EPA/600/R-08/069
September 2008
Characterizing Emissions from the Combustion of
Biofuels
Prepared by
C. Andrew Miller
U.S. Environmental Protection Agency
Office of Research and Development
National Risk Management Research Laboratory
Air Pollution Prevention and Control Division
Research Triangle Park, NC
-------
Abstract
Emissions from two biofuels, a soy-based biodiesel and an animal-based biodiesel, were
measured and compared to emissions from a distillate petroleum fuel oil. The three fuels
were burned in a small (3.5xl06 Btu/hr) firetube boiler designed for use in institutional,
commercial, and light industrial applications. Emissions were measured for carbon
monoxide (CO), carbon dioxide (CO2), oxides of nitrogen (NOX), and sulfur dioxide
(SO2) using continuous emission monitors. Concentrations and size distributions of
particulate matter (PM) were also measured. Flue gas samples were collected and
analyzed to determine concentrations of aldehydes and other volatile organic compounds,
polycyclic aromatic hydrocarbons, and polychlorinated biphenyls. The boiler efficiency
was also determined for operation using each of the three fuels. The most significant
difference was for PM, where the distillate fuel oil had emissions roughly ten times
higher than for either of the two biodiesel fuels. The particle size distributions
(measuring particle volume) showed a mode near 1 jim for the two biodiesels and near
2.5 jim for the distillate fuel oil. All three fuels also had a mode near 20 nm. SO2 was
nearly four times higher for the distillate petroleum fuel oil than for either the soy or
animal biodiesel. NOX emissions were slightly higher for the distillate fuel oil than for
the two biodiesels, but all three were within 6% of one another. CO and CO2
concentrations were approximately the same for the three fuels. The differences in
concentrations of the organic compounds were relatively small, with the emissions
patterns being similar for all three fuels. Boiler efficiencies were also similar for the
three fuels, with any difference being within the unit's measured variability range. In
general, the two biodiesel fuels emitted less pollutants than the distillate fuel oil, and the
low life-cycle CO2 emissions for the biodiesels results in a net CO2 reduction of nearly
75% when using these fuels compared to the petroleum distillate fuel.
11
-------
Table of Contents
Abstract ii
List of Figures iv
List of Tables v
Nomenclature and Abbreviations vi
Background 2
Biofuel Types for Boilers 4
Review of the Literature 5
Health and Safety Issues 7
Emissions Test Study 9
Approach and Equipment 9
Results 13
Continuous Emission Monitor Measurements 14
Particulate Matter 18
Aldehydes 19
Volatile and Semivolatile Organic Compounds 23
Poly cyclic Aromatic Hydrocarbons 30
Poly chlorinated Biphenyls 34
Boiler Efficiency 39
Life Cycle Analysis 39
QA Discrepancies 42
Conclusions 44
References 45
in
-------
List of Figures
Figure 1. Projected bioenergy use by sector through 2030 3
Figure 2. Photograph of the burner end of the firetube package boiler used in the
biofuel tests 10
Figure 3. Exhaust gas sampling locations 11
Figure 4. Concentrations of Q^ and CC>2 for the three tested fuels in ppm 15
Figure 5. Concentrations ofNOx and 862 for the three fuels tested 16
Figure 6. Concentrations of CO and THC for the three fuels tested 17
Figure 7. PM concentrations from the three fuels tested 20
Figure 8. Particle size distributions for the three fuels tested 21
Figure 9. Concentrations of aldehydes from the three fuels tested 23
Figure 10. Concentrations of detected VOCs and SVOCs, in ppb 26
Figure 11. Concentrations of PAHs from the three fuels, in ppm 31
Figure 12. Concentrations of PCBs for each of the three fuels tested 35
IV
-------
List of Tables
Table 1. Ultimate analyses and thermochemical properties of the three fuels
tested 12
Table 2. Trace element content of the three fuels tested in ppm 13
TableS. Matrix of planned tests 13
Table 4. Average flue gas concentrations of combustion gases for the tested
fuels as measured by CEM (dry conditions) 14
Table 5. Gaseous pollutant emission rates, Ib/hr 14
Table 6. Gaseous pollutant emission factors, lb/1012 Btu 18
Table 7. Measured particulate matter concentrations, mass emission rates,
and emission factors 18
Table 8. Measured concentrations of aldehydes from the three fuels, in
ppm 21
Table 9. Emission rates of aldehydes from the three fuels, Ib/hr 22
Table 10. Emission factors for aldehydes from the three fuels, lb/106 Btu 22
Table 11. Average concentrations of volatile organic compounds in ppb 24
Table 12. Emission rates of volatile and semivolatile organic compounds, in
Ib/hr 27
Table 13. Emission factors for volatile and semivolatile organic compounds,
in lb/1012 Btu 28
Table 14. Compounds not detected in any sample 29
Table 15. Concentrations of PAHs in ng/Nm3 32
Table 16. PAH emission rates in g/hr 33
Table 17. Emission factors for PAHs in lb/1012 Btu 34
Table 18. Concentrations of PCB for the three fuels tested, in pg/Nm3 36
Table 19. Emission rates of PCBs for the three fuels tested, in Ib/hr 37
Table 20. Emission factors for PCBs from the three fuels, lb/1012 Btu 38
Table 21. Results of boiler efficiency measurements 39
Table 22. Fossil fuel input required per unit of thermal energy output for the
boiler tested 41
-------
Nomenclature and Abbreviations
APCS air pollution control system
APS aerodynamic particle sizer
ASTM American Society for Testing and Materials
Bw Biodiesel consisting of n% biodiesel and (100-«)% petroleum diesel
Btu British Thermal Unit
CEM continuous emission monitor
CO carbon monoxide
CC>2 carbon dioxide
DOE U.S. Department of Energy
dscm dry standard cubic meter
EIA DOE's Energy Information Administration
EPA U.S. Environmental Protection Agency
FOR flue gas recirculation
FOG fats, oils, and greases
GHG greenhouse gas
NOX oxides of nitrogen
NRMRL National Risk Management Research Laboratory
O2 oxygen
OAQPS EPA' s Office of Air Quality Planning and Standards, part of OAR
OAR EPA' s Office of Air and Radiation
PAH polycyclic aromatic hydrocarbon
PM paniculate matter
PM2 5 particulate matter smaller than 2.5 jim aerodynamic diameter
PMio particulate matter smaller than 10 |im aerodynamic diameter
ppb parts per billion (109)
ppm parts per million (106)
QA quality assurance
QAPP quality assurance project plan
SMPS scanning mobility particle sizer
SO2 sulfur dioxide
THC total hydrocarbons
TSE transmissible spongiform encephalopathy
UGa University of Georgia
VOC volatile organic compound
VI
-------
Characterizing Emissions from the Combustion of Biofuels
Background
The term "biofuels" typically refers to liquid fuels derived from biomass, although it is
also used to describe solid or gaseous biomass-derived fuels. "Biomass" itself covers a
broad range of biological feedstocks. The majority of these feedstocks tend to be waste
materials from wood products, pulp processes, and agricultural production, or purpose-
grown crops such as corn or soy. Biomass can also include food and municipal solid
waste, solids from sewage treatment, and landfill gas, and liquid fuels from these sources
are also considered to be biofuels. Biofuels are also a subset of the broader terms
"renewable fuels" and "alternative fuels."
Biofuels have been of increasing interest as concern over reducing carbon dioxide
emissions has grown due to CCVs role in global climate change as well as rising concern
over the stability of petroleum resource availability and price. This interest was
emphasized when President Bush, in his 2007 State of the Union Address, called for the
U.S. to reach the goal of producing 35 billion gallons of renewable and alternative fuels
per year by 2017.
The U.S. Department of Energy (DOE) has projected that transportation biofuels, which
are expected to make up the majority of liquid biofuel use, will increase by 5.5%
annually through 2030, with other sectors (except biomass for electricity generation)
increasing at a slower rate or remaining relatively constant, as shown in Figure 1
(U.S. Department of Energy 2007c). The same report projects biomass for electricity
generation to increase by 4.8% annually over the same period. Between 1990 and 1995,
transportation has been the only sector in which biomass consumption increased, and the
rate of increase has been substantial at nearly 20% per year (U.S. Department of Energy
2006). Because transportation use of biomass is entirely via biofuel use, it is a valid
conclusion to note that over the five year period from 2000 to 2005, nearly all of the
increase in biomass-based energy consumption has been an increase in transportation
biofuels. Consumption of ethanol blended into gasoline increased from 139 trillion Btu
in 2000 to 340 trillion Btu in 2005.
The bulk of biofuel production and use has been for ethanol and biodiesel. Ethanol is
currently used in oxygenated motor gasoline, and increasingly is seen as significant
replacement for gasoline in "flex-fuel" vehicles. Nearly all commercial ethanol is made
from corn, although efforts are being made to enable ethanol to be produced from
cellulosic materials such as wood and agricultural wastes and fast-growing crops such as
switchgrass. Biodiesel is generally produced from soy or palm oil, and is usually used in
blends of biodiesel and conventional petroleum -based diesel fuel with little or no
modification of the engine. However, there are processes that produce biodiesel from
waste oils and fats such as used cooking oil, wastes from meat and poultry production,
and other fats, oil, and grease (FOG) that are considered to be waste materials and are
often discarded into sewage systems. In a few instances, engines have been modified for
-------
2.5
2.0 -
o
c DO
CO o
si
fa
LJJ
1.5 -
1.0 -
0.5 -
0.0
Transportation Biofuels
Electric Power from Biomass
\/
2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030
Year
Figure 1. Projected bioenergy use by sector through 2030.
direct use of filtered but unprocessed vegetable oils, but this is not currently seen as a
likely future approach to biofuel use.
As biofuel use has increased, the information on biofuel feedstocks and production
methods has expanded, in terms of both level of understanding as well as extent of
dissemination. This in turn has expanded interest in applying biofuels to applications
beyond transportation to include biofuel combustion in boilers. As the technical
understanding increases, more firms are venturing into biofuel use, including as boiler
fuels.
The incentives to use biofuels in boilers are increasing, due to state-level requirements
being placed on electricity generators. Data collected by the Database of State Incentives
for Renewables & Efficiency (DSIRE) at North Carolina State University indicated that
31 states had some form of standard or regulation to increase the renewable energy use
by electricity generators (North Carolina State University 2007). Although the renewable
requirements can be met using wind, solar, or hydropower, co-firing fossil fuels with
biomass or biofuels can be a technically attractive approach. These requirements
-------
generally apply to electricity generators, but the focus on renewable energy resources
also provides an incentive for increasing renewable energy use in other sectors.
The use of biofuels may have advantages in addition to reducing CO2 emissions. An
analysis by the New York State Energy Research and Development Authority noted that
New York State now has adequate feedstock production capacity within the state to
produce the full expected requirement for B2 fuel (2% biodiesel, 98% petrodiesel) for
highway use. This analysis estimated that producing this relatively small amount of fuel
within New York, rather than relying entirely on out-of-state supplies, would result in
over 1100 new jobs in the state, even though the price of fuel increased due to mandated
biofuel content (New York State Energy Research and Development Authority 2004). In
addition, industries are working to be seen as environmentally responsible by voluntarily
using increasing amounts of renewable energy, and are also facing possible "green
energy" requirements or incentives to use renewable energy. These factors are providing
incentives for growing biofuel consumption in stationary sources.
Biofuel Types for Boilers
Although the largest user of liquid biofuels is, and is projected to remain, the
transportation sector, the expanding technical understanding of biofuel properties,
production methods, and feedstocks, combined with the increased availability of
(currently) low-cost glycerin from biodiesel production, has illuminated numerous
opportunities for biofuel use in other applications, primarily as a fuel for small boilers.
The DOE Energy Information Administration (EIA) estimates that, of the 5.9xl012 Btu
(0.59 Quads) of biofuel energy consumption in the U.S. in 2005, the only non-
transportation use of biofuels was in the industrial sector, and this consumption was in
the form of biofuel losses and coproducts. This compares to the 2.13 Quads of wood
energy and 0.58 Quads of waste biomass energy (which includes municipal solid waste,
landfill gases, agriculture byproducts/crops, sludge waste, tires, and other biomass solids,
liquids and gases) consumed in the U.S. during the same year
(U.S. Department of Energy 2007c).
Even though the total biofuel consumption outside the transportation sector remains very
small on a national scale, the emphasis on using renewable fuels has made the use of
biofuels a relatively quick approach to increasing the fraction of renewable energy for
owners and operators of boilers. Particularly where no significant modifications to fuel
handling or combustion systems are needed, biofuels can be a relatively simple means to
achieving a corporate or state goal for renewable energy production. Biofuels for boiler
applications can include conventional biodiesel; waste fats, oils, and greases (FOGs) that
have been treated but not put through the transesterification process to produce biodiesel;
and glycerol generated from biodiesel production.1
Biodiesel use in boilers is very straightforward, as biodiesel is produced to have flow and
combustion properties very similar to those of petroleum diesel, and both of which are
1 It should be noted that most facilities do not consider the full life-cycle emissions of pollutants and CO2
when substituting biofuels for fossil fuels.
-------
generally similar to No. 2 distillate oil. Biodiesel is produced from oils and fats and its
properties must meet American Society for Testing and Materials (ASTM) standard D
6751 before it can be accepted as an actual biodiesel. These properties are designed to
ensure that any biofuel labeled for sale as biodiesel will perform properly in diesel
engines without causing damage to the engine or the fuel system
(American Society for Testing and Materials 2007). Boilers are typically able to burn
fuels with wider ranges of fuel properties than internal combustion engines, so the ASTM
standard is not as critical for boiler fuels as it is for vehicle fuels. Even so, using
biodiesel in boilers does provide the operator with some assurance of fuel quality and
performance, and therefore the operator may consider the additional cost of biodiesel
compared to other biofuels to be a worthwhile premium.
Although it is technically possible to burn ethanol in a boiler, there is very little
information about such use. It is likely that biodiesel is a preferred option because its
properties are very similar to distillate fuel oil, so little adjustment to the boiler is
required. Both ethanol and biodiesel are significantly more expensive than distillate fuel
oil on a per unit energy basis. According to the EIA, wholesale distillate fuel oil prices
were approximately $17.30/106 Btu in late October 2007 (U.S. Department of Energy
2007b), compared to $23.00/106 Btu for ethanol (priced on the Chicago Board of Trade
spot market) (Ethanol Market 2007). These costs compare to $26.20/106 Btu for
biodiesel, as reported by DOE in July 2007 (U.S. Department of Energy 2007a).
