Chemical Engineering Research and Design 1 4 6 ( 2 0 1 9 ) 60–70

Contents lists available at ScienceDirect

Chemical Engineering Research and Design

journal homepage: www.elsevier.com/locate/cherd

Process simulation and techno-economic analysis

of bio-jet fuel and green diesel production —
Minimum selling prices

Elias Martinez-Hernandez ∗ , Luis Felipe Ramírez-Verduzco,

Myriam A. Amezcua-Allieri, Jorge Aburto ∗
Biomass Conversion Division, Mexican Institute of Petroleum, Mexico City, 07730, Mexico

a r t i c l e i n f o a b s t r a c t

Article history: The simulation and techno-economic analysis for the production of bio-jet fuel and green
Received 19 February 2019 diesel from hydrotreatment of vegetable oil is presented in this paper. Monte Carlo simula-
Received in revised form 25 March tions are carried out to find the minimum selling price that allow the processes coping with
2019 uncertainty in capital investment, feedstock and product prices. The green diesel production
Accepted 28 March 2019 involved hydrodeoxygenation (HDO) of vegetable oil, producing propane as a co-product.
Available online 5 April 2019 The bio-jet fuel process involved HDO and isomerisation/hydrocracking to yield bio-jet,
green diesel, naphtha, and propane. Based on a proposed reaction scheme and model, the
Keywords: reaction conversions for the isomerisation/hydrocracking reactor are estimated consider-
Bio-jet fuel ing constraints in observed bio-jet fuel composition and freezing point. The simulation and
®
Green diesel analyses were carried out in SuperPro Designer , linked to Excel Visual Basic for Applica-
Process simulation tions to perform Montecarlo simulations and obtain robust estimations of minimum selling
Techno-economic analysis price (MSP). The results showed that a 75 thousand barrels/y bio-jet fuel plant can be prof-
HEFA itable, but a MSP of 1.35 US$/L is required to lower the risk of failure due to uncertainty.
In the case of the green diesel, for a 63 thousand barrels/y production plant, the required
MSP is 1 US$/L. Bio-jet fuel would need incentives or subsidies while the green diesel can be
competitive with current fossil diesel prices. Therefore, under current price trends, green
diesel production would be a more viable investment. However, bio-jet fuel production ben-
efits from multi-fuel production and if large plant sizes and lower feedstock prices can be
obtained, the risks can be mitigated.
© 2019 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.

1. Introduction and climate change is relatively small, other transport sectors are con-
tinuously reducing carbon and energy intensities, and the aviation
The aviation sector is one of the fastest growing sources of greenhouse transport sector is not the exception. Furthermore, at the Paris climate
gas (GHG) emissions and is responsible for about 2% of anthropogenic conference (COP21) in December 2015, 195 countries adopted the first-
GHG emissions (IATA, 2018). By 2020, global international aviation emis- ever universal, legally binding global climate deal. The agreement sets
sions are projected to be around 70% higher than in 2005 and the out a global action plan to put the world on track to avoid dangerous
International Civil Aviation Organization (ICAO) forecasts that by 2050 climate change by limiting global warming to well below 2 ◦ C. With a
they could grow by a further 300–700% (ICAO, 2017). Direct emissions carbon intensity between 30 to 110 g-C per passenger-km, air transport
from aviation account for about 3% of the EU’s total GHG emissions emissions are comparable to car transport and much higher than the
(EU Commission, 2017). In Mexico, aviation contributed to 5% of total 20 g-C per passenger-km from bus transport (Penner et al., 1999). To
transport sector emissions. While the contribution to GHG emissions mitigate such emissions, the International Air Transport Association

∗
Corresponding authors.
E-mail addresses: emartinez@imp.mx (E. Martinez-Hernandez), jaburto@imp.mx (J. Aburto).
https://doi.org/10.1016/j.cherd.2019.03.042
0263-8762/© 2019 Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
 Chemical Engineering Research and Design 1 4 6 ( 2 0 1 9 ) 60–70 61

