2
96AA "NO" 26.50 1.542E+00 1.933E+00 6.592E-01 -3.218E-03 4.276E-01 2.08KE-03
96AA "NO" 19.00 1.105E+00 2.636E+00 9.694E-01 -5.G43E-03 8.770E-01 5.105E-03
96AA "NO" 9.50 5.527E-01 2.071E+00 7.282E-01 -3.695E-03 1.318E+0:i 6.685E-03
96AA "N02" 26.50 1.512E+00 2.40i>E+07 1.699E+01 -l.OOOE+05 1.102E+01 6.486E+OU
96AA "N02" 19.00 1.105E+00 4.800E+00 1.569E+00 -1.583E-02 l.m9E+0;i 1.432E-02
96AA "N02" 9.50 5.527E-01 2.00IE+00 6.931E-01 -4.167E-03 1.254E+00 7.539E-07,
96BA
96BA
96BA
96BA
96BA
96BA
96CA
96CA
96CA
96CA
96CA
96CA
97AA
97AA
97AA
97AA
97AA
97 AM
97BA
97BA
97BA
97RA
97RA
97BA
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
26.50
19.00
9.50
1
1
5
1
1
5
1
1
5
1
1
5
(*
3
1
4
3
1
3
1
U
3
1
.542E+00
. losE-t-no
.r)27E-01
.542E+00
.105E+On
.527E-01
.5U2E+00
.105E+00
.527E-01
.542E+00
.105E+00
.527E-01
.720E-01
.384E-01
.692E-01
.720E-01
.384E-01
.692E-01
.720E-01
.384E-01
.692E-01
.720E-01
.384E-01
.692E-01
3
2
1
2
1
5
2
9
4
2
2
1
1
1
1
2
2
3
2
1
.143E+00
.4U4E+00
.486E+01
.600E-I-01
.368E+00
.onoE+oo
.857E+00
.onoE+oi
I286E+00
.250E+00
.250E-t-00
.800E+00
.600E+00
.436E>00
.273E+00
.273E+00
.OOOE+OO
.226E+00
.222E+00
.695E+00
1.145E+00
8.938E-01
S.938E-01
2.698E-I-00
3.258E+00
3.137E-01
1.609E+00
1.492E+00
1.050E-I-00
4.50!)E-»-0;)
ll455E+00
8.109E-01
8.109E-01
5.878E-01
4.700E-01
3.618E-01
2.412E-01
8.210E-01
6.931E-01
3.567E-01
1.171E+00
7.985E-01
5.276E-01
-3
-2
-2
-2
-7
-14.
-3
-1
-9
-1
-3
-5
-5
-3
-2
-1
-9
-2
-2
-8
-2
-1
.896E-03
.62i-E-03
.626E-03
.808E-02
.085E-Ot
.0: E-03
.857E-03
t8o9E-02
.OOOE-02
.651E-03
.5!)6E-03
.556E-03
.143E-03
.5157E-03
.545E-03
.OOOE-03
.571E-01
.452E-03
.390E-03
7.428E-01
8.006E-ni
1.617E+n
1.750E-t-0
5.675E-01
1.3U9E+0
2.919E+0.1
2.083E+0'/
1.718E-t-0",
2.397E-I-O.J
9.959E-01
1.069E+0.
1.425E+00
2.048E+01!
2.108E+0!i
2.360E-I-OI)
3.119E+OU
2.5^7E-0.1>
1.728E-02
1.282E-n7,
2.595E-03
3.i:f,E-03
3.360E-03
9.0n7E-03
1.17 E-i)?
1 .6<*2E-02
2.102E-02
4.540E-03
4. 601 E-03
5.737E-03
5.393E-03
5.9UE-03
5.06- E-03
9.432E-03
7.2 4E-03
8.215E-03
VU-14
-------
RUN
Table VII-7. (Cont.)
