Sustainability Assessment of Bioethanol from Food Industry Lignocellulosic Wastes: A Life Cycle Perspective

Abstract

Second-generation bioethanol from food industry lignocellulosic residues offers a promising route toward low-carbon, circular bioenergy systems. However, the reported environmental impacts differ markedly across studies, challenging efforts to assess the true sustainability of these waste-derived bioethanol routes. This review synthesizes current knowledge on the production of bioethanol from key agro-industrial wastes including oil palm empty fruit bunches, sugarcane bagasse, brewers’ spent grain, spent coffee grounds, tea waste, citrus residues, and potato peel waste. We outline feedstock characteristics, availability, and prevailing management practices, and map the principal biochemical conversion routes to identify process steps that drive environmental performance. A systematic comparison of life cycle assessments reveals substantial methodological heterogeneity across functional units, system boundaries, allocation procedures, and impact assessment methods. Nonetheless, consistent hotspots emerge, particularly associated with pretreatment severity, enzyme production, thermal energy demand, and co-product handling. The review highlights robust cross-study trends, pinpoints methodological gaps, and proposes recommendations for harmonized LCA practice. By integrating technological and methodological perspectives, this work aims to support the development and policy uptake of sustainable, waste-based bioethanol within circular bioeconomies.

1. Introduction

The Sustainable Development Goals (SDGs) call for cleaner energy systems in the transportation sector. Therefore, the energy transition requires practical and feasible ways to replace fossil gasoline in the transportation sector. System outcomes depend on coupled process and material choices, not single-point performance claims [1]. Bioethanol remains one of the most established renewable liquid fuels for partial substitution of gasoline, yet its climate and resource benefits cannot be inferred from tailpipe emissions or from the renewable origin of carbon alone [2,3,4,5]. The environmental performance of ethanol is determined by upstream energy use, chemical inputs, and emissions across feedstock supply, conversion, and supporting utilities, which are distributed throughout the life cycle rather than concentrated at the point of use [6]. This system-level character is particularly consequential for lignocellulosic residue pathways, where pretreatment intensity, enzyme production, distillation heat, and wastewater management can dominate impacts [7,8]. Life cycle assessment (LCA) is therefore required to quantify net greenhouse gas and resource outcomes and to prevent burden shifting among climate, energy, water, and land dimensions when evaluating residue-derived ethanol as a decarbonization option [9,10].

Food industry lignocellulosic wastes and residues are carbohydrate-rich by-products generated continuously within established agro-food supply chains, which makes their supply relatively stable and geographically anchored [11,12,13]. Many streams are still managed through low-value or environmentally problematic routes, including uncontrolled dumping, open burning, or disposal pathways that generate methane and local pollution [14,15,16,17,18,19]. Unlike dedicated energy crops, these residues arise from primary production whose land, fertilizer, and processing infrastructures are already in place, which creates an opportunity to supply second-generation ethanol without expanding cultivated area [20]. Their sustainability role is conditional because they also carry a waste management function and often have existing uses, such as process fuel, animal feed, or soil amendments [14,15]. As a result, residue-to-ethanol LCAs depend strongly on the counterfactual baseline and on how displaced treatments and competing co-products are represented, which can shift net burdens and credits substantially [21,22,23,24,25,26].

Lignocellulosic residues are structured by cellulose and hemicellulose as fermentable carbohydrate sources, while lignin forms a protective matrix that limits enzymatic access [27,28]. Most residue-to-ethanol routes therefore rely on a biochemical sequence of pretreatment, enzymatic hydrolysis, fermentation, and distillation [29,30]. Pretreatment uses physical, chemical, or biological actions to disrupt or reconfigure the lignin–carbohydrate network, reduce recalcitrance, and moderate inhibitor formation. Enzymatic hydrolysis then cleaves glycosidic bonds to release fermentable sugars, including glucose from cellulose and pentoses from hemicellulose. Residues differ in moisture content, ash and mineral load, inhibitor profiles, and carbohydrate accessibility, which shifts pretreatment severity, enzyme demand, and attainable ethanol titer [31,32,33]. These steps also define key life cycle inventory drivers, because pretreatment steam, enzyme production, distillation heat, and wastewater management can dominate environmental impacts in lignocellulosic residue systems [34,35].

Although most systems follow a similar conversion train, their life cycle performance is governed by system configuration rather than by the biochemical steps in isolation [29]. Utilities often dominate, because steam for pretreatment and distillation and electricity for solids handling and liquid processing are major energy demands. Heat integration and the carbon intensity of supplied utilities therefore become primary determinants of impacts. The handling of lignin-rich solids and liquid effluents further defines outcomes: these streams can be combusted for process heat and power, anaerobically digested to biogas, upgraded to materials, or treated as wastes requiring energy- and chemical-intensive management [36,37]. Many food industry settings also offer integration options, including shared steam networks, cogeneration units, and co-product recovery lines, so ethanol may operate as a marginal add-on, a co-product within a biorefinery, or a reallocation of existing residue uses [38]. These configuration choices directly shape system boundaries and co-product credits, and they condition whether studies rely on allocation or system expansion when modelling multifunctionality.

Published LCAs of ethanol from food industry wastes report highly dispersed outcomes, ranging from low or even net-negative greenhouse gas results to values that approach or exceed those of conventional fuel comparators in some scenarios [39,40,41,42]. This dispersion is also observed within the same residue class under broadly similar biochemical routes, which indicates that differences cannot be interpreted as conversion effects alone [39]. A first driver is how the functional unit frames the question, either as a waste management service or as a transport fuel substitute [43,44,45]. A second driver is system boundary choice, including whether residue generation and collection, liquid effluent treatment, and use-phase processes are included. A third driver is multifunctionality modelling, because electricity export, biogas recovery, extracted oils, fertilizer products, and displaced conventional waste treatments can be represented through allocation or substitution assumptions, which can shift burdens and credits substantially [46]. Results further diverge through inventory specification, such as utility mixes, boiler efficiencies, enzyme production inventories, drying requirements, and transport logistics [47], and this reinforces the need for a transparent and robust harmonization workflow when synthesizing heterogeneous evidence [48]. Accordingly, this review consolidates peer-reviewed LCA studies within a defined search window and inclusion criteria, harmonizes results to a common functional unit of 1 kg anhydrous ethanol using explicit conversion rules and a boundary classification scheme, and identifies dominant hotspots and key “noise sources,” especially co-product handling and counterfactual baseline choices, to inform both research design and policy interpretation.

2. Literature Search and Data Harmonization Approach

This narrative review synthesizes peer-reviewed literature on LCA studies of bioethanol produced from food industry lignocellulosic residues. The literature search focused on studies published primarily in the last 10–15 years to capture recent advancements in the field. Searches were conducted across key databases, including Web of Science, Scopus, and ScienceDirect, using terms related to bioethanol production, food industry residues, and LCA.

3. Feedstock Characteristics and Conversion Pathways

3.1. Food Industry Lignocellulosic Wastes

Food industry lignocellulosic residues are generated across different processing chains, including oil, sugar, brewing, and fruit starch industries. As shown in Figure 1, these residues vary in their physicochemical properties such as recalcitrance, moisture content, and chemical composition, which directly influence their suitability for bioethanol production and affect their LCA outcomes. Differences in moisture content, cellulose and lignin content, and the presence of inhibitors significantly condition both the conversion design and interpretation of LCA results.

As shown in

Table 1. Overview of food industry lignocellulosic wastes considered for bioethanol production.

