EPA-R2-73-084a
(API Publication 4180}
June 1973
Environmental Protection Technology Series
Field Investigation of Emissions
from Combustion Equipment
for Space Heating
I
55
\
01
CD
American Petroleum Institute
1801 K Street, NW
Washington, D.C. 20006
Office of Research and Monitoring
U.S. Environmental Protection Agency
Washington, D.C. 20460
-------
EPA-R2-73-084a
(API Publication 4180)
Field Investigation of Emissions
from Combustion Equipment
for Space Heating
by
R.E. Barrett, S.E. Miller,
and D.W. Locklin
Battelle - Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
Contract No. 68-02-0251
Program Element No. 1A2015
EPA Project Officer: Robert E. Hall
Control Systems Laboratory
.National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
AMERICAN PETROLEUM INSTITUTE
1801 K STREET, NW
WASHINGTON, D.C. 20006
and
OFFICE OF RESEARCH AND MONITORING
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
June 1973
-------
This report has been reviewed by the Environmental Protection Agency arid
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.
AUTHORS' COMMENTS
It has been brought to the authors' attention that some confusion exists between the
terms "emission factors" and "emission standards". The values reported and suggested in this
report are emission factors. Emission factors, as developed in this study, are suitable for use in
estimating mean emissions from a class of fuel burning equipment and jre useful in compiling
areawide emission inventories; they are not suitable for predicting emissions from any one unit
or as regulatory limits or emission standards. Use of emission factors .as omission standards
would tend to place about 50 percent of the units in noncqmpliance.
11
-------
TABLE OF CONTENTS
Chapter
I. SUMMARY 1-1
Objectives and Scope of Phase II
Objectives
Scope
Residential Units
Mix of Units
Conditions Investigated
Varied-Air Runs
Cyclic Runs
Summary of Residential Emissions and Comparison With
EPA Emission Factors
Suggested Emission Factors for Residential Units
Conclusions for Residential Investigation ....
Commercial Boilers
Mix of Boilers
Conditions and Fuels Investigated . . ,
Base-line Condition
Fuels
Typical Emission Trends
Effects of Fuel
Effect of Fuel Nitrogen on NOX
Summary of Oil-Fired Commercial Boiler Emissions ....
Suggested Emission Factors for Oil-Fired Commercial Boilers .
- 2
- 2
- 2
- 4
- 4
- 5
- 5
- 7
-10
-11
-11
-14
-14
-15
-15
-15
-15
-16
-19
-19
-19
Summary of Gas-Fired Boiler Emissions and Suggested
Emission Factors I -22
Conclusions for Commercial Boiler Investigation I -23
II. EMISSIONS FROM RESIDENTIAL UNITS II - 1
Units Included in the Phase II Investigation 11-1
Basis for Selection of Equipment 11-1
Selection of Equipment Mix and Individual Units 11-2
Description of Residential Units 11-2
Procedures Used in the Residential Field Investigation 11-4
Burner Conditions Investigated 11-4
Operating Conditions Investigated 11-6
Emission Measurements - Instruments and Techniques .... 11-9
Emission Results for Cyclic Runs II -13
Summary of Emission Data and Emission Factors II -13
Effect of Tuning and Fuel 11-13
Comparison of Emissions for Different Equipment Features . . II -18
Statistical Ranking of Equipment and Fuel Variables .... II -23
Conclusions Related to Equipment Features and Fuel .... II -36
Measurements on Follow-Up Units II -36
Experiments on the Effect of Cycle II -40
Emission Results for Varied-Air Runs II -44
Emission Characteristics Related to Excess Air II -44
Limits of Acceptable Adjustment II -59
Trial Correlations of Smoke Versus Particulate . 11-59
iii
-------
TABLE OF CONTENTS
(Continued)
Chapter
III. EMISSIONS FROM COMMERCIAL BOILERS Ill- 1
Boilers Included in the Phase if Investigation Ill- 1
Basis for Selection of Equipment . Ill - 1
Selection of Equipment Mix and Individual Boiler Ill - 1
Description of Commercial Boilers Ill - 4
Procedures Used in the Field Investigation Ill- 4
Conditions Investigated Ill- 4
Emission Measurements Instruments and Techniques , . . Ml - 7
Emission Results for Commercial Boilers . Ill 8
Summary of Emission Data and Emission Factors Ill - 8
Influence of Various Parameters on NOX Emissions . . . 111-11
Influence of Operating Condition on Smoke. CO, and
HC Emissions . Ill -27
Influence of Fual on Smoke Emissions ... ... Ill -27
Factors Influencing Participate Emission Ill -36
Trial Correlation of Smoke Versus Particuiate Emissions . . . Ill -39
Particle Si?e Measurements III-39
Emission Factors Ill -42
Emission Factors Related to API Gravity .... ... Ill-42
Suggested Emission Factors for Commercial Boilers .... Ill -48
APPENDICES
A. BACKGROUND DATA ON RESIDENTIAL OIL-FIRED EQUIPMENT
POPULATION A - 1
B. BACKGROUND DATA ON COMMERCIAL-INDUSTRIAL BOILER
POPULATION B - 1
C. FUEL ANALYSES C- 1
D. DETAILS OF FIELD PROCEDURES- D- 1
E. SAMPLING AND ANALYTICAL PROCEDURES FOR GASEOUS
EMISSIONS E- 1
F. SAIWLING AND ANALYTICAL PROCEDURES FOR PARTICULATE
AND SMOKE AND DETAILED PARTICULATE DATA F - 1
G.* DATA FOR RESIDENTIAL UNITS G 1
H.* DATA FOR COMMERCIAL BOILERS H- 1
I.* CALCULATED EMISSION FACTORS FOR RESIDENTIAL UNITS 11
J.* CALCULATED EMISSION FACTORS FOR COMMERCIAL BOILERS J-1
K. DETAILS OF EMISSION-FACTOR CALCULATIONS AND
CONVERSION FACTORS K _ 1
L. ERKATA TO PHASE I REPORT L _ -,
M. REFERENCES M _ .,
"These appendices appear m tfip Data Supplement Volumb which is available as a separate report
-------
ABSTRACT
Information on air-pollutant emissions from residential and commercial heating equipment was developed
in a 2-year field investigation conducted by the Battelle Columbus Laboratories. Emissions measured were CO,
HC, NOX, SO2, particulate, and smoke. The program, jointly sponsored by the American Petroleum Institute and
the U.S. Environmental Protection Ayericy, covered emissions from 33 residential heating units and 13 commer-
cial boilers including effects of various combustion parameters and fuel-oil compositions.
Residential Units
Although performance similarities among units were noted during runs with varied air adjustments, each
unit had unique emission characteristics. For the representative cross section of residential units investigated, it
was found that the largest reduction in emissions could be achieved by replacing pooriy performing unus, as
identified from smoke measurements made during service inspections. Tuning of the units that were generally
performing well reduced smoke levels but produced only slight reductions in emissions on a gravimetric basis,
Tuning to lower smoke by increasing the air adjustment alone couid result in sharply rising emissions of CO and
HC for some units, unless established tuning practice is followed using field-type instruments for smoke and CO*,
measurements.
Burners equipped with flame-retention-type combustion heads produced lower emissions than those with
conventional heads, and newer burners performed with lower emissions than did older burners.
Commercial Boilers
For the sample of typical commercial boilers, which ranged from 40 to 600 boiler horsepower in
capacity, emissions were measured for 33 different combinations of boilers and fuels at various loads and excess
air settings. Although some performance differences were noted, operating parameters within the normal range of
adjustment had minor effect on gaseous emissions.
Fuels included in the investigation were natural gas and five grades of fuel oil, including a 1-percent-sulfur
residual oil that was transported to each boiler site for use as a reference fuel. Fuel characteristics had a
significant effect on emissions, especially particulate and NOX. Particulate emissions with the 1-percent-sulfur oil
averaged about 30 percent of those for conventional No. 6 oil; particulate emissions for No. 2 oil averaged only
about 3 percent of that for No. 6 oil. NOX emissions were lowest with gas and the lighter grades of fuel oil and
increased with increasing fuel nitrogen content.
Emission Factors
On the basis of this investigation, new emission factors are suggested for use in emission inventories and
in evaluating control strategies. Separate emission factors are suggested for residential oil fired units u,id frr
gas-fired and oil-fired commercial boilers, with a distinction between fuel-oil grades or types.
-------
ACKNOWLEDGMENT
The authors wish to acknowledge the assistance and helpful comments of the EPA
Project Officer, Robert E. Hall, and the API SS-5 Task Force during the course of this program.
Membership on the SS-5 Task Force was as follows:
E. Landau (Chairman) Asiatic Petroleum Corporation
R. C. Amero Gulf Research & Development Company
S. P. Cauley Mobil Oil Corporation
H. E. Leikkanen Texaco Inc.
B. L. Mickel American Oil Company
R. E. Paterson Chevron Research Company
C. W. Siegmund Esso Research & Engineering Company
R. A. Beals National Oil Fuel Institute, Inc.
J. R. Gould American Petroleum Institute.
The assistance of the Commercial-Industrial Air Pollution Committee of the American
Boiler Manufacturers Association is acknowledged for their cooperation in the selection of
commercial units and assistance in defining appropriate test conditions. The authors also wish to
acknowledge the enthusiastic cooperation of those manufacturers who made available their
commercial boilers and plant facilities and who donated the services of skilled service technicians
to assist in the burner adjustments heeded to achieve the desired matrix of test conditions.
Acknowledgment is made for the cooperation of the homeowners who granted permission
for the field team to set up its battery of instruments and for their patience under the
inconvenience of having normal comfort control of their heating system interrupted during the
two or three days of measurements.
Special acknowledgment is due W. H. Axtman, Assistant Manager of the American Boiler
Manufacturers Association, and Margaret Mantho, Statistical Editor of Fueloil & Oil Heat, for
their assistance in preparing the equipment population questionnaires, soliciting replies, and
consolidating the statistical data.
The authors are also indebted to numerous Battelle-Columbus staff members who partici-
pated in various aspects of the program. Special adknowledgment is made to Technician J. J.
Fancelli and J. H. Faught, who served on the field team making the measurements, and to
various supporting staff; to R. D. Fischer, D. R. Hopper, and R. E. Thomas, who advised or
assisted in the computer and statistical interpretations of data; to P. R. Webb for particle-size
measurements; to the secretaries who assisted in the preparation of this report; and to R. B
Engdahl, L. J. Hillenbrand, J. A. Gieseke, and A. Levy, who advised on various sampling and
analytical aspects of the program.
VI
-------
FIELD INVESTIGATION OF EMISSIONS FROM
COMBUSTION EQUIPMENT FOR SPACE HEATING
to
AMERICAN PETROLEUM INSTITUTE
Project SS-5 Task Force
and
ENVIRONMENTAL PROTECTION AGENCY
Control Systems Laboratory
by
R. E, Barrett, S. E. Miller, and D. W. locklin
SUMMARY
Information on air-pollutant emissions from combustion equipment is essential to
environmental quality considerations - including development of emissions inventories, control
strategies, and equipment design criteria. Under sponsorship of the American Petroleum Institute
and the U.S. Environmental Protection Agency, the Battelle Columbus Laboratories has con-
ducted a 2-year investigation to develop more comprehensive and up-to-date information on
emissions from representative residential and commercial fuel-burning equipment for space
heating.
The American Petroleum Institute sponsored Phase I of the investigation during the
1970-71 heating season, in which field measurements of gaseous and particulate emissions were
made on 20 oil-fired and 2 gas-fired residential units, plus 7 commercial boilers. Those findings
were reported in API Publication 40991.
The Phase II investigation was conducted during the 1971-72 heating season under the
joint sponsorship of the American Petroleum Institute and the Environmental Protection Agency.
In the Phase II investigation, more detailed emissions measurements were made over a wider
range of conditions on 13 residential units and 6 commercial boilers. This report describes
procedures and .results for the Phase II investigation and conclusions based on both the Phase I
and Phase II studies.
This Executive Summary provides an overview of the finding;, and conclusions from both
Phases I and II.
1"A Field Investigation of Emissions from Fuel Oil Combustion for Space Heating", by A. Levy, S. E. Miller,
k. b. Barrett, E. J. Schulz, R. H. Melvin, W. H. Axtmarj, and D. W. Locklin. API Publication 4099 (November 1,
J971), available from the API Publications Section, 1801 K Street, N.W., Washington, D. C. 20006.
-------
1-2
OBJECTIVES AND SCOPE OF PHASE II
Objectives
The objectives of the Phase II investigation were to develop additional information on
aii-pollutant emissions from combustion equipment used for residential and commercial heating,
including the effects of combustion parameters and fuel composition on emission levels.
Scope
Measurements made on the 13 residential units covered a range of excess-air adjustments
for both as-found and tuned operating conditions. For the six commercial boilers, combustion
parameters included a range of loads and excess air levels; natural gas and three different grades
of fuel oil were fired in five of the commercial boilers to determine effects of fuel characteristics
on emissions,
Emission Measurements. The following gaseous pollutants were measured by continuous
monitoring equipment sampling from the stack:
Carbon monoxide, CO
Total hydrocarbons, HC
Sulfur dioxide, SO2 (only for commercial boilers)
Nitrogen oxides, NOX (NO + NO2)
In addition, oxygen and carbon dioxide concentrations were measured to define the operating
conditions in terms of excess-air levels,
Particulate measurements were made on a gravimetric basis using the EPA sampling train,
with a slight modification of procedures specified in EPA Method 52. The particulate emission so
determined is termed "filterable particulate" in this Summary and report; the additional con-
tribution of "condensable particulate" is discussed in the report and in Appendix F.
Smoke density measurements were made by the Bacharach hand-pump smoke meter, the
standard method used in the oil-burner industry. The smoke spots were evaluated photomet-
rically by a reflectance meter.
Figure 1-1 shows the field instrumentation as set up for measurements on a commercial
boiler. Details of measurement procedures for both gaseous and particulate emissions are
described in Appendixes E and F of this report.
2"Standards of Performance for New Stationary Sources", Federal Register, Vol 36 No 139 Part II n
24895, December 23, 1971. ' ' ' "> PP
-------
1-3
Figure 1-1. Field Instrumentation Used for Measuring Emissions
From Commercial Boilers
-------
1-4
RESIDENTIAL UNITS
A variety of residential oil-fired heating units were selected for the investigation so that
the sample would be representative of the field population of equipment currently in service for
heating single-family homes in the United States. Criteria for this selection of units were: type of
oil burner by atomizing method, burner capacity, type of combustion head for air-fuel mixing,
burner age, and type of heating system or unit.
Mix of Units
Table 1-1 shows the mix of residential units for the total sample of oil-fired units
investigated in Phases I and II.
Table 1-1. Mix of Residential Oil-Fired Units in Sample
Number of
Units
By Burner Type
By Burner Capacity
By Burner
Combustion-head Type
(high-pressure burners
only)
By Burner Age
By System Type
High-pressure
Low-pressure
Other
1.0 gph or less
1,01-1.35 gph
1.36-2.00 gph
Above 2.00 gph
Conventional head
Flame-retention head
Shell head
5 years or less
6-iQ years
11-15 years
Older than 15 years
Warm-air furnace
Hot-water or steam boiler
Service water heater
29
1
1
6
15
3
7
18
9
2
17
1
7
6
12
18
1
-------
1-5
Conditions Investigated
Measurements of gaseous and particulate emissions, obtained with the heating equipment
operating on the homeowner's fuel supply, were made in the as-found condition of adjustment
and the tuned condition.
Tuning. Normal service and adjustment practices were followed in tuning, short of major
replacement or renovation. Most units were tuned to a smoke level below No. 2 on the
Bacharach scale; the mean smoke was reduced from No. 3.2 for the as-found condition to No.
1.3 for the tuned condition.* Smoke was reduced for over half the units; for most of the others,
a slight sacrifice in smoke level permitted an increase in CO2 for higher overall thermal
efficiency. As-found, the mean CO2 level was 7.9 percent; this was increased to 8.1 percent on
tuning.
Types of Runs. Two ypes of runs were made with the residential units in Phase II.
1. Varied-air runs (Phase II only) - Gaseous emissions and smoke were
determined during steady-state operation for a range of excess-air
settings and plotted against CO2 to determine sensitivity of emissions
to air adjustment.**
2. Cyclic runs (Phases I and II) Gaseous and particulate emissions
and smoke were measured during repeated cycles of 10 minutes on
and 20 minutes off.
Varied-Air Runs
Emission characteristics measured during the varied-air runs were generally similar for
most of the residential units. Figure 1-2 illustrates the smoke-vs-CO2 characteristics for steady-
state firing of a typical unit, as determined in the varied-air runs for the as-found and tuned
conditions. A distinct improvement in performance is noted on tuning, because lower smoke
levels could be achieved at the same CO2 (or excess-air setting).
Figure 1-3 snows the gaseous emissions for this typical unit in the tuned condition,
leaving a low-emission operating range between about 7 and 9 percent CO2 (corresponding to a
range of excess air from about 65 to 110 percent). Although each unit yielded unique
characteristic curves, these curves are typical of burners where a relatively wide range of
excess-air levels could be achieved. For this unit and several others, CO and HC emissions
increased sharply at low CO2 levels, while smoke levels remained low. This demonstrates that CO
and HC emissions can be high if units are adjusted by smoke criteria alone. (NOX was not highly
sensitive to air settings.)
*These mean smoke values do not include units in need of replacement, as explained subsequently.
**This procedure was similar to ASTM-D2157-65, "Standard Method of Test for Effect of Air Supply on Smoke
Density in Burning Distillate Fuel".
-------
.a
e
o
E
CO
-C
o
D
O
a
CD
As-Found
7 8 9 10
COg, percent
11
.
o
CO
o
o
a
m
0
Operating
Range for
Low Emissions
32-
28-
24-
E
Q.
Q.
Smoke
16
12-
8
160
140
120
100
80
60
40
20
0
E
Q.
Q.
o
O
12
7 8 9 10 r
CO 2, percent
12
Figure 1-2. Typical Smoke Versus CO2 Curves for a
Residential Unit in the As-Found and
Tuned Conditions
Figure 1-3. Typical Smoke and Gaseous Emission Characteristics
for a Residential Unit in the Tuned Condition
-------
1-7
Cyclic Runs
Figures 1-4 and 1-5 show the distribution of emission factors and smoke levels measured
on both the Phase I and II residential units for cyclic runs in the as-found and tuned conditions.
It should be noted that 3 units were in poor condition and in obvious need of replacement;
these units yielded oily smoke spots and generally produced emissions of CO, HC, and partic-
ulate that were much higher than those for the remaining units.* The oily smoke spots would
reveal these units as needing burner replacement; this condition would be detected by an
experienced serviceman using field-type instruments.
Effect of Tuning on Emissions. While tuning generally resulted in lower smoke levels, it
produced variable effects from unit to unit on gaseous and particulate emissions in the cyclic
runs, causing some pollutant emissions to increase and others to decrease.
In terms of the mean emission factors, this investigation showed that the major reduction
in CO, HC, and particulate is achieved by identifying and eliminating units in poor condition
(those in need of replacement or major repair). For the remaining units, tuning further reduces
mean smoke and CO emissions but has little effect on the mean values of other pollutants.
Table 1-2 summarizes the reduction in mean pollutant emissions that could be accom-
plished by the following steps**:
1. Identifying and replacing the units obviously in poor condition.
2. Completing Step 1 and, in addition, tuning the remaining units.
Table 1-2. Effect on Mean Emissions of Identifying and
"Replacing" Residential Units in "Poor"
Condition and Tuning
Reduction in Emissions, percent
Step 1 Step 2
Identifying and "Replacement"
Pollutant Replacing of Poor Units Plus Tuning
Smoke
CO
HC
NOX
Filterable Particulate
-
>65
87
No Change
17
59
>81
90
No Change
24
*For normally performing units (those not showing oil on smoke spots), CO and HC levels were very low. In fact,
the family automobile that meets the 1975 EPA emissions standards (as of June 1973) and is driven an average
of 12,000 miles would emit annually 25 times as much CO and HC as would be emitted by firing 1000 gallons
of oil, a typical annual consumption for a residential unit.
**Values are based on distribution of "poor" units found in the Phase I and II studies - that is, 3 poor units in a
total sample of 34 units.
-------
O)
-iC
O
CD
Smoke
,
_ ^ £
-
-
Mean Values 9
«- As Found ^^ (yyx
A
OOO
AA
oo
-*- Tuned A
A9A* AOOOOOOO
A^rl i 1 i 1 i 1 i
D 20 40 60 80 10
ouw
t
4
110
70
a 30
CTV
o 1
8 24
^
c/f 20
o
en
'E 16
LU
0
» ° 12
8
4
0 °<
m
Carbon Monoxide
Phase IH .
_ As found A *°
Tuned o A A
AA
r f
o
Mean Values
+ As Found A o
A
^£f&£&l ^° | , | ,
120
+t
!00
80
- 60
0
o 40
^
wf 20
0 '
|2.0
UJ
o
1 1.5
1.0
0.5
n
HC Emissions ,
c
A
» /
r r
A*
O
o^
Mean Values _.A«
A r- -, A** OOO
*- As Found _ Ooo
&AA* OO
-r j A*>?AAO
c- luned . ^.^AAA
majjafjTi
^°°l ,1,1,1,
) 20 40 60 80 100 0 20 40 60 80 10
Percent of Units Having Emissions Less Than or Equal to Stated Value
(Mean values exclude units in need of replacement.)
Figure 1-4. Distribution of Smoke, CO, and HC Emission for Residential Units
-------
15
12
II
10
_ 9
(u
o
t_
9J r>
CL 8
cu
O
0 7
6
5
4
7,
Carbon Dioxide ,
Phase I H
_ As found A
Tuned o A (
0
~^^*
o^
£«AA**
^°
Mean Values OAA
O*"^**
*- Tuned ^
* As Found A ^
cd****
>OgJP
gj^8*
A
_A
O
_
O*
^
1 1 1 1 , I 1 i 1 .
40
>
36
, 32
o 28
en
0
§ 24
in 20
g
CO
CO
'£ 16
UJ
i '2
8
4
0
z
NOx Emissions
-_
A
A
A
AAO°
~ *A°
dft^
Mean Values -.^SA^
'J^**-'^
n. Tuned otP^**^
~^ As Found o° ««
o00^**
0°^k**
-/aQt*
^^'
-i«
, 1 , 1 , 1 , i ,
1
k
16
"a
0 t4
O
0
£ 12
Q>
O i *-»
10
y
Fj O
a>
JD
o
^ 6
LZ
4
2
0
Filterable Particulate
«
i
0
^
_ A
Mean Values
A A
. Tuned A?AA
^ As Found 9-£&Qr
i 1 i 1 i I i 1 i
20 40 60 80 100
0 20 40 60 80 100
0 20 40 60 80 100
Percent of Units Having Emissions Less Than or Equal to Stated Value
(Mean values exclude units in need of replacement.)
Figure 1-5. Distribution of CO2, NOX, and Filterable Particulate Emissions for Residential Units
-------
no
Follow-Up Measurements. For four of the residential units, follow-up measurements were
made twice during the heating season at 2-month intervals to investigate the effect of operating
time since tuning. Except for one unit which experienced nozzle clogging, only minor shifts in
smoke, NOX, and particulates were noted during the 4-month operating period. Emissions of CO
and HC for one unit increased significantly but not to serious levels.
Summary of Residential Emissions and
Comparison With EPA Emission Factors
Table 1-3 provides a summary of mean emissions for each of the pollutants measured for
the residential units during cyclic runs, Mean values of emission factors* for the Phase I and II
investigations are shown. Also shown for comparison are the emission factors published by
EPA3.
Table 1-3. Comparison of Mean Emissions for Cyclic Runs on Residential Oil-Fired Units
Units IViaan Mean Emission Factors, lb/100Q gal
in Smoke Filterable
Units Condition Sample3 No.b CO HC NOX Particulate
Mean Values From Phase I and II Battelle/API/EPA Investigation:
All units
All units, except
those in need of
reolacement
As-Found
Tuned
As-Found
Tuned
32
33
29
30
(c)
(c)
3.2
1.3
>22.1
>16.4
7.8
4.3
5.7
3.0
0.72
0.57
19.4
19.5
19.6
19.5
2.9
2.3
2.4
2.2
Current EPA-Published Emission Factors:3
5.0 3.0 12.0 10.0
Suggested Emission Factors: For residential units in areas having regular service and inspection
10.0 1.5 20.0 2.5
a One unit was a furnace installed in the laboratory and was not included in deriving the mean emissions values for
the as-found condition.
^ Smoke data ai 5-minute point of on-cycle.
c Oily smoke spots prevented valid mean values.
*Emission factors represent average emission levels for a category of equipment and are used in area emission
inventories and in evaluating control strategies. Emission factors should not be confused with emission standards.
3"Compilalion of Air Pollutant Emission Factors", U. S. Environmental Protection Agency Office of Air Pro
grams Publ. No. AP42 (February 1972).
-------
Ml
Suggested Emission Factors for Residential Units
This investigation was conducted over two heating seasons and comprised measurements
on 33 oil-fired residential units selected to be generally representative of the existing equipment
population. Because this study has provided more extensive data than available at the time the
present EPA compilation of emission factors was prepared, it is suggested that the emission
factors applying to oil-fired residential equipment be updated as shown in Table 1-3.
CONCLUSIONS FOR RESIDENTIAL INVESTIGATION
The principal conclusions reached from this investigation on a representative sample of
oil-fired residential heating units are outlined below, considering findings from both Phases 1 and
II.
1. Inspection and Tuning. Air pollutant emissions from oil-fired residential heat-
ing units can be significantly reduced on an area-wide basis by the use of
available service procedures.
The most effective step in reducing area-wide emissions is the identification
and replacement (or major renovation) of units in poor condition. Such units,
identified by the presence of oil on smoke measurement spots during this
investigation, were found to be contributing a disproportionately large share
of the CO, HC, and particulate emissions.
Tuning the remaining burners by use of good service practices achieved a
significant further reduction in smoke and CO emissions. Mean smoke levels
were reduced in tuning from 3.2 to 1.3 Bacharach smoke number and mean
CO values were reduced from 7.8 to 4.3 lb/1000 gal. Tuning produced little
change in NOX, while filterable particulate and HC emissions decreased
slightly.
Tuning increased the overall thermal efficiency for 60 percent of the Phase II
units: the average incremental increase of overall efficiency for all urtits was
1.7 percent. This has the effect of decreasing total annual emissions due to
resulting savings in fuel input.
2. Range of Emissions and Their Distribution. A broad range of emission levels
was observed for the entire sample of units in the cyclic runs, but the
distribution of emission levels was relatively uniform except for those units in
need of replacement. NOX levels showed the least variability.
3. Emission Characteristics of Individual Burners. Varied-air runs in Phase II
revealed that each burner had its own unique' emission characteristics over a
range of air settings. Similarities in pattern were noted, but the permissible
operating ranges for low emissions and the adjustment sensitivity were widely-
different among burners. For some burners, an attempt to set a minimum
-------
1-12
smoke level by increasing the excess air would result in sharply increasing CO
and HC emissions. Hence, both smoke and CO2 measurements should be used
in adjusting a burner to insure low smoke at a high CO2 level.
4. Smoke Measurements as Indicator for CO and HC. The varied-air runs showed
that low smoke readings were a good indicator of low CO and HC emissions
when operating at the low-excess-air side of the adjustment range; however,
smoke was not a good indicator of these emissions at the high-excess-air side
of the range.
5. CO and HC Levels. Emissions of CO and HC from all units (except those
identified as requiring replacement or major repair) were so low during cyclic
runs as to be insignificant in terms of contribution to air pollution.
6. Effect of Combustion Head and System Type. Burners with flame-retention
combustion heads operated with lower mean values of CO, HC, and filterable
particulates than did burners with conventional combustion heads. No signif-
icant trends were observed relative to other factors of burner or system type.
7. Effect of Burner Age. Burner age had a significant influence on emissions,
with newer burners yielding lower emissions. Newer burners produced signif-
icantly less CO and filterable particulate for the tuned condition and slightly
lower levels of HC, NOXJ and smoke for all conditions. The greatest difference
between emission levels for new and old burners occurred between the burner
age groups divided at 15 years. (These observations as to the effect of burner
age are believed to be due to improved burner designs to meet higher
performance goals recognized voluntarily by equipment manufacturers in
recent years, rather than to a deterioration in service that cannot be restored
by competent tuning.)
8. Effect of Operation Since Tuning. Follow-up measurements, made twice dur-
ing the heating season at 2-month intervals for four units, were not consistent.
Emissions remained nearly constant for two units. CO and HC emission
increased for another unit, while other emissions remained constant. Smoke,
CO, and HC increased for the fourth unit, which had malfunctioned due to
improper tank filling.
9. Smoke vs Particulate. Bacharach smoke readings taken at different points in
the firing cycle typically showed that smoke density was greatest during the
first minute and reduced as the cycle proceeded. A continuous-tape smoke
indicator confirmed that higher levels of smoke frequently occurred on start-
ing and shutdown for brief periods.
Smoke readings taken at relatively steady-state conditions during the cyclic
runs were not a good measure of filterable particulate emissions which
included sampling during starting and shutdown transients. Smoke measure-
ments may be more indicative of emissions of fine particulates, which are the
principal concern in health effects and atmospheric visibility. Additional labo-
ratory investigations are planned to examine the relationship between smoke
-------
1-13
and particulate for steady-state and cyclic conditions over a range of excess-air
settings.
10. Suggested Emission Factors. Emission factors for use in emission inventories
and implementation plans were suggested on the basis of this investigation and
are shown in Table 1-3. Suggested emission factors are higher than EPA-
published factors for CO and NOX, but are substantially lower for HC and
filterable particulates.
-------
1-14
COMMERCIAL BOILERS
The 13 commercial boilers sampled in Phases I and II were selected to be generally
representative of the commercial boiler population in terms of boiler type, capacity, and burner
type. Boiler sites were selected with the aid of a committee of the American Boiler Manu-
facturers Association (ABMA) to ensure that boiler toad and burner adjustment conditions could
be controlled. For Phase II, another selection criterion was the feasibility of firing as many as
four different fuels in a given boiler; these fuels included natural gas, No. 2 heating oil,
conventional grades of residual oil, and a low-sulfur residual oil (1 percent sulfur) that was used
as a reference fuel and transported from site to site in a tank truck. A total of 33 boiler/fuel
combinations were sampled.
Mix of Boilers
Table 1-4 summarizes the mix of commercial boilers covered in this investigation for both
Phases I and II.
Table 1-4. Mix of Commercial Boilers in Sample
By Boiler Type
By Boiler Capacity
(Boiler Horsepower)8
By Oil-Burner Typeb
Scotch firetube
Firebox firetube
Cast iron
Watertube
10-100 BHP
101-300 BHP
301-600 BHP
Air atomizing
Pressure atomizing
Rotary atomizing
Number
of Boilers
8
2
2
1
5
6
2
8
4
1
a Boiler capacity rating; one boiler horsepower is equivalent to
approximately 33,500 Btu/hr output.
Seven of these boilers had dual-fuel burners capable of firing natural gas.
-------
1-15
Statistics developed in a special survey conducted for this program in cooperation with
ABMA confirmed that the Phase II sample was reasonably representative of the current field
population of commercial boilers. Detailed data from the survey are presented in Appendix B.
Conditions and Fuels Investigated
All measurements on the commercial boilers were made at steady-state conditions.
Gaseous emissions and smoke were measured in Phase II while the boiler was operated at four
different loads and at about Five excess-air levels at each load. Nominal loads were full, 80
percent, 50 to 60 percent, and low-fire setting, which was generally from 30 to 40 percent.
Because of the long sampling time required at each setting, particulate emissions were measured
at only about five selected points for each boiler. (Phase I measurements generally consisted of
gaseous emissions and smoke measurements, at four loads and particulate emission measurements
at two points.)
Base-Line Condition
To provide a common basis for comparing emission levels from various boilers, a
"base-line condition" was defined after consultation with ABMA to identify a condition typical
of boiler operation most frequently encountered in the field. This base-line condition was
defined nominally as 80 percent load with an air setting to achieve flue-gas compositions of 12
percent CO2 for oil firing and approximately 10 percent CO2 for gas firing. This corresponds to
the following excess-air levels: approximately 28 percent for No. 6 oil, 25 percent for No. 2 oil
or light residual oil, and 15 percent for natural gas.
Measurements of gaseous, and particulate emissions and smoke were made for this
base-line condition on all boilers in Phases I and II.
Fuels
Table 1-5 identifies the fuels fired in each of the boilers for both Phases I and II. The
fuels normally used as "house fuels" are identified. LSR refers to the low-sulfur residual oil used
as a reference fuel in Phase II; a similar fuel was fired in one boiler during Phase I.
Typical Emission Trends
Each of the commercial boilers generally exhibited the expected pattern of emission
characteristics, with smoke sharply increasing as low excess-air settings were approached from the
normal operating range. Figures 1-6 and 1-7 illustrate typical curves of smoke and NOX emissions
vs excess air for one boiler fired at 80 percent load with three fuel oils and natural gas. Botl
smoke density and NOX levels were strongly influenced by fuel grade. Effects of load on smoke
and NOX were relatively minor. Emissions of CO and HC were consistently low for the
commercial boilers in normal operating ranges.
-------
1-16
Table 1-5. Summary of Fuel Types Fired in Various Commercial Boilers
Boilers for both Phases I and II Listed in order of capacity.
Boiler Horsepower
and Type
40-hp Cast iron
60-hp Scotch
80-hp Scotch
80-hp Firebox
90-hp Cast iron
100-hp Scotch
125-hp Scotch
125-hp Firebox
150-hp Scotch
200-hp Scotch
300-hp Scotch
350-hp Scotch
600-hp Watertube
Burner
Atomizing
Type Phase
Pressure 1 1
Pressure I
Air I
Pressure II
Air II
Air II
Rotary I
Pressure I
Air I
Air I
Air II
Air I
Air II
Fuels Fired"
No. 2 No. 4 No. 5 No. 6 LSRb Gas
A _ _ _ _
A _ _ _ *
_ _ A _ _ _
. _ * _ e
* _ _
_ _ *
A _ _
*___
* « BC
_ A _ _ _ _
- - * m .
_ A _ _ _ _
- - *
a Legend: ^ House fuel oil normally used.
Alternate house fuel.
Low-sulfur residual oil (Phase II reference oil),
" Low-sulfur residual oil commercially available on East Coast (1.0 percent sulfur content).
c A low-sulfur residual oil (0.5 percent sulfur) was fired in this unit in addition to the 1.0 percent sulfur oil.
Effects of Fuel
Significant effects of fuel properties were observed in the investigation, especially as
related to paniculate and NOX emissions.
Particulate Emissions, Particulate emissions and smoke increased with increasing fuel-grade
designations when several fuel oils were fired in the commercial boilers at a given excess-air
settingpthese emissions were consistently low when firing natural gas.
The effect of fuel grade on filterable particulate emissions is summarized in Figure 1-8 for
all boilers from Phases I and II firing oil at base-line conditions. Particulate is plotted against API
gravity as an indicator of burning characteristics and fuel grade - the heaviest fuels having lowest
API gravity. Approximate ranges of API gravity for different fuel grades are shown.
-------
$
o
E
to
u
O
o
s^
o
o
CO
10 20 30 40 50
Excess Air, %
50
70
Figure 1-6. Typical Characteristics of Smoke Versus
Air for a Commercial Boiler Firing Different
Fuels at 80-Percent Load
JQ
E
LU
x
O
0.5
0.4
0.3
0.2
O.I
No. 6
No. 2
Gas
I
I
10 20 30 40 50
Excess Air, %
60 70 80
Figure 1-7. Typical NOX Emissions Versus Excess Air
for a Commercial Boiler Firing Different Fuels
at 80-Percent Load
-------
IOO
O
P 75
_
.a
o
50
f*- No. 5
No. 4
No.2-~|
Fuel Oil Grades
No. 6,
Mo. 5
15
20 25 30 35 40
API Gravity
Figure 1-8. Relation of Filterable Particulate and
API Gravity for the Commercial Boilers
Firing Different Fuels at Base-Line
Conditions Phases I and II
c
tu
D>
b
"cE
Q.
ro
o
o
"in
en
"E
UJ
x
O
450
400r-
3501
300
250 H
o C2002
x C2003
+ C2004
C2005
C2006
200 f-
0.1 0.2 0.3 0.4
Nitrogen in Fuel, percent
Figure 1-9. Relation of NOX Emissions and Fuel Nitrogen
for the Commercial Boiler Firing Different
Fuels at Base-Line Conditions Phase II
-------
1-19
Ash content tends to be higher for fuels of low API gravity but is not sufficient to
account for higher particulate levels with heavier fuels. The band of ash content for the fuels in
this investigation is shown in Figure 1-8.
The 1-percent sulfur residual oil used as a reference fuel was closer in performance to a
No. 4 or No. 5 oil; it yielded filterable particulate levels about equal to those from No. 4 oil and
only one-third of those from the No. 6 oil.
Effect of Fuel Nitrogen on NOX
Figure 1-9 shows the strong effect of fuel nitrogen on NOX emissions from the commer-
cial boilers in which several fuel oils were fired at base-line conditions. The zero-nitrogen
intercept is indicative of NOX formed by thermal fixation, which is dependent upon flame
temperature and other combustion parameters influenced by boiler/burner design. The slope of
the curve of NOX versus fuel nitrogen reflects the conversion of fuel nitrogen to NOX.
Both the thermal component of NOX and the conversion of fuel nitrogen to NOX vary
from boiler to boiler. The equation that best fits the data from boilers in which fuels of
different nitrogen content were fired is NOX (ppm at 3% O2) = 97 + 420-N0-6. This indicates that
about 62 percent of the fuel-bound nitrogen was converted to NOX for a 0.2 percent nitrogen
fuel, 47 percent for a 0.4 percent nitrogen fuel, and 38 percent for a 0.7 percent nitrogen fuel.
Other investigators have reported conversion of fuel nitrogen to NOX in the same range.
It should be pointed out that other fuel properties, such as gravity and viscosity, also
varied from fuel to fuel in this program; thus, the NOX levels shown in Figure 1-9 may hiclude
effects of factors other than fuel nitrogen.
