Techno-economic assessment of bio-oil aqueous phase-to-liquids via Fischer-Tropsch synthesis and based on supercritical water reforming

Depósito de Investigación de la Universidad de Sevilla https://idus.us.es/ This is an Accepted Manuscript of an article published by Elsevier in Energy Conversion and Management, Vol. 154, on December 2017, available at: https://doi.org/10.1016/j.enconman.2017.10.096 © 2017 Elsevier. En idUS Licencia Creative Commons CC BY-NC-ND 1 Techno-economic assessment of bio-oil aqueous phase-to-liquids via Fischer-1 Tropsch synthesis and based on supercritical water reforming 2 F.J. Campanario, F.J. Gutiérrez Ortiz 3 Departamento de Ingeniería Química y Ambiental, ETS de Ingeniería, Universidad de Sevilla 4 Camino de los Descubrimientos, s/n. 41092 Sevilla, Spain 5 Phone no: + 34 954 48 72 60 / 68 6 Corresponding author email: “Francisco Javier Gutiérrez Ortiz” [email protected] 7 Abstract 8 High energy demand along with large capital costs have been the main drawbacks of 9 Fischer-Tropsch plants, which may call into question the economic viability of the Fischer-10 Tropsch process. The second issue is the focus of this paper, which presents a techno-economic 11 assessment of biofuels production by a low-temperature Fischer-Tropsch synthesis with 12 electricity as a co-product from supercritical water reforming of the bio-oil aqueous phase. A 13 plant size of 60 t/h was considered and a heat-integrated process was designed to be energy 14 self-sufficient, which includes syngas production and upgrading, as well as liquid fuels 15 production by Fischer-Tropsch synthesis and refining. The simulation and optimization was 16 performed with the aid of Aspen Plus, and some case-studies were performed. Using a feeding 17 concentration of 25 wt.%, 2.74 t/h biofuels and 5.72 MWe were obtained. In this case, by 18 performing a discounted cash flow analysis, with 10% rate of return and 100% equity financing, 19 the minimum selling prices for the refined FT-gasoline, FT-diesel and FT-jet fuel were 1.20, 20 0.93 and 0.26 €/kg (0.84, 0.75 and 0.20 €/L), respectively, which are competitive prices with 21 respect to the market values of the equivalent fossil fuels. Likewise, the decrease in the selling 22 prices as the plant capacity increases was also analyzed. 23 Keywords 24 Fischer-Tropsch, Supercritical Water, Techno-economic, Bio-oil, Aqueous-phase, Biofuel 25 2 1. Introduction 26 Although there are many renewable emerging technologies for large-scale electricity 27 production, without or with low carbon emissions, the transport sector is entirely based on the 28 use of fossil fuel. The rising price of crude oil until a few years ago and the potential fluctuations 29 of oil market prices in future, along with global environmental problems, have promoted new 30 efforts to find alternative processes to produce transportation fuels with improved energy and 31 environmental efficiencies. This way, we have proposed a new process of Fischer-Tropsch (FT) 32 fuels production from the syngas obtained by supercritical water reforming (SCWR) of the bio-33 oil aqueous phase. In the previous paper [1], Fischer-Tropsch synthesis was designed and 34 analyzed, which provided a novel and efficient valorization for the waste-aqueous phase of bio-35 oil. This previous study has been completed by carrying out a techno-economic assessment of 36 the overall process, once designed and revised some parts of the original process. 37 FT synthesis is a technology to produce a variety of liquid transportation fuels, such as 38 diesel, gasoline and jet fuel from syngas, obtained by other technologies, such as gasification. 39 FT-biofuels exhibit attractive advantages: they are well-suited to the existing market regarding 40 technical specifications (for example, a high cetane number for FT-diesel and the absence of 41 sulfur or nitrogen) and compatible with available vehicle engine designs [2]. 42 The interest in FT-liquids from gas-to-liquids (GTL), coal-to-liquids (CTL) and biomass-43 to-liquids (BTL) process has been increasing. Currently, there are several commercial plants 44 around the world that produce FT-liquids from GTL [3] and CTL [4] process. However, FT-45 liquids production from biomass is still under development due to the great investment costs 46 required, the high feedstocks prices and the low energy density of biomass, among other 47 reasons. As an alternative, solid biomass might be converted into liquids bio-oils by a fast 48 pyrolysis process, with a higher energy density as compared with solid biomass. However, this 49 process generates a waste-aqueous stream (20-30 wt.