Hydrotreatment Upgrading of Hydrothermal Liquefaction Biocrude toward Advanced Fuels: A State-of-the-Art Review

Abstract

The nonrenewable resource depletion and associated environmental concerns have intensified research into advanced biofuel technologies. Biomass-derived biofuels present a promising solution for meeting energy needs while supporting a circular economy and reducing carbon emissions. Among the various conversion methods, hydrothermal liquefaction (HTL) has attracted significant attention due to its ability to process both wet and dry biomass feedstocks into a liquid biocrude product. Although this biocrude exhibits improved characteristics compared with raw biomass, it still requires further upgrading prior to its use as a fuel or refinery intermediate. Catalytic hydrotreating is a crucial process for upgrading HTL biocrude, improving its stability, reducing the oxygen content, and enhancing overall fuel quality. Recent research has focused on optimizing hydrotreating processes using both batch and continuous-reactor systems. This Review provides a comprehensive overview of current advancements in HTL biocrude upgrading through catalytic hydrotreatment. It examines various biomass feedstocks to understand how their compositions affect hydrotreatment efficiency and final product outcomes. Additionally, it discusses commonly used catalysts and evaluates the influence of key operating conditions, such as temperature, pressure, and reaction time, on the product quality. By compiling and analyzing existing experimental data, this Review aims to fill the gap in dedicated literature on HTL biocrude upgrading. Ultimately, it offers practical guidelines and a technological assessment to support future research and development in sustainable biofuel production.

1. Introduction

In the last 150 years, the worldwide population has grown exponentially, leading to an increase in global energy and product consumption. As a result, the industry, agriculture, and technology sectors are expanding their production to address consumer demands. This development has led to an increase in the amount of industrial, municipal, and agricultural wastes. Nowadays, our planet faces massive population growth, alongside a worrying level of environmental pollution. For this reason, researchers are seeking ways to reuse wastes to produce fuels, energy, and high-value-added products, termed biofuels, advanced fuels, renewable energy, and bioproducts, with the aim of reducing global contamination while satisfying population demands. In this context, biofuels derived from biomass offer a promising pathway to meet energy demands while contributing to a circular economy and reducing the carbon footprint. However, the efficient development and scalability of biofuel production processes remain major challenges in both industrial and academic sectors. −

Among the various technologies explored for biofuel production, thermochemical conversion methods such as pyrolysis and hydrothermal liquefaction (HTL) have gained considerable attention due to their ability to process diverse biomass feedstocks, including lignocellulosic materials, algae, plastics, and organic and inorganic wastes, among others. Pyrolysis, which involves the thermal decomposition of biomass in the absence of oxygen, and HTL, which occurs in high-pressure subcritical water, are both capable of producing bio-oil and biocrude (biological crude oil), respectively, which can be upgraded into transportation fuels. Nevertheless, both the bio-oil and HTL biocrude obtained from these processes are often heteroatom-rich and chemically unstable, requiring further upgrading. ,

Catalytic hydrotreatment (HDT) has emerged as a critical process for upgrading HTL biocrude to high-quality biofuels or intermediate refinery streams. The main targets of this catalytic process involve the removal of heteroatoms, such as sulfur, nitrogen, and oxygen, significantly improving the stability and energy content of the final liquid product. To date, catalysts such as NiMo/Al₂O₃ and CoMo/Al₂O₃ have demonstrated efficacy in the hydrotreatment of HTL biocrude derived from diverse biomass sources. These catalysts facilitate the breakdown of complex oxygenated, sulfur-containing, and nitrogen-containing compounds, thereby improving the overall fuel properties. However, challenges such as catalyst deactivation due to coking, high hydrogen consumption, and incomplete removal of nitrogen-containing compounds persist, limiting the commercial viability of these processes. As HTL is a relatively emerging technology, studies investigating biocrude upgrading via hydrotreatment remain limited. Thus, several challenges still need to be addressed in order to optimize this technology. ,

To this end, the present Review aims to collect, evaluate, and analyze studies focused on HTL biocrude upgrading via hydrotreatment. The evaluation of these studies is challenging due to the wide variation in HTL biocrudes produced from different biomass or waste feedstocks, the use of various hydrotreating catalysts, differing operating conditions, and variations in pilot-plant setups. Therefore, different potential feedstocks for HTL biocrude production are described in order to analyze the effect of the initial biocrude feedstock on the hydrotreatment efficiency. In addition, the literature is categorized into batch operation units and continuous operation units. Furthermore, two-stage hydrotreatment of biocrude, as well as the coprocessing of biocrude with petroleum-derived feedstocks, is also explored and presented. As a result, this Review aims to provide clear guidelines on the hydrotreatment of different HTL biocrude feedstocks and to deliver a technological assessment of existing research efforts. It should be noted that throughout this review, the term upgraded biocrude yield refers to the liquid organic product yield reported by the original authors. Since the literature employs different definitions (mass yield, carbon yield, energy yield, or organic phase recovery), all values are presented by using the terminology reported in the corresponding studies. When necessary, this distinction is explicitly stated to avoid direct comparison of non-equivalent yield metrics.

2. HTL Biocrude

HTL is a thermochemical process that converts different types of biomass wastes into biocrude oil (a renewable fuel precursor) under moderate temperatures (250–400 °C) and high pressures (5–25 MPa), using water as the reaction medium. In comparison to other relevant processes, such as pyrolysis, HTL is more suitable for treating wet biomass, since water acts as an effective medium for the cracking reactions of biomass structural compounds. Therefore, water not only acts as a solvent but also as a catalyst, favoring biomass degradation into valuable liquid molecules. The underlying HTL reaction mechanisms, hydrolysis and depolymerization of biomass macromolecules followed by repolymerization/recondensation of the resulting fragments into biocrude, together with their kinetic modeling, have been comprehensively reviewed elsewhere. , Moreover, studies have shown that the organic product generated from HTL (synthetic biocrude) is improved in terms of both quantity and quality compared to pyrolysis oil. The ability of HTL to process wet biomass makes it a prominent technology for converting various types of waste into biocrude. Typical biomass feedstocks for the HTL process include lignocellulosic materials, agricultural residues, animal manure, microalgae, macroalgae, food waste, municipal solid waste (MSW), sewage sludge, and others. Each type of biomass presents unique challenges and opportunities in terms of conversion efficiency, biocrude yield, and economic feasibility.

During HTL, reaction conditions critically influence product distribution. Higher temperatures (>350 °C) favor gas formation, while lower temperatures (<350 °C) maximize biocrude yields (wt %), although yields vary significantly depending on feedstock composition. For example, lipid-rich algae can yield up to 60 wt % biocrude, whereas lignocellulosic biomass typically produces less than 25 wt %. The HTL process generates four distinct product phases: (i) biocrude, a viscous oil with a heating value of 25–40 MJ/kg, requiring upgrading due to its sulfur, nitrogen, and oxygen content; (ii) an aqueous phase, rich in organics and nutrients, requiring treatment or recycling; (iii) hydrochar, a carbon-rich solid with applications in soil amendment or catalysis; and (iv) gas, consisting primarily of CO₂ and CH₄.

The scalability of HTL faces several challenges, including reactor corrosion, feedstock heterogeneity, and the energy-intensive separation of stable Pickering emulsions. Nevertheless, continuous-flow systems, such as Steeper Energy’s Hydrofaction technology and the Silva Green Fuel demonstration plant located in Tofte, Norway, with a production capacity of 4000 L/day, are advancing industrial deployment. Pilot plants have achieved biocrude yields of 42–55 wt % and energy recoveries exceeding 60%.

There are different ways to classify feedstocks for HTL biocrude production based on their origin, availability, sustainability, and potential for large-scale energy conversion. In this review, HTL feedstocks are classified based on their origin in order to compare properties such as density, viscosity, acidity, heteroatom content, water content, and compound distribution, among others. Table presents a summarized classification of biomass feedstocks used in HTL, detailing their biochemical composition and the properties of the resulting biocrude. This classification highlights the versatility of HTL in processing diverse organic materials for biofuel production while emphasizing the relationship between feedstock composition and biocrude quality. It also underscores the critical role of feedstock composition in determining the biocrude quality and the need for tailored preprocessing and upgrading strategies. Lipid-rich feedstocks, such as microalgae and waste oils, generally yield more favorable biocrude for fuel production, while carbohydrate- and protein-rich materials require advanced treatment to mitigate their high heteroatom content. This data synthesis provides a basis for classifying the characteristics of oils obtained through the HTL process from various biomass sources, serving as a guide for their upgrading to biofuels or for coprocessing in existing refineries.

1. HTL Feedstocks, Biomass Type, Composition, and Biocrude Features.

Biomass type	Feedstock main composition	Biocrude features	refs
Forest/agricultural residues
Forestry residues: Rice straw. Corn stover. Wood (pine, poplar, spruce, beech), walnut shells	Cellulose, hemicellulose, and lignin. Low N/S. Low moisture contents.	Moderate viscosity. Requires alkaline pretreatment. Yields oxygenated compounds. Phenolic compounds. Furans and cyclic ketones. Carboxylic acids. Aromatic hydrocarbons. Alcohols and ethers. Low N. High O (phenolics).	−
Wheat: stem, straw.	Cellulose. Hemicellulose. Lignin. Moisture. Ash	High oxygen content, low nitrogen content, viscous, phenolic, and ketone-rich. Ketones/aldehydes. Cyclic hydrocarbons. Heavy fractions.	,

Abattoir waste
Cow Manure	Lignocellulose, proteins, lipids.	Lower carbonyl content, higher lipid-derived hydrocarbons, and less viscous than straw wheat
Streaky pork	Lipids, proteins, high water, low carbohydrate, and ash.	Fatty acids and derivatives, aliphatic amides, hydrocarbons, esters, nitrogenous heterocyclic compounds, and phenolic and oxygenated aromatic compounds

Algae cultivation residues
Scenedesmus almeriensis. Nannochloropsis gaditana	High lipid content and protein High lipid content and protein	Fatty acid (n-hexadecanoic acid), polyunsaturated fatty acids (oleic acid), polysubstituted alkanes, straight alkanes (C≥20), aromatic compounds (phenol), amide, indole derivatives, pyrrole derivatives, and pyrrolidine derivatives
Chlorella vulgaris; Tetradesmus obliquus, Limnothrix planctonica	Lipid, carbohydrate, and protein	Fatty acid (palmitic, linoleic), hydrocarbon, nitrogen, and oxygen compounds
Spirulina	Protein. Lipids. Carbohydrate. Ash. Moisture.	Viscous. High N/O content Aromatic (indoles, pyridines). Low stability
Microalgae	Lipids. Proteins. Carbohydrates	High nitrogen content complicates combustion/hydrotreating. Higher freezing point limits blending higher than Jet A1. High lipid yield.
Macroalgae	Polysaccharides. Proteins	High moisture tolerance; yields biocrude with potential for nutrient recycling.
Chlorella sorokiniana (CS)	High carbohydrates. Lipids and proteins.	Lower yield, higher O/C and N/C ratios, more cyclic N/O compounds, and higher aromaticity.	,
C. vulgaris (CV)	High lipids, moderate carbohydrates, and proteins.	Higher yield, lower O/C and N/C ratios, more fatty acids/amides, and higher aliphatics.

Municipal Solid Waste (MSW)
Food waste	Proteins, lipids, carbohydrates, lignin, fiber, and ash.	Fatty acid, olefin, saturated hydrocarbon, nitrogen and oxygen compounds, aromatics compound, alcohols, ketones, and carboxylic acids	−
Restaurant and household waste	Little proteins, high lipid, and moderate carbohydrates; low ash.	Long-chain acid. Monounsaturated fatty acids. Polyunsaturated fatty acids. Saturated fatty acids.	,
Garden waste, food waste, mixed plastics (PP, PS, PET).	Cellulose, hemicellulose (biodegradable fraction), and synthetic polymers (C/H chains).	Hydrocarbons (alkanes, aromatics), minimal oxygen.
Sewage sludge	Organic matter, fats. Lipids, proteins, carbohydrates. High ash/metals Cellulose remnants.	Lower energy demand (no drying). Fatty acids, alkanes, N-compounds. High heteroatom content (S, N). Metals (Fe, Ca). High TAN (acidity)	,,

Although the HTL process can handle a wide variety of raw materials, the produced oils exhibit different characteristics and all require post-processing for upgrading. The properties of HTL biocrudes (Table ) reveal critical challenges for direct fuel application, including high heteroatom content in most feedstocks, with particularly high levels observed in sulfur-rich sludges, chemical instability, and suboptimal heating values compared to petroleum diesel. These limitations necessitate catalytic upgrading to meet the fuel standards. HDT emerges as the most effective pathway to address these issues through heteroatom removal (hydrodeoxygenation reactions (HDO), hydrodenitrogenation reactions (HDN), and hydrodesulfurization reactions (HDS)) and hydrocarbon saturation, with sulfur removal being particularly crucial for meeting ultralow sulfur diesel (ULSD) specifications (<10 ppm). The following section evaluates HDT processes, focusing on catalyst performance, operating conditions, and their impact on the biocrude upgradability across different feedstock categories. Special emphasis is placed on reconciling batch-scale findings with those of continuous systems in order to bridge the gap between laboratory research and industrial implementation.

3. Biocrude Hydrotreatment

Hydrotreatment is an established technology originating in the mid-20th century within the petroleum industry. It has been adapted to renewable feedstocks in recent decades as part of the transition to sustainable energy sources. Its long-standing use in fossil fuel processing provides a strong foundation for its application in biofuel production. The process involves hydrogenation reactions in which heteroatoms (oxygen, sulfur, and nitrogen) are removed from hydrocarbons, typically in the form of water, hydrogen sulfide (H₂S), and ammonia (NH₃). The process employs catalysts such as NiMo/Al₂O₃, CoMo/Al₂O₃, or noble metal catalysts (Pt, Pd) under high pressures (1–10 MPa) and temperatures (300–400 °C) to facilitate these reactions. The H₂/oil ratio is critical for oxygen removal and carbon hydrogenation, while the Liquid Hourly Space Velocity (LHSV) affects contact time and reaction efficiency.

In the biocrude context, HDT stabilizes the product by removing oxygen, nitrogen, sulfur, and other impurities, converting it into “drop-in” fuels such as gasoline, diesel, and jet fuel. Upgraded biocrude can also be coprocessed with petroleum streams or used directly as a drop-in fuel, enabling seamless integration into existing infrastructure. , Hydrotreatment for upgrading different types of biocrude has been studied in both batch and continuous reactor systems. The former is widely used in laboratory-scale studies due to its simplicity and controllability, allowing detailed investigation of reaction kinetics and catalyst performance. The latter is preferred for industrial applications due to its scalability, efficiency, and ability to ensure steady-state operation and high throughput, which are critical for commercial viability.

This section summarizes available studies on the upgrading of different types of biocrudes via hydrotreatment. The main objective of this review is to collect, evaluate, and analyze all relevant data available in the literature to date. The evaluation of these studies is challenging due to the wide variation in HTL biocrudes produced from different biomass or waste feedstocks, the use of various hydrotreating catalysts, differing operating conditions, and variations in pilot-plant setups. Therefore, an effort has been made to summarize the stabilization of biocrudes derived from different feedstocks and to analyze the effect of hydrotreating operating parameters on the product quality. In addition, a comprehensive list of tested hydrotreating catalysts is presented along with their effectiveness for different biocrude feedstocks. Finally, a comparison between batch and continuous operation systems is provided.

3.1. Bench-Scale Research of HTL Biocrude Catalytic Hydrotreatment

The current section presents a state-of-the-art literature review on HTL biocrude upgrading in bench-scale operation systems. As there are many different feedstocks for biocrude production, an attempt was made to categorize research based on the initial biocrude origin. As a result, all recent studies that were performed in batch hydrotreating units are presented in categories based on the initial feedstock via the following order: lignocellulosic biomass, livestock feedstocks, algae, food waste, municipal solid waste (MSW), and sewage sludge. A summary of all studies performed in batch units is presented in Table .

