&EPA
United States
Environmental Protection
Agency
:PA-600/7-86-003a
February 1986
Research and
Development
ENVIRONMENTAL ASSESSMENT OF
AN ENHANCED OIL RECOVERY
STEAM GENERATOR EQUIPPED
WITH A LOW-NOx BURNER
Volume I. Technical Results
Prepared for
Office of Air Quality Planning and Standards
Prepared by
Air and Energy Engineering Research
Laboratory
Research Triangle Park NC 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-86-003a
February 1986
ENVIRONMENTAL ASSESSMENT
OF AN ENHANCED OIL RECOVERY
STEAM GENERATOR EQUIPPED
WITH A LOW-NOX BURNER
Volume (
Technical Results
By
C. Castaldini, L. R. Water-land, and H. I. Lips
Acurex Corporation
Energy & Environmental Division
555 Clyde Avenue
P.O. Box 7555
Mountain View, California 94039
EPA Contract No. 68-02-3188
EPA Project Officer: R. E. Hall
Air and Energy Engineering Research Laboratory
Research Triangle Park, North Carolina 27711
for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
ACKNOWLEDGEMENTS
The authors wish to extend their gratitude to Craig Jackson of the
Getty Oil Corporation and to Gary Sams of CE-Natco for their interest and
cooperation in this test program. Special recognition is also extended to
the Acurex field test team of M. Chips, P. Kaufmann, G. Chips, J. Oblack,
M. Murtiff, and G. Murphy, under the supervision of B. C. OaRos.
ii
-------
CONTENTS
Acknowledgements ii
Figures iv
Tables v
1 Introduction 1-1
2 Source Description 2-1
3 Emission Results 3-1
3.1 Performance/Emission Mapping Tests 3-1
3.2 Environmental Assessment Testing 3-11
3.2.1 Burner and Steamer Operation 3-14
3.2.2 Criteria Pollutant and Other Gas Phase
Species Emissions 3-14
3.2.3 Organic Species Emissions 3-19
4 Quality Assurance Activities 4-1
4.1 Accuracy Determinations 4-1
4.2 Precision Determinations 4-2
5 Summary 5-1
A Appendix A-l
A. Sampling and Analysis Methods A-l
A.I Continuous Monitoring System A-l
A.2 Particulate and Sulfur Oxide Emissions A-3
A.3 Organic Emissions A-6
A.4 Particle Size Distribution A-10
A.5 Ci to Cg Hydrocarbon Sampling and Analysis . . . A-12
A.6 M20 Emissions A-14
iii
-------
FIGURES
Figure Page
2-1 The MHI PM burner nozzle 2-2
i
2-2 Schematic of test steamer . 2-4
3-1 MOX emissions versus flue gas 63 3-4
3-2 CO emissions versus flue gas 02 3-8
3-3 Effect of OFA rate on NOX and CO emissions from the
low-NOx burner 3-9
3-4 Effect of F6R rate on NOX and CO emissions from the
low-NOx burner 3-10
3-5 NOX emissions versus CO for the MHI low-MOx 3-12
3-6 Emitted particle size distribution 3-18
3-7 M20 versus NOX emissions for external combustion
sources 3-20
A-l Continuous monitoring system A-2
A-2 Schematic of Method 5/8 sampling train A-5
A-3 Source assessment sampling system schematic A-7
A-4 Flue gas analysis protocol for SASS samples ....... A-8
A-5 Flue gas sample analysis protocol A-9
A-6 Organic analysis methodology . A-ll
A-7 Ci to Cg hydrocarbon sampling system A-13
IV
-------
TABLES
Table page
1-1 Completed Tests During the Current Program 1-4
2-1 Fuel Ultimate Analysis 2-5
2-2 Fuel Trace Element Concentrations 2-6
3-1 Flue Gas Emissions Summary: Conventional Burner .... 3-3
3-2 MHI Burner Performance Test Results 3-6
3-3 Steamer/Burner Operating Conditions: Comprehensive
Tests 3-15
3-4 Steamer Thermal Efficiency 3-15
3-5 Flue Gas Emissions 3-16
3-6 Total Organic Emissions Summary 3-22
3-7 Summary of Infrared Spectra of Total Sample Extracts . . 3-22
3-8 Compound Classes and Fragment Ions Searched for by
Direct Insertion Probe LRMS 3-24
3-9 SASS Particulate Extract LRMS Results 3-24
3-10 Compounds Sought in the GC/MS Analysis and Their
Detection Limits (ng/ul Injected) 3-25
3-11 Compounds Detected in the GC/MS Analyses 3-26
4-1 XAD-2 Resin Spike and Recovery Results 4-3
4-2 Duplicate GC/MS Analysis Results for the XAD-2 Extract. . 4-3
A-l Continuous Monitoring Equipment in the Mobile
Laboratory i A-4
A-2 Gas Chromatograph Specifications A-15
-------
SECTION 1
INTRODUCTION
This report describes and presents results for a set of environmental
assessment tests performed for the Environmental Protection Agency's
Air and Energy Engineering Research Laboratory (EPA/
AEERL) under the Combustion Modification Environmental Assessment
(CMEA) program, EPA Contract No. 68-02-3188. The CMEA started in 1976 with a
3-year study, the NOX Control Technology Environmental Assessment (NOX EA,
EPA Contract No. 68-02-2160), having the following four objectives:
« Identify potential multimedia environmental effects of stationary
combustion sources and combustion modification technology
9 Develop and document control application guidelines to minimize
these effects
o Identify stationary source and combustion modification R&D
priorities
Disseminate program results to intended users
During the first year of the NOX EA, data for the environmental
assessment were compiled and methodologies were developed. Furthermore,
priorities for the schedule and level of effort to be devoted to evaluating
the various source/fuel/control combinations were identified. This effort
revealed major data gaps, particularly for noncriteria pollutants (organic
emissions and trace elements) for virtually all combinations of stationary
1-1
-------
combustion sources and combustion modification techniques. Consequently, a
series of seven environmental field test programs was undertaken to fill
these data gaps. The results of these tests are documented in seven
individual reports (References 1-1 through 1-7) and in the NOX EA final
report summarizing the entire 3-year effort (Reference 1-8).
The current CMEA program has, as major objectives, the continuation of
multimedia environmental field tests initiated in the original MOX EA
program. These new tests, using standardized Level 1 sampling and analytical
procedures (Reference 1-9) are aimed at filling the remaining data gaps and
addressing the following priority needs:
« Advanced NOX controls
» Alternate fuels
» Secondary sources
« EPA program data needs
Residential oil combustion
Wood firing in residential, commercial, and industrial sources
High interest emissions determination (e.g., listed and
candidate hazardous air pollutant species)
« Nonsteady-state operation
The petroleum reserves which can be recovered through primary production
methods have been essentially exhausted in the oil fields in Kern County,
California. These fields still contain significant reserves, although the
remaining crude is too viscous to be produced by normal means. This crude is
currently being produced using what has been termed enhanced oil recovery
(EOR). In one popular process, near saturated (80 to 90 percent quality)
1-2
-------
steam is injected into a field. This steam heats the oil, thereby decreasing
its viscosity and allowing it to be pumped.
The steam for injection is raised by crude oil-fired steam generators
(often termed steamers) which have uncontrolled NOX emissions in the 300 ppm
range. Since Kern County is only in borderline attainment of the NOg ambient
air quality standard, EOR steamers have received close regulatory attention
with respect to reducing NOX emissions.
One approach to reducing NOX emissions from these steamers incorporates
a low-NOx emission burner design. One such burner was developed in Japan by
Mitsubishi Heavy Industries (MHI) and is currently marketed in the United
States by CE-Natco (a steamer manufacturer).
A steamer equipped with an MHI low-NOx burner was tested in the current
CMEA program. These tests, described in this report, were conducted to
quantify a broad emissions spectrum from the burner and to compare selected
species emissions to those from a steamer equipped with a "standard" burner.
Thus, a similar unit with a standard burner was also tested (in less depth,
however) in this program.
In addition to the tests described in this report, another EOR steamer,
this one equipped with a low-NOx burner developed under EPA contract by the
Energy & Environmental Research Corporation, was also tested. Results from
these tests are documented in Reference 1-10.
i
Table 1-1 lists all the tests performed in the CMEA program, outlining
the source tested, fuel used, combustion modifications implemented and the
level of sampling and analysis performed in each case. Results of these test
programs are discussed in separate reports.
1-3
-------
TABLE 1-1. COMPLETED TESTS DURING THE CURRENT PROGRAM3
Source
Spark-Ignited, natural-
gas-flred reciprocating
Internal combustion
engine
Compression ignition,
dlesel-flred,
reciprocating Internal
combustion engine
Low-N0x, residential,
condensl ng-hea t1 ng
system furnished by
(Carlsons Blueburner
Systems Ltd. of Canada
Description
Large bore, 6-cy Under,
opposed piston, 186-kH
(250 Bhp)/cyl, 900-rpm
Model 38TDSB-1/S
Large bore, 6-cyllnder
opposed piston, 261-kH
(3SO Bhp)/cy). 900-rpm
Model 36TDD8-1/8
Residential hot water
heater equipped with
M.A.N. low-NO. burner.
0.55 ml/s (0.5 gul/hr)
firing capacity, con-
densing flue gas
Test points
unit operation
Baseline (pre-HSPS)
~ Increased air-fuel
ratio aimed at
meeting proposed
NSPS of 700 ppm
corrected to 15
percent 02 and
standard atmospheric
conditions
Baseline (pre-HSPS)
~ Fuel Injection retard
aimed at meeting pro-
posed NSPS of 600 ppm
corrected to 15 per-
cent 02 and standard
atmospheric conditions
Low-NO. burner design
by M.A.N.
Sampling protocol
Engine exhaust:
- SASS
Method 5
Gas sample (Cj-C* HO
Continuous NO, NO,, CO,
C02, 02, CH4, TUHC
Fuel
Lube oil
Engine exhaust:
-- SASS
Method 8
Method 5
Gas sample (C^Cj HC)
Continuous NO. NOX. CO.
C02, 02, CH4, TUHC
Fuel
Lube oil
Furnace exhaust:
~ SASS
Method 8
Method 5
Gas sample (CI-CA HC)
Continuous HO, NOX. CO,
C02, 02, CH4. TUHC
Fuel
Waste water
Test collaborator
Fairbanks Morse
Division of Colt
Industries
Fairbanks Morse
Division of Colt
Industries
New test
Rocketdyne/EPA
low-NOx residential
forced warm air furnace
Residential warm air
furnace with modified
high-pressure burner and
firebox, 0.83 ml/s
(0.75 gal/hr) firing
capacl ty
Low-H0x burner design
and Integrated furnace
system
Furnace exhaust:
SASS
Method 8
Controlled condensation
Method 5
Gas sample (Ci-Cs HC)
Continuous NO. NOX, CO.
