CO₂ Capture-Integrated Gasification of Hazelnut Shells: Process Performance Investigation via a Hybrid MATLAB–Aspen Modelling and Techno-Economic Evaluation

Abstract

This work presents a techno-economic assessment of hydrogen production via sorption-enhanced gasification (SEG) of hazelnut shells across three plant scales (100 kWth, 1 MWth, and 10 MWth). The overall model is developed through the integration of Aspen Plus^® process simulation, coupled with MATLAB^®-based kinetic reactor modelling, enabling the assessment of the entire process chain. The kinetic SEG model, validated against experimental literature data, was implemented to describe the fluidized bed gasifier behaviour at the three scales. The resulting process streams were subsequently integrated into Aspen Plus^® for downstream upgrading and overall system analysis. The simulations show that the SEG process produces a hydrogen-rich syngas with H₂ contents around 80 vol.%_dry-basis, which is further upgraded to a hydrogen purity of 99.95% with a recovery of 90% via pressure swing adsorption. The process exhibits stable performance across scales, with Cold Gas Efficiency values around 60% and hydrogen yields close to 1 Nm³/kg_Biomass. The economic analysis highlights a decrease in the Levelized Cost of Hydrogen (LCOH) from 41.3 €/kg at 100 kWth to 6.8 €/kg at 10 MWth. These results indicate that SEG represents a promising pathway for low-carbon hydrogen production, while enabling the valorisation of biogenic residues within a sustainable energy framework.

1. Introduction

Fossil fuels still represent the most widely used source of energy worldwide. In conventional energy systems, chemical energy is converted into useful energy through combustion, a process in which fuels react with oxygen to release heat, which is subsequently exploited in thermodynamic cycles for power production. This process results in the emission of CO₂, a major contributor to the greenhouse effect. Consequently, human activities continuously extract carbon stored in geological reservoirs and re-emit it into the atmosphere through irreversible processes. As a result, fossil fuels are neither renewable nor sustainable, as their formation timescales are not compatible with current consumption rates. For these reasons, increasing attention has been directed toward renewable energy sources and sustainable energy technologies.

As a renewable resource, biomass represents a key option for bioenergy production through several thermochemical, biological or physicochemical conversion processes [1]. Among thermochemical processes, gasification converts biomass into a gaseous mixture primarily composed of hydrogen, carbon monoxide, and carbon dioxide. This process involves multiple stages, including drying (100–200 °C), pyrolysis (200–600 °C), combustion, and gasification. During pyrolysis, biomass decomposes into gases, tar, and char, while subsequent reactions further convert these products into syngas. Due to its versatility, syngas can be used for power generation or as an intermediate for the production of fuels and chemicals. Among the products that can be derived from syngas, hydrogen has gained increasing attention as a strategic energy carrier. It can be used in high-efficiency systems such as fuel cells, in the production of synthetic fuels [2,3,4,5], or as a reducing agent in industrial applications such as steelmaking processes.

Hydrogen can be produced from biomass-derived syngas through integrated processes such as sorption-enhanced gasification (SEG), coupled with pressure swing adsorption (PSA) [6]. In SEG systems, in situ CO₂ capture shifts the equilibrium of the Water Gas Shift (WGS) and reforming reactions toward hydrogen production, enabling higher hydrogen yields (up to ~90 vol.% on a dry basis under ideal conditions) [7].

CO₂ capture can be performed using calcium-based solid sorbents, among which dolomite represents a cost-effective and widely available option [8,9,10,11]. The carbonation reaction occurs at temperatures around 650–700 °C, while sorbent regeneration (calcination) requires temperatures of 900–950 °C. These conditions can be achieved in a dual fluidized bed (DFB) system, where gasification/carbonation and sorbent regeneration are carried out in separate but interconnected reactors [12]. This configuration enables continuous operation, producing both a hydrogen-rich stream and a concentrated CO₂ stream suitable for storage or utilization. The produced syngas can be further upgraded via PSA, a mature technology based on selective adsorption, typically operating at 10–30 bar and 30–60 °C and achieving hydrogen purities above 99.9% with recoveries between 70% and 90% [13,14].

In addition to hydrogen production, the SEG process can also serve as an intermediate platform for the synthesis of value-added products. Recent studies have explored the feasibility of small-scale ammonia production via biomass SEG, reporting efficiencies of 44.5–46.7% [15]. Similarly, SEG-based configurations integrated with carbon utilization strategies have been investigated for methanol synthesis, achieving yields up to 0.70 kg_MeOH/kg_biomass [16]. Moreover, when coupled with appropriate pre- and post-treatment technologies, such as hydrothermal carbonization (HTC) and hot gas cleaning (HGC), SEG systems can also be applied to the conversion of low-grade and high-moisture biogenic residues, as well as for power generation through a solid oxide fuel cell. Overall, these applications highlight the versatility of SEG-based configurations for biomass valorisation into multiple energy carriers and chemical products.

The economic viability of hydrogen is a decisive factor for its large-scale deployment. Current production pathways present several limitations. Fossil-based routes are associated with high emissions, while low-carbon alternatives such as electrolysis and Steam Methane Reforming (SMR) with carbon capture and storage (CCS) are strongly affected by the variability of electricity and natural gas prices, as well as infrastructure requirements. These factors introduce significant uncertainty in hydrogen production costs. In this context, the Levelized Cost of Hydrogen (LCOH) is widely used to compare different technologies, representing the average production cost over the plant lifetime and corresponds to the break-even hydrogen selling price. Target values set by several countries range between 1.00 and 2.40 $/kg by 2030. However, current costs remain significantly higher: electrolysis-based hydrogen is estimated to be between 6 and 13 $/kg, while biomass gasification ranges between 3 and 14 $/kg, depending on scale [17].

Although sorption-enhanced gasification has been widely recognized as a promising pathway for low-carbon hydrogen production [18], several limitations still affect its assessment at the process and system level. Most of the studies currently available in the literature rely on thermodynamic equilibrium or quasi-equilibrium approaches [19], which may overpredict hydrogen production under idealized equilibrium assumptions. In addition, limited efforts have been devoted to the integration of detailed reactor-scale modelling with downstream process simulation and hydrogen upgrading units, preventing a comprehensive evaluation of the overall system performance.

In this context, the present work proposes an integrated modelling framework for the techno-economic assessment of hydrogen production via SEG of hazelnut shells across three plant scales (100 kWth, 1 MWth, and 10 MWth). A MATLAB-based kinetic model of the fluidized bed SEG reactor with in situ CO₂ capture using CaO-based sorbents, validated against experimental data available in the literature, was coupled with an Aspen Plus^® process simulation to evaluate the entire conversion chain, including syngas conditioning and hydrogen purification via PSA. Compared to conventional equilibrium-based approaches, the adopted methodology enables a more representative description of reactor behaviour and process integration effects, allowing the simultaneous assessment of hydrogen production performance, process efficiency, and scale-up effects. The proposed approach therefore provides a consistent framework for evaluating the industrial scalability of SEG systems and the energetic valorization of biogenic residues for sustainable hydrogen production. Furthermore, the economic analysis was performed to assess the competitiveness of the proposed technology within the low-carbon hydrogen market based on the Levelized Cost of Hydrogen (LCOH) as the key economic indicator. This allows addressing the research question of whether CaO-based sorbents represent a viable and economically competitive decarbonization pathway. To account for policy uncertainty and market variability, a sensitivity analysis was performed on the most volatile economic parameters, providing a quantitative assessment of their influence on the LCOH under different market scenarios.