GAS HEIGHT AHG-1
PO/PX
LNtPO/PX) 1/PO-l/PX
Kl
K2
97CA
97CA
97CA
97CA
97CA
97CA
"NO"
"NO"
"NO"
"N02"
"N02"
"N02"
26.50
19. On
9.50
26.50
19.00
9.50
<+.720E-01
3.384E-01
1.692E-01
f .720E-01
3.3R4E-01
1.692E-01
4.955E+00
3.825E+00
2.180E+00
2.817E+00
2.050E+00
1.737E+00
1.600E+Ofi
1.341E+0)
7.793E-01
1.046E+00
7.178E-01
5.523E-01
-3
-2
-1
-1
-1
-7
.62RE-03
.591E-03
.083E-03
.802E-03
.024E-03
.193E-04
3
3
H
2
2
3
.391E+00
.964-E+OO
,60.'>E + Oi)
.217E+0(
.12iE*m,
.265E + 0^i
7
7
6
3
3
H
.687E-03
.f>50E-OJ
.39 'E-JI3
.019E-03
.(T27E-03
.251E-03
IN STATEMENT 11.21: END OF FILE ON INPUT
vn-15
-------
Table VII-8. Rate Constants from the GAP Data10
DO PART 15
OUP = »GAROUT
RUN
GAS HEIGHT AHG-1
PO/PX
LN(POXPX) 1/PO-l/PX
Kl
K2
14-1
11-1
._. lt-2
14-2
11-3
11-3
14-4
14-4
14-5
14-5
14-6
14-6
15-1
15-1
15-2
15-2
15-3
15-3
15-4
15-4
15-b
15-5
15-6
15-6
17-1
17-1
17-2
17-2
17-3
17-3
17-4
17-4
17-5
17-5
17-6
17-6
18-1
18-1
18-2
18-2
18-3
18-3
18-4
18-4
18-5
Tfr-5
18-6
18-6
"NO"
"N02"
•LNQ1
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"N0lf
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
"NO"
"N02"
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
.50
.50
,50.
.50
.50
.50
.50
.50
.50
.50
.50
.50
.50
.50
.50
.50
.50
.50
".50
.50
.50
.50
.50
f50
.50
.50
.50
.50
.50
.50
.50
.50
.50
.50
.50
.50
.50
.50
.50
.SO
.50
.50
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.213E-01
1.213E-01
1.213E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
1.243E-01
.50 1.243E-01
.50 1.243E-01
.50 1.243E-01
.50
.50
.50
1.243E-01
1.243E-01
1.243E-01
1.268E+00
1.243E+00
1_.38_QE±P_0
1.351E+00
1.439E+00
1.409E+00
1.561E+00
1.563E-t-OU
1.602E+00
1.578E+00
1.469E+00
1.913E+00
.254E+OU
1 .239E+00
1.330E>00
1.350E+00
1.361E+00
1.395E+00
1.412E+O.I
1.445E-KK)
1.43JE+0!)
1.465E+00
1.176E+OU
i.biie+oo
1.351E+00
1.514E+00
1.379E+00
1.516E+00
1.396E+OU
1.580E+0!)
1.477E+OH
1.653E+00
1.528E+OU
1.652E+00
1.561E+00
1.684E+00
1.277E+OU
1.202E+00
1.326E+OU
1.363E+0(J
1.342E+00
1.318E+OU
1.381E+OM
1.413E+00
1.453E+00
1.129E+00
1.172E+00
1.185E1-00
2.
2.
3_.
3.
3.
3.
4.
4.
4.
376E-01
177E-01
220E-01
009E-01
637E-01
128E-01
452E-01
463E-01
716E-01
4.561E-01
3.842E-01
6.486E-01
2.
2.
2.
2.
3.
3.
3.
3.
3.
3.
3.
Hi
3.
4.
3.
4.
3.
4.
3.
5.
4.
5.
4.
5.
2.
1.
2.
