Techno-economic Analysis of Sustainable Biofuels for Marine Transportation

Renewable, low-carbon biofuels offer the potential opportunity to decarbonize marine transportation. This paper presents a comparative techno-economic analysis and process sustainability assessment of four conversion pathways: (1) hydrothermal liquefaction (HTL) of wet wastes such as sewage sludge and manure; (2) fast pyrolysis of woody biomass; (3) landfill gas Fischer–Tropsch synthesis; and (4) lignin–ethanol oil from the lignocellulosic ethanol biorefinery utilizing reductive catalytic fractionation. These alternative marine biofuels have a modeled minimum fuel selling price between $1.68 and $3.98 per heavy fuel oil gallon equivalent in 2016 U.S. dollars based on a mature plant assessment. The selected pathways also exhibit good process sustainability performance in terms of water intensity compared to the petroleum refineries. Further, the O and S contents of the biofuels vary widely. While the non-HTL biofuels exhibit negligible S content, the raw biocrudes via HTL pathways from sludge and manure show relatively high S contents (>0.5 wt %). Partial or full hydrotreatment can effectively lower the biocrude S content. Additionally, co-feeding with other low-sulfur wet wastes such as food waste can provide another option to produce raw biocrude with lower S content to meet the target with further hydrotreatment. This study indicates that biofuels could be a cost-effective fuel option for the marine sector. Marine biofuels derived from various feedstocks and conversion technologies could mitigate marine biofuel adoption risk in terms of feedstock availability and biorefinery economics.


List of Tables
. RCF reactor operating assumptions for lignin-ethanol oil production (Bartling et al., 2021)

S1. Process flow diagram and process description of wet waste HTL
As shown in Figure S1, the process begins with the high moisture content wet wastes collected from animal farms and water treatment plants. Prior to processing, the wet waste is dewatered to 25 wt% solids for minimizing the capital and operating cost of the HTL plant. In the base case, both two wet waste (sludge and manure) cost at the gate of the HTL plant is assumed to be zero.
However, there exist potential significant savings in avoided disposal costs to farms and water treatment plants and the potential feedstock credits are detailed below. In addition, transportation cost for collecting wastes from multiple locations is considered to support the HTL plant scale.
The HTL plant processes 1000 dry metric tons of wet waste a day and is briefly described below.
The HTL reactor operating condition is near the subcritical water status, which has a high solubility for organic compounds. To reach such operating conditions, the feed slurry is first pumped to 20 MPa and then heated in the heat exchanger and trim heater to reach 350 ℃. The HTL reactor has a shell-and-tube structure, having feed slurry on the tube side while hot heating oil on the shell side. After the reaction, the wet waste is converted into biocrude, an aqueous phase, and a small number of solids and gases. In a solid-liquid-gas three-phase separator, solid and gas are separated from the liquid and then the liquid effluents are cooled for further aqueous-biocrude phase separation. The gas combined with natural gas is sent to a burner for generating heat to supply HTL heating requirements via the hot oil system. The aqueous phase needs a series of treatment steps before recycling back to the wastewater treatment (WWT) plant. Specifically, it is first treated with quicklime to raise the pH to ~11 and then stripped with air to remove ammonia and volatile organics (VOCs) from the aqueous stream. The ammonia and VOCs can be completely destroyed in a thermal oxidizer (THROX) with the help of natural gas and catalyst. At the same time, the liquid at the bottom of the stripper is further treated to decrease chemical oxygen demand (COD) before recycling back to WWT plants.
The produced biocrude is cooled to be used as marine fuel blendstock or can be further mildly-or fully-hydrotreated to improve the fuel properties. Specifically, the hydrotreating equipment includes a guard bed for metals and mild heteroatom removal and the main hydrotreating reactor for removing oxygen, nitrogen, and sulfur as much as possible. Different hydrotreating steps could be required for HTL biocrude from different wet wastes due to the variability of feedstock compositions. Figure S1 shows the potential minimum processing requirements for marine fuel S4 blendstocks. Note that the biocrude upgrading is co-located with the HTL plant in this scenario (i.e., no transportation cost for biocrude). Table S1 summarizes the key process technical assumptions associated with HTL of two feedstocks. Figure S1. Wet waste HTL and biocrude upgrading process flow diagram. 1.0 0.5 H2 consumption in guard bed (g/g dry feed) 0.027 a 0.027 a H2 consumption in main bed (g/g dry feed) 0.019 b 0.016 b a an average H2 consumption in the guard bed is based on the data with the H2 consumption range of 0.007-0.035 g/g dry feed for algae HTL at different hydrotreating conditions. b is based on sludge HTL and manure HTL operating data. c gasoline gallon equivalent S5

