Proceedings of ASME Turbo Expo 2017: Turbomachinery Technical Conference and Exposition

GT2017
June 26-30, 2017, Charlotte, NC, USA

GT2017-64181

COMBUSTION BEHAVIOR OF JET A DROPLETS AND ITS BLENDS WITH

BUTANOL

Álvaro Muelas, Pilar Remacha, Adrián Martínez, Javier Ballester

Laboratory for Research on Fluid Dynamics and Combustion Technologies (LIFTEC)

University of Zaragoza / CSIC
María de Luna 10, 50018 Zaragoza, Spain
E-mail address: ballester@unizar.es (J. Ballester)

ABSTRACT INTRODUCTION

In light of the potential of butanol as an alternative fuel for
blending with petroleum fuels such as gasoline, diesel or Jet A,
experimental data regarding the burning characteristics of these
blends are required in order to better understand their
combustion process. In this study, freely-falling droplets of
butanol, Jet A, and their mixtures (10, 20 and 50% butanol by
volume) were examined in a combustion chamber which
provides representative conditions of real flames, both in terms
of temperature and oxygen availability. The combustion
characteristics reported here include evolution of droplet sizes,
burning rates, soot measurements, and the occurrence of
microexplosions and soot shells. Results show that the
evolution of droplet diameter for butanol, Jet A and their blends
are very similar, regardless of the obvious compositional
differences. Sooting behaviors are found to be quite different,
with a clear reduction in the sooting propensity as the butanol
content in the fuel increases. These results are consistent with a
previous study in a gas turbine showing similar performance
among Jet A and its blends with butanol, suggesting that such
mixtures are promising alternative fuels with very similar
combustion characteristics to Jet A, but with much less
propensity to soot. Moreover, this study provides new results
on the combustion properties of Jet A/butanol blends, for which
very scarce data exist in the open literature.
[Keywords: Droplet combustion, Butanol, Jet A, Soot]
Initiatives to replace fossil fuels by fuels derived from
biofeedstocks (i.e. biofuels) are recently gaining importance
due to environmental and sustainability reasons. Even though
fossil fuels are likely to remain dominating transportation and
energy systems in the near future [1, 2], renewable biofuels
have the potential to reduce their consumption, and therefore
their environmental footprint, especially through usage as
additives to conventional fuels. In contrast to fossil fuels, the
biomass used to produce biofuel is renewable, carbon-neutral
and domestically available [3, 4]. Furthermore, biofuels are
reported to have negligible sulfur content and to inhibit PAH
and soot formation [4, 5].
The most widely used biofuel today is by far bio-ethanol
[2, 5], which is commonly added to gasoline for its use in
spark-ignition engines. Ethanol production via fermentation of
plants and starches was among the very first developments in
biofuels because of the relatively low cost of production due to
the existence of a well-established alcohol industry and
infrastructure [1]. Although ethanol production from edible
crops poses serious concerns, such as competition with the food
industry or indirect land-use change, it remains nowadays as
the major source for bio-ethanol feedstock [2]. These concerns
over the so-called first generation biofuels have motivated
efforts to find new feedstocks which do not compete with the
food industry, such as lignocellulosic biomass or residual waste
[1, 2]. Biofuels derived from such feedstocks are called second
generation biofuels, and although ethanol can also be produced
from cellulosic biomass [6, 7], there has been an important