C02. 02, CH4, TUHC
Fuel
New test
(continued)
-------
TABLE 1-1. (continued)
i
in
Source
Pulverized coal-fired
utility boiler,
Cones vi lie station
Nova Scotia Technical
College Industrial
boiler
Adelphl University
industrial boiler
Description
400-MH tangentially
fired; new NSPS
design aimed at
meeting 301 ng/J
HOX limit
1.14 kg/s steam
(9.000 Ib/hr) f1 re tube
fired with a mixture
of coal-oll-water (COW)
1.89 kg/s steam
(15,000 Ib/hr)
hot water
f1 re tube fired with a
mixture of coal-oll-
water (COM)
Test points
unit operation
ESP Inlet and outlet,
one test
Baseline (COW)
Controlled SO?
emissions with
limestone addition
Baseline (COW)
Controlled SO?
emissions with
soda ash (NaoCCM
addition
Sampling protocol
ESP inlet and outlet
- SASS
Method 5
Controlled condensation
~ Gas sample (Cj-Cg HC)
Continuous HO, NO., CO,
C02, 02
Coal
bottom ash
ESP ash
Boiler outlet
~ SASS
Method 5
Method 8
Controlled condensation
~ Gas sample (C^Cs HC)
Continuous 02, C02,
CO, NO,
Fuel
Boiler outlet
SASS
Method 5
Method 8
Controlled condensation
-- Gas Sample (Cj-Cj HC)
Continuous 02, C02, NOX,
Test collaborator
Exxon Research and
Engineering (ERSE)
conducting cor-
rosion tests
Envlrocon per-
formed particulate
and sulfur
emission tests
Adelphl University
Fuel
CO
Pittsburgh Energy
Technology Center (PETC)
industrial boiler
3.03 kg/s steam Baseline test only
(24,000 Ib/hr) water tube with COM
fired with a mixture of
coal-oil (COM)
Boiler outlet
-- SASS
Method 5
Controlled condensation
-- Continuous 02, C02, MOX,
TUHC. CO
Fuel
PETC and General
Electric (GE)
(continued)
-------
TABLE 1-1. (continued)
Source
Description
Test points
unit operation
Sampling protocol
Test collaborator
TOSCO Refinery vertical
crude oil heater
2.54 Ml/day
(16.000 bbl/day) natural
draft process heater
burning oil/refinery gas
Baseline
Staged combustion
using air Injection
lances
Heater outlet
SASS
-- Method 5
Controlled condensation
-- Gas sample (CpCe HC)
-- Continuous 02, NO., CO.
KVB coordinating
the staged com-
bustion operation
and continuous
emission monitoring
CUjj, HC
-- NsO grab sample
Fuel oil
Refinery gas
Mohawk-Getty Oil
Industrial boiler
8.21 kg/s steam
(65.000 Ib/hr)
water tube burning
mixture of refinery gas
and residual oil
Baseline
Ammonia Injection
using the noncatalytlc
Thermal DeNOx
Process
Economizer outlet New test
SASS
- Method 5, 17
Controlled condensation
Gas Sample (Ci-C6 HC)
Ammonia emissions
HjO grab sample
-- Continuous 02. NO..
CO. C02
Fuels (refinery gas and
residual oil)
Industrial boiler
2.52 kg/s steam
(20.000 ]b/hr) watertube
burning wood waste
Baseline (dry wood)
Met (green) wood
Boiler outlet
SASS
Method 5
Controlled condensation
Gas sample (CrC6 HC)
~ Continuous 02, NO.. CO
Fuel
Flyash
North Carolina
Department of
Natural Resources,
EPA IERL-RTP
Industrial boiler
3.16 kg/s steam
(29.000 Ib/hr)
firetube with refractory
firebox burning wood waste
Baseline (dry wood)
Outlet of cyclone partlculate
collector
- SASS
Method 5
-- Controlled condensation
Gas sample (C^-Cg HC)
Continuous 02, NO., CO
Fuel
Bottom ash
North Carolina
Department of
Natural Resources,
EPA IERL-RTP
(continued)
-------
TABLE 1-1. (continued)
Source
Enhanced oil recovery
steam generator
Description
IS MM (SO million Btu/hr)
steam, generator burning
crude oil equipped with
HHI low-HOx burner
Test points
unit operation
Performance mapping
Low NOX operation
Sampling protocol
Steamer outlet:
SASS
Method 5
Method 6
Test callaborator
Getty Oil Company.
CE-Natco
Andersen Impactors
Gas, sample (Cj - C& NO
Continuous 02. NOX, CO,
C02
1120 grab sample
Fuel
Pittsburgh Energy
Technology Center
(PETC) industrial
boiler
Spark-ignited, natural
gas-fuel reciprocating
internal combustion
engine nonselective
NOX reduction catalyst
3.03 kg/s steam
(24,000 Ib/hr) watertube
fired with a coal-water
slurry (CHS)
610 kH (818 hp) Haukesha
rich-burn engine equipped
with DuPont NSCft system
Baseline test only
with CHS
Low NOX (with catalyst)
15-day emissions
monitoring
Boiler outlet:
SASS
Method 5
Method 8
-- Gas sample (Cj - Cg HC)
Continuous 02, NOX. CO,
CO?. TUHC
NjO grab sample
Fuel
Bottom ash
Collector hopper ash
Catalyst inlet and outlet
SASS
HH3
IICN
IfyO grsb sample
PETC and General
Electric
Southern California
Gas Company
Continuous Op. CO?. Hox
TUHC
Lube oil
Industrial boiler
180 kg/hr steam
(400 Ib/hr) stoker fired
with a mixture of coal
and waste plastic
beverage containers
Case line (coal)
Coal and plastic
waste
Boiler outlet
SASS
VOST
Method 5
Method 8
~ HC1
Continuous 02, NOX,
CO?, TUHC
NjO grab sample
Fuel
Bottom ash
Cyclone ash
Vermont Agency of
Environmental
Conservation
CO,
ICOntinUGuj
-------
TABLE 1-1. (concluded)
Source
Industrial boiler
Description
7.6 kg/s steam
(60.000 Ib/hr) water tube
retrofit for
coal -water-slurry (CMS)
firing
Test points
unit operation
-- Baseline test with
CHS
~ 30-day emissions
monitoring
Sampling protocol Test collaborator
Boiler outlet EPRI, OuPont
SASS
VOST
~ Method 5
-- Method 8
Gas sample (Ci-C6 HO
~ N20 ?rab sample
~ Continuous NOX, CO. C02.
Fuel
02. TUHC, S02
Enhanced oil recovery
steam generator
I
CD
IS m (50 million Btu/hr)
steam generator burning
crude oil, equipped with
the EPA/EER low NOX
burner
Low NOX (with burner)
30-day emission
monitoring
Steamer outlet
~ SASS
- VOST
Method 5
Method 8
Controlled condensation
Andersen Impactors
Grab sample (CrC6 HO
N20 grab sample
~ Continuous NOX. CO, 0)2,
Fuel
Chevron U.S.A.,
EERC
02, S02
Spark-Ignited, natural
gas-fired reciprocating
Internal combustion
engine -- selective
NOX reduction catalyst
1490 k« (2000 hp)
Ingersoll-Rand lean burn
engine equipped with
Englehard SCR system
-- Low NOX (with catalyst)
-- 15-day emissions
monitoring
Catalyst inlet and outlet
SASS
~ VOST
HCN
-- N20 grab sample
Continuous 0?, CO?, CO,
HO. NOX. NOX + NHi
Lube oil
Southern California
Gas Company
aAcronymns used In the table: EERC, The Energy and Environmental Research Corporation; EPA IERL-RTP, The Environmental Protection
Agency's Industrial Environmental Research Laboratory-Research Triangle Park; EPRI, The Electric Power Research Institute;
HC, hydrocarbons; NSCR, nonselectfve catalytic reduction; NSPS, new source performance standard; SASS, source assessment sampling
system; SCR, selective catalytic reduction; TUHC, total unburned hydrocarbon; VOST, volatile organic sampling train
-------
REFERENCES FOR SECTION 1
1-1. Larkin, R. and E. B. Higginbotham, "Combustion Modification Controls
for Stationary Gas Turbines: Volume II. Utility Unit Field Test,"
EPA-600/7-81-122b, NTIS PB82-226473, July 1981.
1-2. Higginbotham, E. B., "Combustion Modification Controls for Residential
and Commercial Heating Systems: Volume II. Oil-fired Residential
Furnace Field Test," EPA-600/7-81-123b, NTIS PB82-231175, July 1981.
1-3. Higginbotham, E. B. and P. M. Goldberg, "Combustion Modification NOX
Controls for Utility Boilers: Volume I. Tangential Coal-fired Unit
Field Test," EPA-600/7-81-124a, NTIS PB82-227265, July 1981.
1-4. Sawyer, J. W. and E. B. Higginbotham, "Combustion Modification NOX
Controls for Utility Boilers: Volume II. Pulverized-coal Wall-fired
Unit Field Test," EPA-600/7-81-124b, NTIS PB82-227273, July 1981.
1-5. Sawyer, J. W. and E. B. Higginbotham, "Combustion Modification NOX
Controls for Utility Boilers: Volume III. Residual-oil Wall-fired
Unit Field Test," EPA-600/7-81-124c, NTIS PB82-227281, July 1981.
1-6. Goldberg, P. M. and E. B. Higginbotham, "Industrial Boiler Combustion
Modification NOX Controls: Volume II. Stoker Coal-fired Boiler Field
Test -- Site A/ EPA-600/7-81-126b, NTIS PB82-231085, July 1981.
1-7. Lips, H. I. and E. B. Higginbotham, "Industrial Boiler Combustion
Modification NOX Control: Volume III. Stoker Coal-fired Boiler Field
Test -- Site B, EPA-600/7-81-126c, NTIS PB82-231095, July 1981.
1-8. Waterland, L. R., et al., "Environmental Assessment of Stationary
Source NOX Control Technologies -- Final Report," EPA-600/7-82-034,
NTIS PB82-249350, May 1982.
1-9. Lentzen, D. E., etal., "IERL-RTP Procedures Manual: Level 1
Environmental Assessment (Second Edition)," EPA-600/7-78-201,
NTIS PB293795, October 1978.
1-10. Castaldini, C., et al., ("Environmental Assessment of an Enhanced Oil
Recovery Steam Generator Equipped with the EPA Low-N0x Burner," Acurex
Draft Report TR-85-174/EED, January 1985.