(IATA) seeks to reduce the carbon footprint by promoting the use of analysis (Yao et al. (2017). A study based on a reaction scheme derived
a more Sustainable Alternative Fuel (SAF) instead of the conventional from experimental data for process simulation, evaluating the mini-
fossil jet fuel (IATA, 2013). mum selling prices and assessing the risk due to data uncertainty in
Alternative jet fuels are drop-in fuels produced from renewable key economic parameters are needed to better understand the eco-
sources such as vegetable oils and biomass (Wang and Tao, 2016). nomics of bio-jet or green diesel and establish informed policies to
Several process technologies have been developed for this purpose support commercial production. In order to fill this gap, this work
including Fischer-Tropsch (FT) process (Hanaoka et al., 2015), alcohol presents a comprehensive process simulation and techno-economic
and sugar to jet fuel process (ATJ and STJ) (Yao et al., 2017), hydropro- study developed for the production of bio-jet and green diesel fuels
cessing of esters and fatty acids (HEFA) (Starck et al., 2016), and others using a HEFA process under a Mexican context. The present study pro-
are under development (Fu et al., 2015; Wu et al., 2017; Yang et al., poses a new simulation approach employing a reaction scheme derived
2018; Li et al., 2018a). These processes produce hydrocarbons forming from observed results for the hydrodeoxygenation and the isomerisa-
a jet fuel type of mixture called biojet fuel or bio-synthetic paraffinic tion/hydrocracking reactors using proprietary catalysts developed at
kerosene (bio-SPK). However, the specifications for aviation fuels are the Mexican Institute of Petroleum (Rocha et al., 2019). Furthermore,
more stringent than for other sectors due to the wider range of oper- the proposed simulation approach allows getting detailed hydrocarbon
ating conditions during flights. Specially, the freezing point is a critical composition of intermediate and final streams, adjusting conversion
property that must be met by alternative jet fuels in order to be certi- parameters according to experimental conditions or to different oil
fied through the standard ASTM D7566-17a (ASTM, 2017), for their use feedstock, or target reaction conversions that allow the final product
in commercial flights. to comply with the ASTM freezing point specification. The simula-
®
The HEFA process is one of the accepted production processes for tion was implemented in Super Pro Designer in order to perform a
aviation fuel, with the UOP/Eni process being the most widely known techno-economic analysis. Sensitivity and uncertainty analysis using
McCall et al., 2019. The HEFA process comprises two catalytic reactor Monte Carlo simulation was carried out by linking this software with
stages and require hydrogen as an additional feedstock. In the first a spreadsheet through Excel Visual Basic for Applications. This part
stage, the oil feedstock is fed to the hydrodeoxygenation reactor for of the work also allows targeting the minimum selling price so that
oxygen removal, and this produces linear chain hydrocarbons with the economics of the process can cope with the uncertainty of key
high freezing point as well as propane. The former intermediate prod- parameters, thus leading to a more robust estimation. The simulation
uct is fed to an isomerisation/hydrocracking reactor where the long approach and techno-economic analysis method are presented in the
linear chain hydrocarbons are cracked and isomerised to produce jet next section.
fuel range hydrocarbons with a freezing point suitable for blending.
The coproducts are naphtha, bio-jet fuel, diesel and a gaseous product
containing light hydrocarbons such as propane, which are separated 2. Process simulation and techno-economic
by common distillation columns. Depending on market conditions, the analysis methods
main product can be green diesel for which the last separation stage
is avoided. Another option is to use the intermediate product from the The bio-jet fuel production process was simulated in the
hydrodeoxygenation reactor as a diesel blend stock. ®
SuperPro Designer software as shown in Fig. 1. The flowsheet
One of the major constraints for the widespread use of bio-jet
consists of three main sections: (1) hydrodeoxygenation (with
fuel is the high production costs, while open information about the
gas and water separation), (2) isomerisation/hydrocracking,
details of the process using renewable oil feedstocks is scarce. Some
and (3) product separation. Based on their fatty acid profile,
efforts in optimising the reaction conditions for aviation fuel pro-
duction from Jatropha oil have been made (Anand et al., 2016). To
the oil feedstocks were simulated as a mixture made up of
identify optimum conditions, it is also important to study possible triacylglycerides. Stoichiometric reaction schemes were for-
reaction networks of triacylglycerides and their respective fatty acids mulated as presented for each reaction stage in the following
from a thermodynamic perspective (Valencia et al., 2018). Simulation sections of the paper. The specified bio-jet fuel process sim-
studies using Camelina, Brassica carinata and waste cooking oil as feed- ulation parameters are presented in Table 1. In the case of
stocks have also been carried out (Chu et al., 2017); however, they the green diesel hydrodeoxygenation process, the green diesel
only focused on the hydrodeoxygenation stage, neglecting the isomeri- obtained is not isomerised and thus only Section 1 and the
sation/hydrocracking stage. A techno-economic review of bio-jet fuel Section 3 of product separation involved.
production, showed that the gate price using a HEFA process ranged
between 1 and 1.16 US$/L for a production capacity of 2.4 million
barrels per year (Pearlson et al., 2013). Another simulation and techno- 2.1. Hydrodeoxygenation reaction section
economic study of large scale plants for bio-jet fuel production from
Camelina oil in Canada, found a production cost of between 0.4 and
The first section consists of a reactor to remove the oxygen
1.71 US$/L for a production capacity of 1.4 million barrels per year
from tryacylglycerides by reacting with hydrogen to produce a
(Li et al., 2018b). New configurations such as a biorefinery combining
mixture of linear hydrocarbons, water, carbon monoxide and
hydrogen production from renewable electricity and the HEFA process
can have environmental benefits; however, techno-economic results
carbon dioxide. The reactions taking place are the following:
showed that provisioning hydrogen in this configuration adds 26–34% saturation, hydrocracking, depropanation, hydrodeoxygena-
to production costs (Zech et al., 2018). Other researchers explored the tion (DO), decarboxylation (DCx) and decarbonylation (DC).
integration of jet fuel production in sugarcane biorefineries employing Based on the model mixture used to represent the oil
a HEFA, FT or ATJ process with on-site hydrogen production, show- feedstock, a simplified reaction scheme for the hydrodeoxy-
ing that HEFA provides the highest jet fuel production potential (Klein genation reactor was formulated as shown in Eqs. (1)–(4).
et al., 2018). The minimum selling price for profitability have been esti- Depropanation
mated for ATJ and FT processes (Atsonios et al., 2015) and the pyrolysis
and HDO process (Yang et al., 2018), while studies on the green diesel
production from HDO (Zech et al., 2018) and FT process (Im-orb and Cn H(2n-6m-1) (COOH)3 + 3H2 → 3Cn H2n+1 COOH + C3 H8 (1)
Arpornwichanop, 2016) have also been reported.
These studies set a reference for techno-economics of various pro-
cess configurations, however they tend to use a fixed capacity or Hydrodeoxygenation reaction (DO)
process yield or conversions mainly based on data or assumptions from
literature, with less consideration to the reaction scheme or fuel prop-
erties. Of the studies reviewed, only one did a stochastic uncertainty Cn H2n+1 COOH + (m+3)H2 → Cn+1 H2n+4 + 2H2 O (2)
62 Chemical Engineering Research and Design 1 4 6 ( 2 0 1 9 ) 60–70

Fig. 1 – Simulation flowsheet of the Bio-jet fuel production process in the SuperPro Designer® software.