1 Copyright © 2017 ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/10/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 research on potential new fuel molecules which can overcome
the significant drawbacks associated with ethanol properties.
One of the most promising candidates is butanol, which has
several well-known advantages when compared with ethanol
[1, 2, 8-15]:
- Higher energy density.
- Lower propensity for water absorption.
- Higher miscibility with hydrocarbons.
- Boiling point closer to the gasoline/diesel fuel range:
lesser impact on the fuel distillation curve.
- While its octane number is similar to that of gasoline,
its cetane number is high enough to make blending
with diesel for compression-ignition engines viable.
- Better material compatibility: corrosion towards
ferrous metals and elastomers swelling are reported to
be much less severe than in the case of ethanol.
- Lower vapor pressure, resulting in safer manipulation
and less volatile organic compounds emission.
- Lower heat of vaporization, which leads to less
ignition problems.
- Current engines can run on more enriched butanol
blends without any modification.
These features have motivated a recent interest in bio-
butanol, which can be both produced by fermentation of edible
crops (first generation biofuel) or through advanced
technologies using lignocellulosic biomass or algae as
feedstock (second generation biofuel) [1, 8, 9]. Even though
there is currently much less butanol production capacity
compared to ethanol, several companies are developing
technologies for bio-butanol synthesis and equipping existing
ethanol plants with butanol production capabilities through
cost-effective retrofitting [8, 11].
In recent years a comparatively large number of studies
have analyzed the performance of both, pure butanol
(particularly the isomer 1-butanol) and its blends with gasoline
and diesel in spark-ignition and compression-ignition engines
(e.g. [16-20]). Much less work has been performed regarding
its possible use in gas turbine engines, being a recent study by
Mendez et al. [11] the only reference found in the open
literature. In this work, the performance of pure 1-butanol and
its blends with Jet A (25, 50 and 75% by volume) in a 30 kW
gas turbine was studied, concluding that mixtures of butanol
with Jet A are promising alternative fuels with similar
performance to that of Jet A, but with less CO and NOx
emissions. It is important to note that this potential of butanol
as fuel for gas turbines is mostly related to its use in stationary
gas turbines for power generation, being its use as fuel for
aviation much less straightforward. Dziegielewski et al. [15]
reviewed some important properties of butanol and Jet A blends
with regard to their possible use in aviation, concluding that
practically any addition of butanol to Jet A results in a flash
point under the minimum dictated by the regulation, while
blends above 10% butanol in volume can present conductivities
and calorific values out of specification too. The aviation sector
2 Copyright © 2017 ASME
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/10/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use
is far more restrictive than any other regarding the evaluation
of new alternative fuels, both for safety and logistical reasons.
Moreover, its main research direction is the development of
fully interchangeable, 100% hydrocarbon fuels such as Fischer-
Tropsch Synthetic Paraffinic Kerosene (FT-SPK) or
Hydroprocessed Esters and Fatty Acids (HEFA) [15, 21, 22].
However, the situation changes significantly when it comes to
gas turbines for energy generation, where restrictions for
alternative fuels are much lower and quite diverse bio-fuels
have been successfully tested in experimental gas turbines:
biodiesel [23, 24], ethanol [25, 26], wood-derived
lignocellulosic alcohols [27] or the already cited butanol [11]
are some examples which illustrate the flexibility of stationary
gas turbines.
Successful use of a new fuel in complex applications
requires a substantial knowledge of its combustion
characteristics. Even though engine studies provide very useful
information regarding fuel performance under realistic
conditions, results can be dependent on several non-
controllable variables (e.g. fuel spray patterns, temperature
profiles inside the combustion chamber, etc.), and therefore
results can depend to some extent on the experimental facility
used. On the contrary, in the much more simplified single
droplet configuration all the parameters affecting the results are
completely known and controllable, and therefore the
combustion characteristics observed are intrinsically
attributable to the fuel. Given the increasing interest in butanol,
recent studies have examined the combustion characteristics of
butanol droplets, both pure [13, 14, 28-30] and blended with
different fuels such as gasoline [12] or soybean oil [10]. In
particular, butanol/Jet A blends have been proposed as a
promising option for gas turbine engines [11]. The detailed
characterization of the combustion behavior of these blends
seems most desirable but, to the authors’ knowledge, has not
been addressed in previous studies. The main objective of this
work is therefore to provide new experimental data on the
combustion and sooting characteristics of unsupported, free-
falling droplets of butanol/Jet A blends.
EXPERIMENTAL
Experiments were conducted in the Droplet Combustion
Facility (DCF) developed at LIFTEC and shown in Fig. 1.
Details of the facility can be found in a previous work [31], and
therefore only the most important features of the experimental
method will be explained here.
Droplet combustion facility
Fuel droplets were generated in the DCF by means of a
piezoelectric device with initial diameters (D0) ~ 150 microns
(150 ± 0.7 µm for all the cases studied), which is considered to
be a good compromise between actual sizes in practical
applications and a good accuracy of the experimental results.
Droplet spacing always exceeded 120 droplet diameters, so that
 interactions between droplets were negligible. As droplet
generation stability is perhaps the most critical variable in these
experiments, initial droplet size was thoroughly checked by
means of repeated size measurements at a given location (2 mm
after injection), at least at the beginning and end of each test;
Standard deviation calculated from repeated tests was found to
be as small as 0.74 µm (rms value), indicating negligible drifts
in D0 during the running of the experiments.
Fig. 1. Schematic of the droplet combustion facility.
The monosized droplets were injected along the axis of a
quartz tube, coaxially with the combustion products generated
by a flat-flame premixed methane-air burner (McKenna). This
coflow provided a controllable and realistic environment to
study the evolution of fuel droplets under representative
conditions of real flames, both in terms of oxygen availability
and temperature. As in real flames a droplet might be subjected
to different oxygen conditions, the droplet combustion process
was studied both at 3 and 10% of oxygen (by volume, dry
basis) in the coflow. These two oxygen concentrations are
believed to be good representatives of the range of conditions
seen by burning droplets in real flames. Additionally to
combustion, the vaporization process (i.e. with 0% oxygen)
was also studied. The axial temperature profile along the tube
centerline was measured with a fine wire thermocouple (50 µm,
type S). The temperatures recorded in the first 30 mm were
within the interval 1684-1600 K for the 0% O2 condition and
1618-1560 K for the 3% O2 (these intervals are the measured
variations along the axis due to heat losses). The actual
temperatures were not measured for the 10% O2 case, but are
expected to be slightly lower due to the increased excess air.
Different imaging methods were used to record visual
information describing the various aspects of droplet
combustion. Size and velocity evolution of droplets were
3 Copyright © 2017 ASME
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/10/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use
recorded through a black and white (BW) high sensitivity CCD
camera (QImaging Retiga SRV, 12-bit Mono) fitted with a long
distance microscope and backlighted via LED strobe. This
optical system was programmed to record two sequential shots
of a droplet in the same frame with a time lapse of 500 µs
between them. As it can be seen in Fig. 2a, this imaging method
not only gives information about the droplet size and motion,
but also about the existence of a soot shell for certain
conditions. A color camera (Teledyne DALSA Genie HC1024,
8-bit) perpendicularly placed to the BW camera and faintly
backlighted was used to record the flame images. The faint
backlight was provided in order to make the droplet
distinguishable from the self-illuminated flame, capturing in the
same image the free-falling droplet surrounded by its flame, as
shown in Fig. 2b.
a) b)
Fig. 2. Images taken by both cameras for the same droplet (Jet A, 10%
O2, 17 mm after injection): a) Double-exposure BW picture showing
the falling droplet surrounded by its soot shell; b) Color picture
displaying the droplet, the soot shell and a faint flame.
The practically spherical shapes of the soot shells
displayed in Fig. 2 are thought to confirm that convective
effects are very weak and do not appear to disturb the spherical
symmetry. The soot tail observed in the wake of the droplets is
ascribed to the shedding of some soot agglomerates, which are
dragged away due to the slip velocity of the gas. Therefore, the
droplets can be assumed to burn under conditions of spherical
symmetry, which facilitates the comparison with one-
dimensional droplet combustion models.
Pictures taken by both imaging methods were post-
processed in order to extract the relevant droplet combustion
features in the most precise and repeatable way. As the BW
images aimed to gather close-up information of the droplet (i.e.
droplet diameter and velocity, soot shell morphology and
microexplosions occurrence), their spatial resolution was fixed
at 1.4 µm/pixel, suitable to determine with good accuracy the
size of droplets in the range 25-150 µm. For every picture, a
background image was taken and subtracted prior to the image
analysis. Droplet diameter and velocity were automatically
calculated by means of an edge detection algorithm. Color
pictures had a spatial resolution of 6 µm/pixel and were
processed to gather information regarding flame size and
morphology. A background image was subtracted prior to the
manual measurement of the flame diameter. However, for the
 fuels and conditions examined in this study, the flames
registered were so weak that the involved uncertainties are
considered too high to extract reliable quantitative data.
In addition to the optical setup, a soot sampling technique
was also applied for quantitative measurement of the amount of
soot generated with the different fuels and blends. The soot and
gases surrounding the droplet were laterally suctioned,
quenched and cooled by a nitrogen flow. The aspirated soot
was retained on a quartz filter, which was weighed after drying
for 24 hours in a furnace at 150 °C to determine the amount of
soot present at the probe position. A soot index was calculated
in terms of the weight of soot per unit of mass of injected fuel.
Fuels
Commercial Jet A (obtained from a local airport) and 1-
butanol (99.4% purity) were used for the tests. Additionally,
three mixtures of these fuels were prepared with 10, 20 and
50% butanol by volume (B10, B20 and B50 respectively).
The main fuel properties for both pure fuels are shown in
Table 1. Butanol properties are extracted from the literature,
whereas a specific analysis of the Jet A sample was performed
at the Instituto de Carboquímica (ICB-CSIC) in order to obtain
its composition and most significant properties.
droplet diameter value at injection (D0). Even though Jet A is a
highly multi-component fuel, results displayed in Fig. 3 are in
good agreement with the ‘d2 law’, and after a short heat-up
period, the normalized square diameter decreases linearly with
time with a nearly constant slope (the burning rate K). As
oxygen concentration increases, so does the burning rate,
accelerating substantially the droplet evaporation process. This
is ascribed to the higher temperature in the shell flame around
the droplet, leading to an enhanced heat transfer to the droplet.
With the pure Jet A curves as a baseline for comparison,
tests were performed for neat butanol and their mixtures (B10,
B20 and B50). These fuels were examined for the three oxygen
conditions considered in the study: 0% (Fig. 4), 3% (Fig. 5)
and 10% (Fig. 6).