2. Summary of Biocrude Bench-Scale Catalytic Hydrotreating Studies.

Biocrude feed	Cat.	Oper. Cond.	Biocrude/Cat	Organic product yields	HDT biocrude composition	Gas products	Refs
Lignocellulosic	Ru/BEA, Ru/ZSM-5, Ru/SiO2, and Ru/Al2O3	330 °C, 4 MPa, 200 min retention time	0.2 g cat and 0.4 g biocrude	67–92 wt %	-	Methane, CO2, ethane, propane, and iso-butane
Miscanthus	NiMo/Al2O3	350 °C and 400 °C, 4 and 8 MPa, and reaction time 4 h	4 g biocrude and 2 g cat.	81–92 wt %	n-C15 n-C16 n-C17 n-C18	-
Sweet sorghum bagasse	Ru/C	350 °C, 3.4 MPa, and holding time 4 h		96 wt %	Phenolics, hydrocarbons, and alkanes
Animal carcass	NiMo/Al2O3, CoMo/Al2O3, Pt/Al2O3, Pd/Al2O3, Mo2C, and Mo2S	250–400 °C, 6 MPa, and 1–4 h reaction time	0.8 g biocrude and 0.16 g catalyst	58.6–68 wt %	Alkanes C16 (18.8 wt %) and C18 (37.1 wt %)	-
Microalgae	CoMo/γ-Al2O3	375 °C, 10 MPa, and 1 h reaction time	3/1 (g)	70.5% wt	C17: 5.1% C16: 3.6% C15: 4.7% Toluene:2.4%, ethylbenzene:1.8%, hexadecanenitrile:5.1%, and octadecanenitrile:3.5%	30.9% CH4, 30.8% CO2, 15.9% C3H8, and 13.6% C2H6
Algae: C. sorokiniana and vulgaris	NiWS/Al2O3	375 °C, 10 MPa, and 5 h reaction time	3 g biocrude, 0.6 g cat., and 25 g n-heptane	56–64 wt %	Alkanes C15 to C18	-
Spirulina	NiMo/Al2O3	350 and 400 °C, 4 and 8 MPa, and reaction time 4 h	4 g biocrude and 2 g cat.	81–92 wt %	n-C15 n-C16 n-C17 n-C18	-
Algae: C. vulgaris and N. gaditana	NiMo/Al2O3 and NiW/Al2O3	250 °C at 1.07 MPa, and 400 °C at 1.38 MPa, and 4 h	0.2 g catalyst per 1 g biocrude	54 (at 400 °C)-90 (at 250 °C) wt %	C14-C50	CO, CO2, CH4, C2H6, C3H8, i-C4H10, and n-C4H10
Algae: Nannochloropsis	NiMo/Al2O3	400 °C, 5 MPa, and 1 h reaction time	10 wt % catalyst	96.6 wt %	-	-
	Ru/Al2O3			63.5 wt %	-	-
	Pt/Al2O3			∼64 wt %	-	-
	Pd/Al2O3			∼65 wt %	-	-
	Pt/C			∼65 wt %	-	-
	Ru/C			∼70 wt %	-	-
	Pd/C			∼80 wt %	-	-
Food waste	NiMo/γ-Al2O3	340 and 400 °C, 2–3.5 MPa, and reaction time 5 h	1 g cat., 55 g of hexadecane, and 10 g of pre-treated biowaste	97 wt % at 400 °C	Alkenes, alcohol-ether-ketone, cyclic alkanes, aromatic naphthalene, alkylated phenolics, and N-components	mostly CH4 and CO2
Sewage sludge	NiMo/Al2O3	350 and 400 °C, 4 and 8 MPa, and reaction time 4 h	4 g biocrude and 2 g cat.	81–92 wt %	n-C15 n-C16 n-C17 n-C18
MSW	NiMo/Al2O3 CoMo/Al2O3	200–370 °C, 15–22 MPa, and 10–60 min residence time		14–45 wt %	Fatty acids, esters, phenols, aliphatic and aromatic hydrocarbons, and nitrogenated and oxygenated compounds.	H2O, NH3, H2S, CH4, H2, CO, and CO2,

3.1.1. HDT of Lignocellulosic Biomass-Derived Biocrude

Lignocellulosic biomass includes agricultural residues, forestry wastes, and dedicated energy crops. It is rich in cellulose (36–43 wt %), hemicellulose (10–20 wt %), and lignin (18–33 wt %), and its high availability and low cost make it an attractive feedstock. However, due to its high ash content and complex structure, a pretreatment step is often necessary to improve HTL performance. Agricultural wastes include crop residues such as corn stover, rice husks, sugarcane bagasse, wood chips, bark, sawdust, wheat straw, pine wood, barley straw, cocoa pods, spent coffee grounds, and coffee husks, among others. Among these feedstocks, corn stover and wheat straw have been extensively studied due to their high availability and favorable energy content. ,

Lignocellulosic biomass exhibits high moisture content (60–80 wt % wet basis) and low energy density (10–20 MJ/kg) in its raw form. This biomass is characterized by its fibrous structure, which contributes to its low bulk density and associated processing challenges. During the HTL reaction, lignocellulosic biomass undergoes decomposition in subcritical water, yielding biocrude with high oxygen content (17–23 wt %) and a significant quantity of heteroatoms, mainly nitrogen. The resulting biocrude is acidic, chemically unstable, and has a viscosity higher than that of conventional petroleum fuels. Its higher heating value (HHV) ranges from 29 to 33 MJ/kg, which is lower than that of petroleum diesel (44 MJ/kg). The presence of oxygenated compounds, such as cyclic ketones and phenolics, further complicates its direct use as a fuel without upgrading. Catalytic HDO is often required to reduce oxygen content and improve fuel properties, yielding upgraded oil with HHVs of up to 42 MJ/kg and hydrocarbon content as high as 96 wt %. ,

The upgrading of HTL biocrude produced from lignocellulosic biomass, specifically radiata pine softwood, was investigated by Yan et al. Their experiments were performed in a small-scale 35 mL batch reactor, using a catalyst-to-biocrude ratio of 0.2 g of catalyst per 0.4 g of biocrude. The aim was to investigate the efficiency of Ru catalysts supported on different materials (Ru/BEA, Ru/ZSM-5, Ru/SiO₂, Ru/Al₂O₃). The results showed that Ru/Al₂O₃ and Ru/SiO₂ exhibited high hydrogenation activity during biocrude hydrotreatment, but virtually no deoxygenation. Another study by Yue et al. focused on the HDO upgrading of biocrude produced from lignocellulosic biomass, specifically sweet sorghum bagasse. Upgrading experiments were conducted in a 100 mL batch micro-stirred reactor (Model Parr 4598) using a Ru/C catalyst. HDO was performed at 350 °C and 3.4 MPa, with a holding time of 4 h. According to their results, the oxygen content of the biocrude was significantly reduced after hydrotreatment. In addition, the upgraded biocrude exhibited a heating value similar to that of petroleum fuels (∼44 MJ/kg). Overall, their findings demonstrated that lignocellulosic biocrude can be upgraded to drop-in fuels via hydrotreatment. As demonstrated by these studies, biocrude produced from lignocellulosic biomass can be effectively upgraded or stabilized via hydrotreatment to produce an oxygen-free liquid product with a high heating value comparable to that of petroleum-based fuels. The upgraded biocrude can be further utilized as a potential blending component for biofuel production. However, further research is required to optimize the process, as available literature data on lignocellulosic biocrude upgrading remain limited.

3.1.2. HDT of Livestock Feedstock-Derived Biocrude

Annually, approximately 194 million animal carcasses are disposed of, rendering their valorization particularly important. As several studies have focused on evaluating the potential of HTL for various animal carcasses, the hydrotreating upgrading of the resulting biocrudes is also worthy of further investigation. Yang et al. examined the upgrading of biocrude produced via HTL of animal carcasses, focusing on the effect of different commercial hydrotreating catalysts, including NiMo/Al₂O₃ and CoMo/Al₂O₃, noble metal catalysts (Pt/Al₂O₃ and Pd/Al₂O₃), and unsupported Mo-based catalysts (Mo₂C and Mo₂S₂). Hydrotreating experiments were conducted in an autoclave reactor with an internal volume of 8 mL, using 0.8 g of biocrude and 0.16 g of catalyst in each experiment. The results showed a reduction in upgraded biocrude viscosity compared to the initial biocrude, while upgraded biocrude yields ranged between 58.6 and 68 wt %. Among the tested catalysts, CoMo/Al₂O₃ presented the highest yield, reaching 68 wt %, while Pt/Al₂O₃ and Pd/Al₂O₃ yielded approximately 60 wt %. Mo₂C and Mo₂S₂ achieved yields of approximately 67.6 wt % and 60.2 wt %, respectively. Overall, all tested catalysts demonstrated good performance in heteroatom removal reactions, with CoMo/Al₂O₃ showing the best performance. Furthermore, an increase in heating value was observed after hydrotreatment, reaching approximately 44.9–46.7 MJ/kg. A wide operational window was investigated, consisting of operating temperatures ranging from 250 to 400 °C, reaction times from 1 to 4 h, and catalyst loadings from 5 to 20 wt %. According to the study, the highest heteroatom removal was observed at the highest temperature (400 °C), indicating that heteroatom removal reactions are favored at higher temperatures. In addition, increasing the catalyst loading in the reactor led to higher yields and improved heteroatom removal. Regarding reaction time, the authors concluded that longer reaction times favor heteroatom removal but also lead to lower upgraded biocrude yields (organic liquid product) due to increased cracking reactions. Overall, the CoMo/Al₂O₃ catalyst demonstrated superior performance in terms of heteroatom removal and energy recovery compared to the other tested catalysts. Although large quantities of animal carcasses are disposed of annually, and this feedstock appears to be promising for biocrude production via hydrothermal liquefaction, research on the stabilization and upgrading of this type of biocrude remains limited. Therefore, further studies are clearly necessary to advance understanding and optimization in this field.

3.1.3. HDT of Algae-Derived Biocrude

Algal biomass is a promising feedstock for biofuel production due to its high photosynthetic efficiency and ability to grow in diverse environments, including freshwater and marine systems. This biomass type typically consists of proteins, lipids, and carbohydrates, which varies significantly by species and cultivation conditions. HTL of microalgae produces biocrude with carbon content above 70 wt % and high heating value (HHV) of 35–38 MJ/kg. However, biocrude exhibits high viscosity, acidity (measured as Total Acid Number TAN), and heteroatom content, including nitrogen (4–7 wt %), oxygen (5–15 wt %), and sulfur (0.1–1.5 wt %). These heteroatoms contribute to harmful emissions as NO _x and SO _x upon combustion and complicate storage and transport due to corrosion promotion and unstable behavior. Microalgae-derived biocrude also contains water, which stabilizes it during HTL but requires removal or further processing; also, polymerization issues during storage due to unsaturated compounds may exist. Hydrodeoxygenation aims to reduce heteroatoms, saturate molecules, and improve fluidity, though challenges like catalyst poisoning and mass-transfer limitations persist. Biocrude density and viscosity are higher than conventional fuels, necessitating refinement to address transportation fuel standards. Microalgae, such as Nannochloropsis and Spirulina, have been popular due to their high lipid and protein content. In addition, macroalgae, as seaweeds, are attractive for coastal regions but often contain high ash content, which may reduce biocrude yield. Algae offer the advantage of not competing with food crops for arable land, , but it is important to note that algal biomass is a dedicated energy crop rather than a waste product.

All in all, microalgae are considered as a very promising feedstock for 3rd generation production of advanced biofuels. Due to their high moisture content and significant lipid content that can be up to 70 wt % (depending on the species), HTL is considered as an effective technique to convert microalgae into a high-added-value liquid product (biocrude). However, this biocrude requires further upgrading to meet fuel standards. Today, there are some studies in the literature that have investigated the possibility to upgrade biocrude from microalgae to high-quality advanced fuels via hydrogenation. The most recent study in this field was performed by Silva et al., who investigated the upgrading via hydrogenation of an HTL biocrude produced from microalgae. Hydrogenation was performed in a small autoclave 0.16 L reactor from Parr Instruments. The ratio between biocrude and catalyst was 3 g of biocrude per 1 g of CoMo catalyst supported on γ alumina. Operating conditions were at 375 °C and 10 MPa pressure with a reaction time of 1 h. Hydrogenation was able to reduce biocrude’s initial viscosity and to remove heteroatoms. However, oxygen content still remained at high levels (initial biocrude oxygen was 8.3 wt % and after the upgrading step reduced to 8.0 wt %), which is the main reason for the small heating value increase (from 38.3 MJ/kg initial biocrude to 38.9 MJ/kg upgraded biocrude). The upgraded organic liquid product was approximately 70.5% wt of the initial biocrude mass. Furthermore, an analysis of the gas product has shown that it consists of 30.9 wt % CH₄, 30.8 wt % CO₂, 15.9 wt % C₃H₈, and 13.6 wt % C₂H₆. Biocrude upgraded organic liquid product mainly consists of heptadecane C17 (5.1 wt %), pentadecane C15 (4.7 wt %), hexadecane C16 (3.6 wt %), toluene (2.4 wt %), ethylbenzene (1.8 wt %), hexadecanenitrile (5.1 wt %), and octadecanenitrile (3.5 wt %). As far as catalyst activity is concerned, authors reported that catalyst deactivation occurred after prolonged use due to coke formation and metals such as S and N-compounds of biocrude. They have concluded that further studies are required to improve biocrude properties.

Another investigation was performed by Magalhaes et al., who upgraded two types of biocrude, one derived from C. sorokiniana and the other from C. vulgaris. Hydrogenation experiments were performed in a 300 mL Parr stainless steel batch reactor, employing the same NiW/Al₂O₃ extrudate catalyst for both biocrudes. The efficiencies of nitrogen and oxygen removal were 55% and 90% for C. vulgaris and 75% and 94%, respectively, for C. sorokiniana. The hydrotreated C. vulgaris -derived biocrude presented lower heteroatom content and higher heating value compared to the C. sorokiniana -derived biocrude, showing precedence of C. vulgaris species for advanced fuels production. A probable reason is that the C. vulgaris -derived biocrude contains a higher number of aliphatic compounds that can be associated with carboxylic acids and fatty amides, while the C. sorokiniana-derived biocrude presented a higher aromaticity degree. In addition, the authors have noticed that hydrotreating was able to reduce boiling range hydrocarbons of the final product below 500 °C, increasing the diesel content. The jet fuel range hydrocarbons yield was 31 wt % for the hydrotreating of C. sorokiniana biocrude and 24 wt % for C. vulgaris biocrude. The authors have observed coke formation on the catalyst, similar to a previous study. The main metals found in the used catalysts were iron and potassium. A very important finding here is that more iron accumulation was noticed in the catalyst used to upgrade biocrude from C. vulgaris, in which the HDN degree was higher. This means that the algal feedstock nature could also affect catalyst efficiency in terms of hydrodemetallization reactions.

Another interesting work was published by Castello et al., who tested hydrotreatment of three different biocrudes, from miscanthus, from microalga Spirulina, and sewage sludge. In addition, they have also studied the temperature and pressure effects on product quality during hydrotreating of different biocrudes. They found that biocrude from different origins presents different upgrading challenges. Upgrading was performed on a small autoclave of 25 ml, and the catalyst was a NiMo/Al₂O₃ at a ratio of 4 g of biocrude per 2 g of catalyst. According to their results, product yields were higher at 8 MPa pressure compared with 4 MPa pressure, showing that higher operating pressures are preferable in terms of liquid product yields. In addition, higher temperatures favor cracking reactions, leading to lower product yields.

HTL biocrude upgrading produced from algae C. vulgaris and N. gaditana utilizing NiMo/Al₂O₃ and NiW/Al₂O₃ hydrotreating catalysts was studied by Guoa. Biocrude upgrading was performed in a 10 mL micro autoclave reactor at two temperatures (250 °C and 400 °C) for a holding time of 4 h. It was found that hydrotreating of biocrude can lead to yields up to 90 wt % at 250 °C but only to 54 wt % at 400 °C, showing that higher hydrotreating temperatures favor cracking reactions, reducing in that way biocrude yields. Hydrotreatment was able to remove heteroatoms, leading to an oxygen content of 1 wt %, a nitrogen content of 52 wt %, and a sulfur content of 25 wt %, resulting in an increase of the heating value from 33.7 MJ/kg to 42 MJ/kg. It was observed that the NiMo catalyst had better performance on HDO and HDN reactions, while the NiW catalyst was more efficient in HDS reactions. In general, it was observed that the use of conventional deoxygenation catalysts can be recommended.

Patel et al. investigated hydrotreatment of a biocrude produced from algae, utilizing an autoclave reactor and commercially available catalysts (NiMo/Al₂O₃, Ru/Al₂O₃, Pt/Al₂O₃, Pd/Al₂O₃, Pt/C, Ru/C, Pd/C). They have noticed that hydrotreatment was able to remove oxygen content from raw biocrude, achieving similar values with fossil crude oil (1.6–6.0 wt %). Oxygen reduction content has led to an increase in upgraded biocrude heating value in the range of 38.36 to 45.40 MJ/kg, depending on the catalyst type. The authors have also concluded that higher yields were obtained from the NiMo/Al₂O₃ catalyst, followed by Pd/C.

In summary, even though microalgae render a good quality HTL biocrude, a subsequent hydrotreatment step is still necessary to reduce oxygen content and to remove nitrogen and sulfur heteroatoms as well as to remove any impurities, rendering a higher heating value product. Among the hydrotreating catalysts that have been investigated, NiMo/Al₂O₃ was the most common and promising one.

