EPA-650/2-73-029
September 1973 ENVIRONMENTAL PROTECTION TECHNOLOGY SERIES
-------
INTERACTIONS
OF STACK GAS SULFUR
AND NITROGEN OXIDES
ON DRY SORBENTS
by
J.W. Brown, D.W. Pershing, J.H. Wasser, andE.E. Berkau
Control Systems Laboratory
National Environmental Research Center
Research Triangle Park, N.C. 27711
Project No. 21ADG42
Program Element No. 1A2014
Prepared for
NATIONAL ENVIRONMENTAL RESEARCH CENTER
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
RESEARCH TRIANGLE PARK, NORTH CAROLINA 27711
September 1973
-------
This report has been reviewed by the Environmental Protection Agency and
approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the Agency, nor does
mention of trade names or commercial products constitute endorsement
or recommendation for use.
ii
-------
ABSTRACT
It was noted that the standard analytical system used for
the measurement of NO emissions gave incorrect NO concentrations
in the presence of SO.. The problem was traced to the dry sorbents
used to remove water vapor prior to the NO analysis. A brief
test series demonstrated that both Drierite and molecular sieve
sorbents can cause incorrect NO results if SO. is present. Further
testing revealed that the materials are capable of simultaneous
removal of both NO and SO. even in low concentrations. More work
is needed to define the actual fate of these species; however,
it appears that this might offer a possible NO /SO control
X X
technique since the data indicate that the sorbent effect is
thermally regenerable.
iii
-------
CONTENTS
Abstract ill
List of Figures v
List of Tables v
Introduction 1
Isothermal Testing with Known Gas Concentrations 3
Hot-Flow Testing Propane Combustion Doped with SO. ^
Hot-Flow Testing Residual-Oil-Fired Package Boiler 13
Conclusions/Recommendations 17
Bibliography 18
FIGURES
No., ^
1 Isothermal Test Facility 4
2 Hot-Flow Test Facility 8
3 Recorded Traces for Tests B-l, B-5, B-7, and B-8 H
4 Package Boiler Hot-Flow Test Facility 12
5 Emission Reduction Across a 65-Gram Molecular 16
Sieve Bed
TABLES
Page
No.
1 Series 1 Results
2 Series 2 Results
3 Series 3 Results Emissions After 5 Minutes
of Sampling
-------
INTRODUCTION
During recent experimental testing, the Combustion Research
Section (CRS) of EPA's Control Systems Laboratory observed that
sampling and analytical procedures previously utilized for measuring
NO emissions from natural gas and light oil combustion were not
x i
satisfactory for measuring nitrogen oxides (NO ) in sulfur oxide (SO )
j X X
laden gases. For example,.sulfur dioxide.(SO ) levels of 1600 ppm
(equivalent of 2.5 percent sulfur fuel) resulted in measured NO
X,
concentrations of less than half the actual level. The problem
was finally isolated to the dry sorbent system used to remove water
from the gas sample prior to the analytical Instrumentation.
2
It is well known (e.g., Sundaresan et al.) that certain
drying materials such as commercial zeolites and silica gel have
the ability to selectively adsorb 110 from nitric acid tail gases
X
where the NO concentration is in excess of 2000 ppm. Previous
x rr
CRS work had shown that NO adsorption and/or reaction did not
occur with the molecular sieve and Drierite drying agents utilized
in the standard CRS analytical train; however, all prior work
had been done with essentially s.ulfur free flue gases. The
purpose of the work reported herein was to briefly investigate the
observed SO /NO /drying agent interactions and to define what further
X X
work, if any, should be done in the area.
A three part approach was utilized. First, controlled mixtures
of ambient air, nitric oxide (NO), and S0_ were prepared, passed
through various types of drying systems, and then sampled to de-
termine exactly what interferences and/or interactions should be
expected. Next, the tests were repeated using actual flue gas
from a propane flame with arid without SO- present. Finally, the
work was extended to full scale testing on a package boiler burning
residual oil with 0.9 percent sulfur.
-------
ISOTHERMAL TESTING WITH KNOWN GAS CONCENTRATIONS
Under combustion conditions it is not possible to measure the
amount of NO in a given flue gas stream without first removing the
water of combustion. (Failure to do so will result in water vapor
condensation in the sample cell and/or instrument interference.)