Feedstock	Food Industry Source/Process	Typical Availability (Scale & Region)	Typical Composition Range (%)	Key Physicochemical Features Relevant to Bioethanol	Current Disposal	Typical Pretreatment in Bioethanol Studies	Representative Studies
Oil palm empty fruit bunches (OPEFB)	Palm oil milling; sterilization and threshing residues (stalks after fruit removal)	≈80–100 Mt yr−1 globally; ≈21 Mt yr−1 in Malaysia (2018); ≈1.07 t OPEFB per t crude palm oil	Cellulose ≈50–58; hemicellulose ≈21–30; lignin ≈18–24; ash ≈3–5; extractives ≈3–5	Fibrous residue with high holocellulose and moderate lignin contents; very high moisture content (>60% wet basis), low bulk density and silica-rich fibres that are abrasive and prone to biological degradation during storage	Mainly combusted on site for process steam and power; also used as mulch and soil amendment; open dumping or simple composting still occurs at some mills	Dilute acid, alkaline, hydrothermal/steam explosion, organosolv and biological pretreatments, typically followed by enzymatic hydrolysis and ethanolic fermentation	[49,50,51,52,53,54,55]
Spent coffee grounds (SCG)	Instant coffee and brewed coffee production; solid residue after hot water extraction of roasted ground coffee	Global SCG estimated at ≈6–18 Mt yr−1 (wet basis), with ≈6 Mt yr−1 most frequently cited; ≈0.65 t SCG per t green coffee	Cellulose ≈20–30; hemicellulose ≈20–30; lignin ≈10–25; lipids ≈10–20; proteins ≈10–17; ash ≈1–5	Fine, particulate residue with high oil and protein contents, substantial volatile and phenolic fractions and relatively high energy density; particles tend to be hydrophobic and to agglomerate during handling	Predominantly landfilled or incinerated; minor use as low-grade solid fuel, compost, animal feed additive, adsorbent and filler in construction or polymer composites	Dilute acid or hydrothermal pretreatment, often combined with alkali-assisted delipidation and enzymatic hydrolysis; subsequent dark fermentation or ethanolic fermentation for fuel production	[56,57,58,59,60,61,62]
Sugarcane bagasse (SCB)	Sugar milling; fibrous residue after juice extraction from sugarcane stalks	≈279–700 Mt yr−1 SCB globally depending on estimation; sugarcane ≈1.6 Gt yr−1 with bagasse ≈30% of cane mass	Cellulose ≈32–50; hemicellulose ≈20–32; lignin ≈17–32; ash ≈1–9; minor extractives	Coarse, fibrous matrix with relatively high lignin and moderate ash contents; high moisture at mill outlet, good handling properties but prone to compaction in bales; ash contains silica and alkali metals that influence combustion and pretreatment	Mainly combusted on site for steam and electricity; surplus sometimes exported as fuel; smaller fractions used for pulp and paper, fibreboards and composite materials	Steam explosion, dilute acid, alkaline and organosolv pretreatments, followed by enzymatic hydrolysis and fermentation; widely studied in integrated first- and second-generation ethanol concepts	[63,64,65,66,67,68,69,70,71,72]
Brewers’ spent grain (BSG)	Brewing; insoluble residue of malted barley and adjuncts after mashing/lautering	≈30–40 Mt yr−1 BSG globally; ≈20 kg BSG per hL beer; 2020 regional production: Americas ≈12.3 Mt, Asia ≈11.0 Mt, Europe ≈10.0 Mt	Cellulose ≈15–25; non-cellulosic polysaccharides (mainly arabinoxylans) ≈20–30; lignin ≈20–30; proteins ≈15–30; lipids ≈5–10; ash ≈2–5	Moist, highly biodegradable fibrous residue with high protein and fibre contents; rapidly spoils under ambient conditions; sticky mash-like consistency but a suitable substrate for microbial growth and enzymatic saccharification	Predominantly used fresh or ensiled as cattle feed; also composted or digested anaerobically; emerging higher-value uses in food ingredients, biopolymers and adsorbents	Hydrothermal and dilute acid pretreatments, alkaline fractionation and enzymatic hydrolysis, followed by separate or simultaneous saccharification and fermentation for bioethanol and/or biogas	[73,74,75,76,77,78,79,80,81]
Tea waste	Tea beverage and instant tea production; spent tea leaves and factory tea dust after hot water extraction	Global tea consumption increased ≈2.1-fold between 1995 and 2015; spent tea leaf sludge accounts for ≈90% of total tea waste; total tea waste generation is in the multi-Mt yr−1 range	Cellulose ≈20–35; hemicellulose ≈15–30; water-soluble fraction (including phenolic, tannin and other extractives) ≈25–40; ash ≈3–7	Heterogeneous fine particles rich in polyphenols, caffeine and other extractives; relatively high ash and mineral contents; exhibits good sorption capacity and a lignocellulosic backbone that can be saccharified if phenolic inhibition is controlled	Typically landfilled or used as low-grade compost; more recently explored as feedstock for adsorbents, biochar, catalyst supports and biofuel production	Acid, alkaline or hydrothermal pretreatments followed by enzymatic hydrolysis and ethanolic fermentation; some studies investigate seawater-based fermentation systems	[82,83,84,85,86]
Citrus processing waste	Citrus juice, marmalade and essential oil production; peel, pulp and rag remaining after juice extraction and oil recovery	Global citrus production ≈124 Mt yr−1; industrial processing generates ≈50% waste, corresponding to tens of Mt yr−1 of peel–pomace, mainly in Brazil, China, the Mediterranean region and US citrus belts	Cellulose ≈8–55; hemicellulose ≈0.3–26; lignin ≈0.5–21.6; rich in pectin and soluble sugars; minor protein and lipid fractions	Residue with high pectin and soluble sugar contents, substantial essential oils (notably limonene) and polyphenols; relatively low lignin compared with many agro-residues; high moisture and rapid spoilage potential	Traditionally used as low-value cattle feed, soil amendment or compost; large fractions still discarded; increasingly examined in integrated biorefineries for pectin, essential oils, bioethanol and biogas	Dilute acid or hydrothermal pretreatment to solubilise hemicellulose and pectin, often coupled with detoxification or limonene removal, followed by enzymatic hydrolysis and ethanolic fermentation; some schemes integrate pectin/essential oil recovery with ethanol production	[87,88,89,90,91,92,93,94]
Potato peel waste (PPW)	Potato processing for chips, fries, mashed and dehydrated products; outer peel and cortical tissues removed during washing, peeling and trimming	Global potato production ≈370 Mt yr−1; industrial peeling generates ≈15–40% of incoming tuber mass as peel and trimmings, corresponding to several Mt yr−1 of PPW, especially in Europe, North America and East Asia	Starchy–lignocellulosic matrix: starch ≈20–40; cellulose ≈10–20; hemicellulose ≈5–15; lignin ≈10–20; proteins ≈8–15; lipids ≈1–5; ash ≈3–8	High-moisture, rapidly putrescible residue with high loads of starch, soluble sugars and phenolic compounds including glycoalkaloids; thin heterogeneous particles that are well suited to enzymatic saccharification once starch and structural carbohydrates are accessed	Commonly used as low-value animal feed or compost; significant quantities still discharged with solid waste or wastewater; increasingly considered as a biorefinery feedstock for polyphenols, bioethanol and biogas	Thermal, dilute acid, alkaline and thermo-chemical pretreatments, typically combined with amylolytic and cellulolytic enzymatic hydrolysis and yeast-based ethanolic fermentation; recent studies focus on multi-enzyme cocktails and process integration	[95,96,97,98,99,100,101,102]

, oil-processing residues are represented by oil palm empty fruit bunches, whereas sugar-processing residues are represented by sugarcane bagasse. Brewing and beverage residues include brewers’ spent grain, spent coffee grounds, and tea waste, arising from both centralized plants and distributed outlets. Fruit residues, including starch-rich streams such as potato peel waste, originate from juice, essential oil, and potato-processing facilities. Across these residues, annual generation spans orders of magnitude, and spatial concentration ranges from mill-scale point sources to urban collection networks, with direct implications for storage, logistics, plant integration, and the definition of collection and baseline assumptions in LCA.

3.1.1. Oil-Processing Residues

Oil palm empty fruit bunches (EFB) are generated at palm oil mills in large, site-concentrated flows and are often available beyond immediate on-site uses [103,104]. Their counterfactual fate varies across regions and mills, including unmanaged decay, field application, composting, and partial combustion [105]. This variability makes LCA outcomes for EFB-derived ethanol highly sensitive to baseline definitions, because avoided emissions from decay, displaced waste handling, and electricity substitution credits can outweigh differences caused by conversion yields [106].

EFB also imposes distinctive process burdens that propagate into environmental indicators. High moisture content, elevated ash and silica, and heterogeneous fibre morphology increase handling losses, drying demand, and pretreatment severity, which in turn raises steam demand and can increase enzyme requirements when cellulose accessibility is not achieved efficiently [107,108,109]. As shown in Figure 2, the multilayered cell wall structure and lumen of EFB fibres help explain why pretreatment rigour and enzymatic hydrolysis performance are often decisive in EFB ethanol production systems.

