Energy 164 (2018) 99e111

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Experimental studies and techno-economic analysis of

hydro-processed renewable diesel production in Taiwan
Keng-Hao Hsu, Wei-Cheng Wang*, Yu-Cheng Liu
Department of Aeronautics and Astronautics, National Cheng Kung University, Tainan 70101, Taiwan

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

Article history: For the purpose of promoting hydro-processed renewable diesel as the local replacements of traditional
Received 20 February 2018 diesel and transestification biodiesel, the experimental investigation and techno-economic analysis were
Received in revised form carried out in this study. The experiment was conducted with varying reaction temperature, pressure,
6 August 2018
weight hourly space velocity (WHSV) and H2-to-oil ratio over the NiMo/g-Al2O3 catalyst and the con-
Accepted 29 August 2018
version, selectivity and yield were determined. The experimental conditions and results were input into
Available online 30 August 2018
process simulation model for determining the mass and energy flows at the scale of 600 tonnes per day.
Furthermore, the techno-economic analysis was conducted based on the mass and energy balances
Keywords:
Hydro-processing
obtained from process simulation, and the locally quoted capital and operating costs, yielding to the
Renewable diesel minimum renewable diesel price (MRDP) of $1.72/L. The effects of hydrogen, plant capacity, catalyst as
Oil upgrading well as the operating conditions on MRDP were demonstrated and discussed.
Hydro-deoxygenation © 2018 Elsevier Ltd. All rights reserved.
Techno-economic analysis
Process simulation

1. Introduction or a combustor which leads to better ignition performance. Lower

Due to the emissions released from diesel-powered vehicles or
machines, alternative diesels such as biodiesel, syndiesel and
hydro-processed renewable diesel have been developed. It is re-
ported that the greenhouse gas emissions was reduced by 6.4
million metric tons as the renewable diesel was applied [1]. As one
of the renewable diesels, hydro-processed renewable diesel refers
to a petrodiesel-like fuel derived from biological sources which are
chemically not esters and thus distinct from biodiesel. Table 1
shows the differences of property of renewable diesel, traditional
biodiesel, Fischer-Tropsch (FT) syndiesel and ultra-low sulfur diesel
(ULSD) [2e4]. Compared to traditional biodiesel, hydro-processed
renewable diesel is more favorable according to its heating value,
cloud point, cetane number, oxygen content, density, viscosity and
stability. High heating value and high cetane number make
renewable diesel competitive in engine performance. Hydro-
processed renewable diesel also has relatively lower cloud point,
which makes it survive in severe climate. Lower viscosity compared
to biodiesel and FT diesel helps in atomization process in an engine
* Corresponding author.
E-mail address: wilsonwang@mail.ncku.edu.tw (W.-C. Wang).
sulfur and aromatics contents compared to ULSD and lower oxygen
content compared to biodiesel helps reduce the emissions from
engine as feeding hydro-processed renewable diesel. Better sta-
bility than biodiesel solves its limitation of storage. It was reported
that 100% of hydro-processed renewable diesel reduces the exhaust
emissions of PM (particulate matter), NOx, THC (total hydrocar-
bons) and CO by 28%e46%, 7%e14% and 5%e78%, respectively,
compared to regular diesel [4]. Currently, hydro-processed
renewable diesel is commercially demonstrated in some oil re-
fineries such as UOP Ecofining™ process [5], Diamond Green Diesel
[5] and Neste Oil [6]. In Taiwan, B2 biodiesel (2% biodiesel blended
in regular diesel) was suspended in April 2014, due to the quality of
the fuel. The problems of storage and engine horse power were the
main reasons. In order to reinitiate the use of biofuel in vehicle
transportation in Taiwan, hydro-processed renewable diesel would
be a good target.
Feedstock play an important role for evaluating the production
of biofuel. Plant oils, algal oil and waste cooking oil have been used
as feedstocks for producing hydro-processed renewable diesel. For
the situation in Taiwan, plant oils such as palm oil and jatropha oil
which can be easily imported from Southeast Asia are the best
choices. Waste cooking oil is another option but the quality of
source is doubtful. Before the quality of waste cooking oil is fully

https://doi.org/10.1016/j.energy.2018.08.208
0360-5442/© 2018 Elsevier Ltd. All rights reserved.
100 K.-H. Hsu et al. / Energy 164 (2018) 99e111

Table 1
Comparison of renewable diesel, biodiesel, FT diesel and ultra-low sulfur diesel.

Renewable Diesel Biodiesel FT-diesel Ultra-low Sulfur diesel

Sulfur (ppm) <10 <10 0.05 46

LHV (MJ/kg) 44 ~38 43.9 42.7
Cloud point (C) 20 5 17 12
Distillation (C) (10e90%) 265e320 340e355 295e342 264e329
Cetane Number >80 50 79 53.9
Oxygen (%) 0 11 0 0
Density at 15 C (kg/m3) 780 880 785 840
Viscosity at 40 C (mm2/s) 2.5 4.1 3.5 2.5
Aromatics (wt%) 0 0 0.3 24.4
Stability Good Marginal Good Good

*Note: Current regulation in Taiwan for sulfur content in diesel fuel is 10 ppm for land vehicle and 5000 ppm for marine vehicle.

controlled, crude palm oil imported from other Asian countries
such as Malaysia and Indonesia would be favorable. Based on the
fatty acid profile of palm oil, the hydro-processing reaction takes
place under the hydrogen environment over a heterogeneous
catalyst. As shown in Fig. 1, the glycerides are saturated and con-
verted into free fatty acids through the propane cleave process.
High value propane product is yielded from this process. The FFAs
are then turned into straight-chain alkanes via the deoxygenation
process, which includes the hydro-deoxygenation and decarbox-
ylation/decarbonylation routes. Co-products such as water emitted
from hydro-deoxygenation, CO2 emitted from decarboxylation and
CO emitted from decarbonylation can be found in the deoxygen-
ation process. Previously, Sotelo-Boyas et al. [7] studied the pro-
duction of hydro-processed renewable diesel through hydro-
processing of rapeseed oil over 3 different catalysts: Pt/H-Y, Pt/H-
ZSM-5 and presulfided NiMo/g-Al2O3 in a batch reactor at the
temperature ranging from 300
C to 400
C and initial hydrogen
pressure ranging from 5 to 11 MPa. The results showed that NiMo/
g-Al2O3 provides the highest yield of n-paraffins from C15 to C18.
The highest yield was found at the temperature of 350
C and initial
H2 pressure of 8e10 MPa. High initial H2 pressure favors the yield of
n-octadecane. As the temperature increased over 375
C, cracking
of the intermediate carbenium ions dominated and this reduced
the yield of hydro-treated renewable diesel. Zhou and Lawal [8]
conducted the hydro-deoxygenation of microalgae oil over pre-
sulfided NiMo/g-Al2O3 in a micro reactor at the reaction tempera-
ture and pressure respectively ranging from 300
C to 360
C and
2.07 MPae3.45 MPa and the optimum conditions for this hydro-
treatment were found at 3.45 MPa and 360
C. Approximately
56.2% of C13 to C20 hydrocarbons was yielded with a 62.7% carbon
yield and a nearly complete conversion (98.7%) was achieved.
Kiatkittipong et al. [9] carried out the hydro-processing of crude

