Assessing the Theoretical Biohydrogen Potential from Agricultural Residues Using Togo as an Example

Abstract

Hydrogen is key to achieving a net-zero carbon future, yet current production remains predominantly fossil-based. Biohydrogen derived from agricultural residues represents a sustainable alternative aligned with circular economy principles. While several studies have assessed the bioenergy potential from agricultural residues in various African countries, their potential in Togo remains largely unexplored. This study employed an exploratory mixed-methods approach to quantify residue availability, evaluate production pathways, and estimate potential biohydrogen yields. Secondary data on crop production from the Food and Agriculture Organization (FAO) and theoretical conversion factors were used to assess the availability of agricultural residues from the eight major crops in Togo, resulting in a residue potential of 7.95 million tons per year. Considering ecological and competing aspects of residue utilization, a sustainable share of 3.1 to 6.6 million tons was estimated to be available for biohydrogen production, depending on the residue recoverability assumptions. A multi-criteria decision analysis (MCDA) was used to evaluate different biohydrogen production processes, identifying dark fermentation as the most suitable due to its low energy requirements and decentralized applicability. The theoretical biohydrogen potential was estimated at 20,991–42,293 tons per year (2.5–5.1 PJ per year) based on biochemical residue composition data and stoichiometric calculations. This study established a baseline assessment of biohydrogen potential from agricultural residues in Togo, offering a methodological framework for assessing biohydrogen potential in other regions. The results also underscore the need for site-specific data to reduce uncertainty and support evidence-based energy planning.

1. Introduction

Hydrogen is central to the global drive toward a net-zero carbon future [1]. Its versatility allows it to be utilized across multiple industries, making it a crucial component of modern economies. In refining, hydrogen is essential for hydrocracking and desulfurization, processes that convert crude oil into usable fuels. The chemical industry also relies on hydrogen, particularly for ammonia synthesis, a fundamental step in fertilizer production [2,3,4]. Beyond these conventional applications, hydrogen is increasingly recognized as a vital driver of renewable energy integration and decarbonization, particularly in hard-to-abate sectors such as heavy-duty transport and steel production [3,4]. Additionally, its role in fuel cells and the production of hydrogen derivatives (e.g., methanol, synthetic fuels) positions it as a key pillar of global climate strategies [5]. However, hydrogen’s environmental impact is determined by its production method. Realizing its full decarbonization potential requires shifting away from fossil-based production [6].

Despite growing interest in low-emission hydrogen technologies, fossil-based methods continue to dominate global production. In 2023, about 99% of the ~97 million tons (Mt) [1,2] of hydrogen produced worldwide stemmed directly or indirectly from fossil fuels, resulting in roughly 920 Mt of CO₂ emissions, comparable to the combined annual emissions of Indonesia and France [7]. Low-emission pathways such as water electrolysis (via renewables) and natural gas reforming with carbon capture collectively account for only 0.7% of production (2022) [2], underscoring the urgent need for alternative solutions.

Biohydrogen offers one such alternative [8]. Biohydrogen refers to hydrogen produced either directly from biomass or through biological processes [9]. It is a renewable and sustainable alternative to conventional fossil fuel-based hydrogen, as it can be produced from various organic feedstocks such as agricultural residues and organic municipal waste, simultaneously solving waste management problems and providing a renewable and environmentally friendly resource for energy production [10,11]. From an environmental standpoint, biohydrogen can achieve near net-zero emissions, given that the CO₂ released during processing is part of the natural carbon cycle [9,11,12,13]. In practical terms, it also enables efficient utilization of waste streams, aligning with circular economy principles by converting residues into a high-value energy carrier [10,11]. Although biohydrogen is promising, the production processes face challenges in terms of cost, scalability, and reaction yield [13].

Across Africa, substantial research highlights the bioenergy potential of agricultural residues, demonstrating their capacity to reduce reliance on imported fossil fuels and tackle energy poverty [14]. For instance, bioethanol-based electricity from crop residues in Ghana could meet up to 91.2% of the nation’s electricity demand [15], while Uganda’s annual 149 PJ crop-residue energy potential [16] equates to 21% of its final energy consumption [17]. Despite these promising developments, Togo remains relatively understudied. Agriculture is a mainstay of Togo’s economy [18], employing around 31% of the population and contributing ~21% to GDP [19]. Effectively tapping into these residues for biohydrogen could advance Togo’s renewable energy goals while addressing waste management challenges [20].

Accurately quantifying these residues and evaluating their biohydrogen potential are vital steps for low-emission hydrogen initiatives and for strengthening Togo’s renewable energy portfolio [20]. Yet data on residue availability and technical feasibility remain limited, hampering large-scale deployment.

Against this backdrop, the present study pursues three objectives:

• Estimate the quantities and availability of agricultural residues in Togo.
• Identify the most suitable biohydrogen production process.
• Quantify the theoretical biohydrogen potential from these residues.

By addressing these objectives, we provide a baseline for future research on biohydrogen in Togo and contribute practical insights and recommendations for further studies. While the focus is country-specific, the approach, particularly the stoichiometric yield estimation and residue assessment, offers a transferable framework applicable to other countries with similar agricultural profiles. In doing so, this work supports broader global efforts to diversify renewable energy sources and implement decentralized, sustainable hydrogen strategies in line with international climate goals such as the UN Sustainable Development Goals.

2. Materials and Methods

2.1. Study Area

Togo is located on the southern edge of West Africa and borders Ghana to the west, Burkina Faso to the north, and Benin to the east. It extends mainly between 6° and 11° north latitude and 0° and 2° east longitude, covering an area of 56,785 km², and is administratively divided into five regions: Maritime, Plateaux, Centrale, Kara, and Savanes [21]. Figure 1 illustrates the administrative divisions of Togo. The capital Lomé is located in the south and serves as the nation’s economic hub, featuring a major seaport [22].

The estimated population of Togo in 2022 was 8.8 million [24], with around 31% employed in the agriculture sector in total employment (2021) [19]. Agriculture contributes around 21% to the national GDP [19], underscoring its economic significance [18]. The country experiences diverse climatic conditions, with a tropical humid climate in the south and a drier, semi-arid climate in the north [22]. These climatic variations influence agricultural practices and crop distribution, affecting the availability and composition of agricultural residues [18].

2.2. Data Collection

Due to the absence of a dedicated database for agricultural residues in Togo, residue quantities were estimated using secondary crop production data combined with theoretical factors. Crop production data were obtained from the FAOSTAT database of the Food and Agriculture Organization of the United Nations [19]. While FAOSTAT data are widely used, they may be subject to reporting lags or regional uncertainties [25]. Theoretical factors, such as the Residue-to-Product Ratio (RPR) and biochemical composition data, were obtained from peer-reviewed studies and specialized databases such as Phyllis [26].

The study focused on the eight most significant crops by production weight in 2022, as reported by FAOSTAT. Residues were categorized based on their stage of occurrence:

• Primary residues—Generated on farms during harvesting.
• Secondary residues—Produced during post-harvest processing [27].

Residues primarily generated at the consumer level, such as yam peel, were excluded due to their widespread dispersion, making collection impractical.

Table 1. The eight major crops by production weight and their associated residues.

No.	Crop	Types of Residue
1	Cassava	Stalk, Peel
2	Yam	Straw
3	Maize	Stalk, Cob, Husk
4	Oil Palm	Empty fruit bunches (EFBs), Kernel Shells, Fibers, Fronds
5	Sorghum	Stalk, Husk
6	Soybean	Straw, Pod
7	Bean, dry	Straw
8	Rice	Straw, Husk

provides an overview of the eight selected crops and their 17 associated residues.

