Proceedings

Proceedings of
of ASME
ASME Turbo
Turbo Expo
Expo 2010:
2010: Power
Power for
for Land,
Land, Sea
Sea and
and Air
Air
GT2010
GT2010
June
June 14-18,
14-18, 2010,
2010, Glasgow,
Glasgow, UK
UK

GT2010-
GT2010-22494

EMISSIONS FROM A GAS TURBINE SECTOR RIG OPERATED

WITH SYNTHETIC AVIATION AND BIODIESEL FUEL

Greg Pucher and William Allan Pierre Poitras

Royal Military College of Canada Quality Engineering Test Establishment
Kingston, Ontario, Canada K7K 7B4 Gatineau, Quebec, Canada J8L 1A1

ABSTRACT Council (NRC). Potential candidate synthetic fuels for these air
Differences in exhaust emissions, smoke production, force applications are derived from various sources and
exhaust pattern factor, deposit build-up and fuel nozzle spray production processes and generally possess properties quite
characteristics for various blends of conventional commercial similar to those of conventional Jet A-1 [1]. The complete or
jet fuel (Jet A-1) with synthetic and biodiesel formulations near absence of aromatic content has rendered synthetic fuels
were investigated. Three synthetic fuel formulations and four as superior to conventional blends in terms of smoke /
Fatty Acid Methyl Esters (FAME) were evaluated as such. The particulates production and combustion system deposit build-
synthetic fuels were tested in both neat (100%) and 50% by up [2]. Aromatic content in jet fuel however serves to prevent
volume blends with Jet A-1, while the FAME fuels were fuel leaks by facilitating the swelling of seals in fuel systems.
blended in 2% and 20% proportions. The 50/50 blend objective ensures a minimum quantity of
The Combustion Chamber Sector Rig (CCSR), which aromatic content to satisfy this requirement.
houses a Rolls Royce T-56-A-15 combustion section, was In a Canadian context, there has also been interest in the
utilized for emissions, deposits and exhaust pattern factor addition of biodiesel fuels (Fatty Acid Methyl Esters-(FAME))
evaluation. A combustion chamber exhaust plane traversing to conventional jet fuel formulations in low concentrations to
thermocouple rake was employed to generate two dimensional ease dependence on conventional feedstocks and to promote
temperature maps during operation. Following combustion renewable industries. FAME fuels have enjoyed considerable
testing, several combustion system components, including the success in land-vehicle diesel applications due to their neutral
combustion chamber, fuel nozzle and igniter plug were effects on engine systems designed for conventional fuels and
analyzed for relative levels of deposit build-up. A Phase they can safely be blended in varying proportions.
Doppler Anemometry (PDA) system was employed to In this study, using neat Jet A-1 fuel as a baseline, the
determine differences in droplet size distributions while an combustion and atomization related impacts of blending
optical spray pattern analyzer was used to compare the spray various FAME and synthetic jet fuel formulations with Jet A-1
pattern for the various fuel blends as they emerged from the T- was investigated. Additionally, the impacts of utilizing neat
56 nozzle. synthetic fuel as opposed to neat Jet A-1 were investigated.
This was facilitated by combusting these fuel blends in a gas
INTRODUCTION turbine sector rig to assess exhaust emissions, smoke
The significance of aviation fuels derived from non- production, exhaust pattern factor and combustion chamber
petroleum sources continues to intensify as issues related to the soot build-up. Four different FAME fuels in concentrations of
security of supply and the depletion of conventional feedstocks 2% and 20% in Jet A-1 fuel were evaluated. Three types of
persist. The Canadian Air Force (CAF) is participating in a synthetic fuel were tested, including Gas to Liquid (GTL), a
U.S. Air Force (USAF) led international effort to fuel military blend of GTL and Coal to Liquid (CTL) and a prototype
aircraft with a 50/50 blend of conventional and synthetic jet Hydrotreated Renewable Jet (HRJ) blend. Concentrations of
propellants within the next decade. To this end, most of the Hydrocarbons (HC), Carbon Monoxide (CO), Oxides of
USAF aircraft engine families have already been certified and Nitrogen (NOx), Oxygen (O2), Carbon Dioxide (CO2) and
the CAF has recently certified the General Electric F-404 (F- smoke within the exhaust gas were sampled and measured. A
18) gas turbine engine through the Canadian National Research sweeping thermocouple rake provided information on the