3.1.2. Sugar-Processing Residues

Sugarcane bagasse (SCB) is structurally embedded in sugar–ethanol complexes and commonly functions as the primary combined heat and power (CHP) fuel that supplies steam and enables electricity export [111,112]. Under this industrial baseline, bagasse is not a low value residue awaiting diversion but an energy carrier already allocated to mill utilities. Ethanol pathways that consume bagasse therefore reallocate an existing energy service. This can create steam and electricity deficits that must be compensated within the system boundary [113].

Reported LCA outcomes for bagasse-derived ethanol are dominated by how this reallocation is represented. As shown in Figure 3, process-integrated co-production layouts explicitly couple pretreatment-derived residues to CHP and electricity export, making results sensitive to boiler and turbine assumptions, heat recovery performance, and the choice of displaced electricity mix used for crediting [113]. Results also shift with the allocation or system expansion rules used to partition burdens between ethanol and co-products [114,115]. These methodological choices often outweigh differences among pretreatment options or fermentation configurations when the process train is embedded within an integrated mill.

3.1.3. Brewing Residues

Brewers’ spent grain (BSG), spent coffee grounds (SCG) and tea waste represent the main brewing- and beverage-linked residues considered for waste-derived ethanol, and their sustainability interpretation is conditioned by strong non-lignocellulosic contributions and established competing uses [81,116].

BSG is generated in large, relatively stable volumes during brewing, mainly after mashing and lautering, as shown in Figure 4. It is partially deconstructed during malting and mashing, yielding a lignocellulosic matrix enriched in proteins and arabinoxylans. This composition supports high utilization potential but often requires enzyme systems beyond cellulases and can constrain high-solids handling and downstream separation. It further illustrates that BSG is commonly situated within multi-path biorefinery configurations, where ethanol is one option alongside energy and material co-products. Because BSG is widely valorized as animal feed and, in some regions, as a substrate for anaerobic digestion, ethanol production typically displaces existing nutrient and energy services. LCA outcomes for BSG-derived ethanol therefore depend strongly on how substituted feed and displaced biogas or electricity are represented under allocation or system expansion, and these modelling choices can outweigh differences in conversion yields for climate and energy indicators.

SCG is shaped by roasting and hot water extraction, resulting in a fine, hydrophobic residue that retains structural carbohydrates but contains substantial lipids, proteins, and phenolic fractions. These properties increase drying and handling demands, motivate delipidation and/or detoxification in many schemes, and can suppress ethanol titers when inhibitory compounds are not adequately removed or tolerated [118,119]. As shown in Figure 5, SCG also has a broad application landscape across energy, chemicals, materials, and food-related products, which implies strong competing uses and multi-product configurations. Consistent with this, lipid-rich fractions and high volatile solids enable integrated oil extraction, direct combustion, or anaerobic digestion, so ethanol is often only one output within a wider valorization system [120,121]. Reported SCG-to-ethanol pathways therefore range from energy-intensive drying and acid hydrolysis of wet grounds to liquid hot water pretreatment at high solids-to-liquid ratios with co-production of biodiesel and heat, and LCA variability is closely linked to whether oil recovery, energy co-products, and wet waste management are included within the system boundary and how associated credits are assigned [122].

Tea waste is introduced characterized by a lignocellulosic scaffold coupled to high levels of polyphenols, tannins, caffeine and mineral ash. These constituents can inhibit enzymatic saccharification and fermentation and make ethanol performance sensitive to extraction history and detoxification strategy, even when structural carbohydrates are present at moderate levels. Figure 6 shows that tea residue can be used for various value-added utilization pathways [123]. However, the standardization of its resource utilization is lower than that of brewer’s grains, so the counterfactual baseline may vary greatly in different regions. In LCA terms, the comparative attractiveness of tea waste ethanol is therefore sensitive to local waste management practice and to whether alternative valorization routes, such as biochar or sorbent production, are represented as competing or co-produced functions.

3.1.4. Fruit and Starch Processing Residues

Citrus processing waste is characterized by high proportions of soluble sugars and pectin-rich cell wall material together with terpenoid compounds, notably limonene [124,125]. As shown in Figure 7, citrus by-products also contain extractable phytochemicals and support multiple valorization routes, including essential oils, pectin recovery, and bioethanol. The resulting matrix is typically lower in lignin and richer in soluble and colloidal fractions than classical lignocellulosic residues, which can enable high sugar release under relatively mild pretreatment conditions [126]. The same chemistry imposes distinct constraints. Limonene and related volatiles can inhibit fermentation, while pectin influences slurry rheology and downstream solid–liquid separation [127,128,129]. Citrus residues are embedded in established value chains that include animal feed, pectin recovery, and essential oil production, so ethanol pathways are usually evaluated within multi-product systems rather than as stand-alone waste valorization. Reported LCA outcomes are therefore strongly contingent on co-product modelling: system expansion credits for displaced pectin, feed, or energy can dominate climate and energy indicators, whereas allocation-based results depend sensitively on the chosen mass, energy, or economic basis [130,131].

Potato peel waste represents a starch–fibre hybrid that combines first-generation-like starch fractions with structural carbohydrates. As shown in Figure 8, it is also a nutrient-rich by-product containing starch, dietary fibre, proteins, minerals, and bioactive compounds, which underpins competing valorization routes and affects counterfactual baseline selection in LCA. From a conversion perspective, this composition enables rapid initial saccharification via amylolytic enzymes but typically requires additional cellulolytic hydrolysis to access cellulose and hemicellulose. Process design therefore depends on multi-enzyme strategies and the integration of starch and lignocellulose conversion steps. Potato peel waste also has high moisture and high organic loads, which can shift environmental burdens toward wastewater handling. In addition, decentralized generation across processing facilities can constrain economies of scale and increase logistics complexity. Configurations that recover resources across streams, such as anaerobic digestion of liquid effluents and valorization of residual solids, tend to improve overall energy and environmental performance, indicating that system-level integration is often more decisive than marginal gains in ethanol yield [132].

Citrus and potato residues therefore exemplify fruit and starch sector wastes for which low lignin does not automatically translate into low environmental burden. Their comparative performance is shaped by inhibition control, rheology and wastewater management, together with competition and synergy among co-products in highly integrated food-processing chains.

3.2. Conversion Pathways

Conversion of food industry residues to bioethanol is dominated by biochemical routes following pretreatment, enzymatic hydrolysis, fermentation, and distillation. As shown in Figure 9, the reviewed systems fall into three route archetypes: stand-alone second-generation (2G) ethanol plants using residues as the primary feedstock; embedded configurations integrated into existing sugar or juice complexes; and multi-product biorefineries in which ethanol is one of several outputs [134].

Table 2. Representative conversion pathways for bioethanol from food industry residues.