2. Methods

2.1. Process Modelling

The kinetic model used in this work is based on an experimentally validated kinetic model developed by Di Carlo et al. for the simulation of biomass gasification within a DFB system [20]. The present study extends that framework by integrating carbonation kinetics, enabling the simulation of sorption-enhanced gasification and allowing performance evaluation of calcium looping-based system for hydrogen production. The updated model was validated against experimental data and subsequently employed to investigate hydrogen production performances and the impact of plant scale on the LCOH.

Figure 1 presents the workflow adopted in the present study, highlighting the integration between the MATLAB kinetic model and the Aspen Plus^® process simulation. The kinetic simulations were performed using MATLAB R2025b (The MathWorks, Inc., Natick, MA, USA), whereas the process modelling activities were conducted in Aspen Plus^® V14 (Aspen Technology, Inc., Bedford, MA, USA).

Hazelnut shells were selected as the reference feedstock because they represent an abundant agro-industrial lignocellulosic residue generated during hazelnut processing. Literature data indicate that hazelnut shell accounts for a substantial fraction of the processed nut residue, with approximately 50 wt.% of the total hazelnut production being generated as solid waste [21]. In addition, hazelnut shell has been reported to exhibit favourable properties for thermochemical valorisation, including low moisture content, low ash content, high volatile matter, relevant fixed-carbon content, and a higher heating value in the range typically reported for lignocellulosic solid biofuels. These characteristics make hazelnut shell a suitable case-study biomass for evaluating sorption-enhanced gasification and hydrogen production through a calcium looping-based configuration [22,23].

However, it should be noted that the present analysis is based on a single biomass feedstock and on a single reference gasification temperature. Therefore, the results should not be interpreted as universally representative of all biomass types, since variations in lignocellulosic composition, ash content, moisture, and devolatilization behaviour may affect gasification performance, carbonation efficiency, syngas composition, and hydrogen yield. For this reason, the selected feedstock is considered representative of a relevant class of dry agro-industrial residues, while the extension of the model to other biomass feedstocks is identified as a specific direction for future work.

The experimentally derived pyrolysis composition is used as the primary input to the kinetic model, alongside thermodynamic parameters and sorbent properties. The devolatilization process is assumed to be instantaneous and is described as follows:

$$Bio \to x_1{H}_2+x_{2}CO+x_3{CO}_2+x_4{CH}_4+x_{5}CHAR+x_6{C}_6{H}_6+x_7{C}_7{H}_8+x_8{C}_{10}{H}_8+x_9{C}_6{H}_{5}OH$$

(1)

Experimental data obtained in a bench-scale fluidized bed reactor [24] were used to define realistic values for the pyrolysis products (x1…x9) and to validate the SEG model. The pyrolysis composition refers to hazelnut shells at 670 °C; light hydrocarbons heavier than methane (below 5 wt.%) are lumped into methane.

The model output is then integrated into an Aspen Plus flowsheet, which simulates the overall process, including combustion, steam generation, CO₂ compression, and a hydrogen production pathway.

In this study, a DFB system was considered for the gasification reactor configuration. A typical DFB system consists of two interconnected reactors: a gasifier and a combustor. In the gasifier, drying, pyrolysis, and gasification reactions take place under steam atmosphere, producing a nitrogen-free syngas. The circulating bed material, together with residual char and sorbent, is transferred to the combustor, where oxidation reactions provide the heat required to sustain the endothermic gasification process.

Accordingly, the Aspen Plus^® model includes a combustion section that accounts for the combustion of unreacted char and the regeneration of the sorbent, consisting of CaCO₃ and unconverted CaO. An O₂ stream, calculated within Aspen, is supplied to ensure complete combustion. The resulting flue gas, primarily composed of CO₂, is subsequently compressed and stored.

The syngas produced from SEG, together with tar and ash, is directed to the hydrogen production section which includes hot syngas cleaning and conditioning units followed by a PSA system for hydrogen separation. The PSA tail gas is used as a fuel for both the combustor reactor and steam generation.

Finally, the simulation outputs are post-processed in MATLAB to evaluate the LCOH.

2.1.1. Kinetic Gasification Model

To describe the fluid dynamics of the fluidized bed, an adapted model based on the two-phase approach of Kunii and Levenspiel was used [25]. In this model, the fluidized bed is represented by two interacting phases: the bubble phase and the emulsion phase, which exchange mass through a transition zone governed by a mass transport coefficient. The MATLAB kinetic model consists of coupled mass and energy balances for the bubble and emulsion phases, including fluid-dynamic correlations for bubble behaviour and gas interchange, together with kinetic expressions for homogeneous gas-phase reactions, heterogeneous char gasification reactions, tar reforming reactions, and CO₂ capture by CaO particles. Within the reactor, the solid particles are fluidized, ensuring highly homogeneous mixing characterized by uniform temperature and concentration profiles throughout the bed. Consequently, the solid phase (bed materials and unreacted char) behaviour can be modelled using the equations of a Continuous Stirred-Tank Reactor (CSTR) [26]. Through morphological analyses, it can be noted that the calcined particles are characterized by a void fraction (${\epsilon }_{p,0})$, which quantifies their porosity. For fully calcined particle, the value, ${\epsilon }_{p,0}=0.4$. Calcined dolomite particles consist of CaO grains dispersed within MgO grains (inert under the operating temperatures considered). For this reason, the Particle Grain Model (PGM) [27] was adopted to describe the carbonation kinetics.

In this model, CO₂ reacts at the external surface of the CaO grain, forming a CaCO₃ layer and leaving an unreacted CaO core inside. As conversion proceeds, the lower porosity of CaCO₃ compared to CaO introduces diffusion limitations, reducing the reaction rate. This effect is explicitly accounted for in the formulation of the carbonation reaction rate (R10).

The following reactions are considered in the MATLAB kinetic model for the steam gasification process:

$$C+C{O}_2\to 2CO$$

(2)

$$C+H_{2}O\to CO+H_2$$

(3)

$$C{H}_4+H_{2}O\leftrightarrow CO+3H_2$$

(4)

$$CO+H_{2}O\leftrightarrow C{O}_2+H_2$$

(5)

$$C_6{H}_6+6H_{2}O\leftrightarrow 6CO+9H_2$$

(6)

$$C_{10}{H}_8+10H_{2}O\leftrightarrow 10CO+14H_2$$

(7)

$$C_7{H}_8+7H_{2}O\leftrightarrow 7CO+11H_2$$

(8)

$$C_6{H}_{5}OH+5H_{2}O\leftrightarrow 6CO+8H_2$$

(9)

Reactions (2) and (3) corresponds to the solid phase Boudouard reaction and Water Gas reaction, respectively, while (4) and (5) represent the gas phase reactions (Steam Methane Reforming and Water Gas Shift, respectively), and (6)–(9) describe tar reforming reactions.

As mentioned before, gasification can mitigate CO₂ emissions if integrated with carbon capture technologies such as calcium looping, where CaO represents the regenerable sorbent to remove CO₂ from gas streams through the following reactions:

$$CaO+{CO}_2\to {CaCO}_3$$

(10)

$${CaCO}_3\to CaO+{CO}_2$$

(11)

The carbonation reaction (10) is favoured at approximately 600–700 °C; consequently, when gasification is integrated with this technology, the process is typically conducted at 650 °C. In the calcination reaction (11), the formed CaCO₃ is decomposed to release pure CO₂ at 900–950 °C [28]. All the equations implemented in the MATLAB code are reported in the Supplementary Materials, including fluid-dynamic correlations (Equations (S1)–(S4)), species and energy balances (Equations (S5)–(S10)), kinetic expressions for gasification and reforming reactions (Equations (S11)–(S18)) [29,30,31,32,33], and the carbonation model for CO₂ capture by CaO particles (Equations (S19)–(S21)).

