Valorization of Textile Cotton Waste and Textile Sludge into High-Quality Torrefied Biofuel Pellets: Fuel Characteristics and Optimization

Abstract

This study investigates the conversion of textile wastewater sludge (TWS) and textile cotton waste (TCW) into solid biofuels through pelletization and torrefaction, addressing the growing need for sustainable waste management and alternative fuels in the textile sector. Blended feedstocks were conditioned to ~10% moisture, pelletized into 8 mm cylinders, and thermally upgraded at 200–240 °C for 30–90 min. Proximate and ultimate analyses, calorific value measurements, compressive strength testing, bulk and true density assessment, and TGA–DTG were used to evaluate fuel properties, while response surface methodology (RSM) optimized torrefaction parameters. The TCW-rich 20:80 (TWS:TCW) blend with 5% starch exhibited the most favorable overall performance, achieving a calorific value of 3377 kcal kg⁻¹, ash of 10.3%, bulk density of 554 kg m⁻³, and maximum compressive strength of 14.9 N mm⁻². Torrefaction at 200 °C for 60 min increased the GCV to 4083 kcal kg⁻¹ with a high mass yield of 92%, indicating mild thermal decomposition and good energy retention. Further Torrefaction at 220–240 °C increased GCV to 4362–4565 kcal kg⁻¹, accompanied by expected mass-yield reductions due to increased devolatilization. TGA–DTG confirmed faster and cleaner decomposition for TCW-rich pellets and higher residues for sludge-rich blends. RSM indicated temperature as the dominant factor governing mass and energy yields. These findings demonstrate that optimized co-pelletization and mild-to-moderate torrefaction can effectively transform textile residues into energy-dense, mechanically stable biofuels suitable for industrial heat applications.

1. Introduction

The textile industry is a cornerstone of the global economy, but it generates enormous quantities of waste that impose serious environmental burdens. Global textile production has risen sharply in recent decades, growing from around 58 million tonnes in 2000 to over 100 million tonnes by 2020 (projected to reach ~145 million tonnes by 2030) [1]. At the same time, textile manufacturers are under intense pressure to phase out coal-fired energy sources while managing their own waste streams more sustainably [2]. Major apparel brands and regulators are increasingly pushing for coal-free supply chains, making the transition to alternative fuels in dyeing, finishing, and boiler operations an immediate compliance concern rather than a distant goal [3,4]. In parallel, textile production yields large volumes of solid residues, notably textile wastewater sludge (TWS) from effluent treatment and textile cotton waste (TCW) from yarn and fabric processing, both of which can strain landfill capacity and pose air–water pollution risks if mismanaged. For example, in the EU about 1.38 million tonnes of textile waste were treated in 2022, yet roughly 26% of collected textiles still ended up incinerated or landfilled despite recycling efforts [5]. Converting these waste streams into a densified, energy-rich solid biofuel via pelletization and torrefaction offers a promising circular-economy solution that simultaneously diverts waste from landfills and replaces a portion of coal in industrial energy supply [6,7].

Cotton-rich textile waste (TCW) and textile sludge (TWS) have complementary fuel properties that can be leveraged in such a waste-to-energy approach. Cotton waste is typically low in ash and high in volatile matter, with a fibrous cellulose structure that aids in forming strong pellets with minimal binder addition [8]. By contrast, sludge from textile effluents contains substantial fixed carbon and inorganic matter (e.g., Ca, K, Na, Fe), which can help catalyze char combustion but also introduce fouling and acid-gas emission challenges if burned untreated. Combining these materials and pre-treating them through co-pelletization and co-torrefaction can harness their strengths while mitigating liabilities. In a blended pellet, sludge-derived alkali minerals catalyze biomass devolatilization and char burnout, cotton’s volatiles improve ignition, and the combined matrix can immobilize a fraction of heavy metals in the ash [8,9]. When the blend ratio and torrefaction conditions are optimally chosen, the result is more complete fuel burnout at lower combustion temperatures and lower NOx and SO₂ emissions compared to firing either waste stream alone [10]. This synergy not only improves energy recovery but also helps address pollutant emissions in a co-firing scenario.

Furthermore, many textile waste fractions (especially synthetic fibers and chemically treated fabrics) are extremely slow to decompose, persisting for decades or even centuries in landfills [11,12]. In regions with large textile industries, these wastes are accumulating rapidly; for instance, Taiwan generates over 80,000 metric tons of textile waste annually, only a small fraction of which is recycled [13]. Such trends underscore an urgent need for sustainable waste management and resource recovery solutions in the textile sector. Utilizing TWS and TCW as feedstocks for biofuel production directly addresses this need by turning difficult waste into valuable energy while minimizing long-term environmental liabilities.