Table 1. Selective properties of the fuels investigated

Jet A 1-butanol a
Molecular formula C10.60 H19.63 C4 H10 O
Molecular Weight (g/mol) 146.83 74.12
% C (mass) 86.62 64.86
% H (mass) 13.38 13.51
% O (mass) - 21.62 Fig. 3. Evolution of normalized droplet size for Jet A at three
Boiling point (°C) 99.1 – 285.8 117.6 different oxygen conditions: 0, 3 and 10% O2 in the coflow.
Lower Heating Value (MJ/kg) 42.60 33.10
Density at 20 °C (kg/m3) 791 810
Viscosity at 20 °C (cP) 0.91 2.54
Latent Heat of Vaporization (kJ/kg) 330 b 584
Adiabatic flame temperature (K) 2360c 2233
a
From ref.[12, 14, 32, 33]
b
From ref. [11]
c
From ref. [34]

RESULTS AND DISCUSSION

Droplet size evolution and burning rate

As stated above, tests were performed for three different
oxygen conditions: 0, 3 and 10% O2 in the coflow. The first one
corresponds to a pure evaporation process, whereas the last two
are believed to be representative conditions of oxygen
availability around droplets in real flames. Fig. 3 shows the Fig. 4. Evolution of normalized droplet size for Jet A, B10, B20,
evolution of Jet A droplet sizes for these three conditions. B50 and pure Butanol with 0% O2 in the coflow.
Based on the quasi-steady theory of droplet burning,
measurements are displayed as normalized square diameter
versus normalized time. Normalization is performed with the

4 Copyright © 2017 ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/10/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use