Summary of Oil-Fired Commercial Boiler Emissions
Emissions for the commercial boilers firing oil at the base-line conditions are summarized
in Table 1-6. Mean emission values are shown separately for each grade of fuel; the mean fuel
properties shown for each grade of fuel pertain to the fuels fired in this field investigation.
(Since broad ranges in properties exist within each grade, these values cannot be considered to be
representative of all fuels within these grade designations.)
Suggested Emission Factors for Oil-Fired Commercial Boilers
This investigation has included emission measurements on 13 commercial boilers with 26
boiler/fuel-oil combinations and with a variety of operating conditions. The emission levels
measured in this investigation show sufficient effects of fuel properties to justify the conclusion
that emission factors intended for use in compiling emission inventories or in evaluating control
strategies should discriminate between fuel-oil grades where these are known.
Accordingly, Table 1-7 presents suggested emission factors by fuel grade.
-------
1-20
Table 1-6. Summary of Emissions from Commercial Boilers
Mean emissions by fuel-oil grade measured in Phases I and 11 at base-line
conditions3 for 13 boilers ranging in capacity from 40 - 600 boiler hp.
Fuel
Oil
Grade
No. 2
No. 4
No. 5
No. 6
LSRb
Number of
Boilers
8
3
3
5
6
API
Gravity,
mean
35
22
19
16
23
Sulfur
%.
mean
0.2
1.7
1.8
1.9
1.0
Viscosity
SSU @ 100 F,
mean
34
120
220
3600
400
Smoke
No.,
mean
0.9
2.6
3.0
3.9
2.9
Mean Emission Factors, lb/1000 gal
CO
0.5
0.8
1.9
1.1
0.3
HC
0.15
0.15
0.22
0.30
0.27
NOX
18
73
54
78
49
SO2C
132-S
149-S
175-S
173-S
157-S
Filterable
Paniculate
1.5
7.1
T3.
38.
13.
a Base-line operating conditions: 80 percent load and air adjustment for 12 percent COj flue-gas concentration.
b Low-sulfur residual oil, typical of 1.0 percent sulfur residual oil being .marketed on the East Coast in 1971 - 1972.
0 S = multiplication factor equal to percent sulfur content of fuel.
These values were derived from the findings of this investigation in such a way as to
normalize the effects of fuel properties for the different grades of fuel that were "encountered"
in the field during the investigation but may have been atypical of that grade on a national basis.
The procedure used to establish the suggested values in Table 1-7 was to plot CO, HC, and
participate emission factors for each boiler/fuel combination at base-line conditions vs API
gravity as an indicator of fuel characteristics. (Although not a universal index of emission
performance, API gravity has the practical effect of reflecting some of the other important
factors and is one of the most commonly available fuel properties.)
"Best curves" were fitted by a least-squares technique; Figure 1-8 shows the curve for
particulates. Emission factors were then obtained from the curves by assuming API gravity values
typical of the various grades of fuel as follows:
Distillate Oil
Conventional Residual Grades
Low-Sulfur Residual (1.0% S)
Fuel Grade
No. 2
No. 4
No. 5
No. 6
LSR
Typical
API Gravity,
degrees @ 60 F
34
22
17
14
23
-------
1-21
Table 1-7. Suggested Emission Factors for Oil-Fired
Commercial Boilers and Comparison
With EPA-Published Factors
Fuel
Oil
Grade3
Emission Factors, lb/1000 galb
CO HC NOXC SO2d
Filterable
Paniculate
Suggested Emission Factors: Based on Battelle/API/EPA Investigation5
No. 2
No. 4
No. 5
No. 6
LSRf
0.5
0.9
1.1
1.2
0.9
0.17
0.24
0.28
0.30
0.23
20 + 78-N0-6
20 + 85-N0-6
20 + 87-N° 6
20 + 89-N0-6
20 + 84-N0-6
142'S
154-S
159-S
162-S
153-S
1.2
14.0
27.0
36.0
12.0
Current EPA-Published Emission Factors3
Distillate
Residual
0.2
0.2
3.0
3.0
40 to 80
40 to 80
142-S
157'S
15.0
23.0
a These values are based on mean emission data for the identified fuel grades
having typical API gravity as follows: 34 degrees API for No. 2; 22 for No. 4;
17 for No. 5; 14 for No. 6; and 23 for LSR. Where actual API gravity is
known, interpolated values should be used.
To convert to emission factors in lb/10^ Btu, multiply these values by 0.0069.
(The actual multiplier varies slightly with fuel grade, being about 0.0071 for
No. 2 fuel oil and 0.0066 for No. 6 fuel oil.)
c N = multiplication factor equal to percent nitrogen in fuel oil. If concentration
is unknown, the following values are suggested: 0.01 percent for No. 2;
0.2 percent for No. 4; 0.3 percent for No. 5; 0.4 percent for No. 6; and
0.2 percent for LSR.
d S = multiplication factor equal to sulfur content in fuel.
e Assuming steady-state base-line operating conditions: 80 percent load and air
adjustment for 12 percent CC>2 flue-gas concentration.
LSR: low-sulfur residual oil (1.0 percent S).
-------
1-22
These typical values were estimated considering values suggested by API representatives and
average values from the field fuels. The resulting emission factors are shown in Table 1-7 as
"suggested emission factors".
Suggested NOX emission factors are based on fuel nitrogen content and reflect decreasing
conversion of the fuel nitrogen to NOX as fuel nitrogen levels increase. Emission factors
recommended for SO2 are proportional to sulfur content of the fuel.
For comparison, Table 1-7 also shows the current EPA-published emission factors, which
distinguish between only two categories of fuel oil: distillate and residual oil.
Summary of Gas-Fired Boiler Emissions
and Suggested Emission Factors
Table 1-8 summarizes emissions as measured from seven commercial boilers firing natural
gas and gives suggested emission factors. The major changes suggested are to reduce emission
factors for HC and filterable particulates.
Table 1-8. Suggested Emission Factors for Gas-Fired Commercial Boilers
With Comparison of EPA-Published Factors
Emission Factors, lb/106 cu ft3
Filterable
CO HC NOX S02 Particulate
Mean Emission From 7 Boilers:
Battelle/API/EPA
Investigation15 16.7 3.7 105 Nil 5.7
Suggested Emission Factors3 20.0 4.0 100 0.6 6.0
Current EPA-Published
Emission Factors3 20.0 8.0 100 0.6 19.0
3 To convert to emission -factors in lb/106 Btu, multiply these values by 0.00098.
b Base-line Operating Conditions: 80 percent load and air adjustment for approximately 10 percent
CO2 flue-gas concentration.
-------
1-23
CONCLUSIONS FOR COMMERCIAL BOILER INVESTIGATION
Major conclusions from the findings of this investigation of emissions from a representa-
tive sample of commercial boilers are outlined below, based on the combined results of Phases I
and II.
1. Emission Characteristics of Commercial Boilers. The varied-air runs with the
commercial boilers showed that the smoke curve generally was flat over a
range of excess-air levels but increased sharply as the combustion air was
reduced below normal air settings. For oil firing, smoke levels tended to
increase before CO and HC increased. With gas firing, CO and HC increased
before the smoke density increased.
To adjust the air setting to acceptable smoke (or minimum smoke for that
boiler/fuel combination) would require combustion air settings as high as 30
to 40 percent excess air for most of the boilers operating with the heavier
fuels. It was possible to fire gas at 10 percent excess air without exceeding
No. 1 smoke; however, 20 percent excess air was required to avoid rapidly
increasing CO emissions with gas. Smoke increased slightly at high excess air
in only a few cases with oil firing and in one case with gas firing.
These observations suggest that field-type instruments for measuring smoke
and CO2 can be used satisfactorily in adjustments for oil firing to minimize
CO and HC as well as smoke; however, a CO indicator is needed in air
adjustments for gas firing. (CO and HC levels at base-line conditions were
generally low for all boilers firing either oil or gas.)
NOX was relatively insensitive to air setting.
2. Effect of Load. Load generally had little influence on emission levels, including
NOX. Smoke and particulate increased with load on several boilers when firing
No. 6 oil, suggesting that limitations in mixing or combustion volume were
being reached at high load. (For minimum smoke at low fire, some boilers
require a higher excess-air setting at this point, and this adjustment is fre-
quently made in the field when setting the control linkage.)
3. Effect of Boiler and/or Burner Design on Emission Levels. Considerable
variation in emissions was noted from boiler to boiler, even when firing the
same fuel at base-line conditions. Particulate emissions with the reference fuel
varied by a factor of 4.4 from lowest to highest. Smoke varied only by a
factor of 1.4. Emission levels of CO and HC were judged to be sufficiently
low to be insignificant for all oil firing.
NOX emissions for various boilers varied from lowest to highest by a factor of
1.3 at base-line conditions when firing the reference fuel. Boiler and/or burner
design variables did not show a consistent influence on NOX emissions, even
when firing low-nitrogen fuels.
-------
1-24
No attempt was made in this investigation to optimize emission performance
by modification of the burner design or atomizer; thus, the data are not
adequate to indicate what minimum emission levels might be achieved by
optimizing burner design for a given boiler/fuel combination.
4. Effect of Fuel Oil Properties on Smoke and Paniculate. Fuel oil grade had an
important bearing on smoke and particulate levels. For the boilers operating at
base-line conditions, the heavier fuels (i.e., lower API gravity) yielded higher
smoke and particulate, although the heavier grades were fired at conditions
used in normal practice.
A regression analysis indicated that the single most important fuel property
influencing filterable particulate at base-line conditions was carbon residue
(correlation coefficient of 0.68). API gravity also had a significant effect
(correlation coefficient of 0.55). Viscosity at firing temperature was relatively
insignificant (correlation coefficient of 0.25).
An index combining carbon residue, viscosity at firing temperature, carbon
content, and API gravity yielded a good correlation with filterable particulate
(correlation coefficient of 0.86).
The low-sulfur residual oil having 1-percent sulfur was closer to a No. 4 or 5
grade in viscosity, gravity, and other burning characteristics than to conven-
tional No. 6 oil. Particulate emissions with this fuel averaged one-third those
of the conventional No. 6 oil.
NOX levels tended to be higher with the heavier fuels, but this is traceable to
generally higher fuel-bound nitrogen with the heavier grades.
5. Effect of Fuel Nitrogen on NOX. NOX emissions increased nearly linearly with
increasing fuel nitrogen content when different fuel oils were fired in the same
boiler. The slope of NOX versus fuel nitrogen curves indicated conversion of
about 40 to 60 percent of the fuel nitrogen to NOX.
A combination of other factors influenced thermal components of nitrogen
fixation, but a consistent pattern was not detected; firing rate did not have as
consistent an influence in Phase II as observed in Phase I.
6. Gas Firing vs Distillate Oil, Smoke and particulate emission levels with gas
firing were generally lower than with No. 2 oil when firing the same boiler at
base-line conditions. On a Btu-input basis, mean particulate emissions at
base-line conditions for gas firing were one-third of those for No..2 oil.
Mean NOX levels with gas firing were three-fourths of those for No. 2 oil. For
half the boilers, gas fking yielded slightly higher NOX levels than did No. 2 oil
at base-line conditions.
CO and HC emissions were higher for gas than No. 2 oil: CO over five times
higher, and HC about three times higher.
-------
1-25
7. Correlation of Smoke vs Particulate. For the commercial boilers operating at
steady-state conditions, Bacharach smoke numbers showed consistent trends in
relationship to filterable particulate when firing different fuels in the same
boiler. This suggests that families of correlation curves may be possible, with
each curve characteristic of a given boiler firing a given fuel. If this relation
could be further defined for a series of similar boiler designs, predictions of
particulate performance could be made using simpler field measurement tech-
niques.
8. Suggested Emission Factors. Because of the large differences in emission levels
in commercial boilers due to the effects of fuel properties, separate emission
factors should be used for different grades of fuel oil in emission inventories
or implementation plans. Suggested emission factors were established by fuel
grades using the trends identified in this investigation; these are shown in
Table 1-7. Suggested emission factors for boilers firing natural gas are shown in
Table 1-8.
Other than to account for the fuel-oil grade discrimination, the major sug-
gested changes of EPA-published values are reductions in emission factors for
HC and particulate from firing No. 2 oil. In addition, it is suggested that
emission factors for NOX be based partly on fuel nitrogen content. Suggested
changes for natural gas firing are reductions in emission factors for HC and
particulate.
-------
II-l
EMISSIONS FROM RESIDENTIAL UNITS
This chapter describes the Phase II investigation of emissions from residential oil-fired
heating units. The principal effort of Phase II was devoted to more detailed field measurements
of emissions covering more conditions than was possible for the Phase I study. The discussion is
presented as follows:
Units Included in the Phase II Investigation
Procedures Used in the Field Investigation
Emission Results for Cyclic Runs
Emission Results for Varied-Air Runs
Correlation of Smoke and Particulate.
Where applicable, results of Phase I emission measurements have been included in the analyses of
the influence of various factors on emissions.
UNITS INCLUDED IN THE PHASE II INVESTIGATION
The scope of equipment in the Phase II investigation included 13 oil-fired residential
heating units with firing rates from 0.6 to 1.75 gph.
Basis for Selection of Equipment
Selection of residential oil-burning equipment installations to be studied in Phase II of
this program was directed toward making the total sample of the Phase I and II units reasonably
representative of the population of equipment in service, considering such factors as
Burner type
atomizer type and combustion head
Capacity or firing rate
Installation type
matched unit (burner-boiler unit or burner-furnace unit) or conversion
burner
Heating system type
hot water, steam, or warm air
Age of installation.
Obviously, within the limits of a program of this size, it was not possible to include a large
number of units in each category. However, an attempt was made to include a distribution of
oil-fired units similar to the distribution of units now in service.
As an initial step in the selection of the residential units to be included in the Phase II
investigation, a questionnaire on the existing oil-burner population was submitted to a number of
-------
II-2
burner service organizations, Replies to the questionnaires were slow in being returned, hence,
the Phase II investigation was launched using previously published oil-burner population data for
selection of residential units. Sources of information on oil-fired equipment that were consulted
included Fueloil & Oil Heat magazine and the National Oil Fuel Institute. This information,
combined with the experience background of the project team, was used in establishing a
suitable mix of units. The Battelle-Columbus recommendations as to the general equipment mix
were then examined and approved by the EPA Project Officer and the API SS-5 Task Force.
Results of the questionnaire confirmed that the equipment mix selected was representa-
tive of the existing equipment population.
Selection of Equipment Mix and Individual Units
Table II-l outlines the types of equipment included in the residential equipment mix and
shows the number of units of each type included in this investigation for Phase II and the total
for the two phases. The selected mix of units is compared with the distribution of the U.S.
equipment population.
Background data obtained from the questionnaires on residential equipment mix are
shown in Appendix A. These data represent the best statistical data available covering detailed
categories of equipment now in service on a national basis. High-pressure gun-type burners are
clearly the dominant burner type for residential heating. Also, burners of 1.35 gph or less
account for nearly 70 percent of the residential heating units.
Individual residential field units were selected by the Battelle field team with the
assistance of Consultant W. H. Axtman and with the aid of a qualified oil-burner servicing
organization. The majority of the units investigated were selected from the service contract files
of the organization, and the initial contact with the homeowner was made by the servicing
organization. The servicing organization supplied a skilled serviceman to tune or adjust the
burners following the initial measurements.
Description of Residential Units
Table II-2 shows the number designations for the residential units which are used to
identify emission data elsewhere in this report for the specific residential units. This table also
includes a brief identification of burner type, firing rate, heating system type, and burner and
system age - including whether the installation is (1) a "matched" factory-designed burner-
furnace or burner-boiler unit or (2) a "conversion" installation where the burner has been added
in the field to a basic furnace or boiler, possibly originally designed for coal firing. (Nearly all
furnaces and boilers being installed today are matched units.)
Twelve of the units (Units 23 through 34) were field units in use for residential heating.
An additional unit, identified as Unit 35, was a new residential oil-fired furnace which was set up
in the laboratory at Battelle-Columbus during the Phase I study for the calibration and checkout
of instruments, sampling procedures, and the operating-cycle timing. All thirteen of the units
were high-pressure gun type; five had flame-retention combustion heads and one had a Shell
combustion head.
-------
II-3
Table 11-1. Mix of 31 Residential Units in Sample Compared With Distribution
of U.S. Equipment Population in Service
Units in Sample
By Burner Type
High-Pressure Gun
Low Pressure
Vertical Rotary
Wall-Flame
Vaporizing
By Burner Capacity, 9 gph
<1.0
1.01-1.35
1.36-1.65
1.66-2.00
2.01-3.0
>3.0
By Burner Combustion-Head Type
(High-Pressure only)
Conventional Head
Flame- Retention Head
Shell Head
By Burner Age, years
<5
6-10
11-15
16-20
>20
By System Type
Furnace
Boiler
Water Heater
Number
Phase II
13
0
0
0
4
7
1
1
0
0
7
5
1
9
0
1
3
0
9
4
0
of Units
Tota^
29
1
1
0
6
15
2
1
5
2
18
9
2
17
1
7
3
3
12
18
1
Percent
of Distribution of U.S. Equipment
Totala
94
3
3
0
20
49
6
3
16
6
62
31
7
55
3
22
10
10
39
58
3
Population, percent11
84
10
5
1
34
35
14
8
5
4
81
11
8
19
23
28
19
11
52
48
-
Including units from both Phases I and II.
From References 4 and 5 as follows:
4 Fueloil and Oil Heat, "The Typical Oil Burner", Vol. 31, No. 6 (June 1972), pp 44-45.
5 Fueloil and Oil Heat, Special Study, Vol. 30, No. 1 (January 1971), pp 22, 24.
These statistics are summarized in more detail in Appendix A.
-------
11-4
Table 11-2. Description of Residential Units for Phase II
All units had high-pressure gun burners.
Burner Head
Unit Type
23
24d
25
26e
27
28
29
30
31
32
33
34
35
a
b
c
d
e
Flame retention0
Conventional
Flame retention0
Shell
Flame retention
Flame retention
Conventional
Conventional
Conventional
Conventional
Flame retention
Conventional
Conventional
Firing
Rate3,
gph
1.35
1.00
1.35
1.75
1.35
1.00
0./5
0.60
1.50
1.00
0.85
0.75
1.00
Heating-System Type11
Cl boiler, conversion, water
Steel forced-air furnace unit
Steel boiler unit, water
Cl boiler, conversion, steam
Steel boiler unit, steam
Steel forced-air furnace unit
Steel forced-air furnace unit
Steel forced-air furnace unit
Steel forced-air furnace unit
Steel forced-air furnace unit
Steel forced-air furnace unit
Steel forced -air furnace unit
Steel forced air furnace unit
Nominal firing rate or nozzle capacity as found, gph.
"Unit" describes "matched" burner-furnace or burner-boiler units engineered by
construction.
High-speed fan motor (3450-rpm).
Unit 12 from Phase I study.
Unit 16 from Phase 1 study.
Combustion
Chamber
Material
Firebrick
Metal
Ceramic felt
Firebrick
Hard refractory
Metal
Hard refractory
Hard refractory
Firebrick
Metal
Firebrick
Hard refractory
Ceramic felt
the manufacturer. Cl
Estimated
Burner
<1
4
<1
2
5
2
18
18
20
<5
3
11
2
Age, yr
System
21
4
<1
9
5
8
18
18
20
<5
15
11
2
denotes cast-iron
Nine of the Phase II oil-fired units were warm-air furnaces and four were water or steam
boilers. Only two of the thirteen units investigated this year were conversion units. The ages of
the burners and heating systems ranged from less than 1 to about 20 years.
PROCEDURES USED IN THE RESIDENTIAL
FIELD INVESTIGATION
Measurements on residential heating units were started in December, 1971, and, except
for the last follow-up runs, were completed in February, 1972. The investigation of each
residential unit required from 2 to 3 days, including instrument setup and measurements under
the test conditions. The measurements were made by a three-man field team supported by other
Battelle-Columbus staff and a consultant.
Burner Conditions Investigated
Emissions from residential units were monitored during cyclic operation (repeated cycles
of 10 minutes on and 20 minutes off) and during steady-state operation while the excess air was
varied. Both cyclic and varied-air runs were made at three sets of burner conditions:
-------
II-5
A "As found condition" of burner adjustment using house fuel
T "Tuned condition" using house fuel
R "Reference fuel" run under the same conditions as the tuned run to
provide a base line for comparison of units.
The letter designations (A, T, and R) are used elsewhere in this report to key the condition of
the runs.*
Definition of Tuned Condition. Experience on the first few of the Phase I units had
shown variability among servicemen in their criteria for properly adjusted units. For the
remaining Phase I and II units, the tuned condition was defined as the best adjustment (in terms
of the smoke-CO2 relationship) that could be achieved by a skilled serviceman with normal
cleanup, nozzle replacement, simple sealing, and adjustment procedures with the benefit of field
instruments. It did not include major repairs, modernization, or replacement of major parts that
would require special charges to the homeowner (e.g., replacement of a combustion chamber).
Further details of the adjustment procedure are presented in Appendix D.
Influence of Tuning on Smoke and CO2. Figure II-l shows the influence of tuning on
smoke and CO2 levels for the Phase II units, with unit numbers identified. (Unit 35, the
laboratory unit at Battelle, is not included in discussions of tuning, as it had no normal as-found
condition, but is treated as a tuned unit in data tabulations.) Due to the generally low smoke
levels in the as-found condition, tuning reduced smoke in only 6 of the 12 units. In 9 of the 12
units the CO2 setting was increased.
Figure II-2 shows the distribution of smoke and CO2 levels for the 12 residential units
included in the Phase II investigation together with the 20 oil-fired residential units included in
the Phase I investigation. For the Phase II units only, 75 percent operated with a No. 2 smoke or
less in the as-found condition and all units operated at No. 2 smoke or less in the tuned
condition. The mean smoke level for the 12 Phase II units was 1.5 for the as-found condition
and 0.6 for the tuned condition. The smoke levels measured during the Phase II investigation for
both the as-found and tuned conditions were significantly below smoke levels measured during
the Phase I study.**
The average CO2 was 7.9 percent for the as-found and 8.5 for the tuned condition for
the 12 Phase II units that were tuned. The plot of CO2 values for the as-found condition for Phase
II data were in a narrower range (6.1 to 9.4) than for Phase I data (3.8 to 12.5); 75 percent of
the Phase II units had CO2 levels of 7.9 to 9.9 in the tuned condition.
"The cold start condition investigated in the Phase I procedure was not included in Phase II because the gaseous
emissions measured for these runs during Phase I did not differ appreciably from the emissions obtained after
repeated cycles.
**Extensive inquiry of, the field team and study of the smoke spots and data have not revealed why the Phase II
smoke data were lower than Phase I data.
-------
II-6
c 8
S; 6
_Q
E
o 4
to
.c
o
_
o
a
CD
Legend
As-found
o Tuned
C02 in Flue Gas, percent
Figure II-1. Operating Conditions for 12 As-Found and Tuned Residential Units
Reference Fuel Condition. To obtain a base line or reference for the variety of burner
units, gaseous emissions and smoke were measured for each residential unit during cyclic and
varied-air operation while firing a reference fuel in the tuned condition. This reference fuel was
selected as a high-quality No. 2 hydrotreated fuel.
Properties of all fuels fired in the residential units are tabulated in Appendix C, Tables
C-l and C-2.
Operating Conditions Investigated
Emission measurements were made while the residential units were fired at two different
operating conditions or types of runs, as follows:
1. Cyclic Runs - Measurements were made during repeated cycles of 10
minutes on/20 minutes off.
2. Varied-Air Runs - Measurements were made during steady-state operation
for a range of excess air-settings and plotted against CO2 using procedures
similar to ASTM D2157-656.
6 "Standard Method of Test for Effect of Air Supply on Smoke Density in Burning Distillate Fuel" ASTM
D2157-65(70).
-------
Oily
As found, Phase I
As found, Phase n
Tuned, Phase I
Tuned, Phase TJ
oooo
000
OAO
, AA
ooooo
0~ "20 40 60 80 lOO
Percent of Units With Smoke Less Than or Equal to Stated Value
o
u
0
Legend
As found, Phase
As found, Phase
Tuned, Phase I
Tuned, Phase TJ
AOA
>O.A*
A. °,2 20
_L
I
"0 20 40 60 80 100
Percent of Units With COj Equal to or Greater Than Stated Value
Figure II-2. Distribution of Smoke and CO2. Levels as Measured in As-Found and
Tuned Residential Units for Phases I and II
-------
II-8
Measurements of gaseous emissions and smoke were made for both cyclic and varied-air runs for
the as-found, A, tuned, T, and reference-fuel, R, conditions. Particulate emission measurements
were limited to cyclic runs.
Selection of Operating Cycle. The same operating cycle was used for cyclic runs in
residential units during Phase II as was used during Phase I to provide a consistent operating
mode for emission measurements. The cycle, 10 minutes on and 20 minutes off, was based on the
following criteria:
1. Emission data from different units could be compared on a common basis
independent of the effects of outdoor weather on operating modes.
2. The field program would not be seriously delayed or limited by warm
weather during the normal heating season. (Normal operation during
warm weather produces very infrequent firing.)
3. Multiple cycles of controlled operation would allow the burner and
combustion chamher to reach a repeatable thermal condition.
4. The on-period would be much longer than the response time of the
monitoring equipment and long enough to allow gaseous emissions to
reach equilibrium values.
5. A 10-minute on-time is longer than will be encountered with direct
burner control by a modern heat-anticipating room thermostat, where 5
cycles per hour at 50 percent on-time is the design basis but may be
shorter than encountered for a hydronic system controlled by the tem-
perature of a water circuit or by steam pressure. A choice of 10 minutes
is a reasonable compromise.
6. The 1/3 operating time or "load" of the 10-minutes-on/20-minutes-off
cycle represents a reasonable average load condition during the colder part
of a heating season.
The 10-minutes-on/20-minutes-off cycle is equivalent to 2 cycles per hour at 33 percent
on-time. This cycle is typical of the average recorded for four units observed during a period of
5 months during the 1971-1972 heating season. The operational data for these four units are
presented under "Measurements on Follow-Up Units" (page H-36).
A limited investigation was made of the effect of cycle length on emissions. These data
are reported under "Experiments on the Effect of Cycle" (page 11-40).
Varied-Air Operation, Gaseous emissions and smoke were measured at each of the three
burner conditions (A, _T, and R), while the excess air level was decreased from the full-open
setting to a setting producing about a No. 6 or 7 smoke. These data are intended to provide
information about the sensitivity of emissions from the burners as related to excess-air adjust-
ments in each burner condition.
The procedure for the varied-air runs was similar to ASTM D2157-65(70)6. Steps were
as follows:
-------
II-9
1. The burner air gate was set at the full-open position and emissions were
measured.
2. The air gate was closed and smoke was monitored until the knuckle of the
smoke curve was reached (the point at which smoke increased sharply
with small decreases in excess air), then emissions were measured.
3. Emissions were measured at several points between Points 1 and 2 by
setting the air gate at intermediate positions.
4. The air gate was closed until a No. 5 to 7 smoke was obtained and
emissions were measured.
5. Emissions were measured at one ,or more points between Points 2 and 4.
Emissions were measured at about six points for each burner condition. Measurements were
made after the furnace had been firing for at least 30 minutes and had operated at the test
condition for at least 10 minutes or after the gaseous emission readings became constant.
-Follow-Up Measurements. Follow-up measurements were made on four residential units.
Three sets of measurements were made for these units at the following times:
Initial visit mid-December
1st follow-up visit mid-February
2nd follow-up visit late April.
The complete set of measurements, including gaseous and particulate emissions for the A, T, and
R conditions, were made during the initial visit. For the 1st follow-up visit, only gaseous
emissions were measured (firing both house and reference fuels). During the 2nd follow-up visit,
gaseous emissions were measured for house and reference fuels and particulate emissions were
measured for the reference fuel only. All follow-up measurements were made during cyclic
operation of the unit.
Emission Measurements Instruments and Techniques
Prior to measurements in the field, all the equipment and instruments required for the
investigation were standardized and checked out in several runs on a residential oil-fired furnace
at Battelle-Columbus. Samp ling procedures and test sequences were also checked out at this time.
An established set of; operational procedures was routinely followed for each unit
investigated. Stack gases were sampled and supplied directly to continuous monitoring equipment
set up alongside the heating unit.
Gaseous Emissions. Measurements were made of the following gaseous emissions under
various conditions of operation using the methods noted as follows (Details of instrumentation
and procedures for measuring gaseous emissions are described in Appendix E.):
-------
11-10
Emission Measurement Method
C02 NDIR (nondispersive infrared) and
Fyrite
O2 Amperometric and Fyrite
CO NDIR
Hydrocarbons (total) .... Flame ionization
NOX Dry electrochemical (used for initial and
1st follow-up visits)
NDIR with converter (used for 2nd
follow-up visit)
NO NDIR
Smoke and Other Combustion Parameters. Smoke was measured using the Bacharach
hand pump smoke meter and the standard procedures adopted by ASTM7. In addition to
these measurements, other combustion conditions were measured, including
Firing rate, measured volumetrically
Stack temperature, measured in the flue at the particulate sampling
location
Stack draft, measured in the flue at the particulate sampling location.
Particulate Emissions. Particulate emissions were sampled using the EPA sampling .rig2-8.
This train is described in Appendix F. A special feature of this sampling train is the inclusion of
two water impingers or bubblers (in an ice bath) downstream from the filter. The impingers were
originally intended to collect any condensable material (at 70 F) that would exist as vapor at
filter temperature and, thus, pass through the filter and also collect any solid particulate that
passes through the filter. Later, EPA abandoned use of the impinger catch in determining
particulate emissions for stationary sources8 as there was indication that reactions occur in
the impinger to generate material that is included in the weight measurement of particulate, even
though.the material does not exist as particulate either in the flue gas or in the atmosphere9.
To insure that the most meaningful information was obtained from the particulate
samples collected on this investigation, the probe wash, the filter catch, and the impinger wash
were treated separately and particulate weights were recorded on each. In this report, particulate
data are reported as filterable (including the probe and filter catches) and total (combining the
filterable and the material found in the water impingers). However, it should be pointed out that
even the filterable catch obtained using the EPA sampling train may not be directly comparable
to the particulate catch obtained using other sampling trains because of differences in the
procedures for washing the probes and differences in sampling rates and volumes.
"Standard Method of Test for Smoke Density in the Flue Gases from Distillate Fuels", ASTM D2156-65(70).
o
"Standards of Performance for New Stationary Sources", Federal Register Vol 36 No 139 Part II pp 15704-
15722, August 17, 1971.
9
"Standards of Performance for New Stationary Sources", Federal Register, Vol. 37, No. 55, Part 1 pp 5767
March 21, 1972.
-------
II-11
Previous experience at Battelle suggests that acetone washing may not be adequate to
remove all particulate from the probe. Therefore, for this investigation two washing procedures
were used: (1) the EPA procedure and (2) a modified EPA procedure referred to as the MEPA
procedure in this report. First, the probe, filter holder, and impingers were washed using
procedures specified by EPA. Then the modified procedure was used wherein additional washings
were made to insure complete removal of deposited particulate. The complete procedures and
the resulting data are discussed further in Appendix F.
Timing of Measurements for Cyclic Runs. Measurements of gaseous emissions were made
continuously over the 10-minute-on period of the cycle. Special tests of the response of the
sampling system and instruments to step changes in pollutant concentrations showed a 90
percent response in less than 72 seconds for all gaseous pollutants. All emission factors were
based on time-average values over the 10-minute-on period, including peaks at starting but not
including peaks at shutdown when the combustion air flow would be diminished and the
measured concentrations would not be meaningful.
Bacharach smoke measurements were made at three points in the firing cycle: at 1, 5,
and 9 minutes. The smoke numbers used in this report are generally for the 9-minutc point in
the cycle, as this is most typical of the smoke that servicemen would normally measure.*
However, in the statistical comparisons including Phase I data the 5-minute smoke data are used,
as only that data were obtained on all Phase I units.
Particulate sampling was conducted during the 10-minutc-on period of the burner for six
consecutive cycles beginning with the second cycle. The particulate sampler was started just
before burner start-up and continued to just beyond shutdown for each cycle.
Emission Transients During Cyclic Huns, Figure II-3 shows typical profiles of gaseous
emission transients during cyclic operation of residential oil burners. Peaks are generally noted
for CO and HC at starting and shutdown when there is transient imbalance of the fuel-air
mixture or flame temperatures are low. For most units, the HC concentration in the flue gas
(after the start-up peak) was less than the ambient HC concentration. For NOX there is some lag
in sampling and NDIR analysis on starting, but the gradual rise in NOX is more likely due to
combustion chamber warm-up.
Smoke at starting and shutdown follows a similar transient effect as does CO and HC. On
ignition, an instantaneous over-rich condition can occur which is compounded by the fact that
fuel nozzle delivery rate is generally highest; the nozzle is relatively cool on starting. On
shutdown, a smoke puff can occur if the fuel pump cut-off is not prompt. Smoke transients at
both start-up and shutdown tend to vary widely from unit to unit.
Emission Data Used for Calculating Emission Factors, Emission factors reported for the
cyclic runs of Phase II are the dose average for all gaseous pollutants. The dose average values
were obtained by dividing the integrated area under the emission curve by the cycle "on"-time.
*ASTM Standard D2157-65(70), "Standard Method of Test for Effect of Air Supply on Smoke Density in Burn-
ing Distillate Fuels", states that the burner should operate for 15 minutes before the smoke measurements are
begun.
-------
11-12
o
o
CO
o
_c
O
O
O
CO
Burner
on time
o
o
cn
o
.c
O
CO,
Burner-
"on" time
8
co
o
J=
CJ
o
jQ
HC
Burner
"on" time
o
CO
, J
i_
o
o
x
O
N0>
Burner
on lime
Time
Time
Figure II-3. Typical Gaseous-Emission Profiles of Residential Units During
10-Minutes-On/20-Minutes-Off Cyclic Operation
-------
11-13
For the Phase I study1, lOth-minute readings were used for data summaries and for
calculating NOX emission factors, because the NOX monitors had slow response. However, the
response of the instruments used for Phase II is sufficiently rapid so that dose average readings
are considered more representative of the overall cyclic operation. NOX emission factors for
Phase I units summarized in this report were corrected to the dose average value (from
lOth-minute data) by multiplying the lOth-minute data by the ratio of dose average data to
lOth-minute data obtained on Phase II units (about 0.96).
EMISSION RESULTS FOR CYCLIC RUNS
The following section contains results.of emission and smoke measurements during cyclic
operation of residential units. Emission data are compared for different burner conditions and
burner and installation design features. Results of measurements during cyclic runs on the four
follow-up units are summarized to show the effect on emissions of operating time since tuning.
In addition, the results of laboratory experiments on the effect of cycle on smoke and gaseous
emissions are presented.
Summary of Emission Data and Emission Factors
Table II-3 provides a comprehensive summary of emission measurements for the cyclic
runs on the 13 residential units in Phase II.* This table includes operational data defining
conditions, measured gaseous emissions in ppm, particulate loading, and calculated emission
factors based on fuel input (Ib of pollutant per 1000 gallons of fuel).
Effect of Tuning and Fuel
Tuning. Table II-4 summarizes the tuning of the 12 Phase II field units, showing the
changes in smoke, CO2, and efficiency and showing qualitative effects on emissions.
All units were'tuned to a smoke level below 2.0, with the average smoke reduced from
1.5 to 0.6. Smoke was reduced for 6 of the 12 units; for another three units, smoke was increased
less than 0.5 while increasing CO2 for higher efficiency; CO2 was increased for nine units, with
the average shifting from 7.9 to 8.5 percent.
Tuning of the residential units generally produced an increase in unit efficiency**. The
average incremental increase in unit efficiency was 1.7 percent, with the efficiency increasing for
seven units, decreasing for three units, and remaining essentially unchanged for two units. The
range of the incremental change in efficiency was from +8.8 percent to -3.9 percent. An
incremental increase in efficiency of 1.7 percent would reduce fuel consumption by about 2.4
percent; hence, emissions would be reduced by the same amount.
*Follow-up measurements for Units 23 through. 26 are included separately in Table 11-10.
**Although the flue gas temperature was not measured at the breeching, as should be done for efficiency calcula-
tions, approximate efficiency calculations were made using the gas temperature at the particulate sampling point.
-------
11-14
Table 11-3. Summary of Emissions end Emission Factors for Cyclic Runs - Phase 11 Residential Units
Unit and
Condition*
23
24
26
26
27
28
29
30
31
32
33
34
35
A
T
R
A
T
R
A
r
R
A
T
R
A
T
R
A
T
R
A
T
R
A
T
R
A
T
R
A
T
R
A
T'
R
A
T
R
T
R
C02.
9.2
9.9
9.4
7.1
8.9
9.2
9.4
9.4
8.5
7.1
8.7
8.8
9.1
8.2
8.1
6.6
7.4
7.7
9.4
7,9
7,7
7.5
7.1
6.9
7.9
9.9
9.6
7.0
8.7
8.6
6.1
6.5
6.6
9.4
9.8
10.1
8.8
93
°2.