3.1.4. HDT of Food Waste-Derived Biocrude

In 2022, the economic cost of food waste within the European Union (EU) was estimated at approximately 130 billion euros, with around two-thirds attributable to households. Transforming food waste into potential liquid biofuels is therefore an important avenue of research and is currently being explored in many studies. Unlike other thermochemical processes such as pyrolysis and gasification, hydrothermal liquefaction (HTL) does not require predrying of biomass, as water serves as the reaction medium. Consequently, feedstocks with high moisture content, including food waste and sewage sludge, are particularly well-suited for HTL, offering the additional benefit of reducing energy consumption associated with predrying.

Food waste-derived HTL biocrude was effectively hydrotreated using a synthesized NiMo/γ-Al₂O₃ catalyst by Prabin Dhakal et al. Catalytic tests were performed in a batch reactor under severe conditions, with temperatures ranging from 340 to 400 °C and a hydrogen pressure around 2–3.5 MPa for a reaction time of 5 h. 1 g of NiMo/γ-Al₂O₃ catalyst, 55 g of hexadecane, and 10 g of pre-treated biowaste (namely liquid biocrude and residual solid) were placed in the reactor and sealed. The hydrotreatment process significantly improved the biocrude quality, with the highest upgraded biocrude yield of 97.0 wt % achieved at 400 °C when the extracted biocrude and residual solids were processed jointly. The resulting biocrude composition was predominantly hydrocarbons (78 wt %), with smaller fractions of oxygenates (14.0 wt %), cyclic/aromatics (6.0 wt %), and nitrogen-containing components (<5.0 wt %). This represents a substantial deoxygenation and denitrogenation (De-N ∼83.3%), transforming the initial oxygen and nitrogen-rich biocrude into a hydrocarbon-rich product suitable for fuel applications.

The gaseous products generated during hydrotreatment were analyzed and consisted primarily of H₂, CO, CO₂, and CH₄, along with the C-2 light hydrocarbons. Gas yield (mostly methane and carbon dioxide) showed no significant changes from 3.1 to 4.0 wt % as a function of temperature, respectively. The formation of CO₂ is indicative of decarboxylation pathways, while that of methane suggests the occurrence of methanation side reactions. The gas yield remained relatively stable, not increasing significantly with temperature, which indicates that under these conditions, the hydrotreatment favored liquid product formation over gas production.

3.1.5. HDT of MSW-Derived Biocrude

Municipal solid waste (MSW), including plastic waste, represents one of the most significant environmental challenges today, not only due to its complex composition and impact but also because of the massive quantities in which it is generated. These wastes not only pose a threat to natural ecosystems and public health but also symbolize inefficiencies in current waste management practices and the need for sustainable solutions. MSW encompasses both dry and wet raw materials, including kitchen scraps, poultry waste, dairy products, agricultural residues, and manure, which constitute its biodegradable components. These elements are primarily composed of polysaccharides (such as cellulose and hemicellulose), lignin, proteins, and lipids. In MSW, cellulose and hemicellulose are found mainly as paper, cardboard, and plant debris. Lignin is typically present in the lignocellulosic biomass portion of MSW such as garden waste. Lipids, which are largely nonpolar compounds like oils and fats, are found in used cooking oil, food scraps, and animal wastes. Proteins, composed of amino acid chains, are present in forms such as animal remains, food grains, and algal biomass.

Research on the upgrading of biocrude from the HTL process of MSW is very limited. To the best of our knowledge, only Eni has studied this topic, but only using the organic fraction of MSW, working at temperatures between 250 and 310 °C, a pressure of 10 MPa, and reaction times between 60 and 120 min. The results obtained were a fuel with a yield between 25 and 37 wt % and an HHV of 35 MJ·kg^–1.

3.1.6. HDT of Sewage Sludge-Derived Biocrude

Sewage sludge is a byproduct of wastewater treatment and has gained attention as an HTL feedstock due to its high organic matter content (40–70 wt %), including lipids (5–20 wt %), proteins (20–30 wt %), and carbohydrates (10–20 wt %), along with significant inorganic constituents (ash: 20–50 wt %). HTL of sewage sludge typically yields 25–35 wt % biocrude with a high heating value (HHV) of 30–35 MJ/kg, though the presence of heteroatoms (N: 4–7 wt %, O: 10–15 wt %, S: 0.5–2 wt %) and metals (Fe, Ca, Al) poses upgrading challenges. Biocrude composition is dominated by fatty acids (C16–C18), nitrogenous compounds (amides, pyrroles), and cyclic oxygenates, reflecting proteins and lipid degradation. The high moisture content (80–95 wt %) of raw sludge eliminates the need for energy-intensive drying, making HTL energetically favorable compared to pyrolysis. However, the ash content, particularly silica and phosphates, can lead to reactor fouling and catalyst deactivation during upgrading.

Castello et al. compared hydrotreating upgrading of three different biocrudes, from miscanthus, microalga Spirulina, and sewage sludge. Based on their results, the most unfavorable operating hydrotreatment conditions for biocrude upgrading produced from Spirulina microalga are those at high temperature (∼400 °C), where nitrogen compounds are unstable, although operating pressure was high enough at 8 MPa. On the other hand, biocrude that was produced from lignocellulosic biomass like miscanthus, presented the highest coke formation at low pressure, where hydrogen was not sufficient to stabilize the biocrude. Furthermore, in the case of deoxygenation reactions, an 80% deoxygenation efficiency was achieved at mild operation of 350 °C and 4 MPa pressure for both biocrude produced from sewage sludge and Spirulina. However, it was observed that an increase of pressure from 4 to 8 MPa improved deoxygenation efficiency almost to 100%. Similar was the trend for denitrogenation efficiency, which increased from 30% at 4 MPa to 48 % at 8 MPa. In the case of sewage sludge, deoxygenation efficiency is high even at mild conditions; however, denitrogenation reactions require more severe conditions. The same trend also applies to lignocellulosic biocrude, which presented a better heteroatom removal at 8 MPa compared to 4 MPa operation. At the moment, there are no other studies available for the stabilization and upgrading of an HTL biocrude from sewage sludge, utilizing hydrotreating in batch operation mode.

3.1.7. Summary from Bench-Scale Operation Studies

It is observed that most of the HTL biocrude upgrading studies that were performed in batch reactors utilized primarily microalgae-derived biocrudes, due to the inherent high moisture content, rendering HTL the most suitable conversion technology. Various microalgae strains like C. sorokiniana, C. vulgaris, and N. gaditana have been converted via HTL, rendering biocrude, which has been effectively upgraded via catalytic hydrotreatment. Catalysts such as CoMo/γ-Al₂O₃, NiWS/Al₂O₃, NiMo/Al₂O₃, Ru/Al₂O₃, Pt/Al₂O₃, Pd/Al₂O₃, Pt/C, Ru/C, and Pd/C have been investigated, with CoMo/γ-Al₂O₃ and NiMo/Al₂O₃ being the most efficient in terms of product yields and heteroatom removal reactions. In the case of operating parameters, temperature strongly influences the reactions, as high temperatures favor heteroatom removal and cracking reactions, rendering higher quality products but lower product yields. Furthermore, high pressures are preferable as they favor reconversion and product yields. It should be noted that the literature for catalytic hydrotreatment upgrading of HTL biocrude derived from animal carcass, lignocellulosic biomass, MSW, and sewage sludge is rather limited. Overall, batch catalytic hydrotreatment of HTL biocrude studies have shown that hydrotreating is a necessary and substantial step for HTL biocrude upgrading, no matter what the initial HTL feed is. The resulting product is characterized by high heating value (>40 MJ/kg) and low oxygen (<10 wt %) and nitrogen content (<6 wt %). A summary of the properties of biocrudes upgraded from different HTL biocrude hydrotreating studies performed in batch reactors is presented in Table .

3. Properties of the Biocrudes Upgraded from Different Biocrude Types via Hydrotreating.

		Biocrude origin
Properties	Units	Lignocellulosic	Animal carcass	Algae	Food waste	Sewage sludge
O content	wt %	0.0–9.0	-	0–9	0.4	0.0–9.0
S content	wppm	-	-	0–265	100	-
H content	wt %	9–13	-	10–14	3–15	11–13
C content	wt %	82–88	-	75–87	85	82–87
N content	wt %	0.6–5.4	-	0.6–5.6	0.4	0.8–5.4
HHV	MJ/kg	41–46	44.9–46.7	37.6–46.3	32.4–48.0	41–46
References		,		,,,,

3.2. Pilot-Scale Research of HTL Biocrude Catalytic Hydrotreatment

The studies presented so far represent proof-of-concept investigations utilizing small-scale batch units. However, to scale up and commercialize a technology, its performance must be demonstrated in continuous operation systems at higher technology readiness levels (TRLs). At present, the available literature on continuous hydrotreatment systems remains limited, with most studies conducted in low-TRL pilot plants. An effort has been made to present state-of-the-art studies of HTL biocrude upgrading in continuous hydrotreatment systems. The studies are classified on the basis of the origin of the biocrude. Accordingly, the literature is presented in the following order, based on the initial feedstock: lignocellulosic biomass, livestock-derived feedstocks, algae, food waste, municipal solid waste (MSW), and sewage sludge. This classification enables comparison to batch reaction units, as discussed in the previous section. A summary of all studies conducted in continuous operation systems is provided in Table , including information on biocrude origin, catalyst type, operating conditions, and final product yields.

4. Summary of Biocrude Pilot-Scale Catalytic Hydrotreatment Testing.

Biocrude feed	Cat.	Oper. Cond.	Organic liquid product yields	refs
Forestry residue	1st stage: CoMo 2nd stage: NiMo	320–370 °C, 8–9 MPa, 0.2–0.5 hr–1 WHSV, 900–1300 Nm3/m3 oil	∼90 wt %
Clean pine	CoMo/Al2O3	400 °C and 10 MPa
Wheat straw	CoMo and NiMo	340, 360, 380 and 400 °C, 8 MPa, 0.7 h–1 for CoMo and 0.5 hr–1 for NiMo	Carbon yields 81 wt %
Cow manure	CoMo and NiMo	340, 360, 380 and 400 °C, 8 MPa, 0.7 h–1 for CoMo and 0.5 hr–1 for NiMo	Carbon yield 81 wt %
Blend of wheat straw and cow manure	CoMo and NiMo	340, 360, 380 and 400 °C, 8 MPa, 0.7 h–1 for CoMo and 0.5 hr–1 for NiMo	Carbon yield 73–92 wt % depending on the process Temp.
Microalgae (Chlorella)	CoMo/Al2O3	400 °C and 10 MPa
Algae (Tetraselmis sp. and cyanobacteria)	CoMo/Al2O3	10 MPa and 400 °C	Deposition of iron observed in the catalyst for both feedstocks
Wet wastes (food wastes, sewage sludge, and a blend of food waste and fats, oils, and grease)	CoMo and NiMo	400 °C, 10 MPa, WHSV: 0.5 h–1	>90% deoxygenation, >90% desulfurization, 63–87% denitrogenation, 15 wt % SAF yields
Sewage sludge and food waste	CoMo/Al2O3 and NiMo/Al2O3	350–400 °C, 3.4–10 MPa, WHSV 0.5–1 h–1	70 wt % diesel yields (150°–350 °C)
Sewage sludge	CoMo/Al2O3	400 °C and 10 MPa
Bio-pulp feed (municipal solid wastes)	NiMo/Al2O3	320–330 °C, WHSV 0.7 h–1, H2/oil ratio 1000 NL/L	Gasoline yields 4.6 wt %, jet yields 7.9 wt %, diesel yields 42.7 wt %, residue 42 wt %.

3.2.1. HDT of Agricultural Livestock Residues-Derived Biocrude

Agricultural wastes and livestock residues are abundant, low-cost biomass resources with significant potential for biofuel production via hydrothermal liquefaction (HTL). Agricultural residues (e.g., crop stalks, husks, and straw) are rich in lignocellulosic compounds, while livestock feedstocks (such as manure and slaughterhouse byproducts) contain proteins, lipids, and minerals. Their availability and renewable nature make them attractive candidates for sustainable biofuel generation and integration into circular waste-to-energy strategies. HTL of these feedstocks renders biocrudes with distinct chemical compositions, typically high in oxygenates and nitrogenous compounds, which necessitates catalytic upgrading for conversion into stable drop-in fuels.

Dimitriadis and Bezergianni investigated the upgrading of HTL biocrude produced from wheat straw to improve its stability and energy content. The study was conducted in a TRL 3 continuous hydrotreating using a commercial NiMo/Al₂O₃ catalyst at various operating conditions (temperature 330–360 °C, pressure 6.9 MPa, LHSV 1 hr^–1, and H₂/oil ratio 5000 scfb). The resulting product was a two-phase liquid consisting of an organic phase and an aqueous phase. The effectiveness of HTL biocrude hydrotreatment was proved from this study, as it significantly removed biocrude oxygen, producing a final product with 0 to 3 wt % oxygen content. In addition, a 41% increase in heating value was observed, increasing the biocrude HHV from 27 MJ/kg to 46 MJ/kg after upgrading. The organic liquid product comprised 7–12 wt % gasoline, 20–22 wt % jet fuel, 37–40 wt % diesel, and 28–36 wt % heavy fuel boiling range hydrocarbons. Although the results were promising, rapid catalyst deactivation was observed that progressed linearly with time on stream.

Another study by dos Passos et al. employed a continuous hydrotreating plant to upgrade three different types of biocrude produced from wheat straw, cow manure, and their blends. In this study, a double-bed reactor was used: the upper zone contained a commercial CoMo catalyst, while the lower zone contained a commercial NiMo catalyst; the two reaction zones were separated by glass beads. Experiments were conducted at four different temperatures (340, 360, 380, and 400 °C) under constant pressure (8 MPa) and weight hourly space velocity (WHSV) (0.7 h^–1 for CoMo and 0.5 h^–1 for NiMo). The hydrotreatment produced a two-phase liquid product, with phase separation strongly influenced by process temperature. At lower temperatures, separation between the phases was more difficult. Carbon yields for wheat straw and cow manure-derived biocrudes were approximately 81 wt %, while the carbon yields for the blend of the two feedstocks (co-HTL of wheat straw and cow manure) ranged from 73 wt % at 400 °C to 92 wt % at 340 °C. The total liquid product collected at 400 °C was distilled to separate bio-gasoline, kerosene, and diesel fractions. Gasoline yields ranged from 6 to 20 wt %, depending on the biocrude origin, with the highest yield observed for cow manure biocrude. Biokerosene yields ranged from 15 to 25 wt %, with the highest yield for wheat straw biocrude. Biodiesel yields ranged from 22 to 30 wt %, with the highest observed for cow manure biocrude.

3.2.2. HDT of Algae-Derived Biocrude

Algae represent a highly promising feedstock for hydrothermal liquefaction (HTL) due to their rapid growth rates, high lipid content, and ability to grow on non-arable land using wastewater or seawater. Algal biomass typically produces biocrudes rich in lipids (15–70 wt %) and proteins (20–70 wt %), resulting in high energy density but also significant nitrogen content, which can challenge direct fuel use. Catalytic upgrading, particularly hydrotreating, is required to remove heteroatoms and stabilize the biocrude, producing renewable “drop-in” fuels compatible with existing refinery infrastructure. However, the upgrading step via hydrotreatment remains a challenge. To address this, Jarvis et al. investigated the potential upgrading of two biocrudes produced from Tetraselmis sp. and cyanobacteria, using a CoMo/Al₂O₃ catalyst. For the Tetraselmis sp.-derived biocrude, catalyst plugging was observed after 102 h of continuous operation, forcing termination of the experiment. In contrast, the cyanobacteria-derived biocrude was processed for 219.8 h of continuous operation without any catalyst plugging. Nevertheless, catalyst analysis after the experiments revealed that metal deposition occurred in both cases, with iron being deposited during the hydroprocessing of the algal biocrudes, causing catalyst plugging. This problem is generally observed in HTL biocrudes, which have higher metal contents (Na, K, Ca, Si, and Fe) compared to conventional crude oils. Further research is needed to understand the effect of metal deposition on catalyst pores during the hydrotreatment of HTL biocrudes.