Fig. 5. Evolution of normalized droplet size for Jet A, B10, B20,
B50 and pure Butanol with 3% O2 in the coflow.
Fig. 6. Evolution of normalized droplet size for Jet A, B10, B20, B50
and pure Butanol with 10% O2 in the coflow.
For each oxygen condition, the observed behaviors are
found to be quite similar between fuels, and all of them display
very close burning rates and droplet consumption times,
particularly among Jet A and its blends. Butanol exhibits the
most distinctive behavior, with a shorter heat-up period and a
smaller burning rate. The lower boiling point of butanol
compared to most of the Jet A constituents (97.7% vol. of the
examined Jet A boils above the butanol boiling temperature)
can be responsible for the shorter heat-up period, whereas its
higher latent heat of vaporization could account for the smaller
burning rate. The Jet A–butanol blends, and particularly B10
and B20, display an almost indistinguishable behavior from
pure Jet A throughout all the vaporization process.. As
butanol’s boiling point is located at the beginning of the Jet A
distillation curve and both fuels exhibit quite close burning
rates, this fact can either imply that there is no preferential
5 Copyright © 2017 ASME
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/10/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use
vaporization, or that the similarity between the pure fuels’
behavior is masking it.
Comparing the vaporization curves for the three blends, it
is noted that although the 3 and 10% O2 curves practically
overlap (Fig. 5 and 6), they show slight but noticeable
differences when exposed to the zero oxygen condition (Fig.
4). This may be caused by several effects. One of them could
be related to the differences in the luminosity of the flame
traces observed between the different cases, the 0% O2
condition leading to a much brighter streak than the other two.
This may be an indication of a higher radiation heat transfer
towards the droplets for Jet A tests with lower oxygen
concentration, causing the differential behavior observed in
Fig. 4. The differences could also be related with the fact that
no envelop flame is possible for 0% O2, whereas there is a
spherical flame when oxygen is available; given the important
effects of the flame on radial species and temperature profiles
and, hence, on mass and heat transfer rates, this may explain a
distinct behavior for 0% O2. In any case, further work would be
needed to confirm or discard the actual influence of these or
other effects on the observed behaviors.
In order to gain insight into the slight differences found
between fuels, their burning rates were quantified for the 3%
O2 condition as the instantaneous slopes of the curves
displayed in Fig. 5: K=-d(D2/dt). Since differentiation greatly
amplifies any small experimental uncertainty in D2 curves, a
three-point centered moving average was performed for
smoothing purposes. The resulting curves are displayed in Fig.
7, where the evolution of K with normalized time is shown.
Even though the slopes in Fig. 5 seemed practically constant,
the calculated burning rates plotted in Fig. 7 display clearly
unsteady behaviors. Whereas pure Jet A, B10 and B20 exhibit
very similar burning rates throughout all the droplet lifetime,
B50 and neat butanol show a clearly distinct behavior, with
higher burning rates in the very first stages of droplet
combustion, and lower values thereafter. Additionally, as time
increases the burning rates of B50 and butanol start to diverge,
with B50 approaching to the Jet A, B10 and B20 curves. These
results suggest a preferential vaporization of butanol in the first
stages for the B50 case. On the contrary, differences between
pure Jet A, B10 and B20 are so small that no assertions can be
made in this regard. The slight decay in K for neat butanol at
late stages of droplet lifetime could be ascribed to different
factors such as flame extinction or the decrease of axial
temperature along the quartz tube. As Jet A is a highly
multicomponent fuel, its burning rate variations can be also
dependent on fuel composition evolution, complicating
therefore its analysis.

Fig. 7. Evolution of burning rate K (mm2/s) with normalized time for
Jet A, B10, B20, B50 and pure Butanol with 3% O2 in the coflow.
Three-point centered moving average was performed for smoothing
purposes.
Fig. 8. Global burning rates for all the fuels and oxygen conditions
studied. The selected quasi-steady period for least-squares line fitting
was (D/D0)2 from 0.8 to 0.1.
In spite of the observed unsteadiness in droplet burning
rates, their temporal variations during the quasi-steady period
can be considered to be small enough to extract a global
burning rate for each fuel and condition. These global burning
rates are obtained by fitting the quasi-steady segments of the
D2-t curves to lines by least-squares best-fit. The quasi-steady
segment was arbitrarily defined as the interval from
(D/D0)2=0.8 to (D/D0)2=0.1, excluding therefore initial transient
heating and potential flame extinction. In doing so, a
representative global burning rate was extracted for every fuel
and oxygen condition, allowing an easier comparison among
cases in Fig. 8. It seems clear that an increase in oxygen
availability leads to higher burning rates, as it was stated above.
Between fuels, pure butanol displays the lowest burning rate
for all the conditions studied, whereas the extracted K for neat
6 Copyright © 2017 ASME
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/10/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Jet A, B10 and B20 can be considered to be alike. B50 shows
burning rates slightly closer to Jet A than to butanol, maybe as
a result of the fitting interval choice, which does not include the
first stages of droplet combustion, where the burning rate of
B50 is closer to that of butanol.
Finally, it is noteworthy that the evolution curves obtained
for pure Jet A and butanol for the 10% O2 condition (Fig. 6)
show a very similar qualitative and quantitative behavior (in
particular, burning rate) to the curves obtained in previous
studies under quite different conditions ([12] for butanol and
[35] for Jet A). These studies were conducted under
microgravity conditions in order to maintain a strict one-
dimensional droplet flame configuration, with 570 µm droplets
subjected to a 21% O2 and room temperature (spark-ignited
droplets). In spite of the obvious differences between
experiments, the droplet burning rate does not depend
theoretically on its initial diameter, and the shortest flame-
droplet distance for the 21% O2 condition might compensate
for the much lower ambient temperature. Therefore, and based
on the remarkable similarities between curves for butanol and
Jet A, it would be plausible that both experimental conditions
shared more coincidences than it could be a priori expected.
This can be also considered as an additional proof of the
negligible role of the convective effects in these tests.
Soot shell evolution and microexplosion occurrence
Although the droplet size evolutions and burning rates
presented in Figs.4-8 are found to be quite similar for all the
fuels studied, the sooting behaviors differ considerably, with a
lower propensity to form soot as the butanol content in the
blend increases. This can be clearly seen in Fig. 9, where a
selection of BW images displaying the evolution of the soot
shell for the 10% O2 condition are shown. These soot shells are
usually achieved in totally convection-free environments
obtained through microgravity (e.g. [12-14, 35]). In these
configurations the soot aggregates formed during combustion
remain trapped at the radial location where thermophoresis and
Stefan flux drag forces balance [36]. Soot shells occurrence is
much rarer for free falling droplets, as any small relative
velocity between the droplet and the surrounding gases would
sweep the soot away.
The experimental conditions used in this work provide
very low relative velocities between the droplets and the
coflow, allowing the formation of an almost perfectly spherical
soot shell surrounding the droplets. Even so, as it can be seen
in Fig. 9, some soot aggregates drifted away from the soot shell
due to a small droplet slip velocity, perhaps also with some
contribution of buoyancy-induced convection. Conclusions on
the evolution and the amount of soot visible around droplets
should be therefore taken with caution. However, as droplet
velocities are found to be very close among fuels (Jet A and
butanol have similar densities), the authors consider that a
qualitative comparison between Jet A and their different
mixtures’ sooting propensity can be drawn from their soot shell
evolution.
 L = 13 mm L = 15 mm L = 17 mm L = 19 mm L = 21 mm L = 23 mm L = 25 mm