2.3. Residue Potential Assessment

To estimate the quantities of agricultural residues in Togo, this study applied theoretical factors to the crop production data from FAOSTAT [28]. The assessment was conducted in two steps:

• Gross residue potential (GRP) calculation, which represents the total potential amount of residues generated.
• Sustainable residue potential (SRP) calculation, which accounts for ecological constraints that limit residue availability.

2.3.1. Gross Residue Potential

The GRP for each crop was calculated using Equation (1):

$${\mathrm{GRP}}_{\mathrm{r}} ={ \mathrm{P}}_{\mathrm{i}} \times {\mathrm{RPR}}_{\mathrm{r}},$$

(1)

where GRP_r represents the gross residue potential of residue r (t/year), P_i represents the annual production (P) of crop i (t/year), and RPR_r is the Residue-to-Product Ratio of residue r.

The RPR method is widely recognized as a reliable approach for estimating agricultural residue quantities [14,15,29,30]. However, the RPRs are subject to significant variability due to differences in crop varieties, cultivation practices, and harvesting methods. Even for the same crop, RPR values can fluctuate significantly, introducing uncertainties into residue potential assessments [30].

2.3.2. Sustainable Residue Potential

Not all residues are available for biohydrogen production due to their ecosystemic importance, technical constraints and competing uses (e.g., soil conservation, animal feed, or biofuel) [31,32,33,34]. To address this, a Residue Recoverability Factor (RRF) was introduced, adjusting GRP values based on recoverability constraints. The SRP was calculated using Equation (2):

$${\mathrm{SRP}}_{\mathrm{r}} ={ \mathrm{GRP}}_{\mathrm{r}} \times \mathrm{RRF},$$

(2)

where SRP_r is the theoretical annual sustainable residue potential available of residue r (t/year) and RRF is the Residue Recoverability Factor.

To account for uncertainties in residue availability, this study defines three recoverability scenarios for field-based residues:

• Lower-bound scenario (RRF = 0.25)

The lower-bound scenario is based on the studies of Smeets et al. [35] and Berndes et al. [36] and assumes only 25% of field residues can be sustainably recovered. It accounts for field-level technical constraints, the need to maintain soil organic carbon, and competing uses such as livestock feed and bedding. This value is widely used in global bioenergy assessments to avoid overestimation in data-scarce contexts.

2.

Reference scenario (RRF = 0.50)

The reference scenario draws on studies such as Scarlat et al. [32] and the EU BioBoost project [37], which report sustainable removal rates of 40–50% depending on crop type. These values align with several other studies including the one from Glassner et al. [38] which define sustainable removal rates between 30 and 60% depending on the crop type. These figures account for soil health, harvest losses, and limited suitability of some residues for reuse. A 50% RRF is therefore a balanced estimate reflecting both agronomic sustainability and technical feasibility and is used as the reference scenario in this study.

3.

Upper-bound scenario (RRF = 0.80)

This upper-bound scenario is based on Esteban et al. [39], who estimate that up to 80% of cereal straw can be recovered when only 20% must remain in the field (10% for soil quality and 10% for losses). This scenario represents a theoretical technical maximum and is used here primarily for sensitivity analysis and does not imply universal sustainability.

The selected crops, their associated residue types, Residue-to-Product Ratios (RPRs), and assigned RRF values are listed in

Table 2. Overview of crop residues with their respective Residue-to-Product Ratio and assigned Residue Recoverability Factor.

Crop	Crop Residue	RPR (g/g)	Reference	Field-Based	Process-Based	RRF (g/g)
Cassava	Stalk	0.13	[40]	x		0.5
	Peel	0.34	[31]		x	1
Yam	Straw	0.5	[41]	x		0.5
Maize	Stalk	1.37	[31]	x		0.5
	Cob	0.25	[31]		x	1
	Husk	0.26	[31]		x	1
Oil Palm	EFB	0.23	[40]		x	1
	Shells	0.06	[40]		x	1
	Fibers	0.15	[40]		x	1
	Fronds	2.60	[42]	x		0.5
Sorghum	Stalk	4.75	[31]	x		0.5
	Husk	0.14	[31]		x	1
Soybean	Straw	2.5	[40]	x		0.5
	Pod	1.0	[40]		x	1
Bean	Straw	2.5	[43]	x		0.5
Rice	Straw	3.05	[31]	x		0.5
	Husk	0.23	[31]		x	1

. For transparency, all main figures and tables are based on the central 50% scenario, while national-level SRP values for the low (25%) and high (80%) cases are reported separately to illustrate the potential range and sensitivity.

The SRP values derived in this study served as the basis for estimating the biohydrogen potential.

2.4. Selection of the Biohydrogen Production Process

Biohydrogen can be produced by various biological, thermochemical, and electrochemical processes [44]. To identify the most suitable production process for the agricultural residues in Togo, a multi-criteria decision analysis (MCDA) approach was used in this study. The evaluation considered technical, economic, and environmental factors relevant to the local context.

2.4.1. Evaluation Criteria and Weighting

The selection process was structured around six key evaluation criteria, each of which was assigned a weight based on its relative importance to the specific context in Togo. Given the country’s status as a low-income economy with limited energy access and financial capacity, the MCDA prioritizes criteria that reflect feasibility, affordability and adaptability. The criteria were determined based on a literature review and their alignment with Togo’s national energy strategy, electrification goals, and economic constraints.

Table 3. Evaluation criteria for biohydrogen production processes and their assigned weights.

Criteria	Weight
Investment cost	0.2
Operational cost	0.2
Environmental impact	0.1
Technological Readiness Level	0.1
Energy requirement	0.2
Feedstock flexibility	0.2

shows the criteria and the respective weightings used in this study.

• Investment costs

High upfront costs present a significant barrier in Togo, where public and private investment capacity is limited (GDP per capita < USD 1000) [45]. As large-scale financing remains a challenge, capital-efficient technologies are favored. This criterion was given a higher weight to reflect the importance of financial feasibility in early-stage deployment.

2.

Operational costs

Operating costs were weighted equal to investment costs because they are equally decisive for long-term sustainability and affordability in Togo. The African Development Bank’s 2021–2026 Country Strategy Paper [20] identifies high electricity costs as a binding constraint, especially in rural areas, and stresses the need for financially sustainable infrastructure with robust operation-and-maintenance frameworks. Consequently, technologies with lower ongoing costs better support reliable service and affordability in resource-constrained settings.

3.

Environmental Impact

Togo’s national development strategy emphasizes sustainability, making the environmental impact of biohydrogen production a key consideration [20]. This criterion evaluated associated emissions and their potential environmental effects. The weighing of the environmental impact criterion was set lower (0.10) to reflect both the limited availability of harmonized lifecycle data for all three technologies and the variability in assessment methodologies found in the literature. Many LCA studies are based on differing assumptions (e.g., feedstock type, system boundaries, energy sources), making a direct and objective comparison across all processes difficult. To avoid introducing disproportionate influence from uncertain or non-comparable data, the environmental impact was included but with reduced weighting.

4.

Technological Readiness Level

The Technological Readiness Level (TRL) assesses process maturity on a scale from 1 (conceptual stage) to 9 (fully commercialized) [46]. Since all biohydrogen production processes considered in this study have TRLs below 9 [47], this factor directly impacts feasibility and implementation risk. In low-income settings, higher-TRL options are generally preferred to reduce execution risk. Because part of the maturity effect is already captured in investment and operating cost estimates, we assign TRL a reduced weight to limit double counting while still reflecting non-cost risks (e.g., bankability and reliability).

5.

Energy requirements

With only 25% of the rural population having access to electricity in 2022 [48] and frequent outages due to dependence on imports [49], energy-efficient processes are essential. A higher weight reflects the need to minimize external energy inputs to ensure reliability and independence from grid fluctuations.