Fig. 1. Experimental setup of hydro-cracking/hydro-isomerization. [(1) H2 Tank; (2) N2 Tank; (3) Filter; (4) Mass Flow Controller; (5) Pressure Gauge; (6) Hot Place Stirrer; (7) High
Pressure Pump; (8) Pressure Relief Valve; (9) K-Type Thermocouple; (10) Reactor; (11) Heater; (12) Temperature Controller; (13) Condenser; (14) Coolant Tank; (15) Liquid-gas
separator; (16) Back Pressure Regulator; (17) Empty Gas Cylinder; (18) Pressure Relief Regulator; (19) Metering; (20) product Container; (21) 5 Port valve; (22) Monitor.].
 K.-H. Hsu et al. / Energy 164 (2018) 99e111
palm oil (CPO), degummed palm oil (DPO) and palm fatty acid
distillate (PFAD) using presulfided NiMo/g-Al2O3and Pd/C catalysts.
For the case of CPO, it was revealed that the conversion of CPO was
enhanced with increasing residence time. The highest diesel yield
was found at the reaction time of 3 h. In addition, no conversion
was found within the reaction temperature from 350
C to 375
C.
At the temperature of 400
C, the conversion and diesel yield were
respectively increased to 70% and 51% due to the increase in
decarboxylation (DCO2) and decarbonylation (DCO). As a result, the
diesel yield reaches the maximum value at the reaction pressure of
40 bar. Furthermore, the study carried out by Srifa et al. [10] has
reported that the optimal conditions for producing bio-
hydrogenated diesel derived from palm oil over presulfided
NiMo/g-Al2O3 were at the temperature and the pressure between
270
C to 420
C and 1.5 MPae8 MPa, respectively, and with a liquid
hourly space velocity (LHSV) and a H2-to-oil ratio ranging from 0.25
h1 to 5 h1 and 250 to 2000 (Ncm3/cm3). It was additionally found
that the HDO was significantly enhanced at the reaction tempera-
tures of 270
Ce330
C (from 21.9% to 74.9%), but decreased at the
temperature higher than 360
C due to the dominated cracking and
isomerization as the reaction temperatures between 270 and
300
C favor the decarboxylation and decarbonylation reactions. In
addition to the feedstock and operating conditions, the function of
the catalyst also plays an important role for converting the crude
bio-oil into diesel-like hydrocarbons. In the cases of NiMo/g-Al2O3
catalyst, the Ni/Al2O3 catalyst produces decarboxylated hydrocar-
bon products [11] while the Mo/Al2O3 catalyst mainly produces
hydro-deoxygenated products [12]. Compared to Pd/C, the pre-
sulfided NiMo/g-Al2O3 is more preferred for triglyceride (TG)
feedstock [9]. However, it was found that NiMo/g-Al2O3 was
deactivated due to the accumulation of produced oxygenated in-
termediates in hydro-deoxygenation reaction, which results in a
lower yield of n-paraffins and a reduced quality of the products.
This can be avoided by elevating the H2-to-oil ratio, residence time,
temperature and pressure [8,9].
Techno-economic analysis of hydro-processing for producing
transportation fuels has been performed recently. Miller and Kumar
[13] studied the production cost of hydrogenation-derived renew-
able diesel (HDRD) from canola oil and camelina oil based on the
capacities of 39746e3179746 L. The minimum costs for the fuels
produced from canola and camelina oils were found as $1.09/L and
$0.85/L, respectively. The feedstock cost, solvent price and solvent
recovery were found to be highly sensitive to the production cost.
Glisic et al. [14] conducted a process and techno-economic analysis
of hydro-treated renewable diesel production from waste vegetable
oil, based on the capacity of 100,000 tonnes per year. The produc-
tion cost was calculated to be $0.63/L~0.68/L, similar to the tradi-
tional biodiesels produced through the supercritical condition
(SCA) and conventional alkali homogeneously catalysed FAME
production process (HACA). It was suggested that the hydrogena-
tion unit incorporated in a refinery where the supporting in-
frastructures are available is more economical than a stand-alone
unit ($0.63/L vs $0.78/L). In order to understand the sustainability
of hydro-processed renewable diesel in Taiwan, it is necessary to
evaluate the production cost based on the Taiwan situation.
In this study, hydro-processing for producing hydro-processed
renewable diesel is first performed at the lab scale with various
parameters such as temperature, pressure, H2-to-oil ratio and
weight hourly space velocity (WHSV). Process simulation as well as
techno-economic analysis were then carried out for calculating the
production cost at the scale of 600 tonne per day, with the infor-
mation quoted locally. The minimum renewable diesel selling pri-
ces (MRDPs) were then determined and compared according to
different scenarios such as hydrogen price, plant capacity, catalyst
cost and various operating parameters.
101
2. Experimental
2.1. Materials
The crude palm oil (CPO) used in this study to produce hydro-
processed renewable diesel was provided by a local company
(Mai-Fung Inc.) in Taiwan. It was first analyzed through the gas
chromatography-mass spectrometer (GC-MS-QP2010 SHIMADZU)
and the major fatty acid compositions were detected as 48% satu-
rated compounds (44% palmitic (16:0) and 4% stearic (18:0)) and
50% unsaturated compounds (40% oleic (18:1) and 10% linoleic
(18:2)). The NiMo/g-Al2O3 catalyst used in the experiments was
provided by A-Plus Inc. Experimental gases such as 99.99% N2 and
H2 were provided by the local gas company (Yun-Shan Gas Com-
pany). The liquid standard, C8 to C20 (PC code 101685440), was
provided by Sigma-Aldrich, which contains 40 mg/L for each
composition. The gaseous standard (with the mixture of 1.02% O2,
1.01% CO2, 1.27% H2, 1.01% N2, 1.02% O2, 1.05% CH4, 1.06% C2H6, 1.03%
C3H8 and 1.01% n-C4H10) was provided by Jing De Gases Co. Ltd.
2.2. Experimental
The systematic diagram of converting CPO (crude palm oil) into
hydro-processed renewable diesel was shown in Fig. 1. The ball
valve, needle valve, lift check valve, back pressure regulator, pres-
sure release regulator, connectors, and adapters used in the ex-
periments were made of 316 stainless steel provided by Swagelok.
An industrial pressure gauge (PGI-63B-PG800-LAQX) placed in a
cross union connector where the carrier gases and the oil feedstock
mixed together was used for measuring the reaction pressure. The
reactor was surrounded by a rectangular electric furnace controlled
by a temperature control device (Ching Ying: CI-35E).
Approximately 5 g of NiMo/g-Al2O3 catalyst used for the cata-
lytic reaction was sandwich-packed within a 316 stainless steel
fixed bed reactor (400 mm length and 30 mm inner diameter)
along with the cotton wools. The volume of the catalyst bed is
approximately 184 cm3. In each experiments, the loading of the
catalyst was averagely separated into 10 equal parts in the reactor
with the upper and bottom parts filled with cotton wool to ensure
the stability of catalyst. Two mass flow controllers (MFCs, Brooks
5850E) driven by a LabView program were installed individually to
control the volumetric flow rates of H2 and N2 gases. The flow rates
measured through MFCs were calibrated before experiment. The
CPO was stirred thoroughly via a hotplate stirrer (HTS-1300) at the
temperature around 70
C before injecting into the reactor. Before
experiment, N2 was used to purge the system at a rate of 50 sccm to
remove all the oxygen from the system. NiMo/g-Al2O3 catalyst was
then activated by flowing H2 at a rate of 50 sccm for 2 h at the
temperature of 200
C. After activation, the temperature and
pressure were further increased to the reaction conditions. As the
desired temperature and pressure were reached, the CPO was fed
using a high pressure liquid pump (Eldex) into the reactor. The
liquid and gas products were collected every 30 min through the
liquid-gas-separated sampling system which includes a back
pressure regulator (BPR, Tescom) for controlling the system pres-
sure, a spun sampling cylinder (Gyrolok) with a capacity of 300 ml
to well-function the BPR, two pressure reducing regulators for
regulating the system pressure and a metering valve for collecting
the specific amount of liquid samples desired. In this study, the
experiment was conducted under various conditions: reaction
pressures from 550 psig to 925 psig, reaction temperatures ranging
from 370
C to 460
C, weight hourly space velocity (WHSV, the
ratio of feedstock flow rate to the weight of the catalyst) from 1 h1
to 4 h1, total reaction time 2 h and 30 min and H2-to-oil ratio from
500 to 1250. In addition, a temperature error of ±5
C was taken as
102 K.-H. Hsu et al. / Energy 164 (2018) 99e111

the tolerance for all the experiments. 3. Process simulation and techno-economic analysis