Feedstock	Food Industry Segment	Country/Region	Product Slate	Main Conversion Route	Pretreatment & Hydrolysis	Fermentation Configuration	Scale	System Boundary in LCA/TEA	Co-Product Handling	FU & Method	Reference
EFB	Palm oil mill residue	Malaysia	Fuel ethanol + surplus electricity	Biochemical lignocellulosic ethanol	Dilute H2SO4 pretreatment, neutralization with NH3; enzymatic hydrolysis in stirred reactors, ≈90% cellulose to glucose	SHF with Zymomonas mobilis; beer distillation and rectification to fuel ethanol	Conceptual industrial biorefinery	Cradle-to-gate: nursery, plantation, milling, biorefinery, wastewater, cogeneration	Lignin, sludge and biogas to CHP; surplus power exported	FU 1 t bioethanol; ReCiPe Midpoint (H)–GWP, acidification, eutrophication, ecotoxicity	[135]
EFB	Palm oil mill residue	Brazil	Fuel ethanol	EFB → dilute acid pretreatment → enzymatic saccharification/fermentation → distillation/dehydration	Dilute acid pretreatment and enzymatic hydrolysis; parameters compiled from literature (no new experiments)	Conventional yeast fermentation; ethanol recovery by distillation	Modelled plant (student project)	well-to-tank + well-to-wheel (comparison)	Co-products and residues treated as in standard palm oil mills	FU 1 kg ethanol; CML-style midpoint indicators	[136]
EFB	Palm oil mill residue	Thailand	Ethanol + cogenerated heat & power	Biochemical ethanol integrated in palm biorefinery	Dilute acid pretreatment; enzyme production modelled from literature sources	Conventional yeast fermentation and distillation	Conceptual palm biorefinery scenarios	Cradle-to-gate LCA of several EFB options; one scenario is EFB–ethanol + cogeneration	Electricity and heat from lignin/biogas displace grid power (credit)	FU 1 t EFB; midpoint LCA (GWP, acidification, eutrophication, etc.)	[137]
EFB	Palm oil mill residue	Thailand	Fuel ethanol	Biochemical ethanol (TEA + flowsheet)	Hot water, hot compressed water or alkaline H2O2 pretreatment; enzymatic hydrolysis	SSF with yeast	Commercial (10,000 L d−1, 99.5 wt% ethanol)	Gate-to-gate: EFB at plant gate to anhydrous ethanol; separate TEA	Lignin-rich residues assumed to supply process heat	TEA only; primary metrics are energy use and cost per litre	[138]
EFB	Palm oil mill residue	Indonesia (conceptual)	Fuel ethanol	Biochemical ethanol	NaOH pretreatment of EFB; washing and neutralization; enzymatic hydrolysis to glucose	Batch fermentation with Saccharomyces cerevisiae; distillation	Lab/pilot	No full LCA; mass and energy balances reported for later LCI use	Lignin residue considered as boiler fuel	Not LCA; reports ethanol yield and energy per kg EFB	[139]
EFB	Palm oil mill residue	Malaysia	Fuel ethanol	Biochemical ethanol	Comparison of steam, dilute acid, alkaline and scCO2 pretreatments prior to enzymatic hydrolysis	SSF with yeast suggested, but study focuses on sugar release	Lab	No full LCA; pretreatment energy and chemical use reported	Residues implied as solid fuel	Not LCA; provides yields and energy inputs	[140,141]
EFB	Palm oil mill residue	SE Asia	Sugars + ethanol	Biochemical ethanol; multiple routes	Chemical-free fractionation (water/organosolv) of EFB and palm fibre; enzymatic saccharification	Conventional yeast fermentation assumed	Conceptual	No LCA; highlights potential benefits of chemical-free pretreatment	Lignin-rich solids and fibres proposed for energy or materials	No FU; qualitative comparison of options	[142]
EFB	Palm oil mill residue	Malaysia/Thailand	Fuel ethanol from oil palm biomass (EFB, trunk)	Biochemical ethanol	Dilute acid/alkaline pretreatment and enzymatic hydrolysis of EFB and trunk; consolidated process design	SSF with yeast at high solids under microaerobic conditions	Lab–pilot	No single LCA; process data used in later palm biomass LCA and TEA	Lignin and residues envisaged for CHP	Not LCA; ethanol yield (g kg−1 dry biomass) and energy balances	[143]
EFB	Palm oil mill residue	Malaysia	Sugars + ethanol (conceptual)	Biochemical ethanol with chemical-free pretreatment	Water-based fractionation of EFB and mesocarp fibre (no H2SO4); enzymatic hydrolysis to sugars	Conventional yeast fermentation considered as downstream option	Conceptual	No LCA; focuses on pretreatment severity vs. sugar yield	Cellulose-rich pulp and lignin proposed for materials or energy	No FU; qualitative discussion of reduced chemical load	[144]
EFB	Palm oil mill residue	General (review)	Biofuels (bioethanol as key product)	Biochemical & thermochemical routes	Review of alkaline, dilute acid, organosolv, steam explosion and biological pretreatments before enzymatic hydrolysis	SHF/SSF with S. cerevisiae and co-cultures; distillation assumed	Review	No quantitative LCA; compiles yield and energy data	Co-utilization of lignin for heat/power generally assumed	Not applicable; qualitative synthesis of impact data	[145]
Sugarcane bagasse	Sugar/ethanol industry residue	Mexico	2G ethanol (bagasse)	Biochemical lignocellulosic ethanol	Dilute acid or steam explosion pretreatment; enzymatic hydrolysis (several configurations)	Yeast fermentation of hydrolysate; distillation	Modelled industrial plant	Cradle-to-gate LCA integrated with process design	Lignin-rich residue for power; sometimes excess exported	FU 1 MJ or 1 L ethanol; midpoint LCA incl. GWP, acidification, eutrophication, land use	[146]
Sugarcane bagasse	Sugar/ethanol industry	India	2G ethanol	Biochemical ethanol (dilute acid)	Conventional vs. modified dilute acid pretreatment of rice straw and bagasse; enzymatic hydrolysis	SSF with yeast; distillation to fuel-grade ethanol	Modelled industrial plant	Cradle-to-gate LCA + life cycle costing for alternative dilute acid routes	Lignin-rich residues fired in boiler; excess power credited	FU 1 L ethanol; CML midpoint indicators (GWP, AP, EP, etc.)	[147]
Sugarcane bagasse	Sugar/ethanol industry	Brazil	2G ethanol (vs 1G)	Biochemical ethanol (integrated 1G + 2G)	Steam explosion or dilute acid pretreatment; enzymatic hydrolysis integrated into existing sugar–ethanol mills	Fermentation of C6 and partially of C5 sugars; distillation in common plant	Industrial scenarios	Cradle-to-gate comparison of 1G vs. 2G ethanol; focus on GHG and land use	Surplus electricity from bagasse/lignin to grid; vinasse as fertilizer	FU 1 km driven or 1 MJ ethanol; GWP, fossil energy demand, land occupation	[148]
Brewer’s spent grain (BSG)	Brewing industry	EU (conceptual plant)	Fuel ethanol from BSG	Biochemical ethanol	H3PO4 pretreatment (varied loadings); enzymatic hydrolysis with multienzyme cocktails	Separate hydrolysis and fermentation with yeast; distillation	Lab/pilot	No LCA; detailed mass and energy balances for LCI	Solid residues considered as fuel or animal feed	Not LCA; ethanol yield (L kg−1 dry BSG) and energy efficiency	[149]
BSG	Brewing industry	Europe	Fuel ethanol from BSG	Biochemical ethanol	Steam explosion to disrupt grains; enzymatic hydrolysis	SSF with S. cerevisiae; microaerobic conditions	Lab	No LCA; reports process yields and energy needs	Residual solids proposed for combustion	Not LCA; basis for later environmental assessment	[150]
BSG	Brewing industry	Brazil/global (conceptual)	Ethanol + other biorefinery products	Integrated BSG biorefinery	Layouts including dilute acid + enzymatic hydrolysis for ethanol vs. protein/chemicals routes	Conventional yeast fermentation in ethanol-oriented layout	Modelled industrial biorefineries	Cradle-to-gate comparative LCA of alternative BSG layouts	Protein concentrates and combusted residues treated as co-products	FU per kg BSG; midpoint LCA (GWP, respiratory inorganics, eutrophication)	[151]
SCG	Coffee shops/soluble coffee industry	China	Fuel ethanol from SCG	Biochemical ethanol with LHW pretreatment	Liquid hot water (180 °C, 20 min, S:L = 1:6); soluble sugars and mannans retained in liquor	Batch fermentation of hydrolysate with S. cerevisiae; ethanol ≈15 g L−1	Lab	Not LCA; discusses water savings from high solids loading	Solid residue suggested as fuel or material; not quantified	No FU; process yields and qualitative water use assessment	[152]
SCG	Coffee industry	UK (conceptual)	Fuel ethanol from oil-free SCG	Biochemical ethanol flowsheet	SCG defatting; acid hydrolysis; neutralization and conditioning of hydrolysate	Yeast fermentation; distillation; plant sized to UK SCG supply	Conceptual industrial design	Process flowsheet with mass/energy balances; used for GHG and cost discussion (not full ISO LCA)	Extracted oil to biodiesel; solid residues as boiler fuel	No fixed FU; scenarios compared per tonne SCG	[153]
SCG	Coffee industry	South Africa	Biofuels from SCG (includes ethanol)	Multi-route biorefinery	Route-specific pretreatment; ethanol route uses hydrolysis to sugars and yeast fermentation	Ethanol fermentation modelled alongside pyrolysis, AD and HTL	Modelled industrial plant	Cradle-to-gate LCA for fermentation, pyrolysis, AD and HTL routes	Electricity, biogas and char as co-products; substitution/system expansion	FU 1 MJ fuel energy; GWP, acidification, eutrophication, fossil energy use	[122,154]
Tea waste (black tea)	Tea processing	Egypt/Turkey	Fuel ethanol from tea waste	Biochemical ethanol	H2SO4 hydrolysis of dried tea waste; optimization of acid load, time and temperature	Fermentation with S. cerevisiae or E. coli K011; ethanol ≈8–9% (v/v)	Lab	No LCA; environmental benefit discussed qualitatively as waste valorization	Residual solids proposed as animal feed or soil amendment	Not LCA; yields and basic energy considerations	[85]
Tea waste (green tea spent)	Tea processing	India	Fuel ethanol	Biochemical ethanol	Acid or enzymatic hydrolysis of spent green tea; optimization of fermentation temperature (~50 °C)	Batch fermentation with S. cerevisiae; max ethanol ≈33% (v/v)	Lab	No LCA; process optimization only	Residual solids suggested for composting	Not LCA; kinetic and yield data only	[155]
Potato peel/potato processing waste	Potato processing industry	Global (case studies)	Fuel ethanol from potato peel/waste	Biochemical ethanol	Hydrothermal, dilute acid, alkaline and enzymatic pretreatments; often hydrothermal + cellulase/amylase/hemicellulase	Fermentation with S. cerevisiae or Z. mobilis; high-gravity runs up to ~79 g L−1 ethanol	Lab/pilot	No dedicated LCA; process energy demand and wastewater loads reported	Solid residues after fermentation used as biomanure or feed in some schemes	Not LCA; TEA and energy balances; environmental benefits discussed qualitatively	[156]