Pelletization and torrefaction are key technologies in this waste-to-fuel pathway, as they produce a solid biofuel with improved handling and fuel characteristics. Pelletization compacts loose biomass into uniform, high-density cylinders, greatly enhancing fuel handling, storage, and combustion efficiency [14,15]. Torrefaction—essentially a mild pyrolysis at roughly 200–300 °C in a low-oxygen environment—then further dries and partially carbonizes the material, yielding a coal-like solid with higher energy density, reduced moisture, and greater resistance to biological decay and reabsorption of water [15,16]. In practice, properly torrefied TWS/TCW composite pellets can achieve moisture contents below about 10 wt% and higher heating values in the ~18–23 MJ kg⁻¹ range, comparable to low-grade coal [17,18]. These pellets also exhibit high mass density (typically ≥600–650 kg m⁻³) and strong mechanical durability (often ≥95% durability index), reflecting their robust structure. Cotton’s natural lignocellulosic fibers provide inherent binding during pellet pressing, while a small fraction of sludge may contribute inorganic minerals that potentially enhance inter-particle cohesion without increasing ash content beyond acceptable limits [19]. As a result, the torrefied TCW/TWS pellets can meet international solid-fuel standards (e.g., EN ISO 17225-2 A1/A2 grade specifications for wood pellets) in terms of ash and moisture content [20]. They also generate fewer fines and have better handling characteristics than non-pelleted or raw biomass fuels, improving their suitability for automated feeding systems [20].

Beyond fuel quality, co-firing these optimized textile-derived pellets has shown environmental advantages in combustion. The inorganic constituents of TWS can capture acidic flue-gas species, leading to significantly lower sulfur emissions (SO₂) when co-combusted with high-carbon biomass, compared to sludge-alone baselines [21,22]. Likewise, the presence of cotton-derived volatiles and char promotes more complete conversion of fuel nitrogen to N₂ rather than NO_x, helping to curtail nitrogen oxide emissions [21,22]. Torrefaction and blending also influence trace metal behavior: in optimized sludge–biomass pellets, heavier metals such as Cu, Cr, and Ni tend to remain sequestered in the bottom ash (stabilized by carbon–mineral matrices) instead of escaping with fly ash, though careful control of torrefaction conditions is required to minimize release of volatile metals like Hg and Cd [23]. These emissions and handling benefit further favor the use of torrefied TWS/TCW pellets as a cleaner-burning coal substitute.

For textile manufacturers, using torrefied pellets produced from their own waste offers significant material handling advantages: densification of TWS/TCW into pellets can reduce waste volume by roughly 3–6×, easing on-site storage and transportation, and the pellets can be fed into existing coal-fired boiler systems with minimal modifications or retrofit costs [24]. This combination of meeting fuel quality standards and improving handling efficiency underpins the viability of TCW/TWS torrefied pellets as a drop-in replacement for a portion of coal in textile industry boilers. It provides a practical pathway for mills to fulfill coal phase-out commitments while still delivering stable process heat [25,26]. Recent advances in biomass-to-biofuel research have also emphasized catalytic upgrading routes to improve fuel quality and carbon efficiency. Huang et al. reported room-temperature, carbon-negative biodiesel production enabled by photothermal/catalytic synergy, highlighting progress toward milder and more efficient biofuel synthesis [27]. In parallel, Li et al. reviewed carbon-increasing catalytic strategies (e.g., C–C coupling and chain-growth approaches) to upgrade biomass-derived intermediates into more energy-intensive fuels and chemicals, addressing the intrinsic oxygen-rich and short-carbon-chain limitations of many biomass platforms [28]. In contrast to these predominantly liquid-fuel catalytic pathways, the present study targets textile-industry solid residues (textile cotton waste and textile wastewater sludge) and develops a practical densification–torrefaction route to produce mechanically stable, energy-dense solid biofuel pellets. However, while previous studies have explored the pelletization or torrefaction of sewage sludge or textile waste individually, there is limited research on the combined use of textile wastewater sludge (TWS) and textile cotton waste (TCW) for co-pelletization and torrefaction, particularly with systematic optimization using Response Surface Methodology (RSM). Furthermore, few studies have comprehensively evaluated the mechanical, thermal, and fuel-quality properties of the resulting pellets. This study explores a new approach by using textile wastewater sludge (TWS) from an industrial plant in combination with textile cotton waste (TCW) to produce high-quality biofuel pellets, a combination that has not been systematically examined in such comprehensive detail before. The work identifies an optimal mixing ratio for improved fuel performance and incorporates co-torrefaction to increase energy density and fuel quality, evaluates pellet production rate and practical limitations of each mixture, and performs detailed mechanical, thermal, and energy characterization including TGA and DTG analyses. Additionally, RSM-based optimization of blending ratios and torrefaction parameters provides predictive models for key fuel properties, offering a sustainable approach to textile waste management, sludge valorization, and cleaner energy generation.