3.1. Process simulation

2.3. Product analysis
For the purpose of scaling up the lab-scale renewable diesel
process, a process simulation model was built to compute the mass
The liquid products were analyzed by the GC-MS equipped with
and energy balances, based on the experimental conditions and
a SGE Analytical Science HT5 column (length 15 m, diameter
results, as shown in Fig. 2. The process model was developed
0.25 mm, film 0.1 mm). The samples were diluted by dissolving
through Aspen Plus™, with detailed methods listed in Table 2. The
65 mg of product samples in 4 ml HPLC grade hexane (Fisher Sci-
properties of feedstocks were characterized through the process
entific) and 0.2 mL of diluted sample was injected into the GC. The
type “COMMON” in Aspen Plus. The modules of the base and
carrier gas (helium, 99.99%) flow rate was set at 0.8 ml/min. The
property methods are set as NRTL. The components of the feedstock
injector temperature was 200
C. The initial oven temperature was
were obtained in the APV72 database (AQUEOUS; SOLIDS; INOR-
first set as 35
C and held for 1 min and then increased to 50
C at
GANIC; ASPENPCD; POLYMER; AQU92; COMBUST; ELECPURE; EOS-
5
C/min and held for 2 min. It was then further increased to 120
C
LIT; ETHYLENE; INITIATO; NRTL-SAC; PC-SAFT; POLYPCSF; PURE10,
at 10
C/min and held for 1 min. Finally it was increased to 280
C at
11, 12, 13, 20, 22, 24, 93, 856; SEGMENT). The databanks include all
8
C/min and held for 1 min. Standard n-alkane samples were first
possible species to be present. No pseudo-components was
examined to qualitatively identify the components and quantita-
assumed. The capacity of the plant was selected as 600 tonnes per
tively calculate their concentrations. The concentrations of n-
day for producing renewable diesel, based on the affordable feed-
alkane products were obtained through area normalization be-
stock provided by Taiwan. Experimental operating conditions such
tween standard and experimental results.
as temperature, pressure, WHSV and H2-to-oil ratio were input into
Gas products were analyzed through two equipment: a GC-MS
the simulation. In the process model, oil was sent to the hydro-
and a gas chromatography-thermal conductivity detector (GC-
processing reactor through a high pressure gear pump specified
TCD-GC2014AT.SHIMADZU). The samples were first collected in the
according to the engineering manual [15]. The reaction occurred
Tedlar sampling bag every 30 min and then injected into GC-MS
within the reactor was split into 5 reactors modeled using RStoic
and GC-TCD. Organic gaseous products were analyzed through
(reactions actually happen in one reactor), which are: (1) REAC-1:
the GC-MS equipped with SGE Analytical Science HT5 column
hydrogenation, converting unsaturated TG, DG (di-glyceride), MG
(length 15 m, diameter 0.25 mm, film 0.1 mm). The injector tem-
(mono-glyceride) and fatty acids into saturated ones; (2) REAC-2:
perature was set as 200
C. The oven temperature was initially 35
C
Hydrogenolysis, converting saturated TG, DG, MG and fatty acids
and held for 3 min, and then increased to 75
C at 5
C/min and held
into free fatty acids (FFAs) by removing propane; (3) REAC-3: HDO,
for 3 min. It was then increased to 120
C at 5
C/min and held for
converting FFAs into alkanes by releasing H2O; (4) REAC-4: DCO,
1 min. Inorganic gaseous products, such as CO, CO2, CH4, C2H6 and
converting FFAs into alkanes by releasing CO; (5) REAC-5: DCO2,
H2, were analyzed through GC-TCD with a packed column (TDX-01
converting FFAs into alkanes by releasing CO2, as seen in Fig. 2. The
column, 3 m  3 mm) using helium as the carrier gas. The injector
flow rate of H2 was calculated based on the double bonds within TG,
temperature was set as 200
C. The initial and ultimate oven tem-
DG, MG and fatty acid which need to be saturated, the hydrogen
peratures were maintained at 80
C and held for 22 min. The
requirement for propane cleave and HDO. Slightly more hydrogen
gaseous products were quantified by the external standard method.
was added to the process for the methanation as well as the pro-
In addition, the experimental error of the GC measurements was
tection of catalyst from coking and water contamination. Pd/C or
±5%.
NiMo/g-Al2O3 catalyst was assumed to be the hydro-processing
The conversion, selectivity and yield were calculated based on
catalyst. After reaction, propane was separated from the gas
the equations shown below. The conversion was determined by the
products using the low-temperature distillation method [16]. The
residues of glycerides and free fatty acids in the products. The
unreacted compounds are pumped back to the reactor after distil-
selectivity was determined by the ratio of concentrations of the
lation. The alkane products were distillated in a distillation column
target product, C15 to C18 normal alkanes, to the total concentra-
and lighter species from the top are the hydro-processed renewable
tions of all the species in the product.
diesel product. The heavier species from the bottom, for the pur-
  pose of maintain the fuel cold flow properties, are sent to a hydro-
G:R:ðfeedÞ  G:R:ðproductÞ
Conversion ð%Þ ¼  100% (1) cracker operated at 350
C and 870 psig for cutting off the heavy
G:R:ðfeedÞ
carbon chain length [17]. The catalyst Pt/SAPO-11 was used in the
hydro-cracker with the WHSV of 2 h1 [17]. The consumption of
where G.R. stands for the combination of the concentrations of
hydrogen is calculated based on the balance of hydro-cracking re-
glycerides and free fatty acids.
action for the hydrocarbons with carbon numbers of 18e24. After
  hydro-cracker, the gas products were released to mix with the gas
Ca
Selectivity ð%Þ ¼  100% (2) species from hydro-processing reactor. The liquid products were
Ct
sent to another distillation column. The lighter species from the top
were mixed with the renewable diesel product from the first
where Ca is the concentrations of C15 to C18 straight-chain alkanes
distillation column and heavier species from the bottom can be
in the product and Ct stands for the concentrations of all the species
viewed as heavy-oil.
in the product.
In addition, yield (GC yield) is calculated in accordance with the
following equation:
3.2. Techno-economic analysis
Yield ðGC yieldÞ ð%Þ ¼ Conversion  Selectivity (3)
According to the mass and energy calculations via process
simulation, the capital and operating costs were computed with the
 K.-H. Hsu et al. / Energy 164 (2018) 99e111 103

Fig. 2. Process model for renewable diesel production process. (a) First part; (b) second part.