Table 1 – Oil composition and simulation parameters for biojet fuel and green diesel production.
Oil composition
Component % weight
Tripalmitin 42.61
Triolein 41.9
Trilinolein 10.41
Tristearin 4.24
Trimyristin 0.84
Hydrodeoxygenation reaction section Isomerisation section
H2 feed 0.033 kg/kg oil H2 feed 0.019 kg/kg feed
Reactor temperature 310 ◦ C Reactor temperature 250 ◦ C
Reactor pressure 40 bar Reactor pressure 30 bar
Reaction set Conversion Reaction conversion (see Table 2)
Depropanation 100%
Hydrodeoxygenation 60.1% Production separation section
Decarboxylation 36.1% Flash temperature 100 ◦ C
Decarbonylation 3.8% Flash pressure 25 bar
Cooling temperature 100 ◦ C Naphtha column pressure 1.013 bar
Recovery in 3-phase separator 99% Naphtha column light key Isononane, 51.62% in distillate
Water separation 100% Naphtha column heavy key Decane, 5.44% in distillate
Bio-jet column pressure 1.013 bar
Bio-jet column light key Iso-pentadecane, 99.56% in distillate
Bio-jet column heavy key Heptadecane, 7.96% in distillate

Decarboxylation reaction (DCx) CO2 are measured, the percentage of each reaction type can
be calculated from the following set of equations.
Cn H2n+1 COOH + mH2 → Cn H2n+2 + CO2 (3)

m DO
 even = =A (5)
Decarbonylation reaction (DC) modd DC + DCx

Cn H(2n+1-2m) COOH + (m+1)H2 → Cn H2n+2 + CO + H2 O (4) mCO2 DCx

= =B (6)
mCO DC
In these equations, n is the number of carbons in the
aliphatic chain and m is the number of double bonds in the DO + DCx + DC = 1 (7)
fatty acid molecule. Therefore, the reaction scheme implicitly
considers the hydrogen saturation and hydrocracking steps. where meven is the mass of hydrocarbons with even carbon
It can be observed that the hydrodeoxygenation reaction number, modd is the mass of hydrocarbons with odd carbon
produces hydrocarbons with even number of carbons while number, while mCO2 and mCO is the mass of CO2 and CO pro-
the decarboxylation and decarbonylation reactions produce duced, respectively. DO, DCx and DC are the percentage or
hydrocarbons with odd number of carbons. These allowed extent of the reaction taking place and contributing to the
calculating the contribution of each reaction to the product product distribution which can be found by solving the system
distribution of the HDO reactor, based on experimental data. of Eqs. (5)–(7). A similar approach can be followed if a detailed
For example, if the mass of produced hydrocarbons, CO and analysis of the hydrocarbon product is not available but the
 Chemical Engineering Research and Design 1 4 6 ( 2 0 1 9 ) 60–70 63

total amount of water produced (mH2O ) is known. In that case, Cn H2n+2 + H2 → Cn -7 H2n -12 + C7 H16 forn = 17, 18 (11)
DO, DCx and DC can be found by replacing Eq. (5) with Eq. (8)
and solving the equation system (Eqs. (6)–(8)). Cn H2n+2 + H2 → 2Cn/2 Hn/2+2 forn = 16, 18 (12)

DO mH2O
=   −1=A (8) The remaining set comprises the isomerisation reactions,
DC mCO 18.015
28.01 represented as the conversion of linear alkane to iso-alkane:

Eq. (8) is derived from the stoichiometry of the DC reaction

Cn H2n+2 → iso-Cn H2n+2 forn = 6–18 (13)
which produces 1 mol of water per mol of CO. The water from
DO reaction is the difference between the total amount pro-
duced and the one from DC reaction. The scheme was developed according to the literature
The simulation consists of a feed of oil and hydrogen which proposes a typically symmetric cracking product
which is preheated and fed to the hydrodeoxygenation reac- distribution and that the major isomerisation product is
tor which operates at 310 ◦ C, 40 bar and 1 h−1 of liquid hourly mono-branched cracking products for C8+ hydrocarbons
space velocity (LHSV). The process uses a proprietary cat- (Thybaut and Marin, 2016; Weitkamp et al., 1983; Coonradt and
alyst developed at the Mexican Institute of Petroleum for Garwood, 1964; Flinn et al., 1960).
hydrodesulphurisation of diesel but adapted for hydrotreat- Based on the sets of possible reactions in Eqs. (9)–(12), a
ment of vegetable oils Rocha et al. (2019). A set of 16 reactions non-linear programming problem was formulated to find the
was specified by setting the reaction conversion according to reaction conversions that minimise the sum of the squared
the results from micro pilot plant trials for the degrees of DO, differences between calculated and experimental product
DCx and DC. The conversion of reaction in Eq. (1) was set yields as follows.
at 100%, while the hydrodeoxygenation reaction was 60.1%, Objective function
decarboxylation 36.1% and decarbonylation was 3.8% based
 2
on the corresponding fatty acid. The reactor outlet is a mix- minimise : Yexp,i − Ycalc,i (14)
ture of linear saturated hydrocarbons ranging from C13 to C18 i
(for palm oil feedstock), and gases including mainly water, CO,
CO2 and propane. The HDO reactor outlet is used to preheat the where Yexp,i is the experimental value measured for the mass
reactor feed in a heat exchanger, gets cooled down and then fraction of hydrocarbon i in the isomerisation reactor effluent,
passes through a three-phase separator removing the gases Ycalc,i is the calculated value for the mass fraction of hydro-
and most of the water. A dewatering unit is used to remove carbon i in the isomerisation reactor effluent according to the
the remained water from the mixture of paraffins before being conversion rates for each of the reactions included in the reac-
processed in the isomerisation section. tion scheme.
Subject to the following constraints:
2.2. Isomerisation/hydrocracking section Hydrocracking/isomerisation reactor mass balance:

In order to produce a suitable jet fuel range fraction of hydro-  

mHT + mH2r − mproducts
carbons, with the low freezing point required to comply with < 0.001 (15)
mHT + mH2r
the ASTM standards, an isomerisation reactor using hydrogen
is required. In this reactor, hydrocracking and isomerisation
where mHT is the mass of the hydrotreated stream fed to the
reactions of the hydrocarbons coming from the HDO section
reactor (e.g. kg/h),
take place in the presence of a proprietary catalyst of IMP.
mH2r is the stoichiometric mass of hydrogen for the reac-
Although reaction mechanisms and kinetics of these com-
tions (e.g. kg/h),
plex reaction system have been studied for crude oil derived
mproducts is the mass of products from the reactor (e.g. kg/h).
fractions, studies for mixtures of renewable hydrocarbons are
Bio-jet yield:
rather scarce. Given a definite composition of hydrocarbons
The bio-jet fuel yield (Y biojet ) was constrained for a range
from vegetable oil hydrotreatment fed to the isomerisation
between 12.8% for lower bound (LB) and 49.4% for upper bound
reaction, a model reaction scheme can be derived to fit an
(UB), according to the literature (Pearlson et al., 2013):
expected product distribution and yields.
A set of possible reactions taking place was developed
LB UB
as follows. The production of light hydrocarbons from the Ybiojet ≤ Y biojet ≤ Ybiojet (16)
isomerisation reactor was represented by hydrocracking reac-
tions to form propane from linear hydrocarbons: Based on a given experimental yield and a measured com-
position of the distilled bio-jet fuel fraction, the mass flows
Cn H2n+2 +H2 → Cn -3 H2n -4 +C3 H8 forn= 13–18 (9) of the hydrocarbons making up the bio-jet fuel are then con-
strained as:
Eq. (9) produces C10–C15 hydrocarbons. The production of
exp
C7–C9 has been represented by hydrocracking reactions of the mBiojet,i ≥ mBiojet,i (17)
hydrocarbons present in higher proportions (C15–C18). These
comprise sets of reactions leading to C8 (Eq. (10)) and C7 (Eq. where mBiojet,i is the mass of bio-jet fuel component i calculated
(11)), and another set with hydrocarbons splitting at the cen- from the reaction scheme (e.g. kg/h)
tral bond (Eq. (12)). exp
mBiojet,i is the value of the mass of bio-jet fuel component
i calculated from the experimental values of bio-jet fuel com-
Cn H2n+2 + H2 → Cn -8 H2n -14 + C8 H18 forn = 17, 18 (10) position and yield.
64 Chemical Engineering Research and Design 1 4 6 ( 2 0 1 9 ) 60–70

Component mass balance: acted hydrogen). The gas stream is treated to remove CO2 and
⎛ ⎞ other impurities, and then hydrogen and propane are sepa-

J

J rated. Hydrogen is recycled back to the reactors and propane
mi0 − vi,j mi xj − mi ⎝1 − xj ⎠ = 0 (18) is sold as product. The liquid stream resulting after the flash
j j tank, is sent to a distillation train where naphtha is obtained as
distillate in the first column, and the heavy fraction is sent to a
where mi0 is the initial mass of component i that is fed to the second column where bio-jet fuel is obtained as distillate and
isomerisation reactor (e.g. kg/h), diesel as bottoms. The distillation columns where designed
vi,j is the mass stoichiometric coefficient of the balanced using a short-cut model procedure in Super Pro Designer. The
chemical equation of component i in reaction j, model required specification of relative volatility with respect
mi is the mass of component i in the reaction mixture (mi0 to a heavy key component. The relative volatility was calcu-
plus any production or minus consumption, e.g. kg/h) lated from the boiling point and vaporisation heat (reported
xj is the conversion. It is assumed that the hydrocracking in the appendix Table A1) using the following equation:
reactions occur in parallel while the isomerisation reactions
AB 1 1

occur after hydrocracking. ˛A/B = exp − (23)
Total conversion to hydrocracking reactions R TbA TbB

where ˛A/B is the relative volatility of component A with


J

0≤ UB
xj ≤ XHcrackj for j = hydrocracking reactions (19) respect to heavy component B,
j AB is the average between vaporisation heats of A and B,
R is the universal gas constant,
Isomerisation reaction conversion TbA and TbB are the boiling point temperature (in K) of A
and B, respectively.

J
LB UB
Xisomj ≤ xj ≤ Xisomj for j = isomerisation reactions (20) 2.4. Techno-economic analysis
j

The techno-economic analysis was performed using the eco-

The lower and upper limits were set as 0 and 0.9 for both
nomic analysis feature of SuperPro Designer® . The revenue
hydrocracking and isomerisation reactions. When the yields
streams were the propane, naphtha, diesel, and bio-jet fuel
and compositions are not know, an additional constraint can
in the case of the bio-jet fuel process, and only diesel and
be used to determine optimum composition to achieve a spec-
propane in the case of the green diesel process. The capi-
ified freezing point:
tal costs were determined by the built-in models in SuperPro
Designer® . The main parameters used for economic analysis
FPBiojet ≤ FPASTM (21)
are presented in Table 2, the software default values were used
for other parameters, such as depreciation, installation and
where FPBiojet is the freezing point of bio-jet fuel and FPASTM is
working capital factors. The following analyses were carried
the specification value from the ASTM standard. The freez-
out:
ing point of the jet fuel fraction (C8–C16) should be lower
than the ASTM standard of −47 ◦ C (ASTM, 2017). Therefore,
a) Sensitivity to oil feedstock price. Simulations were run for
the approach may also help to set selectivity targets for cata-
sensitivity analysis of IRR to oil feedstock price and produc-
lyst development such that a bio-jet fuel suitable for blending
tion capacity.
and use in commercial flights is obtained. The freezing point
b) Minimum Selling Price. Simulations at oil market price
of the bio-jet fuel hydrocarbon mixture was calculated using
varying the plant capacity to determine the minimum sell-
the following mixing rule (Riazi, 2005):
ing price for bio-jet fuel that allows obtaining an IRR of
10%.