1-9
-------
SECTION 2
SOURCE DESCRIPTION
Tests were performed on two CE-Natco model STOP steam generators rated
at 50 minion Btu/hr heat output. One unit was equipped with a standard
North American burner; the other had been retrofitted with the PM low-NOx
burner manufactured by Mitsubishi Heavy Industries (MHI). The primary
objective of the tests was to measure the NOX reduction performance of this
burner as a function of its operational parameter settings (when compared to
a standard burner) and to obtain data on emissions of noncriteria pollutant
categories and species at a nominal low-NOx setting.
Figure 2-1 illustrates the physical design of the MHI PM burner. As
shown, the rectangular burner throat is divided into five nozzles. Typically
about 30 percent of the total combustion air is delivered through the central
primary air nozzle. This air is mixed with a centralized oil spray
comprising approximately half the total fuel fired, forming an oxygen
deficient diffusion flame. A premixed flame is obtained by mixing the
remaining fuel with about 60 percent of the total air, evenly delivered
through each of the upper and lower nozzles. This mixing takes place in a
zone offset from the burner which delays ignition until the fuel and air have
mixed.
The remaining combustion air (about 10 percent) is delivered through an
overfire air (OFA) injection system which injects this air approximately
2-1
-------
Premix
Flame
Recir
Gas Blanket
Diffusion
Flame
Premix
Air Nozzle
Flue Gas
Recycle
Nozzle
Diffusion.
Nozzle
Flue Gas
Recycle
Nozzle
Premix
Air Nozzle
Figure 2-1. The MHI PM burner nozzle.
2-2
-------
halfway down the length of the cylindrical furnace through three sets of
three ports equally spaced around the furnace circumference. This OFA is
designed to ensure that sufficient excess air and mixing are achieved before
combustion gas leaves the furnace.
Recirculated flue gas is delivered to the nozzles separating the central
(diffusion) and outer (premix) air nozzles. This gas is used to shape the
diffusion flame and to maintain separation between the diffusion and premixed
flames. Typically about 15 percent of the total combustion product gas is
recirculated.
Figure 2-2 shows a sketch of the steamer retrofitted with the burner
system. The additional flue gas recirculation (FGR) and OFA systems along
with the burner are shown.
In the test program performed, one day of flue gas emission testing was
performed on the steamer equipped with the conventional burner. In these
tests, flue gas MOX emissions were measured at two steamer loads while
varying the excess air fuel. The steamer equipped with the low-NOx burner
was then subjected to two days of performance/emissions mapping tests in
which flue gas composition (NOX, CO, C02, $2 and smoke) was characterized
while varying burner operation at full steamer load. In these tests, the
following were varied: the FGR rate; the relative distribution of combustion
air among the premixed flame noz'zles, the diffusion flame nozzles, and the
OFA ports; and the overall excess air level. Finally, comprehensive
emissions testing (flue gas organics, particulate load, particle-size
distribution, and S02 and $03 emissions) was performed on the low-NOx
burner-equipped steamer with the burner set at a nominal low-NOx condition.
2-3
-------
rss
£>
Ovc'rfire air
duct
H>M burner
Figure 2-2. Schematic of test steamer.
-------
The fuel fired in both steamers for all tests was local Kern County
crude. The fuel ultimate analysis is given in Table 2-1.
The concentrations of 72 trace elements in the fuel were also obtained
using spark source mass spectrometry (SSMS) supplemented by atomic
absorption spectrometry. Results of these analyses are given in Table 2-2.
TABLE 2-1. FUEL ULTIMATE ANALYSIS
Component (wt percent)
Carbon
Hydrogen
Sulfur
Nitrogen
Oxygen
Ash
86.88
10.84
1.06
0.76
0.43
0.03
Higher heating, MJ/kg 43.2
value (Btu/lb) (18,560)
API gravity 13.3
2-5
-------
TABLE 2-2. FUEL TRACE ELEMENT CONCENTRATIONS
Concentration Concentration Concentration
Element (ug/g) Element (ug/g) Element (ug/g)
Aluminum
Antimony
Arsenic
Barium
Beryl 1 i urn
Bismuth
Boron
Bromine
Cadmium
Calcium
Cerium
Cesium
Chlorine
Chromium
Cobalt
Copper
Dysprosium
Erbi urn
Europium
Fluorine
Gadolinium
Gallium
Germanium
Gold
Hafnium
4.0
0.07
0.4
0.5
<0.01
a
0.2
0.3
48
0.2
4
0.8
2
5
--
27
0.5
0.2
~
^^
Holmium
Iodine
Iridium
Iron
Lanthanum
Lead
Lithium
Lutetium
Magnesium
Manganese
Mercury
Molybdenum
Neodymi urn
Nickel
Niobium
Osmium
Palladium
Phosphorus
Platinum
Potassium
Praseodymium
Rhenium
Rhodium
Rubidium
Ruthenium
__
95
0.4
0.3
0.1
14
0.4
<0.01
1
<0.2
90
<0.03
0.8
4
0.06
0.02
^^
Samarium
Scandium
Selenium
Silicon
Silver
Sodium
Strontium
Tantalum
Tellurium
Terbium
Thallium
Thorium
Thulium
Tin
Titanium
Tungsten
Uranium
Vanadium
Ytterbium
Yttrium
Zinc
Zirconium
_..
0.08
0.2
34
0.03
22
0.5
2
0.06
0.8
0.07
3
~
80
0.6
1
0.3
a denotes present at less than the detection limit of 0.02 ug/g.
2-6
-------
SECTION 3
EMISSION RESULTS
As noted in Section 2, the objectives of these tests were to evaluate
the NOX emission reduction performance of the MHI PM low NOX burner system
retrofitted to an enhanced oil recovery steam generator and to quantitate
emissions of noncriteria pollutant species from the retrofit steamer. To
satisfy these objectives a' brief series of flue gas emission measurement
tests was performed on an identical unit equipped with a conventional burner.
A relatively comprehensive series of performance/emission mapping tests was
performed next on the unit equipped with the low NOX burner. Finally, a set
of comprehensive environmental assessment flue gas characterization tests was
performed on the low NOX burner equipped-steamer with burner operation set at
a nominal low NOX setting.
Section 3.1 summarizes results of the tests of the conventional
burner-equipped steamer and the performance/emission mapping tests on the low
NOX burner-equipped steamer. Results of the comprehensive emission testing
of the low NOX burner-equipped steamer are discussed in Section 3.2.
3.1 PERFORMANCE/EMISSION MAPPING TESTS
Flue gas emissions of NOX, CO, 0)3, 02, and stack gas smoke readings
were measured on the steamer equipped with a conventional North American
burner at two loads (full load and about 75 percent of rated capacity) and
several excess air settings. These measurements were performed at the stack
3-1
-------
using a continuous flue gas monitoring system as described in Appendix A. To
supplement these measurements, Getty Oil Company personnel performed
complementary monitoring of the combustion"gas at the steamer furnace exit.
i
Results of these tests are summarized in Table 3-1. The data in the
table clearly suggest that there was some combustion gas dilution through air
inleakage between the furnace outlet sampling location and the stack sampling
location. Stack 02 levels are consistently higher and COg consistently lower
than corresponding furnace outlet levels. CO levels (corrected to 3 percent
02) are comparable at the two locations at full load. However, NOX levels at
full load were generally about 40 to 50 ppm higher at the stack location than
at the furnace outlet. At 75 percent load, NOX levels at both locations were
comparable; however, CO levels were apparently increased. Reasons for both
these apparent increases (if they were indeed real) can only be speculated.
The stack location NOX emissions data are plotted in Figure 3-1 as NOX
versus stack gas 03. The figure shows a steady decrease in MOX emissions as
excess air is reduced until flue gas 03 falls below about 3 percent. Below
this 03 level, the rate of NOX emissions increases. -However, referring to
Table 3-1, as flue gas 02 is decreased below 3 percent, the smoke number
increases to unacceptable levels. For practical operation, then, the
conventional burner appears capable of attaining full load NOX emissions of
about 300 ppm (3 percent 02) with flue gas 02 about 3.7 percent at acceptable
CO emissions and smoke number. At 75 percent load, NOX emissions are reduced
to about 250 ppm (3 percent 02) at flue gas 02 of 4.0 percent and acceptable
CO and smoke number.
Following the conventional burner-equipped unit testing, a relatively
comprehensive series of performance emission mapping tests was performed on
3-2
-------
TABLE 3-1. FLUE GAS EMISSIONS SUMMARY: CONVENTIONAL BURNER
co
CO
Test no.
Full load
1
2
3
4
5
6
751 load
7
a
9
Fuel
(1/s)
0.39
0.39
0.39
0.39
0.39
0.39
0.29
0.29
0.29
flow
(BPO)
210
210
210
210
210
210
. _
16U
160
160
Heat input
(MW)
16.3
16.3
16.3
16.3
16.3
16.5
12.4
12.4
12.4
(million
Btu/hr)
55.5
55.5
55.5
55.5
55.5
56.3
42.3
42.3
42.3
02
(«)
3.7
5.1
6.6
3.7
2.9
2.3
4.2
6.4
6.1
C02
U)
12.8
12.0
10.8
13.1
12.2
14.4
13.3
11.7
12.7
Stack*
CO
(ppm)c
23
34
31
42
54
46
78
96
133
NOX
(ppm)C
302
340
365
305
278
197
246
296
290
Smoke
__
3.0
4.0
4.0
8.0
8.5
4.0
2.5
3.0
02
(2)
3.0
4.4
6.2
2.8
1.9
1.3
2.8
5.3
4.0
Furnace outletb
CO?
W
14.1
12.9
11.5
14.1
14.8
15.3
14.1
12.1
13.2
CO
(ppm)c
42
48
49
42
46
54
40
46
42
NOX
(ppm)c
263
288
305
256
227
206
254
300
277
S02
(ppm)c
580
580
584
593
598
603
584
603
578
^Emission measurements by Acurex.
"Emission measurements by Getty Oil Company.
C0ry at 3 percent 02.
-------
o'aso
H»
| 300
CD
Q_
« 250
E
Q.
Q.
200
150
0* 100
O Full Load MHI Burner Tests
Full Load Conventional Burner
75% Load Conventional Burner
MHI FD Fan Limit
50
4 5
02 (Percent Dry)
Figure 3-1. NOX emissions versus flue gas
3-4
-------
the steamer equipped with the low NOX burner. Recall from Section 2 that
this burner directs combustion air to a central diffusion flame nozzle,
outboard premixed flame nozzles, and overfire air injection ports. Thus, the
distribution of combustion air among these streams, as well as the total air
fired, are adjustable burner operating parameters. In addition, the rate at
which flue gas is recirculated to the burner (to separate the diffusion and
premixed air flames) is a further adjustable parameter. All of these were
varied in the testing performed.