6.

Feedstock flexibility

Agricultural residues are lignocellulosic materials with varying compositions [50], which can influence process efficiency and hydrogen output. The heterogeneous nature of these residues can pose challenges in achieving consistent process performance [51]. Scattered, small-scale farming further complicates feedstock logistics. A high weight was assigned to reward technologies that can process a wide range of residues with minimal pre-treatment, ensuring adaptability to local biomass conditions.

2.4.2. Candidate Biohydrogen Production Processes

Three biohydrogen production processes were considered in the evaluation, each representing a primary conversion pathway (

Table 4. Biohydrogen conversion pathways and selected processes.

Biological Pathway	Thermochemical Pathway	Biochemical Pathway
Dark Fermentation	Gasification	Microbial Electrolysis Cells

).

Each process in Table 4 was assessed using the MCDA framework. Due to the limited availability of harmonized, process-specific quantitative data for all criteria, especially under local conditions, a comparative ranking approach was adopted. Each process was assigned a score of 1 (lowest performance), 2 (moderate), or 3 (best performance) for each criterion based on its relative performance in the literature. For each criterion, scores were based on triangulated evidence from at least two independent sources; “best” indicates consistent dominance relative to the other options, “lowest” indicates consistent inferiority, and “medium” reflects mixed evidence or no clear dominance. This avoids spurious precision, maintains traceability, and is consistent with recognized MCDA practice for data-scarce settings. Robustness was assessed via equal weights. The weighted scores for each process were calculated by multiplying the scores by their corresponding criterion weights. The final score for each process was determined by summing the weighted scores across all criteria. The process with the highest total score was selected for calculating the theoretical biohydrogen production potential in Togo.

2.5. Biohydrogen Potential Estimation

The theoretical biohydrogen production potential from agricultural residues was estimated based on the selected production process and the biochemical composition of the available feedstock. The methodology involved the following steps:

• Determining the dry mass of the residues;
• Quantifying the cellulose and hemicellulose fractions;
• Calculating the theoretical biohydrogen yield using stoichiometric equations;
• Adjusting biohydrogen yields for uncertainties in metabolic pathways and thermodynamic limitations.

The theoretical biohydrogen production potential from agricultural residues was calculated for the dark fermentation (DF) process. The calculation method was developed based on extensive literature research, which provided key parameters and methods for estimating hydrogen yields.

2.5.1. Data Collection

The biohydrogen yield was estimated based on the cellulose and hemicellulose content of each residue type. Moisture content, cellulose, and hemicellulose fractions were obtained from peer-reviewed studies and specialized databases such as Phyllis [26].

Table 5. Overview of the moisture content and biochemical composition of the analyzed residue types.

Crop	Crop Residue	Moisture Content (wt. %)	Reference	Cellulose Dry (wt. %)	Hemicellulose Dry (wt. %)	Reference
Cassava	Stalk	20	[31]	32	20	[52]
	Peel	50	[41]	39	25	[53]
Yam	Straw	15	[31]	29	29	[54]
Maize	Stalk	15	[31]	36	31	[55]
	Cob	8	[31]	37	33	[55]
	Husk	11	[31]	46	36	[56]
Oil Palm	EFB	67	[57]	37	15	[58]
	Shells	12	[57]	42	12	[59]
	Fibers	37	[57]	34	26	[60]
	Fronds	71	[57]	40	30	[60]
Sorghum	Stalk	62	[31]	32	16	[61]
	Husk	15	[43]	32	15	[61]
Soybean	Straw	15	[40]	40	23	[62]
	Pod	15	[40]	52	19	[63]
Bean	Straw	10	[43]	40	19	[64]
Rice	Straw	16	[31]	35	24	[65]
	Husk	13	[31]	36	26	[66]

summarizes the biochemical composition values used in the calculation.

The shown values were obtained from studies using standardized methods for biochemical composition analysis [67].

2.5.2. Calculation of Dry-State Feedstock and Sugar Equivalents

To account for moisture content, the SRP was converted into a dry mass basis using Equation (3).

$${\mathrm{SRP}}_{\mathrm{dry},\mathrm{r}}={ \mathrm{SRP}}_{\mathrm{r}} \times \left(1-{\displaystyle \frac{{\mathrm{MC}}_{\mathrm{r}}}{100}}\right),$$

(3)

where SRP_dry,r is the SRP in a dry state for residue r (t/year) and MC_r is the moisture content of residue r (wt. %).

Cellulose and hemicellulose are the primary sugar polymers available for biohydrogen production. Their mass was determined to estimate the sugar content available for fermentation following Equations (4) and (5) [68].

$${\mathrm{m}}_{\mathrm{Cellulose},\mathrm{r} }={ \mathrm{SRP}}_{\mathrm{dry}, \mathrm{r}} \times \left({\displaystyle \frac{{\mathrm{w}}_{\mathrm{Cellulose},\mathrm{r}}}{100}}\right)$$

(4)

$${\mathrm{m}}_{\mathrm{Hemicellulose},\mathrm{r}}={ \mathrm{SRP}}_{\mathrm{dry},\mathrm{r}} \times \left({\displaystyle \frac{{\mathrm{w}}_{\mathrm{Hemicellulose},\mathrm{r}}}{100}}\right),$$

(5)

where m_Cellulose,r and m_{Hemicellulose,r} represent the mass of cellulose and hemicellulose in residue r (t/year) and w_Cellulose,r and w_{Hemicellulose,r} are the cellulose and hemicellulose content of residue r (wt. %).

It was assumed that cellulose hydrolyzes into glucose (hexose), while hemicellulose breaks down into xylose (pentose). Conversion factors were applied to account for the molar mass differences between the polymers and their monomers, adjusting for water loss during polymerization using Equations (6) and (7) [67].

$${\mathrm{m}}_{\mathrm{Hexose},\mathrm{r}}={ \mathrm{m}}_{\mathrm{Cellulose},\mathrm{r}} \times \left({\displaystyle \frac{{\mathrm{M}}_{\mathrm{Glucose}}}{{\mathrm{M}}_{\mathrm{Cellulose}}}}\right)$$

(6)

$${\mathrm{m}}_{\mathrm{Pentose},\mathrm{r}}={ \mathrm{m}}_{\mathrm{Hemicellulose},\mathrm{r}} \times \left({\displaystyle \frac{{\mathrm{M}}_{\mathrm{Xylose}}}{{\mathrm{M}}_{\mathrm{Hemicellulose}}}}\right)$$

(7)

where m_Hexose,r and m_Pentose,r are the theoretical mass equivalent of glucose and xylose of residue r, M_Glucose = 180.16 g/mol and M_Xylose = 150.13 g/mol are the molar masses of glucose and xylose, and M_Cellulose = 162.14 g/mol and M_{Hemicellulose} = 132.11 g/mol are the molar masses of cellulose and hemicellulose.

2.5.3. Theoretical Hydrogen Yield Estimation

Hydrogen yields were calculated using stoichiometric equations, assuming the primary products of DF are acetic acid and butyric acid [44,69,70]. The simultaneous formation of hydrogen in this process was quantified by following the stoichiometric equations, starting with hexose as the substrate.

The formation of acetic acid during DF is shown by Equation (8) [71].

$${{\mathrm{C}}_6\mathrm{H}}_{12}{\mathrm{O}}_6+2{\mathrm{H}}_2\mathrm{O} \to { 2\mathrm{CH}}_3\mathrm{COOH}+2{\mathrm{CO}}_2+4{\mathrm{H}}_2$$

(8)

Based on this reaction, the stoichiometric yield of H₂ is 4 moles for each mole of hexose.

$${{\mathrm{C}}_6\mathrm{H}}_{12}{\mathrm{O}}_6 \to { \mathrm{CH}}_3{\mathrm{CH}}_2{\mathrm{CH}}_2\mathrm{COOH}+2{\mathrm{CO}}_2+2{\mathrm{H}}_2$$

(9)

The formation of butyric acid during DF is illustrated by Equation (9) [71].

