Journal Pre-proof

Techno-economic analysis and life-cycle assessment of jet fuels production from

waste cooking oil via in situ catalytic transfer hydrogenation

Elena Barbera, Rustem Naurzaliyev, Alexander Asiedu, Alberto Bertucco, Eleazer P.

Resurreccion, Sandeep Kumar

PII: S0960-1481(20)30991-5
DOI: https://doi.org/10.1016/j.renene.2020.06.077
Reference: RENE 13754

To appear in: Renewable Energy

Received Date: 16 October 2019

Revised Date: 23 April 2020
Accepted Date: 15 June 2020

Please cite this article as: Barbera E, Naurzaliyev R, Asiedu A, Bertucco A, Resurreccion EP, Kumar
S, Techno-economic analysis and life-cycle assessment of jet fuels production from waste cooking
oil via in situ catalytic transfer hydrogenation, Renewable Energy (2020), doi: https://doi.org/10.1016/
j.renene.2020.06.077.

This is a PDF file of an article that has undergone enhancements after acceptance, such as the addition
of a cover page and metadata, and formatting for readability, but it is not yet the definitive version of
record. This version will undergo additional copyediting, typesetting and review before it is published
in its final form, but we are providing this version to give early visibility of the article. Please note that,
during the production process, errors may be discovered which could affect the content, and all legal
disclaimers that apply to the journal pertain.

© 2020 Published by Elsevier Ltd.

CRediT Author Statement
Elena Barbera: Methodology, validation, data curation, writing-original draft; Rustem
Naurzaliyev: Software, methodology; Alexander Asiedu: Investigation, methodology, writing-
original draft; Alberto Bertucco: Supervision, writing-reviewing and editing; Eleazer P.
Resurreccion: Methodology, data curation, writing-original draft; Sandeep Kumar:
Conceptualization, project administration, resources, writing-review and editing.
 1 Techno-economic Analysis and Life-cycle Assessment of
2 Jet Fuels Production from Waste Cooking Oil via in situ
3 Catalytic Transfer Hydrogenation
4
5 Elena Barberaa,*, Rustem Naurzaliyeva, Alexander Asiedub, Alberto Bertuccoa, Eleazer P.
6 Resurreccionc, Sandeep Kumarb
7
a
8 Department of Industrial Engineering DII, University of Padova, Via Marzolo 9, Padova 35131, Italy
b
9 Department of Civil and Environmental Engineering, Old Dominion University, Norfolk, Virginia 23529,
10 United States
c
11 Department of Civil Engineering Technology, Montana State University-Northern, Havre, Montana 59501,
12 United States
13
14
15 Submitted to Renewable Energy
16
17
18 October 2019
19 Revised April 2020
20
21
22
23
24
25
26
27
28
29
30
31
32 *Corresponding author: Elena Barbera
33 E-mail: elena.barbera@unipd.it
34 Tel.: +39-0498275462; Fax: +39-0498275461

1
35 Abstract

36 This work evaluates the feasibility of renewable jet-fuel production from waste cooking oil (WCO) via

37 catalytic transfer hydrogenation (CTH) using isopropanol as hydrogen donor. Results were compared to a

38 commercial hydroprocessed renewable jet (HRJ) fuel technology, employing process simulation-based

39 techno-economic analysis (TEA) and life-cycle assessment (LCA). The two routes were compared in terms

40 of product yield, energy consumption, economic and environmental metrics, and allocation methods. The

41 total capital expenditure of CTH plant (7.3M$) was significantly lower than that of HRJ ($149.7M$). The

42 annual revenues were comparable (∼150M$/year), due to similar fuel yields. To be profitable, the liquid fuel

43 should be sold at $3.00/gal and $1.67/gal for CTH and HRJ, respectively. The cumulative fossil energy

44 demand (CED) of HRJ was 1.6 times that of CTH and the total 100-year GWP of CTH was 8% less than

45 HRJ’s, with both systems not sequestering CO2 through co-product offsets. Mass-, energy-, and market-value

46 allocations were utilized. Sensitivity analysis indicated that both systems were driven by transportation

47 factors and not process inputs. Trend analysis on CTH’s energy-return-on-investment (EROI) showed that

48 wide improvements could be made in energy efficiency (EROI=10.30-11.30). From an

49 investment/construction perspective, CTH (95% cheaper) appears to outperform HRJ at similar revenues.

50

51 Keywords

52 Renewable jet fuels; waste cooking oil; process simulation; catalytic transfer hydrogenation

2
53 1. Introduction

54 The progressive depletion of fossil fuels, along with its environmental impacts, has driven the scientific

55 community to search for sustainable and renewable energy sources. The search has been compelled by the

56 burgeoning global population at 0.9% annual rate, reaching over 9 billion by 2050, resulting to 29% increase

57 in world energy consumption [1]. One of the highest energy-consuming sectors is transportation. In the U.S.,

58 92% of transportation fuels are petroleum-derived, while only 5% are biofuels obtained from renewable

59 sources [2]. Among transportation fuels, the development of bio-based jet fuels is of tremendous interest.

60 According to the U.S. Energy Information Administration, an increase of approximately 45% in jet fuels

61 consumption is expected by 2040 [2]. About 278 billion liters of jet fuels were consumed by commercial

62 operators in 2016, leading to 781 million tons of CO2 emissions [3].

63 While other energy sources (e.g. electricity, fuel cells) are being investigated for road transportation, the

64 aviation sector relies on liquid fuels with strict quality requirements to power gas-turbine engines [4]. These

65 are kerosene-range hydrocarbons with C8–C16 carbon atoms. These hydrocarbons can be obtained from

66 renewable sources such as triglycerides. Among the numerous fatty acid sources (mainly animal fats and

67 vegetable oils), waste cooking oil (WCO) is particularly promising. Unlike virgin oils, WCO is not used in

68 food applications and is also a cheaper feedstock. It is largely available in the U.S. and around the globe,

69 with an annual global production of 29 million tons [5]. Moreover, WCO is already largely used in the U.S.

70 for biodiesel production, so that a well-developed transportation and supply infrastructure already exists [6].

71 Several technologies have been investigated for the conversion of fatty acids into jet fuels [4,7]. These

72 involve oxygen removal from the fatty acids via hydrodeoxygenation (HDO) or

73 decarboxylation/decarbonylation reactions using hydrogen. Depending on the nature of catalyst used, its

74 support, and reaction conditions such as temperature and pressure, one reaction pathway can be favored over

75 another [8]. Most commonly used catalysts are sulfided bimetallic materials such as Ni-Mo or Co-Mo, or

76 noble metals like Pt or Pd [9]. Veriansyah et al. [10] investigated the hydroprocessing conversion of soybean

77 oil on different supported catalysts, and found that the most efficient one was NiMo/γ–Al2O3. Clearly, the

78 desired catalyst characteristics are high stability and low cost. For this reason, due to the high cost of noble

79 metals, extensive effort has been directed towards the development of transition metals, and particularly Ni-

80 based catalysts [11]. These catalysts have reached a mature level, and are already widely employed

3
 81 commercially for hydrodesulphurization (HDS) processes [8]. In particular, Hočevar and co-workers

82 reported that the sulfided form appeared the most active compared to the oxide and reduced (metal) forms in

83 the HDO of lignocellulosic derived compounds [12,13]. On the other hand, the catalyst support also plays a

84 major role in the catalyst performances [14].

85 The hydrotreating process for biojet fuels production from triglyceride-based feedstock, known as

86 hydroprocessed renewable jet (HRJ) or hydroprocessed esters of fatty acids (HEFA), is in fact a mature

87 technology and has already been commercialized. For example, Honeywell’s UOP, in collaboration with

88 ENI has successfully commercialized the Ecofining™ process that converts vegetable oils into green jet fuels

89 by deoxygenation [7,15] as shown in Figure 1A. Accordingly, the distillate products are obtained from a

90 series of two-reaction system with an intermediate separation of by-products. Various feedstock types, such

91 as vegetable oils, animal fats, and greases, can be processed using the UOP-Eni Ecofining. The first reactor,

92 the HDO reactor R1, is fed with the feedstock and hydrogen to produce n-alkanes [16,17]. The nature of the

93 product is dictated by the feedstock’s fatty acid composition, catalyst type, and operating conditions. The by-

94 products are separated from the n-alkanes in a flash drum, operated at the reaction pressure. The gaseous

95 stream is a mixture of unreacted hydrogen, CO2, and CO. The liquid is immiscible mixture of organic liquid

96 and water. In the second reactor, the hydroisomerization/hydrocracking (HH) reactor R2, the liquid product

97 is isomerized and cracked in the presence of hydrogen [7,15]. This step is necessary to obtain a kerosene

98 boiling range and jet fuel requirements. Light fuels and residual diesel (if any), are the valuable by-products.

99 The latter might be recycled to increase jet fuel production or directly sold to the market.

100 Despite its maturity level, HRJ has drawbacks associated with the use of hydrogen gas. First, H2 is

101 poorly miscible with oil at atmospheric conditions, thus, 25-100 bar pressure is required to enhance mass

102 transfer. This results in high capital and operating costs. Second, storage and transportation of hydrogen

103 increases cost and creates safety issues due to its high reactivity and flammability. Third, the production of

104 hydrogen in refineries relies on fossil sources resulting into large CO2 emissions.

4
105

106 Figure 1. Block-flow diagram of HRJ (A) and CTH (B) processes for the conversion of triglyceride into
107 biojet fuels
108

109 This study addresses the three issues associated with HRJ by employing catalytic transfer hydrogenation

110 (CTH) as an alternative process to convert WCO into jet fuels (Figure 1B.) [18]. CTH utilizes hydrogen-

111 donating compounds that, in the presence of a catalyst, release H2 to saturate and reduce triglyceride

112 compounds in oils. Suitable hydrogen-donor compounds typically contain hydrogen bonded to elements or

113 groups with similar electronegativity to that of hydrogen itself. Therefore, organic acids, alcohols, amines,

114 and hydrocarbons are suitable hydrogen donors in CTH [19]. Using hydrocarbons as in situ hydrogen donors

115 gives the advantage that the bond energy of C-H is lower than that of the H-H bond in H2 [20]. However,

116 most of these compounds are costly, toxic and volatile. The most available, less toxic and less expensive

117 solvents are common alcohols (methanol, ethanol, n-propanol and isopropanol), which have been reported to

118 be effective hydrogen donors [21]. Methanol has been used to hydrogenate styrene and nitrobenzene [22],

119 while ethanol and 2-propanol were used to deoxygenate levulinic acid over 5% Pd/C [23]. Among these

120 common alcohols, primary alcohols are generally less active than secondary alcohols. The improved

121 hydrogenation ability of secondary alcohols has been attributed to the highest reduction potential of the

122 corresponding ketones or the smaller electron-releasing inductive effect of one alkyl group compared to two

123 alkyl groups [24]. For example, the two methyl groups in 2-propanol donate more electrons to weaken the O-

5
124 H bond compared to the lower electron-donating ability of one methyl group and one ethyl group in

125 methanol and ethanol, respectively. Despite the use of a variety of alcohols, isopropanol remains the most

126 popular donor, because of its simplicity, cheapness, availability, and ease of removal from reaction systems.

127 Therefore, based on the literature, our preliminary studies, and the possibility of generating this solvent

128 from renewable feedstock, we proposed to use and compare the effectiveness of using isopropanol as in situ

129 H2 source for jet fuels production from WCO. To the best of our knowledge, isopropanol has not been used

130 for hydrogenation or deoxygenation of WCO. Hence, in this study, catalytic transfer hydrogenation of WCO

131 to produce jet fuels was analyzed using isopropanol as an H-donor in a continuous-flow reactor over a fixed

132 bed of activated carbon. The experimental study has been previously published by our group [18], proving

133 that the process was effective in producing jet/diesel range fuels. With this work, we aimed at carrying out a

134 detailed, process simulation-based techno-economic analysis (TEA) and life-cycle assessment (LCA) of the

135 proposed CTH process at industrial-scale and compared its market and environmental performance with that

136 of commercial HRJ process. The novelty of this work lies in the fact that CTH of WCO was performed at

137 nearly atmospheric pressure conditions, which cannot be achieved using gaseous hydrogen. Moreover,

138 detailed techno-economic and environmental performance data, which is limited in scientific literature for

139 this novel process, were generated and reported.

140

141 2. Methods

142 Process simulations of renewable jet fuels production from WCO by direct hydrogenation (HRJ via

143 pressurized hydrogen) and CTH were performed using Aspen Plus® software v.9. The Peng-Robinson

144 equation of state and the Non-Random Two-Liquid (NRTL) activity coefficient model were selected as

145 thermodynamic methods. The former was selected as it is suitable to describe the behavior of hydrocarbon

146 systems in high-pressure conditions, while NRTL was used to model specifically the low-pressure vapor-

147 liquid-liquid equilibrium in the products separation following CTH reaction (see section 2.3). The goal of

148 this study was to perform a comparative TEA and LCA (attributional) between HRJ and CTH process. For

149 the TEA, a basis of 1,000 ton/day WCO feedstock was employed because this represents the throughput of a

150 small- to medium- scale refinery in the U.S. [25]. The two processes were compared in terms of economic

151 metrics (internal rate of return (IRR) and net present value (NPV)) and environmental impacts (cumulative

6
152 fossil energy demand (CED) and life cycle greenhouse gas (GHG) emissions). All the assumptions and

153 methods used in the analysis are described in the succeeding sections. The TEA considered a WCO feed

154 flow rate of 1,000 ton/day and the LCA adapted a 1 MJ of usable energy produced embodied in liquid biojet

155 fuel.

156

157 2.1. Waste Cooking Oil Feedstock

158 The fatty acids composition of the WCO feedstock considered is reported in Table 1 [18]. The number of

159 unsaturated bonds in each fatty acid is specified after the length of the carbon chain. Roughly 70 wt% of the

160 oil was composed of unsaturated fatty acids (oleic and linoleic acids), while the remaining compounds were

161 saturated carbon chains. It should be noted that typical waste cooking oils are composed of 80-96 wt%

162 triglycerides (TG), 2-7 wt% diglycerides (DG), 0.3-2 wt% monoglycerides (MG), and free fatty acids (FFA)

163 contents from 1-16 wt% [26–29].

164
165 Table 1. Fatty acids composition of waste cooking oil feedstock.
Compound Formula wt%
C15:0 Pentadecanoic acid C15H30O2 17.9
C19:0 Nonadecanoic acid C H O
19 38 2 11.3
C18:1 Oleic Acid C18H34O2 18.6
C18:2 Linoleic Acid C18H32O2 52.3
166
167
168 2.2. Hydroprocessed Renewable Jet (HRJ) Process

169 The process flowsheet for the HRJ process was built in Aspen Plus, based on the UOP commercial process

170 (Section 1) as shown in Figure 2. The corresponding material and energy balances resulting from the

171 simulations are reported in Appendix A (Table A.1). The WCO feed was pumped and heated up to the

172 operating conditions of the HDO reactor R1 (HDOR1) together with H2, which was fed at 2.6 wt% ratio with

173 respect to WCO [30]. The HDO unit was modelled as a stoichiometric reactor RStoich. For each saturated

174 fatty acid, the stoichiometric reactions of decarboxylation, decarbonylation and hydrodeoxygenation were

175 defined according to:

176 C H O →C H + CO (Eq. 1)

177 C H O +H →C H + CO + H O (Eq. 2)

178 C H O + 3H → C H + 2H O (Eq. 3)

7
179 For the unsaturated fatty acids, additional hydrogen was consumed to hydrogenate the double bonds. The

180 detailed stoichiometry is reported in Table S1. The operating conditions and fatty acids conversions for the

181 three reactions were taken from the works of Chu et al. [30] and Veriansyah et al. [10], which analyzed the

182 conversion of vegetable oil into fuels at 400 °C, 92 bar, and 2-h reaction time over nickel-molybdenum

183 catalyst. Accordingly, the extent of reaction was set equal to 0.68 for decarboxylation, 0.03 for

184 decarbonylation and 0.29 for hydrodeoxygenation, with overall oil conversion equal to 1. The hydrogen

185 consumption associated with the cleavage of triglycerides backbone (3 moles of H2 for each triglyceride

186 producing the corresponding free fatty acids) and the corresponding formation of 1 mole of propane was also

187 considered (Table S4), according to the reaction pathway proposed by the same authors [10,30].