Results of the performance/emission mapping tests are summarized in
Table 3-2. Again, flue gas composition both at the steamer stack and at the
furnace outlet (measurements performed by Getty Oil Company personnel) are
shown. For these tests there was general agreement between the stack 03 and
C02 readings and those at the furnace outlet. This suggests that negligible
air inleakage occurred between these locations for the low NOX
burner-equipped steamer, in contrast to the apparent case for the steamer
with the conventional burner. Corresponding NOX and CO levels were, in
general, similarly comparable (although for a few test points stack CO levels
were significantly higher than furnace outlet levels).
The data in Table 3-2 show that NOX emissions from the unit varied from
95 to 180 ppm (corrected to 3 percent 03) with changes in the parameters
investigated. Certain conditions resulted in NOX emissions at the stack
i
below 100 ppm (3 percent 02, dry), but these were, in general, accompanied by
high CO emissions and high smoke spot. Conditions which resulted in NOX in
the 110 ppm range with moderate CO are also noted.
The variation in NOX emissions with overall excess air (stack gas 02)
for this unit was shown in Figure 3-1. The scatter in the figure results
3-5
-------
TABLE 3-2. MHI BURNER PERFORMANCE TEST RESULTS
co
A1r distribution
Test
no.
1
2
3
4
5
6
7
e
9
10
11
12
13
14
15
16
17
18
19
20
21
(SASS)
FGR
rate
(%)
8.9
9.4
13.4
9.9
9.9
9.4
8.8
9,1
9.8
9.3
8.8
8.2
8.4
8.8
6.6
2.6
8.4
9.1
9.8
9.8
9.5
OFA
(t)
19
18
19
19
13
15
15
11
9
B
8
8
8
8
8
8
7
6
7
3
10
Premix*
flame air
(*)
48
52
52
53
52
51
51
56
57
57
57
SB
55
SS
55
55
59
58
58
62
54
Diffusion*
flame air
(I)
33
30
29
29
35
34
34
33
34
35
35
34
37
37
37
37
34
36
35
35
36
Fuel
(1/s)
0.383
0.385
0.381
0.390
0.379
0.377
0.388
0.386
0.366
0.388
0.3B6
0.396
0.3B8
0.390
0.390
0.390
0.405
0.377
0.386
0.388
0.386
rate
(BPD)
208
209
207
212
206
205
211
210
210
211
210
215
211
212
212
212
220
205
210
211
210
Heat input
(MM)
16.1
16.2
16.0
16.4
16.0
15.9
16.3
16.3
16.3
16.3
16.3
16.7
16.3
16.4
16.4
)6.4
17.0
15.9
16.3
16.3
16.3
(Million
Btu/hr)
55.0
55.2
54.7
56.0
54.5
54.2
55.8
55.5
55.5
55.8
55.5
56.8
55.8
56.0
56.0
56.0
58.2
54.2
55.5
55.8
55.5
°2
III
3.5
2.7
2.6
2.5
2.6
3.4
3.8
3.2
2.2
3.1
4.1
4.2
4.2
3.4
4.2
4.2
4.2
3.6
2.B
2.8
3.0
ffi
13.4
13.9
13.9
14.0
14.0
13.9
13.3
13.2
14.6
13.3
12.5
12.5
12.4
13.0
12.3
12.3
12.6
12.9
13.5
13.6
13.3
Stackb
CO
(ppm)d
99
266
215
236
269
60
60
51
79
141
70
64
51
85
66
54
60
62
80
64
93
HO,
(ppm)d
119
102
99
95
97
119
140
145
110
111
145
180
126
111
131
152
143
116
106
133
106
TUHC
(ppm)d
3.2
4.4
3.2
5.3
3.9
2.2
1.4
1.)
1.0
4.5
8.5
8.6
1.8
1.1
1.4
1.1
1.4
0.5
0
0
0
Smoke
>10
10
9.5
10
8
6
3.5
3.5
8
10
6
3.5
4
8
6
2.5
4
6
8
6
8
0,
(if
4.3
3.3
3.1
3.1
3.0
4.0
4.4
3.3
2.3
2.2
3.2
4.2
4.6
3.6
4.5
4.6
4.1
3.3
2.3
2.5
2.5
ia^jra-j;
Furnace outlet'
CO
(ppm)"
70
BO
80
73
76
59
44
45
58
104
54
46
55
67
58
53
SO
64
87
55
66
II ! 1 1
CO?
m
12.9
13.9
13.9
14.0
13.9
13.0
12.9
13.9
14.5
14.7
13.8
12.9
12.5
13.5
12.7
12.6
13.1
13.8
14.5
14.5
14.5
HOX
(ppm)"
124
113
108
109
105
126
140
144
112
102
126
174
125
114
131
149
144
113
98
126
108
S°2 H
(ppm)"
507
594
582
592
576
573
581
566
585
616
595
570
587
573
572
582
556
55S
574
583
586
"Prenix and diffusion nozzle combustion air flows were not measured. Values shown here were estimated hased on blower discharge
pressure and static pressure readings in the windbox for diffusion and premlx zones.
Emission measurements by Acurex.
^Emission measurements by Getty Oil Company.
"Dry at 3 percent Og.
-------
from changes in MOX emissions with the split in air flowrates among the
diffusion and premixed flames and the OFA ports, and FGR rates at constant
overall excess air. In general, though, NOX emissions with the low-NOx
burner at full load were roughly half those of the conventional burner at a
given flue gas 02.
Figure 3-2 shows steamer stack gas CO emissions versus stack gas 02 for
both burners at full load. Again the scatter in the data for the low-NOx
burner results from variations in air distribution and FGR rate at constant
stack 02. The data in Figure 3-2 show that CO emissions from the low-NOx
burner increased steeply at flue gas 02 below 2.5 to 3.0 percent. This
contrasts with conventional burner behavior where CO emissions were still low
at flue gas 02 down to 2.5 percent. The higher CO levels from the low-NOx
burner, which were accompanied by high smoke spot (see Table 3-2) are
attributed to flame impingement which was observed at the 4 and 8 o'clock
positions at virtually all burner settings. Higher CO levels are attributed
to increased flame impingement and excessively low diffusion zone
stochiometries during low 02 and high OFA tests.
The effect of OFA flowrate on both CO and NOX levels for the low-NOx
burner is illustrated in Figure 3-3. CO levels decrease sharply at OFA rates
below 10 percent. At 3 percent OFA, CO levels are nearly those of the
conventional burner (see Figure 3-2). NOX emissions at minimum OFA, however,
are not significantly higher than those at high OFA rates.
The effect of FGR on NOX and CO emissions from the low-NOx burner is
shown in Figure 3-4. FGR had a greater effect at a higher 02 and lower OFA
levels (4 percent and 8 percent OFA) than it did at lower 02 and higher OFA
levels (02 of 2.6 percent and 19 percent OFA). CO responded in an opposite
3-7
-------
CO
00
280
'cvi
O 240
c
0)
o
0
Q-
(0
Q
200
160
120
I 8°
8 40
0
- o
°o
o Low-NOx Burner Full Load
Conventional Burner Full Load
4 5
02 (Percent Dry)
Figure 3-2. CO emissions versus flue gas 62-
-------
co
10
C\J
O
*-
c
CD
^
-------
CP
H-*
O
CO (ppm Dry, at 3 Percent 02)
_ 190
'cvi
O
tr
CO
o
fc 150
Q.
CO
tc
100
Q
E
QL
a.
X
O
Z
50
240
200
160
120
80
40
0 0
OFA = 8 Percent
Diffa =37 Percent
PrerrP= 55 Percent
02 = 4.2 Percent
OFA = 19 Percent
Diffa = 29 Percent
Prenrr= 52 Percent
02 = 2.6 Percent
FGR Rate (Percent)
*Diff: Diffusion flame air
Prem: Premixed flame air
Figure 3-4. Effect of FGR rate on NOX and CO
emissions from the low NOX burner.
8 10 12 14 16 18
-------
manner. This can be explained in part by the greater mixing occurring at
higher burner stoichiometries combined with lower FGR rates. This mixing
tended to partly cancel the low-NOx properties of the split flame.
Conversely, the higher FGR rates combined with lower burner stoichiometry,
while keeping the flames separate, will tend to cause greater impingement of
the premix flame, which increased the CO levels.
Figure 3-5 shows a crossplot of the NOX/CO emission data for the low-NOx
burner. This figure shows that, as the burner is adjusted to give NOX
emissions below about 110 ppm (3 percent 03), CO emissions (and smoke, see
Table 3-2) increase significantly. Thus, for this burner/steamer
combination, minimum NOX emissions at acceptable operation appear to be
110 ppm.
3.2 ENVIRONMENTAL ASSESSMENT TESTING
Following the performance/emission mapping tests discussed in
Section 3.2, a set of burner operating conditions was selected for
comprehensive emissions testing. The sampling protocol for these
comprehensive tests included:
o Continuous monitoring for NOX, 03, CO, 0)3, and total unburned
hydrocarbons (TUHC)
o Source assessment sampling system (SASS) for particulate size
fractionation, and organic emissions
o EPA Method 5/8 for particulate mass emissions, and S02 and $03
emissions
o Andersen impactor train sampling for particle size distribution
determination
3-11
-------
₯
190
170
CM
o
fc 150
£ 130
25
Q 110
E
Q.
Q.
- 90
O
z
70
50
O
o
0 40 80 120 160 200 240
CO (ppm Dry, at 3 Percent 02)
Figure 3-5. NOX emissions versus CO for the MHI low NOX burner.
280 320
-------
o Grab sample for onsite analysis of Cj to CQ hydrocarbons by gas
chromatography (GC)
o Grab sample for laboratory analysis of ^0
All flue gas sampling for these tests was performed at the steamer
stack. In addition, as for other testing performed in this program, Getty
Oil Company personnel performed continuous flue gas monitoring at the steamer
furnace outlet.
The analysis protocol for SASS samples included:
«» Analyzing SASS train samples for total organic content in two
boiling point ranges: 100° to 300°C by total chromatographable
organics (TCO) analysis and greater than 300°C by gravimetry (GRAY)
o Analyzing the SASS train sorbent module and particulate extracts for
the 58 semivolatile organic priority pollutant species including
many of the PAH compounds
o Performing infrared (IR) spectrometry analysis of the GRAV residue
of organic sample extracts
o Performing direct insertion probe low resolution mass spectrometry
(LRMS) analysis of selected sample extracts
All aspects of the sampling and analysis protocols conformed to a
modified EPA Level 1 protocol (Reference 3-1). Details of the procedures
used are discussed in Appendix A'.
Bioassay testing of SASS samples, a normal part of the comprehensive
testing performed in this project, was not done in these tests due to the
limited amount of particulate sample obtained and the very low organic
content of the sorbent module extract.
3-13
-------
Results of these comprehensive tests are discussed in the following
subsections. Section 3.2.1 further details the steamer operating condition
during the tests performed; Section 3.2.2 presents the criteria pollutant and
other gas phase species emission results; and Section 3.3.3 summarizes
organic category and species emission results.