Notes: EFB: oil palm empty fruit bunches; BSG: brewer’s spent grain; SCG: spent coffee grounds; 1G: first-generation ethanol; 2G: second-generation ethanol; SHF: separate hydrolysis and fermentation; SSF: simultaneous saccharification and fermentation; LCA: life cycle assessment; TEA: techno-economic analysis; CHP: combined heat and power; AD: anaerobic digestion; HTL: hydrothermal liquefaction; LHW: liquid hot water; S:L: solid-to-liquid ratio; FU: functional unit; GWP: global warming potential; AP: acidification potential; EP: eutrophication potential.

compiles representative cases within these archetypes and indicates where studies diverge in pretreatment severity, hydrolysis–fermentation architecture, utility integration (heat and power), system boundaries, and co-product handling assumptions used in life cycle assessment (LCA) and techno-economic analysis (TEA).

3.2.1. Pretreatment Strategies and Feedstock–Pathway Matching

Pretreatment should be treated as residue-specific deconstruction and fractionation that disrupts the lignin–carbohydrate network and increases cellulose accessibility. As shown in Figure 10, pretreatment loosens the protective lignin matrix, partially solubilises hemicellulose, and exposes cellulose microfibrils, but it may also increase inhibitor formation and downstream handling burdens. Lignin-rich, ash- and silica-bearing fibres, typified by oil palm empty fruit bunches and in some cases bagasse, are commonly paired with high-severity schemes including steam explosion and strongly acidic or alkaline treatments, with some studies extending to compressed-water or supercritical media [157,158,159]. These routes increase cellulose accessibility but impose high steam and chemical demand and may require additional solvent handling and recovery when non-conventional media are used [160,161].

Low-lignin residues with substantial soluble or readily hydrolysable fractions, including citrus processing waste and potato peel waste and some tea-derived streams, are frequently processed under milder acid or hot water conditions with tailored enzyme systems such as pectinases and amylases alongside cellulases and hemicellulase. In these matrices, wastewater load (high COD), volatile inhibitors such as limonene, and the preservation of co-products constrain pretreatment selection as strongly as sugar release [164,165]. Protein- and lipid-rich residues introduce an explicit partitioning requirement: brewers’ spent grain often combines deconstruction of non-starch polysaccharides with options for protein valorization, while spent coffee grounds commonly motivate defatting and conditioning and may employ liquid hot water treatment at high solid-to-liquid ratios to manage hydrophobicity and phenolic inhibition [166,167]. Pretreatment severity therefore mediates a linked trade-off among sugar accessibility, inhibitor burden, utility demand and co-product quality across residue archetypes [168].

3.2.2. Hydrolysis and Fermentation Architectures

Hydrolysis and fermentation determine enzyme demand, inhibitor tolerance, and achievable ethanol titer, and they therefore propagate into downstream utilities and life cycle burdens [169]. Enzymatic hydrolysis is initiated by the synergistic action of endoglucanases and cellobiohydrolases that cleave cellulose chains and release glucose and soluble oligosaccharides. As shown in Figure 11, these enzymes attack distinct sites and ends of the polymer, which explains why apparent “enzyme demand” is a structural property of the residue matrix rather than a tunable input alone. Across the literature, residue-to-ethanol systems diverge along three coupled dimensions: enzyme system specification, hydrolysis–fermentation integration (SHF/SSF/SSCF), and feasible solids loading [170]. Enzyme requirements follow residue matrix: cellulase–hemicellulase paradigms dominate for EFB and bagasse; brewers’ spent grain motivates xylanase systems and sometimes protease support; potato peel waste requires combined amylolytic–cellulolytic cocktails; citrus-derived matrices can require pectin-directed enzymes where pectin governs viscosity and sugar release [77]. Enzyme manufacture can contribute materially to GWP and energy demand, yet many process models adopt generic inventory factors, which limit comparability when enzyme needs differ structurally across residues [171].

Process integration choices separate experimental feasibility from plant-oriented designs. SHF is frequently used to isolate hydrolysis performance and inhibition effects, whereas SSF is commonly adopted in conceptual designs to simplify unit operations and manage end product inhibition. SSCF is proposed when pentose utilization is targeted, but it is often simplified to glucose-only fermentation in LCA due to limited robust mixed-sugar data under industrial solids and inhibitor loads. As shown in Figure 12, hexose-to-ethanol conversion follows different central carbon routes in yeast and bacterial systems, which underpins differences in cofactor use and pathway structure when mapping fermentation assumptions to inventory. Solids loading sets a practical ceiling on ethanol titer and controls distillation steam demand and effluent handling [169]. High-solids operation is more feasible for uniform fibrous residues than for matrices with challenging rheology or fine particles, such as spent coffee grounds, brewers’ spent grain and some tea wastes; pretreatment-derived inhibitors, such as furans, phenolics, volatile terpenes, further constrain solids unless conditioning, detoxification or tolerant strains are explicitly represented [173]. Hydrolysis and fermentation should therefore be treated as system-defining elements rather than interchangeable unit operations when interpreting indicator differences across studies.

3.2.3. Energy Systems, Scale and Integration with Existing Plants

Energy supply architecture determines the utilities baseline of waste-to-ethanol systems and is a primary driver of cross-study variability in GWP and NREU. Agro-industrial fibre residues, especially oil palm empty fruit bunches and sugarcane bagasse, are commonly modelled with on-site boilers or CHP where lignin-rich solids provide steam and electricity [175]. Electricity exports and its credited substitution frequently dominate reported impacts in these cases [111]. High-severity pretreatment and distillation can exceed residue-derived heat supply and force fossil backup or grid imports, which reverses the apparent advantage. Boiler efficiency, heat recovery extent and treatment of exported power therefore act as high-leverage assumptions in LCA [176].

Beverage- and food-processing residues, including brewers’ spent grain, spent coffee grounds, citrus waste and potato peel waste, are more often assessed without dedicated CHP, with steam and electricity supplied by natural gas, fuel oil or the grid unless biogas recovery, solids combustion or oil extraction is integrated [177]. High-moisture and lipid-rich residues amplify the trade-off between drying/conditioning energy and recoverable energy from solids, biogas or extracted oils. Scale assumptions separate bench-stage energy accounting from industrial performance: laboratory and pilot studies rarely represent pinch-based heat integration or multi-effect evaporation, whereas conceptual industrial designs typically embed them and report materially lower net utilities. Integration into existing plants, such as sugar mills or citrus facilities, shifts the analysis to marginal changes in host plant energy and co-product structures, making co-product treatment and displaced services central to interpretation [178].

4. Methodological Drivers of Variability in Waste-Derived Ethanol LCA

Life cycle assessments of bioethanol from food industry residues report widely divergent greenhouse gas emissions, energy demand, and resource footprints, and this dispersion is driven primarily by methodological configuration rather than by conversion yields alone [179]. As shown in Figure 13, system definition, inventory specification, co-product modelling, and life cycle impact assessment (LCIA) method selection form a linked chain that conditions the reported indicators.