188
189
190 Figure 2. Process flowsheet for the HRJ process. Black solid lines indicate material flows while gray dashed
191 lines indicate energy/heat flows.
192
193
194 The gaseous products (CO, CO2, H2 and H2O) might give rise to Water Gas Shift (WGS) and

195 Methanation reactions. However, Veriansyah et al. [10] reported that no methanation was observed over Ni-

196 Mo catalyst under the experimental conditions investigated. To allow for possible WGS, a REquil reactor

197 unit (WGS-HDO) was added to calculate the gas phase equilibrium reaction (Eq.4), which was evaluated at

198 the same operating conditions (400°C and 92 bar).

199 CO + H O ↔ H + CO (Eq. 4)

8
200 A high-pressure flash was used to separate the liquid and gaseous products, which were cooled to 40 °C,

201 making the water and hydrocarbons completely immiscible. Hence, a 3-phase adiabatic flash (HIGH-P-F)

202 was used to achieve complete separation of water, organic liquids, and gases (Table S6).

203 The organic liquid products were then sent to a hydrocracking unit (HCC) that cracked and isomerized

204 the paraffins to obtain the desired product mixture quality at 350 °C and 90 bar [31,32]. Because of the

205 complexity of modelling hydrocracking reactions, the reactor was simulated with a RGibbs model,

206 considering a total of 358 hydrocarbon components from the Aspen Plus database, including all isomers

207 ranging from C1 to C18 carbon atoms. The validity of this approach was verified by checking the product

208 yields (LPG, naphtha, kerosene and diesel ranges) with those reported in the literature [30,33–35]. The

209 amount of H2 required for hydrocracking was evaluated based on the properties of the feed, namely °API

210 gravity and Watson factor (Table S7), and was equal to 84.44 Nm3 of H2 per m3 of oil (Table S8) [35].

211 Hydrogen is typically fed in large excess to absorb the heat of reaction by direct quench at different stages of

212 the reactor. Therefore, a higher amount of inert hydrogen is required in the reactor depending on the heat

213 produced. The amount of hydrogen required for quench was estimated through an energy balance to keep the

214 difference between inlet and outlet of reactor in the maximum range of 25°C (assuming adiabatic operation).

215 This amount of H2 was set as inert fraction in the RGibbs unit, so it was not accounted for in the equilibrium

216 calculation. HCC process input parameters in Aspen Plus are shown in Table S9.

217 The products from the HCC reactor were separated by means of two flash units in series. The first one

218 operated at high pressure (HP-SEP) to recover the unreacted hydrogen, while the second one operated at low

219 pressure (LP-SEP) to separate C1-C4 gaseous products from the liquid mixture, which was to be sent to an

220 atmospheric fractionation tower to recover the different product fractions. The higher the operating pressure

221 of the LP-SEP unit, the higher the recovery of liquid fuels (Table S14). However, it was preferred to operate

222 the flash at 1 bar so that a higher amount of propane C3H8 (94%) was recovered in the gaseous stream, to be

223 used as additional fuel.

224 The liquid fuels mixture was then sent to an atmospheric fractionation tower (FRAC) to recover the

225 different product fractions, i.e. naphtha and kerosene. Distillation was modelled as a RadFrac unit, operating

226 with 30 ideal stages and reflux ratio of 0.8.

9
227 The large excess of H2 recovered from the HIGH-P-F and HP-SEP flash units is valuable, so that high

228 recovery from the gaseous streams was desired to maximize recycle. The gas mixture separated from the first

229 flash unit was characterized by large amounts of CO and CO2, while the one recovered after the

230 hydrocracking unit mostly comprises light hydrocarbons, such as CH4, C2H6 and C3H8. Industrially, the

231 recovery of hydrogen is performed by Pressure Swing Adsorption (PSA), with operating pressures varying

232 between 10-40 bar [33,36]. As a higher recovery pressure results in lower compression loads required for H2

233 recycling, 40 bar was chosen for the simulations. For the purposes of the simulations, the PSA units were

234 modelled by simple SEP blocks assuming complete H2 recovery. The H2 make-up stream was compressed up

235 to 40 bars by means of a 3-stage compressor (C2, compression ratio=3.14, which considered the final

236 temperature of each stage, not exceeding 225 °C for H2), with inter-stage cooling at 40 °C. The make-up and

237 recycled H2 streams were then mixed and compressed up to the final pressure of 92 bar in a second single-

238 stage compressor (C1) and delivered to the HDO and HCC reactors respectively.

239 Finally, pinch analysis was performed in order to optimize heat integration (Section 1.6 of SI). The

240 optimized heat exchanger networks are displayed in the process flowsheet (Figure 2), where integrated heat

241 exchangers are connected by energy streams (dashed lines). Accordingly, the external process utilities

242 needed consisted of a fired heater to achieve 400 °C prior to the HDO reactor, and cooling water to cool the

243 products at 40 °C in the units CW1 and CW2. The column reboiler and condenser duties were also provided

244 by external utilities.

245 2.3. Catalytic Transfer Hydrogenation (CTH) Process

246 The process flowsheet of the CTH process is shown in Figure 3, and the corresponding material/energy

247 balances are reported in Appendix A (Table A.2). The simulation was based on the block flow diagram of

248 Figure 1B and on the experimental data obtained in the laboratory using a fixed-bed tubular reactor filled

249 with charcoal catalyst [18]. The WCO feedstock was mixed with isopropanol (ISO-P) at a flowrate of 341.8

250 ton/day (ISO-P/WCO ratio = 0.3418), and heated up to the CTH reactor operating temperature (380 °C),

251 found to be the optimal value [18]. The CTH operating pressure was set to 2 bar, and the reaction was

252 simulated using the RYield model. The product yields, defined as mass of product over inlet mass of

253 reagents, were calculated based on the experimental material balance results (Table S24).

10
254

255
256 Figure 3. Process flowsheet for the CTH process. Black solid lines indicate material flows while gray dashed
257 lines indicate energy/heat flows.
258
259 The gaseous products considered were H2, H2O, CO, CO2, CH4, C2H6 and C3H8, which were formed

260 either by decarboxylation, decarbonylation, or hydrodeoxygenation reactions as well as by isopropanol

261 decomposition. Concerning the liquid products, for the sake of simplicity, only three reference components

262 were taken as representatives for the naphtha, kerosene and diesel range hydrocarbons, respectively. Based

263 on the average properties (boiling point and molecular weight) of the experimental products distribution,

264 C8H16, C12H26, and C17H36 were selected as the three respective liquid fuel fractions (Table S18-S23).

265 Because the reaction leads to coke formation, catalyst regeneration was included by simulating the coke

266 removal using a SEP unit. The composition of the outlet stream from CTH reactor was characterized by a

267 mixture of hydrocarbons (C8H16, C12H26, C17H36), water (H2O) and gaseous products (CO2, CO, C1-C3). The

268 separation of hydrocarbon mixture from gases and water could be achieved at low temperature, since the

269 solubility of water is inversely dependent on temperature. Accordingly, a 3-phase flash model was used to

270 simulate the required separation (VLL-SEP) at atmospheric pressure, after cooling the products to 25 °C. For

271 a correct estimation of the vapor-liquid-liquid equilibrium at low pressure, NRTL was used as

272 thermodynamic model. The solubility of the incondensable compounds (CO, CO2 and H2) was determined

273 from the Henry constants, available in the Aspen Plus database. The validity of the binary interaction

274 parameters of the H2O-hydrocarbon systems was first verified against experimental data (Table S26-S27).

275 Finally, the three liquid fuels fractions were separated by means of two atmospheric distillation columns in

276 series: the first one (FRAC1) separated the naphtha from the heavier fractions, which were then separated in

11
277 the second one (FRAC2). The columns specifications were, respectively: 10 ideal stages and reflux ratio =

278 0.2 for the first separation; 15 ideal stages and reflux ratio = 0.2 for the second one.

279 Heat integration was also carried out between the cold feed and the hot products, as shown by the energy

280 stream (dashed line) in Figure 3. By analyzing the vaporization range of the WCO-isopropanol feed, the feed

281 outlet temperature from the integrated heat exchanger was set to 300 °C. As most of the WCO vaporizes in

282 the range of 300-350 °C (Figure S8), it was preferred to avoid the phase change within the heat exchanger,

283 and to supply the latent heat duty by means of a fired heater. Accordingly, the external utilities were

284 represented by the fired heater required to reach 380 °C and refrigerated cooling water (available at 5 °C) to

285 cool down the products to 25 °C, in addition to the reboiler and condenser duties of the two distillation

286 columns.

287 2.4. Techno-economic Analysis

288 For a 1,000 ton/day WCO feed flow rate (approximately 270,000 gal/day), the economic and profitability

289 analysis of both HRJ and CTH processes were performed according to the method proposed by Towler et al.

290 [37] for chemical plants. Accordingly, the Fixed Capital Investment (FCI) and the Cost of Manufacturing

291 (COM) were estimated together with the revenues. The FCI consists of the capital expenses of the plant

292 including construction and engineering costs (CAPEX), contingency charges, and offsite infrastructure

293 investment (OSBL) (Section 5.1.1 of SI). The CAPEX were estimated based on the Bare Module Cost of

294 each equipment unit, evaluated using the correlations proposed by Guthrie [38]. According to this method,

295 the purchase cost of the equipment was calculated as a function of the size, as well as the construction

296 material and the operating pressure, which become particularly relevant for the HRJ process, characterized

297 by high pressures. Furthermore, the material chosen for HDO and HCC reactors were stainless steel (SS)

298 because of hydrogen gas in compatibility with respect to the normal carbon steel (CS) [39]. Moreover, to

299 account for inflation, all the costs were referred to the year 2017 using the Chemical Engineering Plant Cost

300 Index (CEPCI) [40].

301 A rigorous sizing procedure was applied to determine the size of the HDO and HCC reactors for the HRJ

302 process, and that of the CTH reactor. A Trickle Bed Reactor (TBR) composed of adiabatic multi-stage beds

303 was selected as HDO reactor [41]. Considering the properties of the feed, catalyst hold-up of 60%, and liquid

304 residence time of 2 h, the resulting reactor was determined to have 4 stages, each of diameter D = 3.048 m

12
305 and H = 8.39 m, for a total volume of 245 m3 (Section 4.1.1 of SI). The HCC reactor was sized as a multi-

306 stage packed bed reactor, such as those used at commercial level. Taking a typical gas residence time of 1 h,

307 the reactor volume equaled 392.5 m3, with 5 stages of D = 3.81 m and H = 6.9 m [42] (Section 4.1.2 of SI).

308 Finally, a packed bed reactor was selected also for the CTH reactor. The scale-up was based on the

309 experimental value of the weight hourly space velocity (WHSV) of 6.8 h-1. The resulting volume, assuming

310 60% catalyst hold-up, was equal to 27.4 m3 (Section 4.2.1 of SI). Given the high coke formation, 2 parallel

311 reactors were considered, in order to allow a continuous operation when performing catalyst regeneration.

312 The sieved tray distillation columns were sized according to the Fair Method [43]. In particular, the

313 distillation tower of the HRJ process resulted having a diameter D = 1.8 m and H = 15.7 m, which allowed

314 operating at 74% of the flooding condition. The diameters and the heights of columns of the CTH process

315 were H = 17.5 m, D = 1.35 m and H = 17.5 m, D = 1.7 m, respectively (approximately 78% of the flooding).

316 In addition, the capital cost related to building an internal H2 production facility by Steam Methane

317 Reforming (SMR) was considered for the HRJ process [37]. The capital cost of the PSA unit was instead

318 determined using a capacity factor with respect to a similar reference unit [36]. According to the simplified

319 procedure employed, the OSBL were estimated as 50% of the total CAPEX (with reference to atmospheric

320 pressure purchase costs), while contingencies amount to 15% of the actual CAPEX [37].

321 The COM considers both direct and indirect manufacturing costs, as well as general expenses related to

322 business administration. Direct manufacturing costs are directly correlated with the production rate and

323 comprise raw materials and utilities or operations expenses (OPEX), as well as staff and operating labor. The

324 OPEX were evaluated based on the results of the process simulations. For the estimation of natural gas utility

325 in the fired heater unit, an efficiency of 80% of the thermal system was assumed. Moreover, the utility used

326 for cooling the CTH reactor was hot oil or molten salts due to the high operating temperature, i.e. 380 °C.

327 The price of this utility was estimated based on the reference for Hot Oil/Molten Salts for reactor cooling

328 [44]. More specifically, the cost was based on the price of natural gas that would be required to heat the hot

329 oil/molten salts. The number of operators was calculated based on the number of equipment units and was

330 equal to 29 for the HRJ and 26 for the CTH processes. Indirect manufacturing costs and general expenses

331 were instead estimated as percentages of the FCI [35,37].

13
332 The plant was assumed to operate for 8000 h/year, i.e. a stream factor of 0.913 was applied. The

333 economic profitability of the processes was evaluated by means of discounted cash flow analysis, according

334 to four criteria: discounted payback period (DPBP), net present value (NPV), internal rate of return (IRR),

335 and present value ratio (PVR). The following additional assumptions were made [45]: first, the useful life of

336 the plant was taken equal to 25 years, the first 2 of which were used for construction and start-up, while full

337 regime production started at year 3. In particular, 70% of the capital was invested at year 1, while the

338 remaining 30% in year 2. Second, the discount rate was taken as an interest rate of 8%, hence, it is not based

339 on loan interests or debt ratio. Third, depreciation was evaluated according to the Modified Accelerated Cost

340 Recovery System (MACRS). The depreciation period was 7 years, starting at the beginning of the 3rd and

341 ending at the end of the 10th year. In particular, the double declining balance was applied for the first 5 years

342 after the plant start-up, after which it was switched to the straight-line method. No salvage value of the plant

343 was assumed. Thus, the total capital to be depreciated equaled the fixed capital investment (FCI). Fourth, the

344 income tax rate was assumed to be 35% of the gross profit. Fifth, the cost of the land was not considered, as

345 it was not relevant for the sake of comparison between the profitability of the two processes. Finally, annual

346 operators’ salary was taken from U.S. Bureau of Labor Statistics [33].

347 The revenues were evaluated based on the wholesale prices of the refinery products, which included the

348 production costs and refiner profit, while distribution and transportation costs and taxes were not considered.

349 The wholesale prices corresponding to the different refinery products (naphtha, kerosene, and diesel) were

350 retrieved from the U.S. Energy Information and Administration [46], and were equal to 0.426 $/L, 0.443 $/L,

351 and 0.440 $/L, respectively, as average prices for 2017. The large amount of fuel gases produced by both the

352 HRJ and CTH process was assumed to be sold at the same price as natural gas (0.13 $/m3), considering that

353 the lower heating value (LHV) of these gases was very similar to that of natural gas.

354 2.5. Life Cycle Assessment

355 2.5.1. Goal and Scope

356 The comparative environmental impacts of HRJ and CTH processes using WCO as feedstock were evaluated

357 in terms of CED and life cycle GHG emissions. It was assumed that one biorefinery was located in any

358 medium-sized U.S. city (population: 100,000-300,000 persons) [47]. A biorefinery serves one division for

359 each U.S. region (e.g. one biorefinery for Division 2 [Mid-Atlantic] of Region 1 [Northeast], which includes

14
360 New Jersey, New York, and Pennsylvania). WCO was assumed to be collected from restaurants, hotels,

361 casinos, malls, and other food providers operating in these cities. The volume of WCO generated from

362 household sources was insignificant relative to the food service industry; restaurants still dominate total

363 spent oil output. These cities are ideal test locations for a WCO-based refinery for the following reasons.