3.2.3. HDT of Food Wastes and Sewage Sludge-Derived Biocrude

Food waste represents a high-potential HTL feedstock due to its high moisture content (60–85 wt %) and rich organic composition, including lipids (5–40 wt %), carbohydrates (20–50 wt %), and proteins (10–25 wt %), with minimal lignin content (<2 wt %) compared to lignocellulosic biomass. This composition facilitates higher biocrude yields (25–40 wt %) with lower oxygen content (8–15 wt %) than lignocellulosic-derived biocrudes, although nitrogen content (2–6 wt %) remains a challenge due to protein degradation into heterocyclic compounds (pyrroles, indoles). HTL of food waste converts triglycerides into fatty acids and glycerol, while carbohydrates and proteins yield cyclic ketones, phenolics, and nitrogenous species. The resulting biocrude has a high HHV of 30–38 MJ/kg but requires hydrotreatment to reduce acidity (TAN > 50 mg KOH/g) and heteroatom content for fuel applications. Variability in feedstock composition (household vs industrial waste) necessitates tailored preprocessing to ensure consistent HTL performance. ,− ,

Dimitriadis and Bezergianni investigated the upgrading of HTL biocrude produced from spent coffee grounds in a continuous HDT plant with a capacity of 40 mL/h. A commercial NiMo/Al₂O₃ catalyst was blended with a demetallization catalyst exhibiting moderate HDS activity at a 7.1 v/v ratio. The results showed that the hydrodeoxygenation activity of the catalyst was very high, yielding a product with an almost zero oxygen content. Furthermore, the heating value increased from 36 MJ/kg to 46 MJ/kg, while viscosity decreased from 27.7 cSt to less than 7 cSt, improving the stability of the initial bio-oil. The total duration of the experiment lasted 192 h of continuous operation without any noticeable coke formation or catalyst plugging. The duration of this study exceeded the duration of the study of Jarvis et al., which lasted 102 h due to catalyst plugging.

Subramaniam et al. investigated the upgrading of HTL biocrude produced from food waste and sewage sludge using a continuous operation unit. In their study, a commercial CoMo/Al₂O₃ catalyst operating at 350 °C served as a guard catalyst, while a commercial NiMo/Al₂O₃ catalyst operating at 400 °C served as the main catalyst. The experiment was conducted for 1500 h under steady-state conditions without any catalyst deactivation. The upgraded biocrude consisted of a complex mixture of aromatics, heterocyclics, n-alkanes, and iso-alkanes. According to the distillation curve of the organic liquid product, it comprised 9–12 wt % gasoline boiling range hydrocarbons (<150 °C), 66.2–67.8 wt % diesel-boiling-range hydrocarbons (150–350 °C), and 21–24 wt % heavier-than-diesel hydrocarbons (>350 °C). The study also demonstrated that high pressure is necessary for efficient hydrotreatment.

Jarvis et al. compared hydrotreated biocrudes from different biomass origins, including clean pine, microalgae (Chlorella sp.), and sewage sludge. These biocrudes were upgraded in a continuous operation hydrotreatment plant at 400 °C and 10 MPa. Hydrotreatment improved the properties of all three biocrudes: carbon content increased to 84–87 wt %, hydrogen content to 11–15 wt %, nitrogen and oxygen were reduced to <0.5 wt % and 1.2 wt %, respectively, and sulfur content was reduced to <50 ppm. Density, viscosity, and moisture content also decreased after hydrotreatment. A comparison of hydrotreated products indicated that products derived from sewage sludge and algae were dominated by hydrocarbons. Hydrocarbons in algae biocrude were more aromatic than those in sewage sludge biocrude, suggesting that polymerization or condensation reactions are more prevalent during algae biocrude hydrotreatment. Conversely, the presence of more nitrogen species in hydrotreated sewage sludge indicates a higher occurrence of cracking reactions. Pine biocrude hydrotreatment appeared more similar to microalgae biocrude than to sewage sludge-derived biocrude, with the hydrotreated pine biocrude exhibiting a more aromatic character, attributable to its higher lignin and starch content relative to the higher lipid content of microalgae and sewage sludge.

Cronin et al. explored hydrotreatment of HTL biocrudes derived from wet wastes, including food waste, sewage sludge, and blends of food waste with fats, oils, and grease, in a continuous trickle-bed reactor using CoMo and NiMo catalysts. Hydrotreatment effectively reduced heteroatom content, achieving >90% deoxygenation and desulfurization and 63–87% denitrogenation. The primary goal was the production of sustainable aviation fuel (SAF) in the 150–250 °C boiling range; however, only 15–25 wt % of the upgraded biocrude fell within this cut, which also exhibited high nitrogen content. The remaining product consisted of heavier hydrocarbons. Nitrogen content in the SAF fraction ranged between 0.96 wt % and 1.34 wt %, far exceeding the <0.001 wt % specification for Jet A fuel, indicating the need for additional hydrodenitrogenation. Furthermore, a second hydrocracking step is necessary to increase SAF fraction yields. These results highlight the need for further research to optimize SAF production via hydrotreatment of HTL biocrudes derived from wet wastes.

3.2.4. HDT of Municipal Solid Wastes (MSW)-Derived Biocrude

Municipal solid waste (MSW) represents a heterogeneous but potentially valuable HTL feedstock, composed of organic fractions (40–60 wt %, including food waste, paper, and yard trimmings) and synthetic materials (10–30 wt %, such as plastics including PP, PET, and PS). The organic components, rich in carbohydrates (20–40 wt %), lipids (5–15 wt %), and proteins (5–10 wt %), contribute to biocrude yields of 20–35 wt % under typical HTL conditions, while the plastic fraction can enhance hydrocarbon production through thermal depolymerization. The resulting biocrude exhibits a high HHV of 35–40 MJ/kg, though its composition varies with the feedstock mix, containing both oxygenated compounds (from biomass) and pure hydrocarbons (from plastics). Key challenges include the high heterogeneity of MSW, which leads to inconsistent product quality, and the presence of contaminants (e.g., heavy metals, chlorine from PVC) that may corrode equipment or poison catalysts. The inorganic fraction (20–40 wt % ash) also poses operational difficulties, such as reactor fouling. Catalytic HTL or subsequent hydroprocessing can improve fuel quality by reducing oxygen (5–12 wt %) and nitrogen (1–3 wt %) content, with upgraded oils achieving 80–90 wt % hydrocarbons in the diesel/jet fuel range (C10–C20). While MSW-derived biocrude shows good potential for waste-to-energy applications, effective preprocessing (sorting, shredding) and careful management of halogenated compounds remain critical for commercial viability. ,,

Kohansal et al. investigated the mild upgrading of an HTL biocrude produced from MSW, specifically derived from bio-pulp feedstock. The biocrude upgrading was performed in a continuous hydrotreating plant at a WHSV of 0.7 h^–1, a temperature of 320–330 °C, and a volumetric H₂/oil ratio of 1000 NL/L using a commercial NiMo/Al₂O₃ catalyst. The organic liquid product after hydrotreatment exhibited low gasoline yields (4.6 wt %) and low jet-fuel yields (7.9 wt %), but high diesel yields (42.7 wt %). The remaining residue fraction, with a boiling point above 375 °C, accounted for 42 wt %. Although the residue fraction remained high, total distillate recovery improved significantly, primarily due to hydrodeoxygenation reactions that removed oxygenated compounds, reducing the average molecular weight and lowering the boiling point distribution. Despite these improvements, the hydrotreated biocrude still exhibited high sulfur content (0.07 wt %), viscosity (25.49 mm²/s), micro carbon residue (2.88 wt %), and water content (0.86 wt %). As a result, when blended with commercial diesel at levels exceeding 5 %v/v, the final blend did not meet EN 590 diesel specifications.

3.2.5. Summary from Continuous Operation Studies

Based on previous findings from continuous operation systems, hydrotreatment is recognized as a key technology for successfully upgrading HTL biocrude into liquid fuels. The most commonly explored catalysts with promising results are commercial NiMo/Al₂O₃ and CoMo/Al₂O₃. The total liquid product after hydrotreatment is typically a two-phase mixture, consisting of an organic phase, which serves as the main fuel, and an aqueous phase that is often ∼99 wt % water. Separation of the two phases is generally straightforward and can be achieved via gravity. The organic liquid phase contains hydrocarbons spanning across different boiling ranges, necessitating a subsequent distillation step to separate distinct fuel fractions, such as gasoline, kerosene, diesel, and heavier residues. The boiling ranges of these fractions are not universally standardized and are typically defined by the research group. For example, in the study by dos Passos et al., the gasoline fraction included hydrocarbons with boiling points below 150 °C, the kerosene fraction ranged from 150–250 °C, the diesel fraction from 250–360 °C, and heavier residues were those with boiling points above >360 °C. Other studies suggested alternative ranges, such as 66–200 °C for gasoline, 200–300 °C for kerosene, and 300–360 °C for diesel.

In summary, although some research has been conducted on continuous operation systems, scaling up and commercializing the technology requires demonstration at higher Technology Readiness Levels, potentially reaching TRL 8. It is evident that current technology remains far from commercial scale, highlighting the need for further studies using higher TRL systems to evaluate and demonstrate technological maturity. Table summarizes the properties of biocrudes upgraded in continuous operation units, while Table presents the properties of various drop-in fuels produced via hydrotreatment of HTL biocrude.

5. Properties of Hydrotreated Biocrude Derived from Different Feedstocks.

		Biocrude origin
Properties	Units	Forestry residue	Lignocellulosic	Algae	Organic waste	Sewage sludge	Municipal wastes
Density at 15 °C	g/mL	0.900–1.020	0.88–0.95	0.780–0.927	0.807–0.810	0.790–0.795	0.77–0.87
O content	wt %	2.2–4.7	0–1	1.2	0.15–0.35	0.15–1.2	7.4–0.9
S content	wppm	46	-	17–1859	10–30	10–30	1200–100
H content	wt %	10.4–11.0	-	15–15.5		15	11.0–14.6
C content	wt %	84.2–87	-	75.1–86	84.9–85.5	83.4–84.9	77–86
N content	wt %	<0.05	1–3.5	0.05–7.9	0.9–1.3	0.05–1.3	2.31–3.54
H2O dissolved content	wt %	<0.3	0.1–0.5	<0.2	-	<0.1	0.09–0.54
Viscosity	cp	78–3321	-	-	3.064–3.196	2.535–3.016	-
Cinematic viscosity	cSt	4.5	6–10	2.2	-	2.5	1.0–4.5
HHV	MJ/kg	39.4–42.7	42–45	-	-	-	40–49
Flash point	°C	-	-	-	-	-	55
Cetane index	-	-	-	-	-	-	30–90
Aqueous yields	wt %	5.3–8.9	-	10–11	-	-	1.32
Organic yields	wt %	91.1–94.7	-	-	-	-	98.68
TAN	Mg KOH/g	3.7–13.3	0.05–1	-	-	<0.01	-
References		,		,		,

6. Properties of Fuel Types Produced by Biocrude Hydrotreating .

		Fuel type
Properties	Units	Gasoline (<150 °C)	Kerosene (150–250 °C)	Diesel (250–360 °C)	Heavier HC (>360 °C)
Density at 15 °C	g/mL	0.75	0.85	0.90	1.0
O content	wt %	0.10	0.25	0.30	0.20
S content	wppm	20	40	60	80
N content	wt %	0.1	0.4	0.5	0.7
Viscosity	cSt	1	5	10	1000
HHV	MJ/kg	47	46	45	43

3.3. Explaining Apparent Variability in NiMo/Al₂O₃ Performance across Operating Regimes

NiMo/Al₂O₃ is the most frequently tested catalyst in this review. Taken at face value, the reported biocrude-to-organic liquid product yields for this single catalyst span an extraordinarily wide range, from 14–45 wt % for municipal solid waste (MSW) biocrude to 96.6 wt % for N. gaditana biocrude, with intermediate values of 81–92 wt % for lignocellulosic, algal, and sewage-sludge biocrudes processed under comparable conditions and 54–90 wt % for two algal feedstocks (C. vulgaris and N. gaditana). Rather than indicating inconsistent catalyst behavior, this range is largely explained by confounding operating variables that differ systematically between studies. The lowest yields (14–45 wt %,) correspond to the most severe pressure regime in this data set (15–22 MPa), combined with the shortest residence times (10–60 min), conditions that favor secondary cracking and gas phase losses over liquid product retention. In contrast, the highest yield (96.6 wt %,) was obtained at a comparatively moderate pressure (5 MPa) and longer reaction time (60 min) that favors deoxygenation without excessive cracking. The intermediate yield range (81–92 wt %,) spans a temperature window of 350–400 °C and pressure window of 4–8 MPa across three different feedstocks (miscanthus, Spirulina, sewage sludge) processed in the same study, indicating that once pressure and residence time are held approximately constant, feedstock identity alone accounts for a comparatively narrow yield variation (within about 11 %) relative to the variation introduced by pressure and residence time differences between studies.

A particularly instructive case is provided by ref (N. gaditana biocrude, NiMo/Al₂O₃), which was tested both in batch mode (Table : 400 °C, 5 MPa, 1 h, 96.6 wt % yield) and in continuous pilot-scale mode (Table : 380 °C, 60 wt % yield) using the same catalyst and feedstock. The 36.6 % difference between these two results cannot be attributed to catalyst identity, since it is the same catalyst and feedstock; instead, it illustrates a distinct and frequently overlooked source of apparent inconsistency in the literature: batch reactors are charged with fresh catalyst for every single run and typically report yields from a single pass, well mixed system with no time-on-stream (TOS) effects, whereas continuous units expose the catalyst to progressive deactivation (Section ) and report yields under quasi-steady-state conditions that already reflect some degree of activity loss. This distinction means that batch-scale yields should be interpreted as an upper bound, mechanistic benchmark for a given catalyst feedstock pair, rather than as a directly comparable figure to pilot-scale or industrially relevant continuous yields, and that comparisons of catalyst performance across the literature are only meaningful when reactor mode, time-on-stream, pressure and residence time are considered jointly rather than treating the catalyst identity as the sole explanatory variable. This reframing directly addresses why nominally identical catalysts such as NiMo/Al₂O₃ can appear to perform inconsistently across Table and Table . The inconsistency is largely an artifact of comparing across incompatible operating regimes and reactor configurations, not evidence against the catalyst’s intrinsic activity.

3.4. Quantitative Influence of Feedstock Composition on Hydrotreating Performance

The initial oxygen content of HTL biocrude varies systematically with feedstock chemical class, and this variation only partially predicts hydrotreating performance. Lignocellulosic-derived biocrude carries the highest initial oxygen content (17–23 wt %), reflecting its lignin and cellulose-derived phenolic and carbonyl structures, and requires catalytic HDO to reach hydrocarbon contents of up to 96 wt % and HHVs of up to 42 MJ/kg. , Microalgal biocrude, by contrast, typically carries a lower initial oxygen content (e.g., 8.3 wt % in Silva et al.), yet this lower starting point does not guarantee easier upgrading: under mild conditions (375 °C, 10 MPa, 1 h), Silva et al. achieved almost no deoxygenation (8.3 to 8.0 wt %), whereas Patel et al., testing a comparable algal biocrude under a wider set of catalysts, reduced oxygen content to 1.6–6.0 wt %, comparable to fossil crude oil. This contrast indicates that oxygen content alone is a poor predictor of hydrotreating outcome; reaction severity and catalyst choice interact strongly with the specific chemical form in which oxygen is bound (phenolics and ethers in lignocellulosic biocrude versus fatty-acid and ketone-derived oxygenates in algal biocrude), rather than with its bulk elemental content.

Nitrogen removal consistently lags behind oxygen removal across all feedstocks examined, but the gap is feedstock-dependent. Castello et al. demonstrated this directly for both sewage sludge and spirulina biocrude under identical conditions: increasing pressure from 4 to 8 MPa raised deoxygenation from 80% to nearly 100%, while denitrogenation rose only from 30% to 48% under the same pressure change. This indicates that nitrogen-containing structures in HTL biocrude (heterocyclic amines, amides, and pyrroles) are intrinsically more resistant to hydrotreating than oxygenated structures, regardless of feedstock origin, and that pressure increases alone are insufficient to close this gap. A related, feedstock specific effect was reported for two microalgae strains hydrotreated under otherwise identical conditions with NiWS/Al₂O₃: C. vulgaris biocrude achieved a higher degree of denitrogenation than C. sorokiniana biocrude, but this was accompanied by greater iron accumulation on the spent catalyst, linking the depth of nitrogen removal to a parallel increase in hydrodemetallization side reactions and, by extension, to faster catalyst deactivation (Section ) for feedstocks with more recalcitrant nitrogen content.