Copyright © 2010 by ASME and Department of National

1 Defense, Government of Canada

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/29/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 comparative thermal uniformity of the exit stream during means are employed to obtain the CO + H2 (synthesis gas)
combustion of the different fuels. The build-up of deposits in input stream. Two of the synthetic fuels investigated in the
the combustion chamber system was evaluated by the weighing current study are derived from coal and NG feedstocks. In the
of components and the use of digital photography. Optical Coal to Liquid (CTL) and Gas to Liquid (GTL) processes, a
spray patternation and phase Doppler anemometry were partial oxidation of the feedstocks in the presence of steam,
employed to investigate nozzle fouling effects as well as spray oxygen and a catalyst results in the formation of the synthesis
characteristics for the various fuel types. All of these were gas [1]. Following the FT process, further processing of these
compared to the baseline case of operation with neat Jet A-1. compounds with actions such as cracking, isomerization and
The results of these evaluations were intended to determine fractionation results in compounds which can serve as aviation,
whether observed differences in exhaust emissions, smoke diesel or other fuel substitutes [5]. The CTL and GTL processes
levels and combustion chamber deposits were attributable not are considered indirect conversion processes, due to the
only to chemical differences between fuel formulations, but intermediate step of synthesis gas production. Direct
possibly also to physical properties (related to atomization) conversion processes also exist, such as pyrolysis, carbonation,
which might be observed in the spray analysis and exhaust fermentation and hydrogenation [3]. Hydrogenation processes
pattern factor measurements. are under development, employing both bio-mass and oils/fats.
These processes involve deoxygenation, followed by
NOMENCLATURE hydrocracking and isomerization to produce synthetic fuels. A
CAF Canadian Air Force third synthetic fuel tested in the current study, a Hydrotreated
CCSR Combustion Chamber Sector Rig Renewable Jet (HRJ) fuel, was derived from the hydrogenation
CO Carbon Monoxide of algae oils.
CTL Coal to Liquid
EI Emission Index Synthetic Jet Fuel Performance
FAME Fatty Acid Methyl Ester A significant amount of investigation has been conducted
FSN Filter Smoke Number to determine the relative performance merits of synthetic fuels
FT Fischer Tropsch in gas turbine engine applications [1,2,4-8]. In a landmark
GTL Gas to Liquid study conducted at the USAF Research Labs [1], an FT fuel
H2 Hydrogen was blended in varying proportions with conventional JP-8 and
HC Hydrocarbons combusted in both a T63 turboshaft engine as well as a swirl
HRJ Hydrotreated Renewable Jet Fuel stabilized research combustor operated at atmospheric pressure.
NG Natural Gas With both test platforms, particulate reductions in excess of
NOx Oxides of Nitrogen 90% for neat FT fuel as compared to neat JP-8 fuel were
NRC Canadian National Research Council realized during an idle operation condition. Slightly more
O2 Oxygen modest improvements were realized with a cruise operation
OSPA Optical Spray Pattern Analyzer condition. For different blends of JP-8 and FT fuel, it was also
PDA Phase Doppler Anemometer found that PM emissions decreased proportionally with the
PM Particulate Matter fraction of FT fuel within the blend. The large scale reduction
PND Particle Number Density in particulate matter (PM) production was linked to the
RMC Royal Military College of Canada complete or near absence of aromatic content in the FT fuel.
SMD Sauter Mean Diameter The absence of aromatics was believed to decrease the
SPK Synthetic Paraffinic Kerosene precursors that lead to the soot nuclei formation. The
USAF United States Air Force mechanisms of soot production with FT fuel combustion are
described as polymerization and fragmentation reactions, which
BACKGROUND due to the presence of only normal and aliphatic compounds,
are ultimately less efficient than the condensation reactions of
Synthetic Fuels aromatics in producing soot. The absence of aromatics in
Liquid fuels derived from coal, NG and biomass are synthetic fuels is believed to slow the molecular growth to
considered synthetic fuels [3]. A variety of processes exist to higher ringed polycyclic hydrocarbons which reduces soot
transform raw materials as such. The dominant current methods emissions [5].
for synthetic jet fuel production are variants of the Fischer In a study aimed at determining the effects of specific
Tropsch (FT) process, as employed by large scale producers aromatic types on PM, it was found that PM increased linearly
such as Sasol (South Africa), Shell (Indonesia) and Rentech with increasing aromatic molecular weight as well as
(U.S.A.). In the FT process, a conversion of Carbon Monoxide concentration [2]. The type of the aromatic content was shown
(CO) and Hydrogen (H2) into an iso-paraffinic substance is to have a significant impact on particle size distribution. These
undertaken in the presence of a catalytic material such as iron effects were attributed to issues surrounding the concentration
and cobalt [4]. Depending on the source hydrocarbon, various of soot precursors and increased rates of particle nucleation and

Copyright © 2010 by ASME and Department of National

2 Defense, Government of Canada

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/29/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 surface growth. A higher hydrogen to carbon ratio of synthetic performance and the impact of fuel nozzle and combustion
fuels was also cited as a likely contributor to reduced soot chamber damage on exhaust temperature pattern factor
production with synthetic fuels. Combustion efficiency has also [15,16,17]. While the combustion systems within the rig
been predicted to be slightly higher for synthetic fuel, in the comprise actual T56 components, the inlet, air delivery and fuel
order of 2-3%. For other primary emissions, such as HC, CO, systems are custom designs. The T56 combustion liner is of
CO2 and NOx, synthetic fuel was found to have limited impact can-annular design and there are 6 combustion liners per
in gas turbine applications [8]. engine. Within the CCSR rig, only one liner supports
Synthetic fuels have been noted for their high thermal combustion, while the neighbouring liners serve to provide the
stability characteristics, exceeding the performance attained active liner with a more realistic airflow simulation. Figure 1
with the F37 (JP-8+100) formulation [3]. Thermal stability of shows the CCSR in a partially disassembled state, revealing the
jet fuels is increasingly important for modern aircraft designs, operational (center) and neighbouring liners.
given the growing demands on stored fuels to serve as heat
sinks. The lower freezing point with respect to conventional jet
fuel of FT fuel (< -60oC vs ~-50C) is also deemed to be a
beneficial property for high altitude and cold climate
applications.
The aromatic content in conventional petroleum-based jet
fuels serves to swell elastomer materials of seals within fuel
systems and thereby prevent fuel leaks. Neat FT fuel is deemed
to be deficient in this functionality. The targeted 50/50 blend of
JP-8 and FT synthetic fuel for air force applications is expected
to resolve this issue by ensuring sufficient aromatic content to
provide elastomer sealing, as well as satisfying density
requirements, as the density of neat FT fuel is typically lower
than the minimum accepted value in the JP-8 specification
[10,11].