Therefore, in the first test series ambient air (with appropriate
NO and SO. added) was utilized instead of actual flue gas so
that the sample could be run directly to the appropriate analyzer.
The purpose of this series was to define which elements if any of
the standard drying system needed further investigation.
The experimental facility used for this test series is
shown in Figure 1. Basically, it was designed to allow sampling
a gas flow of ambient air with and without NO and/or SO . Eoth
the NO and SO. came from laboratory cylinders through precalibrated
rotameters into the sampling duct ahead of the mixing section.
Sampling was accomplished via a standard quartz combustion probe.
From the probe the sample went either directly to the analyzer
(baseline tests) or to the system component being checked (e.g.,
the ice bath) then to the analyzer. All NO analysis in this series
was done with a chemiluminescent analyzer.
The results of these tests are shown in Table 1. In Tests
A-l through A-5 the stream being sampled contained only ambient
air and approximately 200 ppm NO. As the data indicate, none of
the common drying schemes had any significant effect on the measure-
ment. In Tests A-6 through A-10 sufficient SO. was added to the
air stream to give about 1600 ppm in the mixture. As the data
indicate both the Drierite and the molecular sieve led to radical
reductions in the measured NO level initially; however, with time
both appeared to "saturate" and the NO asymptotically approached
the correct value. These data suggest that some type of NO /SO
X X
-------
TO
AIR
SUPPLY
QUARTZ SAMPLE
PROBE
DIRECT TO
ANALYZER
TEST
SORBENT
TO
CHEMILUMINESCENT
NO ANALYZER
MIXING
ZONE
" NO
CYLINDER
S02
CYLINDER
Figure 1. Isothermal test facility.
-------
Table 1. SERIES 1 RESULTS
Chemilumlnescent
Test No. Test gas composition Sample conditions measured NO, ppm
A-l
A-2
A-3
A-4
A-5
A-6
Ambient air + :200 ppm NO Direct to analyzer
Drierite,b 15 gc
11 Molecular sieve, d "
CaCl2.2H20,e
" Ice bath
Ambient air + S200 ppm NO Direct to analyzer
194
204
189
195
193
ISA
A-7
A-8
+ =1600 ppm S02
Drierite,
15 g
Molecular sieve, "
CaCl2.2H20,
A-9
A-10 " Ice bath
A-ll Ambient air + =1600 ppm S0~ Direct to analyzer
A-12 " Drierite, 15 R
A-13 '' Molecular sieve, "
A-14
A-15
CaCl2.2H20,
Ice bath
(2 min.) 141,
(10 min.) 176
(2 min.) 67,
(10 min. ) 159
185
183
0.98
0.60
0.25
0.78
0.95 .
Total sample flow was 71 liters/hr (2.5 scfh). The flue gas sample
flowed through a particulate filter to prevent clogging instrument
sampling lines. (There was no evidence that this filter adversely
affected measured NOX readings.) A dry layer air filter pack #99/97
Microsorban made by Delbag Co. was used to remove particulates.
New "Drierite" - anhydrous CaSO_, W. A. Hammond Drierite Co.
^
'Fresh drying agent materials were weighed on a triple-beam balance
for each sample requiring an agent.
New molecular sieve - 0022-006-3A, #1 pellet, Guild Corp. (clay base).
"CaC1..2H 0 - Calcium chloride, reagent grade, Matheson, Coleman & Bell.
-------
interaction was occurring; however, unfortunately during this
series it was not possible to measure SO. to determine its fate.
It was also noted, that if the bulk of the flow (the ambient air)
was replaced by pure nitrogen no NO/SO interaction was observed.
This suggests that 0- and/or water vapor was also involved.
In Tests A-ll through A-15 the NO was turned off to investigate
possible negative interference effects by SO. on the analyzer
and the system components. As the data show SCL had essentially
no positive or negative effect on the chemiluminescent NO analyzer
since all readings were less than 1 ppm. The data do show, however,
that even at the ambient level of about 1 ppm NO, molecular sieve
and Drierite reduced the NO in the presence of S02-
-------
HOT-FLOW TESTING PROPANE
COMBUSTION DOPED WITH S0_
In the second test series flue gas from propane combustion
was artificially doped with S0~ on a controlled basis. The purpose
of this series was to investigate the NO /SO Interaction on as
X X
nearly a practical system as possible and still be able to control
the SO. concentration in the flue gas stream being sampled.