Feedstock composition also determines whether increasing temperature improves or degrades the yield quality trade-off. For food waste-derived biocrude, rich in lipids and proteins rather than lignin, Dhakal et al. achieved both a high yield (97.0 wt %) and substantial heteroatom removal (denitrogenation of ∼83.3%, yielding a hydrocarbon-rich product of 78 wt %) at the most severe temperature tested (400 °C). This is the opposite trend to that observed for algal biocrude under a comparable temperature range: Guo et al., upgrading C. vulgaris and N. gaditana biocrude with NiMo/Al₂O₃ and NiW/Al₂O₃, reported yields falling from 90 wt % at 250 °C to only 54 wt % at 400 °C, attributing the loss to cracking reactions favored at higher temperature. A similar temperature sensitivity, but through a distinct mechanism, was reported by Castello et al. for miscanthus (lignocellulosic), Spirulina (algal), and sewage-sludge biocrude: miscanthus-derived biocrude showed the highest coke formation at low pressure, where hydrogen availability was insufficient to stabilize its recalcitrant aromatic structures, whereas Spirulina-derived biocrude showed its worst performance at high temperature (400 °C) even at adequately high pressure (8 MPa) because its characteristic nitrogen-rich heterocycles become thermally unstable under these conditions. These contrasting outcomes demonstrate that the yield optimal operating window is not universal but is set by the dominant heteroatom chemistry of each feedstock class: lipid/protein-rich feedstocks tolerate high severity well, lignin-rich feedstocks require sufficient hydrogen availability (pressure) to avoid coking, and nitrogen-rich algal feedstocks are constrained by an upper temperature limit before their heterocyclic nitrogen content begins to degrade product yield. A comparative performance heat map of the major HTL biocrude feedstocks reported in the literature is presented in Figure . Relative scores (1–5) were assigned based on the reported upgrading performance, heteroatom-removal efficiencies, fuel properties, catalyst applicability, and scale-up potential discussed throughout the reviewed studies. The scoring system is intended to facilitate qualitative comparison among feedstocks and does not represent a standardized experimental metric.

1.

Comparative performance heat map for HTL biocrude hydrotreatment feedstocks (relative literature-based scores: 1 = low, 5 = high).

Finally, the feedstock heteroatom profile also informs catalyst selection. Guo et al. found that NiMo/Al₂O₃ outperformed NiW/Al₂O₃ in HDO and HDN, while NiW/Al₂O₃ was more efficient for HDS, on C. vulgaris and N. gaditana biocrude, suggesting that catalyst formulation should be matched to the dominant heteroatom burden of the specific feedstock (oxygen/nitrogen-rich versus sulfur-rich biocrudes) rather than selected independently of feedstock chemistry, reinforcing the comparative catalyst discussion in Section . Guo et al. reported also that, for C. vulgaris biocrude upgraded with NiMo/Al₂O₃ at 400 °C, oxygen content dropped from 14.6 wt % (raw biocrude) to 1 wt % (upgraded), while nitrogen and sulfur contents were reduced to 52% and 25% of their original values, respectively (i.e., from 6.9 wt % to ∼3.6 wt % nitrogen, and from 0.8 wt % to ∼0.2 wt % sulfur), raising the HHV from 33.76 to 42 MJ/kg.

3.5. Do Universal Relationships Exist between Reaction Conditions and Product Properties?

Among the studies reviewed, only one explicitly ranked the relative influence of the main operating variables against each other. Haghighat et al., testing a continuous hydrotreater across temperature (320–370 °C), pressure (80–95 bar), WHSV, and H₂/oil ratio, found that temperature had the strongest influence on conversion, followed by WHSV as the second most effective parameter, while pressure and H₂/oil ratio were both found to be less important than temperature and WHSV. This ranking is useful precisely because it was obtained by varying all four parameters within a single, controlled system, rather than being inferred by comparing across different studies with confounded feedstocks and reactor designs (Section ). The findings indicate that, once feedstock and catalyst are fixed, temperature is the primary lever for both conversion and product quality, with WHSV as a secondary lever and pressure and H₂/oil ratio playing a comparatively minor role over the ranges tested.

Reaction condition design governs not only the absolute values of temperature and pressure but also the trade-off between yield and product quality. Castello et al. processed Spirulina biocrude in a continuous hydrotreater under single-stage conditions (200–340 °C, 10 MPa, WHSV 0.50 h^–1), obtaining total yields of 97.0–99.3 wt %, but even the most severe single-stage condition (340 °C) only reduced the nitrogen content from 7.6 wt % (raw biocrude) to 4.2 wt %. Adding a second hydrotreating stage (holding the first stage fixed at 340 °C while varying the second stage’s temperature, WHSV, or pressure) drove nitrogen content down substantially further, to 0.6–3.6 wt % depending on second-stage severity. This shows that, for a nitrogen-rich algal feedstock, a single hydrotreating stage, regardless of the temperature applied within it, reaches a ceiling in denitrogenation that cannot be overcome by further increasing the severity of that same stage; only the introduction of a second, independently controlled stage allows deep denitrogenation to be achieved, at a comparatively modest cost in overall yield.

Different heteroatom-removal and contaminant-removal reactions also show distinct temperature thresholds within the same reaction system, rather than responding uniformly to overall severity. Haider et al., screening a continuous hydrotreater processing Spirulina biocrude at four fixed temperatures (200, 250, 300, and 340 °C) under constant pressure (100 bar) and WHSV (0.5 h^–1), found that demetallization reached 98% already at 300 °C, while complete deoxygenation required the full 340 °C. This staged threshold behavior, metal removal saturating at a lower temperature than oxygen removal, is consistent with the mechanistic expectation that metal-porphyrin decomposition proceeds more readily than full C–O bond cleavage, and demonstrates that a single temperature value cannot be treated as equally rate-limiting for every deactivation-relevant reaction (Section ) discussed elsewhere in this review. This temperature-dependent reactivity mapping directly informed the design of a graded catalyst bed (a low-activity demetallization zone followed by a higher-activity deoxygenation zone within the same reactor, operating nominally at 340 °C), which sustained 335 h of continuous operation without the coking and deactivation losses that a uniform, single-temperature configuration would incur. Process temperature should be selected according to which specific heteroatom or contaminant is the binding constraint for a given feedstock and product specification, rather than as a single value optimized for only one reaction.

Taken together, these findings: the explicit primacy of temperature over WHSV, pressure, and H₂/oil ratio in a controlled comparison; the ability of staged operation to widen the viable temperature window without sacrificing yield; and the staggered temperature thresholds for demetallization versus deoxygenation within a single graded bed, indicate that no single operating variable behaves as a universal, isolated predictor of product properties. Temperature consistently emerges as the dominant variable, but its effect on yield, heteroatom removal, and catalyst lifespan is strongly mediated by reactor configuration (single vs staged operation) and by which specific heteroatom-removal reaction is rate-limiting for a given feedstock.

Throughout Sections –3.5, catalyst and process performance have been quantified almost exclusively in terms of the extent of HDO, HDN, and HDS achieved under a given set of conditions. Despite differing in the heteroatom removed and in the specific reaction network involved, HDO, HDN, and HDS share a common overall logic: hydrogenation and/or hydrogenolysis of the heteroatom-containing functional group or ring, releasing the heteroatom as a small stable molecule (H₂O, NH₃, or H₂S, respectively) and leaving a heteroatom-free hydrocarbon product. For representative oxygen-, nitrogen- and sulfur-containing model compounds, these pathways are illustrated in Figure , Figure , and Figure : hydrodeoxygenation of phenol proceeds via two competing routes, direct deoxygenation and ring hydrogenation, that converge on cyclohexane (Figure ); hydrodenitrogenation of pyridine requires ring saturation and hydrogenolytic ring opening before the C–N bond can be cleaved to release NH₃ (Figure ); and hydrodesulfurization of thiophene proceeds mainly via direct hydrogenolysis of the C–S bonds to release H₂S (Figure ). This shared framework closes the picture built up across the batch and continuous studies reviewed above, in which the extent, rate, and selectivity of these three reactions were shown to depend strongly on feedstock, catalyst, and operating conditions.

2.

Schematic drawn by the authors to illustrate the generally accepted dual-pathway (DDO/HYD) mechanism for hydrodeoxygenation of phenolic compounds over sulfided hydrotreating catalysts.

3.

Simplified HDN pathway for pyridine, a representative N-heterocyclic model compound in HTL biocrude. Denitrogenation proceeds via saturation of the aromatic ring (pyridine → piperidine), hydrogenolytic ring opening to the linear amine (pentylamine), and final C–N bond cleavage to release NH3, yielding the corresponding paraffin (pentane). Adapted from the generally accepted HDN mechanism for pyridine-type model compounds. (Reprinted (adapted) with permission from Duan, P.; Savage, P.E. Catalytic hydrothermal hydrodenitrogenation of pyridine. Applied Catalysis B: Environmental 2011, 108–109, 54–60. Copyright 2011 Elsevier B.V).

4.

Simplified HDS pathway for thiophene, a representative S-heterocyclic model compound in HTL biocrude. Desulfurization proceeds mainly through direct desulfurization (DDS), i.e., hydrogenolysis of the C–S bonds to release H₂S and give a butadiene-type intermediate, followed by hydrogenation to n-butane. Adapted from the generally accepted HDS mechanism for thiophene-type model compounds. (Reprinted (adapted) with permission from Babich, I.V.; Moulijn, J.A. Science and technology of novel processes for deep desulfurization of oil refinery streams: a review. Fuel 2003, 82, 607–631. Copyright 2002 Elsevier Science Ltd.).

4. Catalyst Deactivation Mechanisms in HTL Biocrude Hydrotreatment

Catalyst deactivation is one of the most significant engineering issues in the industrial deployment of HTL biocrude hydrotreatment, yet it has typically been reported only as a secondary observation within individual feedstock- or reactor-focused studies. This section consolidates the available evidence into three principal deactivation pathways: (i) coke/carbon deposition, (ii) metal deposition (hydrodemetallization), and (iii) nitrogen/sulfur poisoning, followed by a comparative discussion of catalyst behavior and a critical assessment of regeneration strategies and catalyst lifespan data.

4.1. Coke and Carbonaceous Deposit Formation

Coking arises from condensation and polymerization of unsaturated, aromatic, and phenolic intermediates on catalytically active sites, in direct kinetic competition with hydrogenation, HDO, HDN, and HDS reactions. This competition intensifies under severe operating conditions, where dehydrogenation/aromatization pathways are favored over saturation, ultimately causing pressure buildup and reactor plugging.

Haider et al. quantified this trade-off directly: single-stage hydrotreatment at 400 °C produced 9.1 wt % coke from Spirulina-derived biocrude and 3.4 wt % from sewage sludge biocrude, whereas a two-stage approach (mild stabilization followed by severe hydrotreating) reduced coke formation to 1.0% and 0.7%, respectively, with second-stage coke yields as low as 0.014–0.05%. Jensen similarly reported that continuous fixed-bed hydrotreating of forestry residue biocrude under severe conditions caused a significant pressure drop and deactivation within the first 10 h of operation, whereas operation at a milder temperature between 300 °C and 330 °C sustained 330 h without plugging.

Feedstock origin and operating severity jointly determine coking propensity. Castello et al. reported coke yields (as wt C%) of up to 14.1% for miscanthus-derived biocrude at low pressure (4 MPa, 350 °C), attributed to insufficient hydrogen availability to stabilize reactive lignocellulosic intermediates, versus 9.3% for algal (Spirulina) biocrude at the most severe condition tested (400 °C, 8 MPa), where nitrogen-containing aromatic structures become thermally unstable. Sewage sludge biocrude was comparatively coke resistant (2.0–4.2 wt C% across all conditions tested). These authors identified operating at higher H₂ pressure as the single most effective lever for suppressing coke formation across feedstocks.

Catalyst comparison studies confirm that coking is both catalyst and feedstock-dependent. Yang et al., upgrading animal carcass biocrude, reported char yields ranging from 2.45 wt % (Mo₂C) to 7.60 wt % (NiMo/Al₂O₃) depending on catalyst identity, and explicitly identified aromatic compounds as the principal coke precursors. In continuous double bed operation upgrading cow manure, wheat straw, and their co-HTL blend, dos Passos et al. observed that the micro carbon residue (MCR) of the upgraded biocrude fell from ∼26 wt % to below 5 wt % after hydrotreatment, but rose again with increasing severity operation condition specifically for the co-HTL blend, and that carbon buildup was consistently higher on the downstream NiMo bed than on the upstream CoMo guard bed, attributed to more recalcitrant nitrogen-containing aromatics reaching the second bed and inducing mass-transfer limitations. Elemental analysis of spent NiWS/Al₂O₃ catalyst used to upgrade C. sorokiniana and C. vulgaris biocrudes showed carbon content increasing from 0.1 wt % (fresh) to 7.9–8.0 wt % (spent), accompanied by a corresponding decrease in catalyst sulfur content (8.5% to 4.3–4.4%), indicating that coke deposition is compounded by a gradual loss of the catalyst active sulfided phase.

Mitigation strategies reported in the literature include (i) operating at higher hydrogen pressure and moderate temperature to favor hydrogenation over condensation; (ii) two-stage hydrotreating, in which a first mild stage removes the most reactive, coke-prone species before a second, more severe stage; and (iii) blending biocrude with hydrogen-donor solvents, for example, used engine oil, which reduced coke yields below 12 wt % relative to neat biocrude processed under identical conditions, with the effect of most pronounced for heavier, more aromatic distillate cuts.

As illustrated in Figure , coke formation proceeds through sequential adsorption, dehydrogenation, and condensation of unsaturated intermediates on the catalyst surface, ultimately forming a polyaromatic layer that blocks both the active site and the pore mouth. This progression is consistent with the quantitative evidence above: the substantially lower coke yields observed under milder, two-stage conditions or at higher H₂ pressure reflect suppression of the initial dehydrogenation/condensation steps, while the catalyst- and feedstock-dependent char yields reflect how readily different aromatic precursors enter this pathway.

5.

Schematic mechanism of catalyst deactivation by coking (carbon deposition): adsorption and dehydrogenation of aromatic/olefinic intermediates, followed by condensation into a polyaromatic coke layer that blocks the active site and pore.

4.2. Metal Deposition (Hydrodemetallization)

HTL biocrudes, unlike petroleum feedstocks, carry substantial concentrations of alkali, alkaline-earth, and transition metals (Na, K, Ca, Mg, Si, and, most critically, Fe) originating from the mineral and pigment content of the parent biomass. These species are not removed by conventional distillation and instead deposit irreversibly on the hydrotreating catalyst, causing pore blockage and active-site coverage that is mechanistically distinct from carbonaceous coking.

Jarvis et al. provided direct evidence of this mechanism during continuous hydroprocessing of Tetraselmis sp. and cyanobacteria-derived biocrudes over CoMo/Al₂O₃. Tetraselmis-derived biocrude caused catalyst plugging after only 102 h of operation, whereas the cyanobacteria-derived biocrude was processed for 219.8 h without plugging. In both cases, however, post-run catalyst analysis confirmed substantial iron deposition, traced to the conversion of iron-porphyrin pigments present in the algal biomass. This iron-porphyrin pathway was independently confirmed by Magalhaes et al., who observed clear XRF evidence of iron accumulation on spent NiWS/Al₂O₃ catalyst and attributed it to the same pigment-conversion mechanism. In a companion continuous-HTL study, the same group quantified this effect: iron content on the spent catalyst reached 0.63 wt % after upgrading C. sorokiniana biocrude versus 1.60 wt % after upgrading C. vulgaris biocrude, nearly a threefold difference, that correlated with the higher hydrodenitrogenation degree achieved for the C. vulgaris feed, echoing the analogous feedstock-dependent iron accumulation trend reported by Guoa et al. for the same microalgae species.

Because metal content is feedstock-intrinsic rather than reaction-condition-dependent, upstream mitigation is the most effective control strategy. Marrakchi et al. showed that the HTL stage catalyst or pretreatment used strongly affects the metal carryover into the biocrude (Na 0–0.53 wt %, Ca 0–0.14 wt %, K 0.01–1.20 wt %, Fe 0.01–0.38 wt %, depending on conditions), directly framing metal control as a means of extending downstream hydrotreating catalyst lifetime. Summers et al. demonstrated a dedicated desalting/dewatering/deashing pretreatment train for food waste-derived biocrude that removed 91.6% of silicon, 88.3% of calcium, and 67.2% of sodium prior to hydrotreatment, explicitly framing uncontrolled salt, water, and ash content as a cause of “catalyst deactivation during hydroprocessing” in addition to fouling and reduced distillation efficiency.

Metal deposition, summarized in Figure , proceeds through adsorption and decomposition of metalloporphyrin-type compounds at the pore mouth, where the released metal accumulates and progressively restricts access to the interior active sites. This pore-mouth-limited mechanism explains why deposition tracks feedstock pigment/metal content rather than reaction severity and why upstream removal of metals prior to hydrotreating, rather than adjusting hydrotreating conditions, is the most effective mitigation strategy.

6.

Schematic mechanism of catalyst deactivation by metal deposition (hydrodemetallization): metalloporphyrin-type compounds adsorb and decompose at the pore mouth, and progressive metal accumulation plugs the pore, rendering interior active sites inaccessible.