Fatty Acid Methyl Ester (Biodiesel) Fuels

FAME fuels are a variety of ester-based oxygenated fuels Figure 1: Partially Disassembled Combustion
made from substances such as vegetable oils, animal fats and Chamber Sector Rig (CCSR) Revealing the Main and
recycled waste cooking oils. Such base substances are Adjacent ‘Dummy’ Combustion Liners
converted to FAME in a transesterification process, whereby
they are combined with methanol or ethanol in the presence of The target inlet conditions of the active combustion liner
a catalyst [12]. The resulting FAME fuels typically have during CCSR rig operation were mass airflow: 0.465 kg/s; inlet
viscosity levels on the order of one eighth of the original temperature: 315 K; and inlet pressure: 16 kPa. These values
feedstock. combine to produce a quasi non-dimensional Mach number, as
There is a general perception that there are significant described in Equation 1. The values of these parameters were
benefits in terms of emissions and PM production to using chosen such that the calculated value of Mach number matched
FAME blends in diesel engine applications [13]. In a recent that for an in-flight cruise condition. Thus, despite significant
comprehensive testing program done with a LAV3 engine, differences in airflow, temperature and pressure, a certain
statistically significant reductions in PM, but neutral effects on similarity is attained between rig and engine in terms of flow
gaseous emissions were realized for 20% concentrations of velocities.
FAME fuels in diesel [12]. In a study on the impacts of FAME 
ˆ  mair T3
Ma (1)
fuel on emissions from a T63 gas turbine engine, it was found
P3
that at cruise and take-off conditions, a 15% reduction in PND
was observed with a 20% concentration of FAME in JP-8 [14]. Given that the air flow through the active combustion liner
in the CCSR rig was much lower than that of a pressurized in-
EXPERIMENTAL flight liner, a much lower fuel flow was also required. As such,
a T56 fuel nozzle was modified by blocking its secondary flow
Combustion Chamber Sector Rig circuit, which was verified to produce a fuel droplet size
The Combustion Chamber Sector Rig (CCSR), which was distribution and cone angle which closely simulated an
jointly designed by the Royal Military College of Canada and unmodified nozzle in an in-flight operation [16]. A target
Standard Aero Limited, has served in various experimental equivalence ratio of 0.23 was chosen for tests.
programs related to the Allison/Rolls Royce T56-A-15 A stepper motor-controlled water-cooled sweeping rake
combustion systems. These studies have focused on fuel housing 4 K-type thermocouples was used to scan the 2-

Copyright © 2010 by ASME and Department of National

3 Defense, Government of Canada

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/29/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 dimensional exit plane (fishmouth) to determine temperature It was ascertained through fuel analysis that in order to remain
maps within a 1 resolution. This rake is shown in Figure 2. within the fuel specification for Jet A-1, a maximum of 2%
FAME could be tolerated. Higher concentrations resulted in
excess gum content in the blended fuel. Despite this limitation,
testing at a 20% concentration was also conducted to further
understand the impacts of FAME addition to Jet A-1. Due to
availability issues, the base fuel used for testing was
commercial Jet A-1 and not JP-8. The Jet A-1 fuel specification
is very similar to that of JP-8 and differs only by an absence of
corrosion inhibitor/lubricity enhancer, fuel system icing
inhibitor and a static dissipater additive [18].
For the Jet A-1 tests, the fuel/air equivalence ratio was set
to 0.23. The precise amount of fuel delivered to the nozzle with
each (non Jet A-1) fuel blend during testing was controlled so
that the exhaust temperature matched the value measured
during the baseline neat Jet A-1 test. In order to meter fuel flow
based on calorific energy delivered, detailed knowledge of the
heating value of each fuel was required, which was not
available at the time of testing. By matching the exhaust
temperature instead, this requirement was bypassed. An
Figure 2: Thermocouple Rake within the CCSR Rig equilibrium analysis of the exhaust emissions (which was
performed in order to determine Emissions Index (EI) values of
Fuel Blends hydrocarbons) ascertained that values of equivalence ratio for
The following is a list of the fuels and blended proportions all fuel blends were within 5% relative to neat Jet A-1, which
as they were evaluated in the current study (percentages are in was considered acceptable. Values of fuel/air ratio were within
terms of volume): 6% of the value for Jet A-1, a higher deviation due to the
• Neat Jet A-1 slightly different stoichiometry between the fuels tested.
• Jet A-1 + Fish Oil FAME (2% and 20%)
• Jet A-1 + Vegetable Oil FAME (2% and 20%) Emissions Measurements
• Jet A-1 + Canola Oil FAME (2% and 20%) Emissions were measured with the devices as listed below.
• Jet A-1 + Bio Oil FAME (2% and 20%)
• 50/50 Jet A-1 / GTL FT Synthetic Jet Fuel Emissions Species Method
• 100% GTL FT Synthetic Jet Fuel Hydrocarbons (HC) Flame Ionization Detector
• 50/50 Jet A-1 / (CTL/GTL) NRC FT Synthetic Jet Carbon Monoxide (CO) Infrared Detector
Oxides of Nitrogen (NOx) Chemiluminescence Detector
Fuel Carbon Dioxide (CO2) Infrared Detector
• 100% Algae Oil HRJ Synthetic Jet Fuel Oxygen (O2) Infrared Detector
Table 1 lists some of the key properties of the base fuels tested, Smoke (FSN) AVL Smoke Meter- filter paper method
along with the acceptable limits for the F-34 specification.
Exhaust gas was sampled from a port in the exhaust
Table 1: Fuel properties of the evaluated fuel blends ducting approximately 1.5 m downstream of the liner
fishmouth exit and transferred via a heated sample line to a
Fuel Heating Value Flash Point Density Viscosity @ Freeze
(42 MJ/kg min) (38oC min) (775-840 -20oC Point rack housing the measurement devices (excluding the AVL
kg/m ) (8 cSt max) (-47oC)
3
Smoke Meter). The temperature of this line was maintained at
Jet A-1 45.54 54.5 808.6 4.61 -50.3 200C to ensure that water content did not condense. A pump
GTL Synth 47.04 40.8 737.0 2.46 <-64
within the rack served to draw sample from the exhaust duct,
HRJ Synth 46.8 38.8 749.5 3.23 <-64
NRC Synth/ which was then passed through a filter to remove particulates.
43.26 46.2 802.1 3.73 N/A The sample was then distributed to the measurement devices,
Jet A-1
Canola
39.65 53.5 810 solid >0oC
each of which was calibrated prior to and between test runs. A
FAME separate sample was extracted from the exhaust duct and routed
Fish Oil
39.27 165.9 874.9 solid >0oC directly to the AVL Smoke Meter, where it was processed (via a
FAME
Veg. Oil filter paper method) to determine Filter Smoke Number (FSN)
39.49 161.9 875.9 solid >0oC values. The AVL smoke meter was deemed an acceptable
FAME
Bio Oil
39.5 174.1 876.3 solid >0oC device for the comparative measurement of FSN for fuel
FAME blends, given its inherent accuracy and repeatability of