264E-01
116E-01
850E-01
999E-01
084E-01
327E-01
151E-01
682E-01
595E-01
822E-01
891E-01
125E-01
OuSE-01
149E-01
215E-01
161E-01
334E-01
576E-01
90UE-01
027E-01
238E-01
022E-01
455E-01
211E-01
4t;2E-01
839E-01
82aE-01
3.098E-01
2.913E-01
2.759E-01
3.
3.
3.
i.
3.
3.
226E-01
461E-01
738E-01
571E-01
867E-01
953E-01
-9
-9
-1
.822E-04
.347E-04
.026E-03
-9.892E-04
-9.232E-04
-9.087E-04
-9.756E-04
-1.023E-03
-ti.bhlE-04
-8.742E-04
-5.765E-04
-1.151E-03
-9
-9
-9
-9
-a
-8
-7
-«
-fa
-7
-6
-6
-1
-2
-1
-1
-a
-i
-7
-1
-7
-9
-7
-8
— J.
-8
-9
~-i
-7
-7
-6
-7
-6
-6
-6
-6
.937E-04
.361E-04
.613E-04
.974E-04
.009E-04
.741E-04
.42:JE-04
.0 '9E-04
.5iaE-04
.005E-04
.186E-04
.384E-04
.38. E-03
.035E-03
.035E-03
.482E-03
.563E-04
.3oOE-03
.973E-04
.163E-03
.749E-04
.817E-04
.402E-04
.&M3E-04
.091E-03
.102E-04
.45:jE-04
. 067E-03
.613E-04
.135E-04
.893E-04
.477E-U4
.9G2E-04
.582E-04
.140E-04
.168E-04
1.912E+OU
1.752E+0-J
2.592E+00
2.422E+OU
2.927E + 0'J
2.759E + 0!)
3.58JEfOii
3.592E+0,)
3.795E+UJ
3.670E+0',)
3.092E-I-0')
5.2 :OE+0 )
1.82, :£ + ()•)
1.727E+0 i
2.^94E+0:,
2.413EtO'f
2.482EtOn
2.67,'E + O !
2.7/,'EfO'i
2.064EtO!'
2.89JE+0
3.076E + 0;;
3.131E+0 ')
3. ^20E MI;;
2.419E+0
3..) 9EfO <
2.58,'.E + 0 i
3..549E+0 i
2.b83EfO!i
3.6fl3EtO-i
3.138EMD '
4.0't6Ei-0 i
3.41.EK)-.
4.042E*0'.
3.506E<-0,i
4.194E+00
1.965E+0 )
1.480EK) ,
2.27P.E + 0.,
2.493E-t-Oii
2.369EfO -..
2.2J1E+00
2.596E4-0 ;
2.785EfOJ
3.0.i9E+U:i
2.874E+OJ
3.112E+OD
3.182E+UO
7.905E-03
7.523E-U3
8.259E-03
7.962E-Ui
7.430E-Ui
7.314E-U3
7.8bl£-03
a.2J'Jt- )j
6.973E-'J^
7.03DE-03
4.b':-UE-(J3
9.2obt-03
7.9 .flt-U3
7.5.'i4E-()^
7.,-'3/t-U.^
«.U27t-!J5
6.4 c,E-U3
7.U3bE-rJ3
b.973E-D3
6.4 bt-0.}
b.i246L-:Jj
b.638E-UJ
4.97')E-03
.. b.l3;it-tu
1.' 7E-DJ
1.638E-02
G.J UL-JJ
1.I93E-U2
6..491E-03
1.D71E-02
f..417E-;)3
y.359E-03
b.;M7L-'J3
7.901t.-;i3
5.9b7E-03
7.149E-U3
f. . ?c;3L-n j
i'.521E-03
7.iilOE-!J3
8.5f>4E-(J3
6.127L-03
5.743E-Q3
5.M7E-03
6.i)lflE-03
5.6U3E-05
5.297E-03
4.9U2E-U3
4.964E-03
UJ
ERROR IN STATEMENT 11.21: END OF FILE ON INPUT
VII-16
-------
APPENDIX VIII. LABORATORY STRIPPER TESTS
VIII-1
-------
Table V1II-1
G
L (cc/min