A molar ratio of 3:2 (butyric acid to acetic acid) was assumed, leading to an average hydrogen yield of 2.5 moles per mole of hexose. This ratio is supported in the literature [44,72]. The overall reaction is represented in Equation (10).

$${{4\mathrm{C}}_6\mathrm{H}}_{12}{\mathrm{O}}_6+2{\mathrm{H}}_2\mathrm{O} \to {{ 2\mathrm{CH}}_3\mathrm{COOH}+3\mathrm{CH}}_3{\mathrm{CH}}_2{\mathrm{CH}}_2\mathrm{COOH}+8{\mathrm{CO}}_2+10{\mathrm{H}}_2$$

(10)

Both xylose and glucose are metabolized via the Embden–Meyerhof–Parnas pathway during DF. Xylose requires additional steps compared to glucose, including conversion to xylulose before entering this pathway. Once xylulose is formed, subsequent metabolic steps align with glucose metabolism [70,73]. The stoichiometric yield of biohydrogen from pentose sugars (xylose) is similar to that of hexose, with adjustments for the sugar structure.

The overall reaction is shown in Equation (11).

$${{4\mathrm{C}}_5\mathrm{H}}_{10}{\mathrm{O}}_5+1.67{\mathrm{H}}_2\mathrm{O} \to {{ 1.67\mathrm{CH}}_3\mathrm{COOH}+2.5\mathrm{CH}}_3{\mathrm{CH}}_2{\mathrm{CH}}_2\mathrm{COOH}+{6.67\mathrm{CO}}_2+{8.33\mathrm{H}}_2$$

(11)

Assuming the same molar ratio of 3:2 for butyric acid to acetic acid, the theoretical yield from pentose metabolism is 2.08 moles of H₂ per mol of pentose.

Hydrogen yields from both hexose and pentose sugars were adjusted using a variability factor (VF) of 50% to account for uncertainties in metabolic pathways and thermodynamic limitations. This adjustment is recommended in the literature [74] and was integrated in Equations (12) and (13):

$${\mathrm{Y}}_{\mathrm{Hexose}} ={ \mathrm{Y}\text{’}}_{\mathrm{Hexose}} \times \mathrm{VF}$$

(12)

$${\mathrm{Y}}_{\mathrm{Pentose} }={\mathrm{Y}\text{’}}_{\mathrm{Pentose}} \times \mathrm{VF},$$

(13)

where Y_Hexose and Y_Pentose represent the adjusted biohydrogen yield for hexose and pentose, ${ \mathrm{Y}\text{’}}_{\mathrm{Hexose}}$ = 2.5 mol H₂/mol hexose and ${\mathrm{Y}\text{’}}_{\mathrm{Pentose}}$ = 2.08 mol H₂/mol pentose are the stoichiometric biohydrogen yields before adjustment, and VF = 0.5 accounts for the variability.

The total hydrogen yield per residue was determined by summing contributions from hexose and pentose sugars following Equation (14):

$${\mathrm{n}}_{{\mathrm{H}}_2,\mathrm{r}} =\left({\displaystyle \frac{{\mathrm{m}}_{\mathrm{Hexose},\mathrm{r}}}{{\mathrm{M}}_{\mathrm{Glucose}}}}\right)\times { \mathrm{Y}}_{\mathrm{Hexose}} +\left({\displaystyle \frac{{\mathrm{m}}_{\mathrm{Pentose},\mathrm{r}}}{{\mathrm{M}}_{\mathrm{Xylose}}}}\right)\times { \mathrm{Y}}_{\mathrm{Pentose}}$$

(14)

where ${\mathrm{n}}_{{\mathrm{H}}_2,\mathrm{r}}$ represents the total moles of H₂ produced from residue r.

3. Results

This section presents the key findings on the theoretical biohydrogen potential from agricultural residues in Togo. The results are organized into three main parts: the theoretical residue assessment, the selection of the biohydrogen production process, and the estimation of the biohydrogen potential.

3.1. Theoretical Residue Assessment

The study estimated the residue quantities from the eight most produced crops in Togo as of 2022. The production of crops is associated with residues. The various residues were determined for the eight selected crops and are listed in

Table 6. The major crops alongside their production quantities [28] and associated residue types.

No.	Crop	Crop Production (kt/Year)	Types of Residues
1	Cassava	1225	Stalk, Peel
2	Yam	985	Straw
3	Maize	957	Stalk, Cob, Husk
4	Oil Palm	588	EFB, Kernel Shells, Fibers, Fronds
5	Sorghum	281	Stalk, Husk
6	Soybean	236	Straw, Pods
7	Bean, dry	217	Straw
8	Rice	170	Straw, Husk

. The crops are sorted in descending order by the weight of their production quantities, and the respective residues are assigned to the respective crops.

A total of 17 residue types were identified for the eight crops analyzed. Table 6 shows that cassava (1225 kt/year) was the most produced crop in 2022, followed by yam (985 kt/year) and maize (957 kt/year). The residues identified vary by crop, with some producing multiple types of residues, such as oil palm, which produces empty fruit bunches (EFBs), kernel shells, fibers, and fronds. These residues serve as potential feedstocks for biohydrogen production.

3.1.1. Gross Residue Potential

The GRP provides an estimate of the total residues generated from crop production.

Table 7. Production quantities, Residue-to-Product Ratios, and the gross residue potential of the 17 analyzed crop residues.

Crop	Crop Production (kt/Year)	Crop Residue	RPR (g/g)	GRP (kt/Year)	GRP per Crop (kt/Year)
Cassava	1225	Stalk	0.13	159	576
		Peel	0.34	416
Yam	985	Straw	0.5	492	492
Maize	957	Stalk	1.37	1311	1799
		Cob	0.25	239
		Husk	0.26	249
Oil Palm	588	EFB	0.23	135	1787
		Shells	0.06	35
		Fibers	0.15	88
		Fronds	2.60	1529
Sorghum	281	Stalk	4.75	1334	1373
		Husk	0.14	39
Soybean	236	Straw	2.5	591	828
		Pod	1.0	236
Bean	217	Straw	2.5	541	541
Rice	170	Straw	3.05	517	556
		Husk	0.23	39
Total	4658			7953

provides a detailed overview of the data on production quantities, RPR values, and the GRP of the 17 analyzed crop residues in Togo. In addition, the GRP for all residues of a single crop was summarized to determine the overall contribution of the crop to the total availability of residues.

Table 7 shows that residue availability is significant but varies widely by crop and residue type. Maize, oil palm, and sorghum residues contribute the highest shares to the total residue potential. The total GRP for all 17 defined residues amounted to 7953 kt/year, exceeding the total crop production of 4658 kt/year.

3.1.2. Sustainable Residue Potential

The sustainable residue potential (SRP) represents the share of agricultural residues that can be considered available for biohydrogen production after accounting for ecological constraints, technical limitations and competing uses. In this study, we distinguish between the following:

• Field-based residues (e.g., stalks, straws, fronds), which are subject to ecological and technical limitations.
• Processing residues (e.g., husks, shells, fibers, pods, peels), which are assumed to be fully recoverable at the point of processing and are not constrained by field-level retention requirements.

To reflect the uncertainty in field residue availability due to varying soil conditions, climatic zones, and farming practices across Togo, we introduce three residue recoverability scenarios for field-based residues:

• Low (RRF = 25%);
• Reference (RRF = 50%);
• High (RRF = 80%).