2. Materials and Methods

2.1. Sample Collection and Pre-Processing

Textile cotton waste (TCW) was collected from textile cutting and weaving operations, and textile wastewater sludge (TWS) was obtained from the centrifuge output of an effluent treatment plant (Al-Rahim Textile Industries, Nooriabad, Pakistan). The as-received TCW had an initial moisture content of about 10% by weight, whereas the TWS contained approximately 85% moisture. To reduce the sludge’s moisture to a comparable level, TWS was sun-dried in thin layers for ~6 days, reaching a moisture content of around 10% (similar to the TCW). The dried TWS was then gently granulated, and the TCW was pre-shredded; both materials were passed through 5 mm and subsequently 3 mm mesh screens to obtain a uniform particle size (≤3 mm) suitable for pelletization. The prepared TCW and TWS feedstocks were thoroughly homogenized (separately) and stored in airtight containers under low-humidity conditions to prevent reabsorption of moisture prior to blending and compaction. A schematic diagram of the experimental setup for pelletization and torrefaction of textile waste is shown in Figure 1.

2.2. Blending and Pelletization

Four blend formulations of the dried TCW and TWS were prepared with TWS:TCW mass ratios of 20:80, 40:60, 60:40, and 80:20. Each 30 kg batch of a given blend was mixed in a rotary blender for ~10 min to achieve a homogeneous mixture and uniform moisture content. The overall moisture of each blend was adjusted to 10–13% (wet basis) for optimal pelletizing: if the initial moisture was below this range, a fine mist of deionized water was added during mixing, followed by a sealed equilibration period to distribute the moisture evenly.

The conditioned blends were then compacted using a flat-die pellet mill (GEMCO ZLSP200B, ~200–300 kg h⁻¹ throughput). The pellet mill operates through mechanical compression without external heating; however, frictional forces during operation generated die temperatures in the range of approximately 65–85 °C. This mill produces cylindrical pellets of 8 mm diameter and approximately 30 mm length, dimensions that are compatible with standard boiler feed systems. During preliminary trials, the 20:80 TWS:TCW blend showed poor pellet formation due to insufficient inherent binders; accordingly, 5 wt.% of corn starch (dry basis) was added to that blend, which greatly improved its pellet throughput and “green” strength (i.e., fresh pellet durability). The starch used was commercial food-grade native (unmodified) corn starch, added in dry powder form during blending prior to moisture adjustment to ensure uniform distribution. Partial gelatinization of starch occurred during pelletization due to frictional heat and compressive forces, contributing to improved interparticle bonding and pellet densification. The other three formulations (40:60, 60:40, and 80:20) were pelletized without any binder, relying on the natural fiber and mineral-binding characteristics of the feedstock. After extrusion, the hot pellets were cooled in ambient air on perforated trays and then sieved to remove any loose fines. The clean, uniformly sized pellets were sealed in airtight containers and set aside for torrefaction and further testing. The mechanical compression during pelletization also induced structural modification of the blends through fiber entanglement and mineral–organic bonding interactions, thereby increasing bulk density and mechanical strength. The pelletization efficiency was tracked by measuring the pellet conversion rate, defined as the fraction of input dry mass that was successfully converted into durable pellets (this metric is quantified in Section 2.6).

2.3. Torrefaction

The pelletized samples were thermochemically treated in a laboratory oven under low-oxygen conditions (to prevent combustion) to produce torrefied biomass. No external inert gas (e.g., N₂ or Ar) was supplied during torrefaction. Instead, the process was carried out in an oxygen-limited configuration in which oxygen access was passively restricted by minimizing air exchange within the furnace chamber during heating. For each run, pellets were arranged in a single layer on aluminum foil-lined trays, and the oven was preheated to the target temperature before inserting the samples. Three torrefaction temperatures (200 °C, 220 °C, and 240 °C) were investigated, each combined with three residence times (30, 60, and 90 min). During heating, the oven atmosphere was oxygen-limited (essentially inert) to ensure thermal decomposition (mild pyrolysis) rather than burning. After the set time elapsed, the hot trays of pellets were immediately transferred to a desiccator to cool under anoxic conditions. Once cooled to near room temperature, the torrefied pellets were sealed in airtight containers to avoid exposure to air and moisture prior to analysis. Because the system was oxygen-limited rather than fully inert (continuously purged), minor oxidative reactions during torrefaction cannot be completely ruled out, particularly at higher temperatures.