C16
FPBiojet = yi (FP)i (22) c) Uncertainty analysis. Monte Carlo simulations were car-
ried out for a given production capacity and oil price for a
i=C8
profitable plant.
where yi is the molar fraction of component i, and (FP)i is the
freezing point of component i. The freezing points of the linear 3. Results
hydrocarbons and their isomers are presented in the Appendix
Table A1. 3.1. Simulation results
The solution to the problem resulted in a set of reactions
and their corresponding conversions were specified in Super The product distribution and yields after each reactor are
Pro Designer to simulate the isomerisation/hydrocracking presented in Table 3. The results from the hydrodeoxygena-
reactor. The reactor was simulated at 250 ◦ C and 30 bar and tion reactor was used to estimate the mass conversion of
uses a proprietary isomerisation catalyst developed by IMP. the reactions in the isomerisation reactor by following the
approach described in Section 2.2. The values after conver-
2.3. Product separation section sion are presented in Table 4. Results show that more than
50% of the C14–C18 hydrocarbons are hydrocracked to shorter
The isomerisation reactor effluent is used to preheat the reac- chain hydrocarbon products. The isomerisation reactions tend
tor feed in a heat exchanger where it is cooled down to to reach the upper bound of 90% for C13 and shorter chain
155 ◦ C. The effluent is further cooled down to 98 ◦ C to flash components. Results also show that hydrocracking of C14–C17
it and remove the gas products (mainly propane and unre- to propane and C8–14 hydrocarbons are key to obtain cer-
 Chemical Engineering Research and Design 1 4 6 ( 2 0 1 9 ) 60–70 65

Table 2 – Parameters for the techno-economic analysis of the bio-jet fuel process.
Parameter Value

General
Annual operating time 7920 h
Dollar exchange rate 1 US$ = 19 Mexico pesos
Interest rate 10%
Project lifetime 15 years
Equipment cost as per year 2016

Feedstock, utility and labour costs

Palm oil cost 0.289 - 0.405 US$/kg (for sensitivity) 0.6 US$/kg (average market price)
Hydrogen cost 2 US$/kg
Electricity cost 0.07 US$/kWh
Cooling water cost 0.05 US$/ton
Steam cost 12 US$/ton
Wastewater treatment cost 0.85 US$/ton
Labour rate 15 US$/h
Product selling prices
Propane price 0.8 US$/kg
Naphtha price 0.55 US$/kg
Bio-jet fuel price 0.62 US$/kg
Green diesel price 1 US$/kg (isomerised product, bio-jet fuel case)0.8 (non-isomerised product, green diesel case)

Table 3 – Results of product distribution and yields from the simulation of the bio-jet fuel production process.
Component Hydrodeoxygenation reactor Isomerisation reactor Bio-jet fuel product

Outlet mass Mass yield Outlet mass Mass yield Molar Error
composition percentage (oil composition percentage (oil composition (simulation-
percentage feed basis) percentage feed basis) percentage experimental)/
(simulation) experimental

CO2 5.5 5.6 – – – –

CO 0.4 0.4 – – – –
Propane (C3) 5.1 5.2 7.5 6.3 – –
C7 – – 2.1 1.8 0 –
i-C7 – – 0 0 0 –
C8 – – 1.3 1.1 0.2 1.43%
i-C8 – – 11.7 9.8 0.2 1.55%
C9 – – 1.2 1.0 3.4 0.015%
i-C9 – – 9 7.5 12.2 0.007%
C10 – – 0.9 0.7 2.1 0.037%
i-C10 – – 7.9 6.6 20.3 0.000%
C11 – – 0.8 0.6 1.5 0.16%
i-C11 – – 6.9 5.8 9.9 0.007%
C12 – – 0.7 0.6 1.2 0.27%
i-C12 – – 6.4 5.4 13.8 0.006%
C13 0.3 0.3 0.8 0.6 1.5 0.068%
i-C13 – – 6.8 5.7 11.6 0.004%
C14 0.4 0.4 1.7 1.4 3.1 0.009%
i-C14 – – 2 1.7 3.6 0.010%
C15 13.13 13.4 6.2 5.2 1.8 0.022%
i-C15 – – 2.3 1.9 4 0.010%
C16 21 21.5 10.8 9.1 4.4 0
i-C16 – – 0.7 0.6 0.9 0.034%
C17 18 18.4 5.8 4.9 0.7 0
i-C17 – – 0.8 0.7 1.3 0.043%
C18 28.6 29.3 4.6 3.8 0.7 0.008%
i-C18 – – 1.1 0.9 1.5 0.028%
Water (H2 O) 7.7 7.9 – – – –

tain product distribution, which determines the freezing point isomerised to i-C8. Although this lowers the freezing point,
of the resulting bio-jet fuel. The component i-C18 is one of care must be taken not to promote a high conversion of that
the major components of the isomerisation and hydrocrack- reaction because of the low flash point of isooctane.
ing reactor. Palm oil is rich in palmitic acid which contains After product separation section, the overall yield of prod-
16 carbon atoms, and this may favour the production of C8 ucts was propane (11.5%), naphtha (16.1%), bio-jet fuel (38.5%)
from the middle bond splitting of hexadecane, which is then and diesel (22.5%), which is in line with observed experimental
66 Chemical Engineering Research and Design 1 4 6 ( 2 0 1 9 ) 60–70

Table 4 – Reaction conversions (in mass percentage) in the isomerisation reactor from the model solution.
Hydrocracking to Hydrocracking Hydrocracking Other Total Conversion to
propane (Eq. 9) to C8 (Eq. 10) to C7 (Eq. 11) hydrocracking conversion to isomers
reactions (Eq. 12) hydrocracking