Table 3. Methodological configurations of life cycle assessments for bioethanol from food industry lignocellulosic wastes.

Feedstock	Functional Unit	System Boundary	Foreground Process Scope	Co-Product Modelling	Energy System Assumption	Key Inventory Datasets/Assumptions	LCIA Method & Indicator Set	Notes on Comparability	Reference
Municipal wet biowaste (food-dominated)	1 t food waste; 1 MJ ethanol	Cradle-to-grave	Collection → sorting → dilute acid hydrolysis → fermentation → distillation; co-products to animal feed/energy	Not specified (credits implied)	Not specified	Avoided landfill baseline included; enzyme + acid inventories implied	GWP, AP, EP, POCP	Waste management framing; multi-output credits likely decisive	[180]
Household food waste + agricultural residues	1 t mixed HFW + ag residues; 1 MJ EtOH	Cradle-to-grave	Pre-sorting → pretreatment → enzymatic hydrolysis → fermentation; multiple baselines	Not specified (scenario baselines)	Electricity mix explicitly influential (stated)	Waste management baselines vary; electricity mix; enzyme inventories implied	GWP, AP, EP, ODP, HTP	Cross-scenario comparability limited by baseline definition	[181,182]
Food-processing & retail waste (syrups, off-spec products)	1 t waste; 1 MJ EtOH	Cradle-to-grave	Direct fermentation of high-sugar waste; compared with AD/incineration	Not specified (credits implied)	Not specified	Avoided conventional feed and fossil fuel appear in scenarios	GWP	Not lignocellulosic; included mainly for methodological contrast	[183]
Banana agricultural/packaging waste (stems, rejected fruits)	1 t banana agro-waste; 1 L EtOH	Cradle-to-gate	Shredding → dilute acid or enzymatic hydrolysis → fermentation → distillation	Not specified	On-site residue heat implied in scenarios	Heat source is a key scenario lever	GWP, CED	FU differs (waste-based + product-based); scenario-driven	[183]
Banana fruit, peel and stalk mixture	1 L fuel-grade ethanol	Gate-to-gate	Pressing → hydrolysis → fermentation (stream comparison)	Not specified	Waste heat assumed in scenarios	Upstream cultivation burden included for fruit stream	GWP	Gate-to-gate; feedstock stream definition drives interpretation	[184]
Brewery spent grain and brewery by-products	1 t BSG; 1 L EtOH	Cradle-to-gate	Wet milling → hydrolysis → fermentation; integrated biorefinery (EtOH + XOS + biogas)	Not specified (multi-product)	Biogas used for process heat (scenario)	Multi-product biorefinery inventories (XOS, biogas)	GWP, CED, AP, EP	Strong multi-product coupling; co-product method must be explicit	[185]
Citrus peel/citrus-processing waste (multi-output)	1 MJ E85; 1 kWh power; 1 kg limonene; 1 kg digestate	Well-to-wheels	Limonene removal → acid hydrolysis → fermentation → AD of residues	Not specified (multi-FU)	On-site power generation implied	Multi-functional outputs; displaced fertilizer implied	GHG	Multiple functional units; not directly comparable to single-FU studies	[186]
Citrus waste multi-product biorefinery	1 t citrus waste; 1 kg of each product	Cradle-to-gate	Hydrolysis → fermentation → syngas, methane, phenolics, essential oils	Allocation method stated as influential (basis not specified)	Not specified	Separation operations included (flash separation mentioned)	GWP, CED, several midpoints	Strong sensitivity to allocation basis; multi-output comparability limited	[187]
Palm oil frond (OPF) juice/fibre	1 t anhydrous ethanol	Gate-to-gate	Transport → milling → juice extraction → mild pretreatment → fermentation → distillation	Not specified	Steam & electricity supply modelled (details not specified)	OPF treated as low-burden residue; transport included	ADP, AP, EP, GWP, ODP, HTP, FEP, MEP, TEP, POP	Gate-to-gate; residue burden assumption is decisive	[188]
EFB – dedicated ethanol plant	1 t fuel-grade ethanol	Cradle-to-gate	Size reduction → pretreatment (steam + chemicals) → enzymatic hydrolysis → fermentation → distillation; boiler fired with EFB	Not specified	On-site boiler/CHP using EFB solids implied	Enzyme inventory + pretreatment steam important (model levers)	GWP, CED; ReCiPe 2016 midpoints	Utility system definition drives GWP/NREU comparability	[135]
Oil palm biorefinery scenarios with EFB ethanol	1 t fresh fruit bunch (FFB) processed	Cradle-to-gate (FFB at mill to products)	Scenario set: conventional mill; EFB co-composting; EFB ethanol + power	Not specified (scenario/system expansion likely)	Surplus power export modelled	Transport of EFB + chemicals included	GWP, AP, EP, ADP, HTP	FU at mill level; not comparable to “per kg ethanol” studies	[137]
Sugarcane bagasse – 2G ethanol (generic)	1 L or 1 t ethanol	Cradle-to-gate	Bagasse extraction → pretreatment → enzymatic hydrolysis → fermentation → distillation	Not specified	Lignin-rich residue used for power implied	Bagasse treated as residue; chemicals + enzymes + distillation steam inventoried	GWP, energy use	FU varies across studies; residue burden assumption must be aligned	[188]
Sugarcane bagasse – Malaysian case study	252 kg ethanol from 1 t bagasse (98.7 vol%)	Cradle-to-gate	Bagasse extraction → transport → pretreatment → hydrolysis → fermentation → distillation	Not specified	Not specified	Transport explicitly included; enzyme + pretreatment steam inventories	18 ReCiPe 2016 midpoints	FU is process-output-specific; good for within-study sensitivity	[189]
Cassava pulp (starch industry residue)	1 L fuel ethanol; 1 km driven with E10	Cradle-to-wheel	Pulp collection → drying → hydrolysis → fermentation → distillation; biogas from WW replaces fuel oil	Not specified	Boiler fuel substitution by biogas (scenario)	Allocation at cassava factory influential (stated)	GWP, AP, EP, POCP, CED	Mixed FU (fuel + mobility); allocation at source drives results	[190]
Whole-plant cassava/cassava waste scenarios	1 t cassava (whole plant); 1 MJ ethanol	Cradle-to-gate	Integrated process using roots, stems, leaves; sensitivity to peel waste vs. full plant	Not specified	Not specified	Fertilizer and field emissions included (stated)	GWP, energy use	Not lignocellulosic waste-focused; agricultural burdens dominate	[191]
Potato processing waste (potato peel)	1 t potato peel; 1 L ethanol	Cradle-to-gate (often streamlined)	Thermal/thermo-chemical pretreatment → hydrolysis → fermentation; potato processing as foreground	Not specified	Boiler fuel sensitivity noted (details not specified)	Streamlined inventory common; steam + enzymes often included	GWP, CED (sometimes energy only)	“Streamlined” scope reduces cross-study comparability	[101]
Mixed potato and banana waste	1 t mixed waste; 1 L ethanol	Cradle-to-gate	SSF; scenarios with/without external energy supply	Not specified	External energy supply as scenario lever	Scenario-driven utilities definition	Impact 2002+ midpoints	LCIA method differs; scenario framing dominates	[188]
Spent coffee grounds (SCG)	1 t SCG; 1 MJ biofuel	Cradle-to-gate	Multi-route comparison: EtOH, biodiesel, pellets, biogas; sequential oil extraction + fermentation	Not specified	Not specified (route-dependent)	Route comparison; oil extraction included in some routes	GWP, CED, AP, EP	Multi-route functional comparability depends on consistent FU	[120]
Food waste to ethanol with co-products	1 t source-separated food waste; 1 MJ EtOH	Cradle-to-grave	Mechanical preprocessing → fermentation → distillation → co-product recovery	Not specified (credits implied)	Not specified	Co-product credits central (stated); enzyme + energy inputs included	GHG, fossil energy	Waste management framing; credit choices dominate	[192]
Multi-feedstock comparative LCA (wastes vs. energy crops)	1 MJ fuel ethanol	Cradle-to-grave	Harmonized LCA across 1G/2G/waste feedstocks	Not specified	Not specified	Harmonization framework; cross-feedstock baseline alignment	GWP, land use, water, CED	Comparative scope; indicator definitions must be aligned across feedstocks	[193]

Notes: ADP: abiotic depletion potential; AP: acidification potential; EP: eutrophication potential; POCP: photochemical oxidant creation potential; GWP: global warming potential; ODP: ozone depletion potential; HTP: human toxicity potential; FEP: freshwater eutrophication potential; MEP: marine eutrophication potential; TEP: terrestrial ecotoxicity potential; POP: photochemical oxidant formation; CED: cumulative energy demand; NEV: net energy value; ER: energy ratio; WU: water use; LU: land use; NREU: non-renewable energy use; ReCiPe 2016 (midpoint): ReCiPe 2016 midpoint method; IMPACT 2002+ (midpoints): IMPACT 2002+ midpoint method.

maps these choices across representative studies, including functional units, system boundaries, co-product handling strategies, key inventory assumptions, and LCIA methods.