S2. Process flow diagram and process description of catalytic fast pyrolysis
As discussed in the manuscript, three process options were considered for evaluating the minimum fuel selling prices (MFSP) for fast pyrolysis systems. Configuration FP1 ( Figure S2a) is the simplest, without the inclusion of (i) catalytic upgrading, (ii) hydrogen use, and (iii) chemical coproducts recovery. It is also the case with the highest yield, but at the cost of a poorer product quality (high O content in the liquid fuel/intermediate usually renders it reactive and unstable during long term storage). All the FP process configurations include the production of electricity from pyrolysis off-gases and any excess heat available from char combustion. The FP1 pathway is near commercial for pyrolysis-oil production, and technology offerings are available for different fast pyrolysis reactor configurations.
Configuration FP2 ( Figure S2b) includes an ex situ catalytic reactor following the fast pyrolysis reactor for upgrading the fast pyrolysis vapors from the fast pyrolysis reactor. The upgrading reactor design in FP2 is a circulating fluidized system with a zeolite catalyst. Hydrogen is not added to the system. The catalytic upgrading reduces the O content of the product fuel and makes it more stable for handling and downstream upgrading.
Configuration FP3 ( Figure S2c) is the most complex among the three configurations (Dutta et al., 2021). In addition to a fixed bed ex situ upgrading reactor with Pt/TiO2 catalyst, it includes the use of hydrogen (generated from process off-gases) to boost yields. Recovery of oxygenated coproducts acetone and methyl-ethyl-ketone (MEK) are also included; nearly 4.3% of the biomass carbon is retained in the recovered coproducts, providing a high value revenue complement for the fuel product, and helps drive the net operating cost negative (FP3 in Figure 1); the coproducts also help improve the sustainability metrics and GHG emissions for the process, when accounted using a product displacement method.   Figure S3 depicts the simplified process flow diagram of the landfill gas Fischer-Tropsch synthesis process. The feedstock is landfill gas (LFG) instead of the more commonly used natural gas. LFG differs in composition from natural gas, with approximately 40 percent of the volume as CO2. A compositional breakdown is provided in Table S2. LFG comes off of the header at the landfill at a pressure of 1.6 psig and must be compressed to the steam reforming operating pressure of 30 psi (2.1 bar). After compression, an iron bed removes the H2S in the feed stream, followed by an activated carbon bed to remove any remaining siloxanes.   Steam reforming, i.e., CH4 + H2O  CO + 3H2, converts methane gas to carbon monoxide and hydrogen. Additionally, unconverted syngas from the FT is partly recycled back to the reformer and partly combusted to provide the heat necessary for the endothermic reforming reactions. The syngas stream, consisting mostly of CO, H2, H2O, and CO2, is cooled and then compressed to 425 psi (29.3 bar) before entering the acid gas removal system, which removes the bulk of the H2S and CO2 from the process gas.

S3. Process flow diagram and process description of landfill gas Fischer-Tropsch synthesis
Fischer-Tropsch (FT) synthesis is a catalytic conversion process, which converts the synthesis gas to a mixture of reaction products such as diesel, gasoline, jet fuel, and wax products. The overall reaction involved in the FT synthesis is represented with (2n + 1)H2 + nCO  CnH2n+2 + nH2O.
The advantages of the FT polymerization process are that it offers the ability to produce liquid hydrocarbon fuels with relatively low sulfur and aromatic content. For this process, the H2/CO ratio entering the FT reactor was maintained at 1.5:1 by sending a portion of the synthesis gas stream to a pressure swing adsorption (PSA) system with the offgas feeding the FT reactor. The impact of the lower H2/CO ratio results in a less than stoichiometric ratio being fed to the reactor, and thus only 78 percent of the CO is consumed in the FT reactor, compared to a typical natural gas feedstock which has an 85 percent conversion of CO.
The FT products are condensed and separated through a multi-cut distillation column to separate the product streams. The purified H2 from the PSA system is used for hydrotreating distillation products to yield blendstocks for gasoline, diesel, and jet-fuel or used for hydrocracking wax. Wax S9 produced from the hydrocracker is sold as a co-product. Any excess H2 not consumed for hydrotreating or hydrocracking is sold as a co-product. Figure S4. illustrates the block flow diagram of a reactive reductive catalytic fractionation (RCF)

S4. Process flow diagram and process description of lignin-ethanol oil
based cellulosic ethanol biorefinery that co-produces lignin ethanol oil (LEO) according to Bartling et al. (2021). This process's configuration is similar to that described by Humbird et al., (2011), with the dilute-acid pretreatment area replaced with RCF. The RCF area produces a ligninrich oil, here also containing a small fraction of residual ethanol (i.e., LEO) as a co-product and a carbohydrate-rich pulp. The latter is the residual biomass solids after the lignin is removed during RCF and is isolated and saccharified to C5 and C6 sugars via enzymatic hydrolysis, fermented to ethanol, and recovered to produce fuel-grade ethanol. A combination of natural gas, sludge from wastewater treatment, and residual solids from ethanol production are burned to generate steam for process heat and electricity via a combined heat and power (CHP). Excess electricity is sold to the grid as a co-product. The RCF reactor operating assumptions for lignin-ethanol oil production are summarized in Table S3.

S6. Capital cost estimates assumptions
HTL Pathway capital costs for each plant area are based on data from wet waste HTL design reports and publications from the Pacific Northwest National Laboratory Snowden-Swan et al., 2021). The assumptions of total direct cost, fixed capital investment, and total capital investment estimation are listed in Table S5 for all the pathways. S12  shows that 64% of sludge in the US has a negative price and the average sludge price is $-36/wet ton (not including the dewatering cost) (Badgett et al., 2019), as shown in Table S7.  Manure management cost includes on-farm nutrient management costs, off-farm transport costs, land treatment costs, manure and wastewater handling storage costs, and recordkeeping cost.

S7. Variable Operating Costs
National Renewable Energy Laboratory's wet waste resources analysis (Badgett et al., 2019) shows that 27% of animal manure in the US has a negative price and the average sludge price is $-48/wet ton ($-125/dry ton), as shown in Table S8. A wide range of $5 -150/dry ton is reported as the swine manure avoided disposal cost based on the U.S. Department of Agriculture (USDA) method for estimating comprehensive nutrient management plans (CNMP) cost (Ou et al., 2022). S15