Run Date Page (cc/min) at SIP) Gas
i
bo
1 27 July 60
2 61
3
4
5
6
7 28 July
8
9
10
11
62
62
64
I
65
12
13 29 July 67
14
15
16
17
18 30 July 69
19 2 Aug 71
20 2 Aug 71
4.5
5.4
4.6
4.2
3.7
3.7
4.3
4.9
4.5
4.3
4.3
4.1
4.9
4.8
4.1
5.5
5.0
8.3
5.1
100
250
100
250
100
250
100
250
100
250
250
250
100
250
100
250
250
100
250
% H2S04 % HNSOg Column Ac
Gas Jn Out In Out" Stripping NO, NO (°F) (°
N
A
2 '
T
)0 80 1
—
—
—
—
id
up
Fl Packing
.0 1.0 0 0.0 0.0 RT RT Sac
0.9 10 0.0 0.0 RT
0.9 10 0.3 0.0 170
0.9 10 0.4 0.0 170
0.6 40 0.7* ~5.0 300
0.3 70 0.7 -3.2 280
— 0.9 0.9 0 0.0 0.0 RT
—
—
—
t
*
N2 82 8
—
N2
N2
2
0 0.0 0.0 RT
0 0.3 0.0 172
1 0 0.3 0.0 174
0.07 90 -2* 0.0 270
0.17 80 -1.2 0.0 271
0.9 0 0.0 0.0 RT
0.9 0 0.0 0.0 RT
0 0.2 0.0 185
0 0.3 0.1 182
0.5 45 ~1 0.6 280
0.82 — — — — — Ca
0.74 0.42 43 0.0 0.0 RT RT
0.74 0.60 19 0.0 1-2 RT RT
Idles
rbon
Cold
Trap
Sieves
Cold
Trap
Sieves
Sieves
Comments
Runs 1-6 of doubtful
validity were, there-
fore, repeated as Runs
13-17
*NO2 fluctuating pro-
bably due to top of
column being cold
*NO2 fluctuating pro-
bably due to top of
column being cold
fS.G. = 1.742 at 82 °F
*S.G.= 1.735 at 84 "F
Column blocked with
fines. Wash with 80%
Heat sample lines at
outlet but not at top of
column
NO2 jumps up to ~1%
every time a sample is
taken even though no
air enters the system
-------
Table VIII-1. (Continued)
i
CO
G
L (cc/min
Run Date Page (cc/min) at STP) C
21 2 Ai
22
23
24
25
26 3 Aj.
27
28
29
30
31 4 Ai
32
33
34
35
36
37 5 At
38
39
40
41
g 71 5.0 100 I
4.0 250
3.9 100
4.9 250
72 ~4 100 t
g 74 8.0 100
6.1 250
6.5 1C
6.6
\ 6.6
g 77 5.7
83 3.3
83 3.3
to
77 3.3 200
2.7
2.8
g 84 2.9
85 2.9
85 2.6
87 3.2
87 2.6
42 6 Aug 88 2,8
43
44
45
46
88 2.3
89 2.8
91 4.1
91 4.7
% H,SO. % HNSOK
24 5 c
as In Out In Out Stri{
2 «i
Lir
7
N2
N2
t
!•
Lir 8
'2
Air
1
2 82 0
1
9 79 0.
81
79 2.
Column Acid
c % % Temp Column
jping NO0 NO (°F) (°F) Packing Dryer Comments
74 0.03 96 0.0 >5 173 B
0.02 97 0.0 >5 172
0.02 97 0.0 >5 320
0.02 97 0.0 >5 317
— — 0.2 0.0 320
62 0.85 8 0.3 0.0 RT
0.86 7 0.2 0.1 RT
0.94 0 0.3 0.0 181
0.03 97 2-14 >5 338
1 0.13 94 ~2 >5 326
80 2.1 0
— 0.85 0
— 0.85
—
—
—
0—0.