Processing residues are held constant at 100% recoverability across all scenarios. The scenario approach enables us to assess the sensitivity of the national SRP to assumptions about ecological and practical constraints.

At the reference scenario (50% recoverability for field residues), the total SRP is estimated at 4715 kt/year, representing a 41% reduction from the theoretical gross residue potential (GRP) of 7953 kt/year.

Table 8. Overview of the sustainable residue potential of all residue types, classified by their crop under the 50% reference scenario.

Crop	Crop Residue	GRP (kt/Year)	RRF (g/g)	SRP (kt/Year)	SRP per Crop (kt/Year)
Cassava	Stalk	159	0.5	80	496
	Peel	416	1	416
Yam	Straw	492	0.5	246	246
Maize	Stalk	1311	0.5	656	1144
	Cob	239	1	239
	Husk	249	1	249
Oil Palm	EFB	135	1	135	1023
	Shells	35	1	35
	Fibers	88	1	88
	Fronds	1529	0.5	764
Sorghum	Stalk	1334	0.5	667	706
	Husk	39	1	39
Soybean	Straw	591	0.5	296	532
	Pod	236	1	236
Bean	Straw	541	0.5	271	271
Rice	Straw	517	0.5	259	298
	Husk	39	1	39
Total		7953		4715

presents the GRP, RRF, and SRP values for each residue type and crop under the reference scenario.

Table 8 reveals significant variability both between and within crops. The results indicate that maize, oil palm, and sorghum contribute the largest shares to the total SRP, with 1144 kt/year, 1022 kt/year, and 706 kt/year, respectively. In contrast, rice, beans, and yam exhibit the lowest SRP values, ranging from 246 to 298 kt/year.

To illustrate the influence of residue recoverability assumptions,

Table 9. National sustainable residue potential under different RRF scenarios.

Scenario	RRF	SRP (kt/Year)
Lower-bound scenario	25%	3097
Reference scenario	50%	4715
Upper-bound scenario	80%	6658

summarizes the national SRP under all three scenarios.

The SRP ranges from 3097 kt/year in the most conservative case to 6658 kt/year under the high-recoverability assumption. This corresponds to a relative deviation of −34.3% and +41.2%, respectively, compared to the reference scenario. The variation highlights the dominant role of field residue recoverability in determining the available feedstock for biohydrogen production. This sensitivity underscores the importance of developing region-specific residue management strategies and conducting field-level studies to obtain more accurate recoverability estimates in the context of Togo.

3.2. Selection of the Biohydrogen Production Process

The selection of the biohydrogen production process was performed using an MCDA framework, evaluating three possible production processes: dark fermentation, gasification, and microbial electrolysis cells (MECs).

Table 10. Criteria, weights, and results of the MCDA.

Criteria	Weight	DF	Gasification	MEC
Investment Costs	0.2	2	3	1
Operational Costs	0.2	2	3	1
Environmental Impact	0.1	3	1	2
TRL	0.1	2	3	1
Energy Requirement	0.2	3	1	2
Feedstock Flexibility	0.2	3	2	1
Total	1	2.5	2.2	1.3

lists the criteria, the assigned weights, and the results of the MCDA.

DF emerged as the most suitable biohydrogen production process for Togo, primarily owing to its low energy requirements [44], moderate costs [75], and adaptability to a variety of agricultural feedstocks [76]. With an overall score of 2.5, DF demonstrates robust performance across multiple criteria and is particularly well-suited for decentralized systems [51,75,77] in rural areas where electrification and energy infrastructure are limited [47,48]. Although gasification exhibits the highest hydrogen yield [78], cost efficiency [75], and technological maturity [47], reflected in its score of 2.2, it is generally considered more cost-effective for centralized, large-scale plants [51,75,77], a scenario that may not align with the fragmented agricultural sector in Togo. In contrast, MECs attained the lowest score of 1.3; they remain experimental and are currently unsuitable for processing agricultural residues due to high costs [51,79], limited feedstock compatibility [50], and technological immaturity [80].

In summary, although the MCDA shows that the DF process is the most suitable option for decentralized biohydrogen production under Togo’s economic and infrastructural conditions, significant data gaps in key criteria such as investment and operating costs, environmental impacts, and energy requirements affect the robustness of this conclusion.

To test the robustness of the MCDA results, a sensitivity analysis was conducted by assigning equal weights to all six criteria. This approach eliminates the influence of context-specific prioritization and allows for an unbiased comparison of the three technologies.

Under equal weighting, DF remained the highest-ranked option with a total score of 2.5, followed by gasification with 2.17, and MECs with 1.33. This outcome indicates that the selection of DF is not solely dependent on the original weighting assumptions and remains a robust choice even under alternative weighting scenarios.

3.3. Biohydrogen Potential Estimation

The theoretical biohydrogen potential for each agricultural residue was estimated based on the sustainable residue potential calculated under the reference scenario, which assumes a 50 percent recoverability rate for field-based residues. Biohydrogen yields were determined using stoichiometric equations that take into account both the moisture content and the biochemical composition of each residue, specifically the proportions of cellulose and hemicellulose. Figure 2 illustrates the theoretical biohydrogen potential (t/year) for each residue type.

Notably, maize stalks have the highest theoretical biohydrogen potential, approaching 6000 t/year. Ten residues, including cassava peel, yam straw, maize cob and husk, oil palm fronds, sorghum stalk, soybean straw and pod, bean straw, and rice husk, fall within the range of 1900 to 3000 t/year. In contrast, residues such as cassava stalk, oil palm EFB, shells, fibers, sorghum husks, and rice husk exhibit biohydrogen potentials below 1000 t/year.

Table 11. Overview of the biohydrogen potential of each residue and the total per crop.

Crop	Crop Residue	SRP (kt/Year)	Yield (mL H2/g Substrate)	BHP (t/Year)	Total BHP per Crop (t/Year)
Cassava	Stalk	80	80	519	2607
	Peel	416	62	2088
Yam	Straw	246	95	1906	1906
Maize	Stalk	656	110	5859	11,126
	Cob	239	124	2419
	Husk	249	141	2849
Oil Palm	EFB	135	33	363	3588
	Shells	35	91	262
	Fibers	88	73	523
	Fronds	764	39	2441
Sorghum	Stalk	667	35	1904	2150
	Husk	39	77	246
Soybean	Straw	296	103	2479	4709
	Pod	236	116	2230
Bean	Straw	271	102	2249	2249
Rice	Straw	259	96	2009	2339
	Husk	39	104	330
Total		4715		30,674

gives a detailed overview of the biohydrogen potential (BHP) for each residue, together with the SRP values and the specific hydrogen yields (mL H₂/g substrate).

The BHP of the individual residues ranges from 246 t/year for sorghum husks to 5859 t/year for maize stalks. Maize residues contribute the highest total BHP at 11,126 t/year, followed by soybean residues at 4709 t/year and oil palm residues at 3588 t/year. Together, residues from maize, soybean, and oil palm account for 63% (19,423 t/year) of the overall estimated BHP of 30,674 t/year.

The yield analysis reveals that maize husks achieve the highest yield (141 mL H₂/g), followed by maize cobs (124 mL H₂/g) and soybean pods (116 mL H₂/g). These high yields are attributed to their favorable biochemical composition of high cellulose and hemicellulose content coupled with low moisture levels. In contrast, residues with a high moisture content (e.g., oil palm EFB at 67% and fronds at 71%) show significantly lower yields (33 mL H₂/g and 39 mL H₂/g, respectively), underscoring the negative impact of moisture on fermentable substrate concentration. Crop residues from maize (11,126 t/year) and soybeans (4709 t/year) stood out for their high overall biohydrogen potential, attributed to a combination of low moisture content and high polysaccharide levels. In contrast, residues from crops like oil palm and sorghum showed lower yields, primarily due to either high moisture content or lower polysaccharide availability.