% organic compounds), named bio-oil 50 3 aqueous phase, which may be valorized through supercritical water (SCW) reforming that is an 51 emerging technology investigated by many authors and ourselves in the last years [5-7]). 52 Supercritical water has properties very different from those of liquid water, which give 53 advantages regarding other reforming technologies. Thus, the dielectric constant of SCW is 54 much lower, the number of hydrogen bonds is much lower and their strength is much weaker. 55 As a result, SCW behaves like many organic solvents, so organic compounds have a complete 56 miscibility in SCW. 57 The aim of the paper is to estimate the total investment and production costs, obtaining the 58 minimum selling price of the products (FT-gasoline, FT-diesel and FT-jet-fuel), by considering 59 the selling of electricity and different scenarios, thus assessing the feasibility of the full process 60 in a hypothetical plant size of 60 t/h of bio-oil aqueous-phase. The techno-economic analysis 61 was performed in terms of net present value, payback period and break-even price, and a 62 sensitivity study was carried out to investigate the effect of several factors, such as the 63 uncertainty of the capital investment or the feedstock price, on the minimum selling price of 64 the biofuels. 65 2. Methodology 66 In this section, a summary of the process design and simulation is given and the methods 67 used for the process economics are described, for a SCWR-LTFT plant with a feeding of 60t/h 68 during 8000 h per annum. 69 2.1. Process design and simulation 70 The design and simulation of the process was extensively described in the previous paper 71 [1], and only a summary is provided below. The process was designed to maximize the biofuels 72 and electricity production for an industrial size of 60 t/h of bio-oil aqueous phase [8,9] under 73 different feeding concentrations. The process is heat integrated and energy self-sufficient. 74 4 The SCWR reactor operates at 240 bar and 800 ºC to increase the CO and H 2 yields [7]. 75 The product gas leaving this unit is expanded to 15-40 bar by an expander (TURBINE) to 76 generate electrical power. A fraction of syngas may enter a high-temperature water-gas shift 77 (HWGS) reactor (operating at 350 ºC) to increase the H 2 yield when the H 2 /CO molar ratio at 78 inlet Fischer-Tropsch reactor is lower than that specified. Then, the outlet gas is mixed with the 79 by-pass stream and cooled down to 35 ºC to condense the water. A fraction of syngas is sent to 80 the furnace to achieve energy self-sufficiency (Figure 1). 81 82 Figure 1. SCWR and HWGS section 83 84 The rest goes into two pressure swing adsorption (PSA) units, to obtain pure H 2 , pure CO 85 and a CO 2 +CH 4 -rich stream [10]. A fraction of H 2 along with the CO stream are conveyed to 86 the FT synthesis loop. The CO 2 +CH 4 -rich stream enters a dry-reforming (DR) reactor to 87 increase the amount of CO and H 2 fed to the low-temperature Fischer-Tropsch (LTFT) reactor; 88 besides, CO 2 emissions are reduced (environmental advantage). DR operating conditions are 89 600 ºC and 1 bar to boost the CO and H 2 production, as verified by simulation [11]. The gas 90 leaving the DR is cooled to remove the water, and then it is compressed before entering the 91 second PSA system to further increase the flow of H 2 and CO fed to the FT reactor (Figure 2). 92 5 93 Figure 2. PSA systems and DR section 94 95 Reactions in LTFT reactor, which operates at 220 ºC, 20 bar and 2.0 inlet molar ratio 96 H 2 /CO, are highly exothermic (about 150 kJ per mole of CO converted [12]). The heat is 97 released by an evaporator, and saturated steam is formed (212 ºC, 20 bar). A fraction of this 98 stream is sent to preheat the inlet stream of the FT reactor, and the rest of the steam is expanded 99 to 1 bar in a turbine (TURB2) to produce additional electricity. The statistical distribution model 100 by Anderson-Schulz-Flory (ASF) was used to obtain hydrocarbon molar fractions [13]. An α101 value (probability of chain growth) of 0.90 was obtained using the equation by Song et al. [14] 102 for cobalt-based catalysts used in LTFT. CO conversion per pass was assumed to be 50 % [15], 103 and the overall CO conversion in the loop was close to 90%. The stream leaving the FT reactor 104 is cooled, and liquid phase is separated from the gas. The former is sent to a decanter to separate 105 water from hydrocarbons, while the gas phase splits in a recycle stream and a purge to avoid 106 the build-up of inert gases (Figure 3). 