3.2.1 Burner and Steamer Operation
The burner operating conditions selected for comprehensive testing
matched those noted for Test 21 in Table 3-2. This operating point was
selected since it represented about a minimum NOX condition with CO emissions
below 100 ppm (see Figure 3-5). The specific steamer and burner operating
conditions for these tests are summarized in Table 3-3.
The steamer efficiency noted in Table 3-3 was calculated based on the
ASME heat loss method (ASME PTC 4.1). The relative contributions to the
calculations are summarized in Table 3-4. As shown in this table, most of
the efficiency loss was through dry gas loss and moisture loss from the fuel
hydrogen. The overall efficiency noted (82.8 percent) compares favorably to
the efficiency of conventional burner-equipped units.
3.2.2 Criteria Pollutant and Other Gas Phase Species Emissions
Table 3-5 summarizes the gaseous and particulate emission levels
measured during the comprehensive tests. Continuous emission monitor
measurements from both the stack location and the furnace outlet location
(obtained by Getty Oil personnel) are noted in the table. As shown in the
table, steamer stack NOX emissions were just below 110 ppm (3 percent 03)
with CO emissions of 93 ppm (3 percent 02), negligible TUHC emissions, and a
smoke reading of 8.
3-14
-------
TABLE 3-3. STEAMER/BURNER OPERATING CONDITIONS:
COMPREHENSIVE TESTS
Fuel flow, 1/s (BPD) 0.386 (210)
Heat input, MW (million Btu/hr) 16.3 (55.5)
Feedwater flow, 1/s (BPD) 6.72 (3,650)
Steam pressure, MPa (psig) 4.55 (660)
Air flows (percent)
Diffusion 36
Premix 54
Overfire 10
FGR rate (percent) 9.5
Steamer efficiency (percent) 82.8
TABLE 3-4. STEAMER THERMAL EFFICIENCY
Heat loss efficiency (percent)
Dry gas loss 6.8
Loss due to fuel moisture
Loss due to water from the 6.3
combustion of fuel hydrogen
Loss due to combustibles in 0.6
the flyash
Radiation loss 2.0
Unmeasured loss 1.5
Total loss 17.2
Efficiency (percent) 82.8
3-15
-------
TABLE 3-5. FLUE GAS EMISSIONS
Stack3
Pollutant
As measured:
02, percent dry
C02, percent dry
NOX, ppm dry
N20, ppm dry
CO, ppm dry
TUHC, ppm dry
S02, ppm dry
Continuous monitor
Method 3
S03, ppm
Method 8
Bacharach smoke number
Corrected to 3% 02
NOX (as N02)
N20
CO
TUHC (as CH4)
S02
Continuous monitor
Method 8
S03 (as H2S04)
Method 8
Particulate
Method 5
SASS
Andersen
Range
2.7 to 3.3
13.1 to 13.5
108 to 115
12.9 to 20. 5C
45 to 135
<1
d
e
e
8
ppm ng/Jf
106 73.7
17 11
93 39
<1 <0.2
d ~d
594 574
3.1 4.5
mg/dscm
""T? 14
118 30
579 219
Average
3.0
13.3
106
17.0
93
<1
c
594
3.1
8
Ib/million
Btuf
0.171
0.026
0.091
<0.001
~d
1.34
0.010
0.033
0.071
0.0489
Furnace
Range
2.4 to 2.7
14.4 to 14.
110 to 112
d
68 to 75
d
550 to 610
d
d
d
ppm ng/J
108 77.2
d d
69 29
d d
584 565
~d d
d d
d d
~d d
d d
outletb
Average
2.5
5 14.4
111
d
71
~d
600
d
d
d
Ib/million
f Btuf
0.179
d
0.069
d
1.31
d
d
d
d
d
^Emission measurements by Acurex.
"Emission measurements by Getty Oil Company.
CRange over duplicate analysis of 6 separate gas samples.
^Measurement not performed at this location.
^Extractive sampling procedure, range not applicable.
fHeat input basis.
^Average of two trains run.
3-16
-------
The data in Table 3-5 show relatively good agreement between the monitor
measurements at the stack and furnace outlet locations. In addition, there
was good agreement between the flue gas SOg levels measured with a continuous
monitor at the furnace outlet and by the Method 8 train run at the stack.
The Method 8 results suggest that $03 represents about 0.5 percent of
the total sulfur oxides present in the flue gas. This ratio is significantly
lower than the 5 to 10 percent range typical for residual oil-fired utility
and industrial boilers (Reference 3-2). The 803 and $03 levels measured by
Method 8 in the flue gas would be as expected for complete conversion of all
the sulfur in a 1.2 weight percent sulfur fuel oil with heating value as
noted in Table 2-1. This compares favorably to the 1.06 percent sulfur
content measured in the fuel.
Particulate emissions were measured at 39 mg/dscm by Method 5,
118 mg/dscm by SASS, and 57 mg/dscm as an average of two Andersen impactor
trains. These are in fair agreement. The Method 5 result is the most
trustworthy, since this reference method involves a multipoint (traverse)
isokinetic sampling procedure.
Particle size distribution results from the two Andersen impactor trains
run are shown in Figure 3-6. Results from the two runs are similar. The
mean particle diameter of emitted particulate was in the 3 to 4 ym range, for
runs 1 and 2, respectively.
Emissions of nitrous oxide were also measured in these tests. The level
noted, at 17 ppm, is about 16 percent of the NOX emission level. Tests of
several other fossil fuel combustion sources have shown that ^0 emissions
are generally in the range of 20 percent of the NOX emission level. These
3-17
-------
01
+->
o>
03
5
a>
o
100
50
20 _
10
5.0
2.0
1.0
0.5
0.2
0.1
D
I I
I
I
Run 2
Run 1
I
12 5 10 20 30 40 50 60 70 80 90 95 98 99
Cumulative wt percent less than diameter
Figure 3-6. Emitted particle size distribution.
3-18
-------
data are summarized in Figure 3-7. The point noted for this study falls on
the curve corresponding to other data. The curve noted in Figure 3-7 was
obtained from a least squares fit of all the data points shown in the figure,
with the constraint that the curve pass through the origin. The relationship
shown, N20 = 0.22 NOX, had a corrrelation coefficient (r2) of 0.88.
3.2.3 Organic Species Emissions
Organic analyses were performed on specified flue gas samples according
to EPA Level 1 protocol (Reference 3-1) as outlined in Appendix A. Volatile
organics having boiling points in the C^ to C6 range of less than 100°C
(212°F) were determined by analysis of flue gas grab samples by onsite gas
chromatography. The SASS train particulate, organic module sorbent (XAD-2),
and organic module condensate (OMC) samples were extracted with methylene
chloride in a Soxhlet apparatus. The extracts (XAD-2 and OMC extracts were
combined) were then subjected to total chromatographable organic (TCO) and
gravimetric (GRAV) analyses to determine species within the 100° to 300°C
(212° to 572°F), and greater than 300°C (5728F) boiling point ranges,
respectively. Infrared (IR) spectra of the GRAV residue of the extracts were
also obtained.
The extracts were also analyzed via gas chromatography/mass spectrometry
(GC/MS) for the semivolatile organic priority pollutant species (including
many polynuclear aromatic hydrocarbons (PAH's)). Other major
chromatogram peaks were identified and approximately quantitated.
Since the total organic contents (TCO and GRAV) of the extract samples
were less than 15 mg, liquid chromatographic separations were not performed.
However, low resolution mass spectrometry (LRMS) analysis of the particulate
extract was performed.
3-19
-------
CM
I
s
O
O
O
ro
o
o
(si
O
o
o
This test
Coal-oil-mixture-fired industrial boiler (Reference 3-3)
Q Oil/refinery gas-fired crude oil heater (Reference 3-4)
0 Coal-water-slurry-fired industrial boiler (Reference 3-5)
A Oil/refinery gas-fired Industrial boiler (Reference 3-6)
O Coal/plastic water-fired commercial boiler (R-eference 3-7)
O Coal-fired commercial boiler (R-oference 3-7)
O Coal-water-slurry-fired industrial boiler (Reference 3-8)
O EOR steamer equipped with the EPA low NO burner (Reference 3-9)
i I I 2
100
200
300
, (PPm.
400
I,, dry)
500
600
Figure 3-7. ^0 versus NOX emissions for external combustion sources.
-------
3.2.3.1 Total Organic Analysis
Table 3-6 summarizes measured organic emissions from the low NOX
burner-equipped steamer by organic boiling point range. The organic
emissions are dominated by the volatile (Cj to C$) fraction, which is further
composed primarily of compounds in the 03 and 04 boiling range. No
semivolatile organics were detected. Nonvolatile organics (nominally CIQ+)
were found in the particulate, though not in the sorbent module. This
confirms the high smoke emissions for the tests and suggests soot formation
was occurring.
The Cj_, 63, and 04 volatile hydrocarbon levels noted in Table 3-6
correspond to 0.3, 4.6, and'0.9 ppm, respectively, as measured. The total
hydrocarbon monitor (which was unheated) read <1 ppm (see Table 3-5) for the
tests. The two measurements are in fair agreement; most 04 would not reach
the total hydrocarbon monitor, and the response factor for 03 hydrocarbon
would be less than 1 ppm (as methane).
3.2.3.2 Infrared Spectra of Total Sample Extracts
The results of the IR spectrometry analysis of total sample extracts are
summarized in Table 3-7. The SASS particulate spectrum suggest only the
presence of aliphatic hydrocarbons in the organic fraction. The XAD-2
extract spectrum is consistent with the presence of oxygenated species such
as carboxylic acids and alcohols.1 An interpretable spectrum for this sample
was obtained despite its low organic content as noted in Table 3-6.
3.2.3.3 Low Resolution Mass Spectrometry Analysis of Total Sample Extracts
The SASS particulate extract was subjected to LRMS analysis via direct
insertion probe to obtain compound category composition information. The
compound categories searched for with the characteristic ions used to
3-21
-------
TABLE 3-6. TOTAL ORGANIC EMISSIONS SUMMARY
Organic category
mg/dscm ng/J
Volatile organics analyzed in the
field by gas chroma tography
Cl
C2
C3
C4
(V
Total Ci-Cg
0.2
0
8.4
2.2
0
0
TO
0.07
0
3.0
0.80
0
0
3.9
Semivolatile organics analyzed by
TCO
Filter
XAD-2
Total C-;
Nonvolatile organics analyzed
by gravimetry
Filter
XAD-2
Total
<0.004
<0.004
0.3
Total organics
11.1
<0.001
<0.001
0.11
<0.04
4.0
TABLE 3-7. SUMMARY OF INFRARED SPECTRA OF TOTAL SAMPLE EXTRACTS
Sample
Parti cul ate
extract
XAD-2
Wave number
(cm-1)
2980-2910
3500-3020
1660
1270-1130
Intensity
Strong
Strong
Strong
Strong
Possible
assignment
C-H stretch
0-H stretch
C=0 stretch
C-0 stretch
Possible compound
categories present
Aliphatic
hydrocarbons
Oxygenated
hydrocarbons such
as carboxylic
acids or alcohols
3-22
-------
identify them are listed in Table 3-8. Table 3-9 notes compound categories
found and their relative abundance (intensity). As noted, aliphatic
hydrocarbons was the major organic category present in the sample. Minor
categories present were ketones and heterocyclic nitrogen compounds.