T — 0.
86 0
90 <0
03 £
02 9
09 9
01 11
19 0.4 >5 280
8 0.4 >5 271
-10 0.2 271
0.7 0.7 265
<0.4 >5 262
0.0 >5 287
T Cai
0 1.2 0.0 280 280
W 0.1 >5 190 185
0.0 >5 237 246
0.0 >5 270 268
0.0 >5 313 317
— 0.83 0.08 90 Ito6 0.0 190 185
—
—
—
0.04 95 ~1 0.0 226 230
<0.01 100 1.8 0.0 277 256
<0.01 100 1.6 0.0 304 296
— 0.01 0
.0 — 0.0 0.0 312 270 1
rbon Si
eves
Air run for only 15 min
NO2 climbing
Cold
Trap
High NC^ occurred on
flooded column. Returned
to lower NO2 when
column stopped flooding
As temp rose from 280
to 304 NO2 rose to about
8% and then decreased to
1.6% within an hour
-------
Table VIII-2
No.
1
2
3
4
5
6
7
8
9
10
11
Feed Acid
Concentration
(% HNSOg) *
0.59
0.59
0.59
0.68
0.83
0.73
0.59
0.83
0.83
0.83
0.83
Air Feed
Rate
(cc/min)
800
800
800
800
800
800
400
400
400
400
400
NO in Effluent (%)
Percent
N00 NO Denitration
1.50
1.52
1.20
3.20
3.52
1.30
1.06
2.16
2.96
3.76
3.36
0.11
0.11
0.10
0.09
0.17
0.63
0.04
0.29
0.27
0.16
0.19
99.0
99.3
**
99.0
98.9
**
**
71.1
99.0
**
**
Material
Balance
Error (%) t
16.5
25.7
41.2
-35.4
2.8
35.1
76.0
37.9
56.5
41.0
53.9
* Acid fed at 10 cc/min.
** ~ 99% denitrated.
f Material balance error =
NO (in) - NO (out)
X A.
NO (in)
A.
VII1-4
-------
APPENDIX IX. INSTRUMENTATION CALIBRATION
A. Calibration of the UV Spectrophotometer for NO2 at Concentrations Above 1%
During most of the early experimentation, the UV analyzer has been used for the anal-
ysis of NO2 at concentrations up to 1% NOg. During some of the stripper experimentation (Sec-
tion VIII of this report), the UV analyzer was used in cases where the expected NO« concentra-
tions were as high as 10%. To allow us to use the UV spectrometer in this NO0 concentration
£t
range, we have checked the response curve by preparing known gas samples. The dependence
of the response curve on flow rate, cell temperature (i.e., span setting), and diluent gas was
determined.
The gas samples were prepared by mixing (at constant flow rate) NO, O,, and Ar or
N2 in the N0x reactor, allowing sufficient time for greater than 98% conversion to NOg. The
residual NO concentration was determined and corrected for. Using properly calibrated flow
meters, it was possible to prepare gas mixtures containing up to 15% NO-. The UV spectro-
meter gain was set at one concentration, with a known gas sample standard mixture from
Matheson. Then the higher concentration samples we prepared were passed through the UV
and IR cells, and a UV spectrometer calibration curve was constructed. The spectrometer
response was found to be sensitive to cell temperature (span setting) flow rates but independent
of the diluent gas (N« or Ar), as expected. Three typical response curves are shown in Fig.
IX-1.
The unusual shape of the response curves indicates that a linear extrapolation from low
concentrations will give inappropriate analyses. The dual-valued curves make it difficult to
determine which of two possible concentrations might be indicated by any given recorder re-
sponse (e.g., 150). Representatives of duPont indicate that these problems can be most readily
overcome by using a shorter analysis cell. Up to this time a 15-in. cell had been used. A new
1.5-in. cell was installed in the spectrometer to overcome the problem.