107 The purge stream of the FT loop and the gas leaving the hydrocracking reactor are mixed, 108 compressed and cooled before entering a flash separator to remove most of the gases from 109 liquid. This latter and the liquid phase of the FT loop are mixed and sent to three distillation 110 6 columns to separate the different biofuels (gasoline, jet-fuel and diesel). The lightest fraction is 111 re-used in the process, and the heaviest fraction is sent to the hydrocracking. Hydrogen required 112 by the hydrocracking is that non-converted in the FT reactor (separated by the PSA4 unit) plus 113 a fraction of H 2 leaving the PSA1 unit. The hydrocracking reactor operates at 360 ºC, 35 bar, 114 and 0.06 kg-H 2 /kg-wax using a platinum-based catalyst (Figure 4). 115 116 Figure 3. FT loop section 117 118 119 Figure 4. Distillation and hydrocracking section 120 121 7 Finally, and regarding the heat integration, several heat exchangers were suitably located 122 in the process. The strategy followed is based on minimizing the exergy losses, so small 123 temperature-driving forces must be achieved by using countercurrent flows and small 124 temperature differences at the ends of the exchangers. In this approach, the high-temperature 125 hot streams heat up the high-temperature cold streams, and the low-temperature hot streams 126 warm up the low-temperature cold streams, by taking the reformer along with the furnace as a 127 central point, where maximum temperature is required. 128 All the simulations were performed by Aspen Plus, and more details of the process and 129 units are given in section 3.2 and elsewhere [1]. 130 2.2. Process economics 131 Purchase costs of some units were estimated from open access reports, by sizing those units 132 and scaling their cost by a factored estimation method described by Eq. (1) [16]: 133        󰇛𝑍󰇜 (1) 134 where Z is the ratio between capacities of the new and reference equipment (units can be m2, 135 kg/h or kW, among others), and g is the scaling factor, which is taken from literature or assumed 136 to be 0.6 [16]. Costs of PSA units and all reactors were obtained using this rule, because of 137 information has been found [17-19]. 138 Purchase costs of other more common units were estimated by Capcost software [20] or by 139 the website of Peters and Timmerhaus [21], when Capcost was not applicable. In this program, 140 the bare module cost (CBM) is the sum of direct and indirect costs associated with equipment 141 purchase and installation. These costs are estimated using a multiplication factor that considers 142 the specific construction materials and the operating pressure with respect to the base unit cost 143 (C0BM), which is based on the use of carbon steel and atmospheric pressure. Thus, CBM includes 144 indirect expenses (freight, insurances, taxes, construction overheads, contractor engineering 145 8 expenses, and so on), structural supports, piping and material required for installation, among 146 others. 147 The total costs of units were updated to the year 2016 by using the averaged chemical 148 engineering plant cost index (CEPCI) that accounts for the effect of inflation by Eq. (2). 149 𝑃𝑟𝑒𝑠𝑒𝑛𝑡 𝑐𝑜𝑠𝑡  𝑂𝑟𝑖𝑔𝑖𝑛𝑎𝑙 𝑐𝑜𝑠𝑡  󰇡         󰇢 (2) 150 CEPCI value is 541.7 for the year 2016, while CEPCI when original costs were obtained 151 from literature used in this paper corresponding to the years 2001, 2002 (January, used in [16]), 152 2011, 2014 and 2015 are 397.0, 390.4, 585.7, 576.1 and 556.8, respectively [22]. 153 All costs were computed in $, updated to the year 2016 and then presented in €, using a 154 reference exchange rate of 1 € = 1.183 $ (October 11, 2017). 155 On the other hand, the bare module cost (CBM) plus contingency and fees costs (assumed 156 to be 15% and 3% of CBM, respectively) is the total module cost (CTM). Finally, the total 157 investment cost (TIC) is the sum of CTM and the auxiliary facilities costs, which were assumed 158 to be equal to 50% of C0BM. 159 Fixed operating costs were estimated by a percentage of total investment costs discounting 160 the contingency and fees costs (TICr), and variable operating costs are due to consumables 161 (cooling water, waste water treatment and catalysts for reactors). Prices for cooling water and 162 waste water were assumed to be 0.010 €/t [18], since it may be used for district heating, and 163 0.304 €/m3 [23], respectively, while catalysts costs (€/year) were