%
9.1
7.8
8.2
12.0
9.0
8.5
9.0
7.3
8.3
11.1
9.1
8.8
B.I
9.4
9.5
11.9
10.3
9.8
7.1
9.7
9,1
10.3
11.1
11.4
10.B
7.9
8.0
11.4
9.4
9.3
12.6
12.2
12.1
7.4
8.0
7.3
8.7
7.5
Excess
Air.%
68
54
60
120
71
64
75
56
72
109
73
70
64
81
82
126
97
89
55
36
83
95
108
114
94
56
58
114
76
75
143
131
127
55
55
50
70
57
Smoke No.
at9Min
0.3
0.5
0.4
0.4
0.7
0,8
1.0
1.7
2.0
0.5
0.4
0.6
1.0
0.4
0.3
1.3
0.4
0.4
9.0
0.5
0.3
0,2
0.6
0.3
0.0
0.2
0.3
0.2
O.I
0.1
Oily
Oily
Oily
2,1
1,0
1 9
02
0.2
Slack
Temperature,
°Fb
470
480
480
645
525
520
490
450
455
520
500
490
5SO
620
-
540
650
-
3/0
440
-
570
470
-
530
535
-
545
500
-
700
690
-
440
460
-
440
440
Firing
Rate,
gph
1.20
1.30
1.04
_
0.86
1,36
1.40
1.78
_
1.80
1.26
_
1.38
0.90
_
1.40
0.75
_
0.74
1.10
-
0.70
1.37
-
1.50
0.98
-
1.00
0.85
O.S5
0.60
~
070
0.99
0.99
Emiuion Data at 02 Level,
dose average ppm
CO
3.7
7.7
2.6
222.0
4.1
9.5
10.6
8.6
12.1
5.5
fl.O
10.7
11.6
12,5
11.1
162.0
8.6
3.9
24.7
12.3
13.6
11.7
>170.0
>200.0
36,1
64
47
2920
90,0
44,0
>1240.0
>1240X)
>1 240.0
19.6
14. B
18.2
2.5
3.8
HC
8.3
5.7
13
122
3.1
US
5.4
5.9
5.7
b.l
4.0
4.4
3.3
4.4
7.3
13.1
1.8
4.7
4.4
4.2
25.0
8.2
9.4
10.7
3,8
'2.4
2.4
12,6
2,8
4,2
_
221,0
166,0
5.4
3.9
5.1
5.5
9.6
NO
67
82
75
39
71
69
42
47
44
54
68
67
S6
58
62
31
60
61
31
63
64
63
44
37
61
83
85
31
63
62
9
20
22
62
58
63
77
84
NOX
35
89
79
78
7S
71
53
47
49
64
71
72
86
78
69
79
62
76
83
66
74
79
65
58
68
87
89
73
69
67
87
105
97
62
5G
62
85
91
Particular
Loading,
mg/sm3 e
Filterable
10.7
7.2
6.4
6.1
19.0
18.8
5.7
7.4
7.9
11.1
9.1
10,7
17.0
12.3
-
14.7
7.1
-
31.1
43.9
-
9.6
13.1
-
8.5
15.1
-
9.3
13.7
-
24.3
25.3
-
22.3
14.5
-'
17,7
13,7
Total
20.5
26.0
20.5
15.4
47.6
32.2
25.0
19.4
17.7
33.2
17.6
21.4
39,7
26.1
-
49.0
22.3
-
89.7
64.8
-
34.2
35.9
-
342
74,1
-
468
41,4
-
113.
119.
-
48.7
41.2
-
69 a
4.1.6
Emission Factors, lb/1000 gal
CO
0.61
1.17
0.41
42.28
0.69
1.53
1.83
1.32
2 05
1.13
1.02
1.78
1.39
2.26
1.98
31.6
1.68
1,65
3.78
2.26
2.43
2.26
>32.8
?38.5
6.97
1.00
0.73
62.63
15.84
7 57
>301.0
>285.0
>277 .0
2.96
2,24
2.68
0.42
0.59
HC
0,79
0.50
0.29
1.52
0.30
0.42
0.53
0.52
055
0.60
0.39
0.42
0.31
0.46
0.75
1.68
0.20
0.50
0.39
0.44
2.58
0.91
1.11
1.29
0.42
0.21
0.21
1,54
0,28
0,41
_
27 72
21.23
0,47
0,34
043
053
085
NO,
as NO2
231
22.2
205
27,9,
21.1
188
15.9
11.8
13.7
21.6
19.9
197
23.0
23.1
20.2
29.0
19.9
23.2
20 9
19.6
21.9
25.0
22.0
20.1
21.6
22.2
22,7
25.7
20.0
13.9
34.6
39.6
35.6
15.4
14,4
15.3
23.3
23.1
Paniculate0
Filterable Total
1.55
0.93
088
1.17
2.82
2,64
0.83
0.97
116
1.96
1.35
1,55
2.40
2.02
-
287
1 19
-
4 12
703
1 6?
2.37
1.43
2.04
-
1.80
2.13
-
5.12
5.04
-
2.94
1.9'.
2. 56
1.34
2.94
3.44
2.82
2.93
7.02
453
3.73
2.57
2,61
5.97
2.61
311
5.65
4.09
~
8.41
3.79
-
11 96
10.32
-
572
6.42
5 77
10.04
8 69
6.34
23.92
23.76
6.37
6.40
-
10.05
5.89
3 Units are described in Table 11-2.
bTemperalure measured at sampling point near end of lO-minuta-on period.
c Modified EPA procedure.
-------
Table 11-4. Qualitative Summary of Tuning Effects on Emissions From Residential Units Phase II Units
Smoke Number,
9th minute
Unit
23
24
25
26
27
28
29
30
31
32
33
34
As Found
0.3
0.4
1.0
0.5
1.0
1.3
9.0
0.2
0.0
0.2
Oily
2.7
Tuned
0.5
0.7
1.7
0.4
0.4
0.4
0.5
0.6
0.2
0.1
Oily
1.0
CO2,
percent
As Found Tuned
9.2
7.1
8.4
7.1
9.1
6.6
9.4
7.5
7.9
7.0
6.1
9.4
9.9
8.9
9.4
8.7
8.2
7.4
7.9
7.1
9.9
8.7
6.5
9.8
Change in Pollutant Emission Factor3
Incremental Change Particulate
in Efficiency, % CO HC NOX Filterable Total
+0.6 + +
+8.8 - + +
+2.6 + +
+3.1 + - +
-2.6 + +
-1.7 b
-3.9 - +
0 b + - +
+2.9 - - + + +
+4.3 - + -
+6.3 c b +
+0.1 -
+ and indicate an increase or decrease exceeding 5 percent.
Change in emission factor not available due to absence of one measurement.
Change in emission not measureable; both values exceeded instrument range.
-------
Table 11-5. Mean Emission Factors and Standard Deviations, for Cyclic Runs - Phases I and II
Emission Factors, lb/1000 gal
Units
All Units
All Unus
Except Those
in Need of
Replacement11
Condition Phase
A 1
II
1+ II
T 1
II
1 + II
R 1
II
1 + II
A 1
II
1 + II
T 1
II
1 + II
R 1
II
1 + II
Numbei
of Units
in
Sample
20
12
32
20
13
33
20
13
33
18
11
29
18
12
30
18
12
30
Bacharacha
Smoke No.
Mean
b
b
b
b
b
b
b
b
b
4.2
1.4
3.2
1.8
0.7
1.3
1.2
0.6
0.9
S.D.
b
b
b
b
b
b
b
b
b
2.2
2.4
2.6
1.0
0.5
1.0
0.6
0.6
0.6
Gaseous Emissions
CO
Mean
>11.9
>39.1
>22.1
> 9.1
>26.9
>16.4
> 8.8
>26.4
>15.7
3.2
153
7.8
3.5
>5.4
4.3
3.5
>5.5
4.3
S.D.
>27.8
>85.1
>56.8
>18.7
>78.1
>51.4
>17.3
>76.2
>49.3
4.5
22.5
15.2
5.2
>10.2
7.6
4.5
>11.8
8.1
HC
Mean
8.4
0.8C
5.7
3.3
2.5
3.0
1.8
2.3
2.0
0.65
0.83
0.72
0.67
0.44
0.57
0.85
0.73
0.80
S.D.
26.1
0.5
21.1
11.2
7.6
9.7
2.8
5.7
4.2
0.40
0.51
0.45
0.42
0.24
0.37
0.59
0.65
0.61
N0x(as
Mean
16.9
23.6
19.4
18.2
21.5
19.5
18,5
21.1
19.6
17.7
22.6
19.6
19.1
20.0
19.5
19.3
19.8
19.5
N02)
S.D.
4.3
5.4
5.7
4.7
6.4
5.6
4.6
5.2
4.9
3.3
4.3
4.4
3.6
3.5
3.5
4.0
3.0
3.6
Participate Emissions
EPA Procedure Modified EPA Procedure
Filterable Total Filterable
Mean S.D. Mean S.D. Mean
3.4
1.4 0.9 6.8 5.9 2.3
2.9
j_ _ _ 22
1.5 1.5 6.3 5.4 25
2.3
_____
_ _ _ _ -
_
2.7
1,2 0.7 5.2 2.5 2.1
2.4
2.2
1.4 1.5 50 2.8 2.3
2.2
_
_____
_ _ _ _
S.D.
4.0
1.3
3.1
1.7
1.7
1.7
_
-
-
3.0
1.0
1.3
1.7
1.6
1.6
_
_
~
Total
Mean
9.6
7.7
8.9
7.1
7.4
7.2
_
-
5.9
6.2
6,0
5.4
6.0
5.7
_
~
S.D.
12.3
5.7
10.2
6.8
5.6
6.2
_
-
-
3.2
2.7
3.0
2.3
2.9
2.5
_
-
~
3 Smoke data at 5-minute point.
b Oily smoke spots prevented obtaining meaningful averages.
c Value low because no data were obtained for HC emissions from Unit 33 in Condition A.
d Units not included were Units 5, 20, and 33.
-------
11-17
The effects of tuning on the gaseous emissions were generally inconsistent. Tuning
reduced CO, HC, and NOX emissions for about one-half of the units. Total particulate was
reduced for eight units, but filterable particulate was reduced for only six units.
Both before and after tuning, raw oil was found on the smoke spot for Unit 33. Also,
this unit operated with CO in excess of the instrument scale limit of 1240 ppm and with
extremely high HC. Hence, this unit is considered to be in need of replacement or major
renovation and would be so identified by competent servicemen. Because of the distortion of
average emissions by the data from this unit, these data are omitted from data summaries and
statistical tabulations where equipment features are examined. (This procedure was also followed
in Phase I, where 2 of the 20 units were identified to need replacement.)
Table II-5 provides a comparison of average emissions for Phase I and II units in the
as-found and tuned condition. Separate summaries are presented showing (1) all units and (2) all
units except those in need of replacement or renovation (identified by the presence of oil on the
smoke spots). Table II-6 shows the percentage reduction in emissions obtained by (1) replace-
ment of poor units and (2) replacement of poor units plus tuning of remaining units.* A large
reduction in average emissions of CO, HC, and particulate is observed when the poorly perform-
ing units are eliminated from the data. Although some shifts are noted with tuning of the
remaining units, the overall effect on emissions of tuning these units is considered minor except
for smoke.
Table 11-6. Reduction of Mean Emissions Upon
Replacement and Tuning of
Residential Units
Replacement Replacement
Only Plus Tuning
Smoke
CO
HC
NOX
Filterable Particulateb
Total Partial lateb
_a
>65%
87%
No change
17%
33%
59%
>81%
90%
No change
24%
36%
Not meaningful, as smoke spots of units eliminated were identified
as "oily" and not evaluated numerically.
By modified EPA procedure.
Fuel. Table II-5 also shows average emissions for the reference fuel runs. (Insufficient
particulate measurements were included for the reference fuel runs to make comparisons of
particulate emissions meaningful.) Only slight shifts are noted by comparing the averages for the
tuned runs and reference fuel runs. Thus, the effect on emissions of differences between the
house fuels and the reference fuel is considered minor.
* Values are based on the ctistribution of "poor" units found in the Phase I and II studies; that is, 3 poor units in
a total sample of 32 oil-fired units.
-------
11-18
Comparison of Emissions for Different
Equipment Features
In order to determine whether certain features of burner design or of installation type
have an important bearing on emissions characteristics, several comparisons were made using the
cyclic runs for Phase I and Phase II as the basis.
The following tables show comparisons of mean emissions from gun-burner units (excluding
units in need of replacement) by equipment categories normally of interest to the oil-heating
industry:
Table 11-7. Matched Units Versus Conversion Units
Table II-8. Furnaces Versus Boilers
Table 11-9. Conventional Versus Flame-Retention Combustion Heads
Table 11-10. Effect of Burner Age.
Pertinent observations regarding each comparison are noted below, especially for the tuned or
reference runs.*
Matched Units Versus Conversion Units. A matched unit is defined as a burner-furnace or
burner-boiler unit that is factory matched and supplied as a combination, whereas a conversion
burner refers to a furnace or boiler installation incorporating a burner other than the model
supplied by the factory. The comparison in Table 11-7 shows that the smoke from matched units
was lower than from conversion units for the as-found condition (2.9 versus 3.7), but not for the
tuned condition. Conversion units had lower CO emissions for all conditions but had higher HC
emissions for the tuned and reference-fuel conditions. Particulate emissions were not significantly
different for the two types of units.
Furnaces Versus Boilers, Table II-8 shows that furnaces operated at slightly lower smoke
levels than boilers. Conversely, mean CO and particulate emissions for the 11 furnaces were
consistently and appreciably higher than for the 16 boilers; however, the mean CO levels for the
furnaces are not considered excessive. The higher CO level with furnaces is surprising, in view of
higher combustion chamber temperatures generally expected for furnaces. Mean NOX emissions
for the furnaces were slightly but consistently higher than for the boilers. The higher NOX levels
with furnaces probably reflects the higher combustion chamber temperatures.
Conventional Versus Flame-Retention Combustion Heads. The comparisons in Table 11-9
indicate higher mean emissions of smoke, CO, HC, and filterable particulate for the 16 burners
having conventional combustion heads than for the eight burners having flame-retention heads.
Although the conventional burners showed the largest reduction in smoke upon tuning (from 3.7
to 1.2), filterable and total particulate emissions from these units showed only a slight reduction
from the tuning. The three Shell head burners had emissions similar to those of the flame-retention
heads.
*Particulate measurements for the reference-fuel condition were only done on the four follow-up units in Phase
II; thus, particulate data from reference runs are not included in statistical tabulations.
-------
Table 11-7. Mean Emission Factors and Standard Deviations for Matched Units and Conversion Units for
Cyclic Runs, Phase 1 and !l Units3
Bacharach
Unit Type
and
Condition
Matched A
T
R
Conversion A
T
R
Number of
Units in
Sam pie
21
22
22
7
7
7
Smoke No.
at 5
Mean
2.Q
1.2
0.9
3.7
1.5
1.0
Win
S.D.
2.6
0.9
0.6
3.2
0.9
0.6
Emission Factors, lb/1000 gai
Gaseous Emissions
CO HC
Mean
9.8
4.4
4.3
1.0
2.9
3.3
S.D. Mean
Units
17.3 0.69
8.2 0.47
9.1 0.72
Units
0.8 0.65
5.8 0.75
4.4 0.90
S.D.
0.45
0.26
0.53
0.25
0.36
0.72
NGx(asPJO2)
Mean
20.1
19.5
19.7
19.0
20.2
20.1
S.D.
4.7
3.5
3.3
3.0
2.8
3.3
Particulate
Filterable
Mean S.D.
2.4 2.7
2.2 1.5
-
2.4 1.2
2.4 2.6
-
Emissions12
Totai
Mean
6.4
5.8
-
5.1
5.6
-
S.D.
3.4
2.5
-
1.5
3.0
Excluding units in need of replacement.
Particulate by modified EPA procedure.
-------
Table 11-8. Mean Emission Factors and Standard Deviations for Furnaces and Boilers for Cyclic Runs,
Phase I and II Units3
Bacharach
Unit Type
and
Condition
Number of
Units in
Sample
Smoke No.
at 5
Mean
Min
S.D.
Emission Factors, lb/1000 gal
Gaseous Emissions
CO
Mean
S.D.
HC
Mean
S.D.
Particulate Emissions^
NOX (as NO2) Filterable
Mean
S.D. Mean S.D.
Total
Mean S.D.
Furnaces
Furnaces A
T
R
11
12
12
2.7
1.0
0.8
3.0
0.8
C.6
17.2
8.5
7.9
21.5
10.7
11.7
0.86
0.43
0.83
0.52
0.25
0.66
22.6
20.5
20.5
4.1 3.4 3.6
2.8 2.6 1.8
2.4
7.2 3.5
6.8 2.6
-
Boilers
Boilers A
T
R
17
17
17
3.4
1.4
0.9
2.6
0.9
0.6
1.4
1.1
1.4
2.7
1.2
1.6
0.56
0.46
0.71
0.27
0.33
0.53
18.0
19.2
19.3
3.4 1.8 1.0
3.6 1.9 1.6
37
5.3 2.5
4.9 2.3
Excluding units in need of replacement.
Particulate by modified EPA procedure.
to
-------
Table 11-9. Mean Emission Factors and Standard Deviations for Burners With Various Combustion-Head Designs for
Cyclic Runs, Phase I and II Units3
Bacharach
Combustion
Head
Conventional
Shell
Flame retention
Conventional
Shell
Flame retention
Conventional
Shell
Flame retention
Condition
A
A
A
T
T
T
R
R
R
Number of
Units in
Sample
17
3
8
18
3
8
18
3
8
Smoke No.
at 5
Mean
3.7
2.8
2.1
1.3
1.7
1.0
0.9
1.0
0.8
Min
S.D.
3.0
2.9
1.9
o.g
1.2
0.6
0.6
0.6
0.6
Emission Factors, lb/1000 gal
Gaseous Emissions
CO
Mean
9.9
0.9
5.3
5.9
0.9
1.1
5.8
1.6
1.1
S.D.
17.7
0.9
12.5
9.3
0.8
0.9
10.0
0.3
0.7
HC
Mean
0.70
0.69
0.63
0.59
0.60
0.39
0.84
0.81
0.54
S.D.
0.42
0.12
0.48
0.33
G
0.18
0.69
0.39
0.18
NOX (as N02)
Mean
19.8
18.9
20.2
20.7
18.3
18.1
20.3
19.2
18.8
S.D.
4.2
3.8
5.0
2.9
3.1
3.8
3.4
4.0
2.9
Particulate Emissions0
Filterable
Mean
2.9
2.3
1.6
2.8
1.3
1.3
S.D.
3.1
0.6
0.8
1.9
c
0.6
Total
Mean
6.7
5.5
4.9
6.8
2.6d
3.9
S.D.
3.3
1.0
2.7
2.4
c
1.6
Excluding units in need of replacement.
1 Particulate by modified EPA procedure.
Less than three data available standard deviation not meaningful.
Based on only one data point.
I
to
-------
Table 11-10. Mean Emission Factors and Standard Deviations by Burner Age, Phase I and II Units3
Bacharach
Number of Smoke No.
Burner Age,
years Condition
Units in
Sample
at 5
Mean
Min
S.D.
CO
Mean
Emission Factors, lb/1000 gal
Gaseous Emissions
HC
S.D.
Mean
S.D.
NOX (as NO2)
Mean
S.D.
Particulate Emissions13
Filterable Total
Mean S.D. Mean
S.D.
5-Year Cut
<5 A
>5 A
<5 T
>5 T
<5 R
^5 R
<15 A
>15 A
<15 T
>15 T
<15 R
>15 R
16
12
17
12
17
12
22
6
23
6
23
6
3.0
3.3
1.1
1.5
0.8
1.0
2.8
4.4
1.3
1.2
0.9
0.8
2.5
2.7
0.8
0.6
0.5
0.5
2.2
4.0
0.3
0.7
0.5
0.6
11.7
2.2
2.8
5.6
2.6
6.1
^ 5-Year
9.0
2.7
2.6
9.1
2.4
10.6
19.6
1.6
4.8
9.5
3.4
10.8
Cut
16.6
2.0
4.1
14.0
3.0
15.8
0.84
0.47
0.51
0.57
0.68
0.87
0.71
0.56
0.47
0.75
0.62
1.28
0.44
0.16
0.25
0.25
0.32
0.47
0.37
0.21
0.21
0.36
0.29
0.65
20.0
19.6
19.7
19.8
19.3
20.5
19.7
20.1
19.3
21.4
19.4
21.4
5.0
3.2
3.6
2.3
2.9
3.5
4.6
3.1
3.4
1.6
3.3
2.1
2.4 2.8 6.2
2.4 1.3 5.8
1.8 1.0 5.2
2.9 1.9 6.4
- -
2.3 2.6 5.9
2.6 1.5 6.6
1.8 0.9 5.0
3.7 2.5 8.1
_ _
3.6
1.9
2.5
1.7
-
3.1
2.6
2.2
2.2
_
Excluding units in need of replacement.
Paniculate by modified EPA procedure.
-------
11-23
Effect of Burner Age. Table II-10 summarizes emission by burner age. for two age cuts: 5
and 15 years. For the 5-year cut, it compares mean emissions for 17 burners less than 5 years
old with 12 burners 5 years old or older. For the 15-year cut, it compares 23 burners less than
15 years old with 6 burners 15 years old or older. Smoke was consistently lower for newer
burners, considering the 5-year cut; for the 15-year cut, newer burners had lower smoke for the
as-found condition only. For the tuned condition, mean CO and particulate levels of the newer
burners were consistently and appreciably less. Conversely, for the as-found condition, CO was
lower for the older burners. Mean HC emissions were generally lower for newer burners in the
tuned condition but higher for newer burners in the as-found condition.
Effect of Burner Age on Smoke and Particulate Emissions When Tuning. Figures II-4 and
II-5 show the change in smoke level, filterable particulate, and total particulate upon tuning for
5-year increments of burner age. Table II-11 presents the same data in the form of percent
reduction of each of these emissions for the tuned condition compared with the as-found
condition. These data show that tuning of the newer burners (those less than 15 years old)
reduced both smoke and particulate emissions. However, for the older burners (which showed
the greatest reduction in smoke), the tuning actually resulted in art increase in particulate
emissions.
Statistical Ranking of Equipment
and Fuel Variables
In addition to the statistical analysis of the data presented above, another statistical
treatment of the data was used to rank the influence of important variables on emissions.
Basically, this technique consists of separating the data into two sets for each variable
considered and then ranking the variables by the differences between the two sets. Where more than
one division of the data was possible (e.g., for continuous variables such as burner age), the data
are divided at the point that produces two sets with the maximum difference. The sets of data
for each variable are compared primarily on the basis of mean values, but the calculation is
weighted by the number of values in each set and the scatter of the data.
The most significant variables insofar as effecting the pollutant being considered can be
regarded as those that produce the largest differences when the data are separated into two sets
on that variable.
An analysis of this type was made on the emission data as follows:
Emission Data
Smoke
CO
HC
NOX
Filterable particulate (by modified EPA procedure)
Total particulate (by modified EPA procedure)
-------
11-24
o
to
Condition
As found
II Tuned
0-4 5-9 10-14 15-19 >ZO
Burner Age, years
Figure II-4. Effect of Tuning on Smoke, by Burner Age
Table 11-11. Effect of Tuning on Smoke and
Particulate Emissions, by Burner
Age
Change of Emissions
Upon Tuning, percent3
Burner Age,
years
0-4
5-9
10-14
15-19
»20
Paniculate
No. of Units
17
2
4
4
2
Smoke
-64
-33
-18
-75
-69
Filterable
-24
-16
-29
+44
+43
Total
-16
-14
_ 7
+ 2
+82
Minus indicates reduction of emission upon tuning; plus indicates an
: in emissions upon tuning.
-------
11-25
10
en 6
O
o
Q_
"o 4
I
O)
Total Particulate
As found
Tuned
Filterable Particulate
0-4 5-9 10-14 15-19
Burner Age, years
>20
Figure II-5. Effect of Tuning on Particulate Emissions by Burner Age
Modified EPA Procedure.
-------
11-26
Conditions
A as found
T - tuned
R reference runs
A and T - together
Variables
Unit type - matched or conversion
System type furnace or boiler
System age by 5-year increments
Burner type - conventional (CH), Shell head (SH), or flame-retention head
(FRH)
Burner age by 5-year increments
Firing rate - by fixed increments, with breaks at 1.00, 1.35, 1.65, 2.00,
and 3.00 gph
Fuel API gravity - by one degree API increments
Condition - A or T (this variable was only included when all as-found and
tuned data were considered together).
For this analysis the following units were not included:
No. 21 and 22 These were gas-fired units.
No. 19 and 20 These were not gun burners, and the one of each type
included in the program was not considered sufficient for statistical
analyses.
No. 5, 20, and 33 These were units in need of replacement, which would
not reflect equipment and fuel variables properly.
No, 35 for Condition A - This was the lab unit, which had no equivalent
prior operation to define a normal "as-found" condition.
Tables 11-12 through 11-15 list the results of the analyses described above for the A, T, R,
and A + T conditions, respectively. For each table and emission, the variables are listed in order
of their decreasing significance.
A difficulty is encountered in interpreting the results of this analysis for variables with
more than two values, because the analysis may split these variables at different points when
considering the different emissions. For example, for the as-found condition, the best split on
firing rate was at 1.35 gph for filterable particulate and at 1.00 gph for total particulate.
As-Found Conditions. Table II-12 shows the ranking of variables in terms of their
significance in affecting emissions for the as-found condition. Examination of this table permits
the following observations:
-------
11-27
System type and firing rate generally received high rankings as having a
major influence on most emissions. Unit type and system age generally
ranked low.
The units having higher firing rates (with a split at 1.0 or 1.35 gph) had
lower emissions of all pollutants except smoke. The smoke data was split
at 3.0 gph, with the smaller units outperforming the two larger units.
High-turbulence combustion heads (flame-retention and Shell heads)
generally had lower emissions than conventional head burners. Exceptions
to this were NOX for flame-retention head units and HC for Shell head
units.
Although fuel gravity appeared near the mid-point on each ranking (3rd,
4th, or 5th), the influence was varied. That is, lighter fuels (33.0 to 36.9
API gravity) produced lower smoke, CO, HC, and total particulate emis-
sions, but heavier fuels (30.0 to 32.9 API gravity) produced lower NOX
and filterable particulate emissions. Hence, the overall influence of fuel
was interpreted as not significant in affecting emissions for units in the
as-found condition.
Tuned Condition. Examination of Table 11-13 showing the ranking of variables for the
tuned condition shows that:
Burner age and system age generally appeared near the top of the lists as
having the greatest influence on emissions; unit type generally had a low
influence on emissions.
Newer units (both burner age, and system age) had lower emissions than
older units. The best splits on age were at 10 years for smoke and 15
years for other emissions.
High-turbulence head burners generally produced lower emissions than
burners with conventional heads. However, smoke emission from Shell
head burners was higher than from the conventional and flame-retention
head burners.
Fuel gravity generally did not have a high ranking and the influence was
varied. That is, units firing lower gravity fuels produced less smoke, CO,
HC, and filterable particulate but more NOX and total particulate.
Reference Condition. Table 11-14 showing the ranking of variables for the reference
fuel runs (in the tuned condition) produced conclusions generally agreeing with conclusions for
the tuned condition.
Combined Data for the As-Found and Tuned Conditions. Table 11-15 shows the ranking
of variables when all as-found and tuned runs are combined and when unit condition (as-found
(Text continues on page 11-36.)
-------
11-28
Table 11-12. Statistical Ranking of Variables for the As-Found Condition
Units for mean values and standard deviations are Ib/1000 gal.
Smoke
Variable Split
Firing Rate
<3.00 gph
>3.00 gph
System Age
<15 years
>15 years
Burner Type
FRH.SH
CH
Burner Age
<15 years
>1 5 years
Fuel Gravity
32.036.9
30.0-31.9
System Type
Furnace
Boiler
Unit Type
Matched
Conversion
Mean
2.9
6.3
2.6
4.3
2.3
3.7
2.8
4.4
2.6
3.9
2.7
3.4
2.9
3.7
S.D.
2.0
0.3
2.1
3.6
2.1
2.9
2.2
4.1
2.6
2.8
2.9
2.5
2.5
3.0
CO
Variable Split
System Type
Boiler
Furnace
Firing Rate
>1 .00 gph
<1.00gph
Burner Age
>5 years
<5 years
Fuel Gravity
33.0-36.9
30.0-32.9
System Age
>10 years
<10 years
Unit Type
Conversion
Matched
Burner Type
FRH.SH
CH
Mean
1.4
17.2
4.5
17.1
2.2
11.7
3.0
12.3
2.7
11.9
1.0
9.8
4.1
9.9
S.D.
2.6
20.5
10.5
21.9
1.9
19.0
3.4
20.9
4.9
20.2
0.7
16.9
10.7
17.2
HC
Variable Split
Burner Age
>5 years
<5 years
System Type
Boiler
Furnace
Fuel Gravity
33.0-36.9
30.0-32.9
Firing Rate
> 1.00 gph
<1.00gph
System Age
>10 years
<10 years
Burner Type
FRH
CH.SH
Unit Type
Conversion
Matched
Mean
0.47
0.84
0.56
0.86
0.57
0.79
0.62
0.86
0.68
0.77
0.63
0.70
0.65
0.69
S.D.
0.43
0.24
0.26
O.BO
0.27
0.49
0.34
0.52
0.29
0.48
0.45
0.39
0.23
0.44
-------
11-29
Table 11-12. (Continued)
Variable
NOX
Split
Filterable Part icu late
Mean
S.D.
System Type
Boiler
Furnace
Firing Rate
>1.35 gph
<1.35gph
18.0
22.6
17.9
21.3
3.3
4.0
3.2
4.5
Fuel Gravity
Unit Type
System Age
30.0-31.9
32.0-36.9
Conversion
Matched
<5 years
^5 years
18.2
20.9
19.0
20.1
19.2
20.2
4.2
4.1
2.8
4.5
5.0
3.8
Burner Type
Burner Age
SH
CH.FRH
3>10 years
<10 years
18.9
19.9
19.5
20.0
3.1
4.4
3.3
4.8
Variable Split
System Type
Boiler
Furnace
Firing Rate
>1.35 gph
<1.35gph
Burner Type
FRH.SH
CH
Fuel Gravity
30.0-32.9
33.0-36.9
Burner Age
<15 years
305 years
System Age
>5 years
<5 years
Unit Type
Conversion
Matched
Mean
1.8
'3.4
1.7
3.0
1.8
2.9
2.0
2.8
2.3
2.6
2.3
2.5
2.4
2.4
S.D.
0.9
3.4
1.1
3.0
0.8
3.0
1.0
3.3
2.6
1.4
1.1
3.4
1.1
2.6
Total Paniculate
Variable Split
Firing Rate
> 1.00 gph
< 1.00 gph
System Type
Boiler
Furnace
Burner Type
FRH.SH
CH
Fuel Gravity
34.0-36.9
30.0-33.9
Unit Type
Conversion
Matched
System Age
>5 years
<5 years
Burner Age
<15 years
>15 years
Mean
5.3
8.3
5.3
7.7
5.1
6.2
4.8
6.5
5.1
6.4
5.6
6.8
5.9
6.6
S.D.
2.8
2.7
2.2
3l3
2.3
3.2
1.2
3.3
1.4
3.3
2.2
4.0
3.1
2.8
-------
11-30
Table 11-13. StatisLical Ranking of Variables for the Tuned Condition
Units for mean values and standard deviations are lb/1000 gal.
Smoke
Variable Split
System Age
<10 years
>10 years
Burner Age
<10 years
>10 years
System Type
Furnace
Boiler
Firing Rate
>1.35 gph
<1.35 gph
Fuel Gravity
30.0-34.9
35.0-36.9
Burner Type
CH.FRH
SH
Unit Type
Matched
Conversion
Mean
1.0
1.6
1.0
1.7
1.0
1.4
1.0
1.4
1.2
1.8
1.2
1.7
1.2
1.5
S.D.
0.7
1.0
0.8
1.0
0.8
0.9
0.5
1.0
0.9
0.9
0.9
1.0
0.9
0.8
CO
Variable Split
System Type
Boiler
Furnace.
Firing Rate
>1.00 gph
O.OOgph
Fuel Gravity
30.0-33.9
34.0-36.9
Burner Age
<15 years
>15 years
Burner Type
FRH.SH
CH
System Age
<15 years
>15 years
Unit Type
Conversion
Matched
Mean
1.1
8.5
1.7
8.9
2.2
8.9
2.6
9.1
1.0
5.9
2.9
6.3
2.9
4.4
S.D.
1.2
10.2
3.5
10.6
4.7
12.6
4.1
14.0
0.8
9.0
4.4
11.9
5.3
8.0
HC
Variable Split
System Age
<15 years
>15 years
Burner Age
<15 years
>15 years
Unit Type
Matched
Conversion
Fuel Gravity
30.0-30.9
31.0-36.9
Burner Type
FRH
CH.SH
Firing Rate
>1.35 gph
<1 35 gph
System Type
Furnace
Boiler
Mean
0.45
0.73
0.47
0.76
0.47
0.75
0.34
0.58
0.39
0.59
0.45
0.60
0.4S
0.56
S.D.
0.23
0.38
0.23
0.44
0.25
0.33
0.20
0.30
0.17
0.32
0.25
0.32
0.24
0.32
-------
11-31
Table 11-13. (Continued)
NOX
Variable Split
Burner Type
FRH.SH
CH
Fuel Gravity
35.0-36.9
30.0-34.9
System Age
<15 years
>15 years
Burner Age
<15 years
3*15 years
Firing Rate
>1.35 gph
<1.35gph
System Type
Boiler
Furnace
Unit Type
Matched
Conversion
Filterable Participate
Mean
18.2
20.7
16.5
20.1
19.0
21.3
19.3
21.4
18.8
20.4
19.2
20.5
19.5
20.2
S.D.
3.4
2.9
1.5
3.3
3.2
3.2
3.5
1.6
3.6
3.0
3.5
2.7
3.4
2.6
Variable Split
Burner Age
<15 years
>15 years
Burner Type
FRH,SH
CH
Firing Rate
>1.00 gph
< 1.00 gph
System Age
<15 years
>15 years
Fuel Gravity
30.0-30.9
31.0-36.9
System Type
Boiler
Furnace
Unit Type
Matched
Conversion
Mean
1.8
3.7
1.3
2.8
1.7
3.1
1.9
3.0
1.6
2.4
1.9
2.6
2.2
2.4
S.D.
1.0
2.7
0.5
1.9
1.4
1.8
1.0
2.6
0.4
1.9
1.6
1.7
1.5
2.2
Total Participate
Variable Split
Burner Type
FRH.SH
CH
Burner Age
<15 years
>15 years
System Type
Boiler
Furnace
System Age
<15 years
>15 years
Firing Rate
JM.OOgph
<1.00 gph
Fuel Gravity
32.0-36.9
30.0-31.9
Unit Type
Conversion
Matched
Mean
3.7
6.8
5.0
8.1
4.9
6.8
5.2
6.9
5.2
6.7
5.4
6.3
5.6
5.8
S.D.
1.6
2.3
3.1
2.3
2.2
2.5
2.3
3.0
2.5
2.3
2.6
2.4
2.7
2.5
-------
11-32
Table 11-14. Statistical Ranking of Variables for the Reference Condition
Units for the mean values and standard deviations are lb/1000 gal.
Smoke
Variable Split
Burner Age
<10 years
>10 years
System Age
<10 years
>10 years
Burner Type
FRH
CH,SH
Firing Rate
>1.35 gph
<1.35 gph
System Type
Furnace
Boiler
Unit Type
Matched
Conversion
Mean
0.8
1.1
0.8
1.0
0.8
0.9
0.8
0.9
0.8
0,9
0.9
1.0
S.D.
0.5
0.7
0.5
0.6
0.5
0.6
0.5
0.6
0.6
0.5
0.6
0.6
CO
Variable Split
Burner Age
<15 years
5=15 years
Firing Rate
>1.00gph
<1.00 gph
System Type
Boiler
Furnace
Burner Type
FRH.SH
CH
System Age
<15 years
>15 years
Unit Type
Conversion
Matched
Mean
2.4
10.6
1.9
8.9
1.4
7.9
1.2
5.8
2.6
7.4
3.3
4.3
S.D.
3.1
16.2
1.8
12.4
1.6
11.2
0.7
9.7
3.3
13.7
4.0
8.9
HC
Variable Split
Burner Age
<15 years
^15 years
System Age
<1f years
>15 years
Firing Rate
>1.35 gph
<1.35 gph
Burner Type
FRH
CH,SH
Unit Type
Matched
Conversion
System Type
Boiler
Furnace
Mean
0.62
1.28
0.59
1.12
0.51
0.95
0.54
0.84
0.72
0.90
0.71
0.83
S.D.
0.31
1.15
0.29
0.86
0.27
0.68
0.17
0.65
0.52
0.67
0.51
0.63
-------
II-33
Table 11-14. (Continued)
Variable Split Mean S.O.
System Age
<5 years 18.6 2.5
>5 years 20.6 3.5
Burner Age
<10 years 19.2 2.9
>10 years 21.0 3.9
Burner Type
FRH,SH 18.9 3.1
CH 20.3 3.3
System Type
Bailer 19.3 3.6
Furnace 20.5 2.3
Firing Rate
>2.00 gph 18.9 2,5
<2.00 gph 20.0 3.7
Unit Type
Matched 19.7 3.3
Conversion 20.1 3.0
-------
11-34
Table 11-15. Statistical Ranking of Variables for the As-Found and Tuned Conditions Combined
Units tor mean values and standard deviations are lb/1000 gal.
Smoke
Variable Split
Condition
T
A
Firing Rate
<3.00 gph
>3.00 gph
System Age
<10 years
>10 years
Burner Age
<10 years
^10 years
Burner Type
FRH
CH.SH
Fuel Gravity
32.0-36.9
30.0-31.9
System Type
Furnace
Boiler
Unit Type
Matched
Conversion
Mean
1.3
3.1
2.0
4.5
1.7
2.7
1.9
2.7
1.6
2.4
1.9
2.6
1.8
2.4
2.0
2.6
S.D.
0.9
2.7
2.1
2.5
1.8
2.5
2.0
2.5
1.4
2.4
2.0
2.5
2.2
2.1
2.1
2.5
CO
Variable Split
System Type
Boiler
Furnace
Firing Rate
>1.00 gph
<1 .00 gph
Burner Type
SH.FRH
CH
Unit Type
Conversion
Matched
System Age
>5 years-
<5 years
Condition
T
A
Burner Age
>5 years
<5 years
Fuel Gravity
33.0-36.9
30.0-32.9
Mean
1.2
12.9
3.1
12.5
2.6
7.9
1.9
7.1
4.2
8.7
4.0
7.6
3.9
7.3
4.4
7.2
S.D.
2.1
16.8
8.0
17.0
7.6
13.9
3.9
13.5
8.4
16.5
7.4
15.1
7.4
14.5
7.3
15.6
HC
Variable Split
Burner Age
>5 years
<5 years
Condition
T
A
Firing Rate
>1.35 gph
<1.35 gph
Burner Type
FRH
CH,SH
Fuel Gravity
33.0-36.9
30.0-32.9
System Age
<15 years
3»15 years
System Type
Boiler
Furnace
Unit Type
Matched
Conversion
Mean
0.52
0.67
0.53
0.68
0.52
0.67
0.51
0.64
0.55
0.67
0.57
0.69
0.56
0.67
0.58
0.69
S.D.