For stationary sources other than internal combustion engines, forms of bioenergy other
than biodiesel tend to be more commonly used because they tend to have lower costs.
These lower costs are largely due to the fact that biodiesel requires considerable
processing of the feedstock to meet the ASTM standards. The ASTM standards are
largely designed to ensure that biodiesel fuels can be used in internal combustion engines.
Other forms of bioenergy commonly used in stationary source combustion processes
include woody biomass, such as waste wood, and waste FOGs. Liquid biofuels other
than FOGs usually require considerably more processing to produce, and are therefore
usually more expensive and less attractive as a biofuel option. However, biodiesel is
being produced in increasing quantities, and because biodiesel fuels have been processed
to meet ASTM specifications, they can have an advantage over other bioenergy sources,
particularly for small boiler owners and operators who do not have the resources to
evaluate fuel properties.
Review of the Literature
There have been relatively little data reported on performance of, or emissions from, the
combustion of liquid biofuels in boilers. Although considerable research has been done
to evaluate the use of solid biomass fuels in boilers, those results are outside the scope of
this study. The vast majority of research to evaluate performance and emissions from
liquid biofuel combustion has focused on the use of those fuels in internal combustion
engines.
-------
Two studies were found that measured emissions from residential and commercial space
heating equipment using biodiesel blends, one published in 2001 and the second in 2003.
Krishna tested emissions from, and performance of, two small boilers burning No. 2 fuel
oil and four blends of biodiesel: BIO, B20, B30, and a 50% biodiesel/50% kerosene blend
designated BK50 (Krishna 2001).2 The two boilers used were a residential wet-base
boiler rated at 0.6 gph fuel flow (approximately 85,000 Btu/hr) and a larger commercial
boiler rated at l.SxlO6 Btu/hr. The residential boiler was used to evaluate emissions
during transient operation from a cold boiler state through steady state conditions. The
initial water temperature at the cold boiler state was 55 °F, significantly lower than the
usual 140-180 °F minimum water temperature maintained during normal operation.
During the cold boiler start, the carbon monoxide (CO) emissions increased to
approximately 250 ppm in about 30 s for the No. 2 oil, then settled down to 50-60 ppm
after about 2 min. The four biodiesel blends followed the same temporal pattern, but the
B30 fuel peaked at about 400 ppm CO and the BK50 at about 300 ppm CO. CO for the
other two blends was about 250 ppm, similar to the No. 2 oil. At steady state, CO
emissions for the biodiesel blends were consistently lower than the No. 2 oil by about 10
ppm for oxygen (©2) concentrations between 4% and 6%. Above 6%, CO emissions
began to increase for all the fuels.
Nitrogen oxide (NOX) emissions from the residential boiler were also consistently lower
for the biodiesel blends, with the differences ranging between about 2 ppm to about 20
ppm. Interestingly, the NOX concentrations decreased with increasing O2 concentration.
For the commercial boiler (which was operated at steady state for these tests), NOX
emissions were measured at between 40 and 46 ppm for the No. 2 oil with O2
concentrations near 8%. NOX emissions from B100 were between 24 and 32 ppm, with
NOX emissions from the different blends falling between the two unblended fuels.
In the second study, Batey tested a series of residential furnaces burning B20 (20% soy,
80% low sulfur highway diesel fuel) and a No. 2 distillate heating oil alone (Batey 2003).
He tested six different furnaces with fuel flow ratings from 0.75 to 2.5 gph of oil (roughly
100,000 to 325,000 Btu/hr), and measured O2, NOX, CO, and sulfur dioxide (SO2) flue
gas concentrations and smoke number. NOX emissions using the B20 fuel were
frequently lower than those measured for the No. 2 oil. Batey reported that NOX was
about 20% lower when using B20, although the data were all in graphical format and the
20% figure could not be verified. CO emissions tended to be consistently lower when
using B20 compared to the No. 2 oil, and for some units, the CO was substantially lower
when using B20. The optimum operations (low NOX and CO) typically occurred at 4-8%
O2. Batey also reported that SO2 emissions were reduced by 83% when using B20
compared to the No 2 oil. He reported that smoke number was usually lower when using
B20 than the No. 2 oil when operating at the same burner setting.
Batey identified several research needs, including additional tests to evaluate a broader
range of blends (different biofuel content and blends with ultralow sulfur distillate fuel),
determine boiler and furnace fouling rates, evaluate cold-flow characteristics of the
2 Biodiesel-gasoline blends are designated by the letter "B" followed by the percent biodiesel content. A
blend of 20% biodiesel and 80% gasoline is thus designated "B20."
-------
biodiesel and biodiesel blends, and conduct tests of furnaces using biodiesel mixes in
real-world settings.
Adams and colleagues at the University of Georgia (UGa) Engineering Outreach Service
conducted a series of tests to evaluate the combustion performance of biofuels in an
industrial boiler rated at 100,000 Ib/hr of 250 psig saturated steam (roughly 100
mmBtu/hr thermal energy output). The steam-atomizing burner nozzle was a late 1950s
design without modification for low NOX operation, but the boiler was modified for flue
gas recirculation (FGR) to reduce NOX emissions. Neither the boiler nor the nozzle were
modified for the biofuel tests. The UGa tests consisted of 173 separate test runs, burning
natural gas, No. 2 fuel oil, choice white grease, tallow, yellow grease, chicken fat, and
four blends, each with 67% No 2 fuel oil and 33% of the fats or greases.
Emissions measured during the tests of the unblended fuels showed PM to be lower for
the yellow grease and tallow, but higher for the choice white grease and chicken fat
compared to No. 2 fuel oil, with the majority of the difference being in the filterable
fraction of the PM. NOX emissions followed the same pattern, with yellow grease and
tallow NOX emissions being lower than from No. 2 fuel oil, and NOX from choice white
grease and chicken fat being higher. FGR reduced NOX emissions for all fuels tested. In
general, NOX emission concentrations for the fuels ranged from 75-120 ppm, with most
of the measurements falling in the 90-100 ppm range. SC>2 emissions were below 5 ppm
for all the fats and greases, increasing to 87 ppm for the No. 2 fuel oil and 127 ppm for
the No. 2 fuel oil with FGR. SC>2 emissions for the blends ranged from 20 ppm to 87
ppm without FGR, and up to 109 ppm with FGR for the No 2 fuel oil/white grease blend.
CO emissions were lower for all the fats and greases compared to the No. 2 fuel oil, with
the maximum being 14 ppm for the white grease. CO emissions increased slightly when
operating with FGR, to a maximum of 39 ppm for the No. 2 fuel oil/white grease blend.
More recently, Duke Energy conducted a series of tests to evaluate the performance of
biodiesel fuel in a large (90 MWe) combustion turbine. The tests were conducted at the
Duke Energy Mill Creek Station in mid-2007, in collaboration with the Electric Power
Research Institute (EPRI), and complete emissions and performance data have not yet
been published.
These studies have shown the potential for biofuels to be used in ways that can be
beneficial from the perspective of direct stack emissions of pollutants. However, it
should be noted that the range of fuels and combustion equipment tested is very limited,
and the studies were conducted under well-controlled conditions. Further work is needed
to fully understand the potential impacts of biofuel combustion in stationary sources, both
in research studies and in longer-term evaluations. The current study is designed to
evaluate emissions from biofuel combustion in a different type of boiler.
Health and Safety Issues
Safety issues for biodiesel are generally the same as for petroleum-based diesel fuels,
even though the health and safety indicators for biodiesel are consistently less hazardous
-------
than for petroleum diesel fuels (Tyson 2001). Even so, biodiesel is a fuel and needs to be
handled with care.
In general, there have been no reported health issues associated with vegetable-based
biodiesel fuels. Oils derived from soybeans or other crops are typically used in foods,
and although the oils may not be food-grade, they are the same feedstocks that are used
for food-grade products. However, there are health concerns associated with animal-
based biofuels. Waste animal products can contain microbial, other organic, or inorganic
contaminants that may pose health risks to the general public if they are present in the
fuel. Greene et al. reviewed the literature associated with potential contaminants in
animal-based biofuel feedstocks to evaluate the potential for human health risks
associated with each contaminant. They concluded that there was "little or no known risk
to human and animal health and to the environment relative to inherent microbial, organic
or inorganic agents in animal fats destined for biodiesel production"(Greene, Dawson et
al. 2007). Although they noted that it is impossible to establish a zero risk assessment for
any fat used in biodiesel production, the authors did find that the literature indicated that
the production processes and the ultimate biodiesel combustion significantly reduced any
risk associated with microbial contaminants. They included evaluations of studies of
bacteria, viruses, fungi, yeast, parasites, and microbial toxins in animal material and the
potential of these microbial contaminants to survive through the production and use
processes. Greene et al. also concluded that metals and metalloids would not lead to
significant safety issues because of the low levels of these elements in animal fats
(Greene, Dawson et al. 2007). It should be noted, however, that long term use of animal-
based biofuels, particularly in one location, could result in measurable increases in any
metals consistently present in waste animal fats, and in instances where such feedstocks
are being contemplated for long term use, the trace element content of the feedstocks
should be carefully monitored.
One issue that may be of particular health concern associated with using waste animal fat
as the feedstock for biodiesel is the possible contamination with proteinaceous infectious
particles, or prions. Prions are the cause of transmissible spongiform encephalopathy
(TSE), forms of which include bovine spongiform encephalopathy (more familiarly
known as mad cow disease) and Creutzfeldt-Jakob disease in humans, and are only
present (although rarely) in brain and nerve tissue of certain animals. The question of
whether prions can contaminate biodiesel has been studied to determine whether this is in
fact a problem. Seidel et al. concluded that each biodiesel processing step resulted in a
significant reduction in prion viability, leading to the end result that biodiesel, "even
from material with a high concentration of pathogenic prions, can be considered as
safe"(Seidel, Aim et al. 2005). A report prepared for the International Energy Agency's
(lEA's) Committee on Advanced Motor Fuels echoed the Seidel et al. report by
concluding that, "Biodiesel produced from animals infected with TSE poses a negligible
risk to animal and public health" (Baribeau, Bradley et al. 2006). Recommendations of
the Baribeau et al. report included approaches to minimize the potential for prion-
contaminated material to enter the feedstock stream and research needs associated with
reducing the uncertainties in their conclusions.
-------
In summary, the literature does not identify any significant health risks associated with
the production and use of animal-based biodiesel fuels.
Emissions Test Study
Because of the increased interest in and use of biofuels in stationary combustion systems,
there has been a similar increase in questions regarding emissions in a regulatory context.
Equipment owners and operators and regulatory agencies have expressed interest in
information that would provide guidance regarding how emissions would change if
petroleum fuels were replaced with biofuels in an effort to reduce net greenhouse gas
emissions. This study was designed to answer some of these questions for a specific
combination of fuels and combustion equipment.
Quality Assurance
The project was conducted under NRMRL's Quality Management Plan, as a Level III
project. A Level III Quality Assurance (QA) Project Plan (QAPP) was developed and
approved prior to beginning measurements. The QAPP describes the tests to be
conducted, the measurements to be used, and the applicability of the measurements;
specifies the QA objectives for the measurements (such as completeness, accuracy, and
precision); defines the requirements for maintaining a chain of custody of samples; and
provides for the means to document discrepancies and deviations from the plan.
Approach and Equipment
The tests were conducted on a firetube package boiler located in EPA's research facilities
in Research Triangle Park, NC. The boiler is a 2.94xl06 Btu/hr (860 kW), 3-pass
wetback Scotch Marine package boiler manufactured by Superior Boiler Works, Inc. (see
Figure 2). The boiler can fire natural gas or a variety of fuel oils from distillate (No. 2)
fuel oil to residual (No. 6) fuel oil. The boiler's burner is a low pressure, air atomizing
nozzle designed to generate a fuel spray to ensure proper air-fuel mixing. The unit can be
set to fire automatically or manually at any desired rate between the minimum and the
maximum firing rates. An electric heater is used to maintain proper fuel oil temperature
and therefore, viscosity. Both fuel and atomizing air flows can be varied to achieve
proper oil atomization. The boiler has 355 ft2 (33 m2) of heating surface and generates up
to 2,400 Ib/hr (1,090 kg/hr) of saturated steam at pressures up to 15 psig (103 kPa). Fuel
flow is measured with a liquid volume totalizing meter and stoichiometric ratios are
determined by measuring stack O2 and carbon dioxide (CO2) concentrations.
The boiler is fully instrumented with continuous emission monitors (CEMs) for
measuring concentrations of CC>2, CO, NOX, C>2, and SC>2. A computerized data
acquisition system (DAS) records CEM measurements as well as exit steam and flue gas
temperatures. The flue gas from the boiler passes through a manifold to an air pollution
control system (APCS) consisting of a natural-gas-fired secondary combustion chamber,
a fabric filter, and an acid gas scrubber to ensure proper removal of pollutants emitted by
the research facility's combustion units. Particle size distributions (PSDs) are measured
using a TSI Inc. model 3080/3022A scanning mobility particle sizer (SMPS) and a TSI
-------
Figure 2. Photograph of the burner end of the firetube package boiler used in the biofuel tests.
Inc. aerodynamic particle sizer (APS). All flue gas concentration and size distribution
measurements are taken prior to the APCS.
The stack of the boiler has several ports available for collecting gas and particle samples,
as shown in Figure 3. The vertical section of the 8 in (20 cm) diameter steel stack is
sufficient in length and free of flow disturbances so that PM can be sampled at an axial
location that meets EPA Method 1A sampling requirements. Several sampling ports are
located along the horizontal section of the duct approximately 9 ft (2.7 m) above the
facility catwalk. The horizontal section of the duct (8-in steel pipe) is sufficient in length
and free of flow disturbances so that particulate matter can be sampled at an axial
location that meets EPA Method 1A parti culate matter sampling requirements.
For the current tests, the boiler was operated at the lowest stack O2 concentration that
would ensure acceptable NOX and CO concentrations over a full day of operation.