The change in shape of the response curves, as slow rate and cell temperature are
varied, deserves special attention. It verifies the importance of heated sample lines, a ther-
mostatted sample warmer by the UV analyzer, and standard flow rates. This was used, in both
laboratory and miniplant experimentation.
B. IR Spectrometer Calibration
The IR spectrometer gain was first adjusted with a known gas mixture from Matheson.
IX-1
-------
300
o
CO
I
CD
OJ
OS
(H
0)
4-J
E
2
200
C/3
100
Hot 0.8 i/min
span 482
Hot 2.0 I /min
span 424
/min
!old 2.0
span 234
10
15
Mole % NO0
Fig. IX-1. UV spectrometer calibration curve for NO2 with the 15-in. cell.
IX-2
-------
The IR cell was calibrated for high NO concentrations (0 to 100%) by feeding known mixtures
of NO and Ar to the IR cell. The calibration curve is mildly dependent on flow rate. A typical
calibration curve is given in Fig. IX-2. The basic conclusion one may draw is that accurate
analysis for NO above the 1% level requires very careful standardization techniques, and such
measurements should be avoided, if possible.
IX-3
-------
200
(100% NO)
150
-------
APPENDIX X. OXIDATION OF NO TO NO2 FOR MINIPLANT GAS FEED
Miniplant experimentation requires a source of NO» for the oxidation of SO,. Sufficient
NO, is required to form 6000 ppm NO, at 12 SCFM. It is preferable to have an easily regulated
NO supply with less than 10% of the NO present as NO. Higher fractions of NO place an undue
** X
loading on the scrubber and the stripper.
Of the three possible sources of NO :
1. Commercial NO0,
£t
2. Homogeneous oxidation of NO, and
3. Catalytic oxidation of NO,
we conclude that the most expedient and satisfactory source is homogeneous oxidation of NO.
Use of commercially available NO™ involves additional handling and control problems which
may be avoided using NO as a source. Catalytic oxidation of NO with air and activated charcoal
(dry) has been shown to involve the potential dangers of thermal runaway at high NO concentra-
tions, e.g., temperature changes of 44 °F/sec have been observed with 11.6% NO in air.
Homogeneous oxidation of NO with O9 was studied using Ar and N- as diluent gases.
The gases were obtained in high pressure cylinders from Matheson and Airco. In-house com-
pressed air was used without drying. The reactor, a 4 ft. length of 4-in. glass pipe, was held
at an ambient temperature of about 70 °F. Gas sampling lines leading to the UV and IR cells
were 3/16-in. I.D. Teflon lines with a total length of about 30 ft. The flow rate from the reac-
o
tor, through the UV cell up to the IR cell, was 2 ft /min. The residence time in the sample
lines was about 0.2 sec. Residence time in the UV cell was less than 1 sec. One may thus
ignore sampling line-residence times in kinetic calculations. Experimental results are pre-
sented in Fig. X-l.
To allow for the performance of a sufficient number of experiments in a reasonable
time, we must pay careful attention to possible sources of analytical errors. In the present
case, we must consider the results obtainable with the UV and IR analyzers. As discussed in
Appendix IX, concentrations above 2% (as measured by either analyzer) may be expected to
give only semiquantitative information about the gas being analyzed. At concentrations between
500 ppm and 1%, low baseline drift and frequent calibrations yield data accurate to ± 2% of the
indicated value. We can most accurately evaluate rate constants by using data corresponding
to the 500 ppm in 1% regions and in concentration regions where the kinetic equations are not
X-l
-------
o
z
a
% NO as input
Fig. X-l. Conversion of NO to N02 at room temperature without a catalyst (argon flow
was 1740 ml/min).
X-2
-------
sensitive to small analytical errors (10 to 90% conversion of NO to NO,). These requirements
are met by the starred data in Fig. X-l, and the rate constants so obtained may be used to cal-
culate results which would otherwise require excessive experimentation.