0.31
0.38
0.29
0.40
0.27
0.41
0.36
0.35
024
0.46
0.37
0.34
0.29
0.43
0.37
0.29
-------
11-35
Table 11-15. (Continued)
NOX
Variable Split
System Type
Boiler
Furnace
Firing Rate
>1.35 gph
<1.35 gph
Fuel Gravity
35.0-36.9
30.0-34.9
Burner Type
SH.FRH
CH
Burner Age
<15 years
>15 years
System Age
<15 years
>15 years
Unit Type
Conversion
Matched
Condition
T
A
Filterable Paniculate
Mean
18.6
21.5
18.3
20.8
17.5
20.0
19.0
20.2
17.9
20.7
19.4
20.5
19.6
19.8
19.7
19.8
S.D.
3.5
3.5
3.4
3.8
2.7
3.8
4.0
3.5
8.8
2.4
4.0
3.2
2.8
4.0
4.2
3.3
Variable Split
Burner Type
SH.FRH
CH
System Type
Boiler
Furnace
Firing Rate
>1.35gph
<1.35gph
Burner Age
<15 years
>15 years
Fuel Gravity
30.0-30.9
31.0-36.9
System Age
<15 years
>15 years
Condition
T
A
Unit Type
Matched
Conversion
Mean
1.6
2.9
1.9
3.0
1.7
2.8
2.1
3.2
1.7
2.5
2.1
2.7
2.2
2.4
2.3
2.4
S.D.
0.7
2.5
1.3
2.7
1.3
2.4
2.0
2.2
0.6
2.3
2.1
1.9
1.6
2.3
2.1
1.7
Total Paniculate
Variable Split
Burner Type
SH.FRH
CH
Firing Rate
>1.00gph
<1.00gph
System Type
Boiler
Furnace
Burner Age
<15 years
>15 years
Fuel Gravity
34.0-36.9
30.0-33.9
System Age
<15 years
>15 years
Unit Type
Conversion
Matched
Condition
T
A
Mean
4.5
6.7
5.3
7.4
5.1
7.0
5.5
7.4
4.9
6.2
5.7
6.4
5.3
6.1
5.7
6.1
S.D.
2.1
2.8
2.6
2.6
2.3
3.0
2.7
2.6
1.6
3.1
2.8
2.7
2.0
2.9
2.6
3.0
-------
11-36
or tuned) is considered as a variable. The interesting point in this table is the ranking of unit
condition relative to other variables. Observations from this table show that:
Unit condition generally appeared low in the ranking except for smoke
and HC. Unit condition would be expected to appear high on the smoke
list, because tuning was based on a smoke criteria. HC emissions were
always low, so the high ranking of condition for this emission is not
significant. Hence, these results confirm the earlier observation that tuning
does not greatly 'influence emissions from generally well-performing units.
Firing rate and burner type generally appeared high in the rankings;
system age, unit type, and fuel gravity generally appeared low in the
ranking.
Units with lower firing rates produced higher emissions except for smoke.
The data usually split at 1.0 to 1.35 gph.
High-turbulence combustion heads produced lower emissions except for
smoke and HC, where Shell head units did not significantly outperform
conventional units.
Conclusions Related to Equipment Features and Fuel
Examination of results reported above (where emissions are compared for various equip-
ment features and fuels) does not reveal firm guidelines for the industry to follow with regard to
equipment design, service requirements, or fuel. Most variables (such as design features, tuning,
and fuel gravity) had little effect on emissions.
The major influence on emissions from oil-fired industrial units was found to be the
elimination (by replacement or major overhaul) of the very poorly performing units. Eliminating
these units (9 percent of the units in this limited sample) reduced emissions of the total
population of oil-fired units by about 50 percent. It was found that these poorly performing
units were characterized by the appearance of oil on the smoke spot. This characteristic should
readily be recognized by burner servicemen. Hence, an annual inspection and cleanup (already
recommended by many oil-heating organizations) that includes a smoke spot evaluation should
identify units producing excessive emissions. Tuning the remaining units to any specified smoke
level, as has been proposed in some localities as an air pollution-control measure, does not appear
to have any additional benefit insofar as reducing the mass of emissions. In fact, smoke
measurements can yield a low smoke reading and, at the same time, show evidence of oil - the
telltale sign of high emissions.
Apart from replacing poor units, the most significant feature insofar as its impact on
emissions was the generally superior performance of units equipped with flame-retention combus-
tion heads. Fortunately, the industry has largely adopted flame-retention head burners as the
preferred modern burner design. As these units replace older conventional gun burners, the
oil-heat industry will be moving in the direction of lower average emissions.
Measurements on Follow-Up Units
Measurements made in the Phase I program were .on a spot basis only, with a single
follow-up visit made only to two units. This approach was extended in Phase II by observing
-------
11-37
emission levels for four units during the heating season to determine the effect of time since the
initial tuning. Two follow-up checks were made on Units 23, 24, 25, and 26 after approximately
2 and 4 months of operation during the heating season.
These units, selected to provide a cross section of burner types, were as follows:
Unit
Burner Type
23 High-pressure gun, FRH*,
3450 rpm
24 High-pressure gun
25 High-pressure gun, FRH*,
3450 rpm
26 High-pressure gun, Shell
head
Burner
Age
< 1
4
< 1
2
System Type
Water, CI boiler
Forced stored air,
steel furnace
Water, steel boiler
Steam, CI boiler
Domestic Hot
Water Coil
No
No
Yes
Yes
The burners were equipped with operating time clocks and cycle counters to record operating
experience between visits.
For these follow-up checks, gaseous emissions were measured for the as-found condition
using both the house fuel and the reference fuel. Particulate measurements were included for
these units as part of the initial measurements and at the end of the season with the reference
fuel.
Table 11-16 shows complete measurements and emission factors for the initial runs and
the two subsequent follow-up checks. Also shown in this table are emissions for Units 12 and 16
from Phase I. Emissions from these units were measured during Phase II and reported as Units 24
and 26, respectively.
Table 11-17 provides a summary of the average emission factors for the follow-up units.
Emission averages are 'shown for the four units and for three units not including Unit 25, which
had unusual nozzle clogging problems.**
For the four follow-up units over the 4-month operating period, only minor shifts were
noted in CO and NOX. Smoke increased by 90 percent, HC increased by 200 percent, and
filterable and total particulate increased by 131 and 76 percent, respectively. However, if the
data from Unit 25 are omitted, the average particulate emissions do not increase over the
4-month period. The increase in HC emissions was primarily associated with Units 24 and 25. As
stated earlier, Unit 25 had unusual problems, which were responsible for increased emissions
from that unit. The increase in HC emissions between the 1st and 2nd follow-up visits for Unit
24 has not been explained.
Examination of the operating and emissions data for the two units that were included in
both the Phase I and II studies showed that these units performed about the same for both
years. Unit 12-24 had a tendency to have somewhat high CO and HC emissions; this was evident
on each visit.
*Flame retention head.
**See explanation in Table 11-19.
-------
Table 11-16. Summary of Emissions and Emission Factors for Cyclic Runs on Follow-Up Units
Operational Data
Unit and
Condition8
23
23.1
232
12d
24d
24.1
24.2
25
25.1
25.2e
16'
26'
26.1
26.2
A
T
R
T
R
T
R
A
T
R
A
T
R
T
R
T
R
A
T
R
T
R
T
R
ft
T
R
A
P
"
R
T
R
C02,
9.2
99
94
ea
8.4
87
88
7.4
7.2
6.7
7.1
8.9
9.2
8.8
9.0
7.8
8.6
84
9.4
S.B
108
10.9
91
il). 1
6.7
69
69
7,1
8.7
88
9.4
94
9.2
9.0
*
9.1
7.8
8.2
9.1
9.3
9.0
8.8
10.3
10.5
10.6
12.0
9.0
8.5
9.0
8.3
10.5
9.2
9.0
7.3
8.8
6.6
6.5
3.4
7.0
11.3
11.1
11.2
11.1
9.1
8.8
8.2
8.0
8.2
B.6
Excess
Air. %
68
54
60
72
77
73
70
97
102
111
120
71
64
/2
65
95
74
75
56
72
42
41
64
49
118
112
111
109
73
70
62
59
62
66
Smoke No.
at 9 Min»
0.3
0.6
0.4
0.1
0.1
0.0
0.0
3.0
1.0
1.0
0.4
0.7
0.8
0.2
0.1
0.2
0.7
1.0
1.7
2.0
0.4
0.1
4.1
5.8
2.0
1.5
1.0
0.5
0.4
0.5
0.2
0.2
0.5
0.4
Stack
Temperature,
F
470
480
480
-
-
435
445
690
600
-
645
520
520
-
610
605
490
450
455
-
-
500
485
520
-
-
520
500
490
-
-
510
510
Firing
Rata,
gph
1.20
-
1.30
-
1.28
-
1.30
0.96
098
-
1.04
-
0.86
-
0.95
-
1.02
1.35
-
1.40
-
1.38
-
1.59
1.60
1.75
-
1.78
-
1,80
-
1 79
-
1 80
D
"CO
3.7
7.7
2.6
7.0
a. 7
4,5
2.9
66.4
76.9
76.1
222 0
4.1
9.5
20.1
18.3
41.4
19.9
10.6
8.6
12.1
15.5
14.7
10.3
11.8
17.0
17.1
17.5
5J5
6.0
10.7
6.9
7.0
55
5.2
Emission Data,
osa Average, ppm
HC
8.3
6.7
3.2
6.8
7.7
8.3
4.8
9.4
5.5
11.7
12.2
3.1
4.5
5.3
11.7
28.7
21.0
5.4
5.9
5,7
8.6
5.9
22.7
22.0
6.8
6.7
6.7
5.1
4.0
4.4
6.8
6.1
7.8
7.2
NO
67
82
75
86
76
85
91
_
-
-
39
71
69
59
62
56
64
42
47
44
49
54
38
37
_
-
-
54
68
67
69
69
66
66
NOx
85
89
79
84
77
85
97
69
78
62
78
76
71
60
64
58
64
56
47
49
47
48
36
39
51
60
60
64
71
72
74
73
68
67
Emission Factors, lb/1000 gal
Paniculate
Loading, mg/sm3 '
Filterable
10.7
7.2
6.4
-
-
_
12.6
10.4
7.0
-
6.1
19.0
13.8
-
-
8.8
5.7
7.4
7.9
-
-
78.8
9.7
-
-
11.1
9.1
10.7
-
-
-
9.1
Total
20.5
26.0
20.5
-
-
_
24.8
33.2
16.9
-
15.4
47.6
32.2
-
-
26.6
25.0
19.4
17.7
-
-
lOOS
22.7
-
-
33.2
17.6
21.4
-
-
-
18.2
CO
0.61
1.17
0.41
1.20
1.51
0.78
0.4B
10.96
13.37
13.90
48.28
0.69
1.53
3.38
2.96
8.07
3.41
1.83
1.32
2.05
2.17
2.03
1.66
1.72
1.68
1.62
1.69
1.13
1 02
1.78
1.09
1.10
0.88
0.85
Paniculate0
NOX
HC |as N02)
0.79
0.50
0.29
0.67
0.77
0.82
0.46
1.86
0.63
1.39
1.62
0.30
0.42
0.61
1.08
3.20
2.06
0.53
0.55
0.55
0.69
0.47
2.10
154
0.84
0.80
0.80
0.60
0.39
0.42
0.62
0.55
071
0.67
23.1
22.2
20.5
24.3
22.0
24.1
266
22.4
25.5
21.5
27.9
2T1
18.8
16.6
17.0
186
18.0
15.9
118
13.7
11.2
12.3
10.1
9.4
20.5
20.4
23.0
21.6
19.9
19.7
19.2
18.8
17.8
18.0
Filterable
1.55
0.93
0.83
-
-
1.82
1.74
1.23
-
1.17
2.83
2.64
-
_
-
1.31
0.83
0.97
1.16
-
-
9.99
1.82
-
-
1.96
1.35
1.55
_
-
_
1.29
Total
294
3.44
2.82
-
-
3.59
5.67
2.98
-
2.93
7.02
4.53
-
-
-
1.27
3.73
2.57
2.61
-
-
-
12.69
4.31
-
-
5.97
2.61
3.11
-
-
_
2.58
00
8 Decimals (.1 and .2) refer to 1st and 2nd follow-up visits.
D Smoke data for Units 12 and 16 are obtained at the 5-minute point.
c Modified EPA procedure.
^ Unit 12 from Phase 1 and Unit 24 from Phase II were the same unit.
e Unit 25 was adjusted between measurements due to the nozzle clogging problem described in Table 11-19.
' Unit 16 from Phase I and Unit 26 from Phase II were the same unit.
-------
11-39
Table 11-17. Average Emissions for Follow-Up Units
Mean Emission Factors, lb/1000 gal
Units
Four Follow-up Units
(Units 23, 24, 25, & 26)
Initial Visit
1st Follow-up
2nd Follow-up
Three Follow-Up Units
(Units 23, 24, & 26)
Initial Visit
1st Follow-up
2nd Follow-up
Emission Ratios Based on
(Units 23, 24, 25, & 26)
1st Follow-up
Initial Visit
2nd Follow-up
Initial Visit
Emission Ratios Based on
(Units 23, 24, & 26)
1st Follow-up
Initial Visit
2nd Follow-up
Initial Visit
Bacharach
Gaseous Emissions Particulate Emis-
Smoke Number,
Condition 9th minute CO
A
T
R
T
R
T
R
A
T
R
T
R
T
R
Four Units
R
R
Three Units
R
R
0.6
0.8
0.9
0.2
0.1
1.2
1.7
0.4
0.5
0.6
0.2
0.1
0.2
0.4
0.1
1.9
0.2
0.7
13.0
1.05
1.44
1.96
1.90
2.85
1.62
16.7
0.96
1.24
1.89
1.86
3.24
1.58
1.32
1.13
1.50
1.27
NO, sions3
HC (or NO2) Filterable
0.86
0.43
0.42
0.62
0.72
1.71
1.26
0.97
0.40
0.38
0.60
0.80
1.58
1.06
1.71
3.00
2.11
2.79
22.1 1.38
18.8 1.62
18.2 1.56
17.8
17.5
17.7
18.0 3.60
24.2 1 .56
21.1 1.70
19.7 1.69
20.0
19.3
20.2
20.9 1.47
0.96
0.99 2.31
0.98
1.06 057
Total
3.89
3.91
3.27
-
5.78
3.95
4.36
3.49
-
3.48
-
1.76
1.00
3 Modified EPA procedure.
-------
11-40
The general lack of any large increase in emissions for Units 23, 24, and 26 during the
Phase II study supports the finding reported earlier that tuning of units that are generally
performing satisfactorily does not result in a large reduction in emissions. However, periodic
servicing should identify problems such as those experienced by Unit 25 and permit correction of
these problems. The high smoke level measured during the last visit to this unit would identify
existence of a problem, using equipment and techniques presently available to burner servicemen.
The incremental changes in efficiency of the follownip units between the first and last
visits were as follow:
Unit 23 + 0.4 percent
Unit 24 - 2.8 percent
Unit 25 +0.1 percent
Unit 26 + 0.5 percent.
Hence, except for Unit 24, the incremental change in efficiency was insignificant. The incre-
mental increase in efficiency upon tuning during the initial visit for Unit 24 was 8.8 percent.
Apparently, during the 4-month interval between the initial visit (and tuning) and the last visit,
performance of this unit had begun to deteriorate. However, its performance (efficiency) was still
superior to that in the as-found condition.
Table 11-18 and 11-19 provide information on the operating cycles and service history of
the units during the follow-up period. Degree-days during the approximate 4-month period was
4100, indicating an average outdoor temperature of 34 F. Table 11-18 shows that the load factor
of the four units was remarkably similar, ranging from 30 to 38 percent on time (compared with
33 percent chosen for the cyclic runs). The-average cycles per hour ranged from 1.2 to 4.7 for
the four units (compared with 2 cycles per hour for the cyclic runs). The steam boiler operated
with longer cycles characteristic of this type of unit.
Experiments on the Effect of Cycle
A series of experiments were made to investigate the effect of cycle with Unit 35, which
was operated in the laboratory at Battelle-Columbus. This is a forced-air furnace unit with a
high-pressure gun burner a conventional combustion head, and a ceramic felt combustion-
chamber liner. Gaseous emissions were measured while the unit was operated in the following
cycles:
Min. On/Min. Off Cycles/Hour Percent Time On
Equilibrium - 100
10/20* 2 33
10/5 4 67
7.5/7.5 4 50
5/10* 4 33
1.5/13.5 4 10
6.7/3.3 6 67
3.3/6.7 6 33
1/9 6 10
*Runs made with and without a solenoid oil valve.
-------
11-41
Table 11-18. Operational Data on Follow-Up Units
Unit
(System Type)
23
(Water)
24
(Warm Air)
25
(Water)
26
(Steam)
a Cycle used in
Dates
Month
12-10 -
1-10-
2-16-
1-10-
12-13-
1-10
1-17
1-22 -
1-25 -
1-29-
2-4 -
2-9
2-15
2-21 -
2-26-
3-10-
3-18 -
3-26 -
4-9 -
4-16 -
1-10 -
12-16 -
1-10-
1-17 -
1-22 -
1-31 -
2-15 -
3-3
3-10 -
3-24 -
4-15 -
1-10 -
12-18
12-26 -
2
-10 -
-14 -
-24-
-30 -
25 -
2-11 -
2-14 -
2-24 -
3-11 -
3-18 -
3-25 -
4-10
4-22
1-10-
Day
1-10
2-16
4-26
4-26
-10
17
-22
-26
-29
2-4
2-9
2-15
2-21
2-26
3-10
3-18
3-26
4-9
4-16
4-25
4-25
1-10
1-17
1-22
1-31
2-15
3-3
3-10
3-24
4-15
4-24
4-24
12-16
1-2
1-10
1-14
1-24
1-30
2-5
2-11
2-14
2-24
3-11
3-18
3-25
4-10
4-22
4-27
4-27
measurements was:
Burner Operation
Hours
221
306
462
768
234
68
41
22
45
59
72
64
66
64
122
76
75
116
39
38
965
175
52
34
66
139
136
62
109
144
36
778
72
55
84
22
105
60
72
78
22
108
127
57
57
123
50
20
901
Cycles
_
4644
7521
12127
,
330
276
165
240
353
289
377
334
306
692
459
460
757
294
292
5624
644
437
803
1555
1600
668
1233
1915
571
9426
-
-
138
322
152
152
166
87
279
479
154
182
433
356
156
3056
Average On
Time, mina
_
4.0
3.7
3.8
12.3
9.0
8.1
11.3
10.0
15.0
10.1
11.8
12.6
10.5
9.9
9.7
9.2
8.0
7.9
10.3
4.8
4.7
4.9
5.4
5.1
5.6
5.3
4.5
3.8
5.0
-
-
9.6
19.5
23.6
28.3
28.3
15.1
23.1
15.9
22.2
18.8
17.1
8.5
7.8
17.7
Average
Percent
On Timea
30
34
28
30
35
40
34
31
47
41
60
44
46
53
39
40
39
35
23
18
38
29
31
28
31
39
33
37
32
27
17
31
38
33
44
23
44
42
50
54
31
45
33
34
34
32
17
17
35
Average
Cycles
per Houra
52
45
4.7
2.0
2.3
2.3
2.5
2.5
2.4
2.6
2.3
2.6
2.2
2.4
2.4
2.3
1.8
1.4
2.2
3.8
3.6
3.7
4.3
3.9
4.0
3.7
3.6
2.6
3.7
_
-
1.4
1.3
1.1
1.1
12
12
12
12
09
1.1
1.1
1.2
1.3
1.2
On Time, min = 10
Percent On
Cycles per
Time - 33
Hour - 2.0
-------
11-42
Table 11-19. History of Follow-Up Units Between Visits
Unit 23. No servicing or adjustments of any kind were made on this unit between visits.
Unit 24. On February 4, 1972, a no-heat service call was made to this unit. The
serviceman diagnosed the trouble as a burned out burner motor, which was
replaced without making any other adjustments to the unit.
Unit 25. A new shipment of oil (376 gal) was delivered to this unit on January 26, 1972.
Because of prior problems associated with oil deliveries, the burner on this
unit is normally turned off and the oil company has specific instructions
for a slow-fill. On the date in question no one was home -and a new driver,
not having been duly informed, fast-filled the tank. The fast-fill stirred up
sediment in the tank and clogged the nozzle, causing shutdown; the home
owner cleaned the nozzle and adjusted the air to achieve good combustion.
Additional problems were .encountered with this unit on February 16, 1972.
Following the conclusion of the series of emission measurements on the
first follow-up visit, the burner would not ignite. The trouble was diagnosed
as a clogged filter and a partially blocked fuel line. The old filter was
replaced and the fuel line from the filter to the tank was blown out with
C02'cartridges. This corrected the problem and the unit was set to the same
CO2 as for the tuned condition of the initial visit and was left on this setting.
Unit 26. No servicing or adjustments of any kind were made on this unit between visits.
Table 11-20 shows the emissions measured by the procedure used for the cyclic runs in
the field, plus equilibrium operation for comparison. As might be expected from combustion
temperature considerations, runs with shorter "on" time generally yielded slightly higher CO and
HC and' lower NOX. (Particulates were not measured during these runs.) Improved cutoff at
shutdown with the addition of a solenoid oil valve resulted in slightly lower HC emissions where
comparable cycles were run with and without the solenoid.
The temperature of the lightweight, highly insulating combustion-chamber liner in this
unit would be expected to respond rapidly to transient operations. A unit with a dense, highly
conductive refractory liner would be expected to have a slow response and, thus, to show larger
cyclic effects on emissions.
-------
Table 11-20. Effect of Cycle on Emissions
Cycle
"On", min/"Off". min
Steady state (2 hr)
10/20
10/20
10/5
7.5/7.5
5/10
5/10
1.5/13.5
6.7/3.3
3.3/6.7
1/9
Operating Conditions3
Percent
On Time
100
33
33
67
50
33
33
10
67
33
10
C02b,
%
10.3
10.1
10.1
10.3
10.3
10.2
10.3
10.1C
10.3
10.2
9.6C
°2b-
%
6.8
7.3
7.2
6.9
6.9
7.0
7.1
7.5C
6.7
7.0
8.4C
Stack Temperature
Just Before Shutoff
488
481
486
478
470
441
448
323
467
405
245
Solenoid
in Use
-
No
Yes
No
No
No
Yes
No
No
No
No
Dose Average Emissions Data, ppm
CO
11.0
13.3
13.3
13.6
13.8
14.5
14.4
16.8
13.6
15.0
17.0
HC
4.0
15.3
14.8
9.8
11.7
11.7
9.9
18.7
8.8
13.5
26.7
NO
77
72
73
77
79
76
74
70.5C
76
74
66C
IMOX
76.5
71
74
76
78
76
74
74.5
76.5
74
66.5C
All smoke readings (made 1 minute after lightoff and just before shutdown) were 0.4 or less on the Bacharach scale.
Average values over "on" period.
Values recorded at end of "on" cycle, as the on-time was too short to obtain meaningful average values.
-------
11-44
EMISSION RESULTS FOR VARIED-AIR RUNS
The purpose of the varied-air runs was to characterize the sensitivity of emissions to air
adjustment for the various residential units. These runs were made at steady-state conditions and
the only adjustment made between points was to reposition the air gate.
This section contains a summary of gaseous emissions and smoke measurements for the
varied-air runs on the residential units. Particulate emissions were not measured for the varied-air
runs.
Appendix G of the Data Summary Volume contains tabulated emission data and plots of
emissions versus excess air (in terms of CO2 reading) for varied-air runs at all operating
conditions (as-found, tuned, and reference-fuel runs). Appendix I reports the same data in terms
of emission factors.
Emission Characteristics Related to Excess Air
Gaseous emissions and smoke were measured at about six steady-state conditions repre-
senting a range of excess air settings between the wide open setting of the air gate and a setting
which gave a smoke level of 5 to 7. Plots of emissions versus excess air or C02 provide an
indication of the operating range producing low gaseous emissions or smoke. Figure II-6 is an
example plot showing the variation of smoke and gaseous emissions with CO2 for one unit.
Figures II-7 through 11-19 show summary curves of smoke, CO, and HC, for all 12
residential units comparing as-found (A), tuned (T), and reference-fuel (R) conditions. NOX
was not highly sensitive to air setting and is not summarized in these curves. Individual data
points are tabulated and plotted against CO2 for each condition in the Data Supplement Volume.
Examination of Figures II-7 to 11-19 show that each of these units is unique in its
combustion characteristics. In fact, it is quite difficult to select any one of these units as having
combustion characteristics which are typical of the group of units. Some units exhibited very
sharp breaks in the smoke curves; others had broad, sweeping curves. Tuning generally improved
the smoke curve (moved it toward higher CO2 values), but, as tuning was primarily oriented
toward reducing smoke, this was not surprising. Most units had very low CO and HC emissions
over the normal range of operation in the as-found condition, so that significant improvement
upon tuning was not possible.
The most significant point that can be made from these data is that smoke is a good
indicator of low CO and HC emissions at high CO2 levels but not at low CO2 levels. That is,
almost without exception, as excess air is gradually decreased from the normal operating range,
the smoke level begins to increase sharply before the CO and HC emissions show significant
increases. However, as excess air is increased, CO and HC emissions increase quite rapidly while
smoke levels remain low. Hence, at high CO2 levels low smoke is a fairly good indicator that a
burner is adjusted for low CO and HC emission levels. Conversely, smoke is not a good measure
of low CO and HC production in the low CO2 range, and adjustment based on smoke alone
could lead to high CO and HC emissions.
(Text continues on Page 11-59.)
-------
11-45
cu
.0
E 6
o
o
k_
o
o 3
o °
CO
0
CO
32-
28-
24-
20-
12-
Q _-..
4-
160
140
120
100
80
60
40
20
E
Q.
D.
O
O
7 8 9 10 II 12
C02, percent
Figure II-6, Typical Smoke and Gaseous Emission Characteristics for a
Residential Unit
Unit 24 in tuned condition firing house fuel.
-------
s
i
CO
9 10
percent
E
a
o"
0
leu
170
160
150
140
130
120
no
100
90
BO
70
60
50
40
30
20
10
n
-
-
-
-
-
-
-
_
_
-
-
-
-
-
-
-
CO-R
CO-A
HC-R/N"
CO
-7
-
-
-
-
-
-
-
_
-
-T
^C-T -
3D
34
J2
30
28
26
24
22
20
18
16
14
12
10
8
6
4
2
8 9
COZ, percent
Figure 11-7. Emissions for Residential Unit 23 as Function of CO2
Lables on curves refer to pollutants (CO or HC) and/
or burner condition (A, T, or R).
-------
Figure II-8. Emissions for Residential Unit 24 as Function of CO2
-------
8 9
C02, uercenl
180
170
160
150
140
130
120
no
IOO
90
80
70
60
50
40
'30
20
10
0
36
34
32
30
28
26
24
22
20
18
16
!4
12
10
8
00
Figure II-9. Emissions for Residential Unit 25 as Function of CO2
-------
8 9
CO^, percent
ISO
70
160
150
140
ISO
120
no
100
90
80
70
60
50
40
30
20
10
0
36
34
32
30
28
26
24
22
20
18
16
14
12
10
8
6
4
2
0
C02, percent
Figure H-10. Emissions for Residential Unit 26 as Function of CO2
-------
-789
COj, percent
ISO
I TO
60
150
140
1301
12O
no
IOO
90
80
70
60
50
40
30
20
10
0
\CO-A
HC-A
36
34
32
30
2B
26
24
32
20
Figure 11-11. Emissions for Residential Unit 27 as Function of CO2
-------
I
z
8
I
8 '
a 9
COj, percent
ISO
170
160
ISO
140
130
120
110
1 100
I
°- 90
3 80
TO
60
50
40
30
20
10
0
HC-A
36
34
32
30
28
26
34
22
20
18
16
14
12
10
8
6
4
2
0
C02, percent
Figure 11-12. Emissions for Residential Unit 28 as Function of CO,
-------
I
8 9
C02, percent
I
10
d
Q.
O
C02, percent
Figure 11-13. Emissions for Residential Unit 29 as Function of CO2
Ul
to
-------
-------
!
170
160
ISO
140
130
120
110
100
90
80
70
60
5O
40
30
20
10
0
34
32
30
?8
26
24
22
20
O
I
CO-A
CO-T
Figure 11-15. Emissions for Residential Unit 31 as Function of CO9
-------
04-
S
CD
ISO
170
160
150
140
130
120
110
100
90
80
70
60
50
40
30
20
10
0
HC-A
36
34
32
30
28
26
24
2?
20
18
16
14
12
10
8
6
4
2
e 9
C0?, percent
Figure II-16. Emissions for Residential Unit 32 as Function of CO2
-------
I
I
CO,
9
percent
10
I
ISO
170
160
150:
I4o'
130
120
110
100
90
80
70
60
50
10
30
20
10
0
CO-R
CO-A and HC-T off scnle
I
C02, percent
Figure II-17. Emissions for Residential Unit 33 as Function of CO2
36
34
32
SO
28
26
21
22
20
18
16
14
12
10
-------
8 9
C02, percent
ISO
170
160
150
140
130
120
NO
100
90
80
70
60
50
40
30
20
10
0
I
HC-R
HC-T
I HC-A |
36
34
32
30
28
26
24
22
20
18
16
14
12
10
a
6
4
2
0
C02, percent
Figure H-18. Emissions for Residential Unit 34 as Function of CO2
-------
s
789
C02, percent
170
ISO
150
140
130
120
110
100
90
80
70
60
50
40'
30
20
10
C02, percent
Figure 11-19. Emissions for Residential Unit 35 as Function of CO2
36
34
32
30
28
26
24
22
20
IB
ie
M
12
10
8
6
4
2
12
00
-------
II-59
In view of these observations, the tuning procedure adopted for this program and used by
many experienced servicemen for normal adjustments appears to be as good a means of achieving
low overall emissions as can be devised using field instruments for smoke and CO2 measure-
ments. A knowledge of the smoke-versus-CO2 curve for a particular burner permits adjusting the
burner to a low smoke-high CO2 level to both minimize emissions and operate efficiently.
Figure II-20 shows smoke versus CO2 curves for all units in the as-found and tuned
conditions, respectively. These figures further illustrate the wide range of combustion character-
istics exhibited by the individual units.
For some burners, the adjustment limits for low-emission operation were very narrow. In
several cases, the fan capacity limited the operating range such that CO2 levels could not be set
below 8 or 9 percent.
Limits of Acceptable Adjustment
Table II-21 shows a summary of the acceptable operating ranges for the various units,
based on the following criteria without regard to CO, HC, or NOX emissions:
Minimum C02:8 percent (8 percent CO2 was arbitrarily selected as the
minimum CO2 desired to achieve acceptable efficiency)
Maximum CO2 :air adjustment for No. 2 smoke.
The acceptable operating range by these criteria averaged 1.26 percent CO2 and exceeded 3
percent in two cases, but many units yielded negative values, which indicated that No. 2 smoke
was exceeded at 8.0 percent CO2.
TRIAL CORRELATIONS OF SMOKE VERSUS PARTICULATE
In the Phase I program, the possible correlation of steady-state Bacharach smoke numbers
with cyclic particulate emissions was investigated, but no direct relationship was found.* The
Phase II residential data exhibit the same lack of correlation, as will be seen by several plots.
Figures 11-21 and 11-22 show the attempted smoke versus particulate emission correlation
for all Phase I and II data combined. No relation was found, considering either filterable or total
particulate.
Figures II-23 and 11-24 show the effect of tuning on the smoke-part'iculate relationship
for Phase II units. A consistent slope of the connecting lines for the different units would suggest
characteristic curves for individual burners, but even this relation was not observed.
One explanation for the lack of correlation between smoke (at 5 or 9 minutes) and
particulate (integrated over the cycle) may be the particulate associated with start-up. Although
data are not available to determine the contribution of particulate generated during start-up to
* Limited Phase I data indicated a correlation between the carbon content of the particulate filter catch and
smoke up to about No. 7 smoke, where the smoke spot apparently became "overloaded".
-------
ON
o
Figure II-20. Emissions for Residential Units Firing House Fuels
-------
11-61
Table 11-21. Evaluation of Unit Performance
Unit
23
24
25
26
27
28
29
30
31
32
33"
34
35
Average
All
Condition
A
T
R
A
T
R
A
T
R
A
T
R
A
T
R
A
T
R
A
T
R
A
T
R
A
T
R
A
T
R
A
T
R
A
T
R
T
R
A
T
R
All
CO2 Level at Which
No. 2 Smoke Occurs, %
9.2
10.5
10.1
8.3
9.3
9.8
9.4
10.0
10.0
10.5
10.6
10.5
9.4
9.7
8.2
7.0
9.1
9.0
6.9
8.0
7.7
8.2
7.4
7.6
9.9
10.7
10.7
8.3
11.0
10.1
6.9
8.2
8.2
9.0
10.5
10.2
11.1
10.6
-
-
Acceptable Range of
Operation, % CO2a
1.2
2.5
2.1
0.3
1.3
1.8
1.4
2.0
2.0
2.5
2.6
2.5
1.4
1.7
0.2
None
1.1
1.0
None
0.0
None
0.2
None
None
1.9
2.7
2.7
0.3
3.0
2.1
None
0.2
0.2
1.0
2.5
2.2
3.1
2.6
0.85
1.75
1.50
1.26
a Acceptable Range of Operation = CC>2 at which No. 2 smoke occurs -8.0.
b All smoke spots from this unit appeared to include raw fuel.
-------
l*f
12
o
0
O
0 10
.0
*"
g 8
"58
E
UJ
o, 6
J5
o
CD
0
£ 2
(z
n
Legend
"o Phase I o
_A Phase n
1
-
- A
_ O
0 0
t
~ £" ^ ^
feo°° ° 8
Ȥ o 8
1,1,1,1
"02
Bacharach Smoke Number at 5 Minutes
12
o
en
,0
o
~ 8
O
a>
5
3
o
Q-
0
Legend °
o Phase I
_A Phase n
A A
- A
1
02468
Bacharach Smoke Number at 5 Minutes
Ox
Figure 11-21. Relation of Smoke Number and Filterable
Particulate Emissions Phase I and II
Residential Data
"Modified EPA Procedure.
Figure 11-22. Relation of Smoke Number and Total
Particulate Emissions Phase I and II
Residential Data
*Modified EPA Procedure.
-------
8
c
o
in
tn
'£
UJ
o
Q_
D
O)
2468
Bochorach Smoke Number at 5 Minutes
Figure 11-23. Effect of Tuning on Smoke and Filterable
Particulate Emissions for Individual
Phase II Residential Units
*Modified EPA Procedure.
14
12
Legend
As-found
o Tuned
I
2468
Bacharach Smoke Number at 5 Minutes
Figure 11-24. Effect of Tuning on Smoke and Total
Particulate Emissions for Individual
Phase II Residential Units
*Modified EPA Procedure.
-------
11-64
the particulate integrated over the whole cycle, examination of the smoke readings taken at the
1,5, and 9-minute points of the "on" cycle (these data are reported in Table 11-22) suggests that
emissions may be extremely high during the first minute or two of firing.
Bacharach smoke samples, generally taken during essentially steady-state operation near
the end of a cycle, do not include starting and shutdown as do the particulate samples. These
transients show as peaks on gaseous monitors, as "smoke puffs" on the Von Brand continuous
tape smoke meter, and as high smoke reading at the 1-minute point in Table 11-22. These
transients vary widely in degree for different units. The variation can be due partly to ignition
delay and fuel pump cutoff characteristics, with the latter believed to be the most critical.
It should be recognized that only one particulate measurement was taken on each unit
for each condition (as found and tuned), so there was no opportunity to determine if a
correlation might exist between smoke and either steady-state or cyclic particulate for individual
units without other changes.
Among the aspects which may be significant in the lack of correlation already observed
between particulate weight and Bacharach smoke are particle size, density, and light reflection
characteristics. It is hoped that particle-size measurements planned for a residential unit as
follow-on to this program will help to clarify this aspect.
The planned experimental program will include particulate measurements taken during
both cyclic and steady-state operation and may provide additional insight on the relation
between smoke and these particulate levels for residential units.
-------
11-65
Table 11-22. Smoke Reading at 1, 5, and 9
Minutes for Cyclic Runs on
Residential Units
Bacharach Smoke
Unit
23
23.1
23.2
24
24.1
24.2
25
25.1
25.2
26
26.1
26.2
27
28
29
Condition
A
T
R
T
R
T
R
A
T
R
T
R
T
R
A
T
R
R
R
T
R
A
T
R
T
R
T
ft
A
T
R
A
T
R
A
T
R
1 Min
0.7
0.6
0.4
_
-
_
0.4
3.0
2.5
-
_
-
2.1
1.7
1.6
_
-
_
0.9
0.9
0.6
_
-
_
-
1.8
0.8
0.4
1.8
1.1
1.0
9.0
2.3
2.9
5 Min
0.3
0.5
0.3
0.1
0.2
0.0
0.0
0.3
0.7
1.4
0.3
0.2
0.5
0.7
1.2
1.4
0.7
0.7
0.2
4.1
5.7
0.4
0.6
0.4
0.2
0.2
0.5
0.3
1.1
0.4
0.4
0.6
0.4
0.4
9.0
0.7
0.6
Number
9 Min
0.3
0.5
0.4
0.1
0.1
0.0
0.0
0.4
0.7
0.8
0.2
0.1
0.2
0.7
1.0
1.7
2.0
0.4
0.1
4.1
5.8
0.5
0.4
0.5
0.2
0.2
0.5
0.4
1.0
0.4
0.3
1.3
0.4
0.4
9.0
0.5
0.3
-------
11-66
Table 11-22. (Continued)
Eacharach Smoke Number
Unit
30
31
32
33
34
35
Condition
A
T
R
A
T
R
A
T
R
A
T
R
A
T
R
T
R
1 Min
0.4
2.0
2.5
0.2
0.3
0.3
0.7
0.2
0.1
Oily
Oily
Oily
3.0
1.7
1.7
0.3
0.2
5 Min
0.2
1.1
0.4
0.0
0.2
0.3
0.1
0.1
0.1
Oily
Oily
Oily
2.6
1.5
1.6
0.2
0.2
9 Min
0.2
0.6
0.3
0.0
0.2
0.2
0.2
0.1
0.1
Oily
Oily
Oily
2.7
1.0
1.9
0.2
0.2
-------
HI-1
EMISSIONS FROM COMMERCIAL BOILERS
This chapter describes the Phase II investigation of emissions from commercial boilers.