4.3. Nitrogen and Sulfur Species Poisoning

Nitrogen-containing heterocycles: pyrroles, indoles, carbazoles, and pyridines, derived primarily from protein degradation, are markedly more resistant to hydroprocessing than oxygenated species and are widely reported as active-site poisons for both sulfided and noble metal catalysts. Magalhaes et al. found that, even at 400 °C, denitrogenation plateaued below 90% while deoxygenation and desulfurization both exceeded 90% under the same conditions; FT-ICR MS analysis of a two-stage continuous hydrotreater processing Spirulina biocrude showed that two-ring nitrogen (N2-class) species were essentially eliminated after the severe second stage, while single-ring N1-class species (pyridine/indole-type) persisted. Castello et al. quantified this lag directly: conditions achieving >80% deoxygenation for algal and sewage sludge biocrudes yielded only 30–48% denitrogenation under the same temperature/pressure envelope, regardless of the temperature used.

Sulfur plays a dual feedstock-dependent role. HTL biocrude typically contains far less sulfur than petroleum fractions, and several reviews note that conventional sulfided NiMo/CoMo catalysts can gradually lose their active sulfided phase in the absence of a continuous sulfur source, requiring periodic H₂S (or sulfur donor) dosing to sustain activity. Conversely, in sulfur-rich biocrudes (e.g., from sewage sludge), sulfur species can themselves act as poisons for unsulfided noble metal catalysts, which several authors note are “readily poisoned” under these conditions. This dual behavior underlines why catalyst selection must be matched to the sulfur content of the specific feedstock rather than being treated as a universal choice.

Figure depicts the reversible nature of nitrogen/sulfur poisoning: basic N-heterocycles adsorb preferentially and more strongly on the active site than sulfur- or oxygen-containing species, temporarily blocking them and lowering the apparent hydrotreating rate. Because this adsorption is reversible, activity can be recovered under more severe conditions rather than requiring catalyst replacement, consistent with the persistent lag between denitrogenation and deoxygenation/desulfurization described above and with the selective persistence of single-ring pyridine/indole-type species even after severe second-stage hydrotreating.

7.

Schematic mechanism of reversible active-site poisoning by nitrogen and sulfur-containing species: basic N-heterocycles adsorb preferentially and more strongly than S-compounds, temporarily excluding them; unlike coking, this loss of activity is reversible.

4.4. Comparative Catalyst Behavior: Conventional Sulfided Catalysts vs Noble Metals

Direct catalyst screening studies supported a consistent qualitative picture. Yang et al., comparing NiMo/Al₂O₃, CoMo/Al₂O₃, Pt/Al₂O₃, Pd/Al₂O₃, Mo₂C, and MoS₂ on identical animal carcass biocrude, found that CoMo/Al₂O₃ gave the best overall combination of deoxygenation (86.2%), denitrogenation (64.4%), and energy recovery, while noble metal catalysts (Pt/Al₂O₃, Pd/Al₂O₃) achieved lower upgraded biocrude yields despite good hydrogenation activity. Patel et al., screening Pt, Pd, and Ru on carbon and alumina supports against NiMo/Al₂O₃ for algal biocrude, found NiMo/Al₂O₃ gave the lowest char yield (1.5 wt %, versus up to 4.45 wt % for Pd/C) but capped denitrogenation at only 42.7% under neat, solvent-free batch conditions, a useful caution that batch, neat feed screening data should not be directly extrapolated to continuous, industrially relevant HDN performance.

Noble metal catalysts, while often exhibiting superior initial hydrogenation activity, show distinct and generally more severe deactivation signatures. Yana et al. demonstrated via XRD and TEM that Ru supported on zeolites (BEA, ZSM-5) undergoes reduction from RuO₂ to metallic Ru during reaction, with catalytic activity declining progressively over four consecutive reuse cycles due to active site blockage and reduced metal dispersion (confirmed by BET/chemisorption); the authors proposed that adsorption of phenolic species poisons the metallic active sites. Alumina and silica supported catalysts of all types are additionally susceptible to support degradation under hydrothermal-like conditions: several reviews note that γ-alumina can convert to boehmite in the presence of liquid water, causing metal oxidation and active site blockage, while sintering of the alumina support under subcritical/supercritical water conditions leads to a loss of surface area. ,

Taken together, evidence indicates that conventional sulfided NiMo/Al₂O₃ and CoMo/Al₂O₃ catalysts remain the most robust default choice for continuous, industrially relevant HTL biocrude hydrotreatment, consistent with their near universal use in the pilot and TRL scale studies reviewed in Section . Noble metal catalysts offer higher intrinsic hydrogenation activity and are valuable for mechanistic bench-scale studies, but their higher cost, greater sensitivity to sulfur/nitrogen poisoning, and support stability limitations currently restrict their use in continuous operation unless combined with guard beds or highly prestabilized feeds.

4.5. Regeneration Strategies and Catalyst Lifespan

In contrast to the relatively rich literature on deactivation mechanisms, dedicated studies on the regeneration of HTL-biocrude-specific hydrotreating catalysts remain scarce. Available catalyst lifespan data are summarized in Table and are limited almost entirely to the time of deactivation (or time to plugging) figures from single, non-regenerated catalyst charges: 102 h before plugging for Tetraselmis-derived biocrude versus 220 h of stable operation for cyanobacteria-derived biocrude; 192 h without coke formation or plugging for spent coffee-ground biocrude; up to 335.4 h for a two-stage continuous hydrotreater processing Spirulina biocrude, within which a repeated experiment after 150 h of additional time-on-stream showed a 45.9% reduction in gas yield and a 5.3 percentage point drop in denitrogenation efficiency, a subtle but measurable deactivation signature despite the absence of any pressure drop; and up to 1500 h of steady-state operation reported for wet-waste biocrudes using a guard bed/main bed catalyst configuration, consistent with the double bed strategy of dos Passos et al.

7. Summary of Catalyst Deactivation Mechanisms, Evidence, and Lifespan across HTL Biocrude Hydrotreating Studies.

Catalyst/support	Feedstock	Deactivation mechanism	Quantitative evidence	TOS/cycles	refs
NiMo/γ-Al2O3 (sulfided)	Spirulina, sewage sludge	Coking; partial desulfidation	Coke: 9.1% (Spirulina)/3.4% (sludge) single-stage 400 °C	Batch
NiMo/Al2O3 (fixed-bed pilot)	Forestry residue (Hydrofaction)	Coking/pressure drop	Severe coking within 10 h at high severity (100 bar, 360 °C); stable 330 h at 300–330 °C	330 h
NiMo/Al2O3	Miscanthus, Spirulina, sewage sludge	Coking, H2-starvation coking	Coke up to 14.1 wt C% (miscanthus, 4 MPa); 9.3 wt C% (algae, 400 °C/8 MPa); sludge most resistant (2.0–4.2%)	Batch
NiMo, CoMo, Pt, Pd/Al2O3; Mo2C, Mo2S	Animal carcass	Coking (feedstock-derived aromatics)	Char yield 2.45–7.60 wt %; CoMo/Al2O3 best overall	Batch, single pass
NiW/Al2O3 (sulfided)	C. sorokiniana/vulgaris	Coking + partial catalyst desulfidation	Elemental analysis: fresh – (used catalyst): C 0.1% – (7.9–8.0%); S 8.5% – (4.3–4.4%).	Batch, 5 h
CoMo + NiMo (double bed, continuous)	Cow manure, wheat straw, co-HTL blend	Differential coking by the catalyst bed	MCR of oil 26% to 5% after HDT; higher C buildup on NiMo bed for co-HTL feed	Not reported
Fe-HZSM-5 (pyrolytic upgrading)	Spirulina	Char/organic deposition (poisoning)	∼20% catalyst mass loss after 5 cycles.	5 reuse cycles
CoMo/Al2O3	Tetraselmis sp.	Fe (porphyrin) deposition, plugging	Catalyst plugging after 102 h; Fe confirmed on spent catalyst	102 h (failed)
CoMo/Al2O3	Cyanobacteria	Fe (porphyrin) deposition, no plugging	Fe deposition confirmed, but no plugging	219.8 h (stable)
NiWS/Al2O3	C. sorokiniana/vulgaris	Fe deposition (feedstock-dependent)	Spent-catalyst Fe: 0.63% (C. sorokiniana) vs 1.60% (C. vulgaris); correlates with HDN degree	Batch, 5 h	,,
NiMo/Al2O3 (two-stage continuous)	Spirulina	Progressive activity loss	Repeat run at +150 h TOS: gas yield decreases 45.9%, HDN slight decrease 5.3 %; no pressure drop over 335 h	335 h
Ru/BEA, Ru/ZSM-5	Model biocrude/guaiacol	RuO2-Ru reduction; active-site blockage	Activity loss over 4 cycles; recovered by air calcination, 500 °C/200 min	4 reuse cycles
NiMo/Al2O3 (TRL 4/5 continuous)	TAGs-type feed	None detected (best case)	No pressure drop or coke after 33 days; catalyst life monitored via product S content	792 h

Only one regeneration protocol with demonstrated success was identified among the reviewed literature. Yana et al. restored the activity of a spent, 4-cycle runs deactivated Ru/zeolite catalyst by calcination in air at 500 °C for 200 min. A related, but not directly analogous, precedent comes from catalytic pyrolytic upgrading (rather than H₂-based hydrotreating) of Spirulina biocrude over Fe/HZSM-5, where the catalyst was reused over 5 cycles with a gradual decline in product quality and approximately 20% cumulative catalyst mass loss, while calcination-based regeneration was proposed but not experimentally verified. Silva et al. similarly note emerging, but as yet unproven for HTL-biocrude catalysts specifically, spent catalyst recovery concepts such as solvent-based extraction combined with ultrasonic-assisted hydrothermal treatment, and selective graphitization of the alumina support to restore surface area.

It can be concluded that no dedicated regeneration study currently exists for sulfided NiMo/Al₂O₃ or CoMo/Al₂O₃ catalysts that dominate continuous HTL biocrude hydrotreatment and that catalyst lifespan data remain confined to single charge time-on-stream figures rather than multicycle regeneration/reuse campaigns. This gap is consistent with broader calls in the literature for the development of durable, cost-effective, and readily regenerable catalysts tailored to HTL biocrude and is identified here as a priority direction for future research.

5. Comparative Analysis of Batch vs Continuous Hydrotreatment Systems

Batch (Section ) and continuous (Section ) hydrotreating systems differ fundamentally in how reactants contact the catalyst over time. In batch/microbatch reactors, a fixed H₂ charge (typically 5–8 MPa initial pressure) and a fixed catalyst/oil ratio are loaded at the start of the run and left to react for a set holding time (commonly 1–4 h), so both H₂ partial pressure and reactant concentration decline as the reaction proceeds. , In continuous fixed-bed systems, by contrast, biocrude and H₂ are fed continuously at a defined WHSV/LHSV, so the catalyst experiences an approximately constant reactant concentration and H₂ partial pressure throughout the run, and the nominal contact time is set independently of total processing duration. ,

This difference in gas–liquid–solid contacting is generally understood to give continuous fixed-bed or trickle-bed reactors better mass transfer than static batch/microbatch vessels, since flow-induced turbulence continuously renews the liquid film at the catalyst surface. Haider et al., working with microbatch reactors, explicitly invoke this point when discussing their own coke-formation results, noting that despite the superior mass transfer expected in a continuous fixed-bed reactor, comparably severe coke formation was still observed by Jensen under such conditions. Jensen also reported high exothermicity, a sharp pressure drop, and catalyst deactivation from severe coking within the first 10 h of operation at 360 °C and 10 MPa H₂, and only achieved stable operation for 330 h once the first-stage temperature was reduced to the 300–330 °C range. On this basis, Haider et al. concluded that, although mass transfer can play some role, the intrinsic thermal instability/coking tendency of HTL biocrude at high temperature is the determinant feature, largely independent of reactor mode. These observations indicate that better mass transfer in continuous operation does not, by itself, guarantee kinetic stability, and comparable coking-related deactivation risks can appear in both batch and continuous configurations once the thermal stability limit of the biocrude is exceeded.

Batch and continuous studies do not just report different deactivation numbers; they measure fundamentally different quantities. In batch/microbatch reactors, each run starts with fresh catalyst, so “deactivation” can only be captured as a single-pass coke yield at a given severity condition, not as a rate over accumulated operating time. Haider et al. illustrate this directly: a single-stage run at 400 °C (4 h, fresh NiMo/γ-Al₂O₃) produced 9.1 wt % coke for Spirulina biocrude and 3.4 wt % for sewage sludge biocrude, whereas splitting the same overall severity into a low-temperature stabilization stage (310–350 °C) followed by a second stage at 400 °C reduced coke make to 0.014–0.05 wt %, a reduction of roughly 98–99.9% depending on feedstock and first-stage temperature. This is a coking-propensity snapshot useful for screening conditions, but it says nothing about how activity evolves over sustained operation, since the catalyst never accumulates time-on-stream.

Continuous operation is what makes catalyst-to-catalyst deactivation-rate comparisons possible in the first place. Bezergianni et al. evaluated NiMo and CoMo catalysts for HAGO/waste-cooking-oil (WCO) cohydroprocessing across three temperatures (330–370 °C); to isolate the effect of days-on-stream (DOS) from temperature, the deactivation-rate study itself was conducted isothermally at 350 °C, tracking HDS efficiency over 26 DOS as the same catalyst charge was progressively fed increasing WCO content (pure HAGO up to DOS 4, then 90/10 HAGO/WCO up to DOS ≈12, then 70/30 HAGO/WCO up to DOS 26). Normalized to their own initial activity, NiMo and CoMo declined similarly under the milder 90/10 feed (to ≈0.83 and ≈0.80 of initial HDS rate, respectively, by DOS 12); the two catalysts only diverged once WCO content increased to 70/30, with the normalized activity of NiMo essentially plateauing (≈0.81 at DOS 26) while that of CoMo continued to decline (≈0.73 at DOS 26), indicating that the deactivation-rate advantage of NiMo becomes apparent specifically as lipid severity increases, rather than being present from the start. This time-dependent behavior is consistent with the differing sensitivity of the two catalysts to feed composition at steady state: at a fixed 330 °C, the HDS efficiency of CoMo dropped sharply as WCO was introduced (95.00% for pure HAGO → 87.12% at 10% WCO → 80.93% at 30% WCO), whereas the HDS efficiency of NiMo at the same temperature remained stable or even slightly improved (89.15% → 92.06% → 91.89%) notably, CoMo actually outperforms NiMo on pure HAGO at every temperature tested, so the advantage of NiMo is specific to lipid co-feeding, not a general superiority. The authors attribute this to active-site character: the predominantly “brim-type” sites of NiMo tolerate the oxygenated WCO co-feed better, whereas CoMo relies on vacancy-type sites that readily adsorb the O-containing compounds introduced by coprocessing.

Taken together, batch experiments bound the coking propensity of a given feedstock/condition in a single pass, while continuous, DOS-tracked operation is what actually reveals which catalyst will hold up over realistic operating campaigns and can distinguish between catalysts (or feedstocks) that look similar in a short batch screening but diverge sharply once time-on-stream accumulates.

Batch and continuous systems answer different questions when it comes to “product quality stability.” In batch/microbatch studies, each condition is typically run in duplicate with fresh catalyst, so the only stability metric available is run-to-run reproducibility at fixed conditions. Haider et al.,for instance, report their two-stage hydrotreating results as mean ± standard deviation from duplicate runs (oil yield 86.5 ± 0.8% and coke 0.04 ± 0.02% for sewage sludge biocrude, first stage 330 °C → second stage 400 °C), confirming the experiment is repeatable but saying nothing about whether that same product quality would be sustained across an extended production campaign, since every batch run starts from an identical, fresh catalyst bed.

Continuous operation, by contrast, allows product quality to be tracked and actively maintained as the catalyst ages over an extended campaign. Dimitriadis and Bezergianni illustrate this directly: across a continuous coprocessing campaign of HTL biocrude with light cycle oil spanning 49 DOS (alternating neat LCO baselines with 10% and 20 vol % biocrude blends), they used a temperature-compensation strategy, raising the reactor temperature as needed to counteract catalyst deactivation, so that product density stayed within 0.891–0.908 g/mL and product sulfur within 85.3–141.1 wppm throughout, despite the underlying catalyst activity declining over the same period. Carbon, hydrogen, and oxygen content of the product likewise remained essentially constant (C 87.6–88.0 wt %, H 11.9–12.4 wt %, O ≤0.07 wt %) regardless of DOS or blend ratio.

The key distinction is this: batch experiments can only confirm that a given condition is reproducible when repeated from scratch, while continuous operation is what actually demonstrates and enables active control over product quality stability as an operating campaign unfolds. This capability is directly linked to the deactivation behavior discussed above: because continuous systems reveal deactivation as it happens, operators can compensate for it (via temperature, in this case) to deliver a stable product specification, an option that is simply not available in single-pass batch operations.