Copyright © 2010 by ASME and Department of National

4 Defense, Government of Canada

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/29/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 measurements. A standard procedure (SAE ARP 1179) for vertical distance of one inch (25.4 mm) from the nozzle face.
turbine engine smoke measurement, also based on filter paper Beginning at the geometric center, measurements were taken in
darkening, requires the use of a probe rake, more suited to a line out to the outer peripheral of the spray at 1/8 (3.175
measurements behind a complete engine than a combustion mm) intervals. Available information from the PDA for each
sector rig. test point included droplet size and velocity distributions, as
well as volume flux.
Deposit Evaluation The OSPA system served to compare differences in spray
Prior to CCSR rig operation with each fuel formulation, a pattern for the various fuels tested. To generate this
thorough cleaning of the active liner and pertinent components information, laser light was passed through a lens which
of the combustion system was undertaken. These components generated a planar sheet. The nozzle was situated above the
included the fuel nozzle shroud, igniter and combustion liner laser sheet so that the spray axis was perpendicular to the sheet.
crossover plugs. The crossover plugs block the primary The nozzle was set up so that its centre intersected with the
combustion zone connections between the active and centre of the image obtained by the camera. The camera was
neighbouring liners in the rig. In a T56 engine, these passages positioned at an angle and received scattered light from the
are normally open, but due to the presence of burning only in droplets as they passed through the light sheet. Multiple frames
the active liner, were blocked in the rig. Incidentally, the were recorded in order to generate an average image, on which
tendency for deposits to build up on these plugs made them software corrections were performed as part of post processing
ideal for the deposit evaluation. Figure 3 show images of the [19]. Images of both the PDA and Optical Spray Pattern
cleaned combustion chamber liner, igniter, crossover plug and Analyzer systems are shown in Figure 4.
nozzle shroud. Following testing, these parts were carefully
extracted from the combustion chamber, upon which they were b)
weighed and photographed, with deposit weight ascertained by a)
a)
subtracting the final mass of the components from their initial )
cleaned mass.

Figure 4: a) Optical Spray Pattern Analyzer and b)

Phase Doppler Anemometer

a) RESULTS
)
b) Smoke Emissions
Figure 5 shows averaged values of FSN for the different
fuel blends. Each value is a composite mean of all
measurements taken. Five measurements were taken at each of
the 0, 1, 2 and 3 hour intervals. The value of FSN for the Jet A-
1 fuel is a composite of the initial and final tests done with this
baseline fuel. Confidence intervals were generated based on all
measured points to determine whether individual fuel blends
c) d) generated smoke quantities which differed from the baseline Jet
A-1 fuel with statistical significance. It is observed that none of
Figure 3: Cleaned Surfaces of the a) Combustion the FAME blends generated FSN values significantly lower or
Liner, b) Igniter, c) Crossover Plug and d) Nozzle higher than the baseline fuel. For all but the Canola Oil FAME,
Shroud the 2% concentration generated higher FSN values than the
20% blend. With Canola Oil, the values were virtually
Spray Diagnostics identical. In most cases, with the exception of the 2% Fish Oil
The open air spray characteristics of the tested fuel and 2% Vegetable Oil, FSN values were marginally lower than
formulations were investigated using a phase Doppler those from the Jet A-1, but again, without any statistical
anemometer (PDA) and an optical spray pattern analyzer significance.
(OSPA) system. The PDA system was used to determine All of the synthetic blends generated FSN values far lower
individual droplet characteristics at 9 equally spaced points at a than those for Jet A-1, with statistical significance. With the
50/50 and 100% GTL synthetic blends, the reductions from