The principal effort of the Phase II program was on conducting a more detailed investigation of
the effect of fuel, load, and excess air on emissions than was possible during the Phase I study.
The discussion is presented as follows:
Boilers Included in the Phase II Investigation
* Procedures used in the Field Investigation
Emission Results for Commercial Boilers.
Where applicable, results of Phase I emission measurements1 have been included in the analyses
of the influence of various factors on emissions.
BOILERS INCLUDED IN THE PHASE II INVESTIGATION
The scope of boilers in the Phase II investigation included six commercial boilers
ranging in size from 40 to 600 boiler horsepower firing various grades of fuel oils and natural
Basis for Selection of Equipment
A survey of the commercial boiler population was initiated to form the basis for the
selection of a representative mix of commercial boilers. A survey form (developed by the Battelle
project team and project consultant W. H. Axtman) was distributed by the American Boiler
Manufacturers Association (ABMA) to veteran observers in the commercial-industrial boiler
industry. The results of this survey, summarized by W. H. Axtman, are contained in Appendix B.
Selection of Equipment Mix and Individual Boiler
The selection of the Phase II boiler mix and specific boiler type/size combinations was
made by the Battelle-Columbus project team after reviewing results of the survey.
Boilers selected for inclusion in the Phase II study were intended to make the Phase II
sample representative of the existing boiler population. (Phase I boilers did not weigh in this
consideration due to the limited data obtained on these units.) It was decided that the Phase II
study of commercial boilers was to provide so much more detailed information about emissions
at various operating conditions, that the Phase II selections alone should be representative of the
existing boiler population. The selection of the equipment mix and specific boiler type/size
combinations was approved by the EPA Project Officer and the API SS-5 Task Force Steering
Committee.
Table HI-1 outlines the mix of commercial boilers included in the Phase I and II
investigations as representative of commercial heating boilers in the field.
-------
III-2
Table 111-1. Mix of Commercial Boilers Sampled in Phases I and
Distribution of
Sample by
Phases
Total, all boilersb
By Boiler Size
10-50 bhp
51-100 bhp
101-300 bhp
301 -600 bhp
By Boiler Type
Scotch
Firebox firetube
Cast iron
Watertube
Miscellaneous firetube
By Burner Type
Air atomizing
Pressure atomizing
Rotary atomizing
Natural gas
By Fuelb
No. 2 oil
No. 4 & 5 oilsc
No. 6 oil
Natural gas
1
7
0
2
4
1
6
1
0
0
0
4
2
1
1
2
4
2
1
II
6
1
3
1
1
2
1
2
1
0
4
2
0
6
6
5
3
6
Total
13
1
5
5
2
8
2
2
1
0
8
4
1
7
8
g
5
7
Percentage
Phase II
-
17
50
17
17
33
17
33
17
0
67
33
0
100
100
83
50
100
Total
-
8
38
38
15
62
15
15
8
0
62
31
8
54
62
69
38
54
Distribution of
U.S. Commercial
Boiler Population,
percent3
-
47
24
29
21
27
33
6
13
15
22
15
56
38
8
4
56
Commercial boilers are defined as between 50 and 300 bhp.
Numbers total more than "Total, all boilers" because some boilers were fired with more than one fuel.
Including Phase II reference fuel.
-------
III-3
After review of the data obtained on boiler population and sales, it was determined that
the six boilers be selected according to the size ranges of the survey: 10 to 50 bhp, 51 to 100
bhp, and 101 to 300 bhp. It was further decided that the two boilers selected in each size range
be representative of the most common types for the size range. Hence, a tentative selection was
made as follows:
Size, 10 to 50 bhp
Cast iron
Firebox firetube
Size, 51 to 100 bhp
Cast iron
Packaged Scotch
« Size, 101 to 300 bhp
Packaged Scotch
Firebox firetube.
These boiler selections were believed to be generally representative of boilers in the field today,
within the limits placed by the size of the sample.
Another factor in the selection of the boilers was their capability for firing more than
one fuel. The program plan was to fire each boiler with fuels as follows:
» No. 2 fuel oil
A low-sulfur (1.0 percent) residual fuel oil typical of that being marketed
on the East Coast
Conventional No. 5 or No. 6 residual fuel oil
Natural gas.
It was difficult to locate small boilers capable of firing heavier fuels; in fact, the smallest
cast-iron boiler that was available could not be fired satisfactorily with fuels heavier than No. 2
grade fuel oil, in spite of several attempts by the manufacturer to successfully fire heavier fuels
by switching burners. The larger cast-iron boiler could not fire conventional No. 5 and No. 6
grades of fuel oil but could fire No. 4 fuel oil and the low-sulfur residual fuel.
Midway in the program, it was learned that the tentatively selected 40-bhp firebox
firetube boiler could not be fired with fuels heavier than No. 2 fuel oil. At that point, a
substitution was made of a 600-bhp watertube boiler for the 40-bhp firebox firetube boiler. Thus,
the scope of the boilers actually measured was expanded in both boiler size and boiler types.
The EPA Project Officer and the API SS-5 Task Force Steering Committee concurred with this
substitution.
For purposes of this investigation, it was important to select boilers for which load could
be controlled at a desired level for extended periods to provide stable conditions for emission
measurements. This requirement was met, through the cooperation of ABMA, by choice of
equipment installed as house boilers or test boilers in plants of boiler manufacturers.* In addition,
competent personnel already familiar with the specific equipment were available to adjust the
boilers to the desired operating conditions. Normally, these boilers were in day-to-day operation
*Boiler C2003 was an exception. It was a newly constructed unit that had been fired for manufacturers tests
only. Emission measurements were made prior to its delivery by the manufacturer.
-------
ni-4
to supply steam, to provide service training, or to provide a burner test facility for the
manufacturing plants. These boilers did not appear to have received any better service attention
than would be reasonably typical of similar boilers in commercial or industrial applications.
Description of Commercial Boilers
Table III-2 shows the number designation for each commercial boiler, with identification
of boiler type and size, burner type, fuel grades, and other descriptive data.
PROCEDURES USED IN THE FIELD INVESTIGATION
The principal effort of Phase II of this investigation was devoted to more detailed field
measurements of emissions covering more conditions than was possible for the Phase I study.
Phase II was conducted during the heating season of 1971-72. The measurements were made by
a three-man field team supported by other Battelie-Columbus staff.
Commercial boiler measurements were conducted in April through June 1972, with each
boiler monitored for approximately 5 to 10 days, depending on conditions and fuels run.
The commercial boilers were all sampled in the as-found condition; i.e., no cleaning
(except for the .exhaust stacks) or other servicing of the units was done prior to measurements.
Wherever practical, the exhaust stacks were thoroughly cleaned prior to sampling to minimize
collection of previously deposited material during particulate sampling.
Conditions Investigated
Emission data from each of the commercial units were obtained under steady-state firing
conditions at several excess air levels, generally while firing each of four fuels at four different
load levels.
Load and Excess Air Conditions. Figure III-l shows the matrix of runs that was the
general plan or target for measurements on each boiler-fuel combination. A base-line condition,
the operating condition most typical of normal boiler operating conditions, was defined as 80
percent load and 12 percent CO2 (10 percent CO2 for gas firing)*. The target loads were
R - rated load
H 80 percent of load, considered the normal high-fire condition for
boilers in use
M - an intermediate load
L - the normal low-fire setting.
Boiler C2001, a relatively small cast-iron boiler, normally operated strieilj In the on-off mode;
hence, intermediate loads could only be achieved by changing nozzle sizes. The remaining boilers
were modulated to the desired loads.
The base-line condition was defined during the Phase I study cftur discussion with the ABMA Commercial-
Industrial Air Pollution Committee.
-------
Table 111-2. Description of Commercial Boilers Included in Phase II Study
Boiler
C2001
C2002
C2003
C2004
C2005
C2006
Boiler
Horsepower
and Type
40-hp
Cast I ron
90-hp
Cast Iron
300-hp
Scotch
80-hp
Firebox
100-hp
Scotch
600-hp
Watertube
Burner
Type
Pressure
Atomizing
Air
Atomizing
Air
Atomizing
Pressure
Atomizing
Air
Atomizing
Air
Atomizing
Normal
Fuel Other
Grade Fuels3
No. 2 Gas
No. 4 No. 2, CR, Gas
No. 6 No. 2, CR, Gas
No. 5 No. 2, CR, Gas
No. 6 No. 2, CR, Gas
No. 6 No. 2, CR, Gas
Rated
Input,
gph
11.35
26.1
33.5
22.5
28.0
165.2
Output
Rating,
MBtu/hr
1258
3000
10043
3341
3348
25000
Method of
Control
LFSC
on/off
Modulating
Modulating
LFSC
on/off
Modulating
Modulating
Operating
Pressure,
psig
15
15
150
15
15
150
Number
of Flue
Passes
2d
2d
3
3
4
2d
Furnace
Volume,
cult*"
19.7
50.3
86.7
40.8
25.4
413.4
a CR used to designate 1 percent sulfur reference fuel fired in commercial boilers.
b Manufacturer's data; volume may .include turn-around volume at rear of first pass.
c Lorn-fire start.
^ Conventional flue pass definition not appropriate. These values were selected as the best approximation for these units.
-------
III-6
M
H
1
Excess
Air
i
-
G
G
(b)
G+P
G
1
Low
Fire
1
G
G
G
G
1
Int.
Load
i
G
G
G
G+P
(Baseline
condition )
G
1
80%
Load
-
G
G
G
G
1
Rated
Load
M
H
R
Load
12%
CO,
Figure III-l. Matrix of Measurement Points for Commercial Boilers3
a. G = Gaseous measurements, plus smoke
P = Particulate measurements.
b. Particulate was run at low load for normal house fuel only.
Air adjustments were made after disconnecting the air-fuel proportioning linkage. To assure
a steady-state condition, the boilers were operated for at least 30 minutes at each load setting
before sampling was started. Particulate samples were obtained for each boiler-fuel combination
at the base-line condition. Additional particulate samples were obtained at the low-fire setting for
each boiler while firing the normal fuel for the boiler. Gaseous emission measurements were
obtained for each boiler-fuel combination at each load.
Fuels. As listed in Table III-2, each boiler was fired with the "house fuel" normally fired
in the boiler, No. 2 fuel oil, a 1-percent sulfur East Coast residual oil designated CR* (for
"commercial reference" oil), and natural gas. The commercial reference oil was typical of the
current commercially available East Coast fuels at this sulfur level. It was purchased in
Connecticut and transported from site to site in a 6000-gallon, steam-traced tank truck.
Properties of all fuel oils fired in commercial boilers are tabulated in Appendix C, Tables
C-3 and C-4.
Boiler C2001, a small cast-iron boiler which could not fire heavy fuel oils, was fired with
only No. 2 fuel oil and natural gas.
*The designation CR is used to denote the particular fuel oil used as the commercial reference fuel in this study.
LSR is used elsewhere in this report as a designator for the general class of low-sulfur residual fuel oils (1 per-
cent sulfur) produced to meet local air pollution regulations^
-------
III-7
Emission Measurements Instruments and Techniques
The instrumentation and emission measurement techniques employed for the commercial
units were similar to those used for the residential units.
An established set of operational procedures was routinely followed for each boiler
investigated. Stack gases were sampled and supplied directly to continuous monitoring equipment
set up alongside the heating unit.
Gaseous Emissions. Measurements were made of the following stack emissions under
various conditions of operation using the methods noted as follows. (Details of instrumentation
and measurement procedures are described in Appendixes E and F.)
Emission Measurement Method
CO2 NDIR (nondispersive infrared) and
Fyrite
O2 Amperometric and Fyrite
CO NDIR
Hydrocarbons (total) . . , Flame ionization
SO2 Dry electrochemical
NOX Dry electrochemical and NDIR with
converter
NO NDIR
Particulate EPA sampling train.
Smoke and Other Combustion Parameters. Smoke was measured using the Bacharach
hand pump smoke meter and the standard procedures adopted by ASTM7. In addition to
these measurements, other combustion conditions measured included:
Firing rate, by weight measurement
Stack temperature, measured in the flue at the particulate sampling location
Stack draft, measured in the flue at the particulate sampling location.
-------
III-8
Paniculate Emissions. Participate emissions were sampled using the EPA sampling rig2'8.
This train is described in Appendix F. A special feature of this sampling train is the inclusion of
two water impingers or bubblers (at 70 F) downstream from the filter. The impingers were
originally intended to collect any condensable material (at 70 F) that exist as vapor at filter
temperature and, thus, pass through the filter and any solid particulate that pass through the
filter. Recently, EPA has abandoned use of the impinger catch in determining particulate emis-
sions for power plants8 as there was indication that reactions occur in the impinger to
generate material that is included in the weight measurement of particulate, even though the
material does not exist as particulate either in the flue gas or in the atmosphere9.
To insure that the most meaningful information was obtained from the particulate
samples collected on this investigation, the probe wash, the filter catch, and the impinger wash
were treated separately, and particulate weights were recorded for each. In this report, particulate
data are reported as filterable (including the probe and filter catches) and total (combining the
filterable and the material found in the water impingers). However, it should be pointed out that
even the filterable catch obtained using the EPA sampling train may not be directly comparable
to the particulate catch obtained using other sampling trains because of differences in the
procedures for washing the probe and differences in sampling rates and volumes.
Previous experience at Battelle suggests that acetone washing may not be adequate to
remove all particulate from the probe. Therefore, for this investigation two washing procedures
were used: (1) the EPA procedure and (2) a modified EPA procedure referred to as the MEPA
procedure. First, the probe, filter holder, and impingers were washed using procedures specified
by EPA. Then a modified procedure was used wherein additional washings were made to insure
complete removal of deposited particulate. The complete procedures and the resulting data are
discussed further in Appendix F.
EMISSION RESULTS FOR COMMERCIAL BOILERS
This section summarizes the results obtained from measurements of gaseous and
particulate emissions and smoke made on six commercial boilers during the Phase II investiga-
tion. Measurements were made on each boiler at several excess air levels at each of several loads;
most were fired with four fuels. Results for the base-line condition (80 percent load and 12 per-
cent CO2) and analyses of the effect of important variables on emissions are presented in this
section. Complete data for all runs for each boiler, each fuel, each load, and each excess air level
are tabulated in Appendix H of the Data Supplement Volume.
In addition to the emission results, this section contains correlations showing the in-
fluence of various parameters (such as load, excess air, or fuel properties) on emissions.
Summary of Emission Data and Emission Factors
Table III-3 contains a summary of emission data for the commercial boilers. Table HI-4
shows emissions in lb/1000 gal oil for the gaseous pollutants (CO, HC, NOX, and SO2) and for
particulate for each boiler-fuel combination at the base-line condition.
-------
Table 111-3. Summary of Emissions From Commercial Boilers at the Base-Line Conditions3
Operational Data
Commercial Fuel
Boilerb Grade
C2001 No. 2
Gas
C2002 No. 2
No. 4
CR
Gas
C2003 No. 2
CR
No. 6
Gas
C2004 No. 2
CR
No. 5
Gas
C2005 No. 2
CR
No. 6
Gas
C2006 No. 2
CR
No. 6
Gas
Fuel
Temp,
F
-
_
_
195
-
88
143
208
-
_
118
120
-
85
110
206
-
70
127
212
Firing
Rate,
gphc
9.3
1.26
21.6
20.6
20.4
3.04
71.5
66.0
65.5
10.0
19.2
18.3
18.4
2.40
23.6
22.5
21.8
3.21
142
143
139
20.1
Load,
%
80
81
78
78
78
82
80
77
79
83
81
82
85
76
76
77
77
77
79
84
84
83
CO2,
%
12.1
9.1
11.9
12.3
12.1
10.0
12.1
12.0
11.9
10.0
12.0
12.1
12.0
10.0
12.0
12.1
12.1
10.0
11.9
12.0
12.1
9.8
02.
%
4.4
5.0
5.3
4.3
4.4
3.3
4.9
4.5
5.0
3.2
4.5
4.2
4.8
3.6
4.3
4.5
5.3
3.0
3.8
4.0
4.1
2.4
Excess
Air,
%
26
27
30
25
26
16
27
27
30
16
26
25
30
17
26
26
31
16
25
25
27
15
Bacharach
Smoke No.
2.4
0.2
0.6
2.3
3.0
0.1
2.0
3.1
4.2
0.0
0.2
2.9
5.0
0.0
0.0
2.5
3.8
0.1
0.4
3.4
4.1
0.0
Emissions
Gases, ppm
CO
14
51e
4.0
2.0
0
1.0
0
2.0
9.0
5.0
0
0
0
8.0
0
0
5.0
67
0
0
0
0
HC
1.5
35e
0.5
-
0.3
3.0
1.2
3.2
1.0
1.8
5.3
3.7
3.0
2.6
1.0
1.0
8.0
4.7
6.2
4.2
3.2
NO
61
71e
75
247
257
93
70
189
250
113
103
251
291
47
125
238
29?
63
IH 4
210
283
95
NOX
62
75e
74
251
257
94
70
190
250
116
104
249
295
50
127
243
301
65
123
208
266
95
S02
142
Oe
140
900
520
0
86
520
1150
0
102
500
1160
0
180
550
930
0
86
520
660
2
Paniculate",
mg/sm3
Filterable
11.0
5.2e
11.3
39.9
106
4.8
15.9
292
738
10.0
15.8
73.7
144
7.7
10.8
49.0
108
8.1
6.6
71.0
246
7.2e
Total
50.1
18.0e
22.7
57.5
122
15.5
29.6
313
762
24.5
24.9
110
210
13.4
43.1
74.6
160
23.0
20.8
99.9
297
14.8e
(a) Base-line condition is 80 percent and 12 percent C02 (10 percent CO2 for gas firing).
(b) See Table 111-2 for descriptions of boilers.
(c) Firing rate is 10^ cfh for gas firing.
(dl Paniculate by modified EPA procedure.
(e) Emission data at 9.0 percent CCb.
-------
111-10
Table 518-4. Summary of Emission Factors for Commercial Boilers
Emission Factors, ib/1000 gal°
Commercial
Boiler b
C2001
C2002
C2003
C2004
C2005
C2006
Fuel
Grade
No. 2
Gas
No. 2
No. 4
CR
Gas
No. 2
(CR
No. 6
Gas
No. 2
CR
No. 5
Gas
No. 2
CR
No. 6
Gas
No. 2
CR
No. 6
Gas
Bacharach
Smoke No.
2.4
0.2
0.6
2.3
3.0
0.1
2.0
3.1
4.2
0.0
0.2
2.9
5.0
0.0
0.0
2.5
3.8
0.1
0.4
3.4
4.1
0.0
Gaseous Emissions
CO
1.7
6.6e
0.5
0.3
0.0
0.1
0.0
0.3
1.3
0.6
0.0
0.0
0.0
1.0
0.0
0.0
0.7
7.8
0.0
0.0
0.0
0.0
HC
0.11
2.606
0.04
-
-
0.02
0.21
0.09
0.25
0.07
0.13
0.40
0.29
0.2C
0.18
0.08
0.08
0.54
0.33
0.46
0.33
0.21
NOX
12.5
16.0s
15.7
53.2
55.3
18.2
14.4
41.2
56.9
22.4
21.0
53.7
67.0
9.8
25.6
52.4
70.2
12.5
24.6
44.9
59.2
18.1
SO2
40.0
O.O6
40.8
265.6
155.8
0.0
24.6
156.8
364.3
0.0
28.6
148.9
366.8
0.0
50.4
165.2
301.8
0.0
23.8
154.6
?'"* 6
0.0
Participated
Filterable
1.2
0.6e
1.3
4.5
12.1
0.4
1.7
33.6
89.1
1.0
1.7
8.4
17.3
0.7
1.2
5.6
13.3
0.9
0.8
8.0
29.1
0.8e
Total
5.4
1.9e
2.5
6.5
13.9
1.5
3.2
35.9
92.1
2.5
2.7
12.5
25.3
1.3
4.6
8.5
19.8
2.2
1.4
11.3
35.1
1.6e
(a) Base-line condition is 80 percent load and 12 percent CO2 (10 percent CO2 for gas firing).
(b) See Table 111-2 for descriptions of boilers.
(c) Emission factors for gas firing are given as an equivalent tharmal basis to Ib/1000 gal of oil.e.i., in lb/(145 x 108 Btu).
(d) Paniculate by modified EPA procedure.
(e) Emission factor at 9.0 percent C02-
-------
III-l 1
CO and HC emissions are generally low except for occasional gas-fired runs, particularly
for Boilers C2001 and C2005. The reason for these occasional high emissions is not evident
and it is not reflected in smoke number, as the smoke was very low for all gas runs listed in this
table. SO2 emissions are related directly to fuel sulfur and are relatively unaffected by combus-
tion conditions.
The emissions that are affected by boiler design or operation (namely, NOX, smoke,
and particulate) are of prime interest and are discussed separately below.
Influence of Various Parameters
on NOX Emissions
NOX emissions from combustion sources are known to be generated by two processes
oxidation of organic nitrogen bound in the fuel and by thermal fixation of nitrogen in the
combustion air in the furnace environment. Hence, NOX emissions are related to nitrogen
content of the fuel1 °'11, combustion temperature, and the relative competition for oxygen atoms
(i.e., whether conditions are oxidizing or reducing and to what extent12. Therefore, it is useful
to examine the influence on NOX emissions of fuel nitrogen and the major factors which
influence combustion temperature and oxygen availability (such as load and excess air). The
effect of these factors on NOX emissions was investigated in this program and the results are
discussed below.
In addition to considering the influence of the obvious parameters on NOX emissions, a
regression analysis was used to determine the best equation relating the NOX data for the
base-line condition to additional fuel and equipment variables. Variables included in this analysis
were as follows:
Fuel Variables Boiler Variables
Fuel nitrogen, N Firing rate
N1 /2 Combustion volume
API gravity Combustion intensity, CI
Specific gravity (firing rate per unit combustion volume)
Carbon residue (Ramsbottom)
Viscosity at firing temperature, V (SSU)
Carbon content; C
Hydrogen content, H
C
H
C
N
C
H+N
, G. B., and Berkau, E. E., "An Investigation of the Conversion of Various Fuel Nitrogen Compounds
to Nitrogen Oxides in Oil Combustion", presented at AIChE Meeting, Atlantic City, N.J., August 30, 1971.
^Turner, D. W., Andrews, R. L., and Siegmund, C. W., "Influence of Combustion Modification and- Fuel
Nitrogen Content on Nitrogen Oxides Emissions From Fuel Oil Combustion", presented at AIChE Meeting,
San Francisco, November 28-December 2, 1971.
12Bartok, W., Crawford, A. R., Cunningham, A. R., Hall, H. J., Manny, E. K., and Skopp, A., "Systems Study of
Nitrogen Oxide Control Methods for Stationary Sources", Final Report, Esso Res. & Engrg. Co., November 20,
1969, NAPCA Contract PH-22-68-55.
-------
Ill- 12
The best equations that could be fit to the NOX data using these variables were as follows:
Using only one variable
NOX (ppm at 3 percent O2) = 89.97 + 834.7-N (percent)
Using more than one variable
NOX (ppm at 3 percent O2) = 9.10 + 579.2-N (percent) + 0.373-V (SSU) + 101.3-CI
(Btu/ft3).
The correlation of NOX with one variable, N, was 0.81. The correlation coefficient obtained
using the three variables listed in the above equation was 0.85.
Correlation coefficients with NOX emissions obtained when each of the above listed
variables was considered are as follows:
Correlation Coefficient
Variable (Absolute Value)
Fuel nitrogen, N 0.81
Ny" 0.79
API Gravity 0.77
Specific gravity 0.77
Carbon residue, Rarnsbottom 0.73
Hydrogen content, H 0.71
I °-70
°-69
°-63
Viscosity at firing temperature, SSU 0.62
Carbon content, C 0.45
Firing rate per unit combustion chamber volume 0.42
Firing rate 0.10
Combustion chamber volume 0.07
It should be noted that the correlation coefficients were higher for all the fuel variables
considered than for the boiler-related variables. Thus, either the fuel is the significant factor in
NOX formation (which is certainly true for NOX formed from fuel nitrogen) or the most
appropriate boiler variables were not considered (true to the extent that flame characteristics
were not available).
NOX and Fuel Nitrogen, Due to the complexity of NOX emissions (with the therm al NOX
varying from boiler to boiler and the fuel NOX varying from fuel to fuel), several different
approaches were used to examine the relationship of. NOX and fuel nitrogen.
Figure III-2 shows the measured NOX emissions plotted against fuel nitrogen for all
Phase I and II runs firing fuel oils in commercial boilers at the base-line conditions (80 percent
-------
111-13
600
500
400
30O
200
100
0-1
02 0.3 04 0.5
Nitrogen in Fuel,weight percent
0.6
0.7
Figure HI-2. Relation of NOX to Fuel Nitrogen for Phases I and II,
Commercial Boilers Fired at Base-Line Conditions
-------
Ill-14
load and 12 percent excess air). Except for a few data points, a strong relationship is evident
between NOX emissions and fuel nitrogen. The curve obtained from the regression analysis (with
one variable term) described above is also shown in Figure II1-2. This curve has a slope
equivalent to a 65 percent conversion of fuel nitrogen to NOX.*
Figure HI-3 shows the relationships between NOX emissions and fuel nitrogen for each of
the six boilers included in the Phase I and II studies that were fired with more than one fuel.
The slopes of these curves are generally similar to the slope of the line showing a conversion rate
of 60 percent of the fuel nitrogen to NOX. No correlation is evident between the shapes of the
curves and/or the zero-fuel-nitrogen intercepts and the boiler variables, such as firing rate or
combustion intensity.
To obtain a better understanding of the influence of fuel nitrogen on NOX emissions, the
following analysis was performed using data for boilers in which No. 2 fuel oil (a low nitrogen
fuel) and heavier oils were fired. The thermal NOX for each boiler was estimated by assuming
100 percent conversion of all fuel nitrogen from firing of No. 2 fuel oil and subtracting this fuel
nitrogen term from the NOX measured when firing No. 2 fuel oil. The fuel nitrogen term for
each boiler and fuel combustion was estimated by subtracting the thermal NOX term for that
boiler from the measured NOX value. The fuel nitrogen components were plotted against fuel
nitrogen, see Figure HI-4, and a curve was drawn to fit the data by eye. The best curve to fit the
data was of the form NOX (ppm) = Thermal NOX + 420-N0 -6. The overall thermal NOX term
was estimated by averaging the thermal NOX terms for each boiler: the average value is 97 ppm.
Hence, the best estimate of NOX from the Boilers C2001 through C2006 (the boilers in which
No. 2 and heavier fuel oils were fired was:
NOx, ppm= 97 +420 N°-6
(In the range of fuel nitrogen from 0.0 to 0.06 percent, this equation predicts NOX from fuel
nitrogen that exceed 100 percent conversion of fuel nitrogen but the maximum error is about 10
ppm, less than the variation in thermal NOX from one boiler to another.)
NOX and Excess Air. Figures III-5 through III-10 show the sensitivity of NOX emissions
to excess air for all fuels and boilers at 80 percent load, NOX emissions are expressed in lb/106
Btu for these figures, as this is the most meaningful value when comparing widely varying fuels:
e.g., natural gas and No. 6 fuel oil.
Examination of the relationship of NOX to excess air, shown in Figures III-5 to 111-10,
suggests that NOX tends to be relatively constant or increases slightly with excess air.
Apparently, the increase in oxygen concentration promotes NOX formation to a greater degree
than the increased excess air tends to cool the flame. In nearly all cases, the variation of NO
with excess air within the normal range of excess air 15 to 30 percent was 12 percent or
less. (The exceptions were firing No. 4 fuel oil and the CR reference fuel in Boiler C2002, where
the variation of NOX with excess air was large inthe 15 to 20 percent excess air range.)
It is interesting to compare the NOX emission shown in Figures III-5 to III-10 with the
EPA standards for NOX emission from new fossil-fuel power plants2. These standards permit
NOX emissions of 0.20 lb/106 Btu for gas-fired boilers and 0.30 lb/106 Btu for oil-fired boilers.
From the commercial boilers and fuels examined, these .standards were met in each case for
No. 2 fuel oil and gas firing. Conversely, for each boiler the NOX emission standard for oil firing
could not be met when firing conventional No. 5 or 6 fuel oils. When firing the commercial
*One percent fuel nitrogen converted to NOX yields about 1280 ppm of NOX at 3 percent O2.
-------
111-15
450
400h
350h
-------
Ill-16
400
O.I
0.2
0.3 0.4 0.5
Fuel Nitrogen,percent
0.6
0.7
Figure 1H-4. Relationship of NOX from Fuel Nitrogen to Fuel Nitrogen for Boilers
Firing No. 2 and Heavier Fuel Oils
-------
o*
-. 025
c
o
3
u>
m 0 20
Gas
I I I I I I
10 2O 30
40 50
Excess Air, percent
00 0.35
"o
^ 0.30
a.
3
B 025
I
6 0.20
ox
CR Oil
Gas
60 70 80 90 0 10 20 30
J I
J I
40 50 60
Excess Alr.percent
70 80 90 100
Rgure III-5. Relation of NOX Emissions to Excess Air for
Boiler C2001 at 80 Percent Load
Figure III-6. Relation of NOX Emissions to Excess Air for
Boiler C2002 at 80 Percent Load
40-bhp Cast Iron Boiler.
90-bhp Cast Iron Boiler.
-------
0.60
i 0 30
(A
S
£
S 0.25
NO. 6 Oil
005
Gas
10 20 30
10 50 60
Excess Air,percent
70 80 90
:, 035
S
0* 030
No. 5 Oil
CROil
No, 2 Oil
10 20 3O
40 50
Excess Air, percent
00
TO 8O 90
Figure III-7. Relation of NOX Emissions to Excess Air for
Boiler C2003 at 80 Percent Load
Figure III-8. Relation of NOX Emissions to Excess Air for
Boiler C2004 at 80 Percent Load
300-bph Scotch Boiler.
80-bhp Firebox Boiler.
-------
040
S
a
0,5
e
.a
J L
No. 6 Oil
30 « 50 SO
Excess Air, percent
J L
70 80 §0
lit
o"
No. 6 Oil
J L
J L
10 20 30 <*O 50 50 70 60 90
Excess Air, percent
Figure III-9. Relation of NOX Emissions to Excess Air for
Boiler C200S at 80 Percent Load
Figure 111-10. Relation of NOX Emissions to Excess Air for
Boiler C2006 at SO Percent Load
100-bhp Scotch Boiler.
600-bhp Watertube Boiler.
-------
HI-20
reference fuel, CR, the standard could be attained at low excess air levels in some boilers but
could not be attained in other boilers.
NOX and Load. Figures III-l 1 through 111-16 show the relationship between NOX
emissions and load for each boiler and each fuel for the nominal 12 percent CO2 condition.
Load does not seem to be a critical factor influencing NOX emissions. For most boilers, the NOX
emissions were relatively constant over the range of load investigated. Boilers C2002 and C2003
exhibited a trend of NOX increasing with load for the heavier fuels. NOX emissions from Boiler
C2005 were quite high at low load.
NOX for Gas Firing. As expected, gas firing of the commercial boilers produced relatively
low NOX emissions. The NOX emissions from boilers fired along the East Coast (Boilers C2001,
C2002, and C2003 were located in New England and Eastern Pennsylvania) averaged 99 ppm at
3 percent O2. NOX emissions from boilers in the Midwest (Boilers C2004, C2005, and C2006
were located within 200 miles of Chicago) averaged 70 ppm.
It is interesting to compare NOX emissions for gas firing with those for firing with No. 2
fuel oil, as shown in Table III-5. For the boilers located in the East, the NOX from gas firing
exceeded that from No. 2 fuel oil. Conversely, in the boilers located in the Midwest, the NOX
from firing No. 2 fuel oil exceeded that from gas. The fuel nitrogen levels of the No. 2 fuel oils
fired in the Midwest were slightly higher than those for oils fired in the East. However, the
difference would only account for 3 ppm at a conversion efficiency of 100 percent.
Table 111-5. NOX Emissions for Phase II Boilers Firing Natural
Gas and No. 2 Fuel Oil
NOX Emissions, equivalent ppm at 3 percent
Oj and 80 percent load
Boiler
C2001
C2002
C2003
C2004
C2005
C2006
Gas
84
95
117
51
65
95
No. 2 Fuel Oil
67
85
77
114
138
134
N Content
No. 2 Oil, percent
0.0084
0.0084
0.0078
0.0113
0.0124
0.0086
The relative differences in NOX emissions from gas- and oil-fuel boilers in the East and
Midwest may be attributable to different burner adjustment practices in the two regions or some
other subtle variable(s).
-------
111-21
O.60
0.55
0.50
0.45
S 0.40
"o
O 0.35
z
(A
2,
§ 0.30
jj
(A*
.2 0.25
1
UJ
o" 0.20
0.15
0.10
0.05
Gas (9 % COZ)
_ c
I
I
10
20
30
40
50 60
Lood, percent
70
80
90
100
no
Figure III-ll. Relation of NOX Emissions to Load for Boiler C2001 at 12 Percent CO2
40-bhp Cast Iron BoUer.
-------
111-22
0.60
0.55
0.50
045
£ 0.40
"Q
-^
d* 0.35
§ 030
-Q
tf>
'E
bJ
O 0.20
z
0.15
0,10
0.05
10
4 Oil
Gas (10% C02)
No. 2 Oil
1
20 30
40
50 60
Load, percent
70
80 90 100 110
Figure III-12. Relation of NOX Emissions to Load for Boiler C2002 at 12 Percent CO2
90-bhp Cast Iron Boiler.
-------
111-23
0.60
0.55
0.50
0.45
S 0.40
"2
"x
5- 0.35
§ 0.30
.0
% 0.25
.<£
LU
C? 0.20
0.15
0.10
0.05
No. 6 Oil
Gas (a1 10% C02)
I
I
I
I
I
I
I
This point at
13.2% C0e
I
10 20 30 40 50 60 70 80 90 100 110
Load, percent
Figure 111-13. Relation of NOX Emissions to Load for Boiler C2003 at 12 Percent CO2
300-bhp Scotch Boiler.
-------
111-24
0.60
0.55
0.50
0.45
£ 0.40
C
O* 0.35
8
g 0.30
.o
vt
I 025
.<£
LJ
cT 0.20
0.15
0.10
0.05
No. 5 Oil
No. 2 Oil
Gas(otlO%C02)
I
J
10 20 30
40
50 60
Load, percent
70 80 90 100 110
Figure 111-14. Relation of NOX Emissions to Load for Boiler C2004 at 12 Percent CO2
80-bhp Firebox Boiler.
-------
111-25
0.60
0.55
0.50
0.45
£ 0.40
"Q
-»
cf 0.35
0.30
M 0.25
e
LJ
O 0.20
0.15
0.10
0.05
I
No. 2 Oil
Gas (at 10% C02)
J I I I
J I
I
0 10 20 30 40 50 60 70 80 90 100 110
Load, percent
Figure 111-15. Relation of NOX Emissions to Load for Boiler C2005 at 12 Percent CO2
100-bhp Scotch Boiler.
-------
111-26
0.60
0.55
0.50
0.45
£ 0.40
"o
& 0.35
tn
_D
§ 0.30
-O
I 0.25
UJ
0.20
0.15
0.10
0.05
I
No. 6 Oil
CR Oil
No. 2 Oil
J I
J I
I
I
10 20 30 40 50 60 70 80 90 100 110
Load, percent
Figure 111-16. Relation of NOX Emissions to Load for Boiler C2006 at 12 Percent COa
600-bhp Watertube Boiler.
-------
IH-27
Influence of Operating Condition on
Smoke. CO, and HC Emissions
In this program, two parameters were examined that defined operating conditions -
excess air and load. The following discussion presents results showing the influence of each of
these parameters on smoke, CO, and HC emissions. (Their influence on NOX emissions was
described earlier.)
Excess Air. Figures III-17 through 111-22 show the relation of smoke and CO emissions to
excess air for each fuel fired hi each boiler at 80 percent load. HC emissions are not shown on
these figures, as they were nearly always low; when high HC emissions were measured, the CO
emissions were also high. Hence, CO was selected to illustrate emission trends for both CO and
HC.
Figures 111-17 through 111-22 show that about 30 percent excess air is required to fire
most fuel oils hi the Phase II boilers to avoid smoke levels above about Bacharach No. 2 or 3. In
a few cases (No. 6 fuel oil in boiler C2003 and No. 5 fuel oil in boiler C2004), the smoke
remained high even at excess air levels above 30 percent.
Gas could generally be fired at 10 percent excess air without exceeding a No. 1 smoke
level. However, 20 percent excess air generally was required to avoid high levels of CO emissions.
Smoke Number as an Indicator of CO and HC Emissions. Examination of the curves of
smoke, CO, and HC emissions plotted against excess air* shows that, as the excess air was
decreased for oil-fired boilers, the smoke level tended to increase before the CO and HC
emissions increased. However, as the excess ah- was decreased for gas-fired boilers, the CO and
HC emissions generally increased before the smoke level increased. Hence, setting the excess air
for an oil-fired boiler by smoke number will tend to keep CO and HC emissions at a minimum.
Conversely, setting the excess air level by smoke number for gas-fired boilers does not insure
minimum CO and HC emissions.
Load. Generally, load did not significantly influence emissions of CO and HC. For two
boilers (C2003 and C2004), smoke tended to increase as load was decreased at constant CO2.
For the remaining boilers, changes in load did not significantly affect smoke.