364 First, they are affordable, offering abundant opportunities for business investments. Second, these cities have

365 numerous restaurants that can provide appreciable amount of WCO. They have high average number of

366 restaurants per 1,000 people compared to the national average of 1.52 for cities whose population is greater

367 than 50,000 [48]. Third, these cities are not as busy as megacities which cuts transportation costs yet were

368 assumed to be located within 150-mile maximum radius from a Metropolitan Statistical Area (MSA) (>1

369 million population) to ensure reliable supply of spent oil. Examples of these medium-sized U.S. cities are

370 Pittsburgh (2019 population: 302, 908) serving Division 2, Region 1; Greensboro, North Carolina (2019

371 population: 292,265) serving Division 5, Region 3; or Fremont, California (2019 population: 238, 281)

372 serving Division 9, Region 4 [36]. Based on the 270,000 gal/day (98.55 Mgal/year continuous operation)

373 TEA basis, the nine biorefineries in the U.S. will require 8.87 x 108 gal/yr. The U.S. generates more than

374 enough WCO at 3.8 Mton WCO annually (9.92 x 1011 gal/yr) [49], with 2.84 Mton yellow and other grease

375 [50].

376 The system boundary for the analysis was “well-to-tank” (WTT), in which all associated energy and

377 material flows were determined and quantified for each unit operation:

378 hydrogenation/hydrocracking/separation for HRJ and hydrogenation/separation for CTH. Fuel combustion

379 was excluded in the analysis. The WTT system boundary starts with the acquisition of all raw materials up to

380 distribution of products to end-users (e.g. airports). For ease of analysis, HRJ was modeled as the base case

381 using the UOP-Eni EcofiningTM process. CTH was considered as the alternative case.

382 2.5.2. Functional Unit

383 The functional unit (FU) chosen for this analysis was 1 MJ/year embodied energy in liquid biojet fuel (main

384 product) produced from either HRJ or CTH process using WCO as feedstock. Such FU is conveniently

385 adapted to provide a quantitative reference on which all calculated materials and energy flows were based.

386 Co-products were light fuels and naphtha from HRJ and naphtha and diesel from CTH.

387 2.5.3. System Boundary

15
388 The HRJ process was divided into eight sub-systems which include WCO transport (WT),

389 hydrodeoxygenation (HY), CO and CO2 flash separation (FS), hydroisomerization and hydrocracking (HH),

390 high pressure flash (HF), low pressure flash (LF), distillation (DT), and fuel transport (FT). Figure 1A

391 represents hydrotreating as HY, separation as FS, hydrocracking as HH, and products recovery as HF, LF,

392 and DT. In the case of CTH, there were six sub-systems identified: WCO transport (WT), catalytic transfer

393 hydrogenation (CH), low pressure flash (LF), distillation 1 (DT1), distillation 2 (DT2), and fuel transport

394 (FT). These sub-systems are represented in Figure 1B as: hydrogenation for CH, and separation for LF, and

395 products recovery for DT1 and DT2. Both HRJ and CTH have identical WT and FT sub-systems.

396 HRJ Sub-System 1: Waste Cooking Oil Transport (WT). This sub-system involved the systematic collection

397 of WCO from restaurants and food service businesses in cities where a biorefinery was to be established.

398 Additional WCO were to be collected from MSA located within 150-mile maximum radius of the biorefinery

399 to meet the 1,000 ton/day biorefinery requirement used in TEA (approximately 270,000 gallons/day WCO).

400 WCO were transported from sources near the biorefinery or from MSAs using diesel trucks with likeliest

401 truck mileage efficiency of 6.5 miles/gallon [51]. Diesel Class 8 heavy trucks were assumed to travel for 6

402 hours in a day at 50 miles/hr. Material input to WT was transportation diesel while material outputs were

403 combustion products CO2 and H2O. Maximum travel distance per truck was 300 miles and the mileage

404 efficiency was taken to be 5.5 miles/gal [52].

405 HRJ Sub-System 2: Hydrodeoxygenation (HY). WCO entered the biorefinery as pretreated spent oil suitable

406 for processing. Depending upon the nature of fatty acids, conversion can be decarboxylation,

407 decarbonylation, or hydrodeoxygenation (Table S1). HY is adapted as general term for hydrotreatment. The

408 chemical reaction was operated at the following conditions: 400 °C, 9.2 MPa and 2 h of residence time [14-

409 15]. A four-stage HY unit was utilized in this sub-system. Material inputs included pretreated crude oil, H2 as

410 hydrogenating agent, and nickel-molybdenum (NiMo) catalyst. The output from HY was a mixture of

411 straight-chain alkanes consisting of tetradecane, pentadecane, heptadecane, and nonadecane. Moreover,

412 gases such as CO, CO2, and propane were generated in addition to water. The amount of total H2 required

413 and propane generated were based on stoichiometry presented in Table S4. The energy consumption for HY

414 was attributed to: (1) pressurization assuming isentropic compression of H2 at 85% efficiency and (2) WCO

16
415 pumping at 95% efficiency. Heat was utilized to raise the oil temperature from 380 to 400 °C, determined as

416 the heat duty to be supplied resulting from heat integration.

417 HRJ Sub-System 3: CO and CO2 Flash Separation (FS). Flash separation was performed to separate the

418 gaseous by-products from the output liquid straight chain alkanes. Separation input parameters for FS were

419 as follows: 9.2. MPa, 10 min liquid hold-up [23], adiabatic operation, and vapor-liquid-dirt water phases,

420 employing the Peng-Robinson property method in Aspen Plus modeling. A single vertical-type separation

421 vessel was used. These gases include CO, CO2, water vapor, propane, and H2. Along with liquid alkanes,

422 sour water was generated. Energy requirement for this sub-system was attributed to the lowering of product

423 temperature from 231 to 40 °C (after heat integration).

424 HRJ Sub-System 4: Hydroisomerization and Hydrocracking (HH). The separated liquid straight chain

425 alkanes were isomerized/cracked to get smaller chains of iso-paraffines. Mild conditions were employed in

426 this sub-system: 350 °C and 9.0 MPa at 1-hour HH reaction time. Detailed process input parameters used in

427 Aspen Plus modeling are tabulated in Table S9. The hydroisomerization and hydrocracking unit comprised 5

428 stages. Properties of feed oil and hydrogen are presented in Table S7 and Table S8, respectively. Material

429 inputs to this sub-system include the alkanes, H2 for isomerization and cracking, and H2 for quenching.

430 Output isoparaffins include N-paraffins (C1-C3, C4-C8, C9-C15, C16-C18) and their methylated

431 counterparts (mono-x-methyl and mono-yy-methyl) (naphtha, jet fuel, and diesel). The compression energy

432 was already accounted for in the HH unit. All input H2 was previously compressed and then split into the two

433 reactors thereby making pressurization energy of this unit to be 0. Direct heat use was also 0 (i.e. reactor is

434 adiabatic), due to heat integration (See Section 1.5 of SI). The heat to reach the operating temperature was

435 recovered from other streams.

436 HRJ Sub-System 5: High Pressure Flash (HF). High-pressure flash was employed to recover H2 and other

437 gaseous compounds from the liquid products. Herein, the process conditions were: 9 MPa, 10-min liquid

438 hold-up, and no heat duty (adiabatic operation, temperature after cooling was stable at 40 °C). There was one

439 vertical-type HF vessel used. Two material streams were produced in this sub-system: (1) the gaseous stream

440 consisting of the recovered H2, CO, CO2, N-paraffins, and methylated paraffins, and (2) the liquid stream

441 consisting of naphtha, jet fuel, and diesel. There were residual H2, CO, and CO2 present in the output liquid.

17
442 Like sub-system FS, the energy requirement was cooling energy use, or the latent heat expelled to lower the

443 product temperature from 196 to 40°C (after heat integration) (exothermic).

444 HRJ Sub-System 6: Low Pressure Flash (LF). Atmospheric sub-system LF was employed following HF to

445 separate light hydrocarbons to be used as fuel gas and residual H2, CO, and CO2 from the liquid fraction in a

446 horizontal vessel. Process conditions for this sub-system were: 0.10 MPa, 10-min liquid hold-up, and no heat

447 duty (temperature was stable at 24 °C). The output liquid was heavy with biojet fuel, diesel, and naphtha

448 with very little CO2 left. There was no cooling energy requirement for this sub-system.

449 HRJ Sub-System 7: Distillation (DT). Final stage of product recovery was atmospheric distillation, to

450 separate the liquid fractions naphtha (C6-C8) from the top and jet fuels (C9-C15) from the bottom. DT

451 process conditions were as follows: 0.10 MPa, 30 stages, feed input at the 15th stage, 0.80 reflux ratio, 164.8

452 °C reboiler temperature, and 65.30 °C condenser temperature. Reflux ratio was optimized by sensitivity

453 analysis to maximize the recovery of N-octane (light-key component) in the distillate. The output distillate

454 (naphtha) was heavy in mono-x-methyl (C5-C8) (44 wt%) while the output residue (biojet fuels) was

455 predominantly mono-x-methyl (C9-C15) (38 wt%) and multi-yy-methyl (C9-C15) (57 wt%). The energy use

456 for this sub-system came from two sources: the heating duty at the reboiler and cooling energy at the

457 condenser, both were function of WCO feed rate.

458 HRJ Sub-System 8: Fuel Transport (FT). Like WT, FT utilized diesel to transport naphtha and biojet fuel to

459 distribution endpoints (i.e. gate). Designated fuel distributors were assumed to be located within 150-mile

460 maximum of WCO collection. The trucks were assumed to have a mileage efficiency of 6.5 miles/gallon.

461 The distance hauled was calculated to be 300 miles/truck-day at 6 h per truck per day at 50 miles/hr. Material

462 input to FT was diesel and material outputs were CO2 and H2O.

463 CTH Sub-System 1: Catalytic Transfer Hydrogenation (CH). CH was modeled based on the experimental

464 work of Asiedu et al. [18]. Hydrogenation and deoxygenation of WCO was accomplished by means of a

465 hydrogen donor solvent. Herein, gaseous H2 was replaced by isopropanol. Details on lab-scale reactor

466 geometry and catalyst properties are presented in Table S16 and experimental conversion at different

467 temperatures are shown in Table S17. From these results, an industrial model was built based on scale-up of

468 experimental reactor and CH reactor material balance at T = 380 oC presented in Figure S5. The reaction

18
469 conditions for CH sub-system were as follows: 380 °C, 2 bar, 6.8 h-1 weight hourly space velocity in a

470 continuous fixed-bed plug-flow reactor (PFR). There were 2 PFRs, each with a total capacity of 27.4 m3, to

471 be operated alternately while the other undergoes catalyst regeneration. Activated carbon catalyst was

472 assumed to be replaced every 90 days. Material input streams consisted of WCO, isopropanol as

473 hydrogenating agent. The main liquid fuel product contained C1-C3 gases, naphtha, diesel (heptadecane),

474 and biojet fuels. Gases were also generated in the reaction: CO, CO2, and H2, in addition to water and coke.

475 Energy was utilized to pressurize WCO and 2-propoanol to 2 bar and heat was used to raise the feed’s

476 temperature from 300 to 380 °C after heat integration.

477 CTH Sub-System 2: Low Pressure Flash (LP). Like HRJ, output liquid from the main chemical reaction

478 (CH) was subjected to LP to separate gases from main products (naphtha, diesel, biojet fuels). However,

479 unlike HRJ, CTH has three output streams: gas (mainly CO and CO2), organic liquid (main products), and

480 aqueous liquid (water and residual main products). Conditions were like HRJ’s sub-system LF, albeit

481 condition was exothermic; heat was expelled from the system to cool the reaction products from 50 to 25 °C.

482 The cooling down to 50 °C was achieved by heat integration.

483 CTH Sub-System 3: Distillation 1 (DT1) and Distillation 2 (DT2). Products recovery involved a series of

484 successive distillations DT1 and DT2, each employing one column. In DT1, naphtha (C6-C8) was recovered

485 from the top, while the heavier fuels (C9-C18) were separated from the residue. The residue was further

486 distilled in DT2 to extract jet fuels as distillate and diesel from the bottom. Process conditions for DT1 were:

487 0.10 MPa, 10 stages, feed input at the 3rd stage, 0.20 reflux ratio, 220 °C reboiler temperature, and 130 °C

488 condenser temperature. Process conditions for DT2 were: 0.10 MPa, 15 stages, feed input at the 9th stage,

489 0.20 reflux ratio, 299 °C reboiler temperature, and 212 °C condenser temperature. For both distillation

490 columns, heat was applied to the reboilers and heat was removed from the condensers.

491 2.5.4. Allocation

492 Allocation is a critical step in LCA as it directly affects results of impacts assessment. Allocation involves

493 the partitioning of the environmental impacts from the process based on products (materials output) flows

494 [53]. Allocation is not a straightforward division of environmental impacts between products and co-

495 products, thus, protocols set forth by ISO 14044:2006 guidelines [54] were adapted. For this analysis, mass-,

496 energy- and market-based allocation methods were necessary. In both HRJ and CTH, the products were
19
497 identified as naphtha/gasoline (C4-C8, e.g. octane C8H18) and jet fuels (C9-C15, e.g. dodecane C12H26)

498 obtained at the very last sub-system (distillation). In some intermediate sub-systems, co-products were

499 generated. For example, the sub-system HF in HRJ produced output liquid (gasoline and jet fuel mixture)

500 that fed into subsequent LF sub-system. Concurrently, it also generated a co-product output gas mixture

501 consisting of short-chain N-paraffins C1-C3 (e.g. ethane, propane), short-chain N-paraffins C4-C8 (e.g.

502 butane, octane), and mono-x-methyl or multi-yy-methyl C9-C15 (e.g. dodecane, hexadecane or cetane),

503 which can be pressurized and distilled into LPG, naphtha, and jet fuel, respectively. In the case of HRJ, the

504 co-products were as follows: FS = propane; HF = [C1-C3, C4-C8, C9-C15] gaseous hydrocarbons; LF =

505 [C1-C3, C4-C8, C9-C15] gaseous hydrocarbons, [C1-C3 (LPG), C16-C18 (diesel)] liquid hydrocarbons, and

506 liquid CO2. For CTH, the co-products were as follows: CH = coke; LF = [C1-C3, octane, dodecane,

507 heptadecane] gaseous hydrocarbons, and [C1-C3] (LPG) liquid hydrocarbons. Table 2 details the energy-

508 and market-based allocation factors for HRJ and CTH processes. Thus, the CED and life cycle GHG

509 emissions for each process were apportioned between products and co-products using mass flows, lower

510 heating values, and unit selling price at prevailing market conditions.

511
512 Table 2. Mass-, energy-, and market-based equivalency (impact)
513 factors used in the impacts assessment phase of HRJ and CTH life
514 cycle. Factors were calculated from Table S69.
Output1 Mass Energy Market5
HRJ Products 0.8176 0.8109 0.0254
Naphtha 0.2934 0.2993 0.0098
Jet fuel 0.5242 0.5116 0.0156
HRJ Co-products 0.1824 0.1891 0.9746
Propane gas 0.0045 0.0047 0.0675
C1-C3 gas2 0.1070 0.1146 0.9043
C4-C83 0.0568 0.0569 0.0022
4
C9-C15 0.0013 0.0013 0.0000
C1-C3 liquid (LPG) 0.0082 0.0087 0.0004
C16-C18 (diesel) 0.0029 0.0028 0.0001
Liquid CO2 0.0017 0.0000 0.0000
CTH Products 0.7899 0.7867 0.0167
Naphtha 0.2428 0.2495 0.0056
Jet fuel 0.4200 0.4129 0.0086
Diesel 0.1270 0.1243 0.0026
CTH Co-products 0.2101 0.2133 0.9833
Coke 0.0307 0.0201 0.0006
C1-C3 gas2 0.1687 0.1821 0.9823
Octane 0.0061 0.0062 0.0003
Dodecane 0.0001 0.0001 0.0000
C1-C3 liquid (LPG) 0.0046 0.0049 0.0002
1
515 Output flow rates obtained from mass balance (tons/day). 2Represented by ethane
516 gas. 3Represented by compressed liquid hexane. 4Represented by compressed liquid

20
517 dodecane. 5Obtained based on wholesale market price in 2019 US dollars excluding
518 taxes.
519
520
521 2.5.5. Inventory

522 Life cycle inventory (LCI) is the most tedious process in LCA as it involves the acquisition of high-quality

523 data essential to accurate environmental impacts assessment. The development of HRJ and CTH models in

524 this study entailed data from various sources and methods. The models were created and developed using

525 Excel spreadsheet with Crystal BallTM add-in functionality. This allows certain material inputs and process

526 conditions to contain probability distributions (Table S70) that were subsequently used stochastically for the

527 Monte Carlo simulation, resulting into probabilistic outputs (“endpoints”). The upstream impacts of

528 electricity and selected materials were calculated using U.S. equivalency factors (or European where U.S.