C7 – – – – – –
C8 – – – – – 90
C9 – – – – – 88
C10 – – – – – 90
C11 – – – – – 90
C12 – – – – – 90
C13 5.8 – – – 5.8 90
C14 61.7 – – – 61.7 53.9
C15 51.3 – – – 51.3 27.2
C16 37.1 – – 18 55.1 6
C17 50.4 17.3 2.0 – 69.6 12.6
C18 4.9 41.9 13.9 23.2 83.9 18.9

values at IMP. The total hydrogen consumption was 0.027 kg Therefore, incentives or subsidies in the order of 0.5–0.9 US$/L
H2 / kg oil. The water, CO2 and CO yields are 7.9, 5.6, and 0.4%, would be needed for bio-jet fuel production to be an attractive
respectively. The hydrocarbon composition of the bio-jet fuel business. The estimated capital investment required for this
produced is shown in Table 4, and the calculated freezing point plant is 13.8 million USD. The annualised costs breakdown is
was −53.3 ◦ C, which is lower than the ASTM specification of shown in Fig. 4a.
−47 ◦ C. In the case of the process producing green diesel as the
main product without isomerisation, the minimum selling
3.2. Techno-economic analysis price of green diesel at various production capacities is plot-
ted in Fig. 3b. It can be observed that in this case, the green
3.2.1. Sensitivity to oil price diesel requires lower selling prices and lower capacities for a
It is well known that the vegetable oil feedstock is one of the 10% IRR. For a capacity of about 31.5 thousand bbl/y the MSP
major cost components of producing renewable diesel or jet of green diesel should be 1.09 US$/L. The MSP can be low-
fuel that are considered as commodities. Thus, low prices are ered to 0.92 US$/L for 63 thousand bbl/y, which is competitive
desirable and the analysis was carried out to determine how with the fossil diesel price in Mexico of 0.95 US$/L. Therefore,
low the price should be to obtain profitability. Fig. 2 shows the no incentives or subsidies would be needed for green diesel
results of IRR and unit production cost at different oil prices plants higher than 63 thousand bbl/y.
and bio-jet fuel production capacities. Fig. 2a shows that for The estimated capital investment required for the green
a minimum of 10% IRR, as a target for a positive net present diesel plant is 5.53 million USD. The annualised costs break-
value, the oil price should be the lowest (0.263 $US/L or 5 MXN $ down is shown in Fig. 4b. As it can be observed, in both cases,
/L) and the annual capacity should be more than 95 thousand the oil feedstock makes up the major cost, followed by hydro-
barrels of bio-jet fuel. The minimum bio-jet fuel production gen and annualised capital cost. This creates a high sensitivity
capacity increases to more than 200 thousand barrels of bio-jet of economic indicators to the oil and hydrogen price, and capi-
fuel at higher oil prices, a capacity too large for a constrained tal costs. These factors are considered for uncertainty analysis
oil feedstock availability. Considering a plant with capacity of in the next section.
95 thousand bbl/y, this would require 218 thousand bbl/y of
vegetable oil (29% of national Mexican production). Since the
3.2.3. Monte Carlo simulation
oil feedstock availability is constrained, a smaller biorefinery
Monte Carlo simulations were carried out for the 75 thou-
size is desirable.
sand bbl/y bio-jet fuel plant and the 63 thousand bbl/y green
Considering an average market price for palm oil of about
diesel plant to analyse the uncertainty of the IRR due to the
0.55 $US/L (Indexmundi, 2018), it is expected that obtaining
uncertainty in key parameters. The key parameters simul-
oil at the low prices required would be difficult. Therefore to
taneously varied within a given interval were: the oil price
achieve profitability, the selling price of the bio-jet product
(−20%, +30%), hydrogen price (−20%, +20%), direct fixed capital
may need to be higher. A second analysis was then performed
cost (−20%, +30%), and jet fuel (−10%, +10%) and diesel selling
to determine the minimum selling prices to achieve a 10% IRR
prices (−10%, +10%). The results from 1000 runs are plotted in
for the average oil market price.
Fig. 5. Figs. 5a and b show the distribution for the IRR for the
bio-jet fuel and green diesel-only cases, respectively. In both
3.2.2. Minimum selling price of bio-jet fuel and green cases, there is a 50% risk of failure to achieve the 10% IRR target,
diesel with more than 20% probability of a negative IRR. The proba-
The MSP of bio-jet fuel at various production capacities is plot- bility distributions reflect their high sensitivity to variations in
ted in Fig. 3a. For a lower capacity of about 45 thousand bbl/y capital investment and oil and hydrogen costs due to the high
the MSP should be 1.37 US$/L for a 10% IRR, this can be lowered contribution of this items to the annual total costs. To reduce
to 1.21 US$/L for 75 thousand bbl/y. For a lower MSP, the plant the risk of the investment, a higher price of the products are
capacity would need to be much higher and even then the required. Therefore, Monte Carlo simulations were carried out
MSP is nearly double than the fossil jet fuel price (0.5 US$/L). in order to determine the price of bio-jet and green diesel that
 Chemical Engineering Research and Design 1 4 6 ( 2 0 1 9 ) 60–70 67

Fig. 2 – Sensitivity of internal rate of return to oil price and production capacity for (a) the process producing bio-jet fuel,
green diesel and naphtha and propane and co-products, and (b) the process producing green diesel and propane only.

Fig. 3 – Minimum selling price for (a) bio-jet fuel and (b) green diesel at various production capacities.