Figure 2 shows the test facility used for this portion of the work.
Basically, it was an upright multi-fuel combustor, with a 40.6 cm
(16-in.) ID refractory combustion chamber, and hot-air heat exchanger.
Combustion air was supplied by several air blowers in a variety
of combinations so that ambient air, preheated air, or flue gas
can be supplied to the primary, axial, or swirl streams. The
combustor, burner, and support facilities were identical to
3
those used in previous studies. In this series a six-hole
radial propane injector was utilized and the combustor was fired
at 75 million cal/hr (300,000 Btu/hr) and 5 percent excess air.
The NO in this test series was the result of the normal pro-
pane combustion (as compared to the first test series where
with no flame the NO level was simulated by injection of concen-
trated NO). The SO. was injected after the combustion zone to
prevent flame zone reactions and at a flow to give a concentra-
tion of 1600 ppra inside the combustor. As before NO was measured
with a chemiluminescent analyzer. Instrumentation was not available
for SO..
The results of this test series are presented in Table 2.
In tests B-l through B-4 the only major pollutant in the flue gas
being sampled was NO (since propane contains no sulfur and this
-------
AIR I * FUEL
/ / / /
/
1*1-*^ ROTAMETER
kP
, SO?
CYLINDER
HEAT
EXHAUST
STACK
EXCHANGER
QUARTZ SAMPLING
PROBE
ICE
BATH
TEST
SORBENT
TO
CHEMILUMINESCENT
NO ANALYZER
Figure 2. Hot-flow test facility.
-------
Table 2.
SERIES 2 RESULTS
Test
No.
B-l
B-2
B-3
B-4
B-5
Test gas Chemiluminescent
composition Sample conditions measured NO, ppm
Combustion gas Ice bath
" Drierite,b 15 g° + ice bath
Molecular sieve/ "
CaCl2.2H20,e
Combustion gas + Ice bath
94
96.5
98.5
94
94
B-6
1-7
B-8
=1600 ppm S02
Drierite, 15 g + ice bath
Molecular sieve
(2 min.) 41,
(30 min.) 89
(2 min.) 33,
(30 min.) 92
94
aThe total sample flow was 71 liters/hr (2.5 scfh) . In all cases '
the flue gas sample flowed through a particulate filter to prevent
clogging the instrument sampling lines. (There was no evidence
that this filter had any adverse effect on measured NO readings.)
A dry layer air filter pack #99/97 Microsorban made by Delbag Co.
was used to remove particulates.
New "Drierite" - anhydrous CaSO,, W. A. Hammond Drierite Co.
r*
Fresh drying agent materials were weighed out on a triple beam
balance for each sample where an agent was required.
dNew molecular sieve - //022--006-3A, #1 pellet, Guild Corp. (clay base)
eCaCl9.2H 0 - Calcium chloride, reagent grade, Matheson, Coleman & Bell,
£, £
-------
combustor does not produce significant carbon monoxide, unburned
hydrocarbons, or NO.). As in the first test series all of the
drying methods gave essentially the same NO concentration level;
there was no evidence of any unusual interactions. (The sample
could not be run directly to the analyzer without the ice bath
due to water condensation in the analyzer cell.)
During Tests B-5 through B-8, S0« was added to the flue gas
stream and as before' the use of both Drierite and molecular
sieve gave incorrect NO readings. Figure 3 shows the recorder
traces for the CaCl. and molecular sieve tests. As these data
indicate 15 grams of molecular sieve material required almost
30 minutes before equilibration occurred, while the CaCl .2H.O gave
the correct reading almost immediately.
10
-------
4 5 6
NO, millivolts (xlOO ppm RANGE)
10
Figure 3. Recorded traces for tests B-1. B-5, B-7, and B-8.