Copyright © 2010 by ASME and Department of National

5 Defense, Government of Canada

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/29/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 baseline Jet A-1 FSN values were 65% and 83%, respectively. These HC reductions were 20% and 26%, respectively. These
The Jet A-1/CTL/GTL (NRC Synthetic) blend resulted in a modest reductions in HC were consistent with other studies,
50% reduction in smoke from baseline values. The 100% HRJ which indicated that HC, NOx and CO emissions
synthetic blend resulted in a 96% reduction in FSN, which was improvements with synthetic fuels tend to be more modest and
the single greatest improvement. The trends observed with the engine/operating condition dependent than PM reductions
synthetic blends are consistent with those observed in previous [1,8]. The measured concentrations of NOx and CO in the
studies and are again attributed to the absence of polyaromatic current study were consistent with these assertions. No
hydrocarbons in the fuels [1,2]. significant differences were observed between fuels for NOx
0.06 concentrations and some modest improvements were observed
in the CO emissions for the synthetic fuel blends, again with
0.05
Upper and Lower the greatest improvements seen with the 100% synthetic
Confidence Intervals
0.04 blends. These improvements with respect to the baseline Jet A-
1 values were 9% and 13% for the GTL and HRJ synthetic
FSN

0.03
fuels, respectively.
0.02
Deposits
0.01
Figure 7 shows post testing images of the inside of the
0 combustion liner for a)Jet A-1, b) 2% Fish Oil FAME, c) 20%
Canola Oil FAME and d) 100% GTL synthetic fuel. A similar
Jet A-1

Fish Oil 2%

Fish Oil 20%

Canola Oil 2%

Canola Oil 20%

Vegetable Oil 2%

Vegetable Oil 20%

Bio Oil 2%

Bio Oil 20%

50% Synthetic

100% Synthetic

NRC Synthetic

HRJ Synthetic

amount of deposition was observed for the Jet A-1 and

2%/20% biodiesel blends. This held true for the other FAME
blends tested which are not shown in Figure 7. The deposits in
the combustion system following testing with the 20% FAME
Figure 5: Composite AVL Smoke Number Values blends had a significant ‘oiliness’ to them. The 100% synthetic
blend produced very low deposits relative to the other fuel
Hydrocarbon Emissions blends. Figure 8 shows post testing deposits on the igniter plug
Figure 6 shows HC emissions for the fuels in a manner for the same four fuel blends featured in Figure 7. A similar
similar to Figure 5 for FSN values. The values are plotted in trend is seen, with the synthetic blend showing a very low
terms of Emissions Index (EI), with units of grams of HC relative amount of deposition and an oiliness with the 20%
generated per kg of fuel. For the FAME blends, higher values Canola Oil FAME.
of HC are observed relative to the baseline Jet A-1 in most
cases, with (marginally) statistically higher values occurring
with the Fish Oil blends and the 2% Vegetable Oil blend. The
Canola Oil Blend at 2% generated slightly less HC than the
baseline fuel, but not with statistical significance. For the
synthetic blends, statistically significant lower HC emissions
than those for neat Jet A-1 were observed only with the 100%
synthetic blends, namely the GTL and HRJ blends. a) b)
23 )
22
21
20
grams HC / kg fuel

19
18
17
16
15 c) d)
14 )
13
12
Figure 7: Combustion Chamber Deposits Following
11 Testing for: a) Jet A-1, b) 2% Fish Oil, c) 20% Canola
10 and d) 100% Synthetic Fuel
Jet A-1

Fish Oil 2%

Fish Oil 20%

Canola Oil 2%

Canola Oil 20%

Vegetable Oil 2%

Vegetable Oil 20%

Bio Oil 20%

50% Synthetic

100% Synthetic

NRC Synthetic

HRJ Synthetic

Figure 9 shows a graph of the combined (from the igniter,

nozzle shroud, crossover plugs) mass of deposits generated for
each of the fuel blends tested over the three hour rig operation.
Figure 6: Hydrocarbon Emissions The Jet A-1 test produced a net mass of deposits of 170 mg.

Copyright © 2010 by ASME and Department of National

6 Defense, Government of Canada

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/29/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 The Fish Oil and Vegetable Oil FAME fuels in a 2% such as the igniter plug in Figure 8d), which showed virtually
concentration produced an approximately equivalent combined no deposit. The Jet A-1/GTL/CTL blend generated much higher
mass of deposits as Jet A-1, while the Canola and Bio Oil deposits than the 50% GTL blend, at 127 mg. Results are not
FAME blends produced a higher amount of about 240 and 215 available for the HRJ fuel, as enough fuel sample was not
mg, respectively. All of the FAME blends at 20% had slightly acquired to run the CCSR rig for three hours.
lower than neat Jet A-1 deposition, at about 150 mg.
Spray Diagnostics
Figure 10 shows OSPA generated spray pattern images for
six of the fuels tested. The nozzle used in this analysis was the
one used in combustion testing, which had been modified from
its stock configuration by having its secondary flow circuit
blocked off. The nozzle supply pressure was 100 psi for the
spray pattern analysis.
a) b)
a) b) )
)

c) d)
c) d)
)
Figure 8: Igniter Deposits Following Testing for: a) Jet
A-1, b) 2% Fish Oil, c) 20% Canola and d) 100%
Synthetic Fuel