529 values were not available) presented in Table 3.

530 The hydrodeoxygenation LCI were largely based on the experimental results of Veriansyah et al. [10],

531 appropriately scaled-up to 1,000 tons/day FU. Gaseous H2 was stoichiometrically supplied to crude WCO

532 using with nickel-molybdenum (NiMo) catalyst. The electrical energy for pressurization was determined

533 from Aspen, assuming isentropic compression with H2 (85% efficiency) and continuous pumping of WCO

534 (95% efficiency). LCI data for the rest of the sub-systems were obtained from various literature sources and

535 detailed technical descriptions were reported in Section 1 of SI. Electrical energy for pressurization , heat,

536 and material flows were determined from Aspen modeling. Compressed hydrogen gas was supplied as make-

537 up contribution because part of the non-reacted H2 is recovered and recycled. As for CTH, large-scale LCI

538 were based on the experimental study by Asiedu et al. [18]. Specifically, isopropanol was supplied in large

539 quantity (341.8 tons/day) following scale-up of CTH reaction of WCO at different temperatures (Section 2.1

540 of SI). The electrical energy was applied to the pump to pressurize the WCO and isopropanol to 2 bar. The

541 heat needed to increase the feed’s temperature from 300 to 380 °C was calculated by Aspen’s heat

542 integration. Catalytic transfer hydrogenation output to succeeding sub-systems were based on experimental

543 yields by Asiedu and colleagues. Detailed technical information for CTH LCI data were outlined in Section 2

544 of SI. Section 3 of SI provides Aspen simulation. A full accounting of energy/materials inputs/outputs is

545 presented in Table 4.

546

21
547 Table 3. Equivalency factors used in modeling HRJ and CTH processes.
Impact Value Unit Notes
Electricity, U.S.
Energy use 12.5747 MJ/kWh 1.43% from surface coal mining, 1.85% from transportation,
and 96.72% from electricity generation. Normal, SD = 1. [55]
Water use 0.00004 m3/kWh Hard coal extraction region in the U.S., low-volatile bituminous
coal at LHV = 33,412 kJ/kg. Normal, SD = 1. [56]
CO2 to air 1.0220 kg/kWh 0.94% from surface coal mining, 1.72% from transportation,
and 97.34% from electricity generation. Normal, SD = 1. [55]
1.06% from surface coal mining, 1.42% from transportation,
SO2 to air 0.0067 kg/kWh and 97.52% from electricity generation. Normal, SD = 1. [55]
Petro Diesel
Energy use 51.0996 MJ/kg 47.75% domestic crude production, 45% foreign crude
production, 0.27% domestic crude transport, 1.09% foreign
crude transport, 5.38% crude refining, 0.51% diesel transport.
Normal, SD = 1. [57]
Water use 0.0004 m3/kg Normal, SD = 1. [57]
CO2 to air 10.0476 kg/kg 86.54% tailpipe fossil (combustion), 13.46% production.
Normal, SD = 1. [57]
SO2 to air 0.0027 kg/kg Normal, SD = 1. [57]
H2, from fossil fuels
(RER), as liquid H2
Energy use
Water use 33.6420 MJ/kg As oil, crude, in ground. Normal, SD = 1. [58]
0.000723 m3 water estimate + 0.0732 m3 cooling water,
0.0739 m3/kg unspecified origin. Normal, SD = 1. [58]
CO2 to air 1.2328 kg/kg 1.23 kg fossil CO2 – 0.00278 kg biogenic CO2. Normal, SD = 1.
[58]
SO42- to water 0.0005 kg/kg Normal, SD = 1. [58]
Isopropanol
Energy use 3.2000 MJ/kg Propylene oxide production via direct hydration demands a split
of 50% natural gas, 38% electricity, and 12% steam. All energy
used for heat of steam was obtained from natural gas.
Lognormal, SD = 1.88. [59]
Water use 0.0271 m3/kg 0.0032 m3 water estimate as 10x stoichiometric amount + 0.024
m3 cooling water. Lognormal, SD = 1.88. [48]
CO2 to air 0.0990 kg/kg From wastewater treatment. Lognormal, SD = 1.88. [60]
SO42- to water 0.1740 kg/kg Calculated from mass balance. Lognormal, SD = 2.11. [60]
Carbon black, at plant
(RER)
Energy use 9.8820 MJ/kg As heat from industrial furnace, >100 kW. Lognormal, SD =
1.24. [61]
Water use 0.0000 m3/kg Uniform = 0.00 (minimum), 0.01 (maximum). [60]
CO2 to air 1.9700 kg/kg Calculated from mass balance. Lognormal, SD = 1.38. [60]
Cryolite (2 AlF3 • 3NaF)*
Energy use 3.2000 MJ/kg 2 MJ/kg natural gas and 1.2 MJ/kg electricity. Normal, SD = 1.
[62]
Water use 0.0300 m3/kg 24 kg/kg + 6 kg/kg cooling water. Normal, SD = 1. [62]
Uniform = 0.00 (minimum), 0.01 (maximum). [62]
CO2 to air 0.0000 kg/kg
Steel
Energy use 2.1826 MJ/kg 18.66% coke-making, 9.59% sintering, 67% blast furnace,
4.75% LD steel converter. Cradle-to-casting plate gate (steel
plates). Normal, SD = 1. [63]
Normal, SD = 1. [63]
Normal, SD = 1. [63]
Water use 0.0003 m3/kg
CO2 to air 1.1586 kg/kg
548 * Due to missing production data for Ni-Mo/Al2O3 catalyst, the impacts of cryolite production was adapted as a rough estimate on
549 energy use, water use, and CO2 emissions according to the overall reaction: 12 HF + Al2O3 • 3H2O + 6 NaOH 2 AlF3 • 3 NaF + 12
550 H2O. To produce 1 kg cryolite, these reactants are needed stoichiometrically: 28.580 mol HF, 2.382 mol Al2O3 • 3H2O, and 14.290
551 mol NaOH at 100% assumed yield. Gendorf (2000) [62].

22
552 Table 4. Global LCI for HRJ and CTH detailing energy/materials inputs/outputs to and from the
553 technosphere/environment based on Table S71. Energy and heat are in MJ/day and materials and chemicals
554 are in tons/day.
HRJ
Energy, chemicals (input from technosphere) Products, co-products (output to technosphere)
Electricity 683,059.20 Naphtha 267.80
Heat 512,599.92 Jet fuel 478.37
Diesel 16.53 Propane gas 4.12
WCO 1,000.00 C1-C3 gas 97.64
Ni-Mo/Al2O3 3.26 C4-C8 51.81
H2 (gross) 83.44 C9-C15 1.19
H2 (net, minus recycle)* 35.20 C1-C3 liquid (LPG) 7.52
Steel 0.0017 C16-C18 (diesel) 2.64
Emissions (output to environment) Liquid CO2 1.52
Air Heat (cooling) 1,207,360.70
CO2 150.71
CO 11.42
H2 25.75
Water
Pure H2O 18.69
Sour water (H2O+CO2) 37.77
CTH
Energy, chemicals (input from technosphere) Products, co-products (output to technosphere)
Electricity 158,696.70 Naphtha 232.86
Heat 485,100.00 Jet fuel 402.88
Diesel 14.44 Diesel 122.13
WCO 1,000.00 Coke 29.41
Isopropanol 341.80 C1-C3 gas 161.85
Activated carbon 0.0913 Octane 5.83
Emissions (output to environment) Dodecane 0.11
Air C1-C3 liquid (LPG) 4.40
CO2 210.79 Heat (cooling) 493,849.44
CO 127.90
H2 6.85
Water
Pure H2O 16.26
Sour water (H2O+CO2) 72.06
555 *Net H2 requirement is 35.20 tons/day, of which 22.93 tons/day is consumed in the hydrodeoxygenation reactor, 11.34 ton/day
556 in the hydrocracking unit, and 0.93 ton/day are lost in the light gas stream separated by the low-pressure flash.
557
558
559 3. Results and Discussion

560 3.1. Process Performance

561 The results of process simulation for HRJ and CTH systems are summarized in Table 5. Aspen simulation

562 resulted in an overall HRJ hydrogen consumption of 35.2 ton/day, of which 22.93 ton/day were consumed in

563 the HY reactor (at hydrogen conversion of 88.2%), 11.34 ton/day in the hydrocracking unit, and 0.93 ton/day

564 lost in the light gas stream separated by the low-pressure flash. Hydrogen was assumed to be completely

565 recycled without any losses, although in practice these may vary between 15-25% for PSA unit. Hence, the

566 amount of hydrogen consumed would be slightly larger than indicated. The overall hydrogen-to-WCO ratio

23
567 of 3.5% w/w is similar to the ones calculated by Chu et al. (2.6-3% w/w), Han et al. (2-3% w/w), and

568 Pearlson et al. (4% w/w) [30,33,45]. Nevertheless, Chu et al. did not consider the amount of hydrogen used

569 for hydrocracking reactions assuming it as a factor of the degree of cracking instead. Pearlson et al. proposed

570 that 4% w/w overall process ratio is appropriate for maximum jet fuel production based on literature data.

571 The consumption of isopropanol in CTH is equal to its supply (341.8 ton/day), since it was completely

572 decomposed inside the reactor and hence, non-recyclable. CTH’s isopropanol use is equal to 45.5 ton/day H2

573 consumption equivalent. Normalized on per ton biojet fuel produced, HRJ consumes 1.4 ton H2 while CTH

574 utilizes 2.4 ton H2. On a per ton WCO consumed, HRJ uses 0.035 tons H2 while CTH utilizes 0.045 ton H2.

575
576 Table 5. Process performances of HRJ and CTH systems. All specific energy consumption of utilities is
577 scaled to the mass of WCO feed.
Performance parameter HRJ CTH Unit
Waste cooking oil basis rate 1,000 1,000 ton/day
H2/isopropanol consumption 35.2 341.8 ton/day
H2/isopropanol consumption 17,600 5,678.2 kmol/day
H2 consumption equivalent 17,600 22,748.8 kmol/day
H2 consumption equivalent 35.2 45.5 ton/day
H2 consumption/ton WCO 0.035 0.045 ton
H2 consumption/ton biojet 1.4 2.4 ton
Liquid fuels products
Total liquid1 770.7 / 40,207.2 763.2 / 40,139.4 ton/day; L/hr
Naphtha2 8,041.5 12,041.8 L/hr
3
Kerosene 32,165.8 24,083.6 L/hr
Diesel 0 4,013.94 L/hr
Energy consumption
Heating energy 0.243 0.298 kWh/kg
Cooling energy -0.517 -0.365 kWh/kg
Electric energy 0.092 7.4·10-5 kWh/kg
Maximum pressure 9.2 0.2 MPa
Maximum temperature 400 380 °C
1
578 Total liquid yield: HRJ = 77 w/w%, CTH = 76 w/w%.
2
579 Naphtha yield: HRJ = 20 v/v%, CTH = 30 v/v%.
3
580 Kerosene yield: HRJ = 80 v/v%, CTH = 60 v/v%.
581

582 The yield of liquid fuel products obtained from both systems were similar, comparable to previous

583 studies [10,30,33,45] (77 wt% for HRJ and 76 wt% for CTH). The liquid mixture obtained from the

584 conventional HRJ system were characterized by a larger amount of kerosene products, roughly 16% higher

585 than that obtained from CTH. However, CTH produced more naphtha and diesel than HRJ. Both systems

586 produced gaseous by-products. Specifically, about 10% of the WCO feed was converted into light gas

587 (mostly propane) in HRJ, comparable to what is reported in other studies, ranging between 7% and 10%

24
588 [30,33,45]. This light gas had low LHV because it contained large amount of CO and CO2. The CTH system

589 produced around 162 ton/day of fuel gas with LHV of 21.2 MJ/kg. The thermal and electricity comparison

590 between the two systems reflected their relative process conditions: temperature and pressure. The thermal

591 energy (heat) consumptions were comparable between the two systems (513 GJ/day for HRJ, 485 GJ/day for

592 CTH), due to the optimized heat integration network and the similar operating temperatures. CTH’s cooling

593 energy requirement was 59% lower than HRJ’s due to the latter’s combined exothermic effect via flash

594 separation and high-pressure flash. In terms of direct electricity, HRJ consumed more than four times

595 electricity than CTH because of its high operating pressure and the handling of gaseous hydrogen via multi-

596 stage compressors. CTH utilized low electricity because it operates at relatively low pressure and pumps

597 were employed in place of compressors to transport liquid streams. In general, CTH consumed far less heat

598 and electricity than HRJ because the number of pieces of equipment necessary in an HRJ reaction was much

599 larger (two reaction systems: HY and HH), whereas only one reactor is used in CTH. Further, CTH does not

600 require large recycle of gaseous streams, which is beneficial in reactor sizing and operation.

601 3.2. Capital Expenditure

602 The results of CAPEX and FCI for both HRJ and CTH systems are summarized in Table 6 and Table 7,

603 respectively. A summary of direct manufacturing costs for both systems are reported in Table 8. Moreover,

604 the results of COM for HRJ and CTH systems are shown in Table 9 and Table 10, respectively using direct,

605 fixed, and general costs of manufacturing data (Table S58 and Table S59).