The reaction 2NO -f 0- - 2NO, is expected to obey third-order kinetics, and this has
been verified by M. Bodenstein [Z. Elektrochem 22, 327 (1916) and Z. Physik. Chem., 100, 68
(1922) ]. If we denote the initial O, partial pressure by a (atm) , the initial NO partial pressure
by b (atm), the instantaneous partial pressure of NO- by x (atm), and time by t (sec), the rate
of formation of NO, is then given by:
tt
§= K(a-J) (b-x)2
The integrated form is:
i
or the equivalent form
1 1 , 1 2a-x b Kt
E ~ B=x
Given a value for a, b, x, and t, one may solve for the rate constant K. In our case, the
values were:
a =
b =
x =
t =
which leads to
0.169 atm O»
0.030 atm NO
0.86 b
44.1 sec
-2 -1
K = 29.3 atm sec .
Now that we have solved for K, it is possible to predict the residence time required for any
degree of conversion of NO to NOg and starting with any desired initial O2 and NO partial pres-
sures.
We know from experimental data of Fig. X-l that it is possible to achieve 95% or greater
conversion of NO to NO, at high O, concentrations. It is of interest to know what might happen
if air were used with NO at varying residence times. This is easily calculated. In Table X-l,
we present the time necessary to reach the specified conversion of NO to NO, starting with
various NO partial pressures.
The information which is of more direct use to us concerns the number of linear feet
required to generate sufficient NO, for 6000 ppm of usable NO, in a 12 SCFM gas stream. At
99% conversion of NO to NO,, one obtains 98% usable NO,, and to get 6000 ppm usable NO,,
one needs:
X-3
-------
Table X-l. Conversion of NO to NO with Air (See Text)
Starting
NO(%)
3
10
20
25
29
J. J.111^
0.8
23.4
8.6
6.50
7.09
9.1
VNS k/LSW^-AJ. 0-W* N.
0.9
53.2
20.2
17.0
20.9
34.2
/v t* r N^A I^JAV** yi^ v
0.95
113.2
44.0
39.8
54.1
119.8
—/
0.99
595
238
235
367
1532
12 SCF,!V °>006 = 0.0735 SCFM of NO.
0.98
Similarly for 95, 90, and 80% conversion, one requires 0.80, 0.90, and 0.120 SCFM of NO,
respectively.
To calculate F, the required number of cubic feet of NO reactor volume, we use the
X.
formula:
Where T = the time in seconds to achieve the desired degree of conversion, S = the number of
SCFM of NO feed required at the specified degree of conversion, 60 = the number of sec/min,
and F = the fraction of gas entering the reactor which is NO. The number of linear feet of re-
actor required is F/0.0873. Using this formula, Table X-2 and Fig. X-2 were prepared.
Fig. X-2 vividly illustrates that greater than 95% conversion of NO to NO, conversion
with air can be expected with 10 ft of piping as long as the portion of NO going into the NO
-A.
reactor is between 10 and 29%. It also illustrates that, if the portion of NO going into the NO
A
reactor is much less than 10%, drastically larger reactors are required to achieve high yields
of N02.
X-4
-------
Table X-2. Required Length of 4-in. NO Reactor to Oxidize
NO to NO2 with Air and Obtain 6000 ppm NO2 in
a 12 SCFM Stream by Dilution
Length of Reactor for Specified Conversion (ft)
Starting
N0(%)
3
10
20
25
29
0.8
17.8
1.97
0.76
0.66
0.71
0.9
30.5
3.44
1.47
1.41
2.02
0.95
57.7
6.72
3.03
3.28
6.32
0.99
278
33.4
16.5
20.6
74.1
X-5
-------
Fig. X-2. Conversion of NO to NO2 at room temperature without catalyst (number of linear
feet of 4-in. pipe required to give 6000 usable ppm NC>2 when diluted into a 12-SCFM
stream; see text).
X-6
------- |