4.1. Functional Units and System Boundaries

Functional unit choice determines whether a study answers a waste management question or a fuel supply question. As shown in Table 3, some LCAs normalize results to a unit mass of residue, which frames the system as a waste diversion intervention and makes results sensitive to the assumed baseline fate of the waste [194]. Other studies normalize to a unit of ethanol (mass, volume or energy), which frames the system as a fuel production pathway and shifts attention to conversion efficiency and utility intensity. Results reported on a waste basis and on an ethanol basis are not interchangeable and should not be compared without explicit conversion of the reference flow and problem framing.

System boundary selection further controls whether avoided burdens and displaced services are inside the model [195]. Gate-to-gate studies isolate the conversion train and typically omit collection, upstream provision of chemicals and utilities, and downstream use-phase effects. Cradle-to-gate studies include residue generation, transport and utility supply, bringing energy systems, pretreatment chemicals and wastewater treatment into the foreground. Cradle-to-grave or well-to-wheel assessments extend to fuel use and end of life, enabling comparison with fossil fuels but increasing dependence on assumptions about substituted products and co-product markets [196]. Differences in FU and boundary scope therefore account for a substantial share of the spread in reported indicators and must be treated as primary comparability conditions rather than as secondary reporting preferences [197].

4.2. Co-Product Structures and Allocation Approaches

Co-product modelling determines how burdens and credits are distributed between ethanol and the other outputs generated by residue-based systems and is often a stronger driver of reported indicators than conversion yield differences [198]. As summarized in Table 3, typical co-products include electricity from combustion of lignin-rich solids, oils routed to biodiesel in spent coffee grounds schemes, pectin and essential oils in citrus biorefineries, protein-rich fractions from brewers’ spent grain, and digestate or biochar when anaerobic or thermal routes are integrated [199,200]. The methodological issue is not whether co-products exist but how the LCA assigns their shares of upstream and process burdens.

Allocation approaches partition burdens among products using mass-, energy- or economic-based rules. The assigned share to ethanol can shift materially when co-products differ strongly in energy density, market value or functional service, which is common in residue biorefineries [198]. Economic allocation is particularly unstable because it depends on price structures that vary by region and time and can transfer burdens away from ethanol when high-value co-products dominate revenue. System expansion instead credits the system for displaced products, such as grid electricity, conventional fuels, feed ingredients or recovered chemicals, but results then depend on the assumed marginal product, market boundary and regional energy mix. Co-product handling therefore defines a comparability condition: reported GWP and NREU values cannot be interpreted across studies without explicit alignment of allocation basis or substitution assumptions.

4.3. Inventory Modelling: Energy Systems, Enzymes, Wastewater and Logistics

Inventory specification controls which unit operations dominate the life cycle profile and is a major source of cross-study spread in GWP and energy indicators. As shown in Table 3, four inventory blocks recur as high-leverage modelling levers: energy systems, enzyme supply, wastewater treatment and logistics/moisture handling [89].

Energy system inventories often decide the direction of net energy and climate performance. For EFB and bagasse pathways, many studies model on-site combustion of lignin-rich solids to supply steam and electricity, sometimes with electricity export; boiler efficiency, heat recovery extent and the treatment of exported power govern the resulting burdens and credits [201]. For residues without established CHP infrastructure, including spent coffee grounds, brewers’ spent grain, citrus waste and potato peel waste, models frequently default to grid electricity and fossil steam, which can dominate GWP and NREU independently of biochemical yields. When anaerobic digestion or oil recovery is integrated, inventories for recovered biogas or biodiesel and their substitution assumptions become additional controlling inputs [202,203].

Enzyme inventories are frequently under-specified relative to residue needs. Many LCAs apply generic cellulase datasets even when conversion requires multi-enzyme systems (xylanases, β-glucanases, proteases, pectinases, amylases) and residue-specific dosing; this simplification can dominate fossil energy and GWP results in high-dosage or high-solids cases and can mask meaningful differences among pathways [204]. Wastewater treatment inventories show comparable divergence: high-COD liquid streams from fruit- and starch-derived residues are variably represented as detailed aerobic/anaerobic treatment, simplified removal factors or omissions, shifting eutrophication- and toxicity-relevant flows and associated energy use [205]. Logistics and moisture handling further modify inventories through drying assumptions for high-moisture residues and through transport system definition (plant gate availability versus explicit collection networks), which can materially change fuel use and emissions in decentralized supply chains [206].

4.4. LCIA Method Selection and Indicator Coverage

LCIA method choice and indicator coverage limit cross-study comparability even when foreground process models are similar. As summarized in Table 3, GWP is widely reported but is not method-invariant: studies differ in IPCC versus ReCiPe/CML characterization, treatment of land use change, and handling of biogenic CO₂ from fermentation and residue combustion. These choices change the accounting relationship between biogenic and fossil carbon and can shift the interpreted climate performance of the same process configuration [207].

Energy indicators show broader inconsistency. NREU and cumulative energy demand are variably defined, with some studies including embodied energy in enzymes, solvent recovery and wastewater treatment and others restricting accounting to direct fuels and electricity. For multi-product systems, the interaction between energy indicators and co-product credits (electricity export, biogas, recovered oils) further depends on how displaced energy carriers are characterized within the chosen method [208,209]. Water- and land-related metrics remain the least standardized: water use ranges from simple process-water totals to scarcity-weighted indicators with inconsistent inclusion of cooling water and boiler makeup; land occupation is sometimes set to zero for residues and sometimes derived via allocation of cultivation burdens, producing method-driven rather than process-driven LU values. Beyond these core indicators, midpoint categories such as acidification, eutrophication, toxicity and photochemical smog are inconsistently covered, which is consequential for residues with high-COD effluents or solvent-intensive pretreatments [210].

Indicator heterogeneity therefore acts as a reporting constraint rather than a secondary methodological detail. Cross-study synthesis should treat LCIA method and indicator definition as explicit alignment conditions, particularly for NREU, WU and LU, where differences in scope and characterization frequently exceed the magnitude of residue-to-residue differences reported in the literature.

5. Comparative Environmental Performance Across Food Industry Waste Ethanol Systems

Reported life cycle indicators for waste-derived bioethanol vary widely across studies and are not uniformly comparable because functional units, boundary scope, and co-product modelling are inconsistent. Figure 14 synthesizes the reported ranges and reporting gaps across indicators.

Table 4. Summary of reported life cycle indicators for bioethanol from food industry lignocellulosic wastes.

Feedstock	GWP (kg CO2-eq/FU)	NREU (MJ/FU)	WU (kg/FU)	LU (m2a/FU)	Reference
EFB	8.06	57.74	118.58	–	[211]
EFB	−0.14	31.16	109.52	–	[211]
EFB	4.34	–	6.75	–	[212]
EFB	0.45	6.5	3.66	1.09	[213]
EFB	0.74	9	3.45	0.59	[213]
EFB	9.95	27.37	0.48	0.19	[214]
EFB	2.76	–	–	1.03	[215]
EFB	0.35	–	–	1.03	[215]
EFB	7.68	–	–	–	[135]
Coffee grounds	8.65	85.8	16.8	–	[122]
Tea waste	2.22	–	–	–	[216]
Bread waste	1.27	–	–	–	[217]
Apple pomace	1.31	13.9	1840	1.16	[218]
Banana rachis	0.84	7.4	–	–	[183]
Brewer’s spent grain + barley straw	7.39	–	–	–	[219]
Citrus waste	−1.04	–	–	–	[220]
Oil palm frond	0.224	–	–	–	[221]

Note: GWP: global warming potential (kg CO₂-eq); NREU: non-renewable energy use (MJ); WU: water use (kg water); LU: land occupation (m²·a). FU: functional unit (1 kg anhydrous bioethanol). “–”: not reported or could not be extrapolated in the original study.

compiles the underlying values for GWP, NREU, WU, and LU on a 1 kg anhydrous ethanol basis where available. Where original studies reported results on alternative bases, values were recalculated only when the source paper provided sufficient information to support transparent, unit-consistent conversion.