Jet A

B10

B20

B50

Butanol

Fig. 9. Evolution of droplet and soot shell for the examined fuels (cropped double-exposure BW images). For a better
comparison, pictures are arranged according to their approximate length after injection (L).

Fig. 9 clearly shows that, as the butanol content in the the surrounding soot cloud. At that point, the droplet diameter
blend increases, the formation of soot shells is delayed, with is estimated to be less than 10 µm. Somewhere between L=24.4
the mixture B50 not displaying any soot shell until practically and 24.7 mm, the microexplosion occurs, with the result as
reaching the droplet depletion length. Likewise, the soot shell presented in the last picture of the sequence, where the soot
registered at a certain location after injection becomes agglomerates are dispersed, and their oxidation gives the
progressively thinner as the mixture is enriched in butanol. The luminosity macroscopically observable as a tiny spark. These
soot shell becomes thicker as the droplet combustion proceeds, last two pictures of Fig. 10 display the same droplet (by means
and near the end of the droplet burning the soot particles of the double-exposure method), with a time lapse between
agglomerate into a self-supported crust which completely shots of 500 µs.
encloses the droplet, as typically happens in convection-free
droplet combustions experiments such as [12]. Droplet
diameter cannot be measured in the presence of this soot and,
for this reason, some curves in Fig. 6 lack a few points at the
end of the droplets lifetimes. Droplet depletion occurs with a
weak microexplosion, macroscopically visible as a tiny spark at
the end of the flame trace. This feature has been reported in
other free-falling droplet combustion works (e.g. [37]). Here, it
was possible to observe those microexplosions in the close-up
BW images due to the surrounding soot shells, which act as a
‘marker’. Microexplosion sequences have been captured for Jet
A, B10 and B50, as displayed in Fig. 9.
A full droplet extinction sequence for the mixture B50 is
presented in Fig. 10, where the faint soot shell surrounding the Fig. 10. Extinction sequence of B50 droplets with 10% O2 in
the coflow.
droplet appears to shrink, becoming progressively thicker as
the droplet diameter rapidly diminishes. 24.4 mm after
injection, the droplet becomes almost indistinguishable from

7 Copyright © 2017 ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/10/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use

 Visual appearance and analysis of the flame traces
As the droplet stream falls along the quartz tube, its
combustion generates a macroscopically visible flame streak.
This streak can provide valuable information regarding the
amount of soot present in the flame region, as incandescent
soot gives a characteristic yellow luminosity. Therefore, long
exposure images of the flame traces were taken with a DSLR
camera for all the fuels and oxygen conditions studied. Fig. 11
shows the significant role played by oxygen availability in the
amount of soot present in the combustion chamber: for the
same fuel (B50) diminishing the oxygen in the coflow increases
substantially the propensity to soot. Analogous behaviors were
found for the other blends and for pure Jet A. By contrast,
butanol does not exhibit any yellowish luminosity for any
oxygen condition, including the 0% O2.
Fig. 11. Long exposure photographs for 150 µm droplets of B50 at
different oxygen conditions in the coflow.
As all the flame traces were captured with fixed camera
settings, a direct comparison of flame brightness can provide a
soot propensity indicator for each fuel and condition. Fig. 12
shows an analysis of the flame traces displayed in Fig. 11 (B50
at 0, 3 and 10% O2). The axial profiles of luminosity shown in
Fig. 12 were calculated by adding the values at all pixels across
a given cross section of the flame trace. Although the
quantitative value of this result is unclear (e.g., due to eventual
differences in soot temperature), the luminosity curves can be
interpreted in terms of the differences in the amount of soot for
the various fuels at fixed experimental conditions. The vertical
lines indicate the droplet consumption lengths for B50 at each
oxygen condition (L=31.0, 28.5 and 24.7 mm for 0, 3 and 10%
O2 respectively). It is noteworthy that droplet depletion length
only corresponds with flame trace extinction for the 10% O2
condition. This is due to the relatively high oxygen availability
in the coflow, which enhances soot oxidation. In the 3% O2
condition, the oxidation of the already generated soot is much
slower, and therefore the flame trace reaches slightly longer
axial distances than the droplets themselves. The limit case is
the pure evaporation condition (0% O2), where the absence of
oxygen in the coflow prevents soot consumption, and all the
produced soot remains in the combustion chamber. In this case,
8 Copyright © 2017 ASME
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/10/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use
the droplet depletion length corresponds with the maximum
luminosity of the flame trace. After that, the luminosity
decreases because of the progressively lowering axial
temperature, which greatly affects the blackbody emission from
incandescent soot. Caution must be taken therefore with this
image-based soot estimation method, as the luminosity not only
depends on the amount of soot, but also on its temperature
(which can significantly vary among conditions). Analogous
results were obtained for the rest of fuels.
Fig. 12. Axial variation of the flame traces’ luminosity for 150
µm droplets of B50 at different oxygen conditions in the coflow.
The vertical dotted lines indicate the droplet consumption
lengths for each condition.
Once the effect of oxygen availability is ascertained, the
flame traces captured for different fuels for the 3% O2
condition are compared in Fig. 13. As stated above, pure
butanol traces are completely non-sooty, and for the exposure
time fixed in these images (200 ms) its trace does not appear
visible in the photograph. By increasing the exposure time to a
few seconds, a faint blue trace corresponding to the actual
flame boundary appears, but with much less intensity than the
yellowish sooty traces shown for the other fuels. Comparing
the flame streaks of Jet A and its blends, it is noted that the
addition of butanol clearly decreases the yellowish soot
luminosity, especially for the B50 case. The other two blends
present a more similar appearance to pure Jet A, although with
a slightly more orangish and less intense traces. Since soot
formation occurs in the fuel side of the envelop diffusion flame,
a thin blue flame layer should be observable surrounding the
yellowish sooty streak. But, as it was observed for the tests
with pure butanol, this blue luminosity due to
chemiluminescence presents much less intensity than the black-
body emission from soot particles, and therefore it remains
eclipsed by this last one.