estimated to be 1% of TIC 164 [24]. The lifetime of these catalysts was assumed to be three years [24]. Feedstock costs are 165 usually high and could be essential in the calculation of selling prices of biofuels. In this study, 166 the feedstock cost is null, because of the SCWR-LTFT plant is installed within the bio-oil 167 production plant. Furthermore, this leads to a null cost of land. 168 15 Table 3 (cont.) 281 Code Equipment and Specifications (*) Design characteristics Capacity (*) LTFT Low-Temperature Fischer-Tropsch reactor (RStoic); Operating temperature: 220 ºC; Operating pressure: 20 bar Slurry phase reactor; 1214 tubes; 3.35 m ID; 6.70 m height; 50 mm ID for each tube; superficial gas velocity: 0.35 m/s; catalyst concentration: 0.20 60.0 m3 SEP Gas-liquid separators (Flash); Cylindrical, vertical; Temperature: 35 ºC; Pressure drop: 0.1 bar 2.00 m ID; 6.25 m height; carbon steel 19.6 m3 SEP2 3.70 m ID; 5.90 m height; carbon steel 63.2 m3 SEP3 1.90 m ID; 3.25 m height; carbon steel 10.4 m3 SEP4 0.80 m ID; 2.50 m height; carbon steel 1.3 m3 DECANT Liquid-liquid separator (Decanter); Temperature: 35 ºC.; Pressure drop: 0.1 bar; Cylindrical, horizontal 0.90 m ID; 2.70 m height; carbon steel 1.9 m3 Code Equipment, specifications and design characteristics Capacity (2) PSA1 Pressure swing adsorption unit (First PSA system); It removes most the H2 (95 %) from the other gases [10]; Outlet pressure: variable (H2 stream, 16-H), 1.1 bar (the rest of gases) 18259 Nm3/h PSA2 Pressure swing adsorption unit (First PSA system); CO-rich stream: 98 % CO, 1 % for CO2 and CH4 [10]; Outlet pressure: 1.1 bar (top and bottom) 10942 Nm3/h PSA1A Pressure swing adsorption unit (Second PSA system); It removes most the H2 (95 %) from the other gases [10]; Outlet pressure: variable (H2 stream, 31-H), 1.1 bar (the rest of gases) 12179 Nm3/h PSA2A Pressure swing adsorption unit (Second PSA system); CO-rich stream (bottom): 98 % CO, 1 % for CO2 and CH4 [10]; Outlet pressure: 15 bar (34 stream), 1.1 bar (35-CO stream) 9347 Nm3/h PSA3 Pressure swing adsorption unit (Second PSA system); CO2-rich stream (bottom): 90 % CO2, 9 % CH4, 0.5 % CO and H2 [10]. Outlet pressure: 1.1 bar (top and bottom) 5371 Nm3/h PSA4 Pressure swing adsorption unit; It removes most the H2 (95 %) from the other gases [10]; Outlet pressure: same SEP4 1631 Nm3/h DEST1 Distillation column (heavy-light); C13H28 recovery (distillate stream): 90 %, C14H28 recovery (bottom stream): 99%; 45 Sieve plates; 1.67 m ID; 21.45 m height; Down-comer area is 10 % of column area; partial condenser (1.3 bar; 44.5 ºC); kettle reboiler (1.5 bar; 319.5 ºC); carbon steel 47.0 m3 DEST2 Distillation column 2 (wax-diesel); C20H42 recovery (distillate stream): 90 %, C21H44 recovery (bottom stream): 90 %; 44 Sieve plates; 1.23 m ID; 21.00 m height; Down-comer area is 10 % of column area; total condenser (1.3 bar; 284.6 ºC); kettle reboiler (1.5 bar; 420.1 ºC); carbon steel 24.9 m3 DEST3 Distillation column 3 (jet fuel-gasoline); C9H20 recovery (distillate stream): 95 %, C10H20 recovery (bottom stream): 95%; 36 Sieve plates; 1.26 m ID; 17.40 m height; Down-comer area is 10 % of column area; total condenser (1.2 bar; 25.9 ºC); kettle reboiler (1.3 bar; 211.2 ºC); carbon steel 21.7 m3 SCWR + FURNACE Supercritical Water Reforming (RGibbs, Operating temperature: 800 ºC; Pressure drop: 0.0 bar) + Furnace-combustor (RStoic, Combustion of everything able to be oxidized; 1000 ºC @ 1 bar Surplus heat flow to SCWR) 60000 kg/h Code Equipment and specifications of splitters (SP), coolers (C) and heaters (H) SP1 Split fraction stream 10: variable to achieve the hydrogen molar flow required to adjust the H2/CO molar ratio. SP2 Split fraction stream T-PSA1: variable (aimed at achieving the energy self-sufficiency constraint). SP3 Split fraction stream T-MIX6: variable on required H2 for hydrocracking. SP4 Split fraction stream 46: variable (aimed at maximizing CO total conversion as much as possible). SP5 Split fraction stream 51-G: variable (with the goal of sending the required hydrogen to the hydrocracking reactor) SP6 Split fraction stream CH4-O: variable because methane could be sold if energy self-sufficiency was achieved SP7 Split fraction stream 06V: variable (with the goal of preheating the LTFT inlet stream, and producing electrical p ower by the steam turbine with the rest of steam (02V stream)) C1-C5 Heat flow for endothermic heat reaction in DRYREF and heat flow from LTFT, respectively C2-C3-C4 Sink of the heat flow coming from the condenser of the DEST3, DEST1, DEST2, respectively H1-H2-H3 Heat flow towards the reboiler of DEST1, DEST2, DEST3, respectively (*) For the 25 wt. % case. 282 283 284 16 Heaters and coolers are cylindrical units with vertical pipes, and the material is carbon steel 285 or chromium alloy depending on the temperature and pressure of the streams. Costs for these 286 units were estimated by the website of Peters and Timmerhaus [21]. 