Torrefaction severity was quantified by two key metrics: the mass yield and the energy yield of the pellets after treatment. The mass yield is the percentage of the initial pellet mass retained after torrefaction, while the energy yield is the percentage of the initial energy content retained in the product. In essence, the energy yield accounts for both the mass retained and the change in heating value due to torrefaction. These performance metrics were calculated for each torrefaction condition (see Section 2.6) to compare the extent of solid fuel upgrading across different temperatures and durations.

Although the oxygen level inside the furnace was not continuously monitored or controlled, all torrefaction experiments were carried out under the same operating conditions, including furnace setup, sample placement, and heating procedure, to ensure consistency and reproducibility within this study. Therefore, comparisons among the different torrefaction conditions investigated here remain valid. However, in contrast to systems operated under a continuous inert gas flow (e.g., N₂ or Ar), the presence of limited oxygen may have allowed slight oxidative effects, especially at higher temperatures. This could potentially influence parameters such as mass yield or devolatilization behavior to a small extent. Accordingly, direct quantitative comparisons with studies conducted under strictly inert atmospheres should be made with caution. At the same time, the oxygen-limited configuration used in this work reflects a more practical and economically realistic approach for potential large-scale or industrial applications.

2.4. Fuel Characterization and Testing

All analytical measurements were performed for each type of sample—including the raw materials (TCW and TWS), the untorrefied (raw) pellets, and the torrefied pellets—following consistent procedures to enable direct comparison. Each reported value represents the mean of three replicate tests (n = 3) to ensure statistical reliability.

Proximate Analysis: The moisture content (M), volatile matter (VM), ash (A), and fixed carbon (FC) of the samples were measured according to ASTM standard methods for solid fuels. The procedures followed ASTM D3172 guidelines (using ASTM D3173 for moisture, ASTM D3175 for volatile matter, and ASTM D3174 for ash determination). Moisture content was determined by oven-drying the sample at 105 °C to constant weight and calculated as the percentage of mass lost (Equation (1)). Volatile matter was measured by heating the oven-dried sample in a covered crucible at 950 °C for 7 min; it is reported as the percentage of mass lost on a dry basis (Equation (2)). Ash content was obtained by igniting the sample residue in an open crucible at 750 °C until a constant mass of inorganic residue remained, and is expressed as the percentage of the initial sample that remains as ash (Equation (3)). The fixed carbon fraction was computed by difference, as the remaining percentage after subtracting M, VM, and A from 100% (Equation (4)). These proximate analysis results characterize the fuel’s basic composition and were later used in interpreting combustion behavior and yields.

$$M(\%)={\displaystyle \frac{m_{\mathrm{as}\text{-}\mathrm{received} }-m_{\mathrm{dried} }}{m_{\mathrm{as}\text{-}\mathrm{received} }}}\times 100$$

(1)

$$VM(\%)={\displaystyle \frac{m_{\mathrm{dried} }-m_{\mathrm{post}\text{-}\mathrm{VM} }}{m_{\mathrm{dried} }}}\times 100$$

(2)

$$A(\%)={\displaystyle \frac{m_{\mathrm{ash} }}{m_{\mathrm{dried} }}}\times 100$$

(3)

$$FC(\%)=100-VM-A$$

(4)

In the above formulas, $m_{as-received}$ is the sample mass before drying, $m_{dried}$ is the mass after drying, $m_{post-VM}$ is the mass remaining after the volatile matter step (before ashing), and $m_{ash}$ is the mass of the residue after complete combustion.

Ultimate Analysis: The elemental composition of carbon (C), hydrogen (H), nitrogen (N), and sulfur (S) in the samples was determined using an elemental analyzer following ASTM D5373. Oxygen (O) was calculated by difference (on a dry basis) as O = 100% − (C + H + N + S + ash). The ultimate analysis provides insight into the chemical makeup of the fuels and informs discussion of combustion emissions and energy content.

Calorific Value: The gross calorific value (GCV, higher heating value) of each sample was measured with an isoperibolic bomb calorimeter according to ASTM D5865 [29]. Each sample’s GCV (reported in kJ kg⁻¹ and equivalent kcal kg⁻¹) was used, in combination with the mass loss data, to evaluate the energy densification achieved through pelletization and torrefaction (see energy yield in Section 2.6).

Thermogravimetric Analysis: Thermogravimetric analysis (TGA) was performed to study the thermal degradation behavior of the raw and torrefied materials. Approximately 10–20 mg of each sample was tested in a TGA 701 analyzer (LECO Corp., St. Joseph, MI, USA). The sample was heated from 50 °C up to 900 °C at a constant rate of 10 °C min⁻¹ under an airflow of 10 L min⁻¹. The TGA recorded the mass loss as a function of temperature, and the derivative thermogravimetric (DTG) curve was obtained to identify key weight-loss events. Characteristic features such as the moisture release, the main devolatilization peak, and the residual mass at 900 °C were noted for each sample. The TGA/DTG results were later used to compare the combustion profiles of different pellet formulations and to assess how torrefaction severity and blend ratio influence the thermal stability of the fuel.