2.1.2. Aspen Plus Model

Combustion Section

Figure 2 shows the Aspen Plus^® simulation flowsheet of the combustion section.

The simulation is designed to close the calcium looping cycle by addressing two main objectives:

• Identify the optimal fuel flow rate for the allothermal operation of the integrated process, while maintaining the combustor temperature above 900 °C for the sorbent regeneration.
• Calculate the oxygen flow rate to ensure complete combustion of char and auxiliary fuel.

Residual char and carbonated sorbent (CaCO₃ together with an unreacted CaO), separated downstream of the gasification section, are fed to the combustion reactor. The char is initially defined as a non-conventional solid; therefore, it is converted into conventional components (C, H, and O) using an RYield reactor (CHARCONV). This approach is widely adopted in Aspen Plus for solids modelling, as it allows equilibrium-based or stoichiometric calculations while preserving the elemental composition derived from experimental characterization [34].

The combustion and calcination processes are simulated in an RGibbs reactor (COMB), operating under adiabatic conditions at 1 bar. The reactor feed includes O₂, and recirculated CO₂ as a diluting and fluidising agent, the residual char from the gasifier (CHARCONV) and the recirculated tail gas (see Figure 1) named FUELIN in Figure 2. The mass flow rate of the tail gas recirculated (FUELIN in Figure 2) was adjusted to have an adiabatic temperature of the reactions (at the outlet of COMB) equal to 980 °C to assure the complete calcination of CaCO₃ to CaO. The oxygen-to-carbon molar ratio (O/C) is set to 1.1. This assumption ensures complete char oxidation and avoids the formation of incomplete combustion products such as CO, which is consistent with experimental observations in well-mixed fluidized bed combustors operating under oxygen-rich conditions [35]. Within the RGibbs reactor, the equilibrium calculation results in complete conversion of residual char, a full regeneration of the sorbent, and a flue gas composed almost entirely of CO₂.

Downstream of the combustor, a solid–gas separator (CYCLONE) is implemented to separate the regenerated sorbent from the flue gas stream. The regenerated CaO is then recycled to the gasification section, closing the calcium looping cycle.

The high-temperature flue gas is partially recirculated (GAS-SEP) to the combustor (stream CO₂-R set equal to CO₂-R3), based on the fluidizing agent requirements and to preheat the O₂, improving the overall thermal efficiency of the process.

All unit operations of the combustion section are summarized in

Table 1. Aspen Plus^® combustion flowsheet unit operation.

Aspen Plus® Name	Block ID	Description
RYIELD	B1	Decomposition reactor: converts the non-conventional char into a conventional component.
RGIBBS	COMB	Combustion and calcination reactor.
SEP	CYCLONE	Gas–solid separators (CaO from flue gas).
SEP	GAS-SEP	Separator for CO2 recirculation.
HEATER	B2	Heat Exchangers.
MIXER	MIXER	Combining O2 with CO2 recirculated.
BLOWER	COMPR	Prevent pressure drop

.

Hydrogen Production

The hydrogen production section is modelled to represent the downstream conditioning required to obtain a high-purity hydrogen stream from the raw syngas. The feed to this section consists of syngas-containing residual and ash before hydrogen production.

Figure 3 shows the Aspen Plus^® simulation flowsheet of the hydrogen production subsystem.

Initially, solid ash is separated from the gas stream using a solid–gas separator (CYCLONE2). Subsequently, the syngas is heated (B3) by recovering sensible heat from the flue gas exiting the calcination section, maintaining a minimum temperature approach of 10 °C between the hot and cold streams. Then, the gas is directed to a stoichiometric reactor (RStoich, TARREF), where heavy hydrocarbons, represented by benzene, toluene, naphthalene, and phenols, are converted into hydrogen and carbon monoxide (Equations (6)–(9)).

The cleaned syngas is then compressed through a multistage compression system (COMPR 1-2-3), with a compression ratio of 2.5, and interstage cooling (B6, B7, B8). This configuration enables a realistic estimation of compression work and heat duties. As a result, the syngas enters the PSA unit at 40 °C and 16 bar.

The PSA unit is modelled as an ideal separator block (PSA), in which hydrogen is selectively recovered as the high-purity product stream, while the remaining components (CO, CO₂, CH₄, and residual light hydrocarbons) are discharged as off-gas. The hydrogen recovery is set to 90% across all scales with product purities of 99.95%. These values are fully consistent with literature reports for PSA systems applied to syngas purification, where hydrogen recoveries of 90% and purities above 99.9% can typically be achieved [36,37,38]. Since the present work focuses on the conceptual process assessment and scale-up analysis of the overall gasification and CO₂ capture-integrated system rather than on PSA optimization, the adopted hydrogen purity and recovery were selected as representative values consistent with industrial PSA performances reported in the literature, while higher values are generally associated with advanced hybrid or multistage purification systems. Although this approach does not account for adsorption kinetics, mass transfer limitations, or cycle optimization, factors that may affect real system performance, it is widely used in conceptual process design and system-level analyses, as it enables a first-order evaluation of hydrogen purity, recovery, and off-gas composition.

Table 2. Hydrogen purity and recovery in the literature.

H2 Purity (%)	H2 Recovery (%)	Reference
~95	Not specified	[36]
99.999	99.7	[37]
>99.999	99.87	[38]
97	77.5	[39]
>99	Not specified	[40]
99.92–99.95	90	This work

compares the hydrogen purity and recovery reported in the literature with those in this work.

As presented in Table 2, recent studies demonstrate that PSA simulations in Aspen Plus and Aspen Adsorption can achieve hydrogen purities from 95% up to 99.999% with recoveries exceeding 99% and even hydrogen purities above 99.999% with an integrated and Electrochemical Hydrogen Purification and Compression configurations (PSA-EHP/C) [38]. Application-oriented studies on syngas from gasification report product purities exceeding 95% [36]. Advanced configurations such as two-stage PSA combined with CO-selective methanation yield H₂ purities > 97% with recoveries in the range of 75–80% [39]. Two-bed PSA simulations in Aspen Adsorption have produced hydrogen purities of ≈99% [40].

The simplified PSA model allows the evaluation of overall process integration, energy consumption, and hydrogen production potential of the proposed gasification-based system, while providing a consistent basis for comparison with alternative process configurations.

All unit operations of the PSA section are described in

Table 3. Aspen Plus^® PSA flowsheet unit operation.

Aspen Plus® Name	Block ID	Description
SEP	CYCLONE2-PSA	Gas–solid separators (CYCLONE2); Rich H2 stream production (PSA).
RGIBBS	TARREF	Tar reformer.
FLASH2	B5–B9	Expands product to 1 bar, separating the vapours and the liquid phase.
HEATER	B3–B4–B6–B7–B8	Heat Exchangers.
COMPR2	COMPR1-COMPR2-COMPR3	Multistage compression of syngas.
MIXER	MIXER1-MIXER2	Combining syngas, ash, and tar.

.

Performance Indicators

To evaluate the effectiveness and efficiency of the process, the following performance indicators are considered.