"Acceptable" in this case was not determined a priori, but was determined by reducing O
10
-------
Figure 3. Exhaust gas sampling locations.
until CO began to increase and then increasing the 62 level by about 0.5%. It was
anticipated that the stack O2 level would be no lower than 3% for the tests. Measured
stack O2 levels will be discussed in more detail below. The boiler can use flue gas
recirculation to reduce NOX emissions, but the recirculation was set to zero for these tests.
The boiler load was set to maintain a steady flame and boiler water level. The boiler is
designed to automatically turn the burner off when steam pressure increases above a set
point (12 psig in this case), and then turn the burner on again as the steam pressure
decreases below a minimum (9 psig). This cycling typically occurs for lower load
conditions, but can also occur when the steam load varies. In situations where the boiler
load and pressure are too high, the steam flow can exceed the capacity of the condensate
sump tank, resulting in a loss of boiler water and a drop in the boiler water level. This
requires additional water to be injected into the boiler at a lower temperature than the
recirculating condensate, which results in transient boiler conditions. After a series of
initial scoping runs to determine the range between these two situations (burner cycling
and condensate makeup), the target boiler load for the tests was set at approximately
2xl06 Btu/hr, roughly 2/3 of the maximum boiler capacity.
Prior to starting the tests, the boiler tubes were physically cleaned to remove residues that
had accumulated during a previous series of tests to evaluate the potential formation of
polychlorinated dibenzo-d-dioxins (PCDDs). The previous study involved co-firing No.
2 fuel oil and a mixture of 1,2 dichlorobenzene and copper naphthenate. The boiler was
11
-------
operated at conditions that were designed to form significant levels of soot and create a
high-soot, high-chlorine residue along the boiler tubes to simulate long-term operation as
a waste incineration unit. Residues were removed from each end of the boiler, where the
combustion gases exit one series of boiler tubes, change direction, and enter the
subsequent series of tubes. Residues were also removed from the boiler tubes. No
analyses were conducted on the removed residues before they were disposed of as
hazardous waste.
For the current tests, three fuels were chosen for comparison - a non-road diesel fuel
(similar to a commercial No. 2 petroleum fuel oil, hereafter referred to as No. 2 fuel oil),
a biodiesel produced from soy oil, and a biodiesel produced from animal fats. The fuels
were delivered in drums to EPA's test facilities, with several drums each of the three
fuels. To account for any possible variability in composition, a sample was collected
from each drum and combined with other samples of the same fuel from the other drums
prior to analysis. Table 1 provides the ultimate analyses and physical properties for the
three fuels. Note that the oxygen content, the flashpoint, and the kinematic viscosity of
the biofuels is considerably higher, and the fuel higher heating value is lower, than that of
the No. 2 fuel oil.
Table
1. Ultimate analyses and thermochemical properties of the three fuels tested.
Karl Fischer Water
Carbon
Hydrogen
Nitrogen
Sulfur
Chlorine
Ash
Oxygen (by difference)
Fuel Higher Heating Value
(Btu/lb)
Flashpoint-Pensky Martens
Pour Point
Viscosity-Kinematic at 37.8 °C
Viscosity-Kinematic at 40 °C
Specific Gravity
No. 2 Fuel Oil
<0.03 %
84.15%
12.67 %
<0.5 %
<0.05 %
<20 ppm
0.25 %
2.93 %
19,318
64 °C
-16°C
2.845 cSt
2.726 cSt
0.8493
Soy Biodiesel
0.087 %
76.30 %
11.95%
<0.5 %
<0.05 %
<20 ppm
0.12%
11.63%
16,972
> 160 °C
4°C
4.475 cSt
4.278 cSt
0.8776
Animal Biodiesel
0.076 %
74.82 %
11.76%
<0.5 %
<0.05 %
<20 ppm
<0.10%
13.42%
17,084
> 160 °C
8°C
4.609 cSt
4.399 cSt
0.8745
The content of trace elements was also determined for the three fuels, as shown in Table
2. The only trace element measured above detection level was iron in the No. 2 fuel oil.
Extensive flue gas sampling was conducted to measure CO, CC>2, NOX, 862, organic
compound, and PM emissions for each of the three fuels. Three separate test runs were
made for each fuel, with each run lasting approximately three hours. Manual samples
were also taken for each test run to determine concentrations of volatile organic
compounds (VOCs), aldehydes, polycyclic aromatic hydrocarbons (PAHs), and
polychlorinated biphenyls (PCBs). Shorter tests were planned to evaluate possible
changes associated with blends of biofuels. During the tests of the blends, only CEM
measurements were planned, with no extractive flue gas samples.
12
-------
Table 2. Trace element content of the three fuels tested in ppm.
Calcium
Potassium
Zinc
Iron
Silicon
Nickel
Magnesium
Copper
Chromium
Manganese
Cerium
Platinum
Lead
Phosphorus
No. 2 Fuel Oil
<23
<51
<9
44
<196
<9
<9
<9
<9
<9
<9
<9
<5
<9
Soy Biodiesel
<23
<52
<9
<9
<183
<9
<9
<9
<9
<9
<9
<9
<5
<9
Animal Biodiesel
<24
<54
<9
<9
<178
<9
<9
<9
<9
<9
<9
<9
<5
<9
Table 3 shows a summary of the tests planned for each of the fuels and blends.
Table 3. Matrix of planned tests.
Test
Number
1
2
3
4
5
6
7
Fuel
#2 Fuel Oil
Vegetable Biodiesel
Animal Biodiesel
B20C
B40
B60
B80
Constituent Composition
(volume%)
#2 Fuel
Oil
100
0
0
80
60
40
20
Vegetable
-based
Biodiesel
0
100
0
20
40
60
80
Animal-
based
Biodiesel
0
0
100
0
0
0
0
Test Duration
(hours at
steady state)
10
10
10
4
4
4
4
GEM'1'
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Manual
Methods'21
Yes
Yes
Yes
No
No
No
No
Notes:
1. CEM measurement for O2, CO2, CO, NOx, SO2, and total hydrocarbons (THC)
2. Manual sampling method for particulate matter, particle size distribution, VOCs, aldehydes, PAHs, PCBs
Results
The tests of the three unblended biofuels were completed without significant problems,
although one sampling train was damaged during the soy biodiesel test and an additional
test run was required to obtain the three semivolatile samples needed (see QA
Discrepancies below). Prior to the scoping runs, the plan was to operate the unit at rated
load (2.9xl06 Btu/hr); however, the boiler was only able to maintain 1.89xl06 to
1.94xl06 Btu/hr fuel input. It is unclear why the fuel feed rate could not be increased, but
the unit's operation was stable at the lower load and there were no indications of unusual
emissions based on the CEM data, so the tests proceeded at the reduced load. Steam
temperatures were steady at between 225 and 229 °F. Three tests were conducted for
13
-------
each of the three unblended fuels, and the required number of samples was collected for
subsequent analysis.
The tests of the blends were not completed because of residues that built up in the fuel
system, apparently from the fuel. Additional information on this problem is provided in
the QA Discrepancies section below.
Continuous Emission Monitor Measurements
Average flue gas concentrations from combustion of the three fuels tested are shown in
Table 4. The average values and standard deviations were calculated using all CEM data
collected during steady state operation at each test condition. Figure 4, Figure 5, and
Figure 6 show the average concentrations of Q^ and CC>2, CO and HC, and NOX and SC>2,
respectively. The C>2 levels for the two biodiesels were slightly lower than for the No. 2
fuel oil, even though the CO levels changed very slightly. The boiler was able to be
operated at very low O2 levels with the biodiesels with no indication of CO increases or
visible sooting. As expected, the CO2 concentrations were approximately the same for all
three fuels. NOX concentrations were nearly identical for the three fuels, at 110 ppm
(corrected to 3% O2). Concentrations of SO2 were found to be quite low, with the No. 2
fuel oil being the highest at nearly 17 ppm and the two biodiesels each being below 5
ppm. Hydrocarbon concentrations were at or below the instrument detection level,
resulting in negative readings for the concentrations.
Table 4. Average flue gas concentrations of combustion gases for the tested fuels as measured by CEM
(dry conditions). All values except 62 are corrected to 3% 62. Values in parentheses are standard
deviations for the CEM data.
No. 2 (Distillate) Fuel
Oil
Soy Biodiesel
Animal Biodiesel
02
(%)
1.9
(0.13)
1.4
(0.40)
1.7
(0.21)
CO2
(%)
13.4
(0.10)
13.6
(0.14)
13.5
(0.35)
CO
(ppm)
2.2
(0.14)
2.7
(1.0)
3.1
(0.35)
NOX
(ppm)
110.
(2.4)
110.
(4.0)
110.
(1.8)
SO2
(ppm)
16.9
(1.6)
4.4
(0.48)
3.1
(0.37)
THC
(ppm)
0.70
(2.1)
-0.40
(7.5)
-2.9
(26.)
With the exception of O2, all the results are corrected to 3% O2. Emission rates are
presented in Table 5 and emission factors are given in Table 6.
Table 5. Gaseous pollutant emission rates, Ib/hr.
No. 2 Distillate
(Standard
Deviation)
Soy Biodiesel
(Standard
Deviation)
Animal Biodiesel
(Standard
Deviation)
02
31.9
(2.24)
28.3
(3.98)
34.2
(4.17)
CO2
317
(1.63)
268
(2.76)
271
(6.93)
CO
3.28E-03
(2.48E-04)
5.39E-03
(1 .99E-03)
6.11E-03
(6.96E-04)
NOX
1.88E-01
(4.13E-03)
2.10E-01
(7.87E-03)
2.21 E-01
(3.53E-03)
NO
1.82E-01
(3.75E-03)
2.08E-01
(1 .85E-02)
2.19E-01
(3.89E-03)
SO2
5.82E-02
(2.68E-03)
8.73E-03
(9.48E-04)
6.13E-03
(7.35E-04)
HC
6.03E-04
(3.68E-03)
-7.94E-04
(1 .47E-02)
-5.77E-03
(5.21 E-02)
14
-------
<1.J
2.0 -
s£
cf
0 1.5-
"00
§
o 1.0 -
O
(N
0
0.5 -
n n -
I
1
c
g &
CO CO
o
O
-a
CD
™ n
o 8
o ^
16
14 -
12 -
10 -
8 -
6 -
4 -
2 -
No. 2 Fuel Oil Soy Biodiesel Animal Biodiesel
Figure 4. Concentrations of 62 and CC>2 for the three tested fuels in ppm.
15
-------
140
E
Q. IN,
°: °
•1 c§
03 O
i— -i—'
^ T3
8 §
§ 2
o
120 -
100 -
60 -
40 -
-1
20 -
0
25
20 -
§1
§ 8
o t
10 -
5 -
No. 2 Fuel Oil Soy Biodiesel Animal Biodiesel
Figure 5. Concentrations of NOX and SO2 for the three fuels tested.
16
-------
CO O
i— -i— *
"c T3
O
OS
O
E
Q.
Q.
00
O T3
O 3
r- O
O 2
s- O
CO O
O -—•
2
T3
4 -
3 -
2 -
1 -
0
30
25 -
20 -
o 15 -
10 -
5 -
_L
•f-
No. 2 Fuel Oil Soy Biodiesel Animal Biodiesel
Figure 6. Concentrations of CO and THC for the three fuels tested.
17
-------
Table 6. Gaseous pollutant emission factors, lb/10 Btu.
No. 2 Distillate
(Standard
Deviation)
Soy Biodiesel
(Standard
Deviation)
Animal Biodiesel
(Standard
Deviation)
02
16.9
(1.18)
15.0
(2.11)
17.6
(2.15)
C02
168
(0.86)
142
(1.46)
139
(3.56)
CO
1.73E-03
(1.31E-04)
2.85E-03
(1 .05E-03)
3.14E-03
(3.58E-04)
NOX
9.92E-02
(2.19E-03)
1.11E-01
(4.16E-03)
1.14E-01
(1 .82E-03)
NO
9.65E-02
(1 .98E-03)
1.10E-01
(9.77E-03)
1.13E-01
(2.00E-03)
S02
3.07E-02
(1 .42E-03)
4.62E-03
(5.01 E-04)
3.15E-03
(3.78E-04)
HC
3.19E-04
(1 .94E-03)
-4.20E-04
(7.80E-03)
-2.97E-03
(2.68E-02)
Particulate Matter
Particulate matter emissions were measured using Method 202, which provides
information on filterable and condensable PM, with the condensable PM further
distinguished between organic and inorganic fractions (U.S. Environmental
Protection Agency 2005). Results of the tests are shown in Figure 7.
Total PM was significantly higher for the No. 2 fuel oil than for the two biodiesel fuels,
with the largest difference in the results seen in the condensable inorganic fraction as
seen in Table 7. This is not surprising, given that most of the condensable inorganic PM
is likely to be composed of sulfur compounds. Although the fuel sulfur contents were all
below the measurement detection level of 0.05% (500 ppm), the emissions of SC>2 were
roughly an order of magnitude higher for the No. 2 fuel oil than for either of the two
biodiesels. This indicates that the sulfur content of the No. 2 fuel oil was also
significantly higher, and the most likely source of the condensable inorganic PM. For the
No. 2 fuel oil, the condensable inorganic PM was found to be 8-16 mg/dscm over the
three test runs, while for the two biodiesels, the condensable inorganic PM was between
0.3 and 2.0 mg/dscm.
Condensible organic PM was the lowest component of the total PM. For the No. 2 fuel
oil, the condensable organic PM was between 0.12 and 0.46 mg/dscm. For the two
Table 7. Measured particulate matter concentrations, mass emission rates, and emission factors.
No. 2 Fuel Oil
Soy Biodiesel
Animal Biodiesel
No. 2 Fuel Oil
Soy Biodiesel
Animal Biodiesel
No. 2 Fuel Oil
Soy Biodiesel
Animal Biodiesel
Mass Concentration, mg/dscm
Filterable PM
5.43(0.43)
2.40(1.61)
2.52(1.91)
Condensable Organic
PM
0.27(0.17)
0.11 (0.00)
0.15(0.07)
Condensable
Inorganic PM
12.3(3.82)
1.08(0.85)
1.37(0.19)
Total PM
18.0(3.88)
3.48 (2.27)
3.89(1.73)
Emission Rate, Ib/hr
6.90E-3(8.18E-4)
3.20E-3(2.13E-3)
3.52E-3 (2.66E-3)
3.36E-4 (2.03E-4)
1.52E-04(2.52E-6)
2.03E-4 (8.89E-5)
1.56E-2(4.65E-3)
1.45E-3(1.13E-3)
1.93E-3(3.00E-4)
2.28E-2 (4.64E-3)
4.82E-3 (3.01 E-3)
5.65E-3(2.41E-3)
Emission Factors, lb/106 Btu
3.65E-3
1.70E-3
1.81E-3
1.78E-4
8.02E-5
1.05E-4
8.23E-3
7.69E-4
9.91E-4
1.21E-2
2.55E-3
2.91 E-3
18
-------
biodiesels, the condensable organic PM was below the method detection limit of 0.11-
0.22 mg/dscm for all runs.