Table 2 Table 3 are the material and utility costs based on the plant capacity
Methods for process simulation. of 600 tonnes/day, in accordance with the mass and energy bal-
Aspen parameters Type ances computed in Aspen Plus™. The feedstock for producing
Property method NRTL
renewable diesel, palm oil, was assumed to be imported outside
Reference method NRTL Taiwan. The price listed in Table 3 includes the original price of oil,
Component Conventional and Solid the shipping cost, import tax and business tax. The shipping cost
Process type Common from the harbor of Thailand to the one of Taiwan for the size of
Binary Interaction parameter RKTKIJ-1
33 m3 is approximately $1096. The catalyst was assumed to be
replaced once a year with the WHSV of 2. Two hydro-processing
reactors were therefore quoted in this case. Propane can be sold
assist of local data. The annual operating cost includes variable as the credit to offset the production cost. Fixed operating cost
(Table 3) and fixed (Table 4) operating costs. The items listed in includes labor cost (Table 4) and maintenance cost. The
104 K.-H. Hsu et al. / Energy 164 (2018) 99e111

Table 3
List of variable operating cost.

Material Consumption Unit cost (US$ per unit) Reference

Feedstock 25000.00 kg/hr 0.80/kg [20e22]

Natural gas 35527.00 MJ/h 0.39/m3 [23]
Cooling water 345700.00 kg/hr 0.41/m3 [24]
Electricity 2496.76 kWh 0.07/kWh [25]
Hydrogen 676.84 kg/hr 3.51/kg [26]
Hydro-processing catalyst 25000.00 kg/year 3266.10/kg [27]
Hydro-cracking catalyst 5991.34 kg/year 5935.39/kg [27]
Wastewater treatment chemicals (Chloride ingot, HClO) 21.75 kg/hr 6.81/kg [28]
Propane 1771.55 kg/hr (production) 1.00/kg [29]
Heavy-oil 809.10 kg/hr (production) 0.60/kg [30]

Table 4 Table 6
List of labor costs in fixed operating cost. List of plant assumptions.

Position Amount Unit cost (US$/year) Reference Description Assumed Reference

value
Plant manager 1 71,437.00 [30]
Vice plant manager 3 62,507.00 Equity 60% [27]
Maintenance supervisor 1 20,091.00 Loan interest 6% [27]
Maintenance engineer 3 18,752.00 Loan term 5 years [27]
Maintenance Technician 5 15,626.00 Working capital (% of FCI) 15% [40]
Plant supervisor 2 22,324.00 Depreciation period
Plant engineer 5 20,538.00 General plant 7 years [41]
Plant technician 7 17,859.00 Steam plant 20 years [41]
Yard employees 2 9822.00 Construction period 3 years [42]
Accountant & Secretaries 3 14,287.00 % spent in year 2 30% deposit þ design
% spent in year 1 30%
% spent in year 0 40% final down payment þ test
run
Start-up time 0.5 year [41]
Table 5 Fuel production/feedstock use 50% of normal
List of equipment and cost. Variable cost 75% of normal [41]
Fixed cost 100% of [41]
Material Number Unit cost (US$) Reference normal
Hydrogen Compressor 1 40,536.00 [31] Discount rate (Internal rate of 10%
Liquid Heater 2 1064.07 [32] return)
Hydrotreating Reactor 3 80,869.32 [33] Income tax rate 30% [43]
Liquid Pump 4 1013.40 [34] Operating hours per year 8088 [44]
Condenser 4 1140.07 [35]
Vapor-liquid Separator 3 1292.08 [36]
Pressure Relief Valve 1 726.27 [37]
Distillation Tower 3 62,628.12 [38]
year of this study is 2017 and all the costs are inflated to the base
Oil Storage tank 2 6384.42 [33] year [19]. The MRDP was calculated through setting the net present
Gas Storage Tank 1 1261.68 [39] value (NPV) to be zero.
Wastewater Treatment 1 170,250.00 [27]

4. Result and discussions

In this study, the effects of reaction parameters such as reaction

maintenance cost is 4% of ISBL (inside battery limits), which is the
cost of procuring and installing all process equipment [18]. The cost
of equipment was quoted from the local vendor with the appro-
priate specifications, including sizing, flow rate, temperature/
pressure limits … etc (see Table 5). The scaling and installation
were based on the engineering handbook [18]. The plant assump-
tions based on the local case is listed in Table 6. The fixed capital
investment (FCI) is the combination of total direct and indirect
costs. The total direct cost (TDC) is the summary of the total
equipment costs after installation, warehouse (5.7% of ISBL), site
development (12.9% of ISBL) and additional piping (6.4% of ISBL).
The total indirect cost includes Engineering and Supervision (13.7%
of TDC) and construction expenses (34% of TDC). The total capital
investment (TCI) includes FCI, land and working capital (15% of FCI).
The discounted cash flow rate-of-return analysis was conducted
based on the values of TCI as well as variable and fixed operating
costs, in accordance with the financial assumptions in Table 6. As
the variable and fixed operating costs as well as TCI are found, a
discounted cash flow rate-of-return analysis was carried out ac-
cording to the financial assumptions shown in Table 9. The base
temperature, pressure, WHSV and H2-to-oil ratio are examined for
the production of renewable diesel. The conversion, selectivity and
yield were calculated with varying the reaction parameters and
these relations were input into process simulations for determining
the production cost. Table 7 shows the concentrations of produced
n-alkanes including n-C8 to n-C14, n-C15, n-C16, n-C17, n-C18 and n-
C19 to n-C20 as well as the appearance of the liquid products with
various reaction parameters. It can be seen that n-C16 obtains the
highest concentration among all the alkane products, followed by
n-C15.
4.1. Effects of reaction condition on the product distribution
4.1.1. Reaction temperature
In this part, the effects of reaction temperature were observed
by varying the temperature from 370
C to 460
C with fixed
experimental parameters such as hydrogen pressure, H2-to-oil ratio
and WHSV of 800 psig, 750 and 2 h1, respectively. The reaction
temperature has been considered as the most influential parameter
on the catalyst performance and deactivation [45] and it leads to a
 K.-H. Hsu et al. / Energy 164 (2018) 99e111 105

Table 7
Liquid product compositions in different experiment parameters.

Parameter (C) P (psig) WHSV (h1) H2-to-oil ratio n-C8-n-C14 (wt%) n-C15 (wt%) n-C16 (wt%) n-C17 (wt%) n-C18 (wt%) n-C19-n-C20 (wt%) Liquid color
T