Cold Gas Efficiency (CGE) evaluates the efficiency of the gasification process as the ratio between the chemical energy of the produced syngas and that of the biomass feedstock:

$$\mathit{CGE}(\%)={\displaystyle \frac{{\dot{\mathit{m}}}_{\mathit{S}\mathit{y}\mathit{n}\mathit{g}\mathit{a}\mathit{s}}\times {\mathit{L}\mathit{H}\mathit{V}}_{\mathit{S}\mathit{y}\mathit{n}\mathit{g}\mathit{a}\mathit{s}}}{{\dot{\mathit{m}}}_{\mathit{B}\mathit{i}\mathit{o}\mathit{m}\mathit{a}\mathit{s}\mathit{s}}\times {\mathit{L}\mathit{H}\mathit{V}}_{\mathit{B}\mathit{i}\mathit{o}\mathit{m}\mathit{a}\mathit{s}\mathit{s}}+{\dot{\mathit{m}}}_{\mathit{A}\mathit{u}\mathit{x}\_\mathit{F}\mathit{u}\mathit{e}\mathit{l}}\times {\mathit{L}\mathit{H}\mathit{V}}_{\mathit{A}\mathit{u}\mathit{x}\_\mathit{F}\mathit{u}\mathit{e}\mathit{l}}}}$$

(12)

where $\dot{m}$ is the mass flow rate (kg/h) and $LHV$ is the lower heating value (MJ/kg).

Hydrogen production efficiency (η) is defined as the ratio between the chemical energy of the produced hydrogen and that of the biomass feedstock:

$$\eta ={\displaystyle \frac{{\dot{m}}_{H_2}\times {LHV}_{H_2}}{{\dot{m}}_{Biomass}\times {LHV}_{Biomass}}}$$

(13)

Carbon Conversion (CC) represents the fraction of carbon in the raw feedstock that is converted into gaseous products:

$$CC \left(\%\right)={\displaystyle \frac{\sum _{i=CO,{CO}_2,{CH}_4}{mol}_i}{{mol}_{C,Biomass}}}$$

(14)

where ${mol}_i$ are the moles of C contained in CO, CO₂, CH₄ in the syngas, and ${mol}_{C,Biomass}$ are the moles of C that enter with the biomass.

Carbon Capture Efficiency (CCE) quantifies the fraction of carbon dioxide captured by the separation process:

$$CCE \left(\%\right)={\displaystyle \frac{{mol}_{C,Ca{CO}_3}}{{mol}_{C,Biomass}}}$$

(15)

In addition to these parameters, the syngas yield (Y) is evaluated:

$$Y ({\mathrm{N}\mathrm{m}}^3/{\mathrm{k}\mathrm{g}}_{Biomass}) ={\displaystyle \frac{{\dot{V}}_{Syngas}}{{\dot{m}}_{Biomass}}}$$

(16)

where ${\dot{V}}_{Syngas}$ is the volumetric flow rate of dry syngas expressed at normal conditions (Nm³/h) and ${\dot{m}}_{Biomass}$ is the biomass flow rate.

Regarding the process carbon footprint evaluation (without utilities), the Emission Index (EI) parameters is evaluated as expressed in Equation (17), where ${\dot{m}}_{CO2,eq,out gas}$ represent the equivalent CO₂ mass flow rate obtained by the complete combustion of the gas out stream and ${\dot{m}}_{H2\_prod}$ represent the amount of H₂ produced [41].

$$EI ({\mathrm{k}\mathrm{g}}_{CO2net}/{\mathrm{k}\mathrm{g}}_{H2})={\displaystyle \frac{{\dot{m}}_{CO2,eq,out gas}}{{\dot{m}}_{H2\_prod}}}$$

(17)

Then, the Net Carbon Footprint (CF_NET) is evaluated. It is defined as the net amount of carbon dioxide-equivalent emissions associated with the process, normalized to the hydrogen production rate. It is calculated as the difference between the direct CO₂ emissions released from the system ($Ai\ast {EI}_i)$ and the CO₂ permanently captured and stored ($R)$, divided by the amount of hydrogen produced [42]. The emissions released are calculated by accounting for the net CO₂-equivalent emissions associated with electricity production and consumption, water consumption and make-up, and biomass supply chain activities within the adopted system boundaries and are reported in Table S1 in the Supplementary Materials [42,43,44,45,46,47,48,49].

$${CF}_{NET}({\mathrm{k}\mathrm{g}}_{CO2eq}/{\mathrm{k}\mathrm{g}}_{H2})={\displaystyle \frac{\sum (Ai\times {EI}_i)-R}{{\dot{m}}_{H2\_prod}}}$$

(18)

EI is reported as a process-level indicator, whereas CF_NET is evaluated as a simplified cradle-to-gate carbon footprint based on the main upstream contributors included in the adopted boundary.

2.2. Economic Analysis

The economic feasibility of the system was assessed through the evaluation of three key economic indicators. The capital expenditure (CAPEX) represents the total investment associated with the procurement and installation of plant components. The operating expenditure (OPEX) encompasses all recurring costs related to operation and maintenance over the plant’s lifetime. The LCOH was selected as the main economic metric, as it integrates both CAPEX and OPEX contributions over the total hydrogen production, enabling a thorough comparison with state-of-the-art hydrogen production technologies.

The CAPEX was estimated following the methodology proposed by [50], as detailed in Equation (8). Starting from the Bare Erected Cost (BEC), the procedure advanced sequentially through the Engineering, Procurement, and Construction Cost (EPCC), the Total Plant Cost (TPC), and the Total Overnight Cost (TOC), ultimately leading to the final CAPEX estimate. The multiplying factors and Equipment Cost (EC) functions adopted in the calculation are summarized in

Table 4. Multiplying factors adopted in the calculation proposed by [50].

Multiplying Factors	Description	Value [%]
${\mathrm{f}}_{\mathrm{O}\mathrm{C}}$	Owner’s costs	20
${\mathrm{f}}_{\mathrm{p}\mathrm{r}\mathrm{o}\mathrm{c}\mathrm{e}\mathrm{s}\mathrm{s}}$	Process contingency	10
${\mathrm{f}}_{\mathrm{p}\mathrm{r}\mathrm{o}\mathrm{j}\mathrm{e}\mathrm{c}\mathrm{t}}$	Project contingency	15
${\mathrm{f}}_{\mathrm{E}\mathrm{P}\mathrm{C}}$	Engineering, procurement, and construction services	17.5

and

Table 5. EC functions and FBM values [50,51].

Component	Cost Function	Currency	Parameter	FBM
Calcinator	$\mathrm{B}\mathrm{E}\mathrm{C}=2.07\times 193000\times {\dot{\mathrm{Q}}}_{\mathrm{t}\mathrm{h}}^{0.65}$	€2014	Thermal power in MWTH	-
Carbonator	$\mathrm{B}\mathrm{E}\mathrm{C}=2.1\times (217000\times {\dot{\mathrm{Q}}}_{\mathrm{t}\mathrm{h}}^{0.65}+3.83)$	€2014	Thermal power in MWTH	-
PSA	$\mathrm{B}\mathrm{E}\mathrm{C}=817000\times {{\displaystyle \frac{{\dot{\mathrm{V}}}_{{\mathrm{H}}_2}}{944}}}^{0.65}$	€2014	Volume flow in Nm3/h	-
TAR reformer	${\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\mathrm{E}\mathrm{C}=3.7957-0.4593\times {\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\mathrm{V}-0.0160\times {\mathrm{l}\mathrm{o}\mathrm{g}}_{10}^2\mathrm{V} \mathrm{c}\mathrm{o}\mathrm{n}\mathrm{V}<1$	$2001	Volume in m3	4.0
	${\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\mathrm{E}\mathrm{C}=4.5587+0.2986\times {\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\mathrm{V}-0.020\times {\mathrm{l}\mathrm{o}\mathrm{g}}_{10}^2\mathrm{V} \mathrm{c}\mathrm{o}\mathrm{n}\mathrm{V}>1$
Heat exchanger	${\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\mathrm{E}\mathrm{C}=3.3444-0.2745\times {\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\mathrm{A}-0.0472\times {\mathrm{l}\mathrm{o}\mathrm{g}}_{10}^2\mathrm{A} \mathrm{c}\mathrm{o}\mathrm{n}\mathrm{A}<10$	$2001	Heat exchange area in m2	6.1
	${\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\mathrm{E}\mathrm{C}=4.3247-0.3030\times {\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\mathrm{A}+0.1634\times {\mathrm{l}\mathrm{o}\mathrm{g}}_{10}^2\mathrm{A} \mathrm{c}\mathrm{o}\mathrm{n}\mathrm{A}>10$
Compressor	${\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\mathrm{E}\mathrm{C}=$2.2897$+$1.3604$\times {\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\dot{\mathrm{W}}$ − 0.1027$\times {\mathrm{l}\mathrm{o}\mathrm{g}}_{10}^2\dot{\mathrm{W}}$	$2001	Mechanical power in kW	7.7