The filterable PM results were slightly higher for the No. 2 fuel oil than for the two
biodiesels. For the No. 2 fuel oil, the filterable PM data were consistent across all three
runs, with an average of 5.4 mg/dscm. The two biodiesels had filterable PM similar to
one another, with each biodiesel having one run at a significantly lower level than the
other two. Even if the low runs are ignored, the biodiesel filterable PM results were still
lower than that for the No. 2 fuel oil.
The PSDs measured during the tests indicate that the majority of particle mass as
indicated by volume are larger than 1 jim in aerodynamic diameter. Figure 8 shows the
PSDs measured for the three fuels tested, and shows that the majority of particle volume
is in the range of 0.5-10 jim. These measurements are of electrical mobility diameter and
not strictly aerodynamic diameter, although the two are related. The mass measurements
using Method 202 and the mass values estimated from the PSDs are consistent in terms
of magnitude for all three fuels, as shown in Figure 7.
The diamond symbols show the estimated mass based on the integrated volume measured
by the APS, assuming a 1 g/cc mass density for the particles. The integrated volume and
the Method 202 results show excellent agreement. There is some discrepancy, however,
between the size distribution results and the Method 202 fractions, particularly for the
No. 2 fuel oil results. The size distributions (see Figure 8) show that nearly all the PM
volume (and therefore mass) is larger than about 0.5 jim in aerodynamic diameter for all
three fuels. For the soy and animal biodiesel fuels, the integrated mass is about the same
as the filterable mass as measured by Method 202. This makes physical sense, as the
filterable particles are likely to be in these larger size ranges. On the other hand, the
condensable mass, whether organic or inorganic, is more likely to be much smaller - near
0.1 jim or less in aerodynamic diameter. The APS data for the No. 2 fuel oil show the
majority of the volume (mass) as being well above what would be expected for
condensable particles. It is unclear why the APS and Method 202 data match as well as
they do for the No. 2 fuel oil.
Aldehydes
Aldehyde concentrations were measured for each of the test conditions, with three
samples collected from the exhaust from each of the three fuels. The highest
concentration levels were for formaldehyde, acetaldehyde, and acetone, each of which
had concentrations of over 1 ppm for one of the three fuels. Formaldehyde and
acetaldehyde concentrations were measured at 3.4 and 2.5 ppm, respectively, from the
soy biodiesel, and acetone was measured at 1.8 ppm from the No. 2 fuel oil. Figure 9
shows the average measured concentrations of the 12 aldehydes. Given the relatively
high variability in the measurements, the significance of the differences in emissions for
the three fuels is questionable. Even though the formaldehyde and acetaldehyde from the
soy biodiesel appear to be significantly higher than the respective concentrations
19
-------
o
g
"CD
25
20 -
15 -
8 10
c
o
O
5 -
Filterable
Condensable Organic
Condensable Inorganic
Estimated from PSD data
VWfeWS
No. 2 Fuel Oil
Soy Biodiesel
Animal Biodiesel
Figure 7. PM concentrations from the three fuels tested.
measured for the other two fuels, the relative standard deviation for these two
measurements were over 100%, making it impossible to state definitively that the
increases would be consistent in the absence of a larger number of measurements.
Table 8 presents the aldehyde concentrations, emission rates are given in Table 9, and
Table 10 gives the aldehyde emission factors for the three fuels. Those values shown as
being less than the stated value are based on using zero for those samples in which the
compound was not detected at a level of about 0.6 ppm.
20
-------
O)
30000
25000 -
20000 -
15000 -
10000 -
5000 -
O No. 2 fuel oil SMPS
n Soy SMPS
A Animal SMPS
• No. 2 fuel oil APS
• Soy APS
A Animal APS
\
0.01
0.1
Dp, urn
Figure 8. Particle size distributions for the three fuels tested.
10
Table 8. Measured concentrations of aldehydes from the three fuels, in ppm.
Formaldehyde
Acetaldehyde
Acetone
Propionaldehyde
Crotonaldehyde
Butylaldehyde
Benzaldehyde
Iso-Veraldehyde
Veraldehyde
Tolualdehyde
Hexanal
2,5-Dimethy-
benzaldehyde
Mass Concentration, ppm
(standard deviation in parentheses)
No. 2 Fuel Oil
8.34x10^(1. 57x1 (T1)
3. 02x1 0^(4. 04x1 0"J)
1. 77x1 Ou (2.87x1 Ou)
< 2.87x1 0"J (4.98x1 0"J)
< 4.31x1 0-03 (3.77x10-
03)
< 2.02x1 0"J (3.49x1 0"J)
< 2.48 x10'J (4.30x1 0"J)
< 1.10x1 0^(1. 13x10^)
< 2.39x1 0"J (4.1 3x1 0"J)
<7.33x10'J (1.27x10^)
<2.31x10'J(4.00x10'J)
<1.70x10'J(2.94x10'J)
Soy Biodiesel
3.40x10" (5.80x1 Ou)
2.53x1 Ou (4.35x1 Ou)
3.97x10"' (1.62x10"')
NDia)
ND
ND
ND
< 1.05x10^ (9.95x1 0"J)
ND
ND
ND
ND
Animal Biodiesel
3.26x1 0"n (2.68x1 0"2)
3.29x10^(2.24x10^)
2.47x10"' (9.23x10^)
ND
< 1. 95x1 0"J (3.37x1 0"J)
ND
ND
< 2.20x1 0"J (3.81x1 0"J)
ND
ND
ND
ND
(a) Not detected
21
-------
Table 9. Emission rates of aldehydes from the three fuels, Ib/hr.
Formaldehyde
Acetaldehyde
Acetone
Propionaldehyde
Crotonaldehyde
Butylaldehyde
Benzaldehyde
Iso-Veraldehyde
Veraldehyde
Tolualdehyde
Hexanal
2,5-Dimethy-
benzaldehyde
Emission Rates, Ib/hr
(standard deviation in parentheses)
No. 2 Fuel Oil
1. 32x1 0'4 (2.1 1x1 0'3)
7.01x1 O'3 (7.1 3x1 O'b)
5.63x1 0"J (9.1 5x1 0"J)
< 9.1 3x1 O'b(1. 58x1 (T)
<1.63x10'3(1.43x10-3)
< 7.97x1 0-6(1. 38x1 0'3)
<1.44x10'3(2.50x10-3)
<5.12x10'3(5.31x10-3)
< 1. 1 3x1 0'3(1. 95x1 0'3)
<4.83x10'3(8.37x10-3)
< 1. 27x1 0'3 (2.1 9x1 0'3)
<1.79x10'3(3.09x10-3)
Soy Biodiesel
5.83x1 0"J (9.93x1 0"J)
6.40x1 0"J( 1.1 0x10^)
1. 30x1 0"J (5.40x1 0"J)
NDla)
ND
ND
ND
<5.13x10'3(4.91x10-3)
ND
ND
ND
ND
Animal Biodiesel
5.73x1 0'4 (5.55x1 0'3)
8.44x1 0'a (5.69x1 0'3)
8.42x1 0'4 (3.26x1 0'3)
ND
< 7.90x1 0'b(1. 37x1 0'3)
ND
ND
< 1. 1 0x1 0'3(1. 90x1 0'3)
ND
ND
ND
ND
(a) Not detected
Table 10. Emission factors for aldehydes from the three fuels, lb/10 Btu.
Formaldehyde
Acetaldehyde
Acetone
Propionaldehyde
Crotonaldehyde
Butylaldehyde
Benzaldehyde
Iso-Veraldehyde
Veraldehyde
Tolualdehyde
Hexanal
2,5-Dimethy-
benzaldehyde
Emission Factors, lb/10B Btu
(standard deviation in parentheses)
No. 2 Fuel Oil
6.96x10'3(1.12x10'3)
3.70x1 0'3 (3.77x1 0'3)
2.98x1 0"J (4.84x1 0"J)
< 4.83x1 0'b (8.36x1 0'b)
< 8.60x1 0'b (7.56x1 0'b)
< 4.21x1 0'b (7.30x1 0'b)
< 7.63x1 0'b(1. 32x1 0'3)
<2.71x10'3(2.81x10'3)
< 5.96x1 0'b(1. 03x1 0'3)
< 2.56x1 0'3 (4.43x1 0'3)
< 6.70x1 0'b (1.1 6x1 0'3)
< 9.45x1 0'b(1. 64x1 0'3)
Soy Biodiesel
3.08x1 0'6 (5.25x1 0'6)
3.38x1 0"J (5.82x1 0"J)
6.88x1 0'4 (2.86x1 0'4)
NDla)
ND
ND
ND
< 2.71x1 0'3 (2.60x1 0'3)
ND
ND
ND
ND
Animal Biodiesel
2. 95x1 0'4 (2. 86x1 0'3)
4. 34x1 0'3 (2. 93x1 0'3)
4.33x1 0'4(1. 68x1 0'4)
ND
< 4.06x1 0'b (7.04x1 0'b)
ND
ND
< 5.64x1 0'b (9.77x1 0'b)
ND
ND
ND
ND
(a) Not detected
22
-------
1 -
a.
CL
Concentrations
o
0)
_c
0.01 -
n nm -
T
IT
'
•
III
^m No. 2 Fuel Oil
i i Soy Blodiesel
^H Animal Biodiesel
1 11
-------
Table 11. Average concentrations of volatile organic compounds in ppb (standard deviation shown in
parentheses).
Chloromethane
Bromomethane
Chloroethane
Ethanol
Carbon disulfide
Isopropyl alcohol
Acetone
Methyl-t-butyl ether
Vinyl acetate
Cyclohexane
Chloroform
Ethyl acetate
Tetrahydrofuran
2-Butanone
Benzene
Trichloroethylene
1,4 Dioxane
Toluene
4-Methyl-2-pentanone
2-Hexanone
Ethyl benzene
Chlorobenzene
m,p-Xylene
o-Xylene
Styrene
Tribromomethane
1,1,2,2-
Tetrachloroethane
No. 2 Fuel Oil
Average
(Standard Deviation)
0.860
(1.49)
0.267
(0.462)
0.557
(0.964)
23.6
(25.8)
21.7
(24.8)
< 2.64W
(1.94)
33.7
(6.08)
ND
ND
< 1.76
(1.09)
0.563
(0.575)
5.65
(5.07)
1.84
(0.637)
6.22
(10.8)
3.15
(2.06)
1.43
(1.24)
ND
0.920
(0.817)
1.35
(0.289)
6.36
(1.27)
< 0.873
(0.530)
1.00
(0.887)
1.56
(0.890)
< 0.517
(0.525)
< 0.500
(0.000)
0.917
(1.59)
ND
Soy Biodiesel
Average
(Standard Deviation)
ND(a)
ND
1.03
(1.78)
72.6
(86.2)
70.6
(81.9)
1.84
(2.49)
37.8
(10.0)
0.383
(0.664)
ND
4.72
(2.96)
< 1.02
(0.553)
3.97
(4.27)
1.26
(0.406)
2.50
(4.34)
3.09
(1.12)
1.61
(2.79)
ND
0.300
(0.520)
11.71
(16.04)
3.08
(2.74)
< 0.853
(0.307)
0.560
(0.970)
1.65
(0.759)
ND
0.727
(0.049)
1.90
(1.66)
ND
Animal Biodiesel
Average
(Standard Deviation)
ND
1.07
(1.85)
0.633
(1.10)
193
(324)
199
(332)
7.90
(9.03)
62.1
(45.3)
<1.00
(1.32)
2.76
(2.39)
3.46
(2.06)
1.12
(0.563)
4.68
(4.14)
5.54
(5.14)
13.1
(22.6)
19.0
(27.9)
2.81
(1.34)
< 0.863
(1.09)
3.89
(1.02)
1.67
(0.644)
ND
1.31
(0.803)
2.13
(2.82)
2.88
(2.55)
1.02
(0.779)
0.250
(0.433)
1.15
(1.18)
1.14
(1.98)
(Continued on following page)
24
-------
Table 11 (Continued)
1-Ethyl-4-methyl
benzene
1 ,3,5-Trimethylbenzene
1 ,2,4-Trimethylbenzene
1 ,3-Dichlorobenezene
1 ,4-Dichlorobenezene
1,2-Dichlorobenzene
< 0.500
(0.000)
< 0.520
(0.530)
1.12
(0.698)
1.05
(1.82)
1.13
(1.96)
293
(88.8)
< 0.167
(0.289)
< 0.167
(0.289)
0.220
(0.381)
ND
ND
546
(271)
< 0.167
(0.289)
0.517
(0.480)
0.800
(0.904)
0.743
(1.29)
1.36
(1.26)
139
(78.3)
(a) Not detected
(b) Concentrations shown as less than the stated value had one or more measurements below the
method detection level. Average values are calculated using the method detection level.
Figure 10 illustrates the concentrations of the detected VOCs and SVOCs. In general, the
animal biodiesel has higher concentrations of the organic compounds compared to the
other two fuels, although the high variability in concentrations results in differences that
are not statistically significant in many cases. The semi-log plot makes the magnitude of
the variability difficult to see, but the data in Table 12 demonstrate the high standard
deviations relative to the average. There were only six compounds (acetone,
cyclohexane, tetrahydrofuran, ethylbenzene, m,p-xylene, and 1,2-dichlorobenzene) for
which the relative standard deviations were less than one for all three fuels.
25
-------
1000
100 -
.0
Q_
Q.
c"
o
s
-I—•
-------
Table 12. Emission rates of volatile and semivolatile organic compounds, in Ib/hr.