Fig. 13. Long exposure photographs for 150 µm droplets of different
fuels at 3% O2.
In order to obtain a single parameter to characterize the total
luminosity of each flame trace, the profiles displayed in Fig. 12
were integrated along the axial coordinate from L=0 until the
corresponding length of droplet depletion. This integration
interval was chosen to avoid the influence of the trace length,
more related to soot consumption rate than to its concentration.
The results are shown in Fig. 14 and, again, might be
interpreted in terms of the relative sooting propensity of
different fuels. The addition of butanol appears to noticeably
reduce the luminosity of Jet A flame traces for all the oxygen
conditions. For instance, for the 3% O2 flame traces displayed
in Fig. 13, B10 shows a 7% luminosity reduction compared to
pure Jet A, B20 a 12% and B50 a 46%.
Fig. 14. Integrated flame luminosities calculated for all the
fuels and oxygen conditions studied. Integration is performed
from injection point until droplet consumption length.
9 Copyright © 2017 ASME
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/10/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use
Soot measurements
Both the BW pictures of the soot shells and the analysis of
the flame traces point to a clear soot reduction when blending
Jet A with butanol. In order to corroborate this reduction, a soot
sampling method was applied for quantitatively measuring the
amount of soot generated by each fuel. For that purpose, a
probe aspirated the soot present in the combustion chamber for
different axial lengths, determining thereby an axial profile of
the amount of soot present in the flame traces. Due to the
relatively small amount of soot generated by the very small fuel
flow rate used in the tests, the coflow conditions were fixed at
0% O2, maximizing in this way soot production (as clearly
displayed in Fig. 14). As expected, butanol gave absolutely no
soot even for the zero oxygen condition, and its quartz filter
ended the test completely white. The results obtained for the
rest of fuels are plotted in Fig. 15. As the blend becomes richer
in butanol, the amount of soot collected at a certain axial
position decreases, with a qualitative difference between the
pure Jet A and the mixtures. It is also noted that the B50 profile
appears to be slightly delayed, with its maximum situated at 45
mm whereas the highest soot collection for the rest of fuels was
located ~35 mm after injection. Note that these soot peaks
occur shortly after droplet depletion (approximately at L=30
mm), with almost no soot collected during the droplet lifetime,
probably due to the cooling effect caused by the metallic probe.
As at 0% O2 the droplet does not have a surrounding flame,
soot production strongly depends on the coflow temperature,
and therefore the introduction of a soot probe near the droplets
can inhibit soot generation. On the contrary, when the probe is
located downstream of the soot generation area (i.e.
downstream of the droplet vaporization zone), it does not affect
soot production, and consequently it collects the soot formed
upstream. Anyhow, these results do not intend to describe the
soot formation process, but rather to compare the propensity to
soot of the examined fuels when subjected to the same
conditions. Due to the strong axial variations in local soot
quantities, the comparison among different cases is better done
in terms of a single parameter, representative of the amount of
soot collected over the whole trace. With that purpose, the soot
profiles were integrated over the first 60 mm, giving the results
displayed in Fig. 16. This integration length was arbitrarily
selected as an attempt to include the soot formation zones but
excluding downstream regions where no more soot is formed
and changes in measured concentration are actually due to
other artifacts (e.g., radial dispersion of soot particles). As
already suggested by the soot profiles, pure Jet A yields a much
higher soot index compared to any of the blends. Among
blends, the collected soot is further decreased with butanol
content.