The total biohydrogen potential of all residues considered is 30,674 t/year. To contextualize this potential,

Table 12. Comparison of the biohydrogen energy equivalent with Togo’s total oil supply (2022) under the reference scenario.

Parameter	Value	Unit	Reference
Biohydrogen Potential	30,674	(t/year)
Energy Content Hydrogen (LHV)	120	(MJ/kg)	[82]
Biohydrogen Energy Equivalent	3681	(TJ/year)
Togo’s Total Oil Supply (2022)	21,295	(TJ/year)	[81]

presents the estimated total biohydrogen potential, the lower heating value (LHV) of hydrogen, and the resulting energy equivalent (biohydrogen energy equivalent). This energy equivalent was calculated by multiplying the total biohydrogen potential by the LHV of hydrogen and is compared with Togo’s total oil supply in 2022 as a reference. The total oil supply represents the energy derived from crude and refined oil products, including imports, but excluding stored or exported amounts [81].

The theoretical biohydrogen production potential using DF, based on 2022 production data, is equivalent to 3681 TJ/year, representing approximately 17% of Togo’s total oil supply in 2022.

Sensitivity of Biohydrogen Potential to Sustainable Residue Potential Scenarios

To evaluate how sensitive the national biohydrogen potential is to uncertainties in residue recoverability, the total hydrogen output was also estimated under two alternative sustainable residue potential scenarios, in addition to the reference scenario presented earlier. The results demonstrate substantial variation in potential production, depending on the assumed recoverability rates for field-based residues, and are shown in

Table 13. Sensitivity of national biohydrogen potential to assumptions about residue recoverability.

Scenario	RRF	Biohydrogen Potential (t/Year)	Energy Equivalent (TJ/Year)
Low recoverability	25%	20,991	2519
Reference case	50%	30,674	3681
High recoverability	80%	42,293	5075

.

Under the scenario assuming low recoverability, with a recovery rate of 25%, the national biohydrogen potential is reduced to 20,991 tons per year. This corresponds to an energy equivalent of 2519 TJ per year, representing a decrease of approximately 32% compared to the reference scenario that applies a 50% recoverability rate.

Conversely, the high-recoverability scenario, which assumes for an 80% recovery rate, results in a total biohydrogen potential of 42,293 tons per year. This is equivalent to 5075 TJ per year and reflects an increase of approximately 38% relative to the reference scenario.

This range illustrates the strong dependency of national hydrogen potential on ecological, agronomic, and management constraints. The absolute difference of over 2500 TJ per year between the lowest and highest estimates highlights the urgent need for more accurate, region-specific data on residue recoverability. If recovery conditions allow for maximum utilization, biohydrogen derived from agricultural residues could contribute up to 24% of Togo’s current oil-based energy supply, emphasizing its strategic potential as a renewable energy source.

4. Discussion

4.1. Theoretical Residue Assessment

The identification and quantification of agricultural residues relied on FAOSTAT production data and literature-derived theoretical factors. A total of 17 residue types were identified among the eight major crops, which include cassava, yam, maize, oil palm, sorghum, soybeans, beans, and rice. Although FAO data are widely recognized, their inherent variability due to the aggregation of multiple national and international data sources [25] introduces uncertainties in absolute production estimates. To increase accuracy, future research should prioritize collecting primary data directly from local sources such as farmers, agricultural institutions, and government agencies. This approach would capture specific local variations, reduce the uncertainties associated with secondary approximations, and potentially improve residue estimates. While FAO crop production data are the most comprehensive source available, they do not provide formal uncertainty margins. Consequently, this study could not quantify the statistical error of the residue estimates. Future research could apply sensitivity or scenario-based analysis using plausible variation ranges (e.g., ±10–20%) to assess the robustness of the results.

This study focuses on the eight most important crops by production quantity, representing the primary agricultural activities in Togo. The selection was based on the fact that these crops dominate the sector, with the next largest single crop only reaching a third of the production volume of rice, the eighth-ranked crop at 170 kt/year [28]. Expanding future analyses to include all agricultural crops in Togo would allow for a more comprehensive assessment of biohydrogen feedstock potential.

4.1.1. Gross Residue Potential

The GRP was calculated by applying RPRs to the production data, resulting in a total GRP of 7935 kt/year. Maize, oil palm, and sorghum contributed 1799 kt/year, 1787 kt/year, and 1373 kt/year, respectively, representing 63% of the overall potential. While RPRs offer a useful basis for estimating residue generation, their inherent variability, driven by differences in crop varieties, cultivation practices, and harvesting methods, introduces uncertainty into the calculations. For instance, while [31] found that field-based RPR values in Ghana are consistent with those reported in the literature, ref. [30] demonstrated that the choice of RPR model (linear, logarithmic, or exponential) can significantly affect residue yield estimates. The accuracy of the calculations can be optimized by developing Togo-specific, field-based RPR values and investigating different RPR models.

4.1.2. Sustainable Residue Potential

To account for ecological considerations such as soil health and ecosystem services, an SRP was estimated by applying a recovery factor of 50% to field-based residues [83,84,85]. The total SRP of 4715 kt/year underscores the substantial feedstock availability for biohydrogen production. The use of a 50% recovery factor aligns with mid-range recommendations from the literature, which vary significantly depending on the focus of the study and the agricultural context. Some studies recommend RRFs as low as 20% [31], especially in regions with fragile ecosystems where soil conservation is paramount. In contrast, other studies suggest values as high as 83% [43], particularly in regions prioritizing biomass use for bioenergy.

For example, Scarlat et al. [32] conducted a comprehensive assessment of crop residue availability in the European Union, recommending recovery rates between 50% and 60%, depending on residue type. These rates are deemed adequate to maintain soil organic matter while allowing for residue utilization for bioenergy. Such recommendations have been widely adopted in bioeconomy studies [86,87] in other countries. However, it is critical to emphasize that these general recommendations may not fully reflect the local conditions in Togo, where factors such as soil fertility, erosion risk, and climatic variability differ considerably across agroecological zones. Therefore, the direct application of externally derived RRFs may introduce uncertainty in national estimates. To further refine the SRP values, future research should focus on developing Togo-specific recovery factors through targeted field studies. Such studies could explore regionally relevant variables, including soil type, climatic conditions, topography, and agricultural practices, to ensure sustainable residue utilization. This would provide a clearer understanding of how much residue can be sustainably removed without jeopardizing soil health or long-term agricultural productivity.

Moreover, future assessments of the SRP should not only consider ecological constraints but also account for the competing uses of crop residues. Agricultural residues are used for purposes such as animal bedding, feed, and fuel [88]. By considering these competing demands, future studies could provide a more accurate estimate of the actual feedstock available for biohydrogen production, thereby offering a firmer foundation for scaling biohydrogen projects in Togo. Ignoring these alternative uses can lead to an overestimation of the available residue pool for energy applications. By integrating both ecological limitations and socioeconomic demands, future research can produce more robust and operational estimates of the residue potential available for biohydrogen production.

4.2. Selection of a Biohydrogen Production Process

The aim of this section was to identify the most suitable biohydrogen production process for Togo, considering the country’s infrastructural and economic conditions. An MCDA was used to systematically evaluate different technologies based on their performance relative to Togo’s context [89]. Although MCDA provides a structured comparison, its outcomes are sensitive to the selection and weighting of evaluation criteria and the quality of the underlying data. In our analysis, we treated operational costs and energy requirements as separate criteria—even though they are closely intertwined—to emphasize the critical role of energy demands in Togo’s context of limited electrification. Future studies should include sensitivity analyses and integrated techno-economic assessments to further explore these interdependencies.