, respectively. Notably, dedicated cost correlations were employed for the calcinator, the carbonator, and the pressure swing adsorption unit [51], whereas for all remaining components the BEC was derived from the corresponding EC and Bare Module Factor ($F_{BM}$), in accordance with Equation (23). Cost actualization to the reference year 2025 was carried out by means of the Chemical Engineering Plant Cost Index (CEPCI), which accounts for both inflation and structural evolution of the chemical process industry, according to the scaling relationship BEC₂₀₂₅ = BEC₂₀₀₁ × (CEPCI₂₀₂₅/CEPCI₂₀₀₁), with CEPCI2001 = 394.3, CEPCI2014 = 576.1, and CEPCI2025= 800 [52]. All costs, originally expressed in US dollars, were converted to euros by applying a yearly average exchange rate of 0.931 €/$.

$$CAPEX=TOC\times 1.154$$

(19)

$$TOC=TPC\times (1+f_{OC})$$

(20)

$$TPC=(EPCC+f_{process}BEC)\times (1+f_{project})$$

(21)

$$EPCC=BEC\times (1+f_{EPC})$$

(22)

$$BEC={\displaystyle {\sum }_{i=1}^{n_{component}}}{F}_{BMi}\times {EC}_i$$

(23)

The OPEX was estimated as the sum of fixed and variable contributions (OPEX = OPEX_fixed + OPEX_var), accounting for operation and maintenance costs and utility costs, respectively. The procurement costs of biomass feedstock, CaCO₃ makeup (assumed at 5% per year [34]), catalyst, adsorbent material for the PSA unit, boiler water required for superheated steam generation in the Heat Recovery Steam Generator (HRSG), and electricity consumption were included in the variable component. The values are reported in

Table 6. OPEX estimate assumptions.

	Account	Value
Fixed OPEX	Specific cost [% CAPEX]	4.5
Variable OPEX	Biomass [€/ton]	60
	CaCO3 [€/ton]	8.0
	Catalyst [€/kg]	30
	PSA adsorbent material [€/kgH2]	0.16
	Boiler water [€/m3]	1.0
	Electricity [€/MWth]	100

.

The LCOH (€/kg) was computed through Equation (24), where $M_{H2}$ represents the annual hydrogen production (kg/year), r denotes the discount rate, set equal to 8% [53], and LT is the plant lifetime, assumed equal to 20 years, in accordance with standard assumptions for chemical process plants.

$$LCOH={\displaystyle \frac{CAPEX+{\displaystyle {\sum }_{i=1}^{LT}}OPEX\times \frac{r{\times (1+r)}^i}{{(1+r)}^i-1}}{{\displaystyle {\sum }_{i=1}^{LT}}{M}_{H2}\times \frac{r{\times (1+r)}^i}{{(1+r)}^i-1}}}$$

(24)

3. Results

3.1. Modelling Results

The MATLAB model was validated using experimental results from SEG tests conducted on a bench-scale gasifier [24].

Table 7. Comparison between experimental data [24], and simulation results.

	Experimental	Simulation
Gas yield (Nm3/kgBiomass)	0.77	0.75
Tar concentration (g/Nm3)	9.5	8.8

presents the comparison between experimental data and simulation results in terms of gas yield and total tar concentration, while Figure 4 compares the composition of the dry, nitrogen-free gas and the tar distribution obtained from both experiments and simulations.

The results demonstrate that the proposed model accurately reproduces the main features of the SEG process, yielding a syngas composition in good agreement with experimental data. The largest deviation is observed in the model’s underestimation of the C₆H₆ and C₇H₈ concentrations. This discrepancy is reflected in the lower total tar concentration predicted in the simulation compared to experimental measurements, with an overall deviation of approximately 8%. In the previous study [24], a limited conversion of methane and tar was observed; enhancing their conversion could further improve gas quality and hydrogen yield. Therefore, the tar reformer is included in the present simulation (TARREF in Figure 3).

Starting from the validated model, three different plant sizes (100 kWth, 1 MWth, and 10 MWth) were simulated and analyzed. This multi-scale approach enables the identification of configurations capable of achieving a LCOH competitive with state-of-the-art hydrogen production technologies.

In accordance with the economies of scale principle, the CAPEX is calculated according to Equation (25) [50].

$$CAPEX=CAPE{X}_{REF}\times {\left({\displaystyle \frac{Scale Parameter}{Reference Parameter}}\right)}^n$$

(25)

where n < 1.

As a result, larger plants benefit from a relative reduction in specific capital costs, making it possible to identify the plant size that minimizes the LCOH and ensures economic competitiveness.

Table 8. Fluid dynamics and operating conditions of the gasification subsystem.

Thermal Input Power (kWth)	100	1000	10,000
τgas (s)	2.1	2.1	2.1
Particle density (kg/m3)	1650	1650	1650
Wet gas (mol/s)	0.45	4.52	46.16
ΧCaO (-)	0.6	0.6	0.6
Generated CaCO3 (mol/s)	0.33	3.32	33.87
Unreacted CaO (mol/s)	0.10	1.03	9.58

shows the fluid dynamics and operating conditions of the gasification subsystem. The simulation is carried out at constant temperature (670 °C), pressure (1 bar), steam-to-biomass ratio (S/B = 1) and residence time (τ_gas) defined as the ratio between the reactor volume and syngas volumetric flow rate.

The figures below report the dry syngas (Figure 5a) and tar composition (Figure 5b).

The dry syngas composition obtained by the simulation depends on the gasifier scale, as reported in Figure 5a. At 100 kWth, H₂ reaches 80%, with CH₄ 15%, CO 3%, and CO₂ 3%. At 1 MWth, H₂ slightly increases to 81%, while CO₂ increases to 4% and CO remains low (2%). At 10 MWth, hydrogen remains stable to 80%, CO₂ increases to 7%, CO remains stable to 3%, and methane decreases to 11%.

Despite small differences in syngas composition, the model predicts stable performance during scale-up. Under real operating conditions, various fluid-dynamic phenomena may arise during scale-up, potentially leading to reduced process efficiency. The parameters that influence gas production and quality in this model are gas velocity, bubble size, and bubble rise velocity. High gas velocities (u₀) lead to the formation of large bubbles that rise rapidly, thereby reducing the contact time between the gas and solid phases. Consequently, if the bubble rise velocity ($u_B$) is excessively high, the model predicts a decrease in efficiency as the gas escapes before reacting. Furthermore, an increase in bubble diameter (d_B) results in a reduced surface area for mass transfer between the bubble and emulsion phases. This reduction in exchange area hinders the gasifying agent from reaching the solid particles, leading to lower conversion rates [25]. According to Davidson’s theory [54,55], the size of the cloud surrounding the bubble—the primary zone for gas exchange—depends on the ratio between the bubble velocity ($u_B$) and the emulsion gas velocity (u_e). If this region becomes excessively thin, the mass transfer is reduced. Ideally, the bubble velocity should be approximately five times the emulsion gas velocity [25]. During scale-up, the fluid dynamics undergo significant changes that affect hydrogen yield. In wider beds, bubbles tend to coalesce, becoming larger and faster; this decreases the surface-to-volume ratio and impairs mass transfer. To control the growth of bubbles and to promote their breakup, various types of internals are frequently used in large-scale fluidized beds. Specifically, horizontal perforated baffles are employed to periodically redistribute the gas phase and fragment large voids, thereby maintaining a high interfacial area for mass transfer. Similarly, the installation of tube bundles, while often serving as heat exchange surfaces, provides a dual benefit by acting as physical barriers that limit maximum bubble size. Furthermore, the use of screens or grids at various bed heights further assists in mitigating the development of large bubbles [25,56,57].