250

225

200
Net Deposit Mass (mg)

175

150
e) f)
125

100

75

50

25

0
JET A-1

Fish Oil 2%

Vegetable Oil 2 %

Canola Oil 2%

Bio Oil 2%

Fish Oil 20%

Vegetable Oil 20%

Canola Oil 20%

Bio Oil 20%

Synthetic 50%

Synthetic 100%

NRC Synth 50/50

Figure 9: Composite Deposit Mass for the Fuel

Blends Tested Figure 10: Spray Pattern Images for a) Jet A-1,
b) 2% Fish Oil, c) 20% Canola and d) 100% Canola,
The oiliness of deposits from these fuel blends was noted e) 50% GTL Synthetic and f) 100% GTL Synthetic
previously and may be attributed to the excessive (above
specification) gum content in these fuel blends, the presence of The first image shows the spray pattern for the baseline Jet
which may have prevented adhesion to surfaces of a greater A-1 fuel, which had a well defined hollow cone pattern. The
amount of deposit build-up. The GTL synthetic fuel blends had second image is the spray pattern for the 2% Fish Oil FAME
the best overall deposit performance, with the 50% blend and looks quite similar to the image for Jet A-1, with a slight
generating 61 mg of total deposit and the 100% blend only 8.4 amount of definition loss in the hollow cone pattern. Figure
mg. This is in agreement with the visual appearance of surfaces 10c), the image for 20% Canola Oil, shows a distinct loss of

Copyright © 2010 by ASME and Department of National

7 Defense, Government of Canada

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/29/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 definition in the spray cone structure, suggesting that at this drop size with increasing FAME concentration. All of the
concentration, there is a definitive effect on spray pattern due to synthetic blends had similar SMD values to the baseline Jet A-1
physical fuel properties. This is exaggerated further in Figure fuel at this location. At the 1.0 (25.4 mm) location, there were
9d), which shows the spray pattern for 100% Canola Oil no real observable trends between fuel types as compared to the
FAME. Although this formulation was not tested in the CCSR, baseline Jet A-1 fuel droplet sizes. In the spray diagnostics, the
it was of interest to see the fuel’s impact on spray pattern at full distinctions between fuel types were more apparent in the
concentration. The trends observed in Figures 10a) through c) optical spray patternation results (geometric distribution) than
were universal for all four FAME blends investigated as the in the PDA droplet sizing results.
concentration was increased from 2% to 100%. The final two
images are for the 50% and 100% GTL synthetic fuel. At 50% Pattern Factor
synthetic fuel concentration, there is a noticeable thickening Figure 12 shows two dimensional exhaust temperature
and reduction in definition in the outer cone of the spray as maps for a) Jet A-1, b) 20% Canola FAME and c) 100% GTL
compared to the neat Jet A-1 fuel. With the 100% synthetic synthetic fuels as produced with data collected via the exhaust
fuel, this effect is amplified even further, with the appearance thermocouple rake in the CCSR rig. The temperature resolution
of more fuel content in the center of the spray. within the charts is 50C. From a strictly qualitative
In Figure 11, PDA generated Sauter Mean Diameter perspective, little difference is observed between images for
(SMD) values are plotted for all of the fuel blends, including these three fuel blends, with no substantial deviations noted.
100% FAME fuels, for points at radial distances of 0.0, 0.75 The Pattern Factor gives a relative measure of thermal
(19.05 mm) and 1.0 (25.4 mm). SMD, or D32, is defined as the uniformity of exhaust flow and is given by the following
ratio of the sum total of volume to that of area of the measured equation:
droplets [20]. These locations correspond to the spray center T T
and two locations within the outer cone structure of the spray, PF  max mean (2)
Tmean  Tinlet
within the center and at the outer peripheral.
Lower values of Pattern Factor reflect a more thermally
80 uniform exhaust flow. Excessively high values of pattern factor
0 in 3/4 in 1.0 in
can be indicative of imminent dangers to turbine components
70
due to localized hot spots.
60
SMD (microns)

50

40

30

20

10

0
Veg Oil 2%

Veg Oil 20%

Veg Oil 100%

Canola 2%

Canola 20%

Canola 100%

Fish Oil 2%

Fish Oil 20%

Fish Oil 100%

Bio Oil 2%

Bio Oil 20%

Bio Oil 100%

Jet A-1

Synth 50%

Synth 100%

NRC Synth

Figure 11: SMD Values for the Fuel Blends at the

a) Spray Center, b) 3/4 (19.05 mm) and 1.0 (25.4
mm); Measured at 1 (25.4 mm) from Nozzle Face

At the center of the spray, the Jet A-1 fuel had an SMD
value of about 30 microns. The Vegetable Oil, Fish Oil and
Canola Oil FAME blends all showed smaller SMD values at the
spray center, with SMD values of about 23 microns for the 2%
FAME concentrations, increasing to about 27-30 microns for
the 100% FAME fuel. The bio oil FAME showed a similar
SMD at the spray center for all concentrations, at about 30
microns. The 50% and 100% GTL synthetic fuels had SMD
values under 20 microns at the spray center, while the Jet A- Figure 12: Exhaust Temperature Thermal Maps for
1/CTL/GTL blend had an SMD in the 23 micron range. At the a) Jet A-1, b) 20% Canola FAME and c) 100% GTL
0.75 (19.05 mm) location, all fuels had SMD values in the 56- Synthetic Fuels (Temperatures in Kelvin)
64 micron size range. There was a weak trend of increasing