Chloromethane
Bromomethane
Chloroethane
Ethanol
Carbon disulfide
Isopropyl alcohol
Acetone
Methyl-t-butyl ether
Vinyl acetate
Cyclohexane
Chloroform
Ethyl acetate
Tetrahydrofuran
2-Butanone
Benzene
Trichloroethylene
1,4 Dioxane
Toluene
4-Methyl-2-pentanone
2-Hexanone
Ethyl benzene
Chlorobenzene
m,p-Xylene
o-Xylene
Styrene
Tribromomethane
1,1,2,2-
Tetrachloroethane
1-Ethyl-4-methyl
benzene
No. 2 Fuel Oil
Average
(Standard Deviation)
2.20E-06
(3.82E-06)
1.39E-06
(2.40E-06)
1.82E-06
(3.16E-06)
5.59E-05
(6.04E-05)
8.45E-05
(9.54E-05)
< 8.23E-06m
(6.03E-06)
1.03E-04
(1.46E-05)
ND
ND
< 7.80E-06
(4.85E-06)
3.60E-06
(3.78E-06)
2.58E-05
(2.33E-05)
7.06E-06
(2.60E-06)
2.28E-05
(3.94E-05)
1.31E-05
(9.06E-06)
9.70E-06
(8.45E-06)
ND
4.39E-06
(3.93E-06)
7.12E-06
(1.34E-06)
3.34E-05
(5.41 E-06)
< 4.88E-06
(2.95E-06)
5.83E-06
(5.20E-06)
8.70E-06
(4.97E-06)
< 2.94E-06
(2.94E-06)
< 2.74E-06
(1.05E-07)
1.22E-05
(2.11E-05)
ND
<3.17E-06
(1.20E-07)
Soy Biodiesel
Average
(Standard Deviation)
ND(a)
ND
3.67E-06
(6.36E-06)
1.87E-04
(2.22E-04)
3.01 E-04
(3.48E-04)
6.22E-06
(8.33E-06)
1.23E-04
(3.24E-05)
1.94E-06
(3.36E-06)
ND
2.25E-05
(1.43E-05)
< 6.83E-06
(3.69E-06)
1.95E-05
(2.11E-05)
5.13E-06
(1.74E-06)
1.00E-05
(1.73E-05)
1.36E-05
(5.13E-06)
1.18E-05
(2.05E-05)
ND
1.53E-06
(2.65E-06)
6.71 E-05
(9.25E-05)
1.72E-05
(1.53E-05)
< 5. 11 E-06
(1.88E-06)
3.52E-06
(6.10E-06)
9.91 E-06
(4.75E-06)
ND
4.26E-06
(2.18E-07)
2.73E-05
(2.39E-05)
ND
< 1.11 E-06
(1.92E-06)
Animal Biodiesel
Average
(Standard Deviation)
ND
6.06E-06
(1.05E-05)
2.37E-06
(4.10E-06)
5.15E-04
(8.66E-04)
8.79E-04
(1.46E-03)
2.75E-05
(3.14E-05)
2.10E-04
(1.51 E-04)
< 5.23E-06
(6.97E-06)
1.37E-05
(1.19E-05)
1.72E-05
(1.05E-05)
7.89E-06
(4.10E-06)
2.39E-05
(2. 11 E-05)
2.36E-05
(2.22E-05)
5.46E-05
(9.45E-05)
8.80E-05
(1.30E-04)
2.17E-05
(1.07E-05)
< 4.41 E-06
(5.57E-06)
2.09E-05
(5.29E-06)
9.75E-06
(3.65E-06)
ND
8.13E-06
(4.93E-06)
1.42E-05
(1.89E-05)
1.78E-05
(1.56E-05)
6.33E-06
(4.76E-06)
1.55E-06
(2.69E-06)
1.72E-05
(1.77E-05)
1.11E-05
(1.93E-05)
< 1.16E-06
(2.01 E-06)
(Continued on following page)
27
-------
Table 12 (Continued)
1 ,3,5-Trimethylbenzene
1 ,2,4-Trimethylbenzene
1 ,3-Dichlorobenezene
1 ,4-Dichlorobenezene
1 ,2-Dichlorobenzene
< 3.25E-06
(3.36E-06)
7.15E-06
(4.44E-06)
8.47E-06
(1.47E-05)
8.77E-06
(1.52E-05)
2.29E-03
(7.76E-04)
<1.11E-06
(1.92E-06)
1.47E-06
(2.54E-06)
ND
ND
4.54E-03
(2.35E-03)
3.64E-06
(3.36E-06)
5.57E-06
(6.29E-06)
6.33E-06
(1.10E-05)
1.16E-05
(1.07E-05)
1.19E-03
(6.59E-04)
(a) Not detected
(b) Concentrations shown as less than the stated value had one or more measurements below the
method detection level. Average values are calculated using the method detection level.
Table 13. Emission factors for volatile and semivolatile organic compounds:
inlb/1012Btu.
Chloromethane
Bromomethane
Chloroethane
Ethanol
Carbon disulfide
Isopropyl alcohol
Acetone
Methyl-t-butyl ether
Vinyl acetate
Cyclohexane
Chloroform
Ethyl acetate
Tetrahydrofuran
2-Butanone
Benzene
Trichloroethylene
1,4 Dioxane
Toluene
4-Methyl-2-pentanone
No. 2 Fuel Oil
Average
(Standard Deviation)
1.16E+00
(2.02E+00)
7.33E-01
(1.27E+00)
9.64E-01
(1.67E+00)
2.95E+01
(3.19E+01)
4.47E+01
(5.05E+01)
< 4.35E+00™
(3.19E+00)
5.45E+01
(7.73E+00)
ND
ND
<4.13E+00
(2.56E+00)
1.90E+00
(2.00E+00)
1.36E+01
(1.23E+01)
3.73E+00
(1.37E+00)
1.20E+01
(2.08E+01)
6.95E+00
(4.79E+00)
5.13E+00
(4.47E+00)
ND
2.32E+00
(2.08E+00)
3.76E+00
(7.06E-01)
Soy Biodiesel
Average
(Standard Deviation)
ND(a)
ND
1.94E+00
(3.36E+00)
9.91E+01
(1.17E+02)
1.59E+02
(1.84E+02)
3.29E+00
(4.41 E+00)
6.53E+01
(1.72E+01)
1.03E+00
(1.78E+00)
ND
1.19E+01
(7.58E+00)
< 3.61 E+00
(1.95E+00)
1.03E+01
(1.11E+01)
2.71 E+00
(9.21E-01)
5.29E+00
(9.17E+00)
7.22E+00
(2.72E+00)
6.25E+00
(1.08E+01)
ND
8.10E-01
(1.40E+00)
3.55E+01
(4.89E+01)
Animal Biodiesel
Average
(Standard Deviation)
ND
3.12E+00
(5.40E+00)
1.22E+00
(2.11 E+00)
2.65E+02
(4.46E+02)
4.52E+02
(7.54E+02)
1.42E+01
(1.62E+01)
1.08E+02
(7.75E+01)
<2.69E+00
(3.59E+00)
7.07E+00
(6.14E+00)
8.83E+00
(5.39E+00)
4.06E+00
(2.11 E+00)
1.23E+01
(1.09E+01)
1.21E+01
(1.14E+01)
2.81E+01
(4.86E+01)
4.53E+01
(6.70E+01)
1.12E+01
(5.52E+00)
<2.27E+00
(2.87E+00)
1.08E+01
(2.72E+00)
5.02E+00
(1.88E+00)
(Continued on following page)
28
-------
Table 13 (continued)
2-Hexanone
Ethyl benzene
Chlorobenzene
m,p-Xylene
o-Xylene
Styrene
Tribromomethane
1,1,2,2-
Tetrachloroethane
1-Ethyl-4-methyl
benzene
1 ,3,5-Trimethylbenzene
1 ,2,4-Trimethylbenzene
1 ,3-Dichlorobenezene
1 ,4-Dichlorobenezene
1,2-Dichlorobenzene
1.77E+01
(2.86E+00)
<2.58E+00
(1.56E+00)
3.08E+00
(2.75E+00)
4.60E+00
(2.63E+00)
<1.56E+00
(1.55E+00)
<1.45E+00
(5.55E-02)
6.45E+00
(1.12E+01)
ND
<1.67E+00
(6.35E-02)
<1.72E+00
(1.78E+00)
3.78E+00
(2.35E+00)
4.48E+00
(7.75E+00)
4.64E+00
(8.03E+00)
1.21E+03
(4.10E+02)
9.09E+00
(8.11E+00)
<2.70E+00
(9.94E-01)
1.86E+00
(3.23E+00)
5.24E+00
(2.51 E+00)
ND
2.25E+00
(1.15E-01)
1.44E+01
(1.26E+01)
ND
< 5.87E-01
(1.02E+00)
< 5.87E-01
(1.02E+00)
7.75E-01
(1.34E+00)
ND
ND
2.40E+03
(1.24E+03)
ND
4.18E+00
(2.54E+00)
7.31E+00
(9.75E+00)
9.15E+00
(8.03E+00)
3.26E+00
(2.45E+00)
7.99E-01
(1.38E+00)
8.84E+00
(9.12E+00)
5.72E+00
(9.91E+00)
< 5.97E-01
(1.03E+00)
1.87E+00
(1.73E+00)
2.87E+00
(3.24E+00)
3.26E+00
(5.64E+00)
5.94E+00
(5.52E+00)
6.11E+02
(3.39E+02)
(a) Not detected
(b) Concentrations shown as less than the stated value had one or more measurements below the
method detection level. Average values are calculated using the method detection level.
Table 14 shows those compounds for which the samples were analyzed, but were not
detected in any of the samples. For these compounds, the emission rate is reported as 0
Ib/hr and the emission factor as 0 lb/1012 Btu. At the detection limit of 0.5 ppb, the
emission rate would be about 2.7xlO"6 Ib/hr and the emission factor would be about 1.5
lb/1012 Btu.
Table 14. Compounds not detected in any sample.
Dichlorodifluoromethane
1,2-Chloro-1, 1,2,2-
Tetrafluoroethane
Vinyl Chloride
1,3 Butadiene
Trichloromonofluoromethane
1,1-Dichloroethene
1,1,2-Trichloro-1,2,2-
trichloromoethane
Methylene chloride
t-1 ,2-Dichloroethene
Hexane
1,1-Dichloroethane
cis-1 ,2-Dichloroethene
1,1-trichloroethane
Carbon tetrachloride
Heptane
1,2-Dichloroethane
1,2-Dibromoethane
1 ,2-Dichloropropane
Bromodichloromethane
cis-1 ,3-Dichloropropene
t-1 ,3-Dichloropropene
Tetrachloroethylene
1,1,2-Trichloroethane
Dibromochloromethane
Benzyl chloride
1,1,2,3,4,4-Hexachloro-1,3-
butadiene
1 ,2,4-Trichlorobenzene
29
-------
Poly cyclic Aromatic Hydrocarbons
Poly cyclic aromatic hydrocarbon concentrations were collected using EPA Method 0010
(U.S. Environmental Protection Agency 1986) and analyzed according to California Air
Resources Board (CARS) Method 429 (California Air Resources Board 1997). Three
samples were collected for each fuel, with one sample collected per test run. The
analyses determined concentrations of 19 PAHs, of which two were found to be below
the detection limit for all three fuels. Figure 11 shows the concentrations of the 19 PAHs
for the three fuels. In general, the relative pattern of PAH concentration is consistent
across all three fuels, with phenanthrene being the PAH with the highest concentration
and perlyene being the PAH with the lowest concentration for all three fuels. In all three
fuels, both fluorene and dibenzo(a,h)anthracene were below detection limits (at
approximately 1 ppm) in the flue gases of the three fuels.
As can be seen in Figure 11, the PAH concentrations for the animal biodiesel are all
lower than those for the soy biodiesel. Of the 17 PAHs measured above the detection
level, 5 were higher for the No. 2 fuel oil than for the soy biodiesel and 12 were lower.
Eight of the 17 detected PAHs were higher for the animal biodiesel than for the No. 2
fuel oil.
Table 15 shows the concentrations in ng/Nm3, Table 16 gives the emission rate results in
g/hr, and Table 17 provides the emission factor results in lb/1012 Btu for the 19 PAHs.
Field blank samples were taken during each of the three test conditions. Field blank PAH
levels were an average of 19% of the sample levels for the No. 2 fuel oil and 16% of
sample levels for the soy biodiesel. The field blank levels were highest for naphthalene
for both these fuels, at 41% of sample value for the No. 2 fuel oil and 65% for the soy
biodiesel. For the animal biodiesel, the field blank levels averaged 54%, with blank
levels exceeding sample levels for ideno(l,2,3-cd)pyrene and benzo(ghi)perylene.
30
-------
10000
1000 -
O)
c
cf
--
CD
O
C
O
O
100 -
10 -
0.1
-J-
r
;[
i
0)
Naphthale
T
.1
\
'.
r
0)
lethylnapthale
^
(N
-
a
c
j
1
-i-
.£
£
0
a
<
-
r
D £
1 1
: o
: i
M
I) C
J 0
<
•
1
-N /-
i i
if "a
- c
2 2
3 C
1 ^
§
.
—
> o
r a
) :£
i S
- J
Q
r
-,
i
- o
3 ~o
: C
3 0
: c.
5 S
1 1
. <
r
7- a
> c
i 1
3 !
: C
_ —
i -
C u
r
i
) a
: c
3 a
L.
3 5
5 D
5
3
r
T
-,;
T ;"|-
1 1 1
) a) a) a
: c c c
) 0) 0) (
O W .£
- CD ^ *
• .c ^ o
•£ 0 J
CD :
S *
N ~C
C h
0) C
00 a
01
^m No. 2 Fuel Oil
K///I Soy Biodiesel
l l Animal Biodiesel
r T"
- MA '. •
0 0) 0) 0) J^> 0) p-
J 0) 0) 0) ~oT 0) ~oT
3 £ >< >< Er >> £
: c 0.0.^ Q. Q5
5 2 'o? 1? £- 6" CD
I § o' o' ® V i
2 = N N Q- n -g
? 2" 0) 0) CM CD
," "o" oo oo c- ^
g N o d;
~ ^" ^C^
JO) Q) °
3 00 T3 N
0)
izo(ghi)peryle
0)
00
0)
Figure 11. Concentrations of PAHs from the three fuels, in ng/Nm . (1) No samples were above the
detection limit; (2) one sample was below the detection limit; (3) two samples (one each for soy
and animal biodiesel) were below the detection limit.