Density Measurements: The density of the pelletized fuels was characterized at both the particle (pellet) level and the bulk level. The true density of individual pellets was determined by measuring each pellet’s mass and physical dimensions. Assuming a cylindrical shape, the true density (solid density) was calculated as the pellet mass divided by its geometric volume (π r² h) (Equation (5)), where r is the pellet radius and h is its length. At least three pellets per sample were measured and averaged. The bulk density of the pellets was measured using a standard container method: a container of known volume was gently filled with pellets (allowing them to settle under gravity), then the total mass of pellets in the filled volume was recorded. The bulk density was calculated as the mass of pellets divided by the container volume (Equation (6)). This provides an indicator of the packing density and handling characteristics of the pelletized fuel.

$${\rho }_{\mathrm{true} }={\displaystyle \frac{m}{V}},V=\pi r^{2}h$$

(5)

$${\rho }_{\mathrm{bulk}}={\displaystyle \frac{m_{\mathrm{filled}}-m_{\mathrm{empty}}}{V_{\mathrm{container}}}}$$

(6)

Compressive Strength: The mechanical durability of the pellets was evaluated via uniaxial compression testing. Pellet compressive strength was measured using a universal testing machine (Zwick/Roell Z010) with flat steel compression platens. Individual pellets (approximately 8 mm in diameter and 25–30 mm in length) were compressed at a constant crosshead speed of 1.0 mm·min⁻¹ until structural failure. The maximum load at break was recorded for each pellet, and the compressive strength σ (in MPa) was calculated as this peak load divided by the pellet’s initial cross-sectional area (Equation (7)). The load–displacement (crush) curves from the tests were also analyzed to assess failure behavior. From each force–displacement curve, the following quantities were determined: the peak compressive strength (${\sigma }_{max}$), the compressive strain at peak load (${\epsilon }_{max}$), the compressive breaking stress (${\sigma }_B$, MPa at failure), and the breaking strain (${\epsilon }_B$, % at failure). For each pellet formulation and condition, at least three pellets were tested to ensure reproducibility of the results. The average strength values (± standard deviation) are reported alongside density and yield data to elucidate how processing and composition affect pellet mechanical integrity.

$${\sigma }_c={\displaystyle \frac{F_{max}}{A_0}},A_0=\pi r^2$$

(7)

In Equation (7), $F_{max}$ is the maximum force at failure and $A_0$ is the pellet’s initial cross-sectional area.

2.5. Statistical Analysis and Optimization

To systematically evaluate and optimize the effects of torrefaction parameters on pellet fuel properties, a three-factor Response Surface Methodology (RSM) approach was employed using Design-Expert^® software (Version 13.0.5.0, Stat-Ease Inc., Minneapolis, MN, USA). A Box–Behnken Design (BBD) was selected due to its efficiency in developing second-order (quadratic) models while requiring a relatively small number of experimental runs and avoiding extreme combinations of factor levels. The Box–Behnken design consisted of 17 experimental runs, including 12 edge points and 5 replicated center points to estimate experimental error and assess model adequacy.

Three independent variables were considered: torrefaction retention time (A, min), TWS:TCW blending ratio (B), and torrefaction temperature (C, °C). Each factor was investigated at three coded levels (−1, 0, +1), corresponding to the actual values presented in

Table 1. Independent variables and coded levels used in the Box–Behnken design.

Factor	Symbol	Unit	−1	0	+1
Torrefaction Retention time	A	min	30	60	90
TWS:TCW Ratio	B	–	20	40	60
Torrefaction temperature	C	°C	200	220	240

. The selected ranges were determined based on preliminary experiments and practical processing constraints.

The Box–Behnken design consisted of 17 experimental runs, including combinations of middle-edge points and replicated center points to estimate experimental error and evaluate model adequacy. The experimental runs were randomized to minimize systematic bias, and no blocking was applied. A quadratic polynomial model was fitted to each response variable.

The general second-order polynomial model used to describe the relationship between independent variables and responses is given by:

Y = β₀ + β₁A + β₂B + β₃C + β₁₂AB + β₁₃AC + β₂₃BC + β₁₁A² + β₂₂B² + β₃₃C²

where Y represents the predicted response (mass yield, gross calorific value, energy yield, volatile matter, fixed carbon, or ash content); β₀ is the intercept term; β₁, β₂, β₃ are linear coefficients; β₁₂, β₁₃, β₂₃ represent interaction effects; and β₁₁, β₂₂, β₃₃ are quadratic coefficients. All variables were analyzed in coded form for statistical evaluation.