Under ideal sorption equilibrium conditions, dry H₂ concentrations between 70 and 85% are commonly reported in SEG systems operating in the 600–700 °C range [58,59]. Compared to conventional DFB gasifiers without CO₂ capture, which typically yield 35–45% of H₂, 20–25% of CO, 20–25% of CO₂, and 8–12% of CH₄ [12], the results clearly show the substantial hydrogen enhancement achieved via in situ CO₂ sorption. The methane fraction (10–13%) across all scales is consistent with the moderate operating temperature (650 °C), where steam reforming is not fully promoted compared to high-temperature DFB systems (750–850 °C), in which CH₄ concentrations are generally lower [60].

The tar distribution is coherent with fluidized bed gasification at intermediate temperature, where thermal cracking reactions are less severe than in catalytic or high-temperature configurations [61,62]. The slight differences in heavier aromatic fractions at different scales may reflect the differences in residence times, as highlighted in Table 8.

In a calcium looping system, the combustor also serves as the regeneration unit for the sorbent. The char is fed to the combustion subsystem together with the carbonated dolomite, oxygen, and recirculated CO₂. The high-temperature flue gas, mainly CO₂, is partially recirculated to the combustor to have a superficial velocity equal to 10 times the u_mf in the combustor to assure an efficient bed material recirculation as reported in the work [63] for a 100 kWth DFBG as biomass input. For the simulation of the 100 kWth gasifier, therefore, the same size (diameter) of the combustor reported in the work [63] was used to determine the volumetric flow rate (O₂ and recirculated CO₂) at the superficial velocity selected. For the simulations of the scaled-up gasifiers (1 MWth and 10 MWth) the volumetric flow rates were scaled accordingly. As a result, the gas mix of O₂ and CO₂ entering the reactor had a composition comparable to that of air, in which CO₂ replaces N₂.

The reactor outputs include hot flue gas and calcined dolomite. The absence of CO in the flue gas further confirms that the operating conditions and oxygen supply are sufficient to achieve complete combustion. In a DFB system, the bed material plays a fundamental role as a heat carrier between the combustor and the gasifier [64]. Heat required for the endothermic gasification reactions is not generated by partial oxidation but indirectly supplied via the circulation of hot bed solids. In the combustion reactor, residual char is oxidized, producing exothermic heat that increases the temperature of the bed material.

The obtained syngas is finally directed to the PSA subsystem to produce pure hydrogen. As previously mentioned, a limited conversion of tar was observed. Before entering the PSA system, the syngas is enriched via a tar reformer, which ensures high-purity hydrogen recovery while significantly enhancing overall process efficiency, and preventing downstream fouling and catalyst poisoning within the PSA unit.

The output data obtained from the PSA unit is shown in

Table 9. PSA unit input and output.

Thermal Input Power (kWth)	100	1000	10,000
Feed
Mole Flow (kmol/h)	1.23	11.78	124.83
Mole Fraction
H2	0.81	0.79	0.76
CO	0.05	0.06	0.06
CO2	0.08	0.10	0.12
CH4	0.03	0.03	0.03
H2O	0.03	0.03	0.03
Tail gas
Mole Flow (kmol/h)	0.31	2.91	17.60
Mole Fraction
H2	0.33	0.29	0.25
CO	0.20	0.21	0.21
CO2	0.33	0.36	0.39
CH4	0.11	0.11	0.11
H2O	0.03	0.03	0.03
Product
Mole Flow (kmol/h)	0.90	8.33	85.42
Mole Fraction
H2	0.9995	0.9995	0.9995
CO	7.07 × 10−6	8.16 × 10−6	9.2 × 10−6
CO2	1.12 × 10−5	1.41 × 10−5	1.7 × 10−5
CH4	3.74 × 10−4	4.4 × 10−4	5.0 × 10−5
H2O	0.00	0.00	0.00
Gas Out
Mole Flow (kmol/h)	0.00	0.57	19.54
Mole Fraction
H2	0.00	0.29	0.25
CO	0.00	0.21	0.21
CO2	0.00	0.36	0.39
CH4	0.00	0.11	0.11
H2O	0.00	0.03	0.03

.

The tail gas composition, characterized by 24–33% residual H₂ and 33–37% of CO₂, aligns with industrial PSA applications, where the off-gas is typically recycled to the combustion reactor of DFB systems for energy recovery.

The Sankey diagrams of the system mass flow rate for the three scenarios considered are shown in Figure 6, Figure 7 and Figure 8.

The gasification subsystem receives the biomass feedstock together with the required process water (S/B ratio equal to 1) and the sorbent CaO-based for the CO₂ capture, producing syngas directed to the PSA subsystem, char, and various by-products. These represent the inputs for the model developed in Aspen.

As presented in the Sankey diagram, the total sorbent requirements for CO₂ capture in the carbonator amount to 87.7 kg/h, 882.2 kg/h, and 8772.1 kg/h, respectively, for the three scenarios. Given the amount of biomass fed, the CaO-to-biomass ratio remains constant at approximately 4.3–4.4 kg_CaO/kg_Biomass, indicating that the scale-up was performed under proportional sorbent loading conditions, thus representing strongly intensified in situ CO₂ capture conditions.

To ensure the allotermicity of the gasification and capture process and to increase overall efficiency, it is necessary to use a small amount of PSA-produced gas. As can be observed from the figures, all the off-gases are required only in the 100 kWth case, together with a small fraction of H₂ (9.6 mol/h), to ensure sorbent regeneration and the consequent CO₂ capture. The flue gas generated in the combustor is subsequently separated, with a fraction recirculated to ensure proper reactor fluidisation.

3.2. Performance Indicators Results

Table 10. Main process erformance indicators.

Indicator	100 kWth	1 MWth	10 MWth
CGE_Gasification	63.48%	69.67%	69.21%
η	59.08%	58.79%	57.63%
CC	19.09%	19.17%	19.53%
CCE	49.83%	48.36%	45.08%
EI (kgCO2,eq/kgH2)	0	1.02	3.57
CFnet (kgCO2,eq/kgH2)	0.39–0.77	1.49–1.95	4.03–4.49
Y_Syngas (Nm3/kgBiomass)	0.842	0.878	0.919
Y_H2 (Nm3/kgBiomass)	1.009	0.989	0.969

shows the indicator values.

The CGE represents the fraction of the inlet energy that is converted into the chemical energy of the produced syngas. The relatively high conversion values observed are also due to internal energy recovery within the system. Since the overall gasification–combustion process operates under allothermal conditions, steam generation requires a significant energy demand.

Table 11. Energy requirements for steam production.

	100 kWth	1 MWth	10 MWth
Economizer (kWth)	4.86	43.63	193.32
Evaporator (kWth)	10.08	106.59	1300.44
Superheater (kWth)	3.28	33.01	328.27

reports the energy requirements for steam production under the three investigated scenarios.

These components represent the three main stages for vapour generation: heating the liquid water (economizer), change in state (evaporator), and increasing the temperature of dry steam beyond the saturation point (superheater).