-------
Table 15. Concentrations of PAHs in ng/Nm .
Naphthalene
2-Methylnapthalene
Acenaphthylene
Acenaphthene
Fluorene (b)
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(e)pyrene
Benzo(a)pyrene
Perylene
ldeno(1 ,2,3-cd)pyrene
Dibenzo(a,h)anthracene (b)
Benzo(ghi)perylene
No. 2 Fuel Oil
Average
(Standard deviation)
469
(109)
341
(168)
34.4
(17.8)
12.8
(6.7)
<1.55(c)
(0.04)
3706
(558)
<10(a)
(3)
393
(62.0)
25.4
(14.6)
1.06
(0.33)
6.48
(1.69)
4.93
(1.24)
0.80
(0.23)
1.73
(0.61)
0.59
(0.26)
0.16
(0.08)
1.18
(0.70)
<1.55
(0.04)
3.32
(2.67)
Soy Biodiesel
Average
(Standard deviation)
297
(54.0)
174
(64.2)
37.0
(34.7)
<8.48 (a)
(2.37)
<1.53
(0.04)
2580
(364)
23.3
(15.6)
353
(52.7)
169
(22.1)
4.74
(2.05)
11.9
(1.98)
7.37
(1.56)
1.97
(0.75)
3.99
(0.98)
1.52
(1.43)
<0.90(a)
(0.63)
1.74
(0.73)
<1.53
(0.04)
5.40
(3.53)
Animal Biodiesel
Average
(Standard deviation)
210
(45.0)
71.0
(13.6)
8.09
(2.15)
3.68
(0.19)
<1.49
(0.01)
1210
(232)
6.32
(4.38)
181
(37.0)
64.0
(23.4)
2.00
(0.65)
5.57
(1.18)
3.90
(1.01)
1.01
(0.15)
3.11
(2.49)
0.63
(0.33)
<0.29 (a)
(0.26)
1.20
(0.66)
<1.49
(0.01)
4.22
(4.54)
(a) One of three samples were below detection limit.
(b) All samples were below detection limit.
(c) Concentrations shown as less than the stated value had one or more measurements below the
method detection level. Average values are calculated using the method detection level.
32
-------
Table 16. PAH emission rates in g/hr.
Naphthalene
2-Methylnapthalene
Acenaphthylene
Acenaphthene
Fluorene (b)
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(e)pyrene
Benzo(a)pyrene
Perylene
ldeno(1,2,3-cd)pyrene
Dibenzo(a,h)anthracene (b)
Benzo(ghi)perylene
No. 2 fuel oil
2.65E-04
(5.40E-05)
1.92E-04
(8.91 E-05)
1.93E-05
(9.46E-06)
7.22E-06
(3.59E-06)
<8.77E-07 (c)
(3.54E-08)
2.10E-03
(2.63E-04)
<5.41E-06(a)
(1.76E-06)
2.22E-04
(3.01 E-05)
1.43E-05
(7.90E-06)
5.96E-07
(1.69E-07)
3.66E-06
(8.59E-07)
2.78E-06
(6.26E-07)
4.53E-07
(1.18E-07)
9.77E-07
(3.17E-07)
3.30E-07
(1.35E-07)
9.18E-08
(4.21 E-08)
6.62E-07
(3.77E-07)
<8.77E-07
(3.54E-08)
1.86E-06
(1.44E-06)
Soy Biodiesel
1.82E-04
(3.05E-05)
1.07E-04
(3.82E-05)
2.26E-05
(2.10E-05)
<5.20E-06 (a)
(1.39E-06)
O.39E-07
(9.46E-09)
1.58E-03
(2.02E-04)
1.42E-05
(9.38E-06)
2.17E-04
(2.92E-05)
1.04E-04
(1.20E-05)
2.90E-06
(1.22E-06)
7.32E-06
(1.11E-06)
4.53E-06
(8.95E-07)
1.21E-06
(4.42E-07)
2.45E-06
(5.70E-07)
9.27E-07
(8.66E-07)
<5.59E-07 (a)
(3.94E-07)
1.07E-06
(4.32E-07)
O.39E-07
(9.46E-09)
3.31 E-06
(2.14E-06)
Animal Biodiesel
1.33E-04
(2.86E-05)
4.48E-05
(8.65E-06)
5. 11 E-06
(1.37E-06)
2.32E-06
(1.23E-07)
O.43E-07
(4.56E-09)
7.62E-04
(1.47E-04)
3.99E-06
(2.77E-06)
1.14E-04
(2.35E-05)
4.04E-05
(1.48E-05)
1.26E-06
(4.14E-07)
3.52E-06
(7.48E-07)
2.46E-06
(6.44E-07)
6.40E-07
(9.81 E-08)
1.96E-06
(1.57E-06)
3.96E-07
(2.10E-07)
<1.83E-07(a)
(1.61E-07)
7.56E-07
(4.18E-07)
O.43E-07
(4.56E-09)
2.67E-06
(2.87E-06)
(a) One of three samples were below detection limit.
(b) All samples were below detection limit.
(c) Concentrations shown as less than the stated value had one or more measurements below the
method detection level. Average values are calculated using the method detection level.
33
-------
Table 17. Emission factors for PAHs in lb/10 Btu.
Naphthalene
2-Methylnapthalene
Acenaphthylene
Acenaphthene
Fluorene (b)
Phenanthrene
Anthracene
Fluoranthene
Pyrene
Benzo(a)anthracene
Chrysene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(e)pyrene
Benzo(a)pyrene
Perylene
ldeno(1,2,3-cd)pyrene
Dibenzo(a,h)anthracene (b)
Benzo(ghi)perylene
No. 2 fuel oil
Average
(Standard deviation)
3.08E-01
(6.29E-02)
2.23E-01
(1.04E-01)
2.25E-02
(1.10E-02)
8.41 E-03
(4.18E-03)
<1.02E-03(c)
(4.13E-05)
2.44E+00
(3.06E-01)
<6.30E-03 (a)
(2.05E-03)
2.59E-01
(3.50E-02)
1.66E-02
(9.20E-03)
6.94E-04
(1.97E-04)
4.26E-03
(1.00E-03)
3.24E-03
(7.29E-04)
5.28E-04
(1.37E-04)
1.14E-03
(3.69E-04)
3.84E-04
(1.57E-04)
1.07E-04
(4.90E-05)
7.71 E-04
(4.39E-04)
<1.02E-03
(4.13E-05)
2.16E-03
(1.68E-03)
Soy Biodiesel
Average
(Standard deviation)
2.12E-01
(3.56E-02)
1.24E-01
(4.45E-02)
2.63E-02
(2.44E-02)
<6.06E-03 (a)
(1.62E-03)
<1.09E-03
(1.10E-05)
1.85E+00
(2.36E-01)
1.66E-02
(1.09E-02)
2.53E-01
(3.40E-02)
1.21E-01
(1.40E-02)
3.38E-03
(1.42E-03)
8.52E-03
(1.29E-03)
5.27E-03
(1.04E-03)
1.41 E-03
(5.15E-04)
2.85E-03
(6.64E-04)
1.08E-03
(1.01E-03)
<6.51E-04(a)
(4.59E-04)
1.24E-03
(5.03E-04)
<1.09E-03
(1.10E-05)
3.86E-03
(2.49E-03)
Animal Biodiesel
Average
(Standard deviation)
1.50E-01
(3.24E-02)
5.08E-02
(9.80E-03)
5.79E-03
(1.55E-03)
2.63E-03
(1.40E-04)
<1.07E-03
(5.17E-06)
8.64E-01
(1.67E-01)
4.52E-03
(3.14E-03)
1.29E-01
(2.66E-02)
4.58E-02
(1.68E-02)
1.43E-03
(4.69E-04)
3.99E-03
(8.48E-04)
2.79E-03
(7.30E-04)
7.26E-04
(1.11 E-04)
2.23E-03
(1.78E-03)
4.49E-04
(2.38E-04)
<2.08E-04 (a)
(1.83E-04)
8.56E-04
(4.74E-04)
<1.07E-03
(5.17E-06)
3.02E-03
(3.25E-03)
(a) One of three samples were below detection limit.
(b) All samples were below detection limit.
(c) Concentrations shown as less than the stated value had one or more measurements below the
method detection level. Average values are calculated using the method detection level.
Poly chlorinated Biphenyls
PCBs were sampled using EPA Method 0010 (U.S. Environmental Protection Agency
1986) and analyzed using CARS Method 428 (California Air Resources Board 1990).
Concentrations of PCBs are shown in Figure 12. The concentrations measured in these
tests were substantially higher than expected, which is likely to be the result of the
previous testing done to evaluate the potential for dioxin formation during co-firing of
chlorinated wastes in small boilers. Total PCB concentrations were as much as two
orders of magnitude higher than the PAHs, which is extremely high given the lack of
significant chlorine in the three fuels (all less than the 20 ppm detection limit as seen in
34
-------
Table 1). As with the PAHs, PCBs concentrations from the animal biodiesel were lower
in each case than those from the soy biodiesel or the No. 2 fuel oil. PCB concentrations
were higher than those from the soy biodiesel for total chlorinated biphenyls up to and
including total pentachlorinated biphenyls, and were lower than those from the soy
biodiesel for total PCBs with six or more chlorine atoms.
i e^o
1e+5 -
\
Q. 1e+4 -
c
g
2 1e+3 -
§
c
£ 1e+2-
DQ
O
Q.
1e+1 -
1e+0 -
S
/
/
',
'
/
/
'/
-T-
pr
'
/
/
/
^
'
^
/
/
^
I
r
/
^
/
/
^
>
^
/
'.
.
',
/
Tj
^
/
/
^
.
^
/
^
^
/
^
I
I
/
/
^
^
/
/
^
/
^
I
I
/
/
/
'
/
/
/
^
/
^
I
^H No. 2 Fuel Oil
r//xi Soy Biodiesel
I I Animal Biodiesel
I
^
/
^
'
/
/
/
/
^
/
I
IT
7^
/
;
/
/
/
/
/
/
/
j^
s
/
/
^
/
/
;
/
I
000000000000000000
OOOOOOOOO
O V^ "^ CO CO CO CO CO CO
Figure 12. Concentrations of PCBs for each of the three fuels tested.
Concentrations (in pg/Nm3) of 12 individual PCBs and total PCBs by chlorine content are
shown in Table 18. Emission rates in Ib/hr are given in Table 19, and emission factors in
lb/1012 Btu are provided in Table 20.
35
-------
Table 18. Concentrations of PCB for the three fuels tested, in pg/Nm .
3, 3', 4, 4' Tetrachlorobiphenyl
3, 4, 4', 5 Tetrachlorobiphenyl
2, 3, 3', 4, 4' Pentachlorobiphenyl
2, 3, 4,4', 5 Pentachlorobiphenyl
2, 3', 4, 4', 5 Pentachlorobiphenyl
2, 3', 4, 4', 5' Pentachlorobiphenyl
3, 3', 4, 4', 5 Pentachlorobiphenyl
2, 3, 3', 4, 4', 5 Hexachlorobiphenyl
2, 3', 4, 4', 5,5' Hexachlorobiphenyl
3, 3', 4, 4', 5, 5' Hexachlorobiphenyl
2, 3, 3', 4, 4', 5,5' Heptachlorobiphenyl
2, 2', 3, 3', 4,4', 5, 5', 6, 6' Decachlorobiphenyl
Total Mono-CB
Total Di-CB
Total Tri-CB
Total Tetra-CB
Total Penta-CB
Total Hexa-CB
Total Hepta-CB
Total Octa-CB
Total Nona-CB
No. 2 Fuel Oil
Average
(Standard
Deviation)
709
(123)
191
(50.9)
263
(39.8)
71.9
(15.4)
594
(192)
63.3
(9.7)
259
(153)
238
(62.6)
121
(28.2)
90.3
(35.9)
107
(40.5)
119
(32.6)
3140
(1210)
11700
(3270)
214000
(47600)
107000
(22400)
17200
(13900)
31700
(35600)
18900
(20200)
3920
(3000)
417
(115)
Soy Biodiesel
Average
(Standard
Deviation)
587
(109)
131
(24.6)
215
(70.0)
50.2
(15.5)
534
(307)
47.1
(20.1)
183
(48.0)
176
(66.1)
101
(53.0)
56.6
(16.3)
72.2
(19.3)
82.3
(19.0)
1420
(344)
8270
(1500)
191000
(64800)
94400
(28800)
16200
(16200)
32000
(43000)
21700
(23400)
4990
(3240)
441
(159)
Animal Biodiesel
Average
(Standard
Deviation)
362
(34.0)
78.1
(9.40)
128
(7.10)
31.0
(4.00)
313
(67.2)
26.8
(4.00)
96.8
(17.5)
105
(41.2)
62.2
(29.9)
28.8
(6.20)
38.5
(17.3)
48.8
(25.3)
760
(146)
4150
(950)
118000
(28100)
69000
(16500)
8250
(5220)
15600
(11600)
12500
(8320)
3700
(4160)
332
(367)
36
-------
Table 19. Emission rates of PCBs for the three fuels tested, in Ib/hr.