Model adequacy and statistical significance were assessed using analysis of variance (ANOVA), including F-values, p-values, coefficient of determination (R²), adjusted R², predicted R², and lack-of-fit tests. A model was considered statistically significant at p < 0.05. The relative influence of each factor was interpreted based on the magnitude of F-values and corresponding significance levels. Response surface plots and contour diagrams were generated to visualize interaction effects and identify optimal operating conditions. Multi-objective numerical optimization was subsequently performed using the desirability function approach to identify optimal process conditions.

2.6. Data Analysis

All experimental data were processed using standard software tools, and results are reported as mean values with associated variability. Each measurement for composition, calorific value, or strength represents the average of three replicate determinations, and the variability is indicated by the standard deviation. Graphical analysis and calculations were carried out using Microsoft Excel 2016, while statistical modeling (ANOVA, regression, and RSM optimization) was performed using Design-Expert. A significance level of α = 0.05 (95% confidence) was adopted for hypothesis testing to determine if differences or factor effects were statistically significant.

In addition to direct measurements, several performance metrics were calculated from the data. The pellet conversion rate was used to quantify pelletization efficiency, defined as the percentage of the initial feed mass that was recovered as successfully formed pellets (Equation (10)). For torrefaction outcomes, the mass yield represents the percentage of the pellet mass remaining after torrefaction (dry basis), and the energy yield represents the percentage of the initial energy retained in the torrefied pellets. The energy yield was computed as the mass yield multiplied by the ratio of the torrefied pellet’s GCV to the raw pellet’s GCV. The formulas for these calculated indices are given below:

$$\mathrm{Mass} \mathrm{Yield} (\%)={\displaystyle \frac{m_{\mathrm{torrefied}}}{m_{\mathrm{raw}}}}\times 100$$

(8)

$$\mathrm{Energy} \mathrm{Yield} (\%)=\mathrm{Mass} \mathrm{Yield}\times {\displaystyle \frac{{\mathrm{GCV}}_{\mathrm{torrefied}}}{{\mathrm{GCV}}_{\mathrm{raw}}}}$$

(9)

$$\mathrm{Pellet} \mathrm{Conversion} \mathrm{Rate} (\%)={\displaystyle \frac{m_{\mathrm{pellets}}}{m_{\mathrm{feed}}}}\times 100$$

(10)

In the above equations, $m_{feed}$ is the mass of the initial feedstock loaded for pelletization, $m_{pallets}$ is the mass of pellets produced from that feed (after cooling and sieving), $m_{raw}$ is the mass of pellets before torrefaction, and $m_{torrified}$ is the mass of the pellets after torrefaction. ${GCV}_{raw} \mathrm{a}\mathrm{n}\mathrm{d} {GCV}_{torrified}$ are the gross calorific values of the pellets before and after torrefaction, respectively.

3. Results

3.1. Fuel Composition (Proximate and Ultimate Analysis)

The proximate and ultimate analyses of the raw materials and pellet samples are summarized in

Table 2. Proximate and ultimate analysis of raw materials and biofuel pellets.

Analysis		Raw Material-1 (TWS)	Raw Material-2 (TCW)	Binder Additive (Starch)	Sample-1 TWS:TCW (20:80 with 5% Starch)	Sample-2 TWS:TCW (40:60)	Sample-3 TWS:TCW (60:80)	Sample-4 TWS:TCW (80:20)
Proximate Analysis	Moisture (%)	6.5	4.5	8.5	10.1	9.7	11.4	9.4
	Volatile Matter (%)	53.2	87.2	88.3	73.1	70.7	56	48.2
	Fixed Carbon (%)	6.7	6.5	2.3	6.5	6.7	12.1	9.7
	Ash (%)	33.6	1.8	0.9	10.3	13.2	20.5	32.7
Ultimate Analysis	Carbon (%)	28.81	42.21	39.35	39.39	37.05	34.59	21.02
	Hydrogen (%)	4.94	6.36	6.21	6.03	5.92	5.64	3.71
	Nitrogen (%)	4.83	0.55	0.62	0.48	1.35	1.54	1.89
	Oxygen (%)	60.39	50.83	53.81	53.92	55.31	57.72	72.61
	Total Sulphur (%)	1.03	0.05	0.01	0.18	0.37	0.51	0.77
Bulk Density (kg m−3)		492	160	400	554	569	572	648
Gross Calorific Value (kcal kg−1)		2773	3826	3557	3377	2968	2916	2690