Specific compounds detected suggested that the ketones were chiefly
fluoren-9-one, and the nitrogen heterocyclics were chiefly ethyl carbazole.
The LRMS results suggest that fluoren-9-one and ethyl carbazole were present
at levels in the 250 ug/g of particulate range. Confirmation of this by
GC/MS is discussed in Section 3.2.3.4.
3.2.3.4 Gas Chromatography/Mass Spectrometry of Total Sample Extracts
Capillary column GC/MS analyses for the semivolatile organic priority
pollutant species, a category which includes several polynuclear aromatic
hydrocarbons (PAH's), were performed on the SASS particulate and XAD-2
extracts. The compounds sought in the analyses and their respective
detection limits are listed in Table 3-10. In addition, major peaks in the
chromatogram, other than these compounds, were identified and quantitated.
Results of the analyses are summarized in Table 3-11.
Of the PAH's, only naphthalene, phenanthrene, and pyrene were found,
and, except for the naphthalene, which was present at the highest
concentration, these were found only in the particulate. The other species
detected were generally oxygenated aromatics and fused aromatics. Benzoic
acid was present at relatively high levels, followed by fluoren-9-one and
ethyl carbazole. The levels of the fluorenone and ethyl carbazole, at 180
and 110 ug/g particulate respectively, confirm the qualitative results of the
LRMS analyses discussed in Section 3.2.3.3.
3-23
-------
TABLE 3-8. COMPOUND CLASSES AND FRAGMENT IONS SEARCHED
FOR BY DIRECT INSERTION PROBE LRMS
Compound class
Fragment ions (m/e")
Polynuclear aromatic
hydrocarbons
Aliphatic hydrocarbons
Halogenated aliphatics
Aromatic hydrocarbons
Ethers
Alcohols
Phenols
Nitriles
Phthalate esters
Amines
Ketones
N-heterocyclics
Mercaptans, sulfides
Benzothiophenes
Carboxylic acids
Amides
178,202,216,228,252,276
57,71
49,63,79,81,93,95,107,109
50,51,77,78,79,91
45,59,73
45,59,61,73,75
51,77,94
54,68,82
149,167
44,58
51,71
117,129,167,179
47,61,75
57,58,59,69,70,85,97,111,125
60,73,149
58,72,86,100
TABLE 3-9. SASS PARTICULATE EXTRACT LRMS RESULTS*
Intensity5
Category
MW range
Major categories
100
10
10
1
Specific compounds
10
10
Aliphatic hydrocarbons 150 to 250
Hetercyclic nitrogen compounds 150 to 200
Ketones 150 to 200
Polynuclear aromatic hydrocarbons 150 to 250
Fluoren-9-one 180
Ethyl carbazole 195
aTotal organic content of this sample is 3.0 mg/g particulate,
GRAY compounds.
"100: major component; 10: minor component; 1: trace component.
3-24
-------
TABLE 3-10. COMPOUNDS SOUGHT IN THE GC/MS ANALYSIS AND THEIR DETECTION
LIMITS (ng/ul INJECTED)
2,4,6-trichlorophenol
p-chloro-m-cresol
2-chlorophenol
2,4-dichlorophenol
2,4-dimethylphenol
1,2,4-tri chlorobenzene
Acid compounds
5 2-nitrophenol
5 4-nitrophenol
5 2,4-dinitrophenol
5 4,6-dinitro-o-cresol
5 pentachlorophenol
phenol
Base neutral compounds
1
1
1
1
1
1
1
1
5
40
1
1
7,12-dimethyl benz(a)anthracene 40
N-nitrosodi-n-propylamine 5
N-nitrosodimethylamine NA
N-nitrosodiphenylamine 1
acenaphthene 1
acenaphythylene 1
anthracene 1
benzo(ghi)perylene 5
benzidine 20
benzo(b)fluoranthene 1
benzo(k)fluoranthene 1
benzo(a)anthracene 1
benzo(a)pyrene 1
1,2-di chlorobenzene
1,2-diphenylhydrazine
(as azobenzene)
1,3-dichlorobenzene
1,4-di chlorobenzene
2,4-di n i trotoluene
2,6-dinitrotoluene
2-chloronaphthalene
3,3'-di chlorobenzi di ne
3-methyl cholanthrene
4-bromophenyl phenyl ether
4-chlorophenyl phenyl ether
benzo(c)phenathrene
bi s(2-chloroethoxy)methane
bis(2-chloroethyl)ether
bis(2-chloroisopropy!)ether
bis(2-ethylhexyl)phthalate
butyl benzyl phthalate
chrysene
di-n-butyl phthalate
di-n-octyl phthalate
di benzo (a, (i)anthracene
dibenzo(c,g)carbazole
diethyl phthalate
dimethyl phthalate
fluoranthene
fluorene
hexachlorobenzene
hexachlorobutadiene
hexachlorocyclopentadiene
hexachloroethane
indeno(l,2,3-cd)pyrene
isophorone
naphthalene
ni trobenzene
perylene
phenanthrene
pyrene
5
20
20
20
5
1
40
1
1
1
1
1
1
1
1
5
40
1
1
1
1
1
1
1
1
5
1
1
1
40
1
1
3-25
-------
TABLE 3-11. COMPOUNDS DETECTED IN THE GC/MS ANALYSES
Species
Filter participate3 XAD-2 extract3,b Total flue gasc
(ug/g) (ug/dscm) (yg/dscm) (ug/dscm)
Semivolatile organic
priority pollutants
Naphthalene
Phenanthrene
Phenol
Pyrene
Other compounds
identified
1.6
2.6
1.9
0.97
0.19
0.30
0.22
0.11
1.2
<0.04
<0.04
<0.04
a27.0 dscm sampled, 3.11g particulate on filter.
bAverage of duplicate injections.
cSum of average of duplicate XAD-2 result plus filter result.
1.4
0.30
0.22
0.11
Benzofurandione
Benzoic acid
Benzothiazole
Di chl orodi benzosul f one
Ethyl benzoate
Ethyl carbazole
Fluoren-9-one
Terphenyl
28
110
180
45
3.3
13
20
5.2
0.44
34
0.52
0.40
w^
0.44
34
3.3
0.52
0.40
13
20
5.2
3-26
-------
REFERENCES FOR SECTION 3
3-1. Lentzen, D. E., et.al., "IERL-RTP Procedures Manual: Level 1
Environmental Assessment (Second Edition)", EPA 600/7-78-201, NTIS
PB293795, October 1978.
3-2. Waterland, L. R., et.al., "Environmental Assessment of Stationary
Source NOX Control Technologies Final Report," EPA-600/7-82-034, NTIS
PB82-249350, May 1982.
3-3. DeRosier, R., "Environmental Assessment of a Watertube Boiler Firing
Coal/Oil Mixture," Acurex Report TR-81-87/EE, March 1984.
3-4. DeRosier, R., "Environmental Assessment of a Crude-Oil Heater Using
Staged Air Lances for NOX Reduction," Acurex Report TR-82-94/EE,
November 1983.
3-5. DeRosier, R., and L. R. Waterland, "Environmental Assessment of a
Watertube Boiler Firing a Coal-Water-Slurry," Acurex Report
TR-84-156/EE, February 1985.
3-6. Castaldini, C., et.al., "Environmental Assessment of NH3 Injection for
an Industrial Package Boiler," Acurex Draft Report TR-83-139/EE,
November 1983.
3-7. Waterland, L. R., et.al., "Environmental Assessment of a Commercial
Boiler Firing a Coal/Plastic Waste Mixture," Acurex Draft Report
TR-85-_/EE, February 1985.
3-8. VanBuren, D., and L. R. Waterland, "Environmental Assessment of a
Coal-Water-Slurry-Fired Industrial Boiler," Acurex Draft Report
TR-84-155/EE, March 1985.
3-9. Castaldini, C., et al., "Environmental Assessment of an Enhanced Oil
Recovery Steam Generator Equipped with the EPA Low NOX Burner," Acurex
Draft Report TR-85-174/EE, January 1985.
3-27
-------
SECTION 4
QUALITY ASSURANCE ACTIVITIES
Specific quality assurance (QA) activities performed to determine the
accuracy and precision of the laboratory analyses performed on samples
collected in this test program included:
Spiking a sample of cleaned XAD-2 resin from the lot used in this
test with TCO, GRAY, and semivolatile priority pollutant compounds
and analyzing the spiked resin to determine the accuracy (recovery)
of the resin extraction and subsequent analyses
» Analyzing NBS flyash for mercury to determine the accuracy of the
atomic absorption technique used
« Performing duplicate TCO and GC/MS injections on the SASS train
XAD-2 extract to determine the precision of these measurements
The following paragraphs discuss results of these QA activities.
4.1 ACCURACY DETERMINATIONS
A sample of XAD-2 resin from the cleaned lot used for these tests was
spiked with 3.0 mg bis(2-ethylhexylJphthalate, 200 ug tetradecane, and
100 ng each of ds-naphthalene, phenanthrene, and pyrene. Thus, this resin
contained 0.3 mg TCO compounds (tetradecane and naphthalene) 3.2 mg GRAV
compounds (the phthalate, phenanthrene, and pyrene) and 100 ug each of the
three polynuclear aromatics for the semivolatile organic priority pollutant
analysis.
4-1
-------
Results of the analyses of this spiked resin are shown in Table 4-1. As
noted, the recovery of the TCO analysis was 125 percent, of the GRAV analysis
was 31 percent, and averaged 77 percent for the GC/MS analyses. If these are
interpreted to be the accuracy of these measurements, all fall within the
project accuracy objective (Reference 4-1) also noted in Table 4-1.
A sample of NBS 1633a flyash was analyzed by the cold vapor atomic
absorption technique used for sample determinations in this project. The
analysis result was 0.18 ppm Hg in the sample; the NBS certified value is
0.16 ppm. Thus, the accuracy of this measurement was within 11 percent,
again within the QA objective of ±20 percent for this measurement.
4.2 PRECISION DETERMINATIONS
The XAD-2 extract samples from the SASS train for this test were
analyzed in duplicate for TCO content, and for the semi volatile organic
priority pollutants and other major peaks by GC/MS. The two TCO measures
were 0.087 and 0.094 ng/injection, giving a relative standard deviation of
5.5 percent. This is within the precision objective of this measurement of
10 percent (Reference 4-1).