Temperature 370 800 2 750 20.02 14.37 45.74 3.79 12.63 3.43 Yellow
400 800 2 750 37.31 12.09 37.05 2.38 7.86 3.28 Yellow
430 800 2 750 53.32 11.88 26.05 1.53 4.03 3.16 Light Green
460 800 2 750 59.93 13.10 18.42 1.25 2.47 4.80 Light Green
Pressure 430 550 2 750 42.95 18.32 20.68 13.14 2.52 2.36 Brown
430 675 2 750 58.55 11.83 20.79 1.72 3.22 3.85 Light Green
430 800 2 750 53.32 11.88 26.05 1.53 4.03 3.16 Light Green
430 925 2 750 53.10 13.42 21.73 5.89 2.7 3.13 Light Green
WHSV 430 800 1 750 43.60 17.31 26.78 2.78 5.82 3.69 Brown
430 800 2 750 53.32 11.88 26.05 1.53 4.03 3.16 Light Green
430 800 3 750 50.36 11.66 27.86 1.58 4.77 3.74 Light Green
430 800 4 750 50.62 12.28 27.89 1.55 4.09 3.54 Light Green
H2-to-oil ratio 430 800 2 500 48.71 16.11 24.24 2.66 4.31 3.95 Yellow
430 800 2 750 53.32 11.88 26.05 1.53 4.03 3.16 Light Green
430 800 2 1000 41.06 13.06 34.37 1.76 5.53 4.19 Clear
430 800 2 1250 44.55 12.98 30.59 2.37 5.95 3.54 Clear

massive difference on the product compositions. At the tempera-
ture of 370
C, small particles and crystalline were observed as the
product was stored at room temperature atmosphere, which were
considered to be long-chain normal alkanes (>n-C18) or unreacted
free fatty acids. As seen in Fig. 3 (a), the concentrations of n-C8 to n-
C14 increased with increasing temperature, along with the decrease
of n-C16 to n-C18, mainly due to the occurrence of cracking reaction
on the species of n-C16 to n-C18. The variances of n-C19 to n-C20 were
not obvious, indicating that this amount of hydrogen was insuffi-
cient to hydro-crack long chain alkanes. As seen in Fig. 3 (b), the
conversions of glycerides and fatty acids reached 100% as the
temperature went to 400
C. However, the selectivity of n-C15 to n-
C18 decreased with increasing temperature. The yield of n-C15 to n-
C18 in the product drops by 52% as the temperature was raised from
370 to 460
C. The results in this study seem to be consistent to
 a
those of Sim cek et al. [46]. The increase of temperature obviously
results in the decrease of cetane number of the produced renew-
able diesel. However, this leads to better cold flow properties
compared to low temperature reaction due to the increased light
species. Consider the climate in Taiwan (yearly average tempera-
ture of 22
C), high cetane number is more favorable than cold flow
properties. Therefore, it is recommended to have the reaction
operated at 400
C, with the similar yields of n-C8 to n-C14 and n-
C16. Fig. 3 (c) shows the gas product compositions with various
reaction temperature. The increased CO2 and CO with increasing
temperature demonstrates the increased decarboxylation (DCO2)
and decarbonylation (DCO) reactions, and the performance of DCO2
was better than DCO. The growths of DCO2 and DCO can also be
reflected by observing the decreased consumption of hydrogen
(hydrogen increasing), since no hydrogen is needed for DCO2 and
DCO. Hydrogenolysis occurred mostly at the temperature of 370
C,
by observing the highest concentration of C3H8. Methane is formed
primarily from the methanation of CO2 and CO, and was also
increased with increasing temperature. Ethane is produced mostly
from the cracking of propane, and it was decreased with increasing
temperature. The hydrogen consumption was reduced with the
raised temperature, which can be attributed to the growths of DCO
and DCO2 and the reduction of HDO, which can also be proved by
observing the decreases of n-C16 and n-C18
4.1.2. Reaction pressure
The effects of pressure were observed within the range of 550
psig to 925 psig with the fixed reaction parameters: the tempera-
ture of 430
C, H2-to-oil ratio of 750 and WHSV of 2 h1. As seen in
Fig. 4 (a), species of n-C8 to n-C14 increased as the pressure was
raised from 550 psig to 675 psig, and slightly dropped as the
pressure was further increased to 800 psig and 925 psig. The sol-
ubility of hydrogen in the oil was raised as the pressure was
increased. High pressure leads to the strong adsorption onto the
surface of catalyst active sites, which enhances the cracking reac-
tion. This can be observed from the decline of n-C17. The increases of
n-C16 and n-C18 and decreases of n-C15 and n-C17 indicate that high
pressure favors the HDO reaction instead of DCO2 and DCO, which
agrees with the study carried out by Srifa et al. [10]. The conversion
reaches 100% as the pressure reaches 550 psig, but the selectivity of
n-C15 to n-C18 drops by 20% as the pressure was raised from 550
psig to 925 psig, as shown in Fig. 4 (b). To obtain the fuel with a
good cetane number, pressures of 550 psig and 800 psig are the
appropriate reaction conditions. However, since hydrogen is also a
source of energy and requires a certain level of cost, the recom-
mended pressure in this case is 550 psig. The decreases of CO2 and
CO prove the statement of the unfavorable DCO2 and DCO at high
pressure above. The rate of DCO2 drops by 34% as the pressure was
increased from 550 psig to 675 psig. The increase of CH4 was
attributed to the increased rate of methanation with high pressure
[47]. As the pressure increases, as described above, more hydrogen
was dissolved in the oil and this promotes the cracking of C3H8,
leading to the decrease of C3H8 and increase of C2H6 and CH4.
4.1.3. H2-to-oil ratio
The H2-to-oil ratio is defined as the volumetric ratio of hydrogen
carrier gas to the oil feedstock. Theoretically, 3 mol of hydrogen are
needed for saturating unsaturated glycerides with 1 double bond
(e.g. C18:1) within the hydrogenation reaction. In addition, 3 mol of
hydrogen are necessary for cutting off the propane from one glyc-
eride compound within the hydrogenolysis reaction. Finally, the
HDO is the only hydrogen consuming reaction within the deoxy-
genation process and the moles of hydrogen consumption can be
calculated based on the stoichiometric balance. Therefore, for
converting 1 ml of oil feedstock, at least 310 ml of hydrogen is
needed for completing the reaction, in accordance with the fatty
acid profile of the oil feedstock. Additionally, more hydrogen is
required for methanation as well as protecting the catalyst from
coking and from water contamination. In this section, the variance
of H2-to-oil ratios were conducted between 500 and 1250 (Ncm3/
cm3), which are 1.6, 2.4, 3.2 and 4.0 times of theoretical hydrogen
requirement (3e5 times were suggested by Srifa et al. [10]), as
other parameters were fixed such as the reaction temperature of
430
C, reaction pressure of 800 psig and WHSV of 2 h1. As shown
in Fig. 5 (a), as the H2-to-oil ratio was raised from 500 to 750, the
concentrations of n-C8 to n-C14 were enhanced, which indicates the
increased cracking reaction because more hydrogen was dissolved
106 K.-H. Hsu et al. / Energy 164 (2018) 99e111

Fig. 3. Effects of reaction temperature on (a) liquid product compositions; (b) con-
version, selectivity and product yield; (c) gas product compositions.