Based on conversations with commercial boiler/burner specialists - namely, members of
the ABM A Commercial-Industrial Air Pollution Committee - it was determined that commercial
boilers generally operate most of the time at one of two loads a high load of about 80 percent
of rated load and a low fire condition. Once a boiler is in service, the serviceman is likely to
adjust the burner to produce low smoke at these two normal operating conditions. Hence, the
influence of load on smoke (and to a lesser extent, CO and HC) may be an indication of how
that particular burner was adjusted, rather than a general measure of the capabilities of the
boiler/burner combination. Burners frequently are adjusted to operate at higher excess air levels
at low load than at the normal high load.
Influence of Fuel on Smoke Emissions
Notwithstanding the fact that this study has been the most extensive field investigation
ever conducted of emissions from residential units and commercial boilers, there are relatively
"These curves for each boiler, fuel, and load are presented in Appendix H as the Data Supplement Volume.
-------
I
2 4
s
g
-------
.2 0.150
in
'i
0 10 ZO 50
Excess Air.percent
40 50 60
Excess Air.perccnt
70 90 90 100
Figure 111-18. Relation of Smoke and CO Emissions to Excess Air for Boiler C2002 at 80 Percent Load
90-bhp Cast Iron Boiler.
-------
10 20
30 40
Excess Air,percent
60 70
0.300
0.250
£ 0 150
I
UJ
g 0.125
No.6 Oil
No. 2 Oil all values
were zero
10 20 30 40 50
Excess Air.percent
Figure 111-19. Relation of Smoke and CO Emissions to Excess Air for Boiler C2003 at 80 Percent Load
300-bhp Scotch Boiler.
1=1
T1
O
-------
fe 0.173
XI
0 0-150
E
Ul
_ 0-125
40 30
Excess Air,percent
10 20 30 4O 50 6O 70 80 90 IOO
Excess Air,p«rcent
Figure 111-20. Relation of Smoke and CO Emissions to Excess Air for Boiler C2004 at 80 Percent Load
80-bhp Firebox Boiler.
-------
0 10 2O 30 40 50 6O 7O
Excess Air, percent
30 40
Excess Air, percent
TO
Figure ffl-21. Relation of Smoke and CO Emissions to Excess Air for Boiler C2005 at 80 Percent Load
100-bhp Scotch Boiler.
-------
30 40
Excess Air, percent
u.^uv
O.275
O.25O
0.225
0.200
3
£
"g 0-175
« 0.150
.a
1 0125
8
0.100
O.075
0.050
0.025
0
-
-
-
-
Gas
i
_
CROil
No. Z Oil
1 No. 6 Oil
W\ 1, .0 . . 1, 1
10 20 30 40
Excess Air, percenl
UJ
OJ
Figure 111-22. Relation of Smoke and CO Emissions to Excess Air for Boiler C2006 at 80 Percent Load
600-bhp Watertube Boiler.
-------
111-34
few data available from this program, considering the large number of variables that might
influence emissions. However, examination of the relationship to emissions of the fuel variable
considered most likely to correlate with emissions was valuable. It was decided that API gravity
is probably the best readily available single property by which the combustion characteristics of
fuel oils can be classified. Therefore, several trials were made at correlating emissions with API
gravity of the fuels.
Figures IH-23 and 111-24 show relationships between Bacharach smoke number and API
gravity. Figure IH-23 summarizes data for all Phase I and II boilers for runs at the base-line
condition. Figure 111-24 shows the relationship separately for each Phase II boiler. Both figures
reveal that smoke generated at the 80 percent load, 12 percent CO2 base-line firing condition
increased as the API gravity of the fuel decreased.
8
Legend
o Phase I boilers
A Phase H boilers
20 30
API Gravity at 60 F
40
Figure IH-23. Relation of Fuel Gravity and Smoke Number for the Commercial
Boilers Operating at 80 Percent Load and 12 Percent CC>2
Phases I and II Boilers.
Figures III-17 through 111-22 show that the heavier fuel oils generally tended to produce
greater smoke over the entire range of CO2 investigated.
All smoke spots obtained from firing the reference fuel, CR, were yellow in appearance.
The coloration was not attributable to unburned fuel, but was assumed to result from com-
pounds formed from trace metals in the fuel.
-------
111-35
Boiler CI002
o C2002
x C2003
+ C2004
A C2005
A C2006
O
E
.
o
o
o
o
o
m
C2003
I
10
15
20 25
API Gravity at 60 F
30
35
40
Figure 111-24. Relation of Bacharach Smoke No. and API Gravity for Oil Firing
of Phase II Commercial Boileis for 80 Percent Load and
12 Percent CO2
-------
111-36
Factors Influencing Particulate Emission
A regression analysis was used to determine the best equation relating filterable emissions
to the same fuel and boiler characteristics as considered for the NOX correlation (see page
111-12). The equation that resulted was
Filterable particulate (lb/1000 gal) = 884.6 + 9.099 Carbon residue (percent,
Ramsbottom) - 0.120 Viscosity (SSU) - 10.88 Carbon content (percent) + 1.77 API
gravity
This equation gave an 0.86 correlation with the filterable particulate data. Correlation
coefficients with particulate emissions obtained when each variable was considered separately are
as follows:
Correlation Coefficient
Variable (Absolute Value)
Carbon residue, Ramsbottom 0.68
Fuel nitrogen, N 0.62
N% 0.57
Specific gravity 0.55
API gravity 0.55
Carbon content, C 0.52
Hydrogen content, H 0.44
g 0.43
H 0.40
C 0.33
H+N
Viscosity at firing temperature, SSU 0.25
Firing rate 0.21
Firing rate per unit combustion 0.14
chamber volume
Combustion chamber volume 0.05
Figure 111-25 shows the filterable particulate emissions plotted against carbon residue and the
curve that correlated at the 0.68 level. As for the NOX emissions, filterable particulate emissions
correlated better with fuel characteristics than with boiler characteristics.
Figure 111-26 shows the relationship of filterable and total particulate to API gravity for
each of the Phase II boilers at 80 percent load and 12 percent CO7. As might be expected, each
Phase II boiler showed increasing particulate emissions for heavier fuels. However, the range of
emissions obtained when firing fuels of similar gravities in different boilers was large. Thus, boiler
and/or burner characteristics may have as much or more influence on emissions than fuel
properties.
-------
too
111-37
8 «o
10 II
Figure 111-25. Correlation of Filterable Particulate Emissions With Carbon Residue
Participate Emissions by Modified EPA. Procedure.
-------
S 50
20 25 30 35 40
API , groi/ily
S.
I.
20 25
API, gravity
3O 55 40
Figure 111-26. Relation of Participate Emissions to Fuel API Gravity for Base-Line Conditions
Particulate emissions by modified EPA procedure.
00
-------
111-39
Relation of Particulate Emissions to Load. Figure 111-27 shows the change in filterable
particulate emissions with load for each Phase I and II boiler and fuel where data at two loads
are available. In general, increasing load did not significantly affect filterable particulate emis-
sions, except for firing of No. 6 fuel oil. For the four boilers fired with No. 6 fuel oil, emissions
showed a substantial increase with increasing load. Particulate emissions increased from 0.15 to
1.3 lb/1000 gal for each 1 percent increase in load over the range for which data are available.
Trial Correlation of Smoke Versus
Particulate Emissions
Figure 111-28 shows the relationship of filterable particulate (as determined by the
modified EPA procedures) to smoke number for each Phase II boiler at the base-line condition.
It must be pointed out that the fuel varies for each point on each boiler curve. Hence, several
variables are included in these plots and the relationship of particulate emissions to smoke should
be examined further.
These data appear to provide a better correlation between smoke number and particulate
for the commercial boilers than the residential unit data in that, for each boiler, particulate
emissions increased with smoke number. This may relate to larger spread in particulate emissions
for the commercial boilers or, more likely, to the fact that the particulate samples for the
commercial boilers were taken under steady-state conditions, whereas the samples for residential
units included start-up and shutdown during cyclic operation.
Particle Size Measurements
Emission samples were collected from Boilers C2005 and C2006 during operation with
the low-sulfur reference fuel (CR) at 12.1 percent CO2, 78 percent load and 12 percent CO2, 80
percent load, respectively.
Procedure. Sampling was performed isokinetically at the center of each stack using the
Battelle cascade impactor. The impactor classifies particles in the range of 0.25 16.0 microns
into 7 size categories. For the sampling done for this study, a 5/8-inch-diameter sample probe
about 20 inches in length (from probe tip to impactor inlet) was used. It had a 90 degree, 6-inch
radius bend such that the probe tip pointed upstream. The impactor was operated horizontally
and was heated to stack gas temperature (about 400 F) prior to probe insertion in the stack.
Sampling at 12.5 liters per minute was initiated immediately on probe insertion into the stack.
Sampling times for Boiler C2005 were 0.5, 1, 2, 5, or 10 minutes and for Boiler C2006 were 1
and 2 minutes. After sampling, the impactor slides were removed immediately and returned to
the laboratory for weighing.
The impactor slides were covered with a disk of silver membrane material so that the
entire glass slide did not have to be weighed. The possibility of weight change of the silver
membrane filters by reaction during the sampling operation was calibrated by baseline measure-
ments made with an all-glass filter placed in-line ahead of the impactor. In these calibration runs
no particles were collected in the impactor and the small weight changes of the silver substrates
provided corrections for use during particle size measurements.
-------
60
50
Fuel
O No. 2
X No. 4
+ No. 5
No. 6
111-40
40 50 60
Boiler Load, percent
Figure 111-27. Effect of Load on Filterable Particulate Emissions From
Phases I and II Commercial Boilers
Particulate Emissions by Modified EPA Procedure.
-------
m-4i
IOO
90
C200I
O C2002
X C2003
+ C2004
A C2005
C2006
3 4
Bacharock Smoke Number
Figure 111-28. Relation of Filterable Particulate and Smoke Numbers
for Commercial Boilers
Particulate Emissions by Modified EPA Procedure.
-------
\ll-42
Results. The weights of particles collected on each stage provided the basis for size
distribution determinations, and the total particulate weight collected in the impactor provided a
measure of the particulate emission rate. The small quantities of particulate collected indicated
that sampling times of five minutes or more were required for reliable results. This situation
resulted in discarding the 1 and 2-minute samples for Boiler C2005 and both samples (1 and
2-minute) taken from Boiler C2006.
The results for particule size distributions measured for the 5 and 10-minute sampling
periods for Boiler C2005 are shown in Figure 111-29. It can be seen that the agreement for the
two samples is reasonably good with the mass mean size of the particulate emissions estimated to
be about 0.5 micron for both. As the particle masses collected were near the lower limit of
sensitivity for the balance, these data must be considered as approximate. Improved weighing
techniques based on the experience gained in achieving these approximate results would be likely
to produce data in which more confidence could be placed.
The total weight of particulate collected in the impactor was used to estimate the mass
concentration of particles in the effluent stack gas. On this basis, the particulate loadings for
Boiler C2005 were calculated to be 38.4 and 43.1 mg/sm3 for the 5 and 10-minute sampling
periods, respectively. These compare reasonably well with the emission concentration of 44.6
mg/sm3 as determined independently with the EPA sampling procedures.
EMISSION FACTORS
Emission factors published by EPA3 do not provide discrimination relative to fuel
properties. Generally, EPA fuel categories are limited to "distillate oil" and "residual oil" as
classes. Many industry observers believe that a finer discrimination of fuel properties needs to be
considered. For example, the present practice would include in the same "residual" category
both No. 6 fuel oil with 3 percent sulfur (API gravity about 15) and the newer low-sulfur heavy
oils being marketed on the East Coast (API gravity about 24). However, it is well known that
these low-sulfur fuels are much easier to fire (about like No. 4 fuel oil) and generally give much
cleaner combustion.
Emission Factors Related to API Gravity
In recognition of the need for finer discrimination of fuel oils when considering their
emission potential, a correlation was made of emissions relative to API gravity of fuels. Figures
111-30 to 111-33 show Phase I and II boiler-emission data at the base-line conditions (80 percent
load and 12 percent C02) plotted against fuel gravity for CO, HC, filterable particulate, and
total particulate emissions. These figures also show the "best curve" that could be fit to these
data by a polynominal regression analysis that utilizes least square procedures.* It is suggested
that, although considerable scatter exists in the data, emission factors based on the API gravity
and the curves shown in Figures 111-30 to 111-33 give the best available estimate of likely
emissions from any one boiler firing any one fuel.
*The curves were limited to second-order polynomials to avoid negative values and peaks.
-------
III-43
en
c.
_
»
_O
1
O
99
98
95
90
80
70
60
50
40
30
20
10
5
2
I
0.5
0.2
O.I
0.05
0.01
O 5 min sample
A 10 min sample
0.2 0.3 0.4
0.6 0.8 I 234
Particle Diameter, microns
8 10
20
Figure 111-29. Particle Size Distributions for Boiler C2005 Firing the Commercial
Reference Fuel
-------
111-44
4.0
3.5
3.0
B 2.5
8
o
v.
-O
§ 2.0
in
in
O
U
1.5
1.0
0.5
4 points
25
API,gravity
30
35
30
25
CM
O
20
to
o
E
Q.
Q.
IE
D"
UJ
in
O
in
in
10
40
Figure 111-30. Relation of CO Emissions to API Gravity for Commercial Boilers
Equation for curve: CO = 1.572-0.0278 (API)-O.OOOIS (API)2.
-------
111-45
1.4
1.2
1.0
§>
o
Q 0.8
in
O
£ 0.6
UJ
u
I
0.4
20
,5
c
4)
U
l_
0>
o.
ro
o
E
Q.
O.
10 -
a>
o
cr
c
o
O
I
0.2
10
15
1
20
25 30
API, gravity
35
40
Figure 111-31. Relation of HC Emissions to API Gravity for Commercial Boilers
Equation for curve: HC = 0.439-0.0110 (API) + 0.00009 (API)2.
-------
HI-46
100
90
80
70
O
O
O
60
E
LJ
V
S 50
40
25
API .gravity
Figure 111-32. Relation of Filterable Participate Emissions (by Modified EPA Procedure)
to API Gravity for Commercial Boilers
Equation for curve = Filterable Participate = 100.5-5.75 (API) +
0.0832 (API)2. This curve does not apply at API gravities above 34.5.
-------
ni-47
100
90
80
70
60
£
o
50
30
20
10
10
15
20 25 30
API, gravity
35
40
Figure ffl-33. Relation of Total Particulate Emissions (by Modified EPA Procedure)
to API Gravity for Commercial Boileis
Equation for curve: Total Particulate = 106.2 - 4.96 (API) +
0.0597 (API)2.
-------
111-48
Suggested Emission Factors for
Commercial Boilers
The curves presented in Figures HI-30 through HI-33 can form the basis for recom-
mended emission factors to be used in emission inventories, where an average emission level is
needed for developing total emissions for oil-fired commercial boilers based on fuel usage.
To establish values for API gravity which are typical of different grades, results of an
informal survey of fuel specialists on the API Task Force were combined with average values for
fuels encountered in this field investigation. The resulting values are as follows:
Typical
API Gravity,
Fuel Grade degrees at 60 F
Distillate oil No. 2 34
Conventional resid. No. 4 22
No. 5 17
No. 6 14
New low-sulfur 23
residual (1.0% S)
Oil-Fired Boilers. Table III-6 shows the resulting emission factors which are suggested for
use in emission inventories applying to oil-fired commercial boilers. These factors are based on
emission measurements of 27 combinations of boilers and fuels, a fairly broad cross section of
conditions encountered in the field. They should be updated as more comprehensive data
become available, particularly to keep up with trends in fuels.
Gas-Fired Boilers. Table HI-7 shows average emission factors obtained from firing seven
commercial boilers with natural gas.
Emission Factor for /VOX. Assuming that the single variable relating best to NOX
emissions was fuel nitrogen, it was determined that NOX emissions should not be expressed in
terms of API gravity. (However, for many fuel oils there is a fairly strong relationship between
fuel nitrogen and API gravity.) Hence, using the equation relating NOX to fuel nitrogen described
earlier (Page HI-14), NOX emission factors were calculated for fuels of various API gravities.
These results are included in Table I1I-6.
When fuel nitrogen data are lacking, the following values are suggested:
Fuel Fuel Nitrogen, percent
No. 2 0.01
No. 4 0.2
No. 5 0.3
No. 6 0.4
LSR 0.2
-------
111-49
Table 111-6. Suggested Emission Factors for Oil-Fired Commercial Boilers3
Emission Factors, lb/1000 galb
Fuel Grade3
CO
HC
NO..
SO,
Filterable
Participate
Suggested Emission Factors From This Study9
No. 2
Nq. 4
No. 5
No. 6
LSRf
0.45
0.89
1.06
1.15
0.85
0.17
0.24
0.28
0.30
0.23
20 + 78 N°-6
20 + 85 N°-6
20 + 87 N°-6
20 + 89 N°-6
20+.84N0-6
142 S
154 S
159S
162S
153S
EPA Published Emission Factors3
1,2
14.0
27.0
36.0
12.0
Distillate Oil
No. 2 0.20 3.00
Residual Oil
Nos. 4,5,6 0.20 3.00
40 to 80
40. to 80
142 S
157S
15.0
23.0
a These values are based on average emission data for the identified fuel grades having typical API gravity
as follows: 34 degrees API for No. 2, 22 for No. 4, 17 for No. 5, 14 for No. 6, and 23 for LSR. Where
actual API gravity is known, interpolated values should be used.
To convert these values to emission factors in lb/10" Btu, multiply values shown by 0.0069. (The
actual multiplier varies slightly with fuel grade, being about 0.0071 for No. 2 fuel oil and 0.0066 for
No. 6 fuel oil.)
c N = multiplication factor equal to percent nitrogen in fuel.
S = multiplication factor equal to percent sulfur in fuel. These emission factors were determined by
calculation. Measured S02 emission factors were within 10 percent of these values.
Based on firing at 80 percent load and 12 percent COo-
LSR: low-sulfur residual oil (1.0 percent SI.
Table II1-7. Suggested Emission Factors for Gas-Fired
Commercial Boilers
Emission Factors, lb/106 cu ft3
CO
HC NOX SO2
Filterable
Particulate
Suggested Emission Factors From This Studyb
17.7 3.7 103. Nil 5.6
EPA Published Emission Factors3
8 100 0.6
20
19
a fi
To convert these values to emission factors in lb/10 Btu,
multiply these values by 0.00098.
Based on firing at 80 percent load and 10 percent COj.
-------
A-l
APPENDIX A
BACKGROUND DATA ON RESIDENTIAL OIL-FIRED EQUIPMENT POPULATION
This appendix presents information on the current field population of oil-fired residential
heating equipment in the United States by important characteristics, including firing rate, burner
and heating system type, and age of burner and heating system.
RESIDENTIAL OIL-BURNER EQUIPMENT SURVEY
In the selection of residential units for this investigation, pertinent historical data were
reviewed, including information compiled previously by Fueloil & Oil Heat magazine. To obtain
more comprehensive information not previously recorded in the industry, a questionnaire was
submitted to a large number of oil-burner service organizations through the statistical editor of
Fueloil & Oil Heat. The same questionnaire also was submitted directly to the members of the
API SS-5 Task Force and to a number of servicing companies in major cities. The pertinent
portion of the questionnaire is reproduced as Table A-l.
Profile of Residential Oil heating
Equipment Population
Tables A-2 through A-9 present information on the profile of oilheating equipment by
type, age, and geographic distribution. The information was compiled and later published by
Fueloil & Oil Heat4.
The summary is organized in tables as follows:
Table A-2 Firing Rates of Oil Burners in Use
Table A-3 Oil Burners in Use, by Type
Table A-4 Average Age of Oil Burners in Use, by Type
Table A-5 Types of Central Oilheating in Use, by System Type
Table A-6 Average Age of Central Oilheating Equipment in Use, by System Type
Table A-7 Combustion Chambers of Oil Burners in Use, by Type
Table A-8 Oil Burners With Solenoid Valves
A number of general observations and trends in residential oilheating equipment that can
be noted from these data include the following:
Smaller size burners, with firing rates of 1.35 gph and below, make
up the majority (about 70 percent) of residential units.
Gun burners (with conventional, Shell, or flame-retention heads)
represent a large majority of burners (about 84 percent).
-------
A-2
High-turbulence combustion leads (Shell and flame retention)
represent less than 20 percent of all gun burners.
The average age of all oil burners is slightly over 12 years, about the
same as for gun burners with conventional combustion heads. Average
age of burners with high-turbulence combustion heads is only about 6
years, while other types of burners (low-pressure, rotary, and vaporiz-
ing units) average nearly 17 years.
Warm-air furnace and hydronic systems (hot water and steam) are
about equal in number.
In spite of the increased popularity of ceramic fiber combustion
chambers in recent years, refractory brick combustion chambers still
outnumber the ceramic fiber chambers by about four to one.
These findings are familiar qualitatively to those of long-time observers of the oilheating
industry; however, the statistics contained in this appendix provide more quantitative informa-
tion than available previously.
Residential Unit Selection for Phase II
Due to delays in receiving replies to the questionnaire, this information was not complete
in time for planning the details of the Phase II residential equipment investigation. Thus, to keep
this program on schedule it was necessary to rely on the previously published data to select units
for Phase II of this program, as outlined in Chapter II.
The composition of the field sample, chosen prior to the comprehensive survey, is
compared in Chapter II to this profile of the total field population. The mix of units chosen for
the combined program of Phases I and II was generally representative of the total U.S.
population of oil-fired residential equipment.
-------
A-3
Table A-1. Sample Questionnaire
Battelle is doing some work for API and the Government on emissions of residential oilburners, and have asked
your help on sizes and types of equipment in the field. Please give us your best estimates or guesses about the
burners on your customer list.
What percent of your residential No. 2 burners are fired at less than 1.0 gph %; 1.0 to
1.35 gph %; 1.36 to 1.65 gph %; 1.60 to 2.0 %; 2.01 to 3.0 gph %;
over 3.1 gph %
To further profile these existing burners, please fill in as much of the following data as possible:
Less than 1.0 to Average
1.0 gph 3.0 gph Age
High Pressure*
Conventional combustion head % % years
Shell head
Flame retention head
Low Pressure**
Rotary Wall Flame
Vaporizing
100% 100%
'Pressure atomizing (100 psi or greater!.
**Air atomizing.
What percent of your customers have furnaces? % What percent have gravity hot water? '
Have forced hot water with a tankless coil? % Forced hot water without tankless coil? "/,
Have steam heat with tankless coil? %; steam without tankless coil? %
What would you guess would be the average age of your customers' furnaces? years. What is
the average age of your customers' boilers? years. What percent of your customers have space
heaters? %
Table A-2. Firing Rates of Oil Burners in Use4
<1.0
1.0 to
1.35
Firing
1.36 to
1.65
Rate, gph
1.66 to
2.0
2.01 to
3.0
Over
3.0
Percent of Total
New England
Mid-Atlantic
South Atlantic
Midwest
West
19.6
24.3
51.8
44.6
75.4
43.9
39.7
29.8
27.1
14.2
12.8
17.8
9.3
12.8
5.9
8.9
8.6
5.7
8.8
3.3
7.0
5.7
1.6
4.4
0.9
7.8
3.9
1.8
2.3
0.3
All Sections 34.6 34.7 13.9 8.0 4.9 3.9
-------
A-4
Table A-3. Oil Burners in Use by Type
13
High Pressure
Shell
Conventional Head
Retention
Head
Low
Pressure
Rotary
Vaporizing
Percent of Total
Less Than 1.0 gph
New England
Mid-Atlantic
South Atlantic
M idwest
West
All Sections
New England
Mid-Atlantic
South Atlantic
Midwest
West
All Sections
44.9
66.3
74.8
63.0
80.3
63.1
61.4
69.0
89.4
73.3
80.3
71.2
7.2
8.4
1.9
9.4
0.6
7.2
7.9
5.2
2.3
13.1
0,6
6.9
14.7
3.8
13.0
13.1
5.2
9.1
1.0 to
13.6
8.2
7.3
6.2
2.9
8.3
26.4
10.6
0.8
5.6
8.1
11.5
3.0 gph
9.0
10.7
0.8
6.0
13.2
8.6
5.8
8.5
0.1
8.6
0.5
6.6
7.6
6.8
0.2
1.3
1.8
4.7
1.0
2.4
9.4
0.3
5.3
2.5
0.5
0.1
-
0.1
1.2
0.3
All Oil burners up to 3.0 gph
All Sections
68.3
13 Private communication: to D. W.
data were published in Reference
Table A-4.
7.0
8.6
9.6
5.4
Locklin, Battelle-Columbus, from Margaret Mantho, Fueloil & Oil Heat,
4.
Average Age of
High Pressure
Shell
Conventional Head
Oil Burners
Retention
Head
in Use, by Type'
Low
Pressure
l
Rotary
1.1
May, 1972. Similar
Vaporizing
Years
New England
Mid -Atlantic
South Atlantic
Midwest
West
13.1
12.4
10.1
14.1
12.7
9.5
8.9
7.0
8.8
7.0
4.1
4.2
3.0
3.7
3.7
14.4
16.2
19.3
17.8
16.6
16.8
17.7
-
17.8
_
15.6
13.7
17.0
16.7
All Sections
12.6
8.7
3.9
16.5
17.5
15.9
-------
A-5
Table A-5. Types of Central Oilheating Equipment in Use, by System Type4
Forced Hot Water
With Without
New England
Mid-Atlantic
South Atlantic
Midwest
West
All Sections
Furnaces
35.6
36.3
74.1
70.9
91.4
51.9
Gravity
Hot Water
2.4
2.7
1.5
0.8
-
1.9
Tankless Tankless
Coil
Percent of Total
39.4
36.9
12.9
6.9
1.7
25.5
Coil
2.2
9.0
4.8
6.8
5.5
6.5
Steam
With
Tankless
Coil
17.1
9.2
1.5
12.0
0.5
10.0
Without
Tankless
Coil
3.3
5.9
5.2
2.6
0.9
4.2
Table A-6. Average Age of Central Oilheating Equipment in Use,
by System Type4
New England
Mid-Atlantic
South Atlantic
Midwest
West
Furnaces
14.9
14.7
11.1
14.1
13.0
Boilers
18.0
17.4
17.1
17.0
16.5
Combined
Boilers/Furnaces
Years
16.9
16.4
12.5
14.9
13.3
Burners
11.9
12.4
9.5
13.0
12.8
All Sections
13.8
17.5
15.6
12.2
-------
A-6
Table A-7. Combustion Chambers of Oil Burners in Use, by Type14
Firing Rate
Below 1.0 gph
Refractory
Brick
Ceramic
Fiber
Steel
1.
Refractory
Brick
0 to 3.0 gph
Ceramic
Fiber
Steel
Percent of Total
New England
Mid-Atlantic
South Atlantic
Midwest
West
All Sections
61
66
44
61
38
60
19
22
13
25
27
22
20
12
43
14
35
18
78
84
63
61
66
74
12
12
7
31
10
16
10
4
30
8
24
10
14 "Combustion Chambers by Types", Fueloil &ON Heat, Vol. 31, No. 5, March 1972, p. 94.
Table A-8. Oil Burners in Use With Solenoid Valves4
New England
Mid-Atlantic
South Atlantic
Midwest
West
Firing Rate
Below 1.0 gph
Percent of Total
9
25
34
34
6
1.0 to 3.0 gph
21
34
35
35
4
All Sections
23
29
-------
B-l
APPENDIX B
BACKGROUND DATA ON COMMERCIAL-INDUSTRIAL BOILER POPULATION
This appendix presents information on the profile of commercial and industrial boilers,
with distributions and trends by boiler type, capacity, and burner and fuel types, which were
used in selecting the equipment mix for this investigation.
Boiler industry statistics frequently combine commercial and industrial boilers; the
distinction is sometimes based on application and sometimes on size. (For purposes of this
investigation, the "commercial" range was defined as sizes between 10 and 300 boiler horse-
power, or approximately 0.3 to 10 million Btu/hr output.) Thus, to discriminate in considerable
detail on aspects which are common to the two major boiler size categories, both commercial
boilers and industrial boilers up to 500 million Btu/hr output are included in the profile
presented in this appendix.
BOILER SURVEY
As an initial step in the survey of commercial-industrial boilers, available data were
reviewed. These data were obtained from
1. "Systematic Study of Air Pollution from Intermediate-Size Fossil-
Fuel Combustion Equipment", J. R. Ehrenfeld, et al., Final Report
on Contract No. EPA 22-69-85, July, 1971.
2. "Stationary Watertube Steam and Hot Water Generator Sales, 1970",
American Boiler Manufacturer's Association (ABMA).
3. Fueloil & Oil Heat
4. Various other sources.
Although these publications contained data on the boiler population and/or sales, data were not
available with the detailed distributions of size, boiler type, fuel type, etc., as needed for this
study.
Therefore, through the assistance of ABMA, a questionnaire was submitted to members
of the ABMA Air Pollution Committee and a number of other key persons in the boiler
industry. Additionally, the same questionnaire was submitted to the members of the API SS-5
Task Force. The questionnaire was designed to solicit percentage estimates of the present
population of boilers and of boiler sales trends over the last 40 years by pertinent boiler, burner,
and fuel characteristics.
Summary of Boiler Survey
Tables B-l through B-8 summarize the replies to the questionnaire and the boiler
population study; they represent composite data from the various replies and are presented here
-------
B-2
as the "best estimate" of the population as determined from the various sources. The summary is
organized in tables as follows:
Current Equipment Population
Table B-l. By Boiler Type
Table B-2. By Fuel Capability (including oil type)
Tab]., ?-3. By Burner Type (for oil and coal)
Trends in Sales (by Years for '30, '50, '70, and Forecast for '90)
Table B-4. By Boiler Type
Table B-5. By Fuel Capability (including oil type)
Table B-6. By Burner Type (for oil and coal)
Annual Usage Factor for Load Factor)
Table B-7. By Boiler Application
Boiler and Burner Life Expectancy
Table B-8. By Boiler Type and Burner Type (for oil and coal).
General observations on the field population statistics are as follows:
A high portion of smaller boilers are packaged boilers, while most of
the larger boilers are field erected units.
Fire-tube and cast-iron boilers are more prevalent in the smaller sizes
and water-tube boilers are more prevalent in the larger sizes.
In the commercial-boiler size range, fire-tube boilers are the most
numerous type, with cast iron being next in popularity. In the
smaller commercial sizes (10-50 hp), firebox boilers are the most
numerous of the fire-tube types but are about equal in number with
packaged Scotch boilers in the larger sizes (100-300 hp).
Smaller boilers mostly fire a single fuel (gas or No. 2 fuel oil), with
an increasing percentage of duel-fuel boilers uscu "or the larger sizes.
For oil-fired boilers, the smaller boilers are generally fired with No. 2
fuel oil; the grade of fuel fired generally increases with boiler size.
Pressure atomizers are most common for small, oil-fired boilers,
rotary and air atomizers for intermediate-size boilers, and steam
atomizers for larger boilers.
Coal-fired boilers are rare in the commercial size range.
Boiler Selection for Phase II
Based on the estimates of the present boiler population presented in Tables B-l through
B-3, a selection of six boilers was made for inclusion in the Phase II studies. This selection was
discussed in Chapter II.
-------
Table B-1. Population
Breakdown by Boiler Types (Percentage Basis)
All Commercial-Industrial Boilers Now in Service in U.S.
Commercial
Industrial .
RATED
CAPACITY,
SIZE RANGE
Id6 Btu/hr or
1Q3 Ib stm/hr
Bailer Horsepower
10-50
51-100
10-16
101-300
301-500
17-100
101-250
251-500
WATER TUBE
Industrial Type >104 # Steam/hr
Packaged
Field erected
Commercial Type <104 # Steam/hr
Coil
Firebox
Other
FIRE TUBE
Packaged Scotch
Firebox
Vertical
Horizontal Return Tubular (HRT)
Misc. (Locomotive type, etc.)
CAST IRON
MISC. (Tubeless, etc.)
TOTAL
COMMERCIAL-INDUSTRIAL
BOILERS
(22)
15
15
1
5
2
1
100%
20
25
0.5
10
3
33 -
0.5
100%
30
30
nil
15
5
15
nil
100%
100%
100%
100%
-------
Table B-2. Population
Breakdown by Fuel Capability (Percentage Basis)
All Commercial-Industrial Boilers Now in Service
RATED
SIZE RANGE
FUELS
Oil Only
Gas Only
Coal Only
106 Btu/hr or
ID3 Ib stni/lir
Boiler Horsepower
Oil & Gas and Gas & Oil
Oil & Coal and Coal & Oil
Gas & Coal and Coal & Gas
Misc. Fuels
(alone or with alternate fuels)
OIL
Total
Distillate, No. 2
Resid
No. 4 and
Heavy No.
Light No. 5 (No preheat)
5 and No. 6 (Preheated)
Total Oil
Commercial "
10-50
42
50
2
5
'$(%%%%%(%%(%!>
M%%%%%%%f%%
1
100%
95
(5)
4.5
0.5
100%
51-100
42
50
1
6
mmm?//m
!
100%
85
(15)
14
1
100%
101-300
40
50
1
8
W///^///////?////.
!
100%
50
(50)
30
20
100%
Industrial -
10-16
301-500
35
45
3
16
^/////////////////,
w^imfr,
i
100%
10
(90)
20
70
100%
17-100
35
35
10
18
2
100%
2
(98)
2
96
100%
101-250
30
22
18
25
0.5
0.5
3
100%
2
(98)
nil
98
100%
251-500
22
22
22
23
3
3
5
100%
2
(98)
till
98
100%
-------
Table B-3. Population
Breakdown by Burner Type (Percentage Basis)
All Commercial-Industrial Boilers Now in Service in U.S.
RATED
CAPACITY,
SIZE RANGE
OIL BURNERS
106 Btu/hr or
1Q3 Ib stm/hr
Boiler Horsepower
(approx. gph)
Air Atomizing
Steam Atomizing
Pressure or Mechanical Atomizing
Rotary
Total Oil
COAL BURNERS (approx. Ib/hr)
Spreader
Underfeed
Overfeed
Pulverized
Other
Total Coal
10-50
3-15
15
70
15
100%
33-160
60
15
5
'%%ffi^%%%%,
20
100%
51-100
15-30
35
im^^^^
25
40
100%
160-330
60
15
5
20
100%
101-300
30-90
40
20
40
100%
330-1000
55
20
5
20
100%
Industrial
10-16
301-500
90-150
40
20
10
30
100%
1000-1600
60
15
5
lillllllii
20
100%
17-100
150-900
15
70
10
5
100%
1600-10,000
45
15
5
35
100%
101-250
900-2250
5
85
10
100%
10,000-25,000
45
15
5
15
20
100%
251-500
2250-4500
1
94
5
100%
25,000-50.000
20
nil
nil
65
15
100%
-------
Table B-4. Estimated Trends of Boiler Types (Percentage Basis)
All Commercial-Industrial Boilers Installed in Years Noted
- Commercial -
Industrial-
RATED
CAPACITY.
SIZE RANGE
106 Btu/hr or
103 Ib stm/hr
Boiler Horsepower
WATER TUBE
Industrial Type > 104 # Steam/Hr
Packaged
Field erected
Commercial Type <104 # Steam/Hr
Coil
Firebox
Other
FIRE-TUBE
Packaged Scotch
Firebox
Vertical
Horizontal Return Tubular !HRT)
Misc. (Locomotive type, etc.)
CAST IRON
MISC (TUBELESS, ETC)
TOTAL
COMMERCIAL-INDUSTRIAL
BOILERS
10-50
30 '50 '70 '90
30 '50 '70 '90
nil 10 11 10
b 11 18 20
5666
5 1 nil nil
8532
60 50 45 40
3211
100 100 100 100
51-100
I 14 22 26
20 15 17 23
3532
5 2 nil nil
5111
50 45 40 36
3 2 1 nil
100 100 100 100
10-16
101-300
301-500
30 '50 '70 '90
30 '50 '70 '90
(25)(17)(19)(20)
0 2 18 20
25 15 10
2 21 41 40
20 30 35 3D
5 2 nil nil
60 25 1 nil
852 nil
nil 5 10 15
nil 111
100 100 100 100
nil 35 40 45
20 40 40 35
100 100 100 100
17-100
30 '50 '70 '90
(94) (97) (94) (90) J.OO) (100) (100)(1O>) (LOO) (100) (100)(LOO
0 8 80 89
94 89 14 1
100 100 100 100
101-250
30 '50 '70 '90
0 0 80 90
100 100 20 10
251-500
30 '50 '70 '90
001
100 100
100 100 100 100
100 100 100 100
-------
Table B-5. Estimated Trends by Fuel Capability (Percentage Basis)
All Commercial-Industrial Boilers Installed in Years Noted,
Including Conversions
RATED
CAPACITY,
SIZE RANGE
106Btu/hror
103 Ib stm/hr
Boiler Horsepower
FUEL CAPABILITY
Oil Only
Gas Only
Coal Only
Oil & Gas and Gas & Oil
Oil & Coal and Coal Be Oil
Gas & Coal and Coal & Gas
Misc. fuels
(alone or with alternate fuels)
Tota
OIL
Distillate, No
Resid
2
No. 4 & Light No. 5 (No preheat)
Heavy No. 5 & No. 6 (Preheated)
Total Oil
-« Commercial *
10-50
'30 '50 '70 '90
20 40 30 15
10 30 45 50
65 20 5 nil
nil 5 18 25
liiiiiiiiiiiiP
^iitiiiiiiiii
5 5 2 10
100 100 100 100
40 50 70 100
(60) (501(30) (nil)
40 40 25 nil
20 10 5 nil
100 100 100 100
51-100
30 '50 '70 '90
10 30 25 15
10 30 38 50
75 15 5 nil
nil 10 30 25
{%%%%%%%%%%?
5 5 2 10
100 100 100 100
20 30 40 70
(80) (70) (60) (30)
50 50 50 30
30 20 10 nil
100 100 100 100
101-300
30 '50 '70 '90
10 40 30 30
5 25 30 30
80 10 5 nil
lil 20 30 30
%%%%!>:%%%%%
55 5 10
100 100 100 100
10 10 20 40
(90) (90) (80) (60)
30 40 30 5
60 50 50 55
100 100 100 100
-« Industrial *]
10-16
301-500
'30 '50 '70 '90
17 43 30 30
5 20 30 30
75 10 5 nil
nil 25 30 30
^^^^^^.