-------
EXHAUST
STACK
STEAM
to
AIR
COMBUSTION
CHAMBER
FUEL
WETS02
COLLECTOR
. STAINLESS STEEL
'SAMPLING PROBE
REFRIGERATION
COOLER
TEST
SORBENT
TO
CHEMILUMINESCENT
NO ANALYZER
f
TONDIR
NO
ANALYZER
Figure 4. Package boiler hot-flow test facility.
-------
HOT-FLOW TESTING « RESIDUAL-
OIL-FIRED PACKAGE BOILER
The final test series was conducted in a 60-hp residual-
oil-fired package boiler. The purpose of this series was to
confirm the previous results with flue gas from a typical
commercial system and to quantify the actual effects. Figure 4
shows the experimental setup used for this series. Basically,
it consisted of a 64-liter/hr (17-gph) Scotch-Marine boiler
operating at 25 percent excess air. NO was measured by both
chemiluminescent and NDIR analyzers. S0~ was determined by
4 *
wet chemical analysis (Shell method ). During each experiment
the flue gas sample from the boiler stack was passed through' a
cooler and the test bed before going to the appropriate analyzer.
The results of Tests C-l through C-5 are shown in Table 3.
In Test C-l no drying agent (other than the refrigeration cooler)
was used to remove water vapor prior to the emission analysis.
The 232 ppm NO is considered to be the baseline emission for this
unit. In the next four tests new and regenerated molecular
sieve and Drierite samples were used for final sample drying.
(New implies the material had just been received from the
manufacturer; regenerated implies the material had been used on
several prior occasions for water removal and then "regenerated"
by heating to drive off absorbed water.) The data indicate that:
1. With both molecular sieve and Drierite some reduction in
NO does occur across the drying material, confirming the Series
A and B results.
2. The process involves simultaneous reaction of both
NO and SO..
3. Regenerated molecular sieve gave the largest reductions
in both NO and SO .
13
-------
Table 3. SERIES 3 RESULTS ~ EMISSIONS AFTER
5 MINUTES OF SAMPLING3
Test
No.
C-l
C-2
C-3
C-4
C-5
NO by NO by S02 by
Drying chemiluminescent, NDIR, wet chemistry, d
agent" ppmc ppm ppm
None 232
New molecular 138
sieve
Regenerated molecular 99
sieve
New Drierite 122
Regenerated 183
Drierite
ND6 421
143
105 25
185
174
60-hp residual-oil-fired package boiler at 25 percent excess air.
In all cases refrigeration cooling was used to remove the initial
portion of the water.
As measured, dry.
d
Shell method (for details see Reference 1).
'since water vapor strongly interferes with NDIR analyzers it
was not possible to use these analyzers without moisture removal.
14
-------
It should be noted, however, that eventually the measured
emissions approached the baseline level as the drying material
"saturated". To quantify this phenomenon three tests were
run where flue gas was drawn through 65 grams of regenerated
molecular sieve and the emissions were monitored in turn for
NO by chemiluminescent analysis, SO. by NDIR, and NO by NDIR,
all as a function of cime. The reduced results are shown in
Figure 5. C /C is the ratio of the concentration of pollutant
t o
after a given number of liters of gas had passed through the
sieve bed to the correct (baseline) emission level. (Thus a
C /C of 1.0 means no reduction in pollutant concentration is
occurring across the bed and a C /C of 0.0 indicates complete
reduction: zero pollutant concentration after the bed.) The
data show that for the test case of 65 grams of molecular sieve
a 75 percent reduction in S02 and a 55 percent reduction in NO
occurred for the first 25 to 30 liters of gas. From these data
an SO. removal of 0.59 mg/g of sieve was estimated; the NO removal
was 0.077 mg/g of sieve. The relative volumes also appear to
be 3 parts of S02 for 1 part of NO.
To investigate the possible use of dry sorption as a
possible simultaneous NO /SO control technique the data were
X *V
used to estimate a system for a 1000 mw power plant burning 1
percent sulfur fuel. (It should be clearly noted that this
type calculation is crude at best because no attempt was made
to experimentally optimize conditions, sorbent, bed configuration
etc., and all work was done on a very small scale system, 65
grams of material.) The calculations indicated that 5.9 x 10 Kg/hr
(13 x 10 Ib/hr) of molecular sieve would be required to reduce
SO- concentration from 421 ppm to <25 ppm and simultaneously
reduce NO concentration from 232 ppm to 99 ppm.