in the oil. As the H2-to-oil ratio was further increased to 1000,
stronger carrier gas flow rate occurred and resulted in shorter
residence time and less cracking level. The concentrations of n-C8 to
n-C14 were therefore decreased. This can be seen from the
increasing concentrations of n-C16 and n-C18. The conversion rea-
ches 100% as the H2-to-oil ratio goes to 750. The yield approaches
the highest as the H2-to-oil ratio becomes 1000. It is recommended
Fig. 4. Effects of reaction pressure on (a) liquid product composition; (b) conversion,
selectivity and product yield; (c) gas product compositions.
to operate the reaction at the H2-to-oil ratio of 1000, for the pur-
pose of obtaining high quality renewable diesel. As seen in Fig. 5 (c),
DCO2 slightly decreases and DCO slightly increases as the H2-to-oil
ratio was enhanced. This can be reflected from the concentrations
 K.-H. Hsu et al. / Energy 164 (2018) 99e111
Fig. 5. Effects of reaction H2-to-oil ratio on (a) liquid product composition; (b) con-
version, selectivity and product yield; (c) gas product compositions.
of n-C15 and n-C17 shown in Fig. 5 (a). As the amount of hydrogen
increases, propane cracking becomes stronger which results in the
increase of ethane and decrease of propane. As described in the
previous studies [10,47], methanation was affected by H2-to-oil
107
ratio. As seen in Fig. 5 (c), the concentration of CH4 in the product
gas was increased with increasing H2-to-oil ratio, which is in
agreement with those researches. Hydrogen usage should be
considered minimizing because it is also a source of energy.
Therefore, the H2-to-oil ratio higher than 1250 was not concerned
in this study.
4.1.4. WHSV
WHSV is defined as the ratio of feedstock volumetric flow rate to
the weight of the catalyst within the reactor. It is an indicator of
reaction efficiency and life expectancy of the catalyst [45]. Exceed
number of WHSV results in short contact of feed to the catalyst and
the occurrence of white waxy intermediates (unreacted FFA) in the
product [ 48 ]. In this part, the experiment was conducted with the
variance of WHSV ranging from 1 to 4 h1 with other parameters
fixed (the temperature of 430
C, the pressure of 800 psig and the
H2-to-oil ratio of 750). It can be seen from Fig. 6 (a) that increasing
WHSV from 1 h1 to 2 h1 results in slightly higher propane
cracked from long-chain alkanes (shown in Fig. 6 (c)) due to the
lower vapor-liquid equilibrium and enhanced n-C8 to n-C14 species.
As WHSV increased, less residence time leads to less cracking re-
action and lower concentrations of n-C8 to n-C14 alkanes. In addi-
tion, increasing WHSV has little effect to HDO, observed from the
concentrations of n-C16 and n-C18, but decreased the performances
of DCO2, observed from the concentrations of n-C15 and n-C17
shown in Fig. 6 (a) and CO2 in Fig. 6 (c). The conversion maintains at
100% with all WHSVs but the yield drops with increasing WHSV, as
seen in Fig. 6 (b). It is interesting to see that increasing WHSV from
2 h1 to 4 h1 has almost no influence to the main product yield. As
shown in Fig. 6 (c), WHSV has no significant effects on methanation
and propane cracking, as seen from the concentrations of CH4 and
C2H6. The increase of propane from WHSV of 1 h1 to 2 h1 was
suspected to be from cracking reactions. It is noted that DCO
increased with increasing WHSV, which was expected to consume
more hydrogen. For the purpose of obtaining high productivity
with fixed catalyst while maintaining the quality of the produced
fuel, it is suggested to operate the process with WHSV of 3 h1 to 4
h1.
4.2. Properties of liquid product
The properties of renewable diesel produced in this study were
measured and compared with traditional NO.2 diesel, as shown in
Table 8. The density at 15
C was shown as 0.79 g/ml, slightly less
than traditional diesel fuel. The dynamic viscosity, which affects the
characteristics of fuel spray inside the diesel combustor and
furthermore influences the quality of combustion, was measured as
2.4 cSt, within the range of traditional diesel. The flash point is less
than 40
C and heating value is 46.4 MJ/kg, higher than the one of
traditional diesel fuel. Based on the properties measured above, it
can be expected that the produced renewable diesel has better
combustion characteristics and potentially has better horse power
and reduced fuel consumption compared to traditional diesel.
4.3. Economic analysis
The results from economic analysis are shown in Tables 9 and
10. It can be seen that almost 98% of the operating cost is spent
on material, mostly on the purchases of feedstock, hydro-
processing catalyst and hydrogen. The utilities, including elec-
tricity and source of heating duty, cost $5.6 MM annually. The
maintenance, included in the other overhead section, accounts for
$0.26 MM per year. For the plant capacity of 600 metric tons per
day, the TCI and direct cost are $19.50 MM and 6.8 MM, respec-
tively. The annual operating cost was calculated as $301 MM,
108 K.-H. Hsu et al. / Energy 164 (2018) 99e111

$16.02 MM annually, allocated as $0.09/L of the target fuel.

Fig. 7 illustrates the price differences with various prices of
hydrogen, since hydrogen is important for hydro-processing and
further hydro-cracking. The prices of hydrogen were obtained from
Zhang et al. [ 51 ], Pedersen et al. [26], Thailand, Malaysia and
Taiwan, which show the prices of $3.33/kg, $3.51/kg, $4.14/kg,
$4.33/kg and $5.44/kg. The prices of hydrogen in Thailand, Malaysia
and Taiwan are from vendor quotations. With the lowest hydrogen
price mentioned in Zhang et al., the MRDP can be as low as $1.66/L,
4% lower than the one obtained for the Taiwan case. In addition,
since hydrogen is also a source of energy, reducing the consump-
tion of hydrogen is considered not only economical, but energy
sustainable. The effects of plant capacity were demonstrated in
Fig. 8. It can be seen that a significant drop was found as the plant
capacity is increased from 600 tonnes/day to 1000 tonnes/day.
However, the MRDPs are quite stable as the plant capacity was
increased from 1500 tonnes/day to 2500 tonnes/day. The MRDP
drops for 0.6% as the plant capacity increases from 600 tonnes/day
to 2500 tonnes/day. The high productivity of renewable fuel with
the increasing plant capacity can offset the increased costs from
feedstock, hydrogen and hydrogen. Further increase of plant ca-
pacity is not allowable since the imported palm oil is unavailable at
this level. Fig. 9 shows the variation of MRDPs with various oper-
ating conditions conducted experimentally, as described in the
previous section. As shown in Fig. 9 (a), MRDP increases from $1.65/
L to $1.74/L as the reaction temperature increases from 370
C to
460
C. Higher temperature results in more cracking species and
reduces the yield of the target product, which increases its selling
price. Enhancing the pressure from 550 psig to 800 psig increases
the price for 3%. The reaction pressure has less effect on the pro-
duction of renewable diesel therefore has insignificant influence on
the MRDP. However, the higher reaction pressure and temperature
increase the difficulties of reactor manufacturing and the con-
straints of their regulations, resulting in higher equipment costs. As
shown in Fig. 9 (c), with the same feeding rate, higher WHSV leads
to lower loading of catalyst and reduces the cost on the hydro-
processing catalyst. The maximum MRDP ($2.17/L) was found at
WHSV of 1 h1, around 30% higher than the one at WHSV of 4 h1.
Further increase of WHSV might result in the occurrence of more
unreacted species due to the insufficient reaction time, which
lowers the product yield and increases the MRDP. Increasing the
H2-to-oil ratio not only enhances the production cost of the process
but also causes the energy imbalance since hydrogen is also an
energy source. The MRDP is increased for 14% as the H2-to-oil ratio
was raised from 500 to 1250, mainly due to the change of pro-
duction yield and the increase of hydrogen cost. Note that the
change of operating conditions varies the selectivity of heavy and
light species in the product, which effects the usages hydrogen and
catalyst in the hydro-cracker. This additionally influences the pro-
duction cost. Fig. 10 shows the MRDPs with different catalysts for
producing renewable diesel. The product distributes apparently
different with the catalysts of NiMo/g-Al2O3 and Pd/C. As referred
from the literature [10], Pd/C is more effective on converting fatty
acids and NiMo is more effective on converting glycerides. Different
product distributions lead to the differences of product yield and
MRDPs. In addition, according to the quotations from the local
vendor, the price of Pd/C is currently 2.5 times of NiMo/g-Al2O3,
which obviously affects the production cost and MRDP. It can be
Fig. 6. Effects of reaction WHSV on (a) liquid product composition; (b) conversion,
selectivity and product yield; (c) gas product compositions.
seen that 20% variation of MRDP was found in-between NiMo/g-
Al2O3 and Pd/C. The techno-economic research conducted in this
study was compared with other studies carried out by Cho et al. [ 50
] and Glisic et al. [14], including the productions of renewable diesel
accounting for $1.67/liter renewable diesel. The MRDP was calcu-
and transesterification biodiesel, as shown in Table 11. It can be
lated as $1.72/liter renewable diesel. The co-product credits, which
observed that the MRDP produced in Taiwan is $0.36 higher than
obtained from the sales of propane and heavy oil, deducted
the prices of hydro-processed diesel found by Glisic [14] and $0.25/
 K.-H. Hsu et al. / Energy 164 (2018) 99e111 109

Table 8
Comparison in properties of Renewable Diesel and Diesel.