Results of the duplicate GC/MS injections are summarized in Table 4-2.
The relative standard deviations for all compounds quantitated were well
within the project precision objective of 50 percent for this measurement.
4-2
-------
TABLE 4-1. XAD-2 RESIN SPIKE AND RECOVERY RESULTS
Measurement
Spiked Recovered
amount amount Percent Implied Accuracy
(mg) (ing) recovery accuracy objective*
Total chromatographable 0.3
organics (TCO)
0.4
125
+25
±50
Gravimetric organics
(GRAV)
3.2
2.6
81
-19
±50
Semi volatile organic
priority pollutants:
d8-Naphthalene 0.1 0.077 77
Phenanthrene 0.1 0.077 77
Pyrene 0.1 0.077 77
Reference 4-1.
-23
-23
-23
Average
77
-23
-50
+100
TABLE 4-2. DUPLICATE GC/MS ANALYSIS RESULTS
FOR THE XAD-2 EXTRACT
Compound
Phenol
Benzofurandione
Benzoic acid
Dichl orodibenzosul tone
Ethyl benzoate
Run 1
ug/ train
26
10
960
13
10
Run 2
ug/ train
37
14
900
15
12
Relative
standard
deviation
U)
24.7
23.6
4.6
10.1
12.9
4-3
-------
REFERENCE FOR SECTION 4
4-1. "Quality Assurance Plan for the Combustion Modification Environmental
Assessment," Acurex Corporation for EPA Contract 68-02-2160,
September 10, 1982.
4-4
-------
SECTION 5
SUMMARY
A comprehensive emissions testing program was performed on an enhanced
oil recovery steam generator (EOR steamer) equipped with an MHI PM low NOX
burner, with less detailed comparison testing performed on an identical unit
equipped with a conventional North American burner.
Full load NOX emissions from the conventional burner-equipped boiler
varied from 365 ppm (corrected to 3 percent 63) with stack 03 of 6.6 percent
to 197 ppm with stack 02 of 2.3 percent. However, smoke emission levels were
unacceptably high at the lower 02, lower NOX levels. A practical NOX
emission limit (acceptable CO and smoke emissions) of about 300 ppm
(corrected to 3 percent 02) with flue gas 02 of 3.7 percent could be
maintained. At 75 percent load MOX emissions were reduced to about 250 ppm
with stack 02 of 4.0 percent and acceptable CO and smoke emissions.
Full load NOX emissions from the low-NOx burner-equipped steamer varied
from 95 to 180 ppm (corrected to 3 percent 02) with variations in the overall
excess air level (as measured
-------
level of about 110 ppm (3 percent Og) could be maintained with acceptable CO
and smoke.
Comprehensive emissions testing of the low-NOx burner-equipped boiler
was performed with the burner operation at a nominal low-NOx setting. With
54 percent of the combustion air supplied to the burner's premix flame,
36 percent to the diffusion flame, and 10 percent to the OFA ports, and with
9.5 percent FGR and 3.0 percent stack 03, NOX emissions were 106 ppm, CO
emissions were 93 ppm, and Bacharach smoke number was 8. At this condition
S02 and SQ^ emissions were 594 ppm and 3.1 ppm respectively, and particulate
emissions were 39 mg/dscm. The mean particle size of the particulate was in
the 3 to 4 ym range (two separate impactor train runs).
Total organic emissions were 11.1 mg/dscm, 97 percent of which was in
the volatile (Cj to Cg) boiling point range; the remainder was in the
nonvolatile (>Cig) boiling point range. The nonvolatiles were condensed on
flue gas particulate and consisted largely of aliphatic hydrocarbons,
heterocyclic nitrogen compounds (ethyl carbazole), and ketones (fluorenone).
Of the polynuclear aromatic hydrocarbons specifically analyzed for in
flue gas emissions, only naphthalene (1.4 yg/dscm), phenanthrene
(0.3 yg/dscm), and pyrene (0.11 yg/dscm) were detected. Other compounds
identified as comprising the flue gas organfcs included benzoic acid, ethyl
carbazole, and fluoren-9-one with emission levels in the 13 to 34 yg/dscm
range, and phenol, benzofurandione, benzothiazole, ethyl benzoate, and
terphenyl with emission levels in the 0.1 to 5.2 yg/dscm range.
5-2
-------
APPENDIX A
SAMPLING AND ANALYSIS METHODS
Emission test equipment was provided primarily by Acurex Corporation.
Onsite equipment included a continuous flue gas monitoring system; the source
assessment sampling system (SASS) for particulate mass, semivolatile, and
nonvolatile organic emissions measurement; a combined EPA Method 5 and 8
train for measuring particulate, S02 and $03 emissions; an Andersen cascade
impactor train for measuring emitted particle size distribution; gas grab
sampling equipment and an onsite gas chromatograph equipped with a flame
ionization detector (GC/FID) for determining flue gas Cj to CQ hydrocarbon
emissions; and gas grab sampling equipment for laboratory determination of
N20 emissions by gas chromatography using an electron capture detector
(GC/ECD). All the above flue gas emission sampling was performed at the
steam generator stack.
In addition, Getty Oil Company provided flue gas monitoring of 02, C02,
CO, NOX, and S02 at the steam generator furnace outlet location.
The following sections summarize the equipment sampling and analysis
procedures used by Acurex in the evaluation of the steam generator/low NOX
burner.
A.I CONTINUOUS MONITORING SYSTEM
Rack-mounted monitors and recorders located in a mobile emission
laboratory were used for continuous measurement of NOX, CO, total unburned
A-l
-------
1. In situ filter, 0.7u, D9.999 percent efficient
2. Exhaust duct
3. 316 stainless steel probe
4. Four pass conditioner-dryer. 3)6 stainless steel internals
5. 3/8-inch unheated Teflon tubing
6. Teflon-lined sample pump
7. 3/8-inch heated Teflon tubing
8. Rotameter
9. 1/4-inch Teflon tubing
10. Calibration gas manifold
11. Calibration gas selector valve
12. Calibration gas cylinders
13. Backpressure regulator
Exhaust
duct
Figure A-l. Continuous monitoring system.
-------
hydrocarbon (TUHC), C02, and 03. Figure A-l illustrates the continuous flue
gas extractive sampling system and monitors arrangement. Flue gas was drawn
through an in-stack filter and a heated stainless steel probe to a gas
conditioning and refrigeration system designed to remove water. An unheated
line was then used to bring the conditioned gas to the monitors. Calibration
gases were used to monitor and correct the drift in the instruments. The
calibration gases follow the same path as the flue gas being monitored in
that both are conditioned at the stack prior to analysis. Table A-l lists
the instrumentation constituting the continuous monitoring and flue gas
extractive sampling system used in this test program.
A.2 PARTICULATE AND SULFUR OXIDE EMISSIONS
Particulate mass emissions were measured in accordance with
EPA Reference Method 5 and S02 and $03 emissions were measured in accordance
with EPA Reference Method 8. A combined Method 5/8 train employing the
Acurex High Volume Stack Sampler (HVSS), illustrated schematically in
Figure A-2, was used in this program. A glass-lined stainless-steel probe
was used to isokinetically extract the gas sample from the stack.
Particulate was removed by a heated 142 mm (5.6 in.) diameter glass fiber
filter. Both the filter and the sampling probes were maintained at 1208C
(2508F) as specified by Method 5.
The impinger train consisted of four glass impingers with a fritted
glass filter placed between the first and second impingers as specified by
Method 8. The first impinger contained 100 ml of 80 percent isopropanol
(20 percent water); the second and third impingers contained 100 ml of
3 percent ^2 i" water5 and tne fourtn impinger contained 200g of silica
gel.
A-3
-------
TABLE A-l. CONTINUOUS MONITORING EQUIPMENT IN THE MOBILE LABORATORY
Instrument
NO
NOX
CO
TUHC
C02
02
Sample gas
conditioner
Strip chart
recorder
Principle of
operation Manufacturer
Chemi luminescence Thermo Electron
Non dispersive ANARAD
infrared (NDIR)
Flame ionization Beckman
detector
Nondispersive ANARAD
infrared (NDIR)
Fuel cell Teledyne
Refrigerant Hankinson
dryer-condenser
Dual pen Linear
analog
Instrument
model Range
10 AR 0-100 ppm
0-500 ppm
0-1,000 ppm
0-5,000 ppm
500R 0-1,000 ppm
400 0-10 ppm
0-100 ppm
0-1000 ppm
AR500 0-20 percent
0-5 percent
0-25 percent
E-4G-SS 10 scfm
400 0-10 mV
0-100 mV
0-1V
0-10V
A-4
-------
I'rcbo
M>: liuu (diameter)
F Till or-
Teflon
Ik P f
V - -'
:> 1
\ 1
V "S" type
pilot tube
N
,-f filter
j |_ oven
<=>r^='
A
Jl
Oven
T.C.
i
ws
LlMIIH-'l-l. HHJ
/ line
/
Ice/v/ater ~l
bath ~\.
100 ml ->.
«0^ II'A ^\
*"^
Sinuli-tireenber-'j _^z,
"" j IUU INI
'roportional ( 34 11 0
temperature | '22
controllers |_
Al> Magnchelic
gauge
All orifice
plate
Check
Vdl VG
Impiinjer
thermocouple
Silica gel
dcssicant
Modified
SmitJi-Greenbery
i nip i 11 (jer
Gas meter thermocouples
Fine adjustment |
bypass valve
Digital tei.iperaturc
indicator
Vacuum line
Vacuum gauye
Coarse adjustment vulv
7Airtight vacuum pump
Figure A-2. Schematic of Method 5/8 sampling train.
-------
S03 (H2$04 mist) is collected in the first impinger and S02 (oxided to
$04) in the second and third. These were determined in the laboratory by
titration with 0.01N barium perchlorate using thorin indicator.
A.3 ORGANIC EMISSIONS
Emissions of organic compounds and compound categories were sampled
using the source assessment sampling system (SASS). Designed for Level 1
environmental assessment (Reference A-l), the SASS collects large quantities
of gas and solid samples required for subsequent analyses of inorganic and
organic emissions.
The SASS, illustrated in Figure A-3, is generally similar to the system
utilized for total particulate mass emission tests (a high volume Method 5
train) with the exception of:
« Particulate cyclones heated in the oven with the filter to 2308C
(450'F)
The addition of a gas cooler and organic sampling module
The addition of necessary vacuum pumps to allow a sampling rate of
2 1/s (4 cfra)
The particulate cyclones shown were not used for these tests because of
the low particulate loading in the flue gas.
Schematics outlining the standard sampling and analytical procedures
using the SASS equipment are presented in Figures A-4 and A-5. The inorganic
analyses of SASS train samples noted in the figures were not performed for
these tests.