Copyright © 2010 by ASME and Department of National

8 Defense, Government of Canada

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/29/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 Figure 13 shows calculated values of Pattern Factor for all DISCUSSION AND CONCLUSIONS
of the fuels tested. The baseline Jet A-1 had a pattern factor of In agreement with results from previous studies [1,2], it
about 0.675. All of the Fish Oil, Vegetable Oil and Canola Oil was found that the smoke (PM) emissions levels from synthetic
FAME blends had pattern factors higher than the baseline Jet gas turbine fuels were substantially lower than those generated
A-1 fuel, by between 9 and 16%. For the Vegetable and Canola with conventional jet fuels, in this case, Jet A-1. For the
Oils, the 20% blends had higher pattern factors than the 2% synthetic fuels tested, these reductions were as high as 96%, for
blends, while for the Fish Oil, the reverse was true. The Bio-Oil the HRJ Algae Oil based fuel.
had similar values of Pattern Factor to Jet A-1 for both Compared to the values produced with neat Jet A-1 fuel,
concentrations. All of the synthetic blends had slightly higher modest reductions in HC emissions were measured with neat
values of Pattern Factor than the baseline Jet A-1 fuel. To Synthetic Fuels, with reductions as high as 26%. Statistically
explore the potential relation between Pattern Factor and HC significant reductions were not found with any of the FAME
emissions, Figure 14 plots the values of Pattern Factor from blends. Most of the other FAME blends caused a slight increase
Figure 13 against composite values of HC for each of the fuels. in HC emissions, but not with statistical significance.
This was done to investigate whether there was further The addition of FAME fuel to Jet A-1 did not have a
correlation between HC emissions and any of the physical significant effect on the mass of engine deposits, although in a
nozzle effects as seen in the Spray Diagnostics section, which 20% concentration, deposits were greasy in appearance. This
might reveal themselves in the Pattern Factor. As such, no may have been caused in part by excessive gum content in
correlation was observed. these blends. It may also be attributed to likely nozzle effects,
0.8
where minor or significant numbers of fuel droplets emerging
from the fuel nozzle (likely of larger relative size, due to
0.75 density and viscosity effects), did not combust and wound up
coating combustion chamber surfaces. Some evidence of larger
0.7 droplets with 20% FAME blends was observed in the PDA
Pattern Factor

results. This effect was also noted in a previous study [14].

0.65
Engine deposits were relatively low with synthetic fuels,
0.6 especially in the neat formulation. With the GTL synthetic fuel,
significant reductions were realized with the 50/50 blend (65%)
0.55 and very significant reductions with the 100% fuel (95%). The
Jet A-1/GTL/CTL blend produced much higher deposits,
0.5
resulting in only a 25% reduction relative to the neat Jet A-1.
Jet A-1

Fish Oil 2%

Fish Oil 20%

Canola Oil 2%

Canola Oil 20%

Vegetable Oil 2%

Vegetable Oil 20%

Bio Oil 20%

50 % Synthetic

100 % Synthetic

NRC Synthetic

HRJ Synthetic

Optical Spray Patternation revealed significant differences

in spray characteristics of the fuels as they emerged from the
T56 nozzle. As FAME content was increased, the well defined
hollow cone observed with neat Jet A-1 became less distinct.
Figure 13: Pattern Factor for the Fuels Tested With synthetic blends, more droplets were apparent within the
hollow cone of the spray and a denser outer perimeter of the
25 spray was also observed.
24 Phase Doppler Anemometry revealed that synthetic blends
23 generated smaller droplets than Jet A-1 in the center region of
Hydrocarbons (g HC/kg fuel)

22 the spray. This is likely attributed to the lower density and

21 viscosity of these fuels. In the outer cone region of the spray,
20 where most of the volume flux occurs, there was little apparent
19
effect on droplet SMD as a function of fuel blend type. As
18
mentioned, an increase in FAME content translated into a slight
17
16
increase in droplet size in the core of the fuel ring emerging
15
from the fuel nozzle.
14 Pattern Factor revealed little to correlate with the effects
13 seen in the spray pattern analysis, or the HC emissions seen in
12 the combustion testing. No attempt was made to relate Pattern
0.62 0.64 0.66 0.68 0.7 0.72 0.74 0.76 0.78 0.8 Factor to PM emissions, as it was expected that the PM values
Exhaust Pattern Factor
with synthetic fuel would be significantly lower than the other
Jet A-1 based blends.
Figure 14: Pattern Factor vs HC Emissions The current study has verified reported positive impacts of
synthetic fuel utilization [1,2] as compared to conventional
petroleum derived blends in gas turbine applications with