287 Table 4. Bare module costs of the process units at different feeding concentrations. 288 COST (k€ from (k$)2016) Equipment 35 (Case1)* 35 (Case1)** 30 (Case2) 25 (Case 3) 20 (Case 4) 15 (Case 5) P1 47.54 47.54 47.62 47.67 47.72 47.76 SCWR + FURNACE 16 141.44 16 141.44 16 141.44 16 141.44 16 141.44 16 141.44 TURB 12 075.54 12 075.54 9 070.52 9 090.99 9 108.99 9 122.67 HWGS 741.72 741.72 491.00 0.00 0.00 0.00 SEP1-4 1 184.34 1 184.34 821.73 579.95 387.41 218.41 DECANT 22.31 22.31 19.58 17.08 14.80 12.72 PSA systems + PSA4 20 392.06 20 392.06 17 044.41 14 169.63 10 514.25 5 882.19 COMP1-4 6 723.73 6 727.22 5 225.36 4 065.31 2 764.73 1 574.86 DR Reactor 13 011.75 13 011.75 10 852.94 8 405.78 6 095.83 2 938.64 COMP5-7 130.56 130.73 121.05 110.24 67.59 66.08 LTFT Reactor 6 555.28 6 555.28 5 787.92 4 805.73 3 464.92 1 643.92 DEST1-3 4 725.24 4 725.24 4 212.11 3 730.70 2 920.33 2 311.70 P2 17.78 17.78 16.74 15.43 13.57 11.57 HYDR Reactor 4 699.24 4 699.24 4 145.30 3 425.14 2 464.16 1 196.89 Pumps (P3-7) + FAN1 71.34 71.20 68.45 64.27 58.82 52.89 TURB2 1 216.82 1 216.86 1 161.28 1 066.22 907.11 526.69 HE01-18 6 898.15 6 845.08 5 482.88 4 002.15 2 512.00 1 922.23 Heaters and coolers 2 002.54 2 002.54 1 718.22 1 370.02 957.28 417.27 Total 96 657.39 96 607.87 82 428.54 71 107.74 58 440.95 44 087.93 Notes: (*) the expander outlet pressure is 18.5 bar (selling biomethane); (**) the expander outlet pressure is 17.0 bar (only electricity is sold) 289 290 A double-piston pump with a dampener, made of AISI 316, was chosen to pump bio-oil 291 aqueous phase up to 240 bar (P1), and its purchase cost was estimated by the previous website 292 [21]. The other pumps for cooling water are centrifugal and their costs were estimated by 293 Capcost. 294 On the other hand, both centrifugal (C1-C4) and rotary (C5-C7) compressors were used 295 because the former type allows the compression of flow-rates higher than rotary units; likewise, 296 centrifugal compressors have also a higher electrical power. The costs of these units were 297 obtained by Capcost, like the cost of the centrifugal fan made of carbon steel that boosts the air 298 along with the mixture of gases coming from separators and splitters to the furnace. 299 17 Regarding the expanders, three or four units are necessary because the total electrical power 300 is higher than the maximum value accepted by Capcost. The 35 wt.% case results in a higher 301 electrical power than that for lower feeding concentrations, because the expander outlet 302 pressure is lower (17.5-18.0 bar). Additionally, a steam turbine is used in all the cases. Nickel 303 alloy is used for expanders when pressure is 240 bar and temperature is higher than 350 ºC. 304 The cost estimation of reactors was estimated from open access reports and surveys of 305 equipment vendors as follows: SCWR-Furnace [18], HWGS reactor [19], DR reactor [17], 306 hydrocracking reactor [19] and LTFT reactor [33]. The costs were scaled using the rule of ‘six 307 tenths’ (Eq. 1) using the mass flow-rate of the stream entering/leaving the SCWR, HWGS, DR 308 and hydrocracking reactor, and the volume for the LTFT reactor as capacity factors. The slurry 309 LTFT reactor volume was calculated adding the catalyst volume, which was estimated by using 310 the kinetic model by Krishna and Sie [34], to the pipe volume (diameter and spacing are 50 mm 311 and 150 mm, respectively) used to release the reaction heat, and assuming a height-to-diameter 312 ratio of 2.0. The LTFT reactor was designed to operate under similar conditions to those used 313 by other researchers [34] (240 ºC, 30 bar and an α-value of 0.9), with a superficial gas velocity 314 and catalyst concentration of 0.25 m/s and 0.20, respectively, and assuming a CO conversion 315 per pass of 50 %. 316 All the purchase costs estimated were updated to the current date by the CEPCI index. 