Fig. 4 – Annual costs breakdown of the (a) 75 thousand barrels/y bio-jet fuel plant and (b) the 63 thousand barrels/y green
diesel plant.

allows a 95% probability to obtain a positive IRR. The results ucts and if larger plant sizes or lower vegetable oil prices can
are shown in Fig. 5c and d. The price required for the bio-jet be obtained, the risks could be mitigated.
fuel case was found to be 1.35 US$/L, and in the case of green
diesel the price was found to be 1 US$/L. The price for the
4. Conclusions
green diesel is close to the current fossil diesel price which
can make it more competitive in the market. Fig. 5d shows
The simulation and techno-economic analysis of the pro-
that the diesel process has 88% chance to get an IRR equal or
duction of bio-jet fuel and green diesel from hydrotreatment
higher than 10%, while the bio-jet fuel process has 85%. It can
of vegetable oil was presented in this paper. The bio-jet
thus be concluded that the green diesel option is a more attrac-
fuel process involved hydrodeoxygenation and isomerisa-
tive and less risky investment than bio-jet fuel. Furthermore, it
tion/hydrocracking to yield bio-jet as main product and green
requires less feedstock and a lower investment. However, the
diesel, naphtha and propane as co-products. The green diesel
bio-jet fuel production benefits from the credits from coprod-
process involved hydrodeoxygenation only to yield green
68 Chemical Engineering Research and Design 1 4 6 ( 2 0 1 9 ) 60–70

Fig. 5 – Monte Carlo simulation results of IRR for a 75 thousand barrels/y bio-jet fuel plant and a 63 thousand barrels/y
green diesel plant at the minimum selling price (a and b, respectively) and at the price that allows a 95% probability for a
positive IRR (c and d, respectively).

Table A1 – Boiling point, vaporisation heat and freezing point of linear and iso- alkanes present in bio-jet fuel
simulations.
Symbol Hydrocarbon Boiling point, Tb (K) Vaporisation heat,  (kJ/mol) Freezing point, FP (K)

C7 Heptane 371 31.8 183

i-C7 Isoheptane 363 30.6 155
C8 Octane 399 34.4 216
i-C8 Isooctane 389 33.3 164
C9 Nonane 424 36.9 220
i-C9 Isononane 416 36.5 193
C10 Decane 448 38.8 244
i-C10 Isodecane 439 38.2 198
C11 Undecane 469 41.5 248
i-C11 Isoundecane 462 40.3 224
C12 Dodecane 489 43.4 264
i-C12 Isododecane 482 42.7 227
C13 Tridecane 508 45.2 268
i-C13 Isotridecane 501 44.5 247
C14 Tetradecane 527 47.1 279
i-C14 Isotetradecane 518 46.3 248
C15 Pentadecane 544 48.8 283
i-C15 Isopentadecane 535 47.9 265
C16 Hexadecane 560 50.5 291
i-C16 Isohexadecane 550 49.5 266
C17 Heptadecane 575 52.0 295
i-C17 Isoheptadecane 565 51.0 277
C18 Octadecane 589 53.5 301
i-C18 Isooctadecane 578 52.4 279
 Chemical Engineering Research and Design 1 4 6 ( 2 0 1 9 ) 60–70 69