Fig. 15. Soot index axial variation for Jet A, B10, B20 and B50 with
0% O2 in the coflow.
Fig. 16. Integrated soot index for Jet A, B10, B20, B50 and butanol
with 0% O2 in the coflow.
CONCLUSIONS
The droplet combustion characteristics of butanol, Jet A,
and their mixtures (10, 20 and 50% butanol in volume) were
studied under experimental conditions which are thought to be
representative of temperature and oxygen availability in real
flames. At the same time, relative velocities between the
droplets and the surrounding gases were minimized, so that
droplets burnt in a practically 1-D configuration, not disturbed
by convective effects, as proven by the formation of almost
perfectly spherical soot shells. As a consequence, the obtained
experimental results may be compared with theoretical one-
dimensional droplet combustion models. In spite of the highly
multicomponent characteristics of the studied fuels (with the
obvious exception of pure butanol), the droplet vaporization
curves are in good agreement with the d2-law, with quite
constant burning rates after the initial heat-up period. B10 and
B20 show practically indistinguishable burning rates from neat
Jet A, whereas pure butanol displays a slightly different
10
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/10/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use
behavior, probably due to its higher latent heat of evaporation.
B50 shows an intermediate behavior between butanol and Jet
A, presumably with some preferential vaporization of the
former in earlier stages of droplet combustion. The most
significant combustion characteristic of butanol-Jet A mixtures
was found to be their lower soot formation, with a clear
reduction in the soot propensity as the mixture is enriched in
butanol. This was consistently demonstrated through close-up
images of the soot shells, the lower luminosity of the visible
flame traces, and the quantitative results of a soot probe.
The strong similarities in combustion rates between pure
Jet A and its mixtures with butanol, in addition to the lower
propensity to soot of the latter support the potential of butanol
as an attractive additive to Jet A.
ACKNOWLEDGMENTS
This work was supported by the Spanish Ministry of
Education through the pre-doctoral grant FPU15/01866 and the
Ministry of Science and Innovation through the research
projects CSD2010-00011 and ENE2016-76436-R. The analysis
of the Jet A sample was kindly provided by Drs. R. Murillo and
M. Callen, from the Instituto de Carboquímica (ICB-CSIC).
The authors are also grateful to Dr. Álvaro Sobrino for his
assistance with data capture and analysis, and to Luis Ojeda for
his support in the experimental tasks. The work of Mario
Angeloni in the development of the Droplet Combustion
Facility, without which this study would not have been
possible, is also gratefully acknowledged.
REFERENCES
[1] Bergthorson J.M., Thomson M.J., 2015, "A Review of the
Combustion and Emissions Properties of Advanced
Transportation Biofuels and Their Impact on Existing and
Future Engines". Renewable and Sustainable Energy Reviews;
42, pp. 1393-1417.
[2] Sarathy S.M., Oßwald P., Hansen N., Kohse-Höinghaus K.,
2014, "Alcohol Combustion Chemistry". Progress in Energy
and Combustion Science; 44, pp. 40-102.
[3] Demirbas A., 2007, "Progress and Recent Trends in
Biofuels". Progress in energy and combustion science; 33(1),
pp. 1-18.
[4] Saxena R., Adhikari D., Goyal H., 2009, "Biomass-Based
Energy Fuel through Biochemical Routes: A Review".
Renewable and Sustainable Energy Reviews; 13(1), pp. 167-
178.
[5] Tang C., Wei L., Man X., Zhang J., Huang Z., Law C.K.,
2013, "High Temperature Ignition Delay Times of C5 Primary
Alcohols". Combustion and Flame; 160(3), pp. 520-529.
[6] Naik S.N., Goud V.V., Rout P.K., Dalai A.K., 2010,
"Production of First and Second Generation Biofuels: A
Copyright © 2017 ASME
 Comprehensive Review". Renewable and Sustainable Energy
Reviews; 14(2), pp. 578-597.
[7] Nigam P.S., Singh A., 2011, "Production of Liquid Biofuels
from Renewable Resources". Progress in energy and
combustion science; 37(1), pp. 52-68.
[8] Jin C., Yao M., Liu H., Chia-fon F.L., Ji J., 2011, "Progress
in the Production and Application of N-Butanol as a Biofuel".
Renewable and Sustainable Energy Reviews; 15(8), pp. 4080-
4106.
[9] Bruno T.J., Wolk A., Naydich A., 2009, "Composition-
Explicit Distillation Curves for Mixtures of Gasoline with
Four-Carbon Alcohols (Butanols)". Energy & Fuels; 23(4), pp.
2295-2306.
[10] Hoxie A., Schoo R., Braden J., 2014, "Microexplosive
Combustion Behavior of Blended Soybean Oil and Butanol
Droplets". Fuel; 120, pp. 22-29.
[11] Mendez C., Parthasarathy R., Gollahalli S., 2014,
"Performance and Emission Characteristics of Butanol/Jet a
Blends in a Gas Turbine Engine". Applied Energy; 118, pp.
135-140.
[12] Xu Y., Avedisian C.T., 2015, "Combustion of N-Butanol,
Gasoline, and N-Butanol/Gasoline Mixture Droplets". Energy
& Fuels; 29(5), pp. 3467-3475.
[13] Liu Y.C., Alam F.E., Xu Y., Dryer F.L., Avedisian C.T.,
Farouk T.I., 2016, "Combustion Characteristics of Butanol