Test items Renewable Diesel Unit Testing method

Diesel

Density @15
C 0.79 0.82 g/ml ASTM D4052 & CNS

14474 & NIEA
A204.72C
Dynamic viscosity 2.4 2e4.5 cSt ASTM D445 & CNS
@40
C 3390
o
Flash point <40 <52 C ASTM D93 & CNS
3574
Heat value 11097 10200 cal/g ASTM D240 (gross)

Table 9
Annual operating cost for renewable diesel production.

Item Cost (million $US per year)

Material 310.30
Utilities 5.55
Labor 0.88
Other overhead 0.26

Table 10
Cost analysis of renewable diesel production.

Cost Allocated
MRDP ($/liter)

TCI (million $US) 19.50

Operating cost (million $US per year) 300.99 1.67
Renewable Diesel yield (million liter per year) 178.13
Minimum Renewable Diesel Selling Price ($US/liter) 1.72
Co-product credit (million $US per year) 16.02 0.09

Fig. 7. MRDPs with various hydrogen prices [26, 49 ].

Fig. 9. MRDPs with various operating conditions. (a) Temperature; (b) Pressure; (c)
Fig. 8. MRDPs with various plant capacities. WHSV; (d) H2-to-oil ratio.
110 K.-H. Hsu et al. / Energy 164 (2018) 99e111