diesel as the main product and propane as co-product. The EU Commission, 2017. Reducing Emissions From Aviation,
simulation approach considered a detailed reaction scheme European Commission.
and a non-linear programming model to determine the reac- https://ec.europa.eu/clima/policies/transport/aviation en.
Flinn, R.A., Larson, O.A., Beuther, H., 1960. The mechanism of
tion conversions for the isomerisation/hydrocracking reactor
catalytic hydrocracking. Ind. Eng. Chem. 52 (2), 153–156.
so that the calculated yields and compositions represent Fu, J., Yang, C., Wu, J., Zhuang, J., Hou, Z., Lu, X., 2015. Direct
observed yields. This allows directing efforts for process devel- production of aviation fuels from microalgae lipids in water.
opment, especially if bio-jet fuel is the main product. The Fuel 139, 678–683.
approach is especially useful when kinetic data is not avail- Hanaoka, T., Miyazawa, T., Shimura, K., Hirata, S., 2015. Jet fuel
able as it can be calibrated to experimental results at different synthesis from Fischer–Tropsch product under mild
reaction conditions and for a variety of oil feedstocks. The hydrocracking conditions using Pt-loaded catalysts. Chem.
Eng. J. 263, 178–185.
simulation results provide a detailed composition of the wide
IATA, 2013. Sustainable Aviation Fuels (SAF).
range of hydrocarbons produced, which allows estimating fuel http://www.iata.org/whatwedo/environment/Pages/
properties. The process simulation was set up in SuperPro sustainable-alternative-jet-fuels.aspx.
®
Designer to carry out techno-economic and sensitivity anal- IATA, Available at 2018. Fact Sheet Climate Change. International
yses of the IRR to key parameters. The results showed that a Air Transport Association
capacity of 75 thousand bbl/y of bio-jet fuel can be profitable https://www.iata.org/pressroom/facts figures/fact sheets/
Documents/fact-sheet-climate-change.pdf.
at a 10% IRR to cope with the average palm oil market price but
ICAO, 2017. International Civil Aviation Organization (ICAO), UN.
requires a MSP of 1.21 US$/L, which needs to be increased to
https://www.icao.int/Pages/default.aspx.
1.35US$/L to lower the risk of failure due to uncertainty in cap- Im-orb, K., Arpornwichanop, A., 2016. Techno-environmental
ital investment, feedstock cost and product prices. In the case analysis of the biomass gasification and Fischer–Tropsch
of the green diesel process, the MSP required is 0.92 US$/L, and integrated process for the co-production of bio-fuel and
increases to 1 US$/L to cope with uncertainty for a 63 thousand power. Energy 112, 121–132.
bbl/y capacity. The case of green diesel can be more compet- Indexmundi, 2018. Palm Oil Price.
https://www.indexmundi.com/commodities/?commodity=
itive with current fossil diesel prices. This is because green
palm-oil&months=12.
diesel has a higher price and the yield is higher compared Klein, B.C., Chagas, M.F., Junqueira, T.L., Rezende, M.C.A.F., de
to bio-jet fuel, which leads to a lower requirement of veg- Fátima Cardoso, T., Cavalett, O., Bonomi, A., 2018.
etable oil for the same production rate and lower capital and Techno-economic and environmental assessment of
operating costs. Therefore, under current fuel and feedstock renewable jet fuel production in integrated Brazilian
price trends, a more viable and lower risk investment would sugarcane biorefineries. Appl. Energy 209, 290–305.
be welcomed in a green diesel production facility. However, Li, Y., Zhao, C., Chen, L., Zhang, X., Zhang, Q., Wang, T., Ma, L.,
2018a. Production of bio-jet fuel from corncob by
the bio-jet fuel production benefits from multi-fuel produc-
hydrothermal decomposition and catalytic hydrogenation: lab
tion and if larger plant sizes and lower oil feedstock prices analysis of process and techno-economics of a pilot-scale
can be obtained, the risks then can be mitigated. Future work facility. Appl. Energy 227, 128–136.
includes consideration of other oil feedstocks to investigate Li, X., Mupondwa, E., Tabil, L., 2018b. Technoeconomic analysis of
the compositional effect, conduct a life cycle assessment and biojet fuel production from camelina at commercial scale:
compare results to other process technologies. case of Canadian Prairies. Bioresour. Technol. 249, 196–205.
McCall, M.J., Kocal, J.A., Bhattacharyya, A., Kalnes, T.N.,
Brandvold, A.T. (2011). Production of aviation fuel from
Acknowledgements renewable feedstocks. Patent US8039682B2.
Pearlson, M., Wollersheim, C., Hileman, J., 2013. A
This work was supported by SENER-CONACyT Energy Sustain- techno-economic review of hydroprocessed renewable esters
ability Fund project 248090, and IMP project IMP Y. 61023. and fatty acids for jet fuel production. Biofuels Bioprod.
Biorefining 7 (1), 89–96.
Appendix A Penner, J.E., Lister, D.H., Griggs, D.J., Dokken, D.J., McFarland, M.,
1999. Aviation and the global atmosphere. In: IPCC Special
Report. Cambridge University Press, UK.
Riazi, M.R., 2005. Characterization and Properties of Petroleum
References Fractions, vol. 50. ASTM International.
Rocha, E.G., Oviedo, T.A.B., Rivera, B.L.M., Ramos, R.Z., 2002. U.S.
Patent No. 6,383,975. U.S. Patent and Trademark Office,
Anand, M., Farooqui, S.A., Kumar, R., Joshi, R., Kumar, R., Sibi,
Washington, DC.
M.G., Sinha, A.K., 2016. Optimizing renewable oil
Starck, L., Pidol, L., Jeuland, N., Chapus, T., Bogers, P., Bauldreay, J.,
hydrocracking conditions for aviation bio-kerosene
2016. Production of hydroprocessed esters and fatty acids
production. Fuel Process. Technol. 151, 50–58.
(HEFA)–optimisation of process yield. Oil Gas Sci. Technol.
ASTM, 2017. D7566-17a, Standard Specification for Aviation
Revue d’IFP Energies nouvelles 71 (1), 10.
Turbine Fuel Containing Synthesized Hydrocarbons. ASTM
Thybaut, J.W., Marin, G.B., 2016. Multiscale aspects in
International www.astm.org.
hydrocracking: from reaction mechanism over catalysts to
Atsonios, K., Kougioumtzis, M.A., Panopoulos, K.D., Kakaras, E.,
kinetics and industrial application. Advances in Catalysis, vol.
2015. Alternative thermochemical routes for aviation biofuels
59. Academic Press, pp. 109–238.
via alcohols synthesis: process modeling, techno-economic
Valencia, D., García-Cruz, I., Uc, V.H., Ramírez-Verduzco, L.F.,
assessment and comparison. Appl. Energy 138, 346–366.
Amezcua-Allieri, M.A., Aburto, J., 2018. Unravelling the
Chu, P.L., Vanderghem, C., MacLean, H.L., Saville, B.A., 2017.
chemical reactions of fatty acids and triacylglycerides under
Process modeling of hydrodeoxygenation to produce
hydrodeoxygenation conditions based on a comprehensive
renewable jet fuel and other hydrocarbon fuels. Fuel 196,
thermodynamic analysis. Biomass Bioenergy 112,
298–305.
37–44.
Coonradt, H.L., Garwood, W.E., 1964. Mechanism of
Wang, W.C., Tao, L., 2016. Bio-jet fuel conversion technologies.
hydrocracking. Reactions of paraffins and olefins. Ind. Eng.
Renew. Sustain. Energy Rev. 53, 801–822.
Chem. Process Des. Dev. 3 (1), 38–45.
70 Chemical Engineering Research and Design 1 4 6 ( 2 0 1 9 ) 60–70

Weitkamp, J., Jacobs, P.A., Martens, J.A., 1983. Isomerization and Yao, G., Staples, M.D., Malina, R., Tyner, W.E., 2017. Stochastic
hydrocracking of C9 through C16 n-alkanes on Pt/HZSM-5 techno-economic analysis of alcohol-to-jet fuel production.
zeolite. Appl. Catal. 8 (1), 123–141. Biotechnol. Biofuels 10 (1), 18.
Wu, X., Jiang, P., Jin, F., Liu, J., Zhang, Y., Zhu, L., Li, Q., 2017. Zech, K.M., Dietrich, S., Reichmuth, M., Weindorf, W.,
Production of jet fuel range biofuels by catalytic Müller-Langer, F., 2018. Techno-economic assessment of a
transformation of triglycerides based oils. Fuel 188, 205–211. renewable bio-jet-fuel production using power-to-gas. Appl.
Yang, Z., Qian, K., Zhang, X., Lei, H., Xin, C., Zhang, Y., Villota, E., Energy 231, 997–1006.
2018. Process design and economics for the conversion of
lignocellulosic biomass into jet fuel range cycloalkanes.
Energy 154, 289–297.