25
606 Table 6. Total bare module cost in HRJ and CTH jet fuel production processes. For HRJ the cost at
607 atmospheric pressure (base) is reported in addition to the final one.
Equipment Type HRJ CTH
Reaction System Cbm,base Cbm Reaction System Cbm
1
Hydrodeoxygenation $ 1,650,989 $ 40,814,774 Hydrodeoxygenation $ 132,396
Hydrocracking2 $ 2,972,309 $ 76,618,490 2nd hydrodeox. regen. $ 132,396
Steam methane reform. $ 3,441,561 $ 16,313,000
3
Heating $ 1,450,520 $ 1,675,585 Heating3 $ 3,010,654
Compression Cbm,base Cbm Pumps Cbm
Compression 1 $ 448,539 $ 1,704,448 Feed pumping WCO $ 12,046
4
Compression 2 $ 1,234,069 $ 4,689,461 Feed pumping ISOP $ 11,158
Feed pumping WCO $ 73,966 $ 174,098
Separation Cbm,base Cbm Separation Cbm
By-P HP separation $ 74,603 $ 663,506 LP separation $ 74,603
Product HP separation $ 53,829 $ 581,562
Product LP separation $ 63,636 $ 63,636
PSA-hydrogen recov. $ 889,876 $ 2,936,590
Distillation Cbm,base Cbm Distillation Cbm
5 5
Fractionation 1 $ 487,216 $ 487,216 Fractionation 1 $ 387,179
5
Fractionation 2 $ 491,028
Heat Exchangers Cbm,base Cbm Heat Exchangers Cbm
Heat exchange 1 $ 104,310 $ 123,010 Heat exchange 1 $ 469,833
Heat exchange 2 $ 157,024 $ 185,174
Heat exchange 3 $ 387,390 $ 456,232
Heat exchange 4 $ 159,540 $ 187,891
Cooling 16 $ 139,372 $ 164,358 Cooling6 $ 151,479
6
Colling 2 $ 144,417 $ 169,854
3
Heating $ 1,450,520 $ 1,672,574 Heating3 $ 2,446,909
Total (2017) $ 15,383,686 $ 149,681,462 Total (2017) $ 7,319,681
1 2 3 4 5 6
608 4x5 bed, 5x5 bed, Fired heater, Three stages Trays + columns Cooling water
609
610
611 Table 7. Fixed capital investment in HRJ and CTH jet fuel production processes.
Fixed Capital Investment HRJ CTH
ISBL + Engineering $ 149,681,462 $ 7,319,681
OSBL (50% total Cbm base) $ 7,691,843 $ 3,659,841
Contingency (15% total Cbm) $ 26,942,663 $ 1,097,952
Total $ 184,315,968 $ 12,077,474
612
613

26
614 Table 8. Summary of direct manufacturing costs.
Unit Value Reference
Raw Material
Waste cooking oil $ ton-1 150.00
-1
Hydrogen $ kg 1.60 [45,64]
Isopropanol $ kg-1 1.30 [65]
Utility
Electric power $ kWh-1 0.087 [46]
Natural gas (LHV=38.42MJ m-3) $ m-3 0.130 [46]
-1
Cooling tower water (30°C) $ GJ 0.354 [38]
Refrigerated water (5°C) $ GJ-1 4.430 [38]
-1
Medium pressure steam (10 bar, 184°C) $ GJ 6.870 [38]
-1
High pressure steam (41 bar, 254°C) $ GJ 9.830 [38]
Hot oil $ GJ-1 f(kW) [44]
Operators
Salary $ year-2 60,000.00 [66]
615
616
617 Table 9. Total cost of manufacturing in HRJ jet fuel production process.
Raw Material (RM) Value Unit $/year
WCO 1,000 ton/day 50,000,000
Hydrogen 35.2 ton/day 18,773,333
Utilities (UT) Value Unit $/year
Compressor 1 1,196 kW 836,586
Compressor 2 (make-up) 3,191 kW 2,230,902
Feed WCO pump 125 kW 87,540
Cooling water 1 6,197 kW 63,175
Cooling water 2 5,736 kW 58,477
3
Fired heater (LHV = 38.42 MJ/m ) 704 kW 85,807
Condenser 2,042 kW 20,815
Reboiler 4,189 kW 828,901
Waste Treatment (WT) - - -
DMC + FMC + GE (Table S58) 29,251,124
Total COM (without depreciation) 102,236,661
618
619
620

27
621 Table 10. Total cost of manufacturing in CTH jet fuel production process.
Raw Material (RM) Value Unit $/year
WCO 1,000 ton/day 50,000,000
Isopropanol 341.8 ton/day 148,113,333
Utilities (UT) Value Unit $/year
Hot oil CTH 9,508 kW 2,138,041
Feed WCO pump 2.1 kW 1,458
Feed ISOP pump 1 kW 696
Cooling water 1 3,153 kW 32,140
Fired heater (LHV = 38.42 MJ/m3) 5,614 kW 683,904
Reboiler 1 5,034 kW 1,425,077
Reboiler 2 1,747 kW 867,511
Condenser 1 1,058 kW 10,788
Condenser 2 1,505 kW 15,348
Waste Treatment (WT) - - -
DMC + FMC + GE (Table S59) 15,147,753
Total COM (without depreciation) 218,436,050
622
623
624 In both CTH and HRJ systems, CAPEX was controlled by their main reaction sub-systems: 91% in HRJ

625 and 45% in CTH, calculated in reference to the bare module cost assuming a factor of 4.74 to fluid process

626 as single Lang factor, Cbm [22]. It is worth noting that the great difference in the capital costs of reactors

627 between HRJ and CTH in Table 6 is due to a few factors. First factor is reactor volume. HDO and HCC

628 reactors in HRJ operate with a large excess of gaseous H2, resulting in a much larger (about 10 times) reactor

629 volume compared to CTH. Second factor is catalyst price. While conventional hydrotreating processes

630 employ metal-based catalysts (Ni-Mo), CTH uses a sensibly cheaper catalyst that is made of activated

631 carbon. Third factor is operating conditions. HRJ operates at a pressure of 92 and 90 bars for HDO and HCC,

632 respectively, compared to 2 bars for CTH. High pressures require high quality and costly stainless steel as

633 construction material, which increases the Cbm of HDO and HCC by about 25 times relative to the base

634 value. The heating component in CTH’s main reactor costs over 3 M$, about 2% the cost of the heating

635 component in HRJ’s main reactor. CTH required pumping, as the solvent was isopropanol while HRJ

636 required compression, as the source was gaseous hydrogen. Gas compression is prohibitively expensive, in

637 the order of M$. Both systems needed infrastructure for separation, although CTH demanded a single low-

638 pressure separator while HRJ required multiple high- and low-pressure separators including a PSA hydrogen

639 recovery unit.

28
640 The breakdown of CAPEX for HRJ and CTH systems are shown in Figure 4 (top panel), respectively.

641 The total cost of CTH plant (7.3 M$) was significantly lower (around 98%) than the total cost of HRJ plant

642 ($149.7 M$). This was mainly due to the high operating pressure in HRJ, which increased the bare module

643 cost by about 10 times relative to atmospheric conditions. Together, the cost of HY and HH reactors in HRJ

644 represented more than three-fourths of the 2017 total CAPEX or 64% of FCI. This large contribution was

645 attributed to the slow reactions that unavoidably resulted in large reactor volumes and critical operating

646 conditions.

647 The SMR plant for hydrogen production substantially contributed to the overall capital investment, about

648 11%. Fractionation in both HRJ and CTH systems operated atmospherically. HRJ’s single fractionation

649 column had a negligible contribution to total CAPEX (<1%). But unlike HRJ, the CTH system consisted of

650 two fractionation columns in series, thereby doubling the cost of distillation. The contribution of these two

651 columns to the total CAPEX was 12%. The CTH main reactor was comparably priced relative to other units,

652 thanks to the low cost of the catalyst employed. In terms of heat exchange, CTH has one integrated heat

653 exchanger, one cooling unit, and one fired-heating element as opposed to HRJ’s four integrated heat

654 exchangers, two cooling units, and one fired-heating element.

655 The FCI of the two systems, accounting for the offsite and contingency contributions, are 184.3 M$ and

656 12.1 M$ for HRJ and CTH system, respectively (Table 7). The direct manufacturing costs in HRJ and CTH

657 as summarized in Table 8 demonstrated that on a per kg basis, isopropanol and hydrogen’s unit costs were

658 relatively similar. However, when mass flow rates and hours of operation are factored in, the total annual

659 cost of isopropanol ballooned to 148.11 M$, in comparison to hydrogen’s 18.77 M$, regardless of

660 hydrogen’s compression requirement. In fact, isopropanol’s cost represented 67.8% of CTH’s total COM

661 (without depreciation, Table 10). Table 9, on the other hand, shows that the driver in HRJ’s total COM is

662 WCO, not hydrogen. Hydrogen compression was another significant driver, in conjunction with the cost

663 associated with heating the single reboiler in the HRJ system.

664 3.3. Operations Expenditure

665 The breakdown of total OPEX (raw materials, utilities) for the two systems are shown in Figure 4 (bottom

666 panel). It is worth noting that OPEX for HRJ amounted to 73 M$/year, while OPEX for CTH amounted to

667 203 M$/year. This huge difference was determined by the cost of isopropanol (148 M$/year), around 68% of

29
668 CTH’s total operating cost (COM without depreciation). In general, the feedstock supply represented the

669 major input to both systems. The overall COM, accounting for operating labor, fixed cost of manufacturing,

670 and general expenses, amounted to 102 M$/year and 218 M$/year for HRJ and CTH system, respectively.

671 In terms of revenue, the two systems were very similar, i.e. 150.8 M$/year for HRJ and 153.9 M$/year

672 for CTH (Table 11). By comparing the revenues with the total COM, CTH generates a negative gross profit,

673 with most of the income offset by the cost of supplying isopropanol.

674
675 Figure 4. Capital expenses of HRJ (above left) and CTH (above right) systems. Operating expenses for raw
676 materials and utilities of HRJ (below left) and CTH (B) systems.
677
678
679 Table 11. Product incomes and gross profits in HRJ and CTH jet fuel production processes.
HRJ CTH
Fuel Product Produced Unit Income ($/yr) Produced Unit Income ($/yr)
3 3
Fuel gas 6,969 Nm /hr 9,433,018 12,978 Nm /hr 13,497,224
Naphtha 8,041 L/hr 27,405,246 12,042 L/hr 41,038,506
Kerosene 32,166 L/hr 113,969,798 24,084 L/hr 85,333,119
Diesel - L/hr - 4,014 L/hr 13,997,406
Total revenue (R) 150,808,063 153,866,254

30
 Cost of manufacturing (COMd) 102,236,661 218,436,050
Gross profit (GP) 48,571,402 -64,569,796
680
681 3.4. Break-even Analysis

682 Break-even analysis was performed to ascertain the profitability of CTH and HRJ systems using total

683 CAPEX (CTH=$7.3M, HRJ=$149.7M), annual total COM (without depreciation) (CTH=$218.4M,

684 HRJ=$102.2M), revenues from liquid fuel (CTH=$140.4M, HRJ=$141.4M), revenues from gaseous fuel

685 (CTH=$13.5M, HRJ=$9.4M), and depreciation (use MACRS for 7 years). With these inputs, the net present

686 value (NPV) for both processes under 10-year debt financing term was evaluated (CTH=-$216.6M,

687 HRJ=$6.4M). To ascertain how NPV is sensitive to cost inputs (i.e. feedstock and products prices), these

688 were varied by ±10% with respect to the base case values. The overall results are reported in Table 12, while

689 Figure 5 shows the impact of each single factor on the NPV of HRJ and CTH processes, respectively. From

690 the latter, it is shown how jet fuel price plays an important role in both processes, while the cost of

691 isopropanol is the most impacting factor in CTH.

692 Table 12. Sensitivity of NPV on feedstock and products prices.

-10% Base case +10% NPV CTH NPV HRJ
WCO ($/t) 135 150.00 165
H2 ($/kg) 1.44 1.60 1.76
Isopropanol ($/kg) 1.17 1.300 1.43 Mina = -335 M$
Mina = -73 M$
Maxb = -100 M$
Naphtha ($/L) 0.383 0.426 0.469 Maxb = 21.5 M$
Base case = -216.6
Jet fuel ($/L) 0.399 0.443 0.487 Base case = 6.4 M$
M$
Diesel ($/L) 0.392 0.436 0.479
Fuel Gas ($/Nm3) 0.117 0.13 0.143
a
693 NPV value corresponding to a 10% increase in feedstock price and 10% decrease in products sale price

694

31
695 Figure 5. Sensitivity of process NPV to feedstock and products prices for HRJ (left) and CTH (right). Dotted
696 yellow bars indicate +10% change in assumption from base case and full red bars -10% change in
697 assumption from base case
698

699 Thereafter, the break-even point (i.e. time in years at NPV=0) for the 10-year term was calculated for both

700 systems by setting the NPV=0 and generating a new annual cumulative NPV profile (0-25 years) with year

701 10 as the break-even point. This method was repeated for other break-even points: 5, 15, 20, and 25 years for

702 both CTH and HRJ systems. The results are plotted in Figure 6.

703
704 Figure 6. Cumulative NPV profiles for CTH (left) and HRJ (right) systems at various debt financing terms
705 showing corresponding break-even points (5, 10, 15, 20, and 25 years).
706
707
708 The 25-year NPV profiles of CTH (left) and HRJ (right) systems reveal the general trend in cumulative

709 NPV across all financing terms. The initial dip in NPV from 0-2 years indicates the M$ initial investment.

710 Seventy percent of the initial investment was spent on the first year and the remaining 30% on the second

711 year. At year 2, the profile vertically dips even further with magnitude corresponding to the working capital.

712 From 2 years onward, the profile increases until it hits the break-even point. Thus, DPBP coincides from

713 year 2 to break-even point. This discounted cash flow analysis revealed that the cumulative cash flow

714 became positive passed the break-even point. Depreciation was factored in from year 4 to year 11, and the

715 tail end of the profile shows the NPV of each system at the end of its life. For example, CTH’s end-of-life

716 NPVs were as follows: 35.29 M$, 8.01 M$, 3.05 M$, 1.04 M$, and 0 M$ for financing term of 5, 10, 15, 20,

717 25 years, respectively. In both CTH and HRJ systems, the salvage value of the project decreases as the

718 financing term increases.

32
719 Negative cumulative NPV indicates that any system is deemed unprofitable while positive cumulative

720 NPV indicates otherwise. For a system to be viable using the break-even analysis, one must understand that

721 the current cumulative NPV at year n is a function of revenue, total COM, and depreciation at that year,

722 including the initial investment. The break-even point is the time when annual revenue equals total business

723 costs (total COM, depreciation) progressively achieved by paying a fraction of the initial investment year

724 after year. Typically, the break-even point in most economic analyses determines the amount or volume of

725 product sales in a year that equals total dollar sales to total dollar COM (i.e. gross profit=$0). In this analysis,

726 however, the break-even point was calculated as the year beyond which total dollar sales exceeds total dollar

727 COM, coinciding with cumulative NPV>$0. There was no set break-even point for which a business model

728 ensures profitability; however, if total COM is higher than expected total sales, the business plan can be re-

729 examined, and alternative strategies must be implemented. This study determined that selling the biojet fuel

730 at $2.99/gal ($0.79/L) and $1.32/gal ($0.35/L) for CTH and HRJ, respectively, would take 25 years to break

731 even. Since most corporations prefer break-even points to be less than 10 years, it would, therefore, be

732 profitable to sell biojet fuel at $3.00/gal ($0.79/L) and $1.67/gal ($0.44/L) for CTH and HRJ, respectively.

733 At these rates and for a project life of 25 years, the NPV for CTH and HRJ were estimated to be 8 M$ and

734 121 M$, respectively. It was observed that the biojet fuel selling price for the CTH system is higher than that

735 of HRJ due the higher operation cost associated with CTH. However, as the break-even point approaches 5

736 years, the biojet fuel selling price for HRJ ($2.70/gal) and that for CTH ($3.10/gal) converges (12.94%

737 difference) such that in the long run, CTH fuel would be affordable as a result of its low total CAPEX. It

738 appears that CTH would recover its initial investment faster than HRJ through faster payment of its CAPEX.

739 As capital recovery within 5 years appears ambitious, maintaining DPBP (8 years after the construction) and

740 an interest rate of 8% for both processes would be ideal.

741 An IRR analysis was performed for both CTH and HRJ systems covering interest rates from 1-50% for

742 the entire 25-year life. The inputs to this analysis were base case inputs, that is, CTH-NPV was negative and

743 HRJ-NPV was positive at the end of 25 years under a 10-year financing term. The results are presented in

744 Figure 7. The IRR is an index of the investment risk, a method used in capital budgeting to estimate the

745 profitability of investments. It refers to the discount rate that makes NPV of all cash flows equal to 0. IRR

746 calculations followed the same method as NPV. As the interest rate increased, HRJ-NPV decreased while

33
747 CTH-NPV increased. Since HRJ-based biojet fuel production is a mature technology with low risk, its IRR

748 of 15.91% represents an attractive value for the investment. As of current market without any subsidies,

749 incentives, or any of the investigated case scenarios from Table S68, CTH’s IRR profile seem to suggest that

750 the system is untenable.