DCO2. According to the fuel quality demand of Taiwan, the rec-

ommended reaction temperature, pressure, H2-to-oil ratio and
WHSV are 400
C, 550 psig, 1000 and 3e4 h1, respectively, which
gave the conversion, selectivity and GC yield of 85.4%, 81.5% and
69.5%. In addition, based on the fuel property tests, it is expected
that the produced hydro-processed renewable diesel has higher
engine horse power and lower fuel consumption compared to
traditional diesel.
The lab-scale experimental conditions and results were loaded
into process simulation and the techno-economic analysis was
conducted in accordance with the mass and energy flows obtained
from process model. The information of capital and operating costs
was obtained locally in accordance of Taiwan situation. The MRDP
was calculated to be $1.72/L, with the plant capacity of 600 tonne
per day, production of 178.3 MM L per year and base year of 2017.
Fig. 10. MRDPs with different catalysts. Reducing the hydrogen price from $5.44/kg to $3.33/kg provides 6
cents decrease of MRDP. Increasing the plant capacity from 600
Table 11
tonne to 2000 tonne per day reduces the MRDP by 1 cent per liter.
Comparisons with other studies. The MRDP is increased as increasing temperature, pressure and H2-
a
to-oil ratio and decreased when increasing WHSV. The use of NiMo/
This study Cho et al. [ 50 ] Glisic [14] Glisic et al. [14]
g-Al2O3 catalyst obtains 20% reduction of MRDP compared to that
Location Taiwan Korea Serbia Serbia of Pd/C. It is expected that applying waste cooking oil as the feed-
Feedstock Palm Oil Palm fatty acid Waste Waste vegetable
stock will further reduce the MRDP of hydro-processed renewable
distillate vegetable oil
oil diesel according to the situation in Taiwan.
Process Hydro- Transesterification Hydro- Transesterification
processing processing
Acknowledgement
Conversion 99 N/A 96 95
(wt%)
WHSV/LHSV 2 N/A 1 N/A This project was supported by the Ministry of Science and
(h1) Technology, Taiwan, through grant 104-2628-E-006-007-MY3. The
H2-to-oil ratio 500 N/A 1068 N/A
authors are thankful to CSC Research & Development Division T72
Feedstock cost 805 608 260 260
($US/tonne)
of China Steel Corporation for the support of Aspen Plus™.
Plant capacity 202 120 100 100
(kt/yr)
References
Price of the 1.36 0.91e1.11 0.81e1.00 0.67e0.72
product
[1] Neste. What is the difference between renewable diesel and traditional bio-
($/kg)
diesel - if any?. 2016 [06/01/2017]; Available from: https://www.neste.com/
a
Note: Alkali catalysed process for producing biodiesel. en/what-difference-between-renewable-diesel-and-traditional-biodiesel-if-
any.
[2] Kalnes TN, et al. A technoeconomic and environmental life cycle comparison
of green diesel to biodiesel and syndiesel. Environ Prog Sustain Energy
$0.64 higher than the prices of transesterification biodiesel found 2009;28(1):111e20.
by Cho [ 50 ] and Glisic [14]. The main reason is the costs of feed- [3] Gill SS, et al. Combustion characteristics and emissions of FischereTropsch
stock, which are $197 and $545 per metric tonne higher than the diesel fuels in IC engines. Prog Energy Combust Sci 2011;37(4):503e23.
[4] Aatola H, Larmi M, Sarjovaara T. Hydrotreated vegetable oil (HVO) as a
ones quoted by Cho [ 50 ] and Glisic [14]. It is expected that the renewable diesel fuel: trade-off between NOx, particulate emission, and fuel
MRDP of hydro-processed renewable diesel in Taiwan can be consumption of a heavy duty engine. SAE Int J Engines 2009;1(1):1251e62.
further reduced as the feedstock is chosen as waste cooking oil, [5] UOP, H. Honeywell green diesel. 2017 [06/01/2017]; Available from: https://
www.uop.com/processing-solutions/renewables/green-diesel/.
which becomes more compatible and favorable compared to [6] Neste. Neste MY Renewable Diesel e high-performing low-carbon biofuel.
traditional diesel and transesterification biodiesel. 2017 [06/01/2017]; Available from: https://www.neste.com/fi/en/neste-my-
renewable-diesel.
[7] Sotelo-Boyas R, Liu Y, Minowa T. Renewable diesel production from the
5. Conclusion hydrotreating of rapeseed oil with Pt/Zeolite and NiMo/Al2O3 catalysts. Ind
Eng Chem Res 2010;50(5):2791e9.
[8] Zhou L, Lawal A. Evaluation of presulfided NiMo/g-Al2O3 for hydro-
The target of this study is to promote the hydro-processed deoxygenation of microalgae oil to produce green diesel. Energy Fuel
renewable diesel locally in Taiwan for the replacements of tradi- 2014;29(1):262e72.
tional diesel and transestification biodiesel. Both experimental and [9] Kiatkittipong W, et al. Diesel-like hydrocarbon production from hydro-
processing of relevant refining palm oil. Fuel Process Technol 2013;116:
techno-economical studies were conducted to evaluate the quality 16e26.
and sustainability of the produced fuel. The conversion, selectivity [10] Srifa A, et al. Production of bio-hydrogenated diesel by catalytic hydrotreating
and yield were determined based on the variance of operating of palm oil over NiMoS 2/g-Al 2 O 3 catalyst. Bioresour Technol 2014;158:
81e90.
conditions such as temperature, pressure, H2-to-oil ratio and [11] Shi F, et al. Recent developments in the production of liquid fuels via catalytic
WHSV. High temperature promotes DCO2 and DCO reactions and conversion of microalgae: experiments and simulations. RSC Adv 2012;2(26):
the concentrations of n-C8 to n-C14 species were increased due to 9727e47.
[12] Priecel P, et al. The role of Ni species in the deoxygenation of rapeseed oil over
stronger cracking. High pressure increases the adsorption of H2
NiMo-alumina catalysts. Appl Catal Gen 2011;397(1):127e37.
onto catalyst due to the high dissolution of H2 into oil, which results [13] Miller P, Kumar A. Techno-economic assessment of hydrogenation-derived
in better performance of HDO compared to DCO2 and DCO. High H2- renewable diesel production from canola and camelina. Sustain Energy
to-oil ratio enhances the performances of HDO, DCO and propane Technol Assess 2014;6:105e15.
[14] Glisic SB, Pajnik JM, Orlovi c AM. Process and techno-economic analysis of
cracking and reduces that of DCO2. High WHSV varies the vapor- green diesel production from waste vegetable oil and the comparison with
liquid equilibrium and reduces the occurrences of cracking and ester type biodiesel production. Appl Energy 2016;170:176e85.
 K.-H. Hsu et al. / Energy 164 (2018) 99e111
[15] Cherry-Burrell Waukesha. WAUKESHA PUMPS, engineering manual. W.
Cherry-Burrell; 2002.
[16] Gulsby JG. Low temperature process for separating propane and heavier hy-
drocarbons from a natural gas stream. Google Patents; 1981.
[17] Campelo* JM, Lafont F, Marinas JM. Hydroconversion of n-dodecane over Pt/
SAPO-11 catalyst. Applied Catalysis A: General; 1998.
[18] Peters MS. Plant design and economics for chemical engineers. fourth ed. New
York: McGraw-Hill, Inc; 1991.
[19] Muhamad H, et al. A gate-to-gate case study of the life cycle assessment of an
oil palm seedling. Trop Life Sci Res 2012;23(1):15e23.
[20] Office of Agricultural Economics. The price of the palm oil in Thailand. 2017
[16/10/2017]; Available from: http://www.oae.go.th/main.php?
filename¼index.
[21] ZIM j International Shipping Lines, C.S., Cargo Services. The price of the
shipping. 2017 [12/31/2017]; Available from: http://www.zim.com/pages/
default.aspx.
[22] CPT Single Window. The import tariff of Taiwan. 2017 [31/12/2017]; Available
from: http://portal.sw.nat.gov.tw/PPL/.
[23] CPC Corporation, T. The price of the natural gas in Taiwan. 2017 [02/11/2017];
Available from: https://new.cpc.com.tw/division/mb/.
[24] Taiwan Water Corporation. The price of the water in Taiwan. 2017 [31/12/
2017]; Available from: https://www.water.gov.tw/mp.aspx?mp¼1.
[25] Taiwan Power Company. The price of the electricity in Taiwan. 2017 [31/12/
2017]; Available from: https://www.taipower.com.tw/.
[26] Pedersen TH, et al. Renewable hydrocarbon fuels from hydrothermal lique-
faction: a techno-economic analysis. Biofuel Bioprod Bioref 2018;12:213e23.
[27] Aplus-ins Industrial CoLtd. The price of the Hydrotreating catalyst (NiMO).
2017 [31/12/2017]; Available from: https://tw.taiwantrade.com/company/
crm/53461348.
[28] Kam Industrial Co, L. Chloride ingot, HClO. 2017 [30/12/2017]; Available from:
http://asia.rotekwater.com/.
[29] CPC Corporation. Price of propane. 2018 [01/12/2018]; Available from:
https://new.cpc.com.tw/division/mb/oil-more8.aspx.
[30] CPC Corporation, T. Positions of employee. 2017 [31/12/2017]; Available from:
http://en.cpc.com.tw/global_content.aspx?ID¼2.
[31] Yu Sheng Industrial Co Ltd. Compressor. 2017 [31/12/2017]; Available from:
http://www.yushern.com.tw/index.html.
[32] Yancheng Hao Kai Machinery Co Ltd. Liquid heater. 2017 [31/12/2017];
Available from: https://ycshkjx.1688.com/.
[33] Qingdao Xintai Pressure Vessel Co, L. Hydrotreating Reactor,Oil storage Tank.
111
2017 [31/12/2017]; Available from: http://www.qdxintai.com.cn/.
[34] Physical Agriculture Machinery Co, L. Liquid pump. 2017 [31/12/2017];
Available from: http://www.wulipump.com/.
[35] Yancheng Chechen Environmental Engineering Co, L. Condenser. 2017 [31/12/
2017]; Available from: https://shop1490029959023.1688.com.
[36] Xinxiang City, C.a.J.F.C., Ltd. Vapor-liquid separator. 2017 [31/12/2017];
Available from: https://huayufilter.cnal.com/.
[37] Taiwan Tube Co, L. Valve. 2017 [31/12/2017]; Available from: www.tvfco.com.
tw.
[38] Beston (Henan) Machinery Co, L. Distillation plant. 2017 [31/12/2017];
Available from: http://www.bestongroup.com.
[39] Qingdao Xintai Pressure Vessel Co, L. Gas storage Tank. 2017 [31/12/2017];
Available from: http://www.qdxintai.com.cn/.
[40] Plant Design & Economics for Chemical Engineers. Working capital. 2017 [31/
12/2017]; Available from: http://highered.mheducation.com/sites/
0072392665/index.html.
[41] Zhang Y, et al. Techno-economic analysis of two bio-oil upgrading pathways.
Chem Eng J 2013;225:895e904.
[42] Cheah KW, et al. Process simulation and techno economic analysis of
renewable diesel production via catalytic decarboxylation of rubber seed
oileA case study in Malaysia. J Environ Manag 2017;203:950e61.
[43] eTax Portal, M.o.F. Tax. 2017 [31/12/2017]; Available from: https://www.etax.
nat.gov.tw/.
[44] Directorate-General of Personnel Administration, E.Y. Operating hours. 2017
[31/12/2017]; Available from: https://www.dgpa.gov.tw/.
[45] Yang Y, et al. Hydrotreating of C 18 fatty acids to hydrocarbons on sulphide
NiW/SiO 2eAl 2 O 3. Fuel Process Technol 2013;116:165e74.
 
[46] Sim a
cek P, et al. Hydroprocessed rapeseed oil as a source of hydrocarbon-
based biodiesel. Fuel 2009;88(3):456e60.
[47] Gao J, et al. A thermodynamic analysis of methanation reactions of carbon
oxides for the production of synthetic natural gas. RSC Adv 2012;2(6):
2358e68.
[48] Anand M, Sinha AK. Temperature-dependent reaction pathways for the
anomalous hydrocracking of triglycerides in the presence of sulfided CoeMo-
catalyst. Bioresour Technol 2012;126:148e55.
[49] Zhang Y, et al. Techno-economic analysis of monosaccharide production via
fast pyrolysis of lignocellulose. Bioresour Technol 2013;127:358e65.
[50] Cho HJ, et al. Techno-economic study of a biodiesel production from palm
fatty acid distillate. Ind Eng Chem Res 2013;52(1):462e8.