317 Considering the case-studies on the feeding concentration, higher volumes, heat exchange 318 areas, feeding processing streams and powers in process units are found for the more 319 concentrated bio-oil aqueous phase cases. Thus, larger process equipment with higher purchase 320 costs are required, especially regarding compressors, expanders, PSAs, and reactors (except 321 SCWR, where all cases present the same mass flow-rate). Table 5 shows the estimated 322 investment costs for all the feeding concentrations, ranging from 60.5 M€ at 15 wt.% to 134.3 323 M€ at 35 wt.%. In this study, supercritical water (SCWR, heat exchangers, high pressure-pump 324 18 and expander) and syngas conditioning sections (PSAs, heat exchangers, compressors, HWGS 325 and DR reactors) are the most expensive units (72.6 % of total equipment costs for the 35 wt.% 326 case). 327 Table 5 also includes the operating costs for all the feeding concentrations, which are 328 greater at higher concentrations. This is because of fixed operating costs are computed based 329 on a percentage of investment costs (TICr), and variable operating costs involves larger 330 consumptions (especially, utilities) as the concentration of bio-oil aqueous fraction increases. 331 In the variable operating costs, the cost of bio-oil aqueous fraction is considered null, as above-332 mentioned. 333 Table 5. Estimated investment costs and operating costs assuming 8000 h per year for all the feeding concentrations. 334 CAPITAL COST (k€ from (k$)2016) Aqueous fraction (wt. %) 35 (Case1)* 35 (Case1)** 30 (Case2) 25 (Case 3) 20 (Case 4) 15 (Case 5) 𝐶  (under base-case conditions) 40 415.96 40 416.33 34 618.61 29 294.63 23 596.58 16 934.73 CBM (bare module cost) 96 657.39 96 607.87 82 428.54 71 107.74 58 440.95 44 087.93 Contingency and fees 17 398.33 17 389.42 14 837.14 12 799.39 10 519.37 7 935.83 Auxiliary facilities 20 207.98 20 208.17 17 309.30 14 647.31 11 798.29 8 467.37 Grassroots costs 134 263.70 134 205.46 114 574.98 98 554.44 80 758.62 60 491.12 Total investment cost (TIC) 134 263.70 134 205.46 114 574.98 98 554.44 80 758.62 60 491.12 FIXED OPERATING COSTS (k€/year from (k$)2016/year) % TICr Aq. fraction (wt. %) 35 (Case1)* 35 (Case1)** 30 (Case2) 25 (Case 3) 20 (Case 4) 15 (Case 5) 1.56 Labor 1 823.10 1 822.33 1 555.91 1 337.78 1 095.73 819.86 1.50 Maintenance 1 752.98 1 752.24 1 496.07 1 286.33 1 053.59 788.33 3.07 General expenses 3 587.77 3 586.25 3 061.95 2 632.68 2 156.34 1 613.45 0.44 Operating services 514.21 513.99 438.85 377.32 309.05 231.24 1.32 Logistic & others costs 1 542.62 1 541.97 1 316.54 1 131.97 927.16 693.73 0.50 Insurance 584.33 584.08 498.69 428.78 351.20 262.78 8.39 Total 9 805.00 9 800.87 8 368.01 7 194.85 5 893.07 4 409.39 VARIABLE OPERATING COSTS (k€ from (k$)2016) Price Aq. Frac. (wt. %) 35 (Case1)* 35 (Case1)** 30 (Case2) 25 (Case 3) 20 (Case 4) 15 (Case 5) 0.010 €/t Cooling water 29.77 29.77 28.86 25.49 24.44 22.18 0.304 €/m3 Waste water treatment 125.13 125.13 134.49 140.62 148.35 154.27 1 % of TIC Catalysts *** 447.55 447.35 381.92 328.51 269.20 201.64 Total 602.45 602.25 545.27 494.63 441.99 378.08 Notes: (*) the expander outlet pressure is 18.5 bar (selling biomethane); (**) the expander outlet pressure is 17.0 bar (only electricity is sold); (***) 335 after 3 years, catalysts are changed in reactors. 336 337 Figure 5 shows a breakdown of the total annual costs of biofuels production at different 338 feeding concentrations, by annualizing the total capital costs at a 7% rate over the project 339 19 lifetime. In all cases, fixed operating costs and installed purchase units’ costs are the highest of 340 the total annual cost, and they are of the same magnitude order. Both costs decrease as the 341 feeding concentration decreases (by half from 35 to 15 wt.%), as the size is significantly 342 reduced for many units. This also occurs with the rest of investment costs (contingencies, fees 343 and auxiliaries), which depend on the costs of installed purchase units. Variable operating costs 344 are much lower than the former ones. 345 346 347 Notes: * the expander outlet pressure is 18.5 bar (selling biomethane) 348 Figure 5. Breakdown of annual costs of biofuels and electricity production at different feeding concentrations 349 (60 t/h). 350 3.3. Cash flow analysis 351 Cash flow analyses were carried out for all the feeding concentrations to calculate the 352 minimum selling prices of a mass unit of FT-diesel, FT-jet fuel and FTgasoline, in the way 353 above-mentioned, with similar technical specifications products to those previously reported 354 [30] and norms such as UNE-EN-ISO 12185 and ASTM D 4052. Thus, the break-even points 355 of FT-diesel, FT-jet fuel and FT-gasoline are 0.93 €/kg, 0.26 €/kg and 1.20 €/kg (0.75, 0.20 and 356 0.84 €/L), respectively, at a feeding concentration of 25 wt.