Copyright © 2010 by ASME and Department of National

9 Defense, Government of Canada

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/29/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 respect to PM generation and to a lesser extent, gaseous 8. Moses, C., 2008, “Comparative Evaluation of Semi-
emissions such as HC. By investigating additional fuel blends Synthetic Jet Fuels”, Final Report. CRC Project No. AV-2-
containing distinct FAME formulations in differing 04a, pp. 9-29.
concentrations, a comprehensive study ensued on the relative 9. Corporan, E., Monroig, O., Wagner, M. and Dewitt, M.,
impacts of purely chemical vs physical properties of fuels with 2004, “Influence of Fuel Chemical Composition on
regards to their impacts on exhaust emissions. It was ultimately Particulate Matter Emissions of a Turbine Engine”,
ascertained that differences in spray pattern between fuel GT2004-54335, Proceedings of ASME Turbo Expo 2004,
blends were generally neutralized during combustion in the Vienna, Austria, pp.1-8.
liner and not reflected in the Pattern Factor and 10. CGSB 3.24., 2008 “Aviation Turbine Fuel (military
emission/smoke/deposit measurements. Thus it was concluded Grades F-34 and F-44)”, Fuel Specification.
that some chemical properties of synthetic fuels, likely related 11. Goodger, E., 1994, Jet Fuel Supply and Quality, Landfall
primarily to aromatic content, render these fuels superior to Jet Press, pp. 7-13.
A-1 in terms of emissions performance, while the chemical 12. Gardiner, D., LaViolette, M., Pucher, G. and Bennet, L.,
properties of the FAME blends may render some benefits, yet 2007, “Biodiesel Fuel and Emission Testing on Military
ultimately cause performance similar to conventional Jet A-1 Equipment – Phase II”, Mechanical Engineering Report #
fuel. 070401, Royal Military College of Canada, pp. 25-40.
13. EPA, 2002, “A Comprehensive Analysis of Biodiesel
REFERENCES Impacts on Exhaust Emissions, EPA Technical Report,
1. Corporan, E., DeWitt, M., Belovich, V., Pawlik, R., Lynch, EPA420-_02-01”, United States Environmental Protection
A., Gord, J., Meyer, T., 2007, “Emissions Characteristics Agency, pp. 2-30
of a Turbine Engine and Research Combustor Burning a 14. Corporan, E., Reich, R., Monroig, O, DeWitt, M., Larson,
Fischer –Tropsch Jet Fuel”, Energy and Fuels 2007, pp. V., Aulich, T., Mann, M. and Seames, W., 2005. “Impacts
2615-2626. of Biodiesel on Pollutant Emissions of a JP-8 Fueled
2. Dewitt, M.J., Corporan, E., Graham, J. and Minus, D., Turbine Engine”, Journal of the Air and Waste
2008, “Effects of Aromatic Type and Concentration in Management Association, Technical Paper, pp. 1-3.
Fischer-Tropsch Fuel on Emissions Production and 15. Bishop, K., 2009, “Effects of Fuel Nozzle Condition Upon
Material Compatibility”, Energy and Fuels 2008, 22, pp. Gas Turbine Combustion Chamber Exit Temperature
2411-2418. Profiles”, Master’s Thesis, Department of Mechanical and
3. Speight, J.G., 2008, Synthetic Fuels Handbook – Aerospace Engineering, Carleton University, Ottawa,
Properties, Process and Performance, McGraw Hill Ontario, Canada., pp. 10-23.
U.S.A., Chapter 1, 7. 16. Kotzer, C.E., 2008, “Effects of Combustion Chamber
4. Moses, C. and Roets, P., 2009, “Properties, Characteristics Geometry Upon Exit Temperature Profiles”, M.A.Sc.
and Combustion Performance of Sasol Fully Synthetic Thesis. Royal Military College of Canada, April, pp. 12-
Fuel”, J. Engineering for Gas Turbines and Power, Vol. 46.
131, pp. 1-5. 17. Pucher G. and W.D. Allan, 2008, “Emissions from
5. Hermann, F., Klingmann, J., Gabrielsson, R., Joergen, P., Thermal Stability Additives, Phase 3”, Internal Report,
Olsson, J. and Owrand, F., 2006, “Chemical Analysis of RMC Mechanical Engineering Report 080602, pp. 3-42.
Combustion Products from a High Pressure Gas Turbine 18. Goodger, E., 2000, Transport Fuels and Technology-
Combustor Rig Fueled by Jet A-1 and a Fischer-Tropsch- Mobility for the Millenium, Landfall Press, pp. 60-90.
Based Fuel”, GT2006-90600, Proceedings of ASME Turbo 19. Underhill, R., Shafique, H., Fiset, P. and Prociw, L.A.
Expo 2006, May 8-11, Barcelona, Spain, pp. 1-3. 2006, “Optical Patternation for Quality Control of Gas
6. Kahandawala, M., Dewitt, M., Corporan, E. and Sidhu, S., Turbine Fuel Injectors”, ILASS Americas, 19th Annual
2008, “Ignition and Emission Characteristics of Surrogate Conference on Liquid Atomization and Spray Systems,
and Practical Jet Fuels”, Energy and Fuels 2008, 22, pp. Toronto, pp. 2-3.
3673-3679. 20. Lefebvre, A.H., 1999, Gas Turbine Combustion, Taylor
7. Hileman, J., Wong, H., Oritz, D., Brown N., Maurice, L. and Francis Group, pp. 206-208.
and Rumizen, M., 2008, “The Feasibility and Potential
Environmental Benefits of Alternative Fuels for
Commercial Aviation”, Proceedings from the 26th Intl.
Congress of the Aeronautical Sciences, pp. 5-8.

Copyright © 2010 by ASME and Department of National

10 Defense, Government of Canada

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 01/29/2016 Terms of Use: http://www.asme.org/about-asme/terms-of-use