3, 3', 4, 4' Tetrachlorobiphenyl
3, 4, 4', 5 Tetrachlorobiphenyl
2, 3, 3', 4, 4' Pentachlorobiphenyl
2, 3, 4, 4', 5 Pentachlorobiphenyl
2, 3', 4, 4', 5 Pentachlorobiphenyl
2, 3', 4, 4', 5' Pentachlorobiphenyl
3, 3', 4, 4', 5 Pentachlorobiphenyl
2, 3, 3', 4, 4', 5 Hexachlorobiphenyl
2, 3', 4, 4', 5,5' Hexachlorobiphenyl
3, 3', 4, 4', 5, 5' Hexachlorobiphenyl
2, 3, 3', 4, 4', 5,5' Heptachlorobiphenyl
2, 2', 3, 3', 4,4', 5, 5', 6, 6' Decachlorobiphenyl
Total Mono-CB
Total Di-CB
Total Tri-CB
Total Tetra-CB
Total Penta-CB
Total Hexa-CB
Total Hepta-CB
Total Octa-CB
Total Nona-CB
No. 2 Fuel Oil
Average
(Standard
Deviation)
4.01 E-04
(6.22E-05)
1.08E-04
(2.60E-05)
1.49E-04
(2.27E-05)
4.07E-05
(8.39E-06)
3.38E-04
(1.15E-04)
3.58E-05
(5.12E-06)
1.46E-04
(8.21 E-05)
1.34E-04
(3.23E-05)
6.88E-05
(1.62E-05)
5.09E-05
(1.88E-05)
6.02E-05
(2.12E-05)
6.70E-05
(1.65E-05)
1.77E-03
(6.30E-04)
6.60E-03
(1.72E-03)
1.21E-01
(2.70E-02)
6.07E-02
(1.30E-02)
9.82E-03
(8.09E-03)
1.82E-02
(2.06E-02)
1.09E-02
(1.17E-02)
2.24E-03
(1.74E-03)
2.37E-04
(6.75E-05)
Soy Biodiesel
Average
(Standard
Deviation)
3.61 E-04
(7.03E-05)
8.03E-05
(1.46E-05)
1.32E-04
(4.26E-05)
3.08E-05
(9.30E-06)
3.28E-04
(1.86E-04)
2.89E-05
(1.21 E-05)
1.13E-04
(2.84E-05)
1.08E-04
(4.00E-05)
6.19E-05
(3.22E-05)
3.48E-05
(9.68E-06)
4.44E-05
(1.15E-05)
5.06E-05
(1.14E-05)
8.69E-04
(1.99E-04)
5.10E-03
(1.02E-03)
1.18E-01
(4.22E-02)
5.83E-02
(1.88E-02)
9.91 E-03
(9.87E-03)
1.96E-02
(2.62E-02)
1.33E-02
(1.42E-02)
3.08E-03
(1.99E-03)
2.72E-04
(9.93E-05)
Animal Biodiesel
Average
(Standard
Deviation)
2.29E-04
(2.16E-05)
4.93E-05
(6.01 E-06)
8.10E-05
(4.36E-06)
1.96E-05
(2.52E-06)
1.97E-04
(4.21 E-05)
1.69E-05
(2.48E-06)
6. 11 E-05
(1.11 E-05)
6.61 E-05
(2.60E-05)
3.93E-05
(1.89E-05)
1.82E-05
(3.95E-06)
2.43E-05
(1.10E-05)
3.08E-05
(1.60E-05)
4.80E-04
(9.27E-05)
2.62E-03
(6.02E-04)
7.44E-02
(1.78E-02)
4.36E-02
(1.04E-02)
5.21 E-03
(3.28E-03)
9.85E-03
(7.33E-03)
7.87E-03
(5.26E-03)
2.34E-03
(2.63E-03)
2.10E-04
(2.32E-04)
37
-------
Table 20. Emission factors for PCBs from the three fuels, lb/10 Btu.
3, 3', 4, 4' Tetrachlorobiphenyl
3, 4, 4', 5 Tetrachlorobiphenyl
2, 3, 3', 4, 4' Pentachlorobiphenyl
2, 3, 4, 4', 5 Pentachlorobiphenyl
2, 3', 4, 4', 5 Pentachlorobiphenyl
2, 3', 4, 4', 5' Pentachlorobiphenyl
3, 3', 4, 4', 5 Pentachlorobiphenyl
2, 3, 3', 4, 4', 5 Hexachlorobiphenyl
2, 3', 4, 4', 5,5' Hexachlorobiphenyl
3, 3', 4, 4', 5, 5' Hexachlorobiphenyl
2, 3, 3', 4, 4', 5,5' Heptachlorobiphenyl
2, 2', 3, 3', 4,4', 5, 5', 6, 6' Decachlorobiphenyl
Total Mono-CB
Total Di-CB
Total Tri-CB
Total Tetra-CB
Total Penta-CB
Total Hexa-CB
Total Hepta-CB
Total Octa-CB
Total Nona-CB
No. 2 Fuel Oil
Average
(Standard
Deviation)
4.67E-01
(7.25E-02)
1.25E-01
(3.02E-02)
1.73E-01
(2.64E-02)
4.74E-02
(9.77E-03)
3.94E-01
(1.34E-01)
4.17E-02
(5.97E-03)
1.70E-01
(9.56E-02)
1.57E-01
(3.76E-02)
8.01 E-02
(1.88E-02)
5.93E-02
(2.19E-02)
7.01 E-02
(2.46E-02)
7.80E-02
(1.93E-02)
2.06E+00
(7.33E-01)
7.69E+00
(2.00E+00)
1.41E+02
(3.15E+01)
7.07E+01
(1.51E+01)
1.14E+01
(9.42E+00)
2.12E+01
(2.40E+01)
1.26E+01
(1.36E+01)
2.61 E+00
(2.03E+00)
2.76E-01
(7.86E-02)
Soy Biodiesel
Average
(Standard
Deviation)
4.21 E-01
(8.19E-02)
9.35E-02
(1.70E-02)
1.54E-01
(4.97E-02)
3.59E-02
(1.08E-02)
3.82E-01
(2.17E-01)
3.37E-02
(1.41 E-02)
1.31 E-01
(3.31 E-02)
1.26E-01
(4.66E-02)
7.21 E-02
(3.75E-02)
4.05E-02
(1.13E-02)
5.17E-02
(1.34E-02)
5.89E-02
(1.32E-02)
1.01 E+00
(2.32E-01)
5.94E+00
(1.19E+00)
1.38E+02
(4.92E+01)
6.79E+01
(2.18E+01)
1.15E+01
(1.15E+01)
2.28E+01
(3.05E+01)
1.55E+01
(1.66E+01)
3.58E+00
(2.32E+00)
3.16E-01
(1.16E-01)
Animal Biodiesel
Average
(Standard
Deviation)
2.59E-01
(2.45E-02)
5.59E-02
(6.81 E-03)
9.18E-02
(4.94E-03)
2.22E-02
(2.86E-03)
2.24E-01
(4.77E-02)
1.92E-02
(2.81 E-03)
6.93E-02
(1.26E-02)
7.50E-02
(2.95E-02)
4.45E-02
(2.15E-02)
2.06E-02
(4.48E-03)
2.76E-02
(1.24E-02)
3.50E-02
(1.81 E-02)
5.44E-01
(1.05E-01)
2.97E+00
(6.82E-01)
8.43E+01
(2.02E+01)
4.94E+01
(1.18E+01)
5.90E+00
(3.72E+00)
1.12E+01
(8.31E+00)
8.91E+00
(5.96E+00)
2.65E+00
(2.98E+00)
2.38E-01
(2.63E-01)
Results for total PCBs were surprisingly high, particularly for the tri- and tetra-
chlorinated biphenyls (CBs). Comparison with field blank results showed that the values
for these two congener classes were well above blank levels. For the No. 2 fuel oil, the
mono- through penta-CBs all had blank levels lower than 10% of the sample values. The
remaining blank levels were 19% for hexa-CBs, 39% for the hepta-CBs, 65% for the
octa-CBs, and 40% for the nona-CBs. For the soy biodiesel, the blank levels were below
10% for all the total CBs other than hepta-CBs, for which the blank was 10% of the
average sample mass. For the soy biodiesel, only the tri- and tetra-CB blank levels were
38
-------
below 10% of the sample mass. The other blank levels were 14% for the mono-CBs,
57% for the di-CBs, 102% for the penta-CBs, 103% for the hexa-CBs, 61% for the hepta-
CBs, 30% for the octa-CBs, and 18% for the nona-CBs.
Based on these values, only the penta- and hexa-CBs from the animal biodiesel are
questionable in terms of their presence in the flue gases. The high overall levels across
all three fuels is likely a consequence of the fact that the boiler was used as a test bed for
dioxin formation tests prior to use in the current tests. Although over a year had passed
since those tests were performed, and even though the boiler itself was cleaned prior to
the biofuel tests, it is likely that residues from the earlier tests resulted in the high PCB
values reported here. It is not believed that these values are representative of PCB
emissions from No. 2 fuel oil or biofuels in general.
Boiler Efficiency
The thermal efficiency of the boiler was estimated during operation with each of the three
fuels. A complete efficiency analysis was not conducted for these tests, but an
input/output analysis was conducted to estimate differences in boiler efficiency. The
input energy was determined from the fuel heat of combustion (reported as the higher
heating value) and fuel mass flow, and the output energy was calculated from the change
in temperature of the boiler cooling water loop and the cooling water flow.
Measurements of fuel energy input and cooling water energy output were made for each
test run, providing three values for each of the three fuels. Table 21 provides the energy
input and output values, the calculated thermal efficiencies, and the standard deviation of
the efficiency estimates for each of the three fuels.
Table 21. Results of boiler efficiency measurements.
Fuel
No. 2 Fuel Oil
Soy Biodiesel
Animal Biodiesel
Energy Input, 10B
Btu/hr
1,890,000
1,890,000
1,940,000
Energy Output,
10s Btu/hr
1,610,000
1,580,000
1,670,000
Thermal
Efficiency, %
85.1%
83.8%
85.7%
Standard
Deviation, %
5.95%
6.86%
6.31%
The lower average value for the soy biodiesel is not significant, as the range for the soy
biodiesel overlaps the ranges for the other two fuels. The difference is most likely due to
normal measurement variability, as suggested by the overlap in average plus or minus
standard deviation. A brief examination of the combustion gas concentrations does not
provide any indication that other factors are involved in the difference. The C>2 level was
lower for the soy biodiesel compared to the other two fuels, and the CO level was below
3 ppm, both of which would indicate that the difference in efficiency is not due to excess
combustion air or poor fuel burnout.
Life Cycle Analysis
When evaluating biofuels as a means to reduce greenhouse gas (GHG) emissions, it is
important to understand the emissions over the fuel lifecycle, e.g., the entire production
and use of the fuel. This is because GHGs are unlike criteria pollutants, which tend to be
relatively short-lived and at unhealthy concentrations over relatively local spatial scales.
GHGs, on the other hand, tend to be quite long-lived and therefore act over global scales.
Thus, GHG emissions at any point in the production and use cycle contribute to the
39
-------
effects of increased GHG concentration, and emissions from any part of the lifecycle are
therefore of concern.
The lifecycle of biodiesel includes the production of the feedstock, the transport of the
feedstock to the biodiesel production facility, the conversion of the feedstock to biodiesel,
the transport of the biodiesel to the end user, and the combustion of the biodiesel. Each of
the steps from production to transport to the end user requires some level of energy input,
which may be in the form of direct energy (such as fuel to generate steam in a conversion
process) or indirect energy (such as the energy content of the natural gas used to produce
nitrogen fertilizer). These energy inputs are required for both biodiesel and No. 2
distillate oil, although many of the processes will be significantly different.
There have not been any studies that compare distillate heating oil to biodiesel, but there
have been studies that have compared lifecycle emissions of highway diesel fuel (very
similar to No. 2 distillate oil) to biodiesel. Sheehan et al. evaluated the lifecycle energy
inputs for a highway diesel fuel and a soy-based biodiesel in a 1998 report, Life Cycle
Inventory of Biodiesel and Petroleum Diesel for Use in an Urban Bus (Sheehan,
Camobreco et al. 1998). They estimated that biodiesel requires 1.24 units of energy input
to produce fuel with 1 unit energy (including feedstock production energy inputs), while
petroleum diesel fuel requires 1.20 units of energy input to produce fuel with 1 unit of
energy content. However, the fossil energy input requirements were significantly
different, because of the fact that the biodiesel feedstock is soy oil rather than a fossil
energy source (crude oil). Therefore, the ratio of fossil energy input to fuel energy output
is just under 1.20 for the diesel fuel compared to 0.31 for the biodiesel.
More recently, Hill et al. reported similar results, with a "net energy balance" of 0.73,
based on 1 unit of fuel energy (Hill, Nelson et al. 2006). The net energy balance is
calculated by subtracting the total energy used in biofuel production from the energy
present in the fuel. To directly compare the net energy balance to the results of Sheehan
et al. requires one to assume that all the production input energy is in the form of fossil
energy, which may not be the case. However, given the probable increases in process
efficiencies between 2000 and 2006, and the likelihood of agricultural and conversion
processes to be almost entirely fueled by fossil energy, it seems reasonable to directly
compare these two results. Direct comparison shows the Hill et al. value is very close to,
but slightly higher than, the Sheehan et al. result, which would convert to a net energy
balance value of 0.69.
MacLean et al. report that biodiesel vehicles would consume 843 GJ of fossil energy, and
864 GJ of total energy, over the entire fuel lifecycle and vehicle operation (MacLean,
Lave et al. 2000). However, they estimate that the vehicle operation would consume 612
GJ of energy over the vehicle's lifetime, and their calculations evidently assume this 612
GJ is entirely fossil energy. If the fuel being analyzed is B100 (100% biodiesel), then
there should be no fossil energy consumption during vehicle operation. The MacLean et
al. results contradict the results of Sheehan et al. discussed above, although MacLean et
al. cite the Sheehan et al. study as a key source of information on biodiesel lifecycle
energy use. It would appear that the MacLean et al. study erroneously includes the 612
40
-------
GJ as fossil energy, which would result in a fossil fuel input requirement of 231 GJ and a
total energy consumption of 864 GJ, which would result in a net fossil energy input to
fuel energy output ratio of 0.27, much more in line with the Sheehan et al. results. In
terms of the net energy balance, this would translate to 0.73, the same as Hill et al. and
slightly higher than Sheehan et al.
Sheehan et al. is the only study to estimate net (direct and indirect) CO2 emissions
explicitly, as opposed to inferring emissions based on the energy balance. It should be
noted that their analysis is for a different fossil fuel - highway diesel fuel - than what is
used in the boiler tests reported here. However, the lifecycle assessment to the point of
end use will be largely the same, given the similarities between highway diesel fuel and
No. 2 distillate oil. They estimated that highway diesel fuel would emit a net 633 g-
CO2/bhp-h, compared to 136 g-CO2/bhp-h for the soy biodiesel (Sheehan, Camobreco et
al. 1998).