32 5 10
100 100 100 100
5 2 10 30
(95) (98) (90) (70)
20 23 10 nil
75 75 80 70
100 100 100 100
17-100
'30 '50 '70 '90
13 30 30 25
10 30 30 25
75 30 5 nil
nil 5 30 35
iiiiiiiiiiiillf
^HHHHililtii
2 5 5 15
100 100 100 100
ill nil 10 20
(ioo) (too) (90) $o;
til 5 nil nil
100 95 90 80
100 100 100 100
101-250
'30 '50 '70 '90
5 20 24 20
5 20 24 20
90 38 15 nil
nil 10 25 40
nil 555
nil 5 55
nil 2 2 10
100 100 100 100
nil nil 5 10
(LOO) (100) (95) (90]
lil nil nil nil
LOO 100 95 90
100 100 100 100
251-500
30 "50 '70 '90
5 15 20 10
5 15 20 10
90 60 20 10
nil 5 20 30
nil 3 10 20
nil 2 10 20
nil nil nil nil
100 100 100 100
nil nil 5 10
(1 00) (100) (95) (90)
nil nil nil nil
100 100 95 90
100 100 100 100
-------
Table B-6. Estimated Trends by Burner Type (Percentage Basis)
All Commercial Industrial Bailers Installed in Years Noted,
Including Conversions
- Commercial -
Industrial
106 Btu/hr or
RATED W3 Ib stm/hr
SIZE RANGE Boiler Horsepower
OIL BURNERS
Air Atomizing
Steam Atomizing
Pressure or Mechanical Atomizing
Rotary
Total Oil
COAL BURNER
Spreader
Underfeed
Overfeed
Pulverized ;
Other
Total Coal
10-50
'30 '50 '70 '90
10 15 15 15
70 75 85 85
20 10 nil nil
100 100 100 100
% v.-.
nil nil nil nil
5 90 90 nil
nil 5 5 nil
95 5 5 nil
100 100 100 100
51-100
30 '50 '70 '90
15 20 30 30
Wffiffiffiffiftffift
55 60 65 70
30 20 5 nil
100 100 100 100
%
lil nil nil nil
45 90 90 nil
555 nil
50 5 5 nil
100 100 100 100
101-300
30 '50 '70 '90
20 30 55 60
50 40 40 40
30 30 5 nil
100 100 100 100
%
nil 5 5 nil
75 85 85 nil
20 5 5 nil
55 5 nil
100 100 100 100
10-16
301-500
30 '50 '70 '90
10 20 35 40
30 30 35 40
25 20 20 20
35 30 10 nil
100 100 100 100
%
5 10 5 nil
75 75 85 90
15 10 5 5
5555
100 100 100 100
17-100
'30 '50 '70 '90
5321
75 80 88 90
15 14 10 9
5 3 nil nil
100 100 100 100
%
15 50 50 nil
50 35 35 85
30 10 10 10
vX£$//&2///%2&/'&%'
5555
100 100 100 100
101-250
'30 '50 '70 '90
2 1 nil nil
93 94 95 95
5555
100 100 100 100
%
15 40 40 nil
50 30 20 20
25 15 15 15
5 10 20 60
5555
100 100 100 100
261-500
'30 '50 '70 '90
2 1 nil nil
93 94 95 95
5555
100 100 100 100
%
15 30 20 nil
40 20 10 1.0
20 15 10 10
20 30 55 75
5555
100 100 100 100
-------
Table B-7. Typical Annual Usage Factor lor Load Factor)
ACTUAL FUEL USED PER YEAB X 100
For Commereial-lndustria. Boilers UM"~ ' """" >- ^AGE FOR YEAR AT CONSIAN! FULL LOA1> UKHAIMI
For Various Applications
RATED
SIZE RANGE
APPLICATION
106 Btu/hr or
103 Ib stm/hr
Boiler Horsepower
Commercial Heating
Commercial Processing
(Laundry, etc.)
Industrial Heating Only
Industrial Processing Only
General Industrial Use
(Heating, Processing, Power]
Industrial Power
Utility Power
10-60
%
50
80
60
85
/yyfflvvsysfflfflwwft
y///////////////
'Y/y////2s/Y/y//s
61-100
%
50
80
60
85
'<%MM//M///S
101-300
%
50
80
60
88
WT/m/y/mmWm.
'////////////////
w/Mw/y//,
10-16
301-500
%
50
80
60
85
mmm
nMwffiwMMWW.
85
17-100
%
60
80
60
85
85
WM%MM
90
101-250
%
80
60
90
90
95
95
251-500
80
60
90
90
95
95
-------
B-10
Table
B-8. Boiler and
A. Boiler Life Expectancy, Average
Boiler Type
Water Tube
industrial type
Packaged
Field erected
Commercial type
Coil
.
Firebox
Other
Fire Tube
Packaged Scotch
Firebox
Vertical
Horizontal return tubular
Misc.
Cast Iron
Misc.
Years
30
45
20
25
25
20
25
10
40
10-20
40
30
Burner Life Expectancy
B. Burner Life Expectancy, Average
Burner Type
Oil Burners
Air Atomizing
Steam Atomizing
Pressure/Mechanical Atomizing
Rotary
Coal Burners
Spreader
Underfeed
Overfeed
Pulverized
Other
Years
20
30
15
20
20
15
15
18
15
-------
C-l
APPENDIX C
FUEL ANALYSES
Residential Fuels
Table C-l lists properties of the No. 2 oils fired in residential units for the Phase II
investigation. The reference fuel for firing in the residential units (designated RR) was a
high-quality hydrotreated fuel.
Table C-2 reports results of chemical analyses for C, H, and N contents of the fuels fired
in the residential units.
Commercial Fuels
Table C-3 lists the generally reported properties for the fuel oils fired in the commercial
boilers. These include fuel grades Nos. 2,^4, 5, and 6.
The reference fuel for firing in the commercial boilers (designated CR) was a low sulfur
residual fuel typical of the 1.0 percent sulfur residual oils being marketed to the East Coast in
March, 1972. It was purchased from a fuel oil supplier located in Connecticut.
Table C-4 reports results of chemical analyses for C, H, and N contents of the fuel oils
fired in the commercial boilers.
-------
Table C-1. Properties of Fuels Fired in Residential Units3
Fuel Gravity,
Fired
in Unit
23
23 1
23.2
24
24.1
24.2
25
25.1
25.2
26
26.1
26.2
27
28
29
30
31
32
33
34
35
RRC
a Analyses
O Dificpl inr
API at
60 F
34.2
31.7
31.0
32.3
34.4
30.6
34.1
34.2
34.8
33.8
34.1
34.3
31.1
32.7
33.2
34.0
30.9
30.6
32.2
36.9
33.6
34.7
by E. W.
Ani
ipv =
Kinematic
Viscosity
at 100 F,
cs
2.23
242
2.23
2.20
2.47
2.19
2.45
2.31
2.17
2.41
2.42
3.12
2.45
2.39
2.54
2.59
2.46
2.44
2.42
2.34
1.82
2.71
Flash
Point,
F
150
152
144
144
148
138
154
150
158
150
152
154
152
156
156
154
152
156
148
160
152
142
Carbon
Residue,
Ramsbottom
(last 10%),%
0.23
0.24
0.22
0.23
0.14
0.25
0.16
0.15
0.15
0.19
0.12
0.16
0.23
0.22
0.21
0.18
0.24
0.23
0.24
0.14
0.28
0.15
Sulfur,
%
0.19
0.17
0.15
0.27
0.16
0.17
0.11
0.14
0.16
0.17
0.14
0.14
0.24
0.18
0.17
0.10
0.20
0.16
0.20
0.19
0.24
0.05
Aniline
Point,
%
133
130
122
124
140
119
139
139
140
139
141
139
129
132
135
142
133
124
129
150
118
148
Diesel
lndexb
45.5
41.2
37.8
40.1
48.2
36.4
47.4
47.5
48.7
47.0
48.1
47.7
40.1
43.2
44.8
48.3
41.1
37.9
41.5
55.4
39.6
51.4
IBP,
F
356
356
344
344
366
342
354
366
368
360
372
368
370
360
368
364
358
370
358
368
346
354
10%,
F
398
412
406
396
418
394
416
414
416
416
414
414
412
408
416
416
412
424
416
414
392
418
ASTM
50%,
F
490
498
490
490
500
486
498
496
484
494
500
502
504
492
502
510
506
510
504
490
468
504
Distillation
90%,
F
578
592
588
586
582
588
586
582
558
582
582
588
586
578
590
592
604
604
590
578
550
602
End
Point,
F
632
644
640
644
634
640
634
630
614
634
630
636
642
634
642
640
656
656
642
630
588
640
Recovery,
%
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
99.0
Saybolt and Company.
line point in F
x API gravity
Reference luel high-quality hydrotreated fuel.
-------
C-3
Table C-2. Chemical Analyses of Fuels Fired in Residential Units3
Fuel Fired in Unit
23
23.1
23.2
24
24.1
24.2
25
25.1
25.2
26
26.1
26.2
27
28
29
30
31
32
33
34
35
Cb
87.0
86.9
87.4
87.5
87.4
86.7
87.0
86.9
86.7
86.6
87.1
86.3
87.2
86.3
87.1
84.6
84.2
87.4
87.2
85.9
87.2
Weight Percent
Hb
12.8
12.5
12.4
12.4
12.3
12.2
12.7
13.1
13.2
12.7
12.0
13.2
12.3
12.5
12.8
12.5
11.8
12.6
12.8
13.4
12.6
Nc
0,006
0.006
0.007
0.008
0.011
0.009
0.006
0.003
0.003
0.007
0.011
0.005
0.010
0.006
0.005
0.005
0.009
0.008
0.007
0.003
0.007
RRC
87.0
13.0
0.005
Analyses by Battelle-Columbus.
Pregl method.
Kjeldahl method.
Reference fuel - a high-quality hydrotreated fuel.
-------
Table C-3. Properties of Fuels Fired in Commercial Boilers3
Fuel
Fired in
Boiler
C2001 &
C2002
C2002
C2003
C2003
C2004
C2004
C2005
C2005
C2006
C2006
CR"
Fuel
Grade
. No.
2
4
2
6
2
5
2
6
2
6
Gravity,
API at
60 F
(D-287)
34.6
23.5
34.8
17.3
36.4
16.6
35.4
14.7
36.2
16.9
23.5
Viscosity,
at 100 F,
SSU
(D-445)
34.5
101.7
35.0
2746.
34.9
291.
35.0
3982.
34.9
3053.
313.
Flash
Point,
F
(D-93I
145
186
153
176
157
265
159
224
163
199
205
Carbon
Residue
Ramsbottom,
%
(D-524)
0.07
3.18
0.07
9.66
0.04
7.47
0.20
8.44
0.06
6.78
5.49
Ash, %
(D-482)
0.001
0.019
<0.005
0.062
<0.005
0.021
<0.005
0.019
<0.005
0.017
0.032
Sulfur, %
(D-26221
0.27
1.59
0.16
2.19
0.23
1.93
0.3B
1.67
0 18
1.13
0.99
Nitrogen
(Kjeldahl),
%
(C-93)
0.0084
0.17
0.0078
0.37
0.0113
0.29
0.0124
033
0.0086
031
0.22
Aniline
Point,
F
(D-611)
145
-
153
-
157
-
159
-
163
-
-
ASTM Distillation (D 861
IBP.
F
336
376
357
362
366
400
364
332
380
328
368
10%,
F
398
502
419
482
417
485
417
560
429
510
507
50%,
F
489
671
498
677
494
657
494
660
49fi
660
677
End
90%. Point,
F F
586 630
(75% at 700 F
592 630
(60% at 685 F
580 520
(68% at 682 F
587 638
(70% at 680 F
584 629
(70% at 685 F
(78% at 695 F
Metals0
Recovery,
%
98.0
max)c -
98.0
max)c
98.0
max)c
98.0
max)0 -
98.0
max]c -
max)c -
V.
ppm
<1.0
130.
<1.0
320.
<1.0
120.
CI.O
89.
CI.O
54.
170.
Ni. Ma,
ppm ppm
<0.1 <5
23. <5
<0. 1 <5
74. 46.
<0.1 <5
41. <5
<0.1 <5
36. 23.
<0.1 <5
24. 32.
40. <5
a Analyses by American Oil Company.
^ Emission spectroscopy, accuracy ±35 percerit.
c Cracked.
^ Reference fuel for commercial bailers was 1 .0 percent -sulfur residual-fuel
n
oil currently being marketed in Connecticut.
-------
C-5
Table C-4. Chemical Analyses of Fuels Fired in Commercial Boilers8
Weight Percent
Fuel Fired in Boiler
C2001 & C2002
C2002
C2003
C2003
C2004
C2004
C2005
C2005
C2006
C2006
CRd
Fuel Grade
2
4
2
6
2
5
2
6
2
6
-
Cb
86.1
85,0
85.8
85.2
86.9
86.0
86.3
86.8
87.1
86.0
86.3
Hb
12.8
11.8
12.9
11.6
13.1
10.9
13.1
11.1
13.0
11.1
12.5
Nc
0.004
0.15
0.010
0.36
0.010
0.28
0.010
0.31
0.007
0.30
0.22
a Analyses by BatteHe-Columbus.
b Pregl method.
c Kjeldahl method.
d Reference fuel for commercial boilers was a 1.0 percent sulfur residual-fuel oil currently being
marketed in Connecticut.
-------
EM
APPENDIX D
DETAILS OF FIELD PROCEDURES
RESIDENTIAL UNITS
Prior to setting up the monitoring equipment in the field, each homeowner was visited by
a Battelle staff member in advance of the scheduled arrival of the field team. During this visit,
the heating unit was examined for accessibility, space requirements, and material needs for
rerouting the flue pipe to accommodate the particulate sampling.
When the field team arrived to begin measurements, the monitoring instruments were set up
and recalibrated, and stack changes were made to permit installing a test section of duct (needed
for particulate sampling) in the stack. The gaseous sampling trains were set up and the
instruments checked out. Stack velocities were measured with an S-type Pitot tube, stack
pressures and temperatures were measured, and flue-gas flow rate and excess air were calculated.
After making the as-found run, the fuel line was broken and the fuel rate was determined
volumetrically by connecting the pump suction line to a calibrated container and measuring fuel
consumption over a convenient time interval.*
Measurements
Measurements were made of the following gases and stack emissions under various
conditions of operation:
CO2
02
CO
Hydrocarbons (total)
NOX and NO
Particulate loading.
In addition to these measurements, the combustion conditions normally observed by servicemen
were measured. These conditions included: draft (over-fire and stack) and flue-gas temperature
(at the particulate sampling point which was several feet from the breeching) - plus smoke,
CO2, and O2 as measured by standard field-type instruments. Details of instrumentation and
measurement procedures are given in Appendixes E and F.
*Measuring fuel consumption required breaking the fuel line. Thus, some air may have entered the fuel line, caus-
ing the cutoff to become somewhat more sluggish than normally expected for subsequent tuned and reference-
fuel runs.
-------
D-2
Conditions Investigated
Emissions from residential units were monitored under the following sets of conditions:
Cyclic and varied-air runs
- The as-found condition, A
A tuned condition, T
- Firing a No. 2 hydrotreated reference fuel in the tuned condition, R
Cyclic runs only
- Follow-up checks (Units 23 to 26).
Cyclic Runs. Cyclic runs refer to measurements taken at a fixed air setting during
operation on a 10-minute-on and 20-minute-off cycle, as controlled by a timer overriding the
thermostat. For the cyclic runs, gaseous-emission measurements were made for the entire period,
including both the on and off portion of the cycles. Particulate measurements were started at the
beginning of the second or third cycle and measured during each 10-minute firing portion of
four or five additional cycles. A total of 50 to 60 minutes of firing time (about 3 hours' elapsed
time) was required for the particulate sampling to accumulate sufficient material for accurate
weighing.
Varied Air Runs. Varied air runs refer to measurements made during steady-state opera-
tion of the unit to investigate the effect on emissions of excess-air setting. Gaseous emissions and
smoke were measured as the excess air adjustment was varied over the range from full open to
approximately a No. 5 to 7 smoke. The varied air runs were made with the house fuel in the
as-found condition and with the house and reference fuels for the tuned condition.
As-Found Condition. The initial measurements on the heating unit were made in the
"as-found" condition, i.e., no servicing or changes were made except for the rerouting of the flue
pipe. Measurements were made during both cyclic and varied air runs.
Tuned Condition, Following the test run in the "as-found" condition, the serviceman
cleaned and tuned the burner. This was not an "eyeball" adjuc^nent, but was intended to be a
tuning that a skilled serviceman would achieve with normal procedures of good practice with the
benefit of instrument readings of draft, CO2, and smoke, The following steps were included in
the serviceman's procedure to establish the tuned conditions:
Cleaning and adjusting the electrodes
Cleaning the blast tube and blower wheel
* Cleaning or replacing the nozzle (even by a different size or spray
pattern if it were better suited to the installation)
Cleaning or replacing the oil filter
-------
D-3
Simple sealing of air leaks at inspection door, around blast tube, or in
other easily accessible location
Change in draft-regulator setting (replaced regulator when necessary)
Change in combustion air adjustment.
The following items, being major repairs or modernization requiring special charges to the home-
owner, were not included in the tuning procedure.
Replacement of the combustion chamber or liner
Sealing of air leaks that would require disassembly of the boiler or
furnace jacket.
Replacement of the combustion head generally was not done. However, the burner for Unit 27
was intended to include a flame-retention combustion head which apparently had been removed.
To return this unit to its normal operating condition, a flame-retention head was installed during
the tuning.
The field team developed a rough CO2-smoke curve as part of defining the well-tuned
conditions. The air adjustment was then made just off the "knuckle" of this curve. The
instructions for adjusting to the tuned condition consisted of these steps:
(1) Compare the CO2 level in the stack with that obtained in sampling
by traversing above the fire using a simple averaging procedure. An
appreciable difference between stack and over-fire CO2 is indicative
of air infiltration through leaks; easily accessible leaks are to be
repaired.
(2) Establish a smoke-CO2 curve by use of the continuous CO2 reading
and the Bacharach smoke spot reading (with only enough points to
define the general shape, and particularly the location of the
"knuckle").
(3) If No. 1 or less Bacharach smoke can be achieved with air adjust-
ment alone, adjust the burner to the maximum CO2 for No. 1
smoke, but allow a cushion no nearer the "knuckle" than 0.5 to
1.0 percent CO2. (It would be expected that a well-tuned burner
should operate with at least 8 percent CO2.)
(4) If Step (3) cannot be accomplished to reach a No. 1 smoke, carry
out any of the specific tuning steps listed above to achieve that
performance, or approach it as closely as possible. In this pro-
cedure, first priority should be given to "reducing smoke level to at
least No. 2, and second priority to maintaining high CO2 (with 0.5
to 1.0 percent cushion from the "knuckle"). The smoke-CO2 curve
should be repeated to define the burner performance and to locate
the desired setting for the tuned test run.
-------
D-4
All the residential units in the Phase II program were tuned according to the above
tuning steps; adjustments for each unit are shown in Table D-l. Following the tuning, gaseous
and particulate emissions and smoke were measured for the "tuned" condition while firing the
house fuel.
Reference Fuel Condition. After completing the tests for the as-found and tuned condi-
tions, each residential unit was operated on a reference fuel, a high-quality No. 2 hydrotreated
fuel. The purpose of the reference fuel measurements was to provide a baseline in comparing the
variety of burner units on a common fuel basis and, thus, to remove the effect of randomness in
the quality of house fuels. Gaseous emissions and smoke were measured while firing the
reference fuel. In addition, particulate emissions were measured for Units 23 to 26.
Follow-Up Checks. To examine the effect of seasonal operation, four installations (Units
23 to 26) were selected for visits at three times during the course of the heating season an
initial measurement series plus two follow-up checks. The initial measurements were made in
mid-December, the first follow-up in mid-February, and the second follow-up the last week in
April. For the first follow-up, gaseous emissions were measured for the as-found condition using
both the house fuel and the reference fuel. Similar gaseous emission measurements were made
during the second follow-up and, in addition, particulate emissions were measured while firing
the reference fuel for comparison with the particulate measurements made during the initial visit.
These four burners were equipped with operating time clocks and cycle counters to
record operating experience between the first and last visits.
COMMERCIAL BOILERS
Although the instrumentation techniques and the emission measurements used for the
commercial boilers were similar to those employed for the residential units, the conditions under
which the measurements were made were quite different. In addition, SO2 emissions were
measured for the commercial boilers.
The commercial boilers were all investigated in the as-found condition, i.e., no cleaning or
other servicing of th'e boiler was done prior to testing. The exhaust stack from the breeching
through the sampling section was replaced with new stack on all units except Unit C2006. For
this unit the particulate sampling was done from the roof and it was not economically feasible to
replace the entire stack; however, a short section of stack was replaced with a section containing
particulate sampling ports.
The commercial measurements were conducted in late spring, 1972 (March 20 through
June 8), near the end of the heating season. Visual inspection of the units showed the normal
quantities of soot and ash one would expect to accumulate over a typical heating season. (Unit
C2003 was a new boiler that had only been fired briefly as a part of the manufacturer's
checkout operation. Hence, it was essentially clean.)
Gaseous emissions and smoke measurements were obtained for several excess air settings
at each of four loads and for several fuels to yield a fairly complete picture of the various
operating parameters.
-------
Table D-1. Information on Condition and Tuning of Residential Units
Atomizing Nozzle
Unit
23C
24°
25
26
27
28C
29
30
31C
32C
33C
34
35
Size,"
gph Cleaned
1.35
1.00
1.35 X
1.75
1.35 X
1.00
0.75
0.60
1.50
1.00
0.85
0.75
1.00
Pressure, psi
Replaced
X
X
X
X
X
X
X
X
X
X
X
As- Found
100
91
100
100
110
95
90
88
100
99
106
80
100
Tuned
100
100
100
100
100
100
100
100
100
100
90
90
100
Normal
Cleaning &
Electrode
Adjustment
X
X
X
X
X
X
X
X
X
X
X
X
X
Other Steps
_
Replaced fuel pump
Replaced fuel pump filter; heat exchanger
baffle plates cleaned
Replaced strainer and line filter; installed
check valve in fuel line
Installed new flame retention head
Replaced strainer; increased nozzle size
to 1 .35 gph
Replaced strainer; soot cleaner used in
cleaning heat exchanger
Replaced fuel pump, strainer, and two
line filters; increased nozzle size to
0.75 gph
Replaced strainer
Replaced line filter
Replaced strainer; resealed door and
burner assembly
Replaced line filter
-
Remark:6
Good condition; delay valve not working
properly
Average condition
Good condition; 447 gallons of fuel oil
delivered between initial visit and first
follow-up; stirred up sediment in tank
and created nozzle clogging problems
Good condition
Average condition; nozzle 1 month old
Average condition; winds gusting to
65 mph during tuned and reference runs
Average condition; unit had a No. 9 smoke
in spite of servicing only 3 days earlier
Poor condition; original nozzle had been
drilled by homeowner
Good condition
Poor condition; occasional puff, possibly
due to air leak in very old fuel line
Very poor condition
Average condition
Good condition
a Nozzle size, as found. Replacement nozzles were same size except as noted.
b Serviceman's estimate of condition.
c Owner has service contract, thus unit probably received an annual cleanup and tuning.
-------
D-6
The four load levels at which emission data were obtained for commercial boilers were:
R rated load
H a high load for normal operation (selected as 80 percent of rated load)
M an intermediate load
L normal low-fire setting.
Boiler C-2001 was fired with a burner having a fixed firing rate and therefore was only operated
at three loads: 100, 82, and 61 percent. The different loads were obtained by changing the
nozzle size.
For the oil-fired boilers, 12 percent CO2 was used at the baseline excess-air condition
(after discussions with the ABMA Commercial-Industrial Air Pollution Committee). Gaseous
emission measurements were made at several excess air levels, generally between 9 and 14
percent CO2. For gas-fired runs, 9 to 10 percent CO2 was used as the baseline, representing
an excess air level typical of normal boiler operation. Because particulate measurements
required extensive sampling times (approximately 80 minutes at each load setting), particulate
emissions were measured for only one excess-air level (12 percent CO2 for oil firing and 9 or 10
percent CO2 for gas firing) at one or two loads for each fuel fired in each boiler.
To assure the attainment of steady-state conditions when changing load or excess air, the
boilers were operated at the new setting for 30 minutes before particulate sampling was begun.
Each boiler (except C-2001) was fired with four different fuels: the typical house fuel, an
additional fuel oil ranging from No. 2 through No. 6 grades, a 1-percent sulfur reference residual
oil, and natural gas. A sample of each fuel oil was obtained for analysis of its physical properties
chemical composition.
-------
E-l
APPENDIX E
SAMPLING AND ANALYTICAL PROCEDURES FOR GASEOUS EMISSIONS
Sampling
Selection of installations with adequate space and accessibility made it possible to sample
the gaseous emissions from stacks directly through sampling probes into the monitoring equip-
ment (avoiding grab-bag sampling techniques). Direct sampling with the shortest possible lines
rninimized losses by condensation, reaction, and/or adsorption.
A schematic drawing of the sampling train used in monitoring gaseous emissions from
both residential and commercial units is shown in Figure E-l.
Stack
HC(high temp)
NO
N02to NO
converter -
1 _
1 r
L
co"\°2"\c°27 rs°2/~HC
-------
E-2
and a Dry Ice-cooled water trap. When the high-temperature hydrocarbon analyzer was used, a
third probe was employed which consisted of a 1/4-in. heated Teflon line connected directly to
the analyzer.
Analytical Methods
It is generally recognized that monitoring equipment with 100 percent reliability is still
lacking, especially for field measurements. Therefore, special care was taken in checking out,
tuning, and calibrating all instruments prior to each run. Zero and appropriate upscale span gases
were used for calibration.
The monitoring techniques used in the field on this program are listed in Table E-l. The
choice of techniques was based primarily on the emission range anticipated, interference-
correction requirements, and ease of use in the field.
Table E-1. Gaseous Emission Instrumentation Used in Field Survey
Pollutant Instrument
Range a Principle of Operation
Comments
CO BeckmanModel215A 0-1250 NDIR
Continuous, portable; water and
CO2 interference can be
accommodated
Total HC Beckman Model 109A
Beckman Model 402
0-120,000 Flame ionization
0-120,000 Flame ionization
Continuous, fast response, portable
Selectable elevated temperature
sampling line and oven
NO
Beckman Model 315L
0-750
NDIR
Continuous, portable; water and
CO2 interference can be
accommodated
NOX Faristor
Beckman Model 315L
0-2500 Electrochemical (dry)
0-750 NDIR with N02 to
NO converter
Continuous, fast response, portable;
SO2 interference can be
accommodated
Continuous, portable, water and
C02 interference can be
accommodated;
CO interference in converter can-
not be accommodated
SO2 Faristor
C02 Beckman Model 215A
0-2500 Electrochemical (dry)
0-20% NDIR
Continuous, fast response, portable;
no N02 interference
Continuous, portable; water inter-
ference can be accommodated
Beckman Model 715
0-25% Amperometric
Continuous, portable
(a) Ppm except as noted.
-------
E-3
Carbon Monoxide. Carbon monoxide was continuously monitored by nondispersive
infrared using a Beckman Model 215A analyzer. The instrument has two ranges, 0 to 250 and 0
to 1250 ppm. The sensitivity is 0.5 percent of full scale with an accuracy of ±1 percent.
Hydrocarbons, Total hydrocarbons were measured by flame ionization using two
Beckman analyzers - Model 109A and Model 402. The operation of the two analyzers is
basically the same, the primary difference being that the Model 402 utilizes a selectable
elevated-temperature sampling line and analyzer oven. Sampling at elevated temperatures (200 to
400 F) minimizes the loss of higher molecular-weight hydrocarbons. Both analyzers are usable
over a wide range of concentrations and have excellent response time and sensitivity. The most
sensitive range obtainable is 0 to 10 ppm carbon, while the least sensitive range is about 0 to
120,000 ppm carbon.
Nitrogen Oxide. Three techniques were used to continuously measure the nitrogen oxide
concentrations.
NO by NDIR analyzer
NOX by NDIR + converter
NOX by an electrochemical analyzer.
The nondispersive infrared (NDIR) analyzer, a Beckman Model 315L, has three ranges (0-150,
0-450, and 0-750 ppm). The unit in actuality measures nitric oxide (NO). However, when used in
conjunction with a NO2 to NO converter, it can also be used for measuring NOX.
The NO2 to NO thermal converter is a 6-ft-long coil of l/8-in.-diameter No. 316 stainless
steel tubing which is resistance heated to 650-700 C. A bypass valve on the converter permits
rapid switching between NO (bypassing the converter) and NOX modes. The internal surface of
the converter heater coil is stabilized when delivered; however, periodic reconditioning is
necessary. This is accomplished by passing NO or NO2/air through the converter for approxi-
mately 15 minutes while in the NOX mode.
The Faristor, an electrochemical analyzer, operates on the principle of a fuel cell. When
sample gas is passed through the detector, an electrochemical process within the detector
generates an electrical signal proportional to the NOX concentration in the gas sample. This signal
is amplified and then displayed on a meter having a 0 to 100 linear scale. Two range settings are
possible - 0 to 500 ppm for the low range and 0 to 2500 ppm for the high range.
Sulfur Dioxide. The Faristor electrochemical cell was also used to measure SO2. The
Series NS-200 SO2/Nitrogen Oxide Analyzer is bimodular, using two Faristor plug-in detectors,
Type N76H2 for measuring NOX and Type S64H2 for measuring SO2. Two analog outputs are
available on the Faristor permitting simultaneous monitoring of both gases.
-------
E-4
Carbon Dioxide. Nondispersive infrared was also used in monitoring CO2. The instru-
ment, a Beckman Model 215A, has two ranges (0 to 5 percent and 0 to 20 percent CO2 by
volume). The sensitivity is 0.5 percent of full scale with an accuracy of ±2 percent.
Oxygen. The Beckman Model 715 Process Oxygen Analyzer was used to continuously
monitor gaseous oxygen. The analyzer has ranges of 0 to 5 percent and 0 to 25 percent oxygen.
Accuracy is ±1 percent of full scale at a given sample temperature and ±6 percent of full scale
for sample temperature variations within the temperature range of 32 to 110 F.
The amperometric oxygen sensor contains a gold cathode and silver anode. The two
electrodes are separately mounted within the PVC body and are electrically connected by a
potassium chloride electrolyte. A gas-permeable Teflon membrane separates the electrodes from
the process sample and fits firmly against the gold cathode. Oxygen from the sample diffuses
through the membrane and is reduced at the gold cathode. The resultant electrical current flow
between the electrodes is proportional to the partial pressure of oxygen in the sample.
Response Times
Response times (to 90 percent for step changes of gaseous composition) for the various
gases and instruments (including the sampling probe and train) were as follows:
90 Percent Response Time,
Gas Analysis Technique seconds
CO NDIR 63
Total HC Flame-ionization detector 3
NO NDIR 60
SO2 Electrochemical cell 72
CO2 NDIR 57
O2 Amperometric 65
-------
F-l
APPENDIX F
SAMPLING AND ANALYTICAL PROCEDURES FOR PARTICIPATE
AND SMOKE AND DETAILED PARTICULATE DATA
Participate Sampling
The particulate sampling rig used in this investigation was the EPA sampling train2 -8, with
modifications for oil burner emissions measurements. Figure F-l shows the particulate sampling
train. For the residential runs, the sampling probes used were a combination nozzle and probe,
15 inches long, extending out of the top of the heated chamber of the sampling train. The
cyclone included in the EPA rig was not used and the probe was connected directly to the filter.
For the commercial runs, a 36-inch combination probe and nozzle was used and the cyclone was
used. The rest of the train was as shown in Figure F-l, and the procedures of operation (except
probe and impinger washing) were those specified by EPA.
Velocities in the stacks were measured with a S-type Pitot tube. As the firing of the
residential units was relatively constant from cycle to cycle, S-type Pitot tube measurements
were made during the cycle preceding the beginning of sampling. For sampling from commercial
boilers, the S-type Pitot tube and thermocouple were- attached to the probe and positioned
adjacent to the sampling nozzle.
For residential units, it was not always possible to get reliable velocity pressure readings,
as some flue gas velocities were so low that the Pitot tube velocity pressures were in the range of
0.002 to 0.005 inch of water. Hence, to insure isokinetic sampling, the fuel-oil firing rate was
measured, excess ah" was determined, and the flue-gas volume and average velocity were calcu-
lated and compared with the Pitot tube measurements. (For the as-found runs, the fuel rate was
measured after the run as a check on isokinetic sampling. For the tuned and reference-fuel runs,
this measurement was made before the run.)
Traversing was not done during particulate sampling from residential units. In cases where
the velocity profile across the duct was relatively uniform, the sample was collected by isokinetic
sampling at a velocity about equal to the average velocity. Sampling at the average velocity was
not possible where velocity profiles were not uniform.
For particulate sampling from commercial boilers, two traverses were made at 90° angles.
Sampling was done at four points on each traverse.
The sampling rig was operated in accord with the EPA recommendation, i.e., the filter
was kept at a temperature between 230 and 250 F and the impingers were immersed in an ice
bath. The first two impingers contained 100ml of double-dis-tilled water each, the third impinger
was initially dry, and the fourth impinger contained about 175 grams of Drierite, Moisture was
collected in the impingers and in the Drierite. The volume of dry flue gas sampled for the
particulate collection run was measured by a dry gas meter.
-------
F-2
Probe
Impingers
13
1. Stainless steel, buttonhook-type probe tip
2. Stainless steel coupling
3. Probe body, 5/8-inch OD, medium-wall
Pyrex tube logarithmically wound with
25 feet 26 ga, nickel-chromium wire
4. Cyclone and flask (not used for residential
units)
5. Fritted-glass filter holder
6. Electrically heated enclosed box
7. Ice bath containing four impingers
connected in series
8. The Greenburg-Smith type impinger with
tip removal
9. Second impinger with tip
10. Third impinger with tip removed
11. Fourth impinger with tip removed and
containing approximately 175 grams of
accurately weighed dry silica gel
12. Pressure gauge
13. Check valve
14. Flexible-rubber vacuum tubing
15. Vacuum gauge
16. Needle valve
17. Leakless vacuum pump
18. By-pass valve
19. Dry-gas meter
20. Calibrated orifice
21. Draft gauge
22. 5-type pilot tube
Figure F-l. EPA Particulate Sampling Train
-------
F-3
Before the sampler was put in operation, it was leak checked to see that all the joints
were tight. The nozzle was plugged with a rubber stopper and 10 inches of vacuum was drawn
on the system. With this vacuum, no more than 0.02 cfm leakage was permitted; checks showed
that actual leakage usually was much less than this value.
Silver membrane filters were used.* The filter for the residential unit was a 3-inch-
diameter silver membrane with a pore size of 0.8 micron. Because of the higher particulate
loadings expected when firing heavier fuels in the commercial boilers, a 5-inch-diameter silver
membrane filter with a pore size of 0.8 micron was used for those runs.
For the residential units, particulate sampling was initiated just before the start of the
10-minute "on" cycle and was stopped just after the "on" cycle was completed. During the
20-minute "off cycle, there was no particulate sampling. Particulate was sampled for six firing
cycles to collect sufficient particulate for weighing.
PARTICULATE ANALYSIS
Moisture Content of Flue Gas
After the particulate run was completed, the moisture content of the sampled gas was
determined by weighing the Drierite to determine the gain in weight and measuring the amount
of moisture in each of the impingers. These measurements were used to calculate the total
quantity of water in the sample volume and the moisture content of the flue gas.
Particulate Loading
The silver membrane filters were dried and the tare weight obtained before the run was
started. After the run, the filter was removed, dried, desiccated, and weighed to obtain the
quantity of material collected on the filter.
As a result of earlier particulate sampling at Battelle-Columbus, questions have arisen as
to whether the EPA procedure washes all particulate out of the probe. Hence, the particulate
sampling for this project was designed to permit samples to be obtained by both the EPA
method and a modified-EPA method (MEPA) which included more thorough washing of the
probe and impingers.**
Table F-l shows the steps involved in washing the probe and impingers and storing the
samples for later drying and weighing. The washing which generated the samples in Containers
2B and 5B were the additional washings not required by EPA.
*This is a departure from the glass-fiber filters specified by EPA2, but silver filters are superior for detailed
chemical analysis and are not hydioscopic, a characteristic of glass-fibei filters.
**The modified method is similar to the procedure used in Phase 1 (prior to publication of the EPA procedure).
-------
F-4
Table F-1. Procedures for Recovery of Participate Catches
Sample Recovery From Particulate Sampler.
Container No. 1. Remove the filter from its holder, place in this container, and seal.
Container No. 2A. Place loose participate matter and acetone washings from all sample-exposed surfaces prior to
the filter (probe and front half of filter) in this container and seal. Use a razor blade, brush, or rubber police-
man to loosen adhering particles.
Container No. 2B.a Wash the probe and the front half of the filter holder successively with:
1. Methylene chloride
2. Water
3. Acetone.
Place the washings in this container and seal.
Container No. 3. Measure the volume of water from the first three impingers and place the water in this con-
tainer. Place water rinsings of all sample-exposed surfaces between the filter and fourth impinger in this con-
tainer prior to sealing.
Container No. 4. Transfer the silica gel from the fourth impinger to the original container and seal. Use a
rubber policeman as an aid in removing silica gel from the impinger.
Container No. 5A. Thoroughly rinse all sample-exposed surfaces between the filter and fourth impinger with
acetone (but do not wash fourth impinger), place the washings in this container, and seal.
Container No. 5B.a Thoroughly rinse all sample-exposed surfaces between the filter and fourth impinger suc-
cessively with methylene chloride and acetone, place the washings in this container, and seal.
(a) These steps are not included in the EPA procedure but were included in this investigation.
Figures F-2 and F-3 show the drying procedure followed for each container obtained in
the washing procedure. The weights recorded as "EPA-Probe" and "EPA-Impinger" correspond
to the conventional EPA procedure for obtaining particulate weights. The weights recorded as
"MEPA-Probe" and "MEPA-Impinger" include the material obtained by the additional washing;
they were not a part of the conventional EPA procedure.
In each step, washing and drying, the above procedures were designed to follow the EPA
procedure first and then to add any additional steps recommended by the Battelle staff. Hence,
the additional steps in no way interfered with the EPA procedure.