Three biohydrogen production processes were evaluated: gasification, DF, and MECs. These were assessed using six criteria. DF emerged as the most suitable technology for Togo due to its low energy requirements, lowest greenhouse gas emissions [75], and high adaptability to diverse agricultural feedstocks [76]. These characteristics make DF especially appropriate for decentralized systems in rural areas with limited electrification. Although gasification demonstrates the highest hydrogen yield [78], cost efficiency [75], and technological maturity [47], its benefits are largely linked to large-scale, centralized operations [51,75,77], a scenario less applicable to Togo’s fragmented agricultural landscape. MECs received the lowest score because of their high costs [75], limited feedstock compatibility [50], and technological immaturity [80], consistent with recent studies [89,90].

Despite the clear ranking from the MCDA, several criteria presented particular challenges that warrant further discussion, in particular the investment and operational costs, environmental impact, and energy requirement.

4.2.1. Investment and Operational Costs

Assessing investment and operational costs was challenging due to the lack of a unified dataset comparing the three technologies under consistent conditions. None of the studies analyzed evaluated costs using the same feedstock, plant size, and time frame, forcing reliance on a mix of qualitative insights and limited quantitative data. Gasification was ranked as the most cost-effective option based on its technological maturity and industrial adoption, though its cost assumptions derive from large-scale, ideal conditions. DF was moderately ranked, reflecting its suitability for small-scale systems, while MECs scored lowest due to their early development stage and reliance on expensive components, such as electrodes and membranes. These heterogeneous data sources do not fully account for region-specific factors like imported technology costs, local labor rates, and maintenance requirements in Togo. Future research should conduct standardized cost assessments using consistent parameters, collect region-specific capital and operational cost data (including transportation, labor, and maintenance), and develop scenario-based models to evaluate how variations in plant size, feedstock composition, and economic conditions affect costs.

4.2.2. Environmental Impact

Lifecycle assessments (LCAs) served as a theoretical reference for evaluating environmental impact. However, existing comparative studies [91,92] rely on secondary data and inconsistent assumptions [50,93], which limited our analysis to greenhouse gas emissions—the most consistently reported environmental impact category. Future studies should perform LCAs tailored to Togo’s context, using consistent feedstocks and methodologies to enable a more comprehensive environmental assessment.

4.2.3. Energy Requirement

Energy demand is a decisive factor in the evaluation of biohydrogen production technologies. However, in the absence of quantitative data on the processes, this study utilized only qualitative findings. While these findings provided general trends, they did not capture the variability due to feedstock type, scale of operation or regional factors, so the energy requirement ranking is only preliminary. Future analyses should include pilot studies or case-specific modelling to quantitatively assess energy consumption and better evaluate the techno-economic feasibility of the technologies.

4.2.4. Implications and Future Directions

In summary, DF technology appears to be the most promising technology for decentralized biohydrogen production in Togo, while gasification can be considered for centralized applications where the infrastructural conditions are in place. To refine these conclusions, future research should include standardized cost estimates, region-specific lifecycle assessments and quantitative energy consumption studies. These efforts will help provide a more solid foundation for the deployment of biohydrogen production technologies in Togo. Although this study focuses on Togo, the methodological framework developed can be applied to other countries with similar agricultural structures and decentralized farming systems. The modular approach—starting from the estimation of gross and sustainable residue potentials, followed by technology selection using MCDA, and culminating in a stoichiometric calculation of the biohydrogen yield—offers a replicable template. By adapting input data such as RPRs, moisture content, biochemical composition, and recovery factors to local conditions, the workflow can be transferred to other contexts. In particular, the MCDA structure, with clearly defined weighting and scoring criteria, allows for easy modification based on national development priorities and data availability. Moreover, if geo-referenced crop production data are available, this framework can be expanded through GIS to identify spatial hotspots of biohydrogen potential and support future siting analyses.

The weighting used in the MCDA was based on qualitative reasoning informed by the relevance of each criterion to the Togolese context. However, we recognize that this approach remains subjective. To improve the robustness and representativeness of future evaluations, stakeholder consultations, expert elicitation methods, or structured statistical approaches such as Analytic Hierarchy Process could be used to derive more systematic weighting schemes.

4.3. Biohydrogen Potential Estimation

The theoretical biohydrogen potential for the selected DF process was estimated using the SRP as the feedstock basis. Biohydrogen yields were calculated through stoichiometric equations that incorporated the moisture content and biochemical composition of individual residues, providing estimates of hydrogen production (t/year) tailored to each residue’s characteristics.

Under the reference scenario (50% recoverability), the total national BHP for all residues considered in this study was estimated at 30,674 t/year. Among the residues, maize residues provided the highest BHP at 11,126 t/year, followed by soybean residues at 4709 t/year and oil palm residues at 3588 t/year. These high contributions highlight the importance of strategically locating biohydrogen facilities close to the cultivation areas of these crops. Considering the low financial value and high transportation costs of agricultural residues, proximity to feedstock sources can significantly reduce transport expenses and improve logistical feasibility, as noted by Ukoba et al. [94]. The corresponding energy equivalent of 3681 TJ/year represents approximately 17% of Togo’s total oil supply in 2022. Although modest relative to the national energy supply (155,083 TJ/year), this potential still offers a great opportunity to replace a portion of oil-derived energy and could reduce reliance on imports, improve energy security, and lower the environmental footprint of Togo’s energy sector. In a regional context, comparable studies point to substantial residue-based energy potential in Africa, for example, around 149 PJ per year from crop residues in Uganda (21% of final energy consumption) [16,17] and a national analysis for Ghana indicating that residue-derived bioethanol could supply a large share of electricity demand [15].

To reflect uncertainty in feedstock availability, the national biohydrogen potential was also calculated under a 25% and a 80% sustainable residue potential scenario. The ranges reported in Table 9 show that the result is highly sensitive to the assumed recoverability of field residues. In mass terms, the national output spans from 20,991 to 42,293 tons per year. In energy terms, the difference between the low and high scenarios is approximately 2.6 PJ per year, which corresponds to a shift from about 12% to 24% of Togo’s current oil-based energy supply. This underscores the importance of accurate, site-specific field measurements to narrow the uncertainty. Accurate estimation required converting the SRP to a dry basis to align with biochemical composition data [67]. In the absence of specific moisture data for Togo, literature values were used, introducing additional uncertainty, as moisture content can vary significantly across species and conditions and harvest handling [31]. Future studies should prioritize field measurements to refine these estimates.

Fermentable sugars were estimated by converting cellulose and hemicellulose into hexose and pentose equivalents, respectively. The conversion of cellulose is straightforward, as it consists exclusively of glucose, a hexose sugar. For hemicellulose, it was assumed that xylose, a pentose sugar, serves as the primary monomer, even though hemicellulose contains a variety of sugars, including some hexoses. While this assumption slightly oversimplifies the composition of hemicellulose, studies like [50,67] indicate that pentose sugars are predominant, making this approach reasonable in the absence of more detailed data.

Hydrogen yields were calculated using stoichiometric equations, assuming that the primary products of dark fermentation (DF) are acetic acid and butyric acid [44,69,70]. During the degradation of complex substrates such as agricultural residues, diverse microbial communities generate a range of intermediate compounds and by-products, including volatile fatty acids and alcohols [95]. Detailed descriptions of these complex metabolic pathways can be found in the literature [96]. The actual hydrogen yield in DF varies due to the multiplicity of reaction pathways, and the presented stoichiometric equations represent only a subset of the processes involved. Yields are further influenced by factors such as microbial composition and environmental conditions [97].