The literature reviewed in Sections and 3.2 shows a clear asymmetry: essentially every demonstration of scale-up beyond the laboratory uses continuous fixed-bed hydrotreating, never batch. Dimitriadis et al. illustrate why this is a viable path: optimal conditions (330 °C, 6.89 MPa, LHSV 1 h^–1) identified in a small continuous fixed-bed unit (TRL 3, 15.8 mm I.D. tubular reactor and 704 mm in length) were transferred directly, without re-optimization, to a larger two-reactor continuous unit (TRL 5, 27.94 mm I.D. and 954 mm in length, 15 L/day capacity). The TRL 5 campaign ran continuously for 30 days with no pressure-drop buildup, confirming that the same catalyst and kinetics held up at the larger scale. This “dimensional scale-up” works because WHSV/LHSV-based design keeps the per-catalyst-volume contact time constant regardless of absolute throughput, the same principle underlying the continuous studies discussed above.

Scale-up is not automatically smooth, however: Dimitriadis et al. report that when the same fixed-bed hydrotreatment step was fed with pyrolysis bio-oil produced at TRL 5 (industrial-scale pyrolysis) instead of TRL 4 (bench-scale pyrolysis), a fast pressure buildup appeared that had not been observed at the smaller scale, traced not to the hydrotreating reactor itself, but to higher solids and carbonyl content in the TRL 5 produced feedstock. This shows that engineering feasibility at scale depends on the whole process chain, including how upstream feedstock quality (and its variability) shifts when the preceding step is itself scaled up and not just on the scalability of the hydrotreating reactor itself.

No equivalent industrial-scale demonstration of batch HTL-biocrude hydrotreating appears in the reviewed literature, which is itself informative: unlike continuous fixed-bed reactors, batch/microbatch vessels do not scale by simply enlarging the same geometry because mixing intensity and heat-transfer surface-to-volume ratio both decline unfavorably as vessel volume increases, a well-established limitation in reactor engineering that requires redesigning agitation and cooling systems rather than transferring conditions directly, unlike what was shown for the continuous case above. This is consistent with the fact that every scale-up pathway documented for HTL biocrude and related bio-based feedstocks in this review relies on continuous, not batch, technology.

6. Single vs Two-Stage Hydrotreatment of HTL Biocrude

A single-stage severe hydrotreating process for the production of drop-in, bio-, or advanced fuels is illustrated in Figure . In this approach, HTL biocrude is upgraded through a single severe (high temperature 300–400 °C and high pressure >8 MPa operation) catalytic hydrotreating step, in which oxygen is removed into a second aqueous phase, and nitrogen and sulfur species are removed via HDN and HDS reactions. Additionally, cracking and isomerization reactions also occur. The resulting organic phase, which constitutes the main upgraded biocrude product, is a drop-in fuel (gasoline, jet, or diesel boiling range hydrocarbons) that can be blended with the respective fossil fuel fractions at various ratios to produce conventional market fuels containing a small bio-content. Severe catalytic hydrotreating has been shown to effectively upgrade and improve the properties of HTL biocrude. As illustrated in Figure , HTL biocrude is fed into a single severe hydrotreating stage, yielding two streams: a drop-in biofuel that is blended with fossil fuel from a conventional refinery to produce market fuel, and a heteroatom-rich aqueous phase removed as a byproduct. However, significant challenges remain, the most critical being rapid catalyst deactivation due to coking. Under more severe operating conditions, competition between hydrogenation and coking reactions intensifies, leading to rapid catalyst deactivation, lower hydrocarbon yields, coke formation, and, ultimately, reactor plugging.

8.

One-stage severe hydrotreatment for drop-in fuels.

To address these challenges, some studies have proposed a two-stage hydrotreating approach. The first stage involves mild hydrotreatment with lower-severity hydrogenation (low operation temperature <300 °C) to reduce oxygenated compounds and remove most nitrogenous species, yielding a product with lower acidity. This is followed by a second, more severe hydrotreating stage (high operation temperature >300 °C) to achieve fuel-range heteroatom levels, thereby extending downstream catalyst lifetimes. In this scenario, the resulting upgraded biocrude remains a drop-in fuel, miscible with fossil fuels, and can be blended at various ratios to produce market fuels with a small bio-content, as illustrated in Figure . The aim of this section is to summarize and present the literature on two-stage hydrotreating processes.

9.

Two-stage hydrotreatment for drop-in fuels.

As shown in Figure , HTL biocrude first undergoes a mild hydrotreating stage, which removes oxygenated compounds via the aqueous phase, followed by a second, more severe hydrotreating stage that removes remaining heteroatoms; the resulting drop-in biofuel is then blended with fossil fuel from a conventional refinery to produce market fuel.

Two-stage hydrotreatment of a Spirulina-derived biocrude was investigated by Haider et al., who utilized a Mo/γ-Al₂O₃ catalyst in the top zone of the reactor (25% of reactor length) and a NiMo/γ-Al₂O₃ catalyst in the remaining 75% of the reactor length in a continuous operation pilot plant. They studied the effect of reaction temperature over a range of 200–340 °C at a constant pressure of 10 MPa and a constant WHSV of 0.5 h^‑1. Complete demetallization was achieved, reaching 98% at 300 °C, and nearly complete deoxygenation was observed at 340 °C. These staged temperature thresholds are examined further in Section . The same group also investigated catalytic hydrotreating of a nitrogen-rich biocrude derived from Spirulina microalgae via HTL in a dual tubular pilot-scale reactor operated in downflow mode. The system employed multiple catalytic beds: a Mo/Al₂O₃ catalyst (Catalyst A) for hydrodemetallization in the first stage, followed by two NiMo/Al₂O₃ catalysts (Catalysts B and C, with varying metal loadings) for heteroatom removal in the second stage. Key operating conditions included temperatures of 200–400 °C, pressures of 8–12 MPa, and weight hourly space velocities (WHSV) of 0.15–0.50 h^–1, with a constant H₂-to-biocrude ratio of 1000 NL/L. Upgraded products exhibited significant nitrogen removal (up to 99%) under severe conditions (400 °C, low WHSV), but at the cost of high hydrogen consumption, which led to a shift in the boiling point distribution. Diesel-range hydrocarbons (C14–C20) dominated at mild conditions (340 °C), while gasoline-range products (C5–C12) increased at higher severities (>340 °C operation temperature) due to fragmentation of nitrogen-containing aromatic structures, as revealed by FT-ICR MS. GC-MS analysis showed progressive conversion of fatty amides to n-paraffins, followed by cyclo- and iso-paraffins at elevated temperatures. The aqueous phase contained ammonia (up to 295 g/L N), indicating HDN. The single-stage nitrogen-removal ceiling for this system, and how a second stage overcomes it, is examined quantitatively in Section . These findings highlight that the hydrotreatment of algal biocrude is strongly influenced by its high nitrogen content, necessitating severe reaction conditions, particularly elevated temperatures (>350 °C) and low space velocities, to achieve effective denitrogenation. Under such conditions, nitrogen removal is accompanied by extensive hydrogenation and cracking, converting heavy fractions (boiling point >350 °C) into gasoline-range hydrocarbons while preserving diesel-range yields. This phenomenon is attributed to the fragmentation of nitrogen-containing aromatic structures inherent to algal biocrude. The findings also emphasize the importance of tailoring the hydroprocessing strategy based on the desired product slate. For middle distillate production (diesel/jet fuel), a two-stage approach with intermediate fractionation is recommended: initial hydrotreatment under milder conditions to maximize diesel yields, followed by selective upgrading of the middle-distillate fraction to minimize residual heteroatoms. This strategy optimizes hydrogen consumption and avoids excessive cracking of heavy fractions, which could otherwise be used for lower-grade applications (e.g., bunker fuel). By reducing unnecessary processing severity, this approach enhances the economic feasibility of HTL-derived biofuels, positioning them as competitive, sustainable alternatives to conventional fossil fuels.

Haider et al. explored two-stage catalytic hydrotreatment of two HTL biocrudes derived from sewage sludge and Spirulina microalgae. Their results showed that the first hydrotreating stage at 350 °C achieved complete deoxygenation with limited coke formation using a NiMo/γ-Al₂O₃ catalyst in a 25 mL microbatch reactor at 8 MPa and a reaction time of 4 h (mimicking a continuous WHSV of 0.5 h^‑1) with a biocrude-to-catalyst ratio of 4 g:2 g. The upgraded oil, after aqueous phase removal, was then used as feedstock for the second hydrotreatment stage at 400°C. The second stage increased hydrodenitrogenation to 92% and further reduced coke formation compared to direct one-stage hydrotreatment. Their general conclusion was that two-stage hydrotreating is particularly effective for biocrudes with high nitrogen content.

Similar results were reported by Zhao et al., who investigated two-stage hydrotreatment of an HTL biocrude from Nannochloropsis oceanica. Experiments were conducted in a 500 mL stainless steel autoclave using a NiMo/γ-Al₂O₃ catalyst. The product from the first stage served as feedstock for the second stage under similar operating conditions. They observed an increase in hydrogen content in the final product after two stages compared to one-stage hydrotreatment. Furthermore, the heating value was higher for the two-stage process. Oxygen was primarily removed during the first stage, while nitrogen species were removed in the second stage. Additionally, the second stage increased the kerosene fraction (C₈–C₁₆) from 65.5% (one stage) to 79.0% (two stages).

The same group also investigated the use of a guard catalyst to protect the core catalyst. Experiments were conducted at 350 °C and 6 MPa with a biocrude-to-catalyst ratio of 8 g:2.4 g. The core catalyst was NiMo/γ-Al₂O₃, while the guard catalyst was RG-20B (QiMao Catalysts Co., Ltd., Shandong, China). This approach yielded high-quality biofuels with nitrogen content below detectable limits and an HHV of 45.66 MJ/kg.

Finally, Haghihat et al. investigated hydrotreatment of an HTL biocrude derived from forestry residues using a continuous unit with a 730 mL reactor and a biocrude flow of 8–20 mL/h. They compared one-stage and two-stage hydrotreating under various operating conditions. They found that single-stage hydrotreatment with a CoMo catalyst, although improving biocrude properties, was insufficient for complete upgrading. Therefore, the organic phase from the first stage was fed into a second hydrotreating stage using a NiMo catalyst. The upgraded biocrude exhibited improved properties, with negligible TAN, lower oxygen content, and viscosity below 10 cP. More than 65% of the upgraded biocrude was distilled below 350 °C, while only 16% had a boiling point above 454 °C. Reaction temperature had the strongest effect on product yields and quality compared to WHSV, pressure, and H₂/oil ratio. This ranking of operating variables is discussed further in Section . A summary of all published studies focused on two-stage hydrotreating is presented in Table .

8. Summary for Two-Stage Hydrotreating.

Biocrude origin	System	catalysts	Conditions	Remarks	refs
Forestry residue	Continuous	1st step: CoMo/gAl2O3 2nd step: NiMo/γ-Al2O3	320–370°C, 80–95 bar, biocrude flow 0.143–0.358 ml/min, H2 flow 186–465 mL/min	Product with negligible TAN, low oxygen content, and viscosity below 10 CP
N. oceanica	autoclave	NiMo/γ-Al2O3	350 °C, 4 MPa, 2 h, biocrude/cat. 10:0, 10:1, 10:3, 10:5, 10:7	1.9 % nitrogen content, 44.5 MJ/kg HHV, kerosene yields 79%
N. oceanica	autoclave	NiMo/γ-Al2O3	350 °C, 6 MPa, 1 biocrude 8 g and cat. 2.4 g,	N content below detection limits, HHV 46 MJ/kg
Microalgae (Spirulina)	Continuous	1st step: Mo/gAl2O3 2nd step: NiMo/γ-Al2O3	T: 200–400 °C, P: 8–12 MPa; WHSV: 0.15–0.50 h–1; H2/Oil: 1000 NL/L	88.7–100 wt %
Microalgae (Spirulina)	Continuous	1st step: Mo/gAl2O3 2nd step: NiMo/γ-Al2O3	T: 200–340 °C, P: 100 bar; WHSV: 0.50 h–1;	98% Demetallization at 300 °C and complete deoxygenation at 340 °C
Sewage sludge and Spirulina	autoclave	NiMo/γ-Al2O3	350 °C, 8 MPa, 4 h time, 4 g biocrude + 2 g Catalyst	92 % denitrogenation rate

7. Coprocessing Mild Upgraded Biocrude with Petroleum Streams

Over the past several years, hydroprocessing technology has been investigated in both batch and continuous operation units for the direct upgrading of HTL biocrude to advanced fuels. However, this technology entails high investment costs for the new refinery infrastructure. Coprocessing HTL biocrude in existing refineries is one option for refining biocrudes, simultaneously adding renewable content to end-fuel products. By coprocessing a partially stabilized HTL biocrude with petroleum streams within existing refining infrastructure, the need for costly new standalone hydroprocessing facilities can be avoided. This strategy enables the production of low-carbon hybrid fuels (partly bio-based and partly fossil-derived), enhances sustainability, and supports progressive decarbonization in the transportation and aviation sectors through integration of biomass into refineries. The technological pathway is illustrated in Figure . HTL biocrude can be introduced at multiple points within a refinery, with each leading to different final products. Potential insertion options include feeding it into specific processing units after mild upgrading, integrating it into the system before the predistillation phase without treatment, or using it as a drop-in fuel following more extensive upgrading. Among these, the drop-in fuel route poses the least risk to refinery operations, while introduction at the predistillation stage carries the highest risk. The most widely adopted approach, similar to strategies used for pyrolysis bio-oil, is the integration of a stabilized biocrude into designated processing units. Although coprocessing appears to be an attractive strategy, several challenges remain, such as the impact of biocrude on commercial hydroprocessing catalyst performance, the chemical instability of biocrude, and its poor miscibility with petroleum-derived streams. As a result, a first mild upgrading step is essential prior to integrating biocrude into the existing refinery processes.

10.

Coprocessing mild upgraded biocrude with fossil streams for hybrid fuels production.

As depicted in Figure , the HTL biocrude undergoes a single mild hydrotreating step, producing a mildly upgraded biocrude (with aqueous phase removed as a byproduct) that is coprocessed within a conventional refinery alongside fossil streams, yielding a hybrid market fuel.

Hydrodeoxygenation can be used as a first step of mild upgrading to remove oxygenated compounds and render biocrude miscible with fossil streams. However, the optimum level of deoxygenation prior to coprocessing is not well-established. In this context, Badoga et al. investigated the degree of hydrotreatment required to upgrade an HTL biocrude derived from a 50/50 mix of spruce and pine wood, so that it could be miscible with vacuum gas oil (VGO) at 10 wt % for coprocessing within a refinery. They found that mild hydrotreating at 320 °C, 9.6 MPa, a reaction time of 9 h, and a catalyst-to-feed ratio of 0.06–0.11 g/g could yield a product with an oxygen content close to 2 wt %, enabling homogeneous blends with VGO. However, when the oxygen content exceeded 3 wt %, blends began to form deposits on bottle walls after 7 days of storage, indicating instability over time. , These experiments were performed in a batch reactor, and thus, hydrogen requirements for reducing oxygen content to ∼3.0 wt % could not be quantified. Furthermore, the study focused solely on mild hydrotreatment and the miscibility of upgraded biocrude with VGO, without evaluating cohydroprocessing of the blends.

Currently, literature on the coprocessing of biocrude in a refinery is limited to a few studies. Badoga et al. tested coprocessing of raw HTL biocrude, without any pre-stabilization step, with VGO in a continuous operation hydrotreating unit. They implemented a two-step process. In the first step, a blend of 92.5 % v/v VGO and 7.5 % v/v biocrude was hydrotreated over a commercial NiMo/Al₂O₃ catalyst to remove heteroatoms (sulfur and nitrogen). The product from this step was then passed through a commercial bifunctional hydrocracking catalyst in the second step. They observed that the addition of biocrude to VGO required at least an 8 °C higher process temperature to achieve the same conversion as with pure VGO, attributed to the high oxygen content of the biocrude that remained after the first hydrotreatment step. In a follow-up study, they introduced an initial hydrodeoxygenation step, reducing the biocrude oxygen content to 3.6 wt %. After this stabilization step, the low-oxygen biocrude became fully miscible with the VGO fraction (92.5 wt % VGO / 7.5 wt % biocrude). The coprocessing procedure then followed the same two-step sequence as their initial study, consisting of a hydrotreating step for heteroatom removal followed by a hydrocracking step. However, they noted that reactor temperature had to be increased by 10–15 °C compared to the baseline scenario with pure VGO, to counteract the poisoning effect of oxygen compounds. Reactor plugging was also observed during the hydrotreating step. To address this, an additional pretreatment step after mild hydrodeoxygenation was tested. Two approaches were examined: distillation and solvent treatment with toluene, aimed at removing heavier compounds that caused catalyst plugging during coprocessing. When coprocessing solvent-treated biocrude, the reactor temperature had to be increased by only 9 °C to achieve the same desulfurization levels as pure VGO. In contrast, coprocessing distilled biocrude achieved the same desulfurization levels without any need to raise the reactor temperature. No reactor plugging was observed after 650 h of continuous operation.