This is a significant reduction in net CO2 emissions, and it should be expected that a
similar percent reduction would be possible when using biodiesel in a boiler, since all the
fossil energy input is upstream of the end use. If the thermal efficiency of the unit does
not change when using the biodiesel compared to using the No. 2 distillate oil, the net
change in CO2 will be entirely associated with the reduction in fuel production and
conversion. Assuming a similar percent reduction, the net CO2 emissions would drop by
about 79% when using the biodiesel compared to the No. 2 distillate oil.
For these tests in particular, one can evaluate the total conversion of fossil energy to
steam energy as a means to compare the potential CO2 impacts of using biodiesel as a
boiler fuel. Based upon the fossil fuel inputs to fuel energy output ratios estimated by
Sheehan et al., petroleum diesel requires 1.2 units of fossil fuel input per unit of useful
fuel energy output, while soy-based biodiesel requires 0.31 units of fossil fuel input.
Using the boiler efficiency values presented in Table 21, the net fossil fuel input per
steam energy output for this boiler can be seen in Table 22 below.
Table 22. Fossil fuel input required per unit of thermal energy output for the boiler tested.
Fuel
No. 2 Fuel Oil
Soy Biodiesel
Animal Biodiesel
1.41
0.37
0.36ia)
(a) Assumes the fossil fuel input for animal
biodiesel is the same as the soy biodiesel
fossil fuel input.
The values above assume the average fossil fuel input for both this particular soy
biodiesel and animal biodiesel; however, the significant reduction in total fossil fuel
content indicates that the net CO2 emissions using biodiesel will be significantly reduced
in comparison to petroleum distillate fuel.
41
-------
QA Discrepancies
Several differences occurred during the tests between the actual conduct of the tests and
the planned methods or approaches as stated in the QAPP. These are noted below, along
with corrective actions (if any) and impacts on the results of the tests.
Discrepancy: The QAPP states that, "if the 'past-optimum' O2 concentration is below
3%, the tests will be run with flue gas C>2 concentration at 3%" (the "past-optimum" C>2
concentration was defined as the C>2 concentration at which the CO began to increase
significantly as C>2 level was reduced). In practice, the boiler was operated at C>2 levels
well below 3% for all three fuels (see Table 4), at between 1.4% and 1.9%. The CO
concentrations for these Oi levels were very low (less than 5 ppm), indicating that higher
O2 levels would have resulted in reduced boiler efficiency without any improvement in
combustion efficiency due to the increased mass of combustion air being used.
Resolution: No corrective actions were taken. Operation at the lower 02 concentrations
was appropriate and reflects actual boiler operation; therefore the results when using the
lower 02 concentrations are appropriate and represent the conditions the tests were
designed to evaluate.
Discrepancy: The QAPP states that manual sampling will be conducted only when
steady state conditions have been reached (and presumably maintained, although this
second criterion was not explicitly stated), as determined by flue gas O2 concentration
remaining within 0.3%, boiler steam temperatures within 3 °F, boiler steam pressure
within 0.5 psig, and heat exchanger cooling water temperature rise within 2 °F over a
period of 20 minutes. In practice, manual sampling was conducted during periods when
each of these criteria varied outside the stated bounds. Flue gas O2 concentrations
changed by up to 0.7%; boiler steam temperature changed by 4 °F in one instance, boiler
steam pressure changed by as much as 2 psig, and heat exchanger cooling water (outlet)
changed as much as 5 °F within 15 minutes (these changes occurred distinct from one
another).
Resolution: There is no evidence from measurements that the transient conditions had
any major impact on the emissions of concern. Although there are instances where
transient changes in O2 concentration coincide with changes in CO concentration, the CO
concentrations never exceeded 20 ppm, and the highest excursions only lasted for about
90 seconds. The remaining criteria did not appear to have any influence on emissions,
although the boiler and cooling water temperatures will influence boiler efficiency
calculations. Although the conditions under which the sampling was started each day
need to be more clearly communicated in future tests, the samples collected during
periods when boiler was operating outside the steady state criteria still represent valid
data that will be representative of actual unit operation.
Discrepancy: During a ten-minute period on the second test day (November 1, 2007),
several CEM values appear to be at monitor calibration levels for a short period. At 8:39,
the SO2 concentrations rose rapidly from 1.7 ppm to about 140 ppm (the concentration of
the SO2 calibration gas), stayed near that concentration for about three minutes, and then
dropped back down to 2.8 ppm at 8:42. From that point, the SO2 concentration never
rose above 6 ppm. At 8:44, O2, CO2, NOX, and NO concentrations suddenly dropped to
42
-------
near zero, and then returned to their previous levels about five minutes later. During this
period, no significant change was seen in the CO concentration, which would indicate
that the C>2 concentration in the boiler had not changed significantly.
Resolution: Based on the step changes in recorded concentrations to either zero or to the
level of the calibration gas, and the absence of any evidence of operational upset, it is
clear that these transient excursions were due to changes in the valve settings to the
CEMs, and not to any changes in the boiler operation. The data during these transient
periods will not be used in the calculations of average concentrations or concentration
variability. In addition, the brief period of the excursions and the rapid return to pre-
excursion levels indicate that simply excluding these data will result in an accurate
reflection of the boiler's actual emissions.
Discrepancy: During testing on November 5, 2007, CEM data were not stored by the
data acquisition system during the first test run of the day. The data were recorded from
8:00:33 to 8:03:52, at which point the recorded data end until data logging was resumed
at the beginning of the second test run. CEM data were then recorded from 11:10:16
through the end of the day's testing at 21:03:21.
Resolution: The collection of data for nearly 10 hours is adequate to determine the
emissions of those gases measured by CEM (C>2, CO2, CO, NOX, NO, SO2 and HC), and
meets the stated test duration in Table 2 of the QAPP. Comparison of these gas
concentrations to the measurements using other fuels is therefore not compromised.
The largest impact of the loss of the CEM data is the inability to identify short-term
transient excursions in the measured gases. Although hand-written data sheets provide
data on these gas concentrations at 15 minute intervals during the period of data loss, they
do not provide guidance with respect to possible transient excursions that may have
lasted less than 15 minutes. No such excursions were noticed by the operators.
Discrepancy: A Method 10 sampling train was dropped and a glass impinger was
broken as the sampling train was being recovered after the first test run on November 5,
2007 (animal biodiesel test).
Resolution: A fourth test run was conducted on November 5 to allow an additional
Method 10 sample to be collected. This additional run provides the necessary three
samples required in the QAPP, and there was no loss of data for the November 5 tests.
Discrepancy: Tests of the blends of soy biodiesel and No. 2 fuel oil were not conducted
due to problems with the fuel system "gumming up" and preventing adequate fuel flow.
As the unit was being prepared for initial testing of the first biofuel blend (B20),
problems were noted with the fuel flow. A semi-solid, gelatin-like material was found to
be impeding flow through the fuel lines and the fuel filter. Repeated attempts to operate
the unit were not successful, and the tests had to be halted.
Resolution: This problem was not resolved. No data were collected for the biofuel/fuel
oil blends. These data will need to be collected in future testing.
43
-------
Conclusions
In general, use of biodiesel as a replacement for petroleum distillate fuel appears to have
little, if any, disadvantage from an environmental perspective. Emissions of all pollutants
measured in this study are roughly the same or lower for the biofuels than for the
petroleum distillate fuel.
For the gas-phase criteria pollutants, there was little difference in CO or NOX emissions
across the emissions from the three fuels, with the CO emissions being less than 5 ppm
and the NOX emissions near 110 ppm for all three fuels. The largest difference was for
SO2, where the No. 2 fuel oil had emission concentrations of about 17 ppm and the two
biodiesels each had less than 5 ppm. Total hydrocarbon emissions were below the
detection level of the instrument for all three fuels.
PM emissions were highest for the No. 2 fuel oil, at 18 mg/dscm. Most of the PM mass
for the No. 2 fuel oil was in the condensed phase. The two biodiesel fuels each had total
PM mass emission concentrations between 3 and 4 mg/dscm, with the majority of that
(approximately 2.5 mg/dscm) in the filterable fraction.
Aldehydes were significantly higher for the soy biodiesel compared to the No. 2 fuel oil
and the animal biodiesel. However, the variability measured over the three runs was high
enough that significant uncertainty remains with respect to this conclusion.
Other volatile and semivolatile organic emissions suggest that these compounds may be
at higher concentrations in the animal biodiesel compared to the other two fuels. In all
cases, the emissions were roughly of the same order of magnitude, which indicates that
there were no fundamental differences in emissions. However, as for the other organic
compounds (aldehydes, PAHs, and PCBs), the variability in the measurements was
significant for all compounds and all three fuels. The significance of the differences in
VOC and SVOC concentrations remain highly uncertain.
General trends in the distributions of PAHs and PCBs were very similar, indicating that
there are no significant differences in the formation pathways for these compounds that
depend on the characteristics of these three fuels. PCB emissions seem particularly high,
likely to be a consequence of the previous series of tests. In general, there was little
significant difference between petroleum and biodiesel fuels noted for PAH and PCB
emissions.
Potential problems in boiler operation were noted at the start of the tests of
petro/biodiesel blends. It is unclear whether these problems were specific to the fuel
blends, or are a more general problem with biodiesel fuels in general. The length of
operating time on unblended soy and animal biodiesels suggest that the issue may be
specific to the blends, but additional study is needed to determine whether this is indeed
the case or not. Such problems could impact emissions, particularly with respect to
unburned fuel, CO, and organic compounds.
44
-------
Although the average boiler efficiency for the soy biodiesel was lower than for the other
two fuels, the variability in the results was such that it is not possible to determine
whether there were significant differences in boiler efficiency when using the three fuels.
The difference in boiler efficiency was relatively small, although over long-term
operation, such differences can be significant in terms of total emissions.
CC>2 emissions when using biodiesel are estimated to be significantly lower than for
petroleum distillate fuel, based on the previous life cycle analyses evaluated for this
study. The stack concentrations of CC>2 were nearly the same for all three fuels, but the
two biofuels had CC>2 emission factors 15-17% lower than the No. 2 fuel oil. The life
cycle CC>2 emissions for biodiesel are difficult to determine for boiler applications, but
one estimate for life cycle CO2 emissions for mobile sources estimated that the total CO2
emissions for soy biodiesel were only about 22% those for highway diesel fuel.
References
American Society for Testing and Materials (2007). Specification for Biodiesel (B100) -
ASTMD6751-07b.
Baribeau, A.-M., R. Bradley, P. Brown, J. Goodwin, James , U. Kihm, E. Lotero, D.
O'Connor, M. Schuppers and D. Taylor (2006). Biodiesel from Specified Risk
Material Tallow: An Appraisal of TSE Risks and their Reduction. Advanced
Technology Fuels CanadaOttawa, Ontario.
Batey, J. E. (2003). Combustion Testing of a Bio-Diesel Fuel Oil Blend in Residential Oil
Burning Equipment. Massachusetts OilHeat CouncilWellesley Hills, MA.
California Air Resources Board (1990). Method 428: Determination of Poly chlorinated
Dibenzo-p-dioxin (PCDD), Polychlorinated Dibenzofuran (PCDF), and
Polychlorinated Biphenyl Emissions from Stationary Sources.
California Air Resources Board. Sacramento, CA.
California Air Resources Board (1997). Method 429: Determination of Polycyclic
Aromatic Hydrocarbon (PAH) Emissions from Stationary Sources.
California Air Resources Board. Sacramento, CA.
Ethanol Market. (2007). "Live Commodity Prices." Retrieved November 1, 2007, from
http://ethanolmarket.aghost.net/.
Greene, A. K., P. L. Dawson, D. Nixon, J. R. Atkins and G. G. Pearl (2007). Safety of
Animal Fats for Biodiesel Production: A Critical Review of Literature.
Hill, J., E. Nelson, D. Tilman, S. Polasky and D. Tiffany (2006). "Environmental,
economic, and energetic costs and benefits of biodiesel and ethanol biofuels."
Proceedings of the National Academy of Sciences of the United States of America
103(30): 11206-11210.
Krishna, C. R. (2001). Biodiesel Blends in Space Heating Equipment. Brookhaven
National Laboratory. BNL-68852. Upton, NY.
MacLean, H. L., L. B. Lave, R. Lankey and S. Joshi (2000). "A life-cycle comparison of
alternative automobile fuels." Journal of the Air & Waste Management
Association 50(10): 1769-1779.
New York State Energy Research and Development Authority (2004). Statewide
Feasibility for a Potential New York State Biodiesel Industry. New York State
Energy Research and Development Authority Alb any, NY.
45
-------
North Carolina State University. (2007). "Database of State Incentives for Renewables &
Efficiency." Retrieved November 1, 2007, from http://www.dsireusa.org.
Seidel, B., M. Aim, R. Peters, W. Kordel and A. Schaffer (2005). "Safety evaluation for a
biodiesel process using prion-contaminated animal fat as a source."
Environmental Science and Pollution Research 13(2): 125-130.
Sheehan, I, V. Camobreco, J. Duffield, M. Graboski and H. Shapouri (1998). Life Cycle
Inventory of Biodiesel and Petroleum Diesel for Use in an Urban Bus. National
Renewable Energy Laboratory. NREL/SR-580-24089. Golden, CO.
Tyson, K. S. (2001). Biodiesel Handling and Use Guidelines. National Renewable
Energy Laboratory. NREL/TP-580-30004. Golden, CO.
U.S. Department of Energy (2006). Annual Energy Review 2005. Energy Information
Administration. DOE/EIA-0384(2005). Washington, DC.
U.S. Department of Energy (2007a). Clean Cities Alternative Fuel Price Report. Office of
Energy Efficiency and Renewable EnergyWashington, DC.
U.S. Department of Energy. (2007b). "Heating Oil and Propane Update." Retrieved
November 1, 2007, from http://tonto.eia.doe.gov/oog/info/hopu/hopu.asp.
U.S. Department of Energy (2007c). Renewable Energy Trends in Consumption and
Electricity, 2005. Energy Information AdministrationWashington, DC.
U.S. Environmental Protection Agency (1986). Method 0010: Modified Method 5
Sampling Train. Office of Solid Waste and Emergency Response. Test Methods
for Evaluating Solid Waste, Physical/Chemical Methods (SW-846). Washington,
DC.
U.S. Environmental Protection Agency (2005). "Method 202 - Determination of
Condensible Paniculate Emissions From Stationary Sources." Code of Federal
Regulations. Title 40 Part 60 (Appendix M): 426-431.
46
------- |