The presence of specific microorganisms can significantly alter hydrogen production. For instance, Clostridium barkeri metabolizes glucose into ethanol and lactic acid, resulting in no hydrogen production, while species such as Clostridium articum can even consume hydrogen during their metabolic processes [98].

In DF, the acetate pathway yields 4 mol H₂ per mol of hexose, while the butyrate pathway yields 2 mol/mol. Based on a baseline molar ratio of 3:2 butyric to acetic acid, the resulting stoichiometric hydrogen yield is 2.50 mol/mol. For comparison, an equal split (1:1) results in 2.67 mol/mol, and a more acetate-rich scenario (3:2 acetate/butyrate) yields 2.86 mol/mol. The theoretical bounds range from 2.00 mol/mol (100% butyrate) to 4.00 mol/mol (100% acetate).

To reflect pathway variability, microbial diversity, hydrolysis limitations, and environmental factors (e.g., pH, inhibitors), an aggregate variability factor (VF) of 0.5 was applied, consistent with values reported in the literature [97]. This adjustment halves all theoretical yields, with the baseline dropping to 1.25 mol/mol. As a robustness check, assuming a plausible VF range of 0.3–0.7 results in hydrogen yields between 0.75 and 1.75 mol/mol for the same product ratio. Additional pathways (e.g., lactate, ethanol, propionate) would further reduce actual yields. Therefore, the acetate–butyrate pathway serves as a transparent baseline, and the sensitivity analysis illustrates how shifts in acid ratios and efficiency affect the expected yield range.

Comparison of Theoretical and Experimental Hydrogen Yields

Hydrogen yields for each residue were estimated based on their biochemical composition (see Table 5) and expressed in mL H₂/g of substrate to enable direct comparison and practical application. This standardization facilitates correlating residue availability with theoretical biohydrogen yields and supports informed decision-making for biohydrogen projects. Additionally, yields were converted to mL H₂/g total volatile solids (TVS), representing the biomass fraction remaining after moisture and ash are removed, to align with common reporting formats in biohydrogen studies.

Table 14. Estimated biohydrogen yields of maize stalk.

Crop	Crop Residue	Yield (mL H2/g TVS *)	Yield (mL H2/g Substrate)
Maize	Stalk	133	110

* Total volatile solids (excluding moisture content and ash content).

shows the estimated hydrogen yield for maize stalk in both units.

Table 15. Hydrogen yields from experimental studies on maize stalk along operational parameters.

Substrate Type	Microbial Inoculum Source	Strain	Temp. (°C)	H2-Yield	Unit	Reference
Maize stalk	Cow manure	Clostridium sartagoforme	35	87	mL H2/g substrate	[99]
	Cattle manure	Clostridium butyricum	36	93	mL H2/g substrate	[100]
	Cow manure	Clostridium sp.	50	150	mL H2/g TVS	[101]

presents the results of experimental studies on maize stalk, along with operational parameters to allow a comparison:

Comparing our estimated yields with experimental studies reveals that the yield of 87 mL H₂/g substrate reported by [99] and 93 mL H₂/g substrate reported by [100] are approximately 21% and 15% lower, respectively, than our estimated substrate yield. Similarly, the yield reported by [101] at 150 mL H₂/g TVS is about 13% higher than our estimated yield of 133 mL H₂/g TVS. These deviations fall within a reasonable range, suggesting that our theoretical estimates are consistent with experimental findings and reflect achievable yields under typical conditions.

To further refine these estimates and improve yield prediction reliability, future experimental studies should generate empirical data on actual yields and conversion times. Such data would offer valuable insights for techno-economic assessments and support the optimization of biohydrogen production processes. Additionally, evaluating the variability in yields across different agricultural residues would provide a more nuanced understanding of the DF process and its feasibility for each residue type.

5. Conclusions

The aim of this study was to estimate the theoretical biohydrogen potential from agricultural residues in Togo. To fill significant data gaps regarding residue availability and biohydrogen yield, an exploratory mixed-methods approach combining quantitative estimates with qualitative analyses was used. By systematically assessing the availability of agricultural residues, selecting a method to produce biohydrogen, and quantifying the theoretical potential, the study provides insights and serves as a foundation for promoting sustainable energy initiatives in Togo.

• Objective 1: Estimate the quantities and availability of agricultural residues in Togo.

The study identified 17 relevant agricultural residues from eight major crops in Togo, with a theoretical residue potential of 7953 kt/year. Considering aspects such as soil health and technical limitations with a 50% recoverability factor, a sustainable residue potential of 4715 kt/year was calculated. Maize, oil palm, and sorghum accounted for the largest shares, with 1144 kt/year, 1023 kt/year, and 706 kt/year, respectively. However, sensitivity analysis revealed that the SRP could vary significantly, ranging from 3096 kt/year (25%) to 6658 kt/year (80%), depending on the ecological assumptions made.

These findings confirm the considerable resource base for biohydrogen production, but also highlight the importance of site-specific assessments. Future studies should consider additional crops, incorporate competing uses (e.g., for animal feed, bedding, and fuel), and generate localized data to better quantify available feedstock.

• Objective 2: Select a suitable process for biohydrogen production.

The second objective was to identify a suitable process for biohydrogen production from agricultural residues. Using a multi-criteria decision analysis based on economic, environmental, and technical factors, dark fermentation emerged as the most suitable option for decentralized biohydrogen production in Togo. Its moderate investment and operating costs, low energy requirements, and flexibility in processing various agricultural residues make the process particularly compatible with Togo’s rural agricultural landscape and limited energy infrastructure. While gasification offers higher hydrogen yields and greater technological readiness, its high electrical energy requirements limit its feasibility in rural areas. MECs show potential for high hydrogen yields and reduced post-treatment needs, but their early development stage and high costs currently hinder practical applications for agricultural residues. DF is recommended as the most viable process for decentralized biohydrogen production in Togo, with gasification as an alternative for larger-scale applications where sufficient energy infrastructure exists. To validate these findings and gain detailed insights, techno-economic studies and lifecycle assessments tailored to Togo’s context are needed. Such research would provide critical information on the environmental impacts and economic feasibility of biohydrogen production, enabling more sustainable and informed decision-making.

• Objective 3: Quantify the theoretical biohydrogen potential.

Using the sustainable residue potential as the feedstock basis and accounting for residue-specific biochemical composition and moisture content, the theoretical biohydrogen potential was estimated. Under the reference scenario with 50% recoverability for field-based residues, the national potential is 30,674 t/year, equivalent to 3.68 PJ per year and about 17% of Togo’s total oil supply in 2022. Scenario analysis shows strong sensitivity to recoverability assumptions: under a conservative 25% scenario, the potential falls to 20,991 t/year, or 2.52 petajoules per year, which corresponds to roughly 12% of oil supply; under an optimistic 80% scenario, it increases to 42,293 t/year, or 5.08 PJ per year, representing about 24%. Residues with low moisture and high holocellulose content, including maize stalks and cobs and soybean pods, deliver the highest hydrogen yields. Locating production facilities close to these high-yielding residue sources would reduce transport costs and strengthen the feasibility of decentralized systems. To improve the robustness of these estimates, future work should include experimental validation to quantify yield variability, conversion times, and residue-specific performance under local conditions. The assessment of theoretical biohydrogen potential from agricultural residues in Togo underscores a significant opportunity to transform agricultural waste into a valuable energy carrier. By utilizing abundant residues from major crops through dark fermentation, Togo can take meaningful steps towards improving energy self-sufficiency, environmental sustainability, and socio-economic development. Realizing this potential will require coordinated efforts in research, technological adaptation, infrastructure development, and policy support. Despite these challenges, biohydrogen production offers a promising renewable energy solution that aligns with global sustainability goals and Togo’s national development priorities.