751

752
753 Figure 7. IRR profiles for CTH and HRJ systems under 10-year debt financing term.
754
755
756 At the present state, the CTH process based on experimental conditions investigated has potential to be

757 economically profitable despite large input cost associated with isopropanol. Given that isopropanol cost

758 represents 68% of CTH’s total COM, total COM must be reduced by around 30% to even start realizing net

759 profit. This can be done using a combination of two strategies: (1) reduce both isopropanol cost and total

760 direct/fixed/general cost by 60%, (2) reduce isopropanol cost by 55%. It is highly unlikely that a 60%

761 reduction in staff salaries, plant overhead, and distribution/selling costs will be achieved because these

762 variables are set by prevailing market conditions and for an emerging technology such as CTH-based biojet

763 fuel, these costs will be slightly elevated, leaving the latter option to be realistically feasible. It was

764 calculated that to obtain the same Present Value Ratio of HRJ (a more meaningful index than the NPV when

765 two processes with very different capital investments are compared), the price of the isopropanol should not

766 exceed 0.735 $ kg-1 (i.e. a 43.5% reduction from current cost) (Table S68). Alternatively, at the current

767 market price, the maximum isopropanol inlet was calculated to be 196.54 ton/day (isopropanol/WCO feed

768 ratio of 0.204 by weight) to be profitable. However, performances and product yields should be verified in

769 this scenario.

34
770 3.5. Environmental Metrics

771 The results of the inventory analysis described in Section 2.5.5 was used in life cycle impacts assessment, a

772 step in LCA that quantifies environmental impacts, selects priorities from these environmental impacts, and

773 transforms these priorities into meaningful environmental endpoint (damage category or forecast). In his

774 study, two major environmental endpoints were identified: CED and life cycle greenhouse gas GHG

775 emissions using the Eco-Indicator 99 method [67] and the equivalency factors (damage factors) detailed in

776 Table 3. CED represents the cumulative non-renewable fossil energy requirement of the entire process in MJ

777 per 1,000 tons WCO. Life cycle GHG emissions is reported as 100-year global warming potential (GWP) in

778 kg CO2 equivalents per 1,000 tons WCO.

779 Figure 8 and Table 13 depict the characterization results in the production of biojet fuel from WCO via

780 CTH and HRJ processes. The left pane demonstrates environmental endpoints (CED and GWP, positive bar)

781 and environmental credits (negative bars). The right pane shows the energy distribution of fuel products

782 produced from CTH and HRJ processes including the energy-return-on-investment (EROI), defined as the

783 ratio of the total energy output from the system to the total energy input into the system.

784

785

786
787 Figure 8. Results of characterization in the production of CTH- and HRJ-based biojet fuel from 1,000 tons
788 WCO. Left pane corresponds to environmental endpoints (CED and GWP) and right pane corresponds to
789 distribution of output energy. Energy-return-on-investment (EROI) is included in the right pane.
790
791

35
792 Table 13. Characterization results of Figure 7. Cumulative fossil energy demand (CED) and output fuel
793 energy (OUT) are expressed as × 106 MJ/1,000 tons WCO and global warming potential (GWP) is expressed
794 as 106 kg CO2-equivalents/1,000 tons WCO.
Input (CED/GWP) Operations Direct Heat Upstream
CED-CTH 1,625,836 -447,771 583,780 1,489,773
CED-HRJ 2,539,483 -804,010 793,395 2,550,098
GWP-CTH 327,414 192,557 0 134,240
GWP-HRJ 356,414 137,451 0 218,686
Output (Fuel Energy) Naphtha Diesel Jet Fuel Total
OUT-CTH 9,485,084 4,720,114 15,716,052 29,921,250
OUT-HRJ 10,908,355 0 18,660,657 29,569,011
795
796 To facilitate analysis, it is important to note that CTH operation includes pressurization energy required

797 to pump WCO and isopropanol to the CTH reactor at 2-bar pressure, cooling energy (negative energy)

798 needed to cool reaction products in low-pressure flash unit, and distillation units (by heat integration), and

799 heat input. On the other hand, HRJ operations involves hydrogen pressurization energy in HY and HH units

800 assuming isentropic compression (85% efficient) and WCO pumping (95% efficient). It also includes

801 cooling energy in the flash separator, high-pressure flash unit, distillation column (exothermic conditions),

802 and heat input. From a gross operations perspective, HRJ exhibits 1.6 times the CED of CTH. This is not

803 surprising given that it has greater upstream fossil energy brought about by the production, pressurization,

804 and delivery of hydrogen. Gaseous hydrogen is primarily produced from stream reforming of natural gas.

805 Isopropanol is the main upstream driver in CED of CTH; albeit insufficient amount to cause higher CED of

806 HRJ.

807 Direct CED of CTH and HRJ systems are negative, that is, surplus electricity is generated by the system.

808 The energy generated from cooling the reactor is re-captured and utilized as in-process energy. HRJ operates

809 at higher temperature than CTH, enabling larger energy release. It also has more reactor units. Consequently,

810 HRJ’s net heat consumption is 26% higher than CTH’s. HRJ’s direct heat use comes from heating the oil

811 from 380°C to 400°C plus the heat duty of the distillation column. Although CTH consists of two distillation

812 units in series needing heating of WCO from 300°C to 380°C, the combined heat demand is still lower than

813 HRJ’s.

814 Total CED is equal to operations CED plus infrastructure CED. The results suggest that total CED for

815 both systems are equal to their operations CED due to the miniscule contributions of steel. Reactor steel

816 demand for both systems are between 33-177 kg/yr, annualized over a 30-year period.

36
817 The results of operations GWP is consistent with the results of operations CED in both systems, i.e.

818 operations GWP-CTH < operations GWP-HRJ; however, the difference is low at around 8%. This result

819 considers that the net GHG emissions has two components – direct GHG emissions and upstream GHG

820 emissions. Both CTH and HRJ systems do not sequester CO2 through co-product offsets because there is no

821 biogenic carbon embodied in the final fuel unlike jet fuel from algae or switchgrass. While it is true that

822 some WCO are plant-based (olive oil, palm oil, soybean oil, or canola oil), others are animal-derived (butter,

823 lard) or synthetic. In fact, even if this study assumes that all WCO are plant-derived, it is still outside the

824 scope to identify which location produced waste olive oil or canola oil. Thus, this analysis does not consider

825 carbon sequestration. Direct GHG emissions from CTH is higher than direct GHG emissions from HRJ,

826 mainly due to CO2 emissions from low-pressure flash. However, upstream GHG emissions from CTH is

827 26% lower than HRJ’s because of the considerable emissions associated with hydrogen production. Total

828 GWP is equal to operations GWP due to the insignificant GHG emissions associated with steel in the

829 reactors. Further, both CTH and HRJ systems have no heat offset, i.e. total heat use equals net heat use. The

830 hydrodeoxygenation reaction in HRJ consumes more heat than the CTH reaction.

831 The conversion of 1,000 tons/day WCO generates approximately 60 x 106 MJ usable fuel energy in the

832 form of jet fuel, naphtha, and diesel from both CTH and HRJ systems. As evident in the right pane of Figure

833 8, relative magnitudes of these energy are similar, except that HRJ reaction does not produce diesel product.

834 HRJ has a slightly lower total energy than CTH despite its higher jet fuel and naphtha production because

835 HRJ does not produce diesel. The EROI of both systems were calculated using total energy output and total

836 energy input. EROI is the system’s energy efficiency measured in terms of energy productivity normalized

837 by energy expenditure. It is used as a decision tool in adapting any process from an energetic standpoint with

838 EROI>1 to be desirable (i.e. energy break-even=1, total energy output=total energy input). However,

839 different EROIs can be calculated from the same dataset as a result of dissimilar energy accounting methods

840 [68]. In this study, EROI is evaluated as the ratio of the summation of all energy embodied in the main

841 product and co-products plus surplus electricity (if any), and, the summation of net energy (if any), net heat,

842 upstream energy, and infrastructure energy consumption. Although both energy platforms are energetically

843 favorable, CTH’s EROI is 1.8 times higher than HRJ’s, attributable to its lower heat and upstream impacts.

37
844 Figure 9 shows that CTH has lower electricity and heat consumption than HRJ; in fact, CTH’s negligible

845 electricity use is credited to pressurization, while the rest of the sub-processes yield in-process heat. Both

846 systems have comparable direct heat use. Because CO2 separation, high-pressure separation, and

847 fractionation units in HJR operate at high temperatures, large amount of heat is expended to lower product

848 temperature. HRJ’s single fractionation unit consumes more than twice the electricity of CTH’s.

849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869 Figure 9. Electricity and heat use breakdown among sub-processes in the production of CTH- and HRJ-based
870 biojet fuel from 1,000 tons WCO.

871
872 Figure 10. GWP breakdown among sub-processes in the production of CTH- and HRJ-based biojet fuel from
873 1,000 tons WCO.
874
875
876 The relative global warming potential (GWP) contributions reported as kg CO2 equivalents/1,000 tons

877 WCO of the different sub-processes for CTH and HRJ systems are presented in Figure 10. The results

878 indicate that both systems create similar warming within 100 years (CTH=376,403. HRJ=375,420). CTH’s

38
879 GWP is dominated by emissions from low-pressure flash while HRJ’s is driven by WCO transport

880 emissions. This is surprising given that the CED of HRJ is 1.56 times that of the CED of CTH. A closer look

881 at the GWP breakdown reveals that the main reaction in CTH produces half the GHG emissions of HRJ’s

882 main reaction. Nonetheless, the sum of GHG emissions associated with the sub-processes in CTH, except for

883 WCO/fuel transport and diesel emissions, is higher than HRJ’s by 33%. Beyond CTH’s main reactor, the

884 low-pressure flash generates the most emissions among all sub-processes at 169,390 kg CO2-eq/1,000 tons

885 WCO. Such phenomenon is associated with: (1) the atmospheric 3-phase (VLLE) flash unit following CTH

886 that separates gaseous products from the liquid organic fraction and water, and (2) the large temperature

887 differential in cooling the reaction products. The system has neither provision for recouping this large

888 emission for re-use nor method of sequestration. In contrast, emissions associated with low-pressure flash in

889 HRJ is one-fifth that for CTH’s, as exhibited by the stable reactor temperature of HRJ at 24oC.

890 3.6. Effect of Allocation Method

891 The allocation methods were analyzed in the context of understanding the relative effect on CED and GWP

892 (Figure 11). For CTH, CED for main products via mass- and energy-allocations are in the order of millions

893 while CED for main products via market-allocation is in the order of thousands. Conversely, CED for co-

894 products via mass- and energy-allocations are two orders of magnitude greater than CED for co-products via

895 market-allocation. This is due to mass rates (tons/day) corresponding to energy rates (MJ/day) and the LHVs

896 (MJ/kg) among fuel products and co-products are close to each other (min=42.791, max=47.206). Yet,

897 market-allocation is based on price ($/gal) and density (kg/m3). Hence, the resulting allocation is remarkably

898 different from mass- and energy-allocations. In the case of CTH, 28.20 M$ is attributed to co-products, of

899 which 28.17 M$ is generated by the sales of gaseous products (C1-C3 gas). Consequently, 98.33% was

900 allocated to co-products while only 1.67% was allocated to main products. Similar trend is observed in HRJ

901 main products: 2.08 MMJ/1,000 tons WCO and 2.06 MMJ/1,000 tons WCO for mass- and energy-

902 allocations, respectively, while 0.06 MMJ/1,000 tons WCO for market-allocation. In the case of co-products,

903 mass- and energy-allocations are in the order of hundred thousand while market-allocation is in the order of

904 millions. Similar trend is observed in GWP. From an environmental life cycle perspective, price and density

905 have more pronounced effect on impacts assessment than LHV. Because this analysis pertains to comparing

39
906 an emerging energy platform to a more established conventional processing, reporting CED and GWP

907 impacts based on the three allocation methods is vital.

908

909
910 Figure 11. Analysis of CED and GWP in the production of CTH- and HRJ-based biojet fuel from 1,000 tons
911 WCO using mass rate (MRA), energy rate (ERA), and market value (MVA) allocations.
912
913
914 3.7. Sensitivity

915 The construction of CTH and HRJ models requires the incorporation of materials and process inputs with

916 distributions (“assumptions”) obtained from this study’s experimental results and from various literature

917 sources (Table S70). The models also apply the equivalency factors in Table 3 during life cycle impacts

918 assessment. These factors contain distributions. When analyzing a model, it is useful to know which input

919 variables have the greatest impact on the model’s result (endpoint or forecast). The goal is to determine the

920 driver of a forecast while ignoring inconsequential inputs.

921 The tornado plots presented in Figure 12 is the result of independently testing each input variable one at

922 a time against the target forecast, CED and GWP for CTH. The Monte Carlo simulation of the models

923 utilized assumptions and generated several charts that depicted the relationships between the assumptions

924 and the target. It is evident that CED and GWP of CTH are dependent on transportation – hauling of

925 WCO/fuels and not on process parameters. For example, a 9% increase in truck mileage efficiency from 5.50

926 mi/gal base case to 6.05 mi/gal resulted in a 2.4% decrease in CED from the median value of 0.0835 MJ/MJ

40
927 WCO. The results reveal that as the truck mileage efficiency increases, CED and GWP decrease; whereas, as

928 distance hauled increases, both CED and GWP increase.

929 Of special case is the GHG emissions associated with the electricity equivalency factor (U.S. grid). The

930 base case is 1.0220 kg/kWh and SD = 1. A 125% decrease in electricity equivalency factor resulted in a 68%

931 increase in GWP (0.015 to 0.047 kg CO2 eq/MJ WCO). Similarly, a 51% decrease in electricity equivalency

932 factor caused a negligible increase in GWP (0.015 to 0.016 kg CO2 eq/MJ WCO).

933

934 Figure 12. Tornado plots depicting drivers in CED and GWP of CTH. Dotted yellow bars indicate +5-10%
935 change in assumption from base case and lined red bars indicate -5-10% change in assumption from base
936 case. Bar labels show the test range for each input variable. Tornado method used is percentiles of the
937 variables and test range is between 10% to 90%. Explained variation in forecast is cumulative.
938
939
940 Results of the sensitivity analysis highlighted two key findings: (1) fossil energy consumption and GWP

941 are highly sensitive to externalities associated with transportation of WCO and fuels, and (2) the relative

942 contributions of surface coal mining, transportation, and electricity generation to GWP are insignificant

943 because the combined effects of these processes tend to amplify GHG emissions. The first finding has an

944 advantage and disadvantage. The advantage relates to the minimal influence of CTH’s reaction chemistry on

945 the environment which allows any research and development to further enhance energy and emissions

946 performance. The CTH-based biofuel plant does not require reconfigurations of the main reactor to cut costs,

947 increase profits, and improve sustainability other than what was detailed in this analysis. Because the CED

948 and GWP drivers are transportation-related and not process-related, the disadvantage would be the effect of

949 diesel fuel used in heavy trucks on the circular bioeconomy, to the extent that diesel’s volatile market creates

950 a gap between any gain or loss incurred by the plant. Such disadvantage has a more pronounced impact on

951 the TEA aspect than the environmental outcome.

41
952 Given the knowledge that the environmental impacts of CTH are driven by externalities, it is essential to

953 examine the CTH model with regards to variations truck mileage efficiency and distance hauled. There are

954 two types of variations: uncertainty and variability. Uncertainty is the lack of pertinent information about the

955 driver while variability is the characteristic spread of the driver in a set of observations or population. Truck

956 mileage efficiency has variability [39,40] while distance hauled has uncertainty. The current state of trucking

957 industry in the U.S. has been essentially stable and therefore emphasis must be placed on distance traveled,

958 which is a function of the study’s scope. This analysis determined that the most optimal hauling distance is

959 300 mi/truck-d (6 h per truck per day at 50 miles/hr), obtained from the four-region, nine-division scheme

960 (Section 2.5.1).

961 To capture CTH’s energy outlook as a function of distance hauled and truck mileage efficiency across all

962 regions using process conditions and reaction chemistry described herein, a trend analysis on CTH’s EROI

963 was made on a single CTH-based biojet fuel plant. This plant assumes to transition from HRJ to CTH in one

964 year. The results are presented in Figure 13. HRJ’s median EROI of 10.5 lies within the 25% certainty band

965 (range: 10.4-10.6). In other words, HRJ will have a 25% probability of enhancing EROI to 10.6 from the

966 median value of 10.5. It also has a 90% chance of improving EROI to 11.30 from the median value of 10.5.