. The results show a clear trend: increasing the share of textile cotton waste (TCW) in the blend improves fuel quality (higher calorific value, lower ash), whereas increasing textile wastewater sludge (TWS) has the opposite effect. The best-performing formulation was the 20:80 TWS:TCW pellet with 5% starch, which achieved the highest gross calorific value (GCV) of 3377 kcal kg⁻¹ (≈14.12 MJ kg⁻¹) alongside the lowest ash content (10.3 wt%), highest volatile matter (73.1 wt%), stable moisture (~10 wt%), and only trace nitrogen (0.48 wt%) and sulfur (0.18 wt%) content. In contrast, as the TWS fraction rose from 40% to 80% in the pellets, the GCV dropped steadily to 2968–2690 kcal kg⁻¹, ash content climbed from 13.2% to 32.7%, and bulk density increased from 569 to 648 kg m⁻³. Ultimate analysis explains these shifts: with higher TWS, the carbon content of the pellets falls sharply (from ~39.4% at 20% TWS to ~21.0% at 80% TWS) while oxygen rises (from ~53.9% to ~72.6%), and the fuel nitrogen and sulfur increase several-fold (N from 0.48% to 1.89%; S from 0.18% to 0.77%). These compositional changes dilute the combustible organic fraction and introduce more mineral matter and potential emissions precursors. The proportion of ash content rose with an increase in TWS, showing a strong correlation between predicted and experimental values (R² = 0.879), which verifies that ash exhibits a nearly linear additive blending behavior primarily determined by mass fraction contribution. Conversely, while GCV declined as TWS content increased, the predicted values had a poor correlation with the experimental data (R² = −0.668), suggesting that the calorific value does not adhere strictly to linear mixing principles and is affected by non-linear thermochemical interactions during combustion. This indicates that mineral residue behaves in an additive manner, while energy characteristics are influenced by more complex combustion dynamics. (as indicated by the trendlines in Figure 2). Each incremental addition of sludge thus lowers the fuel’s energy density and increases its ash burden, whereas a higher cotton waste ratio produces a cleaner, more energy-dense fuel. Accordingly, the TCW-rich 20:80 (+5% starch) formulation is identified as the optimal blend for energy generation, maximizing calorific value and minimizing ash and pollutant-forming elements.

These composition-driven trends are consistent with literature on biomass–sludge fuels. Cotton-rich biomass provides abundant volatiles and low ash content, which aligns with thermogravimetric evidence on textile wastes showing dominant volatile release and minimal residue. In contrast, sludge-derived materials contain more inorganic matter (e.g., minerals, bound water) and less fixed carbon, leading to lower heating values and higher ash yields. Consequently, many co-combustion studies report that adding sewage sludge to a fuel blend depresses the calorific value and increases the ash, nitrogen, and sulfur content of the fuel. The near-linear decrease in GCV with increasing sludge observed in this work mirrors those reports. For example, Castells et al. [30] found that sludge-rich solid fuels inherently have lower energy content and higher ash due to their composition. Likewise, the rise in pellet nitrogen and sulfur with higher TWS is noteworthy because it portends higher NOx/SO₂ emissions during combustion; this agrees with studies showing that sludge-derived fuels tend to produce more NOx and SO₂ unless mitigated. Conversely, introducing a low-N, low-S biomass like cotton waste can dilute these precursors and even lead to net reductions in NOx/SO₂ at moderate biomass blending ratios [21]. Overall, our findings confirm that a cotton-heavy blend yields superior fuel characteristics (high energy, low ash, low N/S), while excessive sludge loading impairs fuel quality. The selected 20:80 (TWS:TCW) recipe with a small starch binder achieved the best balance of high energy content and acceptable ash levels, in line with recommended practice for optimizing sludge–biomass solid fuels.

3.2. Mechanical and Physical Properties

Physical and mechanical testing revealed that pellet quality depends more on inter-particle bonding and feedstock balance than on density alone. The TCW-rich pellet with binder (Sample-1, 20:80 TWS:TCW + 5% starch) achieved the highest overall performance, with a pelletization conversion rate of ~96%, a maximum compressive strength of 14.9 N·mm⁻², a compressive breaking stress of 4.47 MPa, and exceptional ductility (peak strain ~33.0% and breaking strain 40.3%). These values are markedly higher than those of the same formulation without binder (Sample-0, 20:80 without starch), which reached only 9.6 N·mm⁻² maximum strength, 2.87 MPa breaking stress, and ~30.9% breaking strain. The stark improvement with the addition of 5% starch confirms the binder’s role in enhancing inter-particle bonding, filling voids, and strengthening the pellet matrix. Increasing the sludge content in the blend had the opposite effect on mechanics: as TWS rose from 40% (Sample-2) to 80% (Sample-4), the pellets became denser but mechanically weaker. True density increased monotonically from 1209 to 1244 kg m⁻³ (and bulk density from 569 to 648 kg m⁻³) for Samples 2–4, yet the maximum compressive strength dropped from 12.1 to 10.8 N·mm⁻² and the breaking strain fell from ~25.5% to ~28.3% (a slight decrease in ductility). The higher mineral content in sludge-rich pellets likely creates a more brittle, fragile structure despite the density gain. Thus, the 20:80 TCW-rich pellet with binder emerged as the most robust formulation, attaining superior strength and toughness while still maintaining a moderate density (true 1159 kg m⁻³, bulk 554 kg m⁻³). This combination of high fiber content and binder produces pellets that resist crushing and handle better than either binder-less or sludge-heavy pellets.