RESULTS
Tables F-2 and F-3 present summaries of the results obtained using the above procedures
for both EPA procedures (EPA) and modified EPA procedures (MEPA). These tables include:
Probe catch
Filter catch
Impinger catch
-------
F-5
FILTER
J Dry at 212 F
J Oessicate 24 hr
Weigh; record as EPA-I
7 Seal; lable ; store in refrigerator
PROBE
EPA probe wash
Additional
probe wash
V I Clean empty dish; dry at 212 F; dessicate 24 hr
y I Weigh; record tare weight
f"^ 1 Add acetone wash; dry at
TT water \
PT^\ 1
\/ \o
w y-
J<^e1hyle/, \ I
u ^ -< ?/>, \ T
ambient temperature
Dessicate 24 hr
Weigh; record as EPA-2
V I Add organic extract; dry at 212 F
I
\ I Dessicate 24 hr
I
y 7 Weigh ; record as MEPA - 2
I
V I Cover; lable; store in refrigerator
Figure F-2. Procedures for Drying Filter and Probe Portions of Particulate Samples
-------
F-6
JMPINGER
Water \vOs/,
Chloroform Ethyl ether
Chloroforms
extract A
^ J Clean; dry at 212 F, dessicate 24 hr
I
V 7 Weight; record tare weight
*__,
Add water portion; dry at 212 F
Oessicale 24 hr
T
\ 1 Weigh; record as EPA-3B
Add organic portion; dry at
ambient temperature
Dessicate 24 hr
V 1 Weigh, record os EPA- 3 A
Add acetone wash; dry at
ambient temperature
Dessicate 24 hr
»
^ 7 Weigh; record as EPA-5
. \
\ 7 Add remaining wash, dry at
V / ambient temperature
y 7 Dessicate 24 hr
^ I Weigh; record as MEPA-5
*
\ ~7 Cover; lable; store in refrigerator
Figure F-3. Procedures for Drying Impinger Portions of Particulate Samples
-------
Table F-2. Summary of Paniculate Emission Data for Residential Units9
Mats of Paniculate Collected, mg
Probe Wnh
Unit
23A
23T
23R
23.2 R
24A
24T
24R
24.2R
2SA
25T
2SR
2S.2R
26A
26T
26 H
26. 2R
27A
27T
28A
28T
29A
29T
30A
30T
31A
31T
32A
32T
33A
33T
34A
34T
35T
3SR
EPA
3.9
1.6
1.3
2.0
1.3
3.7
1.7
SB
1.4
3.4
3.1
6.3
1.6
3.8
2.5
1.9
2.0
2.3
2.1
1.7
3.3
2.7
2.3
2.0
2.1
3.1
2.6
2.6
4.0
3.7
2.3
2.5
2.8
1.4
MEPA
9.8
6.2
5.2
10.1
5.3
9.4
9.7
5.1
5.6
6.4
6.8
22.4
5.8
6 .8
8.2
7.1
11.0
11.1
7.6
6.3
10.0
11.5
6.9
10.2
6.2
7.6
7.8
6.7
14.7
15.4
6.3
6.4
I8.G
13.7
Filter
3.1
1.9
2.6
6.1
3.1
11.0
14.4
5.6
0.4
1.1
1.6
67.8
6.0
4.5
4.8
3.4
8.8
4.5
8.7
2.1
2.5
39.3
4.3
5.1
2.7
3.7
3.6
9.8
14.4
11.3
20.0
9.2
2.1
1.9
Impinger Wash
EPA
9.3
19.1
15.6
21.6
10.8
29.2
10.7
20.6
21.0
11.3
9.0
36.5
26.8
9.7
12.1
13.6
27 S
10.1
26.5
12.3
24.9
23.7
28.8
25.1
26.7
46.6
440
31.8
118.6
101.4
32.8
31.0
59.4
39.5
MEPA
11.8
21.8
17.0
23.9
12.9
30.5
17.3
22.9
21.1
12.5
10.5
40.5
27.7
10.6
13.0
15.6
26.6
16.1
31 .6
18.1
23.7
23.8
29.9
26.2
27.1
44.3
43.1
32.6
106.8
99.3
30.8
29.5
59.7
33.9
Totals, mg
EPA
Filterable
7.0
3.5
3.9
8.1
4.4
14.7
16.1
9.4
1.8
4.5
4.7
74.1
7.6
8.4
7.3
5.3
10.8
6.8
10.8
3.8
5.8
42.0
7.1
7.1
4.8
6.8
6.2
12.4
18.4
15.0
22.3
11.7
4.9
3.3
Total
16.3
22.6
19.6
29.7
15.2
43.9
26.8
30.0
22.8
15.8
13.7
110.6
34.4
18.1
19.4
18.9
38.6
16.9
37.3
16.1
30.7
65.7
35.9
32.2
31.5
53.4
50.2
44.2
137.0
116.4
55.1
42.7
64.3
42.8
MEPA
Filterable
12.9
81
7.8
16.2
84
20.4
24.1
10.7
6.0
7.5
8.4
90.2
13.8
11.3
13.0
10.5
19.8
15.6
16.3
8.4
12.5
50.8
11.7
15.3
8.9
11.3
11.4
16.5
29.1
26.7
26.3
15.6
20.7
15.6
Total
24.7
29.9
24.8
32.0
21.3
S0.9
41.4
32.3
27.1
20.0
18.9
114.6
41.5
21.9
26.0
20.9
46.4
31.7
47.9
26.5
36.2
74.6
41.6
41.5
36.0
55.6
54.5
49.1
135.9
126.0
67.1
4S.1
80.4
49.5
Percent
Filterable
EPA
43
15
20
27
23
33
SO
31
8
28
34
67
22
46
38
28
28
40
29
24
19
64
20
22
15
13
12
28
13
13
40
27
8
8
MEPA
52
27
31
51
39
40
58
33
22
38
44
79
33
52
50
60
43
49
34
32
35
68
28
37
25
20
21
34
21
21
46
35
26
32
Grain Loading
EPA Procedure,
mg/sm3
Filterable
5.8
3.1
3.2
6.3
3.2
13.7
12.5
7.7
1.7
4.4
4.4
647
6.1
6.7
6.0
4.6
9.3
5.6
9.7
3.2
14.4
36.3
5.8
6.1
4.6
9.1
5.3
10.5
15.4
14.1
18.9
10.9
4.2
2.9
Total
13.5
19.7
16.1
23.0
11.0
41.0
20.9
24.6
21.0
15.3
12.8
96.6
27.4
14.5
16.0
16.4
33.1
13.9
33.6
13.5
76.0
56.8
29.5
27.8
30.0
71.2
42.9
37.3
114.3
109.7
46.8
38.9
55.2
37.6
Emission Factor, lb/1000 gal
EPA Procedure
Filterable Total
0.84
0.40
0.44
0.91
0.61
2.03
1.76
1.15
0.25
0.58
0.65
8.21
1.08
1.00
0.87
0.65
1.31
0.88
1.90
0.54
1.91
5.81
0.9B
1.10
0.77
1.23
a.98
1.60
3.24
2.83
2.49
1.44
0.61
0.39
1.94
2.60
2.22
3.33
2.09
6.05
2.93
3.97
3.14
2.03
159
12.25
4.91
2.16
2.32
2.33
4.70
2.18
6.55
2.30
10.13
9.09
4.94
4.98
5.05
9.64
8.00
5.71
24.11
21.95
6.15
5.11
8.04
509
MEPA Procedure
Filterable Total
1.56
0,93
0.8B
1.82
1.17
2.82
2.64
1.31
0.83
0.97
1.16
9.99
1.96
1.35
1.55
1.29
2.40
2.02
2.87
1.19
4.12
7.03
1.62
2.37
1.43
2.04
1.80
2.13
5.12
5.04
2.94
1.92
2.58
1.84
2.94
3.44
2.82
3.E9
2.93
7.02
4.53
4.27
3.73
2.57
2.61
12 89
5.97
2.61
3.11
2.58
5.65
4.09
8.41
3.79
11.95
10.32
5.72
6.42
5.77
10.04
8.69
6.34
23.92
23.76
6.37
5.40
10.05
5.89
Bacharach
Smoke No.
at 9Mln
0.3
0.5
0,4
0.0
0.4
0.7
0.8
0.7
1.0
1.7
0.9
5.8
0.5
0.4
0.5
0.4
1.0
0.4
1.3
0.4
9.0
0.5
0.2
0.6
00
0.2
0.2
0.1
1.5
1.8
2.7
1.0
0.2
0.2
a EPA refers to EPAProcedure specified in Reference 2. MEPA refers to "Modified EPA Procedure" described in text.
-------
Table F-3. Summary of Paniculate Emission Data for Commercial Boilers3
(Mass of Particulate Collected
Probe Wash
Boiler
C2001
C2001
G2001
C2002
C2002
C2002
C2002
C2002
C2003
C2003
C2003
C2003
C2003
C2004
C2004
C2004
C2004
C2004
C2005
C2005
C2005
C2005
C2005
C2006
C2006
C2006
C2006
C2006
Fuel
No. 2
No. 2
Gas
No. 2
No. 4
No. 4
CR
Gas
No. 2
CR
No. 6
No. 6
Gas
No. 2
No. 2
CR
No. 5
Gas
No. 2
CR
No. 6
No. 6
Gas
No. 2
CR
No. 6
No. 6
Gas
Load
H
L
H
H
H
L
H
H
H
H
H
L
H
H
L
H
H
H
H
H
H
L
H
H
H
H
L
H
EPA
2.4
2.4
1.5
4.6
1.6
1.2
4.7
3.6
7.8
95.6
205.6
53.1
3.7
7.3
2.4
16.8
38.3
1.4
1.3
2.4
19.7
6.2
1.7
0.8
22.7
76.1
20.8
3.4
MEPA
12.6
11.2
6.6
9.0
8.9
10.5
12.8
8.9
14.0
102.8
225.4
66.2
10.5
14.0
7.1
23.6
48.7
6.9
6.2
6.3
24.7
11.8
7.1
7.1
27.6
83.9
22.0
8.4
Filter
5.4
9.2
4.2
16.5
34.1
0
228.9
2.0
6.6
261.6
695.3
95.6
0.7
1.9
12.2
58.3
97.4
1.0
4.0
41.6
76.1
40.8
0.9
6.2
50.5
93.1
19.5
4.5
Impinger Wash
EPA
68.9
17.0
25.4
23.0
15.4
55.8
32.2
21.4
14,1
24.5
27.0
70.9
11.6
7.2
42.3
40.5
68.6
4.3
31.8
23.5
50.4
sa.o
13.5
30.8
29.8
38.3
83.4
11.8
MEPA
66.4
20.5
27.7
25.6
19.9
61.1
36.1
25.5
17.5
27.2
30.1
67.9
15.2
9.0
42.0
40.9
66.3
6.2
32.4
25.7
49.3
49.2
13.7
30.8
31.6
37.1
67.8
13.4
EPA
Filterable
7.8
11.6
5.7
21.1
35.7
1.2
233.6
5.6
14.4
357.2
900.9
148.7
4.4
9.2
14.6
75.1
13B.7
2.4
5.3
44.0
95.8
47.0
2.6
7.0
73.2
169.2
40.3
7.9
Totals, mg
MEPA
Total
76.7
28.6
31.1
44.1
51.1
57.0
265.8
27.0
28.5
381.7
927.9
219.6
16.0
16.4
56.9
115.6
204.3
6.7
37.1
67.5
146,2
97.0
16.1
37.8
103.0
207.5
123.7
19.7
Filterable
18.0
20.4
10.8
25.5
43.Q
10.5
241.7
10.9
20.6
364.4
920,7
161.8
11.2
15.9
19.3
81.9
146.1
7.9
10.2
47.9
100.8
52.6
8.0
13.3
78.1
177.0
41.5
129
Total
84.4
40.9
38.5
51.1
62.9
71.6
277.8
36.4
38.1
391.6
950.8
229.7
26.4
24.9
61.3
122.8
212.4
14.1
42.6
73.6
150.1
101.8
21.7
44.1
109.7
214.1
109.3
263
Percent
Filterable
EPA
10
41
18
48
70
2
88
21
51
94
97
68
28
56
26
65
66
36
14
65
66
48
16
19
71
82
33
40
MEPA
21
50
28
50
68
15
87
30
54
93
97
70
42
64
31
67
69
56
24
65
67
52
37
30
71
83
38
49
Grain Loading,
EPA Procedure,
mg/sm3
Filterable
4.6
9.2
2.6
9.6
32 .8
1.4
103.
2.4
11.2
286.
722.
207.
4.0
9.3
6.5
67.6
134.
2.2
5.4
44.6
102.
40.9
2.7
3.3
66.6
235
73.5
45
Total
45.5
227
14.2
20.0
46.9
64.1
117.
11.4
22.2
305.
744.
306.
14.7
166
25.2
104.
202.
6.2
37.5
68.5
156.
84.3
16.7
17.8
93.7
288
226.
11.1
Emission Factor, lh/1000 g
EPA Procedure
Filterable
0.5
1.0
0.3
1.1
3.7
0.2
11.7
0.2
1.2
32.9
87.2
27.0
0.4
1.0
0.8
7.7
16.1
0.2
0.6
6.1
12.6
5.0
0.3
0.4
7.5
27.8
8.6
0.5
Total
4.9
2.4
1.5
2.2
5.3
7.2
13.3
1.1
2.4
36.0
89.9
39.9
1.5
1.8
3.0
11. a
24.3
0.6
4.0
7.8
19.3
10.3
1.6
1.2
10.6
340
26.4
1.2
al»
MEPA Procedure
Filterable
1.2
1.8
0.6
1.3
4.5
1.8
12.1
0.4
1.7
33.6
89.1
29.4
1.0
1.7
1.1
8.4
17.3
0.7
1.2
6.6
13.3
5.6
0.9
0.8
8.0
29.1
8.9
0.8
Total
5.4
3.4
1.9
2.5
6.5
90
13.9
1.5
3.2
35.9
92.1
41.7
2.5
2.7
3.2
12.5
25.3
1.3
4.6
8.5
19.8
10.8
2.2
1.4
11.3
35.1
23.3
1.6
Bacharach
Smoke
No.
2.4
2.8
0.2
0.6
2.3
2.8
3.0
0.1
2.0
3.1
4.2
5.7
0.0
0.2
3.4
2.9
5.0
0.0
0.0
2.5
3.8
3.4
0.1
0.4
3.4
4.1
5.2
0.0
71
DO
a EPA refers to EPA Procedure specified in Reference 2. MEPA refers to "Modified EPA Procedure" described in text.
b Emission factor for gas firing for this table is lb/(145 x 106 Btu), about the same on a Btu basis as lb/1000 gal for oil firing.
-------
F-9
Filterable catch, probe plus filter
Total catch
Percent filterable
Grain loading
Emission factor
Smoke number, Bacharach.
The results of Tables F-2 and F-3 show that the revised washing procedure (used in the
"modified" EPA method) did wash more material from the probe than was obtained using
standard EPA washing procedures. Comparative results obtained by using the two washing
procedures are as follows:
Participate Emission Factors, lb/1000 gal
Average for residential units
Average for commercial boilers
firing No. 2
Average for commercial boilers
firing No. 4 CR
Average for commercial boilers
firing No. 5 and No. 6
Filterable
EPA
1.51
0.83
MEPA
2.43
1.35
Difference
+0.92
+0.52
EPA
5.88
2.74
Total
MEPA
6.76
3.30
Difference
+0.88
+0.56
9.83 10.57
26.33 27.53
+0.74
+ 1.20
13.00 13.94
34.87 35.44
+0.94
+0.57
On the average, sufficient additional particulate was washed from the probe by the extra
washings used in the modified procedure to increase emission factors nearly 1.0 lb/1000 gal. The
additional waitings did not increase the particulate associated with the impinger catch.
Impinger Catch
The filterable ft. otion of the total particulate catch averaged 28 percent (EPA procedure)
or 38 percent (MEPA procedure) for residential units and 54 percent (EPA procedure) or 58
percent (MEPA procedure) *br oil-fired boilers. Questions arise as to the composition of the
remainder of the total partifilate catch - that portion associated with the impinger. The
detailed analyses that would be r.. quired to determine the exact nature of the impinger catch was
not a part of this study. Howeve* some analyses of impinger samples were made during the
Phase I study.
Analyses of the impinger samples, for C, H, and N accounted for 11 percent of the
impinger catch for samples from residential units and 33 percent of the impinger catch for
samples from oil-fired commercial boilers. Sulfur analyses were made on impinger samples from
the commercial boilers.* These analyses identified sulfur present in quantities that represented 15
to 99 percent of the impinger catch. Assuming that the sulfur is present as SO3 (possibly in
H2SO4 or sulfates), the quantities of sulfur compounds present were calculated to be 38 to 248
*One-half of the impinger catch was dried (within about 2 months of the sampling) to obtain the weight of the
particulate catch. The remaining one-half of the impinger catch was analyzed for total sulfur about 6 months
after the sampling.
-------
F-10
percent of the impinger catch. The problem associated with finding SO3 in quantities greater
than the mass of dried particulate illustrates the difficulties encountered in making sense out of
impinger catch data. Apparently, large quantities of S02 are retained in the impinger solutions
and, over a period of time, some of the SO2 is oxidized to S0315. It may have been that some
SO2 was driven off those portions of the samples used for the particulate weight determinations
during the drying operation. However, during the interval about 4 months before the sulfur
analyses were made on the remaining portions of the samples, additional SO2 may have oxidized.
Thus, the latter analyses gave sulfur quantities higher than the total particulate from the former
analyses.
SMOKE SAMPLING TECHNIQUES
Two techniques for sampling smoke were used. Use of the standard Bacharach hand-
pump smoke meter and ASTM Procedure D2156-657 was the primary method of smoke spot
sampling. The ports for the gaseous measurements were used as sampling locations for smoke
measurements. If conditions suggested erroneous readings due to greatly disturbed air flow, the
sample was taken through the particulate sampling port. However, in most cases the gaseous
ports were adequate.
For the residential units, Bacharach smoke readings were made at the 1-minute, 5-minute,
and 9-minute points of the "on" cycle. As expected, the smoke number usually decreased as the
cycle proceeded. The decrease between the 1- and 5-minute points generally was much greater
than that between the 5- and 9-minute points.
To enable the field team to monitor the variations in smoke during the cyclic runs, a
Von Brand continuous recording smoke meter was used. During a run, the vacuum pump was
started 15 seconds prior to ignition. This permitted observation .of smoke levels for the air in the
stack prior to firing of the burner. Smoke levels were monitored continuously throughout each
on-cycle until all evidence of smoke associated with shutdown had ceased. This smoke persisted
from 1 second to 30 seconds. The smoke associated with burner startup and shutdown coincided
with the CO and HC peaks noted for the gaseous emissions.
A Photovolt reflectometer with tricolor filter was used to obtain numerical reflectance
values from the Bacharach smoke spots according to ASTM standard procedures7. This provided
a means for accurately reading the smoke spot while reducing the possibility of human error
caused by inadequate lighting, colored material, or oil on the smoke spot.
An attempt was made to correlate reflectometer readings from the Bacharach smoke
spots to similar readings of the Von Brand tape traces. A good correlation was not obtained for
the low smoke levels of interest (below about a No. 3 or 4 Bacharach smoke).
15 Hillenbrand, L. J., Engdahl, R. B., and Barrett, R. E., "Chemical Composition of Particulate Air Pollutants
from Fossil Fuel Combustion Sources", Final Report on EPA Contract EHSD 71-29, March 1, 1973.
-------
K-l
APPENDIX K *
DETAILS OF EMISSION-FACTOR CALCULATIONS
AND CONVERSION FACTORS
Emission factors are commonly defined in terms of the weight of pollutant emissions per
unit of fuel input - either weight, volume, or Btu heating value of the fuel. For purposes of this
report, emission factors for oil-fired equipment are expressed in terms of pounds of pollutant per
thousand gallons of fuel oil input, as this is the basis of the published emission factors compiled
by EPA3, and it also is convenient in developing emission inventories.
Calculation of the emission factors requires data on (1) the concentration of the
pollutant (e.g., ppm for gaseous pollutants or dust loading in mg/sm3 for particulate) and (2) the
rate of generation of flue gas. This latter quantity may be determined by one of several
approaches, depending on data available. Four alternate approaches are feasible, requiring the
following information:
(1) Measurement of flue-gas flow rate, plus flue-gas analysis (used where
fuel composition and firing rate are unknown)
(2) Measurement of flue-gas flow rate, plus fuel firing rate
(3) Fuel firing rate, plus flue-gas analysis (the most frequently used
technique)
(4) Fuel composition (C-H), plus flue-gas analysis.
Of these approaches, the fourth was chosen as being consistently the most accurate for the
conditions of the field investigation. This approach is outlined in the following section.
Combustion Calculations
The combustion equation, assuming that all of the sulfur and nitrogen from the fuel
appear as SO2 and NO in the flue gas and that CO and HC in the flue gas are negligible, is
CHESFNG + (1.0 + 0.25E + F + 0.5G + X)O2
+[(1.0 + 0.25E + F + 0.5G + X) 3.77]N2 = CO2 + 0.5E-H2O
+ F-SO2 + G-NO + X-O2
+[(1.0 + 0.25E + F + 0.5G + X) 3.77]N2
where E = ratio of hydrogen atoms to carbon atoms in the fuel,
(*) Appendices G through J are not included in this volume; they appear in
the Data Supplement Volume which is available as a separate report
(EPA-R2-73-084b).
-------
K-2
£ _ 12.01 (wt% H in fuel)
1.008 (wt%Cinfuel)
F = ratio of sulfur atoms to carbon atoms in the fuel,
F= 12.01 - (wt%S in fuel)
r 32.06 (wt%C in fuel)
and G = ratio of nitrogen atoms to carbon atoms in the fuel
12.01 (wt%N in fuel)
14.008 (wt%C in fuel)
and X = ratio of percent O2 to percent CO2 in flue gas.
From this equation, a predicted CO2 and O2 concentration for the dry flue gas can be
expressed as
i r\f\
C02 in flue gas, percent = 4777T4777XT0.94E + 4.77F~l89G
_ . _ t _ 100X _
02 in flue gas, percent = 4.77 + 4.77x + 0.94E +~4.77F + 2.89G
Solving each of these equations for X gives
100 - CO, [4.77 + 0.94E + 4.77F + 2.89G]
- , . .
X (based on CO2 measurement) = - ^77
O2 [4.77 + 0.94E + 4.77F + 2.89G]
X (based on O2 measurement) = --- J-QQ , , ~~ TT-^T -
where CO2 and O2 are the CO2 and O2 concentrations in the flue gas. Excess air (EA) in terms
of the above combustion equation is defined as
EA, percent = j;0 + 0.25E -TFT"a5G~
Thus, excess air can be calculated either based on CO2 or O2 readings from the flue-gas analysis
as
100-XCo2 __
EAC02, percent I.Q + Q.25E + F + 0.5G
or
100- X0
EA02, percent = or^SE + F + 0.
5G.
-------
K-3
Values for excess air calculated by these two routes did not always agree for this investigation.
(Usually, the difference was small, less than 10 percent). The reliability of the CO2 and O2 field
measurements were considered and judged to be similar. Thus, an average value was used to
define the actual excess air:
EA', percent =
Likewise,
+X°2
The weight (W) of dry flue gas generated in pounds per pound of fuel fired at this
average excess air is
W^(1.Q.44.01)+(F.64.06)+(G-30.01)+(X'.32.0)+[(1.0+Q.25E+F-K).5G+X')3.77-28.02]
(12.01)+(E.1.008)+(F-32.06)+(G. 14.01)
The volume (V) of dry flue gas generated (standard cubic feet per pound of fuel fired) is
W
V =
0.075 '
^12% co and V3% o, are defined as the volume of dry flue gas generated per pound of fuel
fired at 12% CO2 and at 3% O2 in the flue gas, respectively.
Emission Factor Calculations, lb/1000 gal
Gaseous emission factors, EF, can now be calculated as
nr: /iu 11 + W.nnn if n V PPM MW FD
EFa (Ib pollutant/ 1000 gal fuel) = - 3g6 . 1000 -
where PPM = concentration of pollutant in flue gas (dry)
MW = molecular weight of pollutant
FD = fuel density in Ib/gal.
Particulate emission factors, EF', can be calculated as follows:
EF'a (Ib particulate/1000 gal fuel) =
where GL = grain loading measured in sampled gas in mg per standard cubic meter (dry
at 77 F).
-------
K-4
Converting to Other Emission Factor Units
Emission factors in lb/1000 gal can be converted to other units as follows:
Gaseous Emissions.
EFb (kg pollutant/1000 liter fuel) = EFa .0.1198
EFa
EFC (Ib pollutant/1000 Ib fuel) = j^
EF
EFd (g pollutant/kg fuel) = EFC = -^
EF, 1000
EFe (Ib pollutant/106 Btu) = rr , Cf , /P+ < n
e Heating value of fuel (Btu/gal)
It is sometimes convenient to express emissions in ppm at some standard condition, for example:
PPM V
EFf (ppm pollutant at 12% CO2!. dry) = ~
v 1 2 % c o 2
EFa 386 1000
= MW.FD.V12%C02
and
PPM V
EF. (ppm pollutant at 3% O5, dry) = r;
9 V3% o2
EFa 386 1000
MW.FD.V3%02
Paniculate Emissions. Similarly for particulates, EFb', EFC', EFd', and EFB' follow from
EFg'. Additional emission factors are sometimes used for particulate emissions as follows:
EFh' (Ib particulate/106 scf flue gas at 12% CO2, dry)
^ GL 1000 V
15800 'V12%C0
or
EFar - 1000
FD V
12% CO
2
-------
K-5
and EF,' (Ib particulate/106 scf Hue gas at 3% O2 dry)
GL- 1000
15>800 'V3%0.
EF ' 1000
FD 'V3% o2
Conversion Factors for No. 2 Fuel Oil
For a typical No. 2 fuel oil of 33 degrees API gravity,
FD =7.17 Ib/gal
Heating value = 139,900 Btu/gal
Analysis = 87.0% C
12.6% H
0.17% S
0.007% N
Vi2%co2 = 243 cu ft/Ib fuel
vs% o2 = 219 cuft/lb fuel.
Then
EFC (Ib pollutant/ 1000 Ib fuel) = 0.139 EFa
EFd (g pollutant/kg fuel) = 0.139-EFa
EFC (Ib pollutant/ 106 Btu) = 0.0071 5 -EFa
222-EF.
EF, (ppm pollutant at 1 2% CO2 , dry) = jpy
246 -EF
EFg (ppm pollutant at 3% O2 , dry)
and
EFh' (Ib particulate/106 scf flue gas
at 12%C02)dry) = 0.574-EFa'
EF? (Ib particulate/106 scf flue gas
at 3% O2, dry) = 0.637.EFa'
-------
K-6
Conversion Factors for No. 6 Oil
For a typical No. 6 fuel oil of 15 degrees API gravity,
FD = 8.05 Ib/gal
Heating value = 151,200 Btu/gal
Analysis = 86.8% C
11.1% H
1.70%S
0.30% N
vi 2% co = 243 cu ft/lb fuel
vs% 02 = 214 cu ft/lb fuel
Then
EFC (Ib pollutant/ 1000 Ib fuel) = 0.124 EFa
EFd (g pollutant/kg fuel) = 0.124 EFa
EFe (Ib pollutant/106 Btu) = 0.00661 EFa
197-EF
EF, (ppm pollutant at 12% CO2, dry) =
and
224-EF
EFg (ppm pollutant at 3% O2 , dry) =
EFh' (Ib particulate/106 scf flue gas at 12% CO2, dry)= 0.511-EFa'
EF[ (Ib particulate/106 scf flue gas at 3% O2, dry) = 0.580-EF,',
Emission Factor Calculations, Gas Firing
Emission factors can be calculated for gas firing in a similar manner as for oil firing
(although other calculation-procedures may be simpler for gas firing, only). For example,
EFr (Ib pollutant/1000 Ib fuel)
and
V.GT
EFr'(Ib paniculate/1000 Ib fuel) =
-------
K-7
Assuming
Density gas = 0.0458 Ib/ft3, standard conditions
Heating value = 1020 Btu/ft3
Analysis = 75% C
25% H
V3% o2 = 246 cu ft/lb fuel
Then
EFS (lb pollutant/106 cu ft) =45.81 EFr
or
_ V-PPM.MW
8430
likewise,
EFj (lb particulate/106 cu ft)
and
EFt (lb pollutant/109 Btu) = 44.91 EFr
or
_ V-PPM-MW
8590
likewise,
EFt' (lb particulate/109 Btu)
and
EF...386-1000
EFU (ppm pollutant at 3% Cfe, dry) = «,,-,
iVlW-V
or
1570-EFr
MW~
PPM-V
246
-------
K-8
and
(lb particulate/106 scf flue gas at 3% O2> dry) = ^-L
'3% O2
= 4.065 -EFr
or
V-GL
3890
Summary of Conversion Multipliers for Oil Firing
Table K-l summarizes conversion multipliers for some of the common methods of
expressing emission factors.
Table K-1. Multipliers to Convert Emission Factors From
Lb/1000 Gal To Other Units
To Obtain Emission Factor
in These Units
Kg /1000 liter fuel
Lb/1000 lb fuel
Gm/kg fuel
Lb/106 Btu input
Gaseous pollutants'c)
ppm at 12% CO2
Ppm at 3% 02
Particulates
Lb/106 scf flue gas at 12% CO2
Lb/106 scf flue gas at 3% O2
Multiply Emission
lb/1000 Gal Fuel
No. 2 Oil's I
0.1198
0.139
0.139
0.00715
222
MW
246
MW
0.574
0.637
Factor in
-------
L-l
APPENDIX L
ERRATA TO PHASE 1 REPORT
During the interval between the publication of the report on Phase I of this study and
the present time, several points were identified where clarification or corrections are in order for
information contained in the Phase I report1. The following sections discuss these points and
identify corrections as necessary.
Wherever Phase I results have been incorporated in this Phase II report, the corrected
values have been used.
CO Emission Data
Emission values for CO in the Phase I report (Tables IV-1 and VI-1) are high, as the raw
NDIR data were not corrected for CO2 interference. One percent of CO2 produces an equivalent
output from the CO NDIR as 1.25 ppm CO. Hence, about 10 percent CO2 produces CO readings
that are high by 12.5 ppm. Consequently, CO emission factors tabulated in Tables IV-3 and VI-4
are high by the ratio of CO uncorrected to CO corrected. The CO emission factors for oil-fired
residential units and commercial boilers are high by about 2.0 lb/1000 gal.
Table 1-3, Emission Factors From
Residential Furnaces
The NOX emission factors reported in Table 1-3 for the APCO Research Laboratory
studies are incorrect. Correct values are as follows:
Modern burners, Reference 4 Average 16.1
Range 13.4-18.8
Modern burners, Reference 5 Average 12.5
Range 8.5 - 15.5
-------
L-2
Table VI-7, Emission Factors for Gas-Fired
Commercial Boilers
Table IV-1. Corrected Emission Factors for Phase I
Gas-Fired Commercial Boilers
Emission Factors, lb/106 cu ft, natural gasa
Paniculate
Unit
C1004
Condition
H
M
L
Mean
CO
3.6
6.2
2.7
4.2
HC
0.6
0.5
0.6
0.6
SO2
40
38
47
42
IMOX
53
40
39
44
Filterable
9
-
3
6
Total
19
-
17
IS
Emission factor in lb/10^ Btu input.
Nitrogen in Distillate Fuel Oils
Kjeldahl analyses of the No. 2 fuel oils fired in the residential units indicated consistently
lower values for fuel nitrogen for the Phase II fuels compared to the Phase I fuels. The Phase II
fuel nitrogen analyses were confirmed by a round-robin program, with analysis by five different
laboratories. Repeat analyses of two Phase I fuels produced fuel nitrogen levels that were lower
than those reported in the Phase I report by a factor of about 5. Hence, it must be concluded
that fuel nitrogen analyses of the No. 2 fuel oils fired during Phase I are high by a factor of
about 5.
Ratio of NO to NO2
Measurements of NOX for the Phase I study indicated that about 22 percent of the NOX
was emitted as NO2. Phase II measurements, which used improved methods for measuring NO
and NOX emissions, gave a ratio of NO2 to NOX of about 0.2. However, there is a possible
interference of CO with NOX measurement when large concentrations of CO are present. When
the Phase II data for units emitting less than 100 ppm of CO are considered, the ratio of NO2 to
NOX is only 0.09. Hence, it is concluded that NO2 emissions are usually less than 10 percent of
the total NOX for residential furnaces.
-------
M-l
APPENDIX M
REFERENCES
1. Levy, A., Miller, S. E., Barrett, R. E., Schulz, E. J., Melvin, R. H., Axtman, W. H., and
Locklin. D. W., "A Field Investigation of Emissions from Fuel Oil Combustion for Space
Heating", API Publication 4099 (November 1, 1971), available from the API Publications
Section, 1801 K Street, N.W., Washington, D. C. 20006.
2. "Standards of Performance for New Stationary Sources", Federal Register, Vol. 36, No.
139, Part II, pp 24876-24895, December 23, 1971.
3. "Compilation of Air Pollutant Emission Factors", U. S. Environmental Protection Agency,
Office of Air Programs Publ. No. AP-42 (February 1972).
4. "The Typical Oil Burner", Fueloil and Oil Heat, Vol. 31, No. 6 (June 1972), pp 4445.
5. Special Study, Fueloil and Oil Heat, Vol. 30, No. 1 (January 1971), pp 22, 24.
6. "Standard Method of Test for Effect of Air Supply on Smoke Density in Burning Distillate
Fuel", ASTM D2157-65(70).
7. "Standard Method of Test for Smoke Density in the Flue Gases from Distillate Fuels",
ASTM D2156-65(70).
8. "Standards of Performance for New Stationary Sources", Federal Register, Vol. 36, No.
139, Part II, pp 15704-15722, August 17, 1971.
9. "Standards of Performance for New Stationary Sources", Federal Register, Vol. 37, No. 55,
Part I, pp 5767, March 21, 1972.
10. Martin, G. B., and Berkau, E. E., "An Investigation of the Conversion of Various Fuel
Nitrogen Compounds to Nitrogen Oxides in Oil Combustion", presented at AIChE Meeting,
Atlantic City, N.J., August 30, 1971.
11. Turner, D. W., Andrews, R. L., and Siegmund, C. W., "Influence of Combustion Modifica-
tion and Fuel Nitrogen Content on Nitrogen Oxides Emissions From Fuel Oil Combustion",
presented at AIChE Meeting, San Francisco, November 28-December 2, 1971.
12. Bartok, W., Crawford, A. R., Cunningham, A. R., Hall, H. J., Manny, E. K., and Skopp, A.,
"Systems Study of Nitrogen Oxide Control Methods for Stationary Sources", Final Report,
Esso Res. & Engrg. Co., November 20, 1969, NAPCA Contract PH-22-68-55.
13. Private communication: to D. W. Locklin, Battelle-Columbus, from Margaret Mantho,
Fueloil & Oil Heat, May, 1972. Similar data were published in Reference 4.
14. "Combustion Chambers by Types", Fueloil & Oil Heat, Vol. 31, No. 5, March 1972, p. 94.
15 Hillenbrand, L. J., Engdahl, R. B., and Barrett, R. E., "Chemical Composition of Particulate
Air Pollutants from Fossil Fuel Combustion'Sources", Final Report on EPA Contract EHSD
71-29, March 1, 1973.
-------
M-2
BIBLIOGRAPHIC DATA ' Report No.
SHEET EPA-R2-73-084a
Field Investigation of Emissions from Combustic
Equipment for Space Heating
7. Author(s)
R. E. Barrett , S. E. Miller , and D. W. Locklin
9. Performing Organization Name and Address
Battelle - Columbus Laboratories
505 King Avenue
Columbus, Ohio 43201
12. Sponsoring Organization Name and Address
EPA, Office of Research and Development
NERC-RTP, Control Systems Laboratory
Research Triangle Park, North Carolina 27711
15. Supplementary Notes
2- 3- Recipient's Accession No.
5- Report Date
m June 1973.
6.
8. Performing Organization Rept.
No.
10. Project/Task/Tork Unit No
11. Contract/Grant No.
68-02-0251
13- Type of Report & Period
Covered
Final
14.
is. Abstracts The report gives results of a 2-year field investigation of air-pollutant
emissions from 33 residential heating units and 13 commercial boilers. It includes
effects of combustion parameters and fuel-oil compositions, as well as measurements
of CO, HC, NOx, SO2, particulate , and smoke. The largest residential emissions
reduction resulted from replacing poorly perform ing units.
Burners with flame -retention combustion heads had lower overall emissions than those
with conventional heads; newer burners had lower emissions than older ones. Emissior
from commercial boilers (40-600 boiler hp) were measured for 33 combinations of
boilers and fuels at various loads and excess air settings. Generally, operating
parameters within the normal adjustment range had minor effect on emissions. Fuels
investigated were natural gas and five grades of fuel oil, including a 1%-S residual
oil. Fuel characteristics significantly affected emissions , especially particulate and
17. Key words and Document Analysis. 17a. Drscriprars
Air Pollution Nitrogen Oxides
Space Heating Nitrogen Oxide (NO)
Residential Buildings Nitrogen Dioxide
Commercial Buildings Particulate Composites
Combustion Smoke
Emission Carbon Monoxide
Burners Hydrocarbons
Boilers Sulfur Oxides
Furnaces Sulfur Dioxide
17b. [dencif iers/Open-Hnded Terms
Air Pollution Control No. 5 Oil
Stationary Sources No. 6 Oil
Emission Factors
No. 2 Oil
No. 4 Oil
17c. COSATI Field/Group 13B, 7C
18. Availability Statement
Unlimited
INUX. £or oil iiring, INUX
emissions increased nearly
linearly with fuel-N content.
Sulfur Trioxide
Natural Gas
Fuel Oil
19. Security Class (This 21- No, of Pages
Report) 1QQ
UNCLASSIFIED laa
20. Security Class (This 22. Price
Page
UNCLASSIFIED
{[ G. P. O. 1974 747-791 / 3AB. REGION NO. 4
USCOMM-DC 14952-P72
------- |