%. 357 20 The variation of the cumulative non-discounted and discounted (present cash flow) over 358 the lifetime of the plant is shown in Figure 6. At the beginning (year zero in which the plant is 359 built), the cash flow is negative because of 50% of TIC is paid. At the end of the first year, the 360 process is ready to start, but the rest of capital costs (50% TIC) and working capital (5% TIC) 361 are paid, so this is the point with the most unfavorable cash flow. From the second year onwards, 362 money from sales (biofuels and electricity) is received because the production begins, and the 363 investment starts to be recovered with an internal rate return of 10%. The cash flow is recovered 364 more slowly from the seventeenth year because of the depreciation of capital goods is no longer 365 accounted. Regarding the number of years to recover the initial investment of the project 366 (discounting the land and working capital), the simple (non-discounted) and discount payback 367 period are 8.5 and 17.5 years, respectively, from the beginning. As the payback period is lower 368 than lifetime of the project, this may be profitable. 369 370 Figure 6. Cumulative non-discounted and discounted cash flow diagrams over 371 the lifetime, by considering 100 % equity financing (base-case). 372 -120 -80 -40 0 40 80 120 160 01234567891011121314151617181920 M€ Lifetime (year) Non-discounted cash flow Discounted cash flow PB for investment recovering 21 Another scenario would be the debt financing of some fraction of the capital. By using 373 fixed annuities, if 100% of the required capital was borrowed at 7% real interest with a 15-year 374 repayment term (matching the depreciation time), the required selling price for FT-diesel would 375 be reduced (0.78 €/kg (0.63 €/L) versus 0.93 €/kg, both at 25 wt.%). This is because when the 376 access to credit is available, NPV increases when the return on investment (10%) is higher than 377 the interest rate on debt (7%). 378 Additionally, CO2 is produced in SCWR and HWGS reactors, and partially converted in 379 DR reactor. The surplus CO2 is conveyed to PSA3 to be separated from CH4, and then sent to 380 sequestration. CO2 production increases as the feeding concentration increases (Table 2). As 381 CO2 comes from a renewable source, and it is captured and sequestered, the process economics 382 would improve if an international trading system for CO2 emissions allowances was 383 established, but currently even carbon tax is still uncommon in most places of the world. 384 Regarding the other feeding concentrations, Figure 7 shows the cumulative non-385 discounted cash flow for the 15, 25, and 35 wt.% cases, which moves from negative to positive 386 values, and a positive final cash flow is achieved when the project ends. The change of 387 amplitude increases as the concentration increases. However, the minimum selling prices of the 388 biofuels at the lowest feeding concentration are slightly higher than those at the highest feeding 389 concentration (0.98 and 0.90 €/kg FT-diesel at 15 and 35 wt.%, respectively), despite the lower 390 total investment costs for the 15 wt.% case. 391 Results of this study show that the products have competitive selling prices with respect to 392 the market prices in Spain for the same products coming from crude oil (1.41 €/kg (1.14 €/L) 393 FT-diesel, 1.83 €/kg (1.28 €/L) FT-gasoline and 0.40 €/kg (0.30 €/L) FT-jet fuel [26,35]). 394 395 396 22 397 Figure 7. Cumulative non-discounted cash flow diagrams over the lifetime 398 considering 100 % equity financing at 15, 25 and 35 wt.%. 399 3.4. Effect of the plant capacity 400 In order to inspect the effect of the plant capacity on the cost of production and, hence, on 401 the biofuels selling prices, units were re-designed and the process was simulated for seven 402 capacities, expressed by the mass feeding flow-rate: 20, 40, 60, 80, 100, 150, and 200 t/h. Table 403 6 shows the production of biofuels and net electrical energy for each capacity. 404 Table 6. Biofuel production at different plant capacities (25 wt.%) 405 Plant capacity (t/h feeding) 20 40 60 80 100 150 200 FT-diesel (kg/h) 458 916 1 374 1 831 2 289 3 434 4 578 FT-jet fuel (kg/h) 299 598 898 1 197 1 496 2 244 2 992 FT-gasoline (kg/h) 156 312 467 623 779 1 169 1 558 Electrical energy (kWh) 1 908 3 816 5 728 7 653 9 566 14 349 19 132 406 Table 7 and Figure 8 give the same information as Table 5 and Figure 5, respectively, 407 but when changing the plant capacity for the 25 wt.