In a system without internal heat recovery, the energy required for steam generation would be supplied by an auxiliary fuel. This would constitute an additional energy input to the process and would therefore directly affect the overall system efficiency. In the present study, thermal integration strategies were implemented to eliminate the use of auxiliary fuel and enhance the system’s overall thermal performance.

The sensible heat of the flue gas (980 °C) exiting the combustion section is first utilized to superheat the syngas upstream of the tar reformer. This heat exchanger was designed to ensure a minimum temperature driving force of 10 °C between the flue gas outlet temperature and the syngas inlet temperature. As a result, the flue gas temperature decreases to 690 °C. Its remaining sensible heat is then employed for steam superheating up to 400 °C and subsequently in the economizer to preheat the boiler feed water to its saturation temperature. At this stage, only the vaporization remains to be supplied, which, as shown in Table 11, represents the main energy requirement for steam production.

Downstream of the tar reformer, the syngas is first cooled and then sent to a flash unit to remove condensed water before the compression stage of the PSA (see Figure 3). The sensible heat of the syngas is partially recovered and used to support water vaporization. However, to guarantee a minimum temperature driving force of 10 °C, the available syngas sensible heat alone is insufficient to complete the vaporization process.

A high-energy output stream of the system is the PSA tail gas. As shown in the Sankey diagram, in the 100 kWth scenario, the tail gas is entirely directed as an energy carrier to the combustor section, together with a small fraction of the produced hydrogen (0.013 kg/h). In this case, a limited amount of auxiliary fuel is still required to meet the total steam generation demand. In contrast, in the 1 MWth and 10 MWth scenarios, a portion of the tail gas not required in the combustor is used to provide the additional energy necessary to complete the water vaporization step. With this thermal integration strategy, auxiliary fuel was eliminated in the 1 MWth and 10 MWth scenarios, thereby improving the system’s overall thermal efficiency. The slight discrepancy across the scenarios is because scaling up influence syngas composition and modify efficiency, as presented before.

In the literature, CGE values typically range from approximately 50% to above 75%, depending on the process configuration, reactor technology, and operating conditions. For instance, ref. [65] reported CGE values ranging from 51.8% to 71.4% in a 200 kWth fluidized-bed gasifier with energy recovery from solid residues and syngas. Even in the work of [66], where the results obtained during gasification test are integrated to feed a Combined Heat and Power (CHP)-based system, the CGE is approximately of 66%. The study of [67] reported a CGE value of 61.6% in a Computational Fluid Dynamics (CFD) model for biomass gasification in fixed-bed reactors for hydrogen production.

In this study, the primary objective is hydrogen production, which results in low-carbon utilization. This parameter represents the fraction of carbon in the biomass converted into products such as methane or methanol. In the present case, carbon is almost entirely stored as CO₂, while the remain is found as unreacted char and light carbon-based compounds in the syngas.

In addition to the Carbon Capture Efficiency (CCE), which quantifies the fraction of carbon dioxide captured within the process, a specific emission intensity (EI) normalized to hydrogen production was evaluated to provide a process-level indication of the net carbon burden associated with the investigated system. Under the adopted system boundaries, the calculated EI values are 1.02 kg_CO2,eq/kg_H2 for the 1 MWth case and 3.57 kg_CO2,eq/kg_H2 for the 10 MWth case, suggesting that the intermediate scale exhibits the most favourable balance between carbon capture performance and residual emissions. However, it should be emphasized that this indicator is limited to the modelled process boundary and does not constitute a full life-cycle assessment, as upstream and downstream auxiliaries were not included in a complete LCA framework. The higher EI at 10 MWth is mainly attributable to the reduced use of PSA tail gas for internal energy integration, and also a small contribution of the capture of CO₂ which depends on lower flue-gas capture performance in the carbonator, changing from 48.36% to 45.08% at 1 MWth and 10 MWth respectively. The above consideration is bonded to the condition that all the CO₂ contained in the flue gas is supposed to be stored adequately in a long-time or permanent storage site.

When the auxiliary contributions were also included through the simplified net carbon footprint formulation, the resulting CF_net values ranged from 0.39 to 0.77, 1.49–1.95, and 4.03–4.49 kg_CO2,eq/kg_H2 for the small-, intermediate-, and large-scale cases, respectively; the auxiliary burden is largely dominated by electricity consumption, while the contributions of water and biomass supply remain comparatively limited. Nevertheless, this CF_net should be regarded only as a partial indication of the environmental performance of the system and the impact of the produced hydrogen, since it does not represent a complete life-cycle assessment.

Table 12. Hydrogen production and removal efficiency in the literature.

Gasification Technology	Hydrogen Production Efficiency	PSA Removal Efficiency	Ref.
Battelle Columbus Laboratory	55.5 (HHV)	0.8	[68]
Downdraft fixed-bed gasifier	63.8 (LHV)	0.95	[69]
SES technology	49.6–65 (LHV)	0.95	[70]

reports a comparison between the findings of this paper and some literature results in terms of hydrogen efficiency.

In the study of [68], a gasification process was modelled in Aspen Plus, using correlations derived from Battelle Columbus Laboratory operational data and integrated with PSA technology for syngas upgrading. Conversely, the work of [69] implemented biomass gasification within a downdraft reactor using oxygen-enriched air and steam; following gas cleaning, the syngas is fed into a PSA unit for H₂ production. Finally, study [70] employs Synthesis Energy Systems (SES) technology, which is capable of pressurized operation and handling feedstocks with high moisture and ash content. This study evaluates three scenarios of hydrogen production integrated with carbon capture: cases A and B differ in the thermal energy supplied to the steam reformer, while case C is developed without carbon capture, with heat provided externally. Across these scenarios, the removal efficiency significantly affects the performance: a 75% efficiency yields a 51.3% H₂ recovery, whereas increasing the efficiency to 95% raises the productivity to 65%, underscoring the criticality of this parameter. Furthermore, in case B, a portion of the internally produced hydrogen is used as fuel for the steam reformer, leading to a further efficiency reduction to 49.6%. This trend is also pronounced in the other works; the first study yields 55.5 MJ/kg, based on the HHV, which decreases when the latent heat of vaporization is neglected, and in the second work, where a 95% removal efficiency results in overall efficiencies of up to 63%. So, the differences across this study can be attributed both to the PSA removal efficiency and to the scaling effects that influence the overall efficiency.

3.3. Economic Results

In this section, the economic results are presented and discussed. Figure 9 reports the boundaries defined for the techno-economic analysis, and Figure 10 reports the indicators and their breakdown into the main plant sections.