Isomers in Multiphase Droplet Configurations". Combustion
and Flame; 169, pp. 216-228.
[14] Alam F.E., Liu Y., Avedisian C., Dryer F., Farouk T., 2015,
"N-Butanol Droplet Combustion: Numerical Modeling and
Reduced Gravity Experiments". Proceedings of the
Combustion Institute; 35(2), pp. 1693-1700.
[15] Dzięgielewski W., Gawron W., Kaźmierczak U.,
Kulczycki A., 2014, "Butanol/Biobutanol as a Component of an
Aviation and Diesel Fuel". Journal of KONES; 21(2), pp. 69--
75.
[16] Gu X., Huang Z., Cai J., Gong J., Wu X., Lee C.-f., 2012,
"Emission Characteristics of a Spark-Ignition Engine Fuelled
with Gasoline-N-Butanol Blends in Combination with EGR".
Fuel; 93, pp. 611-617.
[17] Galloni E., Fontana G., Staccone S., Scala F., 2016,
"Performance Analyses of a Spark-Ignition Engine Firing with
Gasoline–Butanol Blends at Partial Load Operation". Energy
Conversion and Management; 110, pp. 319-326.
[18] Elfasakhany A., 2016, "Experimental Study of Dual N-
Butanol and Iso-Butanol Additives on Spark-Ignition Engine
Performance and Emissions". Fuel; 163, pp. 166-174.
[19] Zhang T., Jacobson L., Björkholtz C., Munch K., Denbratt
I., 2016, "Effect of Using Butanol and Octanol Isomers on
Engine Performance of Steady State and Cold Start Ability in
Different Types of Diesel Engines". Fuel; 184, pp. 708-717.
[20] Merola S., Tornatore C., Iannuzzi S., Marchitto L.,
Valentino G., 2014, "Combustion Process Investigation in a
High Speed Diesel Engine Fuelled with N-Butanol Diesel
Blend by Conventional Methods and Optical Diagnostics".
Renewable Energy; 64, pp. 225-237.
11
Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/10/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use
[21] Blakey S., Rye L., Wilson C.W., 2011, "Aviation Gas
Turbine Alternative Fuels: A Review". Proceedings of the
Combustion Institute; 33(2), pp. 2863-2885.
[22] Winchester N., Malina R., Staples M.D., Barrett S.R.,
2015, "The Impact of Advanced Biofuels on Aviation
Emissions and Operations in the Us". Energy Economics; 49,
pp. 482-491.
[23] Habib Z., Parthasarathy R., Gollahalli S., 2010,
"Performance and Emission Characteristics of Biofuel in a
Small-Scale Gas Turbine Engine". Applied Energy; 87(5), pp.
1701-1709.
[24] Bolszo C., McDonell V., 2009, "Emissions Optimization
of a Biodiesel Fired Gas Turbine". Proceedings of the
Combustion Institute; 32(2), pp. 2949-2956.
[25] Moliere M., Vierling M., Aboujaib M., Patil P., Eranki A.,
Campbell A., et al., editors. Gas Turbines in Alternative Fuel
Applications: Bio-Ethanol Field Test. ASME Turbo Expo 2009:
Power for Land, Sea, and Air; 2009: American Society of
Mechanical Engineers.
[26] Patra J., Ghose P., Datta A., Das M., Ganguly R., Sen S., et
al., 2015, "Studies of Combustion Characteristics of Kerosene
Ethanol Blends in an Axi-Symmetric Combustor". Fuel; 144,
pp. 205-213.
[27] Seljak T., Oprešnik S.R., Kunaver M., Katrašnik T., 2012,
"Wood, Liquefied in Polyhydroxy Alcohols as a Fuel for Gas
Turbines". Applied energy; 99, pp. 40-49.
[28] Bartle K., Fitzpatrick E., Jones J., Kubacki M., Plant R.,
Pourkashanian M., et al., 2011, "The Combustion of Droplets
of Liquid Fuels and Biomass Particles". Fuel; 90(3), pp. 1113-
1119.
[29] Jangi M., Sakurai S., Ogami Y., Kobayashi H., 2009, "On
the Validity of Quasi-Steady Assumption in Transient Droplet
Combustion". Combustion and Flame; 156(1), pp. 99-105.
[30] Ogami Y., Sakurai S., Hasegawa S., Jangi M., Nakamura
H., Yoshinaga K., et al., 2009, "Microgravity Experiments of
Single Droplet Combustion in Oscillatory Flow at Elevated
Pressure". Proceedings of the Combustion Institute; 32(2), pp.
2171-2178.
[31] Angeloni M., Remacha P., Martínez A., Ballester J., 2016,
"Experimental Investigation of the Combustion of Crude
Glycerol Droplets". Fuel; 184, pp. 889-895.
[32] National Center for Biotechnology Information. Pubchem
Compound Database [cited 2016 Nov. 16]. Available from:
https://pubchem.ncbi.nlm.nih.gov/compound/263
[33] National Institute of Standards and Technology
[cited 2016 Nov. 16]. Available from:
http://webbook.nist.gov/cgi/cbook.cgi?ID=C71363&Mask=4.
[34] Riebl S., Braun-Unkhoff M., Riedel U., editors. A Study
on the Emissions of Alternative Aviation Fuels. ASME Turbo
Expo 2016: Turbomachinery Technical Conference and
Exposition; 2016: American Society of Mechanical Engineers.
[35] Liu Y.C., Savas A.J., Avedisian C.T., 2013, "The
Spherically Symmetric Droplet Burning Characteristics of Jet-
A and Biofuels Derived from Camelina and Tallow". Fuel; 108,
pp. 824-832.
Copyright © 2017 ASME
 [36] Manzello S.L., Yozgatligil A., Choi M.Y., 2004, "An
Experimental Investigation of Sootshell Formation in
Microgravity Droplet Combustion". International journal of
heat and mass transfer; 47(24), pp. 5381-5385.
[37] Li T., Zhu D., Akafuah N., Saito K., Law C., 2011,
"Synthesis, Droplet Combustion, and Sooting Characteristics of
Biodiesel Produced from Waste Vegetable Oils". Proceedings
of the Combustion Institute; 33(2), pp. 2039-2046.

12 Copyright © 2017 ASME

Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 02/10/2018 Terms of Use: http://www.asme.org/about-asme/terms-of-use