5.1. Greenhouse Gas Performance Across Waste-Derived Ethanol Systems

GWP is the most consistently reported indicator for waste-derived ethanol. As shown in Table 4, values range from net-negative results to about 10 kg CO₂-eq per kg anhydrous ethanol. This dispersion appears across feedstocks and within the same feedstock under nominally similar biochemical routes [188].

Oil palm empty fruit bunches (EFB) illustrate pronounced intra-feedstock spread. Some scenarios report near-zero or negative GWP, whereas others report substantially higher values. Lower GWP is typically associated with internal utility closure, where lignin-rich solids supply process heat and power, and where electricity export or avoided waste management credits are applied. Higher GWP is generally reported when pretreatment and distillation depend on steam-intensive utilities with limited internal energy recovery [213].

Other residues have fewer reported data points, but configuration dependence remains evident. Spent coffee grounds and the brewers’ spent grain–barley straw system fall within an intermediate positive GWP range in the available studies. These results are consistent with models that include drying and preprocessing energy and import external utilities [222]. Apple pomace, banana rachis, and bread waste show lower GWP values in Table 4, but these outcomes remain conditional. They depend on the counterfactual handling of the residue and on whether wastewater treatment, detoxification, and utility sourcing are explicitly modelled within the system boundary [223,224].

5.2. Energy Performance and Non-Renewable Energy Demand

Reported NREU values vary widely and track the assumed “energy backbone” for steam and electricity supply more closely than feedstock identity alone. For lignocellulosic agro-industrial residues, lower NREU is generally reported when lignin-rich solids supply process steam and power and when heat integration limits external utility imports. Higher NREU is reported when pretreatment and distillation impose high steam demand that is met by external fossil utilities or when boiler efficiency and heat recovery assumptions are weakly specified [225,226].

Spent coffee grounds (SCG) often show an NREU penalty because preprocessing steps are frequently modelled explicitly [227]. Drying and conditioning, and in some schemes defatting or solvent-related operations, increase thermal and electricity demand, and NREU rises when these utilities are supplied by grid electricity or fossil steam [175]. Lower NREU is reported or implied when models include energy recovery from residual solids, anaerobic digestion of liquid streams, or use of extracted lipids as a co-fuel. This pattern indicates that SCG energy performance is largely governed by whether co-product energy recovery is included within the system boundary and treated consistently in crediting [222].

For fruit- and starch-derived residues, milder pretreatment and higher readily fermentable fractions can reduce direct process energy inputs, but the net effect remains boundary-dependent [228]. Wastewater treatment energy and chemical inputs can dominate NREU in high-COD systems. By contrast, integration into existing plants or explicit CHP modelling can reduce fossil energy demand through shared utilities or electricity export credits.

5.3. Water and Land Implications: Variability and Reporting Gaps

WU and LU are reported inconsistently and remain weaker bases for cross-feedstock comparison than GWP and NREU [229]. Reported WU spans from single-digit values in some EFB scenarios to >100 kg water per kg ethanol in others. Spent coffee grounds fall in an intermediate range, and apple pomace shows an extreme high value in Table 4. This spread reflects accounting definitions as much as process performance [230]. Studies differ in whether WU includes only process and boiler make-up water or also cooling water, dilution water, and water used for wastewater treatment. Some studies also include upstream agricultural water embedded in residue generation through allocation, which can shift WU by orders of magnitude.

A pronounced outlier is observed for apple pomace (1840 kg/FU). This discrepancy is most plausibly interpreted as a system boundary effect, because WU estimates change substantially depending on whether upstream cultivation water, such as irrigation, is included versus reporting only conversion-stage water use. Comparisons across residues therefore require explicit alignment of WU definitions and boundary scope before interpreting apparent differences as technology effects.

LU values are sparse and, when reported, are generally low, consistent with residue-based supply. Non-zero LU reported for some residues indicates that certain studies allocate a share of cultivation or processing land occupation to the residue stream rather than treating the residue as burden-free. LU outcomes therefore depend on the upstream co-product and allocation structure and are not directly inferable from conversion performance.

Overall, the observed WU and LU patterns are dominated by reporting scope and allocation choices, and the current evidence base is insufficient for robust benchmarking across residue classes without harmonized definitions and boundary treatment.

6. Comparative Insights and Future Directions for Waste-Derived Ethanol LCA

Taken together, the comparative evidence reviewed here indicates that life cycle indicators reported for waste-derived ethanol should be interpreted as system-contingent outcomes rather than intrinsic feedstock properties [231]. Across residues, similar numerical values of GWP or NREU often emerge from fundamentally different configurations, while sharply divergent values can arise from nominally similar conversion routes [232]. This observation challenges feedstock-centric narratives that implicitly rank wastes by “environmental suitability” and instead supports a configuration-centric perspective, in which energy integration, co-product structures and baseline definitions determine environmental performance [233,234]. In this sense, the primary comparative insight from the literature is not which residue performs best, but under which structural conditions a residue-to-ethanol system can credibly deliver environmental benefits [235].

From this perspective, cross-study comparison becomes meaningful only when indicator values are read together with their modelling context. Apparent advantages in GWP or energy demand cannot be generalized without reference to functional units, boundary scope and co-product treatment, because these choices effectively redefine the question being asked from waste management optimization to fuel substitution or biorefinery allocation [198,236]. Consequently, synthesis across studies should prioritize pattern recognition over numerical aggregation, identifying recurring combinations of assumptions and system designs that lead to favourable or unfavourable outcomes. This shifts the role of review synthesis away from benchmarking absolute values and toward clarifying the conditional logic that links modelling decisions to reported performance [235].

Future research should therefore focus less on generating additional isolated point estimates and more on building comparability through structured variation. This includes systematically testing alternative co-product treatments, energy system configurations and baseline waste fates within the same modelling framework, as well as expanding consistent reporting of water and land indicators alongside climate and energy metrics [21]. For decision-relevant application, LCA of waste-derived ethanol will benefit from scenario-based analyses that explicitly connect environmental indicators to regional energy systems, logistics constraints and waste management contexts. Such an approach allows LCA results to inform deployment choices without overstating their generality, and positions waste-derived ethanol not as a universally superior pathway, but as a context-dependent option within broader circular bioeconomy strategies [237,238].

7. Conclusions

This review demonstrates that the life cycle environmental performance of bioethanol derived from food industry lignocellulosic wastes is governed primarily by system configuration and methodological choices rather than by feedstock identity alone. Across studies, co-product modelling—allocation versus substitution/system expansion—is the single most influential comparability condition and the dominant “noise source” that reshapes viability in cross-study comparisons, often exceeding the effect of biochemical yield differences. By structuring the literature around feedstock archetypes, conversion architectures and LCA modelling decisions, the analysis shows that differences in functional units, system boundaries, co-product handling and energy system assumptions systematically outweigh variations in conversion performance when explaining the wide dispersion of reported GWP, NREU, water and land indicators. Pathways that achieve internal energy closure through residue-based heat and power generation and credible displacement of baseline waste management consistently report lower climate and fossil energy burdens, whereas steam-intensive pretreatments, external fossil utilities and under-specified effluent treatment and pretreatment-related emissions frequently drive higher impacts across residues. Water use and land occupation results remain insufficiently robust for cross-feedstock benchmarking due to inconsistent definitions and allocation practices. From a policy perspective, the results indicate that incentives framed solely around “waste feedstock” identity are insufficient; robust energy integration, minimized external fossil utilities, and transparent co-product and effluent modelling should be rewarded as the conditions that govern real environmental performance. Collectively, these findings clarify why existing LCA conclusions appear contradictory and provide an interpretative framework that separates structural system effects from feedstock-specific properties, supporting more transparent comparison and more policy-relevant interpretation of waste-derived ethanol LCAs.