% case. The increase in capital costs and 408 operating costs is lower than the corresponding increase in plant capacity, as expected on the 409 -150 -100 -50 0 50 100 150 200 01234567891011121314151617181920 M€ Lifetime (year) 15 wt.% 25 wt.% 35 wt.% 23 economy of scale, so the production cost is reduced when the plant capacity increases. Figure 410 9 shows how the biofuels selling prices vary depending on the plant capacity by considering 411 both equity and debt financing at 7% rate. As can be observed, the higher the plant capacity is, 412 the lower the selling prices are. This makes it possible to estimate a theoretical optimum plant 413 capacity, which would be about 200 t/h as the curves tend to stabilize at such high capacities 414 and become almost flat. However, currently, this plant scale might be too high because the 415 feedstock likely could not be guaranteed, so it would not be realistic; this way, a mass feeding 416 flow rate of 100 t/h could be considered as a maximum [8,9]. 417 418 Table 7. Estimated investment costs and operating costs assuming 8000 h per year for all different plant capacities (25 wt.%). 419 CAPITAL COST (k€ from (k$)2016) Plant capacity (t/h feeding) 20 40 60 80 100 150 200 𝐶  (under base-case conditions) 14 094.37 22 205.90 29 294.63 35 698.67 41 980.86 56 102.93 69 653.23 CBM (bare module cost) 34 221.06 55 640.04 71 107.74 85 486.79 100 809.76 140 299.29 179 395.59 Contingency and fees 6 159.79 10 015.21 12 799.39 15 387.62 18 145.76 25 253.87 32 291.21 Auxiliary facilities 7 047.19 11 102.95 14 647.31 17 849.34 20 990.43 28 051.46 34 826.61 Grassroots costs 47 428.04 76 758.20 98 554.44 118 723.75 139 945.95 193 604.62 246 513.41 Total investment cost (TIC) 47 428.04 76 758.20 98 554.44 118 723.75 139 945.95 193 604.62 246 513.41 FIXED OPERATING COSTS (k€/year from (k$)2016/year) % TICr Plant capacity (t/h) 20 40 60 80 100 150 200 1.56 Labor 643.78 1 041.19 1 337.78 1 612.04 1 900.08 2 626.27 3 341.87 1.50 Maintenance 619.02 1 001.14 1 286.33 1 550.04 1 827.00 2 525.26 3 213.33 3.07 General expenses 1 266.94 2 049.01 2 632.68 3 172.42 3 739.27 5 168.37 6 576.62 0.44 Operating services 181.58 293.67 377.32 454.68 535.92 740.74 942.58 1.32 Logistic & others costs 544.74 881.01 1 131.97 1 364.04 1 607.76 2 222.23 2 827.73 0.50 Insurance 206.34 333.71 428.78 516.68 609.00 841.75 1 071.11 8.39 Total 3 462.41 5 599.74 7 194.85 8 669.90 10 219.04 14 124.63 17 973.24 VARIABLE OPERATING COSTS (k€/year) Price Plant capacity (t/h) 20 40 60 80 100 150 200 0.010 €/t Cooling water 8.76 17.53 25.49 32.62 40.75 61.13 81.51 0.304 €/m3 Waste water treatment 46.78 93.80 140.62 175.72 234.37 351.55 468.74 1 % of TIC Catalysts * 158.09 255.86 328.51 395.75 466.49 645.35 821.71 Total 213.63 367.18 494.63 604.08 741.61 1 058.03 1 371.96 Notes: (*) after 3 years, catalysts are changed in reactors. 420 421 24 422 Figure 8. Breakdown of annual costs of biofuels and electricity production at different plant capacities (25 wt.%). 423 424 425 Figure 9. FT-biofuels selling price versus plant capacity (25 wt.%) with 100% equity financing 426 (solid lines) and debt financing at 7% rate (dash lines). 427 428 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 190 200 FT-biofuels selling price (€/kg) Capacity (t/h) Gasoline_equity Jet Fuel_equity Diesel_equity Gasoline_debt Jet Fuel_debt Diesel_debt 31 [29] M.J.A. Tijmensen, A.P.C. Faaij, C.N. Hamelinck, M.R.M. van Hardeveld. Exploration of 564 the possibilities for production of Fischer Tropsch liquids and power via biomass 565 gasification. Biomass Bioenergy 23 (2002) 129–152. 566 [30] A. Klerk. Fischer-Tropsch Refining. 2011 Wiley-VCH Verlag & Co. KGaA, Boschstr. 567 12, 69469 Weinheim, Germany. 568 [31] R. K. Sinnot. Coulson and Richardson Chemical Engineering Series. 4th ed. Elsevier. 569 Volume 6, 2005. 570 [32] L. Pellegrini, S. Locatelli, S. Rasella, S. Bonomi, V. Calemma. Modeling of Fischer-571 Tropsch products hydrocracking. Chem. Eng. Science 59 (2004) 4781-4787. 572 [33] O.P.R. van Vliet, A.P.C. Faaij, W.C. Turkenburg. Fischer–Tropsch diesel production in 573 a well-to-wheel perspective: A carbon, energy flow and cost analysis. Energy Convers. 574 Manage. 50 (2009) 855-876. 575 [34] R. Krishna, S.T. Sie. Design and scale-up of the Fischer–Tropsch bubble column slurry 576 reactor. Fuel Process. Technol. 64 (2000) 73-105. 577 [35] D. Iribarren, A. Susmozas, J. Dufour. Life-cycle assessment of Fischer-Tropsch products 578 from biosyngas. Renewable Energy 59 (2013) 229-236. 579