The force–displacement (crushing) behavior of the pellets (see Figure 3) further illustrates these differences. The starch-bound 20:80 pellet (Sample-1) sustained the highest peak load (~730 N at about 32–33% compression) and exhibited a gradual failure beyond that point, indicating a tough and ductile pellet. In comparison, the binder-free 20:80 pellet (Sample-0) showed a lower peak force (~500 N at ~27% compression) and failed at a smaller strain, and the sludge-rich pellets (Samples 2–4) also registered lower peak forces (~520–580 N) with earlier drop-offs in the force–crush curve. Notably, although pellet density consistently increased from Sample-0 through Sample-4 (true density ranging ~737→1244 kg m⁻³,

Table 3. Physical and mechanical properties of biofuel pellets (dimensions, density, strength, strain, etc.).

Analysis	Sample-0 TWS:TCW (20:80)	Sample-1 TWS:TCW (20:80 + 5% Starch)	Sample-2 TWS:TCW (40:60)	Sample-3 TWS:TCW (60:80)	Sample-4 TWS:TCW (80:20)
Initial Diameter of Specimen (mm)	8.0	7.8	7.8	7.8	7.8
Initial Cross Section Area (mm2)	50.27	47.78	47.78	47.78	47.78
True Density (kg m−3)	737	1159	1209	1214	1244
Pellet Conversion Rate (%)	70	96	81	83	87
Maximum Compressive Strength (N/mm2)	9.6	14.9	12.1	11.6	10.8
Maximum Compressive Strain (%)	26.9	33	20	19	21
Compressive Breaking Stress (MPa)	2.87	4.47	3.62	3.48	3.24
Compressive Breaking Strain (%)	30.9	40.3	25.5	26.6	28.3

), the peak mechanical performance occurred at Sample-1 rather than the densest sample. This underscores that simply maximizing density does not guarantee strength; instead, an optimal composition with sufficient fibrous content and a small amount of binder yields the best mechanical integrity. In practical terms, the TCW-rich, binder-assisted pellet combines high density with resilient internal bonding, which is advantageous for handling and transportation (fewer fines and breakages during conveying). The 20:80 +5% starch formulation thus provides a durable, transportable solid fuel suitable for energy applications, whereas pellets with excessive sludge or no binder are more prone to crumbling or fracturing under stress.

The observed mechanical performance trends are supported by prior studies on biomass pelletization. Pellet strength and durability are known to depend strongly on inter-particle bonding, pore structure, and the presence of binders. In particular, adding a small amount of starch (or similar organic binder) can significantly improve pellet cohesiveness and compressive strength, as it promotes fiber binding and matrix gelatinization during pressing. Sykorova et al. [18] reported that biomass pellets with an organic binder showed enhanced mechanical properties even after torrefaction, highlighting that binder effects persist through mild thermal treatments. Our results concur: the starch-bound cotton-rich pellet exhibited ~55% higher strength than the same blend without binder, which agrees with evidence that optimally dosed binders increase pellet robustness without overly raising ash content. Meanwhile, the fact that higher sludge (ash) content led to denser but weaker pellets is also explained by literature. Excess minerals in the pellet can disrupt the formation of strong fiber–fiber bonds and create stress concentration points, yielding a more brittle pellet despite higher compaction. High inorganic filler content reduces pellet durability due to weaker interparticle bridges. In contrast, cotton-rich matrices provide ample cellulose and lignin that, under compression heat (~70–100 °C in the die), plastically deform and bind together, resulting in higher cohesion. The superior strength and ductility of our 20:80 + starch pellet, despite not being the most dense, is a direct consequence of this optimized internal architecture. Moreover, surveys of commercial pellets emphasize that high calorific value, low ash, and strong mechanical durability are key quality attributes for end-users. The cotton-rich pellet in this study delivers on all three fronts (high GCV, low ash, high durability), which would translate into reduced fines generation and smoother handling in industrial feeding systems. Overall, the mechanical results underscore that a TCW-rich blend with a modest binder yields a pellet that is not only energy-rich but also physically robust, an important consideration for its practical use as a coal replacement.

3.3. Thermogravimetric Analysis (TGA–DTG)

In this study, TGA was conducted under air atmosphere intentionally because our focus was on evaluating the oxidative decomposition