15
-------
ONO BY CHEMILUMINESCENCE
DNO BY NDIR
ASO? BY NDIR
TOTAL GAS VOLUME, liters
Figure 5. Emission reduction across a 65-gram molecular sieve bed.
-------
CONCLUSIONS/RECOMMENDATIONS
1. Use of a solid sorbent drying material prior to NO
analysis can lead to erroneously measured NO concentrations in
SO. laden flue gases. Drierite and molecular sieve have both
been shown to be susceptible to this problem in varying degrees;
calcium chloride dehydrate appears to be a possible alternative
but needs a more complete analysis. The results are rudimentary
at best but even so they suggest the serious need for a detailed
chemical-analytical study to provide definitive guidelines for
NO sampling.
A
2. Both Drierite and molecular sieve have been shown
capable of simultaneous NO and SO. "removal" when both are present
even in low concentrations. Experimental work should be under-
taken to define the fate of the specie; i.e., is the NO being
retained on the surface of the sorbent in some form or is it
coming through the bed in a form not detected by the analyzer
(e.g., N02)?
3. Since this concept potentially offers simultaneous
NO and SO removal work should be undertaken to assess the
x x
feasibility of using it for flue gas treatment. Future work
should develop data on the effect of sorbent composition and
structure, NO and SO concentrations, and regeneration times so
X X
that a reasonable economic analysis can be conducted.
17
-------
BIBLIOGRAPHY
1. Martin, G. B., D. W. Pershing, and E. E. Berkau. Effects of
Fuel Additives on Air Pollutant Emissions from Distillate
Oil-Fired Furnaces. EFA. Research Triangle Park, N. C.
Office of Air Programs Publication No. AP-87. June 1971. 91.
2. Sundaresan, B. B., C. I. Harding, F. P. May, and E. R.
Hendrickson. Adsorption of Nitrogen Oxides from Waste Gas.
Environ. Sci. Techuol. 1:151-156, February 1967.
3. Pershing, D. V., J. W. Brown, and E. E. Berkau. Relationship
of Burner Design to the Control of NO Emissions Through
Combustion Modification. EPA. (Presented at Coal Combustion
Seminar. Research Triangle Park, N. C. June 19-20, 1973.) 53.
4. Shell Development Company, Analytical Department. Determina-
tion of Sulfur Dioxide and Sulfur Trioxide in Stack Gases.
Emeryville, Calif. 1959.
18
-------
INSTRUCTIONS FOR COMPLETING FORM NTIS-35 (10-70) (Bibliographic Data Sheet based on COSATI
Guidelines to Format Standards for Scientific and Technical Reports Prepared by or for* the Federal Government,
PB-180 600).
1. Report Number. Each individually bound report shall carry a unique alphanumeric designation selected by the performing
organization or provided by the sponsoring organization. Use uppercase letters and Arabic numerals only. Examples
FASKB-NS-87 nnd FAA-RD-68-09-
2. Leave blank.
3. Recipient's Accession Number. Reserved for use by each report recipient.
4. Title and Subtitle. 1 itle should indn.itc clearly and briefly the subject coverage of the report, and be displayed promi-
nently. Set subtitle, if used, in smallir type or otherwise subordinate it to main title. When a report is prepared in more
than one volume, repeat the primary title, .uit1 volume.- number and include subtitle for ihe specific volume.
5- Report Date. I .ich report .shall carry a dan indicating ai least month and year. Indicate the basis on which it was selected
(c.fj., date of issue, date of approval, date of preparation.
6- Performing Organization Code. Leave blank.
7. Author(>). Give name(s) in conventional order (e.g., John K. Doc, or j.Robert Dor). List author's affiliation if it differs
from (he performing organization.
8. Performing Organisation Report Number. Insert if performing organisation wishes to assign this number.
9. Performing Organisation Name and Address, dive name, street, city, state, and zip < <><.U-. List no more than two levels of
.in organisational hierarchy. Display (hi name of the organization exactly as it should appear in Government indexes such
as USGRDR-I.
10. Project 'Task, Work Unit Number. I si the project, task and work unit numbers under winch the report was prepared.
11. Contract'Grant Number. Insert conir.u t >>r grant number under whuh report was prepared.
12- Sponsoring Agency Nome and Address. In* lude /ipcodr.