The SASS train particulate, XAD-2 resin, and organic module condensate i
(OMC) were extracted with methylene chloride in a Soxhlet apparatus. The
A-6
-------
Heated oven
Stainless
steel
sample
nozzle
Stack T.C.
v$
Organic module
Gas temperature T.C.
1/2" Teflon line
Stack
velocity
AP magneheltc
gauges
\/2" Teflor
line
Isolation
ball valve
Stainless steel
probe assembly
^t*-4
^ I I Oven T.
W" Tef ton
Condensate
lector vessep
Imp/cooler trace
element collector
\./f~"eater controller
ll "I
Coarse adjustment
valve
Orifice All
iiiagnehel i c
gauge
Vacuum pumps
(10 ft3/m«n each)
Implnger
T.C.
Ice bath
600 grams
silica gel
desicant
500 ml
0.2 M AgNOi
0.2 M (NH4)z
500 ml
301
Heavy wall
vacuum line
| Control moduli ' Dr
Note: T.C. = Thermocouple
Figure A-3. Source assessment sampling system schematic.
-------
SAMPLE
1- PYTt (INF . ..,
i-- pypi nvF . - ..
PflHQF WACW ffTP
SORSENT CARTRIDGE -
AQUEOUS CONOENSATE
FIRST IMPINGES
M
U
2 z z
< u o g
5* STs " **
=* Iv = 2
y o x x o
* e£ 5 ui s S
5 »H 3 £ 1 «
I §5 * i > I I
£ *t o > J < Q = C 5
x u c c c c u u < S
w U O O w (3 ^- -i a»ui
*v . S * ° "
^> acT SPL|T
ar ^^- em
'0 9 H
*
SPUT x 10
S GRAMS ' "
COMBINE
a AQUEOUS PORTION
\v ORGANIC EXTRACT \
J? g
in
5
*
= <
SECOND AND THIRD
TOTALS
H '(Quired, umoie should bt tn aiidi lor biological >nalyfn IT tha point.
525
Thri mo n quired to ot tetil mm of piniculii* cilch. If ih« »mpl« muni 10% of the teal cyclone i
(rli.t Ufnpii wctgnt procwd to irnlyin. If iht arnpl* n IMI then 10% of the atcti, hold in r
6 1
Figure A-4. Flue gas analysis protocol for SASS samples.
A-8
-------
liui suuncf
IWACIW
I ItlOASSAY I
Figure A-5. Flue gas sample analysts protocol.
-------
XAD-2 and OMC organic extracts were combined for analysis. The extracts were
analyzed for total organic content in two boiling point ranges: 100° to
300eC (nominally Cy to C^s organics) by GC/FID for total chromatographable
organics (TCO) and >3008C (nominally >CIQ organics) by gravimetry (GRAV).
Infrared (IR) spectra were obtained of the GRAV residue of extracts. GC/mass
spectrometry (MS) in accordance with EPA Method 625 for the semivolatile
organic priority pollutant species was also performed on extract samples.
Extract samples containing total organic content corresponding to emissions
of >0.5 mg/dscm were analyzed by low resolution mass spectrometry to identify
the major compound categories present. Figure A-6 illustrates the organic
analysis methodology generally followed.
A.4 PARTICLE SIZE DISTRIBUTION
An Andersen 2000 Mark III in-stack cascade impactor was used to measure
particle size distribution. The impactor was preheated inside the stack for
30 minutes prior to the start of sampling. Sampling was performed
isokinetically at a point of average stack gas velocity.
The Mark III impactor consists of multiple stages which collect
different particle sizes. Each stage consists of orifices of a specific
diameter above a collection plate containing a glass fiber substrate. The
orifice sizes of each stage are different and are arranged in descending
order, the largest being stage 0.
For sampling, the stack gas was drawn in through the stainless-steel
nozzle into the heated preseparator and impactor. The gas flowed through a
stainless-steel probe and a Teflon line into the condensor train consisting
of a series of 3 Lexan impingers. The gas was then pulled through a carbon
A-10
-------
Organic Extract
or
Neat Orqanic 1 iouid
.
Concentrate
Extract
t t
GC/MS Analysis,
POM, and other Infrared Analysis
organic species
i
t t
Repeat TCO
Gravimetric Analysis
if necessary
Aliquot containing
15-100 mg
t
Solvent
Exchange
1
Liquid
Chromatographic
Separation
t M 1
* t *
Seven Fractions
t
Infrared Analysis
- -
T
Mass Snectra
Analysis
TCO
Gravimetric
Analysis
Figure A-6. Organic analysis methodology.
A-ll
-------
vane pump, dry gas meter, and calibrated orifice. The temperature of the gas
leaving the impactor and the impactor temperature were measured during each
test with type K thermocouples.
After a test, the Mark III impactor was carefully disassembled and the
glass fiber substrates returned to their original foil containers. Any
particulate matter which adhered to the impaction plates was brushed onto the
appropriate filter. The samples were desiccated for 24 hours and weighed to
the nearest 0.01 mg. The nozzle and Mark III inlet cone were rinsed
thoroughly with acetone into a labeled amber jar. These washings were
transferred to tared aluminum pans, evaporated, then desiccated for 24 hour
and weighed to the nearest 0.1 mg.
A.5 C]_ TO C6 HYDROCARBON SAMPLING AND ANALYSIS
Samples of flue gas for Cj to 05 hydrocarbon analysis were collected
using a grab sampling procedure employing the apparatus illustrated in
Figure A-7. The equipment consisted of a heated, 0.64-cm (1/4-in.) 00
pyrex-lined, stainless-steel probe fitted with a 7-ym sintered
stainless-steel filter at the probe inlet. The outlet of the probe was
directly attached to a diaphragm vacuum pump which was in turn attached to a
500-ml stainless-steel heated sampling cylinder. The sampling cylinder was
insulated with heat tape powered by a varying voltage controller. The
heating jacket kept the sample gas above the dew point to minimize sample
loss due to water condensation.
Prior to sampling, the gas cylinder was purged with stack gas for
3 minutes and then sealed. The trapped flue gas was then analyzed onsite
with a Varian Model 3700 gas chromatograph (GC) equipped with a flame
ionization detector.
A-12
-------
0.7 inn sintered stainless-steel filter
1/4-in. stainlebs-steel
probe
-TutIon diaphragm pump
Pressure
-------
Table A-2 lists the design specifications of the Varian GC. A 1.85m (6 ft)
long, 0.32 cm (1/8 in.) diameter stainless-steel column packed with Poropak Q
60/80 mesh was used to separate the hydrocarbons into their respective
components (C^ to 5). The GC was calibrated with repeated injections of a
Scott Speciality standard gas containing C^ to 5 hydrocarbons (each having a
concentration of 15 ppm). The chromatographic responses for the standards
and the samples were recorded on a Hewlett Packard Model 3390A reporting
integrator.
A.6 N20 EMISSIONS
Stack gas grab samples were extracted into stainless-steel cylinders
similar to those used for C^ to CQ hydrocarbon sampling (Section A.5) for
laboratory analysis for N20. For the analysis each sample cylinder was
externally heated to 120°C (250T); then a 1-ml sample was withdrawn with a
gas-tight syringe for injection into a gas chromatograph. The analytical
equipment consisted of a Varian 3700 gas chromatograph equipped with a 63|\|-f
electron capture detector and a 3.65m (12 ft) stainless-steel column packed
with Poropak Super Q, 80/100 mesh. The injector temperature was kept at
30°C, the detector at 350°C, and the column temperature at 33"C. Elution
time for ^0 was approximately 5 minutes, with a flowrate of 20 ml/min of
nitrogen.
A-14
-------
TABLE A-2. GAS CHROMATOGRAPH SPECIFICATIONS
Van'an Model 3700 Gas Chromatograph:
Sensitivity
Zero range
Noise (inputs capped)
Time constant
Gas required
1 x 10~12 A/mV at attenuation 1
and range 10"12 A/mV
-10'11 to 10"9 A (reversible
with internal switch)
5 x 10"15 A; 0.5 uV peak to peak
220 ms on all ranges (approximate
1 sec response to 99 percent
of peak)
Carrier gas (helium), combustion
air, fuel gas (hydrogen)
A-15
-------
TECHNICAL REPORT DATA
(Please read laurvctions on the reverse before completing]
1 5£°O^T NO
EPA-600/7-86-003a
2.
3. RECIPIENT'S ACCESSION NO.
i. TITLE AND SUBTITLE Environmental Assessment of an
Enhanced Oil Recovery Steam Generator Equipped with
a Low-NOx Burner; Volume I. Technical Results
. REPORT DATE
February 1986
6. PERFORMING ORGANIZATION CODE
7 AUTHOH(S)
C. Castaldini, L. R. Waterland, and H. I. Lips
8. PERFORMING ORGANIZATION REPORT NC.
TR-84-161/EE
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Acurex Corporation
P.O. Box 7555
Mountain View, California 94039
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-3188
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Air and Energy Engineering Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final; 1/84 - 1/85
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES AEERL project officer is Robert E. Hall, Mail Drop 65, 91S/541-
2477. Volume II is a data supplement.
is. ABSTRACT
repOrt discusses results from sampling flue gas from an enhanced oil
recovery steam generator (EOR steamer) equipped with an MHI PM low-NOx burner.
The tests included burner performance /emission mapping tests, comparative testing
of an identical steamer steamer equipped with a conventional burner, and comprehen-
sive testing of the low- NOx-burner- equipped steamer. Comprehensive test measure-
ments included continuous flue gas monitoring; source assessment sampling system
testing with subsequent laboratory analysis to give total flue gas organics in two boil-
ing point ranges and specific quantitation on the semivolatile organic priority pollu-
tants; Cl to C6 hydrocarbon sampling; Methods 5/8 sampling for particulate and SO2
and SC3 emissions; and emitted particle size distribution tests using Andersen im-
pactors. Full- load NOx emissions of 110 ppm (3% O2) could be maintained from the
low-NCx burner at acceptable CO and smoke emissions, compared to about 300 ppm
O2) from the conventional-burner-equipped steamer. At this low-NOx condition,
CO, SO2, and SOS emissions were 93, 594, and 3.1 ppm, respectively. Particulate
emissions were 39 mg/dscm with a mean particle diameter of 3 to 4 micrometers.
Total organic emissions were 11.1 mg/dscm, almost exclusively volatile (Cl to C6)
organics. Three PAHs were detected at from 0.1 to 1.4 micrograms/dscm.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATl Field/Group
Pollution
Boilers
Oil Burners
Crude Oil
Oil Recovery
Assessments
Pollution Control
Stationary Sources
Low-NOx Burners
Environmental Assess-
ment
13B
13A
11H, 08G
081
14B
12. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
69
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
A-16
------- |