All of the studies above focused on HTL biocrude derived from wood biomass. However, biocrude from other feedstocks, such as algae, adds complexity to coprocessing due to high nitrogen content, which can poison hydrocracking and hydrotreating catalysts. In this context, Alvarez-Majmutov et al. examined coprocessing of algae-derived biocrude with VGO at different blending ratios (2.5, 5, and 10 % v/v) in a continuous hydrotreating pilot plant. They observed that adding biocrude to the VGO hydrotreater reduced the level of hydrodesulfurization, which could be compensated for by increasing the reactor temperature. No significant catalyst deactivation attributable to the biocrude content was observed.

Until now, petroleum feedstock in coprocessing studies has primarily been VGO. Dimitriadis and Bezergianni extended this research by performing mild hydrotreating of an HTL biocrude derived from municipal solid waste and investigating cohydroprocessing with light cycle oil (LCO) at 10 and 20 v/v% ratios to produce hybrid fuels. Experiments were conducted in a continuous TRL 3 HDT plant by using a commercial NiMo/Al₂O₃ catalyst. They found that integrating partially upgraded biocrude into the LCO hydrotreater reduced hydrogen consumption and produced a higher heating value product compared to hydrotreating pure LCO. The only drawback observed was a temporary reduction in the catalyst hydrodesulfurization activity, which was mitigated by increasing the process temperature. Overall, they concluded that partially stabilized biocrude can be integrated into an LCO hydrotreatment plant to produce advanced hybrid fuels consisting of jet- and diesel-boiling-range hydrocarbons. Similar findings were reported by Hansen et al., who coprocessed a stabilized wood-derived biocrude with VGO at a ratio of 90 % VGO / 10 % biocrude. They observed that adding 10 % biocrude required more severe operating conditions to achieve the same conversion as pure VGO. Additionally, the coprocessed product contained more light naphtha-range hydrocarbons, whereas net VGO processing produced mainly middle distillates.

Coprocessing biocrude with fossil streams is a promising strategy for utilizing existing refinery infrastructure and reducing investment costs. However, direct use of raw HTL biocrude in an existing hydrotreating unit is not optimal, primarily due to miscibility issues with fossil streams, while its oxygen content can further cause operation instability and reactor plugging. One mitigation strategy is to use a guard catalyst at the top of the reactor, although this reduces the volume available for the main hydrotreatment and hydrocracking catalysts, thereby lowering the feed rates. Consequently, the most promising approach is to include a first mild hydrotreatment step to partially stabilize the biocrude by removing most of the oxygen content, rendering it fully miscible with fossil streams. Nevertheless, further research on the stabilization of biocrude is needed. Achieving stabilization through hydrodeoxygenation remains challenging, and studies on the reaction pathways involved are essential to improving our understanding of catalyst performance and preventing coke formation. In parallel, technoeconomic analyses are required to evaluate industrial-scale feasibility, and more research is needed to develop catalysts optimized for mild biocrude upgrading.

8. General Discussion

Hydrothermal liquefaction (HTL) is a promising thermochemical process that converts wet biomass into energy-dense biocrude in hot-compressed water (typically 250–374 °C, 10–25 MPa), avoiding energy-intensive drying and achieving good carbon recovery. However, raw HTL biocrude is generally unsuitable for direct use as a transportation fuel because it is chemically complex, containing oxygenates, nitrogenates, and sulfur species, acidic, and viscous, necessitating catalytic upgrading to meet specifications for drop-in use. HTL biocrude typically contains 10–20 wt % oxygen in the form of phenols, ketones, acids, and other heteroatom-containing compounds. These functionalities contribute to poor storage stability, corrosiveness, and low heating value. Hydrotreating remains the most mature route to remove heteroatoms and stabilize HTL biocrude. Recent studies reaffirm hydrotreating as the workhorse step for integrating HTL biocrude into refinery practice. Dominant reaction pathways include (HDO → H₂O), (HDN → NH₃), and (HDS → H₂S), accompanied by hydrogenation/saturation and limited cracking to reduce viscosity and improve distillability.

Conventional hydrotreating catalysts, such as sulfided NiMo or CoMo supported on alumina, are widely used due to refinery compatibility. However, the high heteroatom and water content of HTL biocrude can deactivate these catalysts, prompting research into more robust alternatives (e.g., noble metals such as Ru/C, Pt/Al₂O₃, or tailored bifunctional catalysts), with reaction conditions typically ranging from 300–450 °C and 5–20 MPa hydrogen pressure, depending on the desired fuel specifications.

A two-stage approach (mild stabilization followed by deep hydrotreating) is often favored to mitigate rapid coking and catalyst deactivation. Laboratory and pilot-scale runs demonstrate that staged hydrotreating can reduce oxygen to near-zero and substantially increase higher heating value (HHV), producing refinery-compatible intermediates or blendstocks approaching fuel specifications. For example, a 2024 stabilization–hydrotreatment study reported products with near-zero oxygen content and HHV ≈ 46 MJ kg^–1, while also documenting rapid catalyst deactivation, underscoring the need for more robust, water- and nitrogen-tolerant catalysts and optimized stabilization protocols. An increasingly favored deployment path is to centralize hydrotreating at existing refineries rather than at distributed HTL sites, leveraging established hydrogen infrastructure, hydrotreaters, and logistics. A 2025 PNNL report, explicitly recommends transporting untreated HTL biocrude to refineries for hydrotreating/coprocessing as a reliability- and cost-improving modification to baseline designs, with allied studies demonstrating reaction conditions that can deoxygenate forest-residue HTL biocrude to refinery-ready levels.

Despite its effectiveness, hydrotreating HTL biocrude faces several challenges, including catalyst deactivation, high hydrogen demands, phase separation, and process integration. Catalyst deactivation due to coking, metal deposition, and particularly nitrogen species poisoning remains a primary hurdle. A two-stage hydrotreating strategy is suggested to address catalyst deactivation: a first mild hydrotreating step with lower-severity hydrogenation reduces oxygenated compounds and removes most nitrogen species, lowering acidity, followed by a second, more severe hydrotreating step to reach fuel-range heteroatom levels and extend downstream catalyst lifetimes.

Another suggested approach is mild biocrude upgrading via hydrotreatment, producing refinery-intermediate feeds that are miscible with petroleum-based feedstocks and can be cohydroprocessed to generate hybrid fuels. , This strategy reduces the investment cost of standalone units by leveraging the existing refinery infrastructure. However, further research is needed on biocrude stabilization, and studies on the reaction pathways involved are essential to improving our understanding of their impact on catalyst performance and preventing degradation and coke formation.

Beyond the reaction-engineering aspects discussed above, the practical deployment of HTL biocrude hydrotreatment is ultimately governed by its hydrogen and energy demand, which existing technoeconomic (TEA) and life cycle assessment (LCA) studies consistently identify as a dominant cost and environmental burden, though its relative weight varies with upgrading severity and feedstock. Tzanetis et al. reported a hydrogen requirement of 0.050 kg H₂/kg biocrude (25 mol H₂/kg biocrude) for combined deoxygenation and saturation, with a resulting biojet fuel cost of 1086 EUR/t (25.1 EUR/GJ) and life-cycle GHG emissions of 567 kgCO₂-eq./t (13.1 kgCO₂-eq./GJ), an 85% reduction relative to fossil jet fuel. Jiang et al. reported a closely comparable hydrogen consumption of 0.0467 kgH₂/kg wet biocrude, a fuel blendstock cost of 700-1059 EUR/t (2.28 to 3.45 USD/gge, converted at 1 EUR = 1.157 USD), and GHG emissions of 722–1053 kgCO₂-eq./t (16.6–24.2 gCO₂/MJ), a 73.4 to 81.8 % reduction versus petroleum diesel (91 gCO₂e/MJ). Karka et al., modeling a centralized upgrading plant fed by 10 decentralized sewage-sludge HTL units, derived an equivalent hydrogen consumption of 0.051 kg H₂/kg biocrude (7.1 million kg H₂/year for 140.0 million kg biocrude feed/year), a production cost of 675-900 EUR/t (60–80 EUR/MWh-product), and GHG emissions of 300-2500 kgCO₂-eq./t (0.3–2.5 kgCO₂-eq./kg biofuel blend), corresponding to a 35–90% reduction versus fossil diesel. These three independent studies converge closely on hydrogen consumption (0.047–0.051 kg H₂/kg biocrude) for deep hydrotreating toward diesel/jet-range fuels.

By contrast, Li et al., evaluating marine fuel, renewable diesel, and sustainable-aviation-fuel (SAF) pathways from sludge, manure, and food-waste biocrudes, reported a substantially lower hydrogen consumption of 0.017–0.020 kg H₂/kg feed specifically for the mild-hydrotreating marine-fuel scenario (diesel/SAF-specific consumption was not separately quantified), together with a fuel cost of 792–1108 EUR/t (2.58–3.32 USD/GGE, converted at 1 EUR = 1.157 USD, after feedstock disposal credits) and GHG emissions of −2523 to +392 kgCO₂-eq./t (−58 to +9 gCO₂e/MJ), with manure- and food-waste-derived fuels achieving negative net emissions from avoided methane credits. Notably, their sensitivity analysis found hydrogen consumption to have the most negligible impact on minimum fuel selling price (MFSP) among the tested variables, with HTL plant capital cost, scale, and feedstock-disposal credits emerging as the dominant cost drivers instead, indicating that hydrogen demand scales strongly with upgrading severity and target fuel specification and that its weight on overall cost depends on which other cost components dominate in a given plant configuration.

For the mild-upgrading/coprocessing route, Del Alamo et al. reported a standalone-unit fuel cost of 935–2814 EUR/t (21.5–64.7 EUR₂₀₂₁/GJ) for a centralized refinery diesel/naphtha product across a biocrude feed capacity range of 8.8–88 t/day, identifying biocrude feedstock price, make-up hydrogen production cost, and annualized capital expenditure as the three parameters with the greatest influence on the MFSP; reusing existing refinery equipment instead of building dedicated units lowered the MFSP by only 7%. Their hydrogen consumption was reported on a volumetric basis (nm³ H₂/m³ feed) per reactor stage, and no full life-cycle GHG assessment was performed (only direct process emissions of 0.58 kg CO₂/kg dry sewage sludge feed are given). Badoga et al., by contrast, explicitly did not quantify hydrogen consumption for their batch-reactor mild-hydrotreating experiments, noting that reliable estimates require continuous pilot-scale testing–underscoring a persistent methodological gap between bench-scale mechanistic studies (Section ) and the pilot-scale data needed for robust TEA.

From an environmental standpoint, Zoppi et al. demonstrated that integrating HTL with aqueous-phase reforming can supply hydrogen internally–107% of the upgrading demand for the corn stover case, versus electrolytic make-up hydrogen for the lignin-rich-stream case at a biofuel cost of MFSP 1270 EUR/t (corn stover) and 1230 EUR/t (lignin-rich stream) (1.27 and 1.23 EUR/kg biofuel), with GHG emissions of 2412 kgCO₂-eq./t (corn stover, 56.1 gCO₂e/MJ) and 2511 kgCO₂-eq./t (lignin-rich stream, 58.4 gCO₂e/MJ), corresponding to a 37% reduction versus fossil diesel that rises to 80% when the lignin-rich-stream scenario is combined with renewable electricity for the electrolyzer. Taken together, these TEA/LCA studies confirm that hydrogen demand and capital intensity, rather than catalyst cost per se, are primary levers on both production cost and carbon footprint, directly reinforcing the relevance of the coke-suppression and catalyst-lifetime strategies discussed in Sections and 4.5: extending catalyst life and enabling milder operating severities reduce both hydrogen consumption and the frequency of costly catalyst replacement or regeneration.

9. Challenges and Prospects

Novel catalysts: beyond the conventional sulfided NiMo/CoMo and noble metal catalysts discussed above, sulfur-free alternatives such as transition metal phosphides and nitrides are emerging as promising candidates for HTL biocrude hydrotreating. These materials combine metallic and acidic active sites and avoid the need for continuous sulfiding agents, which is relevant given the variable, often low native sulfur content of many HTL feedstocks. Their performance under water and nitrogen-rich environments characteristic of HTL biocrude remains to be systematically evaluated.

Green hydrogen coupling: hydrogen demand remains a major cost and emission driver for HTL biocrude hydrotreating, as discussed above. Coupling hydrotreating with green hydrogen, produced via renewable electrolysis or generated in situ through aqueous-phase reforming of the HTL aqueous coproduct, offers a pathway to reduce both carbon footprint and logistical burden of external H₂ supply and represents a promising direction for future process integration.

Process intensification: process intensification strategies could help address catalyst deactivation and mass-transfer limitations discussed in Section . Operation under supercritical conditions has been shown to improve hydrotreating of biocrude by bringing hydrogen and oil into a single fluid phase, overcoming conventional gas–liquid mass-transfer limitations. Similarly, membrane reactor technology, which combines reaction and separation in a single unit and has been successfully demonstrated for other renewable fuel production routes such as biodiesel, hydrogen and Fischer–Tropsch synthesis, offers a complementary process intensification strategy that has not yet been explored for HTL biocrude hydrotreating.

Policy and market access: market access for HTL-derived hydrotreated fuels is increasingly shaped by regulatory frameworks. In the European Union, the recast Renewable Energy Directive (RED III) and the ReFuelEU Aviation Regulation establish progressive blending mandates for sustainable aviation fuel. In the United States, the Section 45Z Clean Fuel Production Credit rewards fuel producers according to lifecycle carbon intensity, directly incentivizing low-carbon hydrotreating routes. These instruments create a market pull that could accelerate commercial deployment of HTL biocrude hydrotreatment, provided that technological challenges discussed above are resolved in parallel.

Despite the promising results obtained on a pilot scale, refinery integration of HTL biocrude-derived streams remains challenging. Particular attention should be paid to refinery compatibility issues, including alkali metal contamination, chloride-induced corrosion, catalyst poisoning, sulfur balance disturbances, and hydrogen network management, which may impose additional process constraints and increase the cost of industrial implementation. In conclusion, hydrotreating remains the most technologically mature route for upgrading HTL biocrude to drop-in fuels or refinery intermediate feeds. Nevertheless, challenges such as reducing hydrogen consumption, improving catalyst durability, and integrating renewable hydrogen sources remain and must be addressed to enable a large-scale deployment. If these challenges are overcome, coupled with advances in HTL feedstock flexibility and process efficiency, optimized hydrotreating could make HTL a cornerstone of sustainable biofuel production.

10. Conclusion

This review consolidates the state of the art in catalytic hydrotreatment for upgrading hydrothermal liquefaction (HTL) biocrude into advanced biofuels. The assessment of diverse feedstocks from lignocellulosic biomass and algae to various waste streams reveals that, despite their differing compositions, hydrotreatment is a universally critical step for heteroatom removal (O, N, S) and stability enhancement. Conventional sulfided catalysts (NiMo, CoMo) have proven to be effective, although challenges such as rapid deactivation due to coking, metal deposition, and high hydrogen consumption persist. To mitigate these issues, process strategies such as two-stage hydrotreating and coprocessing of mildly upgraded biocrude with conventional petroleum streams have emerged as promising approaches to enhance catalyst lifetime and enable integration into existing refinery infrastructure.

Looking forward, hydrotreatment remains the most technologically mature route to transform HTL biocrude into “drop-in” fuels. However, to achieve commercial viability, future research must prioritize the development of more robust, poison-tolerant catalysts, optimization of mild stabilization processes to reduce hydrogen demand, and comprehensive technoeconomic assessments. Overcoming these hurdles is essential to fully leverage HTL as a cornerstone of a sustainable and circular bioeconomy, effectively converting a wide range of biomass and waste resources into renewable transportation fuels.

Glossary

Abbreviations

CV

Chlorella vulgaris

CS

Chlorella sorokiniana

EU

European Union

HDN

hydrodenitrogenation

HDO

hydrodeoxygenation

HDS

hydrodesulfurization

HDT

hydrotreatment

HHV

high heating value

HTL

hydrothermal liquefaction

LCA

life cycle analysis

LCO

light cycle oil

LHSV

liquid hourly space velocity

MFSP

minimum fuel selling price

MSW

municipal solid waste

RED

Renewable Energy Directive

SAF

sustainable aviation fuel

TAN

total acid number

TEA

technoeconomic analysis

TON

time on stream

TRL

technology readiness level

ULSD

ultralow sulfur diesel

VGO

vacuum gas oil

WHSV

weight hourly space velocity

The manuscript was written through contributions of all authors. All authors have given approval to the final version of the manuscript. A.D.: writing–original draft, investigation, data curation, review, and editing. F.T.: writing–original draft, investigation, and editing. S.B.: writing–review and editing.

The open access publishing of this article is financially supported by HEAL-Link.

The authors declare no competing financial interest.