The CAPEX exhibits a power-law dependence on plant capacity $\dot{W}$ ($\mathrm{C}\mathrm{A}\mathrm{P}\mathrm{E}\mathrm{X}\approx {\dot{W}}^n$), consistent with the well-established economies of scale in process plant construction. Between the demonstrative and industrial configurations (4.8 and 20 M€, respectively), a scaling exponent of n ≈ 0.6 is obtained, which aligns with values typically reported for chemical process industries [51,71]. A significantly lower exponent of n ≈ 0.3 is observed between the experimental (2.6 M€) and demonstrative capacities, reflecting the disproportionately high specific equipment costs of very small-scale components. Among the CAPEX contributions, the syngas treatment section consistently represents the largest share, ranging from 47% to 58% of CAPEX across all plant sizes. While the elevated share at the experimental scale was due to the very high cost of small-scale compressors (≈1 kWth), at the demonstrative and the industrial scales it is driven by the PSA unit, reflecting the greater amount of hydrogen to be treated compared to the 100 kWth capacity. For similar reasons related to the absence of standardized equipment at small size, the heat exchangers contribution decrease from 25% to 2% of CAPEX across the scale-up from 100 kWth to 10 MWth. The OPEX amounts to 0.14, 0.43 and 3.0 M€/y for the 100 kWth, 1 MWth and 10 MWth plants, respectively. Within the OPEX, the fixed component decreased progressively from 85% to 31% of CAPEX for the smallest and the largest scale, respectively, consistent with the reduction in specific capital expenditure, as it was estimated as a fixed percentage of the CAPEX (see Table 5). Among the variable contributions, the gasifier represented the most significant one, which increased with the plant size, predominantly driven by biomass feedstock consumption. The syngas treatment represented the second most relevant variable cost, increasing with plant size as a result of the greater hydrogen production and the consequent costs associated with PSA operation. The syngas treatment followed as the third largest cost, growing from 2.4% to 13% of OPEX from the 100 kWth to 10 MWth scales, as a result of greater hydrogen production and the consequent costs associated with PSA operation. Finally, the LCOH reflected the trends observed for both CAPEX and OPEX across the investigated scales. At the industrial scale, an LCOH of 4.54 €/kg_H2 was achieved, proving competitive against low-carbon hydrogen production technologies. The demonstrative scale yielded an LCOH of 8.77 €/kg_H2, slightly exceeding the benchmark market prices. At the experimental scale, the LCOH reached 41.3 €/kg_H2, substantially penalized by the high specific CAPEX inherent to very small-scale components, resulting in values far beyond economic feasibility.

Potential valorization of the captured CO₂ stream through compression and subsequent sale was evaluated at the industrial capacity. To this end, the CO₂ compression cost was incorporated into both the CAPEX and the OPEX of the calcination section (see Figure 11), accounting for equipment procurement and installation costs and the associated electricity consumption, respectively. CO₂ compression from 1.00 to 81.0 bar required approximately 300 kW, resulting in a 49% increase in LCOH, reaching 6.77 €/kg_H2. This value exceeds the cost range reported for SMR with CCS technology, owing to the smaller size of the compressors compared to those typically employed in large industrial facilities, resulting in elevated specific capital costs. The obtained LCOH value at industrial scale of 6.8 €/kg_H2, including CO₂ compression cots, places hydrogen production from the SEG + PSA system within the range reported for conventional biomass gasification processes (3–14 €/kg_H2) and comparable to green hydrogen production pathways based on renewable-powered electrolysis, typically ranging between 6 and 13 €/kg_H2. However, the obtained value remains higher than those generally reported for SMR (0.9–3.7 €/kg_H2) and SMR coupled with CO₂ capture (1.7–5 €/kg_H2) [17], mainly due to the higher capital costs and process complexity associated with biomass handling, gasification, and integrated CO₂ capture systems. Nevertheless, the proposed configuration offers the advantage of combining renewable feedstocks with integrated carbon capture and storage, contributing to low-carbon hydrogen production with reduced fossil fuel dependence.

The economic performance is inherently subject to market and policy uncertainties. To unravel their influence on the competitiveness of the proposed technology, a sensitivity analysis was performed, evaluating the impact on the LCOH of the most volatile parameters: CEPCI, biomass price, electricity cost, discount rate, sorbent makeup rate, and H₂ recovery. Each parameter was varied independently around its reference value, assuming no significant correlations among them. The lower and upper bounds were set to the same relative variation in both directions around the reference value, based on literature, ensuring a fair comparison between the impact of an increase and a decrease of each parameter on the LCOH. Specifically, the CEPCI was varied between 580 and 1020, where the minimum reflects the actual variation observed between 2020 and 2025 [72] and assuming a symmetric future variation of the same magnitude. The biomass price spanned between 30 and 90 €/t, consistent with the range reported in the literature for biomass feedstocks suitable for gasification [73]. The electricity cost range was selected based on the minimum and maximum prices registered in Europe between 2020 and 2025, excluding the out-of-range peak of 2022 when prices exceeded up to five times the decade average [74]; the lower bound of 10 €/MWh reflected prices recorded in Spain, Sweden and Finland, while the upper bound of 190 €/MWh referred to prices observed in Italy, Romania and Austria. The discount rate variation was set between 5 and 11% [75], and the sorbent makeup rate between 3 and 7%, in line with values reported for calcium looping processes [76]. Finally, the H₂ yield ranged between 85 and 95%, based on H₂ recovery values from PSA technology reported in the literature (see Table 2). The results are presented in Figure 11 in the form of a tornado chart. Notably, the y-axis reports the examined parameters along with their respective variation ranges, while the x-axis shows their influence on the LCOH with the resulting percentage variation labelled on each bar. Among the examined parameters, H₂ yield demonstrates the strongest influence on economic performance, with a nearly proportional but opposite effect on the LCOH: a decrease in H₂ recovery penalizes the LCOH slightly more (i.e., +6%) than the same increase benefits it (i.e., −5%). The second most influential parameter is the CEPCI, positively correlated with the LCOH, with an impact that decreases with increasing plant size, ranging from ±26% at the experimental scale to ±19% at the industrial scale, reflecting the economies of scale. The discount rate similarly shows a positive correlation with the LCOH, with an impact ranging from ±6% to ±9%, decreasing with increasing plant capacity. Conversely, the sorbent makeup rate, electricity price, and biomass price show a positive influence on the LCOH that increases with plant size, reflecting the higher feedstock and energy consumption at larger scales. Notably, their individual impact ranges from less than ±1% at the experimental scale to approximately ±8% at the industrial scale.

4. Conclusions

This work demonstrates the capability to integrate detailed kinetic modelling with process simulation tools to assess biomass gasification systems. The coupling between the MATLAB^®-based kinetic SEG model and the Aspen Plus^® process enables a direct link between reactor-scale behaviour and plant-scale performance. The validated model predicts SEG performances across operating conditions and scales, accounting for fluid-dynamic and kinetic effects on syngas composition, while its integration in Aspen Plus^® extends the analysis to the overall process chain, including gas cleaning, upgrading, and thermal integration, thus supporting a consistent scale-up evaluation. From a process standpoint, the results confirm that SEG can produce a hydrogen-rich syngas (~80% dry basis), which can be effectively upgraded to high-purity hydrogen, with stable overall process performance across scales. The techno-economic analysis highlights the critical role of scale in determining process feasibility. The expected LCOH decreases significantly from 41.34 €/kg at 100 kWth to 8.77 €/kg at 1 MWth and 4.54 €/kg at 10 MWth, driven by the reduction in specific CAPEX and improved distribution of operating costs. At larger scales, the obtained LCOH falls within the range typically reported for biomass-based systems with carbon capture and green hydrogen production routes, while remaining higher than conventional grey and blue hydrogen pathways. Furthermore, the inclusion of CO₂ compression for transport, storage, or utilization purposes increased the industrial scale LCOH from 4.54 to 6.77 €/kg_H2; however, this increase does not alter the overall positioning of the technology relative to other biomass-based and renewable hydrogen production pathways. In addition, the Carbon Capture Efficiency, specific emission intensity and preliminary carbon footprint indicate that the environmental performance is also scale-dependent, with a possible intermediate configuration showing the most favourable balance under the adopted system boundaries. However, the results presented in the text should be interpreted as process-level indicators rather than as evidence of full life-cycle carbon neutrality. Overall, the results highlight SEG as a promising and scalable pathway for renewable hydrogen production from lignocellulosic biomass residues. In particular, the use of hazelnut shell as a representative agro-industrial residue demonstrates the applicability of the proposed modelling framework to biomass streams. The integrated modelling approach developed in this work therefore provides a robust basis for future process optimization studies and techno-economic assessments of sorption-enhanced gasification systems.