13. Type of Report and Period Covered InJu .itc- im< rim, final, t tc., and, if applicable, date s covered.
14. Sponsoring Agency Code. Leave hl.ink.
15. Supplementary Notes. Kntcr information nut nu luilcd elsewhere but useful, such a-- Prepared in cooperation with . . .
translation of ... Presented at eonfcremc ol . . . 'I o h« published in ... Supersedes . . . Supplements . . .
16. Abstroct. Include a brief (.'00 wortU or less) faetual summary of the most significant information contained in the report.
If the report contains a significant bibliography or literature- survey, mention it here.
17. Key Words and Document Analysis, (a). Descriptors, .Select from the Thesaurus of (.nc.inec-ring and Scientific Terms the
proper authorised terms that idennf> the major concept of the re-search and are sufficiently specific and precise to be used
as index entries for cataloging.
(b). Identifiers and Open-Ended Terms. 1,'se identifiers for protect names, code names, equipment designators, etc. Use
open-ended terms written in descriptor form for those subjects for which no descriptor exists.
(c). COSATI Field/Group. Field and Croup assignments are to be taken from the 1965 COSATI 'subject Category List.
Since- the majority of documents are- multidisc iplmary in nature, the primary Field/Group assignment(s) will be the specific
discipline, area of human endeavor, or type of physical object. The applications) will be cross-referenced with secondary
Field/Group assignments that will follow the primary postmg(s).
18. Distribution Statement. Denote relcasability to the public or limitation for reasons other than security for example "Re-
lease unlimited". Cite- any availability to the public, with address and price.
19 & 20. Security Classification. Do not submit classified reports to the National Technical
21. Number of Pages. Insert the total number of pages, including this one and unnumbered pages, but excluding distribution
list, if any.
22. Price. Insert the price set by the National Technical Information Service or the Government Printing Office, if known.
FORM NTIS-33 IREV. 3-72) USCOMM-DC I4992-P72
-------
BIBLIOGRAPHIC DATA
SHEET
. Report No.
EPA-650/2-73-02 9
3. Recipient's Accession No.
5. Re-port Date
September 1973
4. 1 nlc and Subtitle
Interactions of Stack Gas Sulfur and Nitrogen Oxides
on Dry Sorbents
6.
7. Author(s)
J.W. Brown, D. W. Pershjng, J.H.Wasser. and E.E.Berkau
8. Performing Organization Rcpt.
No.
9. Performing Organization Name and Address
EPA, Office of Research and Development
NERC-RTP, Control Systems Laboratory
Research Triangle Park, North Carolina 27711
10. Project/Task/Work Unit No.
21ADG42
11. Coniract/Crant No.
In-House Report
12. Sponsoring Organisation Name ->nd Address
13. Type of Report & Period
Covered
Final
14.
15. Supplementary Notes
16. Attracts The repOrt describes a brief test series that demonstrates that both Drierite
and molecular sieve sorbents can cause incorrect NO results if SO2 is present. It
was noted that the standard analytical system used for measuring NO emissions gave
incorrect NO emissions in the presence of SO2. The problem was traced to the dry
sorbents used to remove water vapor prior to the NO analysis. Further testing
revealed that the materials can simultaneously remove both NO and SO2, even in low
concentrations. Although more work is needed to define the actual fate of these
species, it appears that this might offer a oossible NOx/SOx control technique
since the data indicate that the sorbent effect is thermally regenerable.
17. Ko\ Words and Document Analysis 17o IU s<
Air Pollution
Flue Gases
Sulfur Oxides
Nitrogen Oxides
Sorbents
Des ice ants
17b. Idenuf icrs/Opcn-l£nded Terms
Air Pollution Control
Stationary Sources
Dry Sorbents
Molecular Sieves
17e. COSATI Field/Group
Q7B
18. Availability Statement
Unlimited
19. Security Class (This
Report)
UNCLASSIFIED
20. Security (".lass (This
Page
UNC1.ASSIFILP
21. No. of Pages
24
22. Price
FORM NTIS-33 IREV. 3-72)
19
USCOMM-OC M932-P7Z
------- |