967 Clearly, the latitude on improvement (or deterioration) of energy efficiency through optimization in the HRJ

968 system is limited (EROI=10.30-11.30). In contrast, the range by which CTH’s EROI could be improved is

969 quite wide, from 18.6 to a maximum of 25. This finding is significant considering that on a per volume WCO

970 basis, CTH yields statistically equivalent biofuel energy as HRJ, expends half the fossil energy as HRJ, and

971 creates 92% GWP of HRJ’s. Despite isopropanol’s high cost in the TEA (Section 3.3), it only accounts for

972 57% of CTH’s total CED and surprisingly, around 9% of CTH’s total GWP.

973 The contributions between the CTH’s main reaction system versus HRJ’s on EROI projections are

974 evident in their individual CED. While isopropanol’s CED represents 99% of CTH’s main reaction system

975 (activated carbon’s CED is merely 1% of CTH’s main reaction system and is assumed to be regenerated and

976 replaced every 90 days), CTH reaction constitutes half the CED of HY and HH reactions combined. More

977 importantly, isopropanol’s CED in producing CTH biojet fuel is less than hydrogen’s CED in producing HRJ

978 biojet fuel by a factor of 0.5. Clearly, the environmental opportunities in using an alcohol-based solvent

979 remarkably outweighs any cost barrier.

42
980
981 Figure 13. Trend chart showcasing the shift in EROI from HRJ to EROI at median EROI values (HRJ-
982 EROI=10.5, CTH-EROI=18.6.
983
984
985 4. Conclusions

986 This study analyzed the techno-economic and environmental life-cycle implications of biojet fuel production

987 from WCO via the CTH process. Results were compared to the conventional HRJ process. TEA results

988 demonstrate that the total CAPEX of CTH plant (7.3 M$) is significantly lower (~95%) than the total

989 CAPEX of HRJ plant ($149.7 M$). In terms of other economic metrics, CTH outperforms HRJ based on

990 revenues from gaseous fuels (CTH=$13.5M, HRJ=$9.4M) but underperforms based on annual total COM

991 (without depreciation) (CTH=$218.4M, HRJ=$102.2M). The revenues from liquid fuels in both systems are

992 comparable (CTH=$140.4M, HRJ=$141.4M). NPV analysis reveals that to be profitable, and for a break-

993 even point to be less than 10 years, biojet fuel can be sold at $3.00/gal ($0.79/L) and $1.67/gal ($0.44/L) for

994 CTH and HRJ, respectively. At these rates and for a project life of 25 years, the NPV for CTH and HRJ are 8

995 M$ and 121 M$, respectively. IRR for HRJ is 15.91% while CTH’s IRR profile is untenable without

996 subsidies, incentives, or any of the investigated case scenarios. This TEA demonstrates that the CTH process

997 based on experimental conditions has potential to be economically profitable despite large input cost

998 associated with isopropanol. Given that isopropanol cost represents 68% of CTH’s total COM, total COM

999 must be reduced to around 30% to realize net profit. This is done using a combination of two strategies: (1)

1000 reduce both isopropanol cost and total direct/fixed/general cost by 60%, and (2) reduce isopropanol cost by

1001 55%.

43
1002 The LCA results indicate that the environmental opportunities in using an alcohol-based solvent

1003 remarkably outweighs any cost barriers. The latitude on improvement (or deterioration) of energy efficiency

1004 through optimization of CTH’s process parameters is wide. Despite isopropanol’s restrictive cost, CTH

1005 outperforms HRJ, regardless of the allocation method employed. The CED of HRJ is 1.6 times the CED of

1006 CTH. Direct CED of CTH and HRJ systems are negative, meaning surplus electricity is generated. In cooling

1007 the reactors, they also produce heat, which is re-captured and utilized as in-process energy. HRJ operates at

1008 higher temperature than CTH, so HRJ consumes four times more electricity than that of CTH. CTH’s

1009 cooling energy requirement is 59% lower than HRJ’s due to the latter’s combined exothermic effect via flash

1010 separation and high-pressure flash. Although EROIs for both CTH and HRJ are energetically favorable

1011 (EROI > 1), CTH’s EROI is 1.8 times higher than HRJ’s. The total 100-year GWP of CTH is 8% less than

1012 HRJ’s, and both systems do not sequester CO2 through co-product offsets. CTH’s GWP is dominated by

1013 emissions from low-pressure flash while HRJ’s is driven by WCO transport emissions. Environmental

1014 endpoints based on mass- and energy-allocations are relatively similar to each other but remarkably different

1015 from market-value allocation. In both systems, price and density have more pronounced effect than LHV has

1016 on impacts assessment. The fossil energy consumption and GWP are highly sensitive to externalities

1017 associated with transportation of WCO and fuels and the relative contributions of surface coal mining,

1018 transportation, and electricity generation to GWP are insignificant because the combined effects of these

1019 processes tend to amplify GHG emissions.

1020 5. Acknowledgements

1021 This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-

1022 profit sectors.

1023 6. Appendix A – Stream Tables

1024 Table A.1. Material and energy balances of HRJ process (Figure 2)
BLOCK COMPRESSION HDO REACTOR FLASH-1
Inlet Outlet Inlet Outlet Outlet
H2- H2- H2- PROD- GAS-
STREAM H2 WCO WATER HC
MAKEUP REC HDO HDO HP
Temperature (°C) 25 36 155 155 25 155 59 59 59
Pressure (MPa) 0.1 4.0 9.2 9.2 0.1 9.2 9.2 9.2 9.2
Total mass flow rate (t/d) 35.2 24.8 60.0 18.7 1053.5 1072.2 37.8 1002.0 32.4

44
 Components mass fraction
Linoleic acid C18H32O2
Oleic acid C18H34O2
Pentadecanoic acid C15H30O2
Nonadecanoic acid C19H38O2
N-tetradecane C14H30
N-pentadecane C15H32
N-Heptadecane C17H36
N-octadecane C18H38
N-Nonadecane C19H40
C1-C3 gases
Napthta (C4-C8)
Jet fuels (C9-C15)
Diesel (C16-C18)
H2
CO
CO2
H 2O
Energy
Heating (MW)
Cooling (MW)
Power (MW)
1025
BLOCK
STREAM
Temperature (°C)
Pressure (MPa)
Total mass flow rate (t/d)
Components mass fraction
Linoleic acid C18H32O2
Oleic acid C18H34O2
Pentadecanoic acid C15H30O2
Nonadecanoic acid C19H38O2
N-tetradecane C14H30
N-pentadecane C15H32
N-Heptadecane C17H36
N-octadecane C18H38
N-Nonadecane C19H40
C1-C3 gases
Napthta (C4-C8)
Jet fuels (C9-C15)
Diesel (C16-C18)
0 0 0 0 0.496 0 0 0 0
0 0 0 0 0.177 0 0 0 0
0 0 0 0 0.170 0 0 0 0
0 0 0 0 0.106 0 0 0 0
0 0 0 0 0 0.097 0 0.104 0
0 0 0 0 0 0.042 0 0.045 0
0 0 0 0 0 0.402 0 0.430 0
0 0 0 0 0 0.237 0 0.254 0
0 0 0 0 0 0.027 0 0.029 0
0 0 0 0 0.051 0.050 0 0.049 0.127
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0 0
1 1 1 1 0 0.003 0 0.001 0.072
0 0 0 0 0 0.005 0 0.002 0.104
0 0 0 0 0 0.098 0.001 0.082 0.694
0 0 0 0 0 0.039 0.999 0.004 0.003
- - E2+Fired Heater 5.93
Intercooling -2.58 Reactor cooling -5.16 CW1 -6.2
Compression
3.71
(C1+C2)
HCC REACTOR HP-FLASH LP-FLASH
Inlet Outlet Outlet Inlet Outlet
H2-HCC HC-IN PROD-HCC V-HP L-HP L-LP LIGHTGAS LIQFUELS
155 350 196 40 40 24 24 24
9.2 9.0 9.0 9.0 9.0 0.1 0.1 0,1
34.0 1002.0 1035.9 141.01 894.9 894.9 124.2 770.7
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
0 0 0 0 0 0 0 0
0 0.104 0 0 0 0 0 0
0 0,045 0 0 0 0 0 0
0 0.430 0 0 0 0 0 0
0 0.254 0 0 0 0 0 0
0 0.029 0 0 0 0 0 0
0 0.049 0.102 0.385 0.057 0.057 0.355 0.010
0 0 0.334 0.063 0.377 0.377 0.351 0.383
0 0 0.448 0.002 0.518 0.518 0.007 0.602
0 0 0.003 0 0.003 0.003 0 0.003

45
 H2 1 0.001 0.023 0.159 0.001 0.001 0.008 0
CO 0 0.002 0.008 0.052 0.001 0.001 0.006 0
CO2 0 0.082 0.080 0.337 0.039 0.039 0.273 0.002
H 2O 0 0,004 0 0 0 0 0 0
Energy
Heating (MW) - - - -
Cooling (MW) - CW2 -5.74 -
Power (MW) - - - -
1026
BLOCK H2-RECOVERY FRAC
Outlet Outlet
STREAM CO2 FUELGAS H2-REC1 H2-REC2 NAPHTHA JETFUELS
Temperature (°C) 51 37 51 37 65 165
Pressure (MPa) 4.0 4.0 4.0 4.0 0.1 0.1
Total mass flow rate (t/d) 30.1 118.6 2.34 22.5 276.4 494.3
Components mass fraction
Linoleic acid C18H32O2 0 0 0 0 0 0
Oleic acid C18H34O2 0 0 0 0 0 0
Pentadecanoic acid C15H30O2 0 0 0 0 0 0
Nonadecanoic acid C19H38O2 0 0 0 0 0 0
N-tetradecane C14H30 0 0 0 0 0 0
N-pentadecane C15H32 0 0 0 0 0 0
N-Heptadecane C17H36 0 0 0 0 0 0
N-octadecane C18H38 0 0 0 0 0 0
N-Nonadecane C19H40 0 0 0 0 0 0
C1-C3 gases 0.137 0.458 0 0 0 0
Napthta (C4-C8) 0 0.080 0 0 0.988 0.016
Jet fuels (C9-C15) 0 0 0 0 0.012 0.984
Diesel (C16-C18) 0 0 0 0 0 0
H2 0 0 1 1 0 0
CO 0.112 0.062 0 0 0 0
CO2 0.748 0.400 0 0 0 0
H 2O 0.003 0 0 0 0 0
Energy
Heating (MW) - Reboiler 4.19
Cooling (MW) - Condenser -2.04
Power (MW) -
1027

1028

1029

1030

46
1031 Table A2. Material and energy balances of CTH process (Figure 3)
BLOCK CTH REACTOR VLL-FLASH
Inlet Outlet Outlet
STREAM WCO ISO-P COKE PROD GASES WATER HC
Temperature (°C) 25 25 380 50 15 15 15
Pressure (MPa) 0.1 0.1 0.2 0.2 0.1 0.1 0.1
Total mass flow rate (t/d) 1000 341.8 29.4 1312.4 483.3 71.2 757.9
Components mass fraction
Linoleic acid C18H32O2 0.523 0 0 0 0 0 0
Oleic acid C18H34O2 0.186 0 0 0 0 0 0
Pentadecanoic acid C15H30O2 0.179 0 0 0 0 0 0
Nonadecanoic acid C19H38O2 0.112 0 0 0 0 0 0
Isopropanol C3H8O 0 1 0 0 0 0 0
C1-C3 gases 0 0 0 0.124 0.340 0 0
Naphtha (C8H16) 0 0 0 0.178 0.012 0 0.307
Jet fuels (C12H26) 0 0 0 0.300 0 0 0.532
Diesel (C17H36) 0 0 0 0.091 0 0 0.161
CO 0 0 0 0.095 0.265 0 0
CO2 0 0 0 0.127 0.351 0.0004 0
H 2O 0 0 0 0.057 0.012 0.9995 0
H2 0 0 0 0.006 0.015 0 0
Coke 0 0 1 0 0 0 0
Energy
Heating (MW) Fired heater 5.61 -
Cooling (MW) - - CW -3.15
Power (MW) Pumps (P1+P2) 0.003 -
1032
BLOCK FRAC1 FRAC2
Outlet Outlet
STREAM NAPHTHA HEAVY JETFUEL DIESEL
Temperature (°C) 130 220 212 299
Pressure (MPa) 0.1 0.1 0.1 0.1
Total mass flow rate (t/d) 233.0 524.9 403.0 121.9
Components mass fraction
Linoleic acid C18H32O2 0 0 0 0
Oleic acid C18H34O2 0 0 0 0
Pentadecanoic acid C15H30O2 0 0 0 0
Nonadecanoic acid C19H38O2 0 0 0 0
Isopropanol C3H8O 0 0 0 0
C1-C3 gases 0 0 0 0
Naphtha (C8H16) 0.985 0.006 0.008 0
Jet fuels (C12H26) 0.015 0.761 0.989 0.008
Diesel (C17H36) 0 0.233 0.003 0.992

47
 CO 0 0 0 0
CO2 0 0 0 0
H 2O 0 0 0 0
H2 0 0 0 0
Coke 0 0 0 0
Energy
Heating (MW) Reboiler 5.03 Reboiler 1.75
Cooling (MW) Condenser -1.06 Condenser -1.51
Power (MW) - - - -
1033

1034 7. References
1035
1036 [1] IEA International Energy Agency, World Energy Outlook, (2018). https://webstore.iea.org/world-
1037 energy-outlook-2018.
1038 [2] (EIA) U.S. Energy Information Administration, Annual Energy Outlook 2017, (2017) 3.
1039 https://doi.org/DOE/EIA-0383(2012) U.S.
1040 [3] Air Transprt Action Group (ATAG), Aviation Benefits Beyond Borders, 2016.
1041 https://aviationbenefits.org/media/149668/abbb2016_full_a4_web.pdf.
1042 [4] W. Wang, L. Tao, J. Markham, Y. Zhang, E. Tan, L. Batan, M. Biddy, W. Wang, L. Tao, Y. Zhang,
1043 E. Tan, E. Warner, M. Biddy, Review of Biojet Fuel Conversion Technologies,
1044 Www.Nrel.Gov/Publications. (2016) 98. https://doi.org/10.1016/j.rser.2015.09.016.
1045 [5] K. Nanou, T. Roukas, Waste cooking oil: A new substrate for carotene production by Blakeslea
1046 trispora in submerged fermentation, Bioresour. Technol. 203 (2016) 198–203.
1047 https://doi.org/10.1016/J.BIORTECH.2015.12.053.
1048 [6] M.G. Kulkarni, A.K. Dalai, Waste Cooking OilAn Economical Source for Biodiesel: A Review, Ind.
1049 Eng. Chem. Res. 45 (2006) 2901–2913. https://doi.org/10.1021/ie0510526.
1050 [7] G. Liu, B. Yan, G. Chen, Technical review on jet fuel production, Renew. Sustain. Energy Rev. 25
1051 (2013) 59–70. https://doi.org/10.1016/j.rser.2013.03.025.
1052 [8] M. Ameen, M.T. Azizan, S. Yusup, A. Ramli, M. Yasir, Catalytic hydrodeoxygenation of
1053 triglycerides: An approach to clean diesel fuel production, Renew. Sustain. Energy Rev. 80 (2017)
1054 1072–1088. https://doi.org/10.1016/J.RSER.2017.05.268.
1055 [9] S. Popov, S. Kumar, Renewable fuels via catalytic hydrodeoxygenation of lipid-based feedstocks,
1056 Biofuels. 4 (2013) 219–239. https://doi.org/10.4155/bfs.12.89.
1057 [10] B. Veriansyah, J.Y. Han, S.K. Kim, S.A. Hong, Y.J. Kim, J.S. Lim, Y.W. Shu, S.G. Oh, J. Kim,
1058 Production of renewable diesel by hydroprocessing of soybean oil: Effect of catalysts, Fuel. 94 (2012)
1059 578–585. https://doi.org/10.1016/j.fuel.2011.10.057.
1060 [11] C. Kordulis, K. Bourikas, M. Gousi, E. Kordouli, A. Lycourghiotis, Development of nickel based
1061 catalysts for the transformation of natural triglycerides and related compounds into green diesel: a
1062 critical review, Appl. Catal. B Environ. 181 (2016) 156–196.
1063 https://doi.org/10.1016/J.APCATB.2015.07.042.
1064 [12] B. Hočevar, M. Grilc, M. Huš, B. Likozar, Mechanism, ab initio calculations and microkinetics of
1065 hydrogenation, hydrodeoxygenation, double bond migration and cis–trans isomerisation during
1066 hydrotreatment of C6 secondary alcohol species and ketones, Appl. Catal. B Environ. 218 (2017)
1067 147–162. https://doi.org/10.1016/J.APCATB.2017.06.046.
1068 [13] B. Hočevar, M. Grilc, M. Huš, B. Likozar, Mechanism, ab initio calculations and microkinetics of
1069 straight-chain alcohol, ether, ester, aldehyde and carboxylic acid hydrodeoxygenation over Ni-Mo
1070 catalyst, Chem. Eng. J. 359 (2019) 1339–1351. https://doi.org/10.1016/J.CEJ.2018.11.045.
1071 [14] M. Ojeda, N. Osterman, G. Dražić, L. Fele Žilnik, A. Meden, W. Kwapinski, A.M. Balu, B. Likozar,
1072 N. Novak Tušar, Conversion of Palmitic Acid Over Bi-functional Ni/ZSM-5 Catalyst: Effect of
1073 Stoichiometric Ni/Al Molar Ratio, Top. Catal. 61 (2018) 1757–1768. https://doi.org/10.1007/s11244-
1074 018-1046-7.
1075 [15] T.N. Kalnes, M.M. McCall, D.R. Shonnard, Renewable Diesel and Jet-Fuel Production from Fats and

48
1076 Oils, Thermochem. Convers. Biomass to Liq. Fuels Chem. (2010) 468–495.
1077 https://doi.org/10.1039/9781849732260-00468.
1078 [16] M.J. McCall, J.A. Kocal, A. Bhattacharyya, T.N. Kalnes, T.A. Brandvold, Production of aviation fuel
1079 from renewable feedstock, US 8,039,682 B2, 2011.
1080 [17] J.A. Petri, T.L. Marker, Production of Diesel Fuel from biorenewable feedstocks, US 7,511,181 B2,
1081 2009.
1082 [18] A. Asiedu, E. Barbera, R. Naurzaliyev, A. Bertucco, S. Kumar, Waste cooking oil to jet-diesel fuel
1083 range using 2-propanol via catalytic transfer hydrogenation reactions, Biofuels. 0 (2019) 1–14.
1084 https://doi.org/10.1080/17597269.2018.1532754.
1085 [19] R.A.W. Johnstone, A.H. Wilby, I.D. Entwistle, Heterogeneous catalytic transfer hydrogenation and
1086 its relation to other methods for reduction of organic compounds, Chem. Rev. 85 (1985) 129–170.
1087 https://doi.org/10.1021/cr00066a003.
1088 [20] K.M. Isa, T.A.T. Abdullah, U.F.M. Ali, Hydrogen donor solvents in liquefaction of biomass: A
1089 review, Renew. Sustain. Energy Rev. 81 (2018) 1259–1268.
1090 https://doi.org/10.1016/J.RSER.2017.04.006.
1091 [21] T. Kilpiö, P. Mäki-Arvela, T. Salmi, D.Y. Murzin, Modeling of the catalytic deoxygenation of fatty
1092 acids in a packed bed reactor, in: Multiph. Catal. React., John Wiley & Sons, Ltd, 2016: pp. 365–376.
1093 https://doi.org/10.1002/9781119248491.ch15.
1094 [22] Y. Xiang, X. Li, C. Lu, L. Ma, Q. Zhang, Water-improved heterogeneous transfer hydrogenation
1095 using methanol as hydrogen donor over Pd-based catalyst, Appl. Catal. A Gen. 375 (2010) 289–294.
1096 https://doi.org/10.1016/J.APCATA.2010.01.004.
1097 [23] A.S. Amarasekara, M.A. Hasan, Pd/C catalyzed conversion of levulinic acid to γ-valerolactone using
1098 alcohol as a hydrogen donor under microwave conditions, Catal. Commun. 60 (2015) 5–7.
1099 https://doi.org/10.1016/J.CATCOM.2014.11.009.
1100 [24] P. Panagiotopoulou, N. Martin, D.G. Vlachos, Effect of hydrogen donor on liquid phase catalytic
1101 transfer hydrogenation of furfural over a Ru/RuO2/C catalyst, J. Mol. Catal. A Chem. 392 (2014)
1102 223–228. https://doi.org/10.1016/J.MOLCATA.2014.05.016.
1103 [25] O.J. Guerra, G.A.C. Le Roux, Improvements in Petroleum Refinery Planning: 2. Case Studies, Ind.
1104 Eng. Chem. Res. 50 (2011) 13419–13426. https://doi.org/10.1021/ie200304v.
1105 [26] K.F. Haigh, G.T. Vladisavljević, J.C. Reynolds, Z. Nagy, B. Saha, Kinetics of the pre-treatment of
1106 used cooking oil using Novozyme 435 for biodiesel production, Chem. Eng. Res. Des. 92 (2014)
1107 713–719. https://doi.org/10.1016/J.CHERD.2014.01.006.
1108 [27] Y. Liu, R. Sotelo-Boyás, K. Murata, T. Minowa, K. Sakanishi, Production of Bio-Hydrogenated
1109 diesel by hydrotreatment of High-Acid-Value waste cooking oil over ruthenium catalyst supported on
1110 Al-Polyoxocation-Pillared montmorillonite, Catalysts. 2 (2012) 171–190.
1111 https://doi.org/10.3390/catal2010171.
1112 [28] L.F. Chuah, A.R.A. Aziz, S. Yusup, A. Bokhari, J.J. Klemeš, M.Z. Abdullah, Performance and
1113 emission of diesel engine fuelled by waste cooking oil methyl ester derived from palm olein using
1114 hydrodynamic cavitation, Clean Technol. Environ. Policy. 17 (2015) 2229–2241.
1115 https://doi.org/10.1007/s10098-015-0957-2.
1116 [29] M. Toba, Y. Abe, H. Kuramochi, M. Osako, T. Mochizuki, Y. Yoshimura, Hydrodeoxygenation of
1117 waste vegetable oil over sulfide catalysts, Catal. Today. 164 (2011) 533–537.
1118 https://doi.org/10.1016/J.CATTOD.2010.11.049.
1119 [30] P.L. Chu, C. Vanderghem, H.L. MacLean, B.A. Saville, Process modeling of hydrodeoxygenation to
1120 produce renewable jet fuel and other hydrocarbon fuels, Fuel. 196 (2017) 298–305.
1121 https://doi.org/10.1016/j.fuel.2017.01.097.
1122 [31] M. Rigutto, Cracking and Hydrocracking, 2010. https://doi.org/10.1002/9783527630295.ch18.
1123 [32] G. Valavarasu, M. Bhaskar, K.S. Balaraman, Mild Hydrocracking—A Review of the Process,
1124 Catalysts, Reactions, Kinetics, and Advantages, Pet. Sci. Technol. 21 (2003) 1185–1205.
1125 https://doi.org/10.1081/LFT-120017883.
1126 [33] J. Han, A. Elgowainy, H. Cai, M.Q. Wang, Life-cycle analysis of bio-based aviation fuels, Bioresour.
1127 Technol. 150 (2013) 447–456. https://doi.org/10.1016/J.BIORTECH.2013.07.153.
1128 [34] L. Tao, A. Milbrandt, Y. Zhang, W.C. Wang, Techno-economic and resource analysis of
1129 hydroprocessed renewable jet fuel, Biotechnol. Biofuels. 10 (2017) 1–16.
1130 https://doi.org/10.1186/s13068-017-0945-3.
1131 [35] J.H. Gary, G.E. Handwerk, M.J. Kaiser, Petroleum Refining: Technology and economics, CRC Press,

49
1132 2007.
1133 [36] S. Peramanu, B.G. Cox, B.B. Pruden, Economics of hydrogen recovery processes for the puri_cation
1134 of hydroprocessor purge and o}!gases, Int. J. Hydrogen Energy. 24 (1999) 405–424.
1135 [37] G. Towler, R.K. Sinnot, Chemical Engineering design: Principles, Practice and Economics of Plant
1136 and Process Design, Second Edi, Butterworth-Heinemann, 2012.
1137 [38] R. Turton, R.C. Bailie, W.B. Whiting, J.A. Shaeiwitz, D. Bhattacharyya, Analysis, Synthesis, and
1138 Design of Chemical Processes, Prentice-Hall, 2012.
1139 [39] C. San Marchi, B.P. Somerday, Technical Reference for Hydrogen Compatibility Of Materials,
1140 Sandia Tech. Rep. (2012) 292. http://energy.sandia.gov/wp/wp-
1141 content/gallery/uploads/SAND2012_7321.pdf.
1142 [40] http://www.chemengonline.com/ (Last access: 06 March 2018), (n.d.).
1143 [41] V. V Ranade, R. V Chaudhari, P.R. Gunjal, Trickle Bed Reactors: Reactor Engineering and
1144 Applications, Elsevier, Amsterdam, 2011. https://doi.org/https://doi.org/10.1016/B978-0-444-52738-
1145 7.10012-9.
1146 [42] G. Eigenberger, Fixed-Bed reactors, 2012. https://doi.org/10.1002/14356007.b04.
1147 [43] C.C. Herron Jr., B.K. Kruelskie, J.R. Fair, Hydrodynamics and mass transfer on three-phase
1148 distillation trays, AIChE J. 34 (1988) 1267–1274. https://doi.org/10.1002/aic.690340805.
1149 [44] G.D. Ulrich, P.T. Vasudevan, How to estimate utility costs, Chem. Eng. 113 (2006) 66–69.
1150 [45] M. Pearlson, C. Wollersheim, J. Hileman, A techno-economic review of hydroprocessed renewable
1151 esters and fatty acids for jet fuels production, Biofuels, Bioprod. Biorefining. 7 (2013) 89–96.
1152 [46] (EIA) U.S. Energy Information Administration, https://www.eia.gov/ (Last access 06 March 2018),
1153 (n.d.).
1154 [47] “World population review: The 200 largest cities in the United States by population 2019,” (2019).
1155 http://worldpopulationreview.com/us-cities/ (accessed April 4, 2019).
1156 [48] J. Miller, Determining the rate of used cooking oil output by the restaurant industry in the Salt Lake
1157 Valley, UT, 2007.
1158 [49] A. Radich, Biodiesel performance, costs, and use, 2006.
1159 [50] N. Brett, S. Searle, Projected availability of fats, oils, and greases in the U.S., 2016.
1160 [51] U.S. Department of Tranportation (DOT), National Highway Traffic Safety Administration,
1161 Commercial medium- and heavy-duty truck fuel efficiency technology study – Report #1, 2015.
1162 [52] “The state of fuel economy in trucking,” (n.d.). https://www.geotab.com/truck-mpg-benchmark/
1163 (accessed August 15, 2019).
1164 [53] M. Brander, R. Tipper, C. Hutchinson, G. Davis, Consequential and attributional approaches to LCA:
1165 a guide to policy makers with specific reference to greenhouse gas LCA of biofuels, (2008).
1166 http://www.globalbioenergy.org/uploads/media/0804_Ecometrica_-
1167 _Consequential_and_attributional_approaches_to_LCA.pdf.
1168 [54] International Organization for Standardization, ISO 14044:2006(E): Environmental management –
1169 life cycle assessment – requirements and guidelines, (2006).
1170 [55] M. Spielmann, C. Bauer, R. Dones, M. Tuchschmid, Transport services, Data v2.0, Ecoinvent report
1171 no. 14., Dübendorf CH, Switzerland, 2007.
1172 [56] R. Dones, C. Bauer, R. Bollinger, B. Burger, T. Heck, A. Röder, M.F. Emmenegger, R. Frischknecht,
1173 N. Jungbluth, M. Tuchscmid, Life cycle inventories of energy systems: Results for current systems in
1174 Switzerland and other UCTE countries, Ecoinvent report no. 5, Dübendorf CH, Switzerland, 2007.
1175 [57] J. Sheehan, V. Camobreco, J. Duffield, M. Graboski, H. Shapouri, An overview of biodiesel and
1176 petroleum diesel life cycles, a joint study by U.S. Department of Agriculture (DOA) and U.S.
1177 Department of Energy (DOE) prepared by National Renewable Energy Laboratory (NREL), 1998.
1178 [58] R. Hischier, Life cycle inventories of packaging and graphical paper, Data v2.0, Ecoinvent report no.
1179 11, Dübendorf CH, Switzerland, 2007.
1180 [59] A.J. Papa, Propanols, Ullmann’s Encycl. Ind. Chem. (2000).
1181 [60] H.-J. Althaus, R. Hischier, M. Osses, A. Primas, S. Hellweg, N. Jungbluth, M. Chudacoff, Life cycle
1182 inventories of chemicals, Data v2.0, Ecoinvent report no. 8., Dübendorf CH, Switzerland, 2007.
1183 [61] M. Voll, P. Kleinschmit, Carbon – 6: Carbon black, Ullmann’s Encycl. Ind. Chem. (2002).
1184 [62] W. Gendorf, Umwelterklärung, (2000).
1185 [63] P.A. Renzulli, B. Notarnicola, G. Tassielli, G. Arcese, R. Di Capua, Life cycle assessment of steel
1186 produced in an Italian integrated steel mill, Sustain. 8 (2016). https://doi.org/10.3390/su8080719.
1187 [64] F. Mueller-Langer, E. Tzimas, M. Kaltschmitt, S. Peteves, Techno-economic assessment of hydrogen

50
1188 production processes for the hydrogen economy for the short and medium term, Int. J. Hydrogen
1189 Energy. 32 (2007) 3797–3810. https://doi.org/10.1016/J.IJHYDENE.2007.05.027.
1190 [65] https://www.icis.com/ (Last access: 06 March 2018), (n.d.).
1191 [66] U.S. Bureau of Labor Statistics, https://www.bls.gov/home.htm (Last access 15 March 2018), (n.d.).
1192 [67] H.-J. Althaus, C. Bauer, G. Doka, R. Dones, R. Frischknecht, S. Hellweg, S. Humbert, N. Jungbluth,
1193 T. Köllner, Y. Loerincik, M. Margni, T. Nemecek, Implementation of impact assessment methods,
1194 Data v2.2, Ecoinvent report no. 3, Dübendorf CH, Switzerland, 2010.
1195 [68] Y. Zhang, L.M. Colosi, Practical ambiguities during calculation of energy ratios and their impacts on
1196 life cycle assessment calculations, Energy Policy. 57 (2013) 630–633.
1197 https://doi.org/10.1016/J.ENPOL.2013.02.039.
1198

51
Highlights

• Jet fuels production from waste cooking oil by CTH and HRJ is investigated by process simulation;

• HRJ fuel product has large amount of kerosene with no diesel while CTH produced significant

naphtha and diesel;

• Isopropanol is the main driver for the total operating cost for CTH

• EROIs for both processes are energetically favorable, but CTH’s EROI is 1.8 times higher than

HRJ’s.

• The environmental opportunities in using an alcohol-based solvent remarkably outweighs any cost

barriers.
Declaration of interests

☒ The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.

☐The authors declare the following financial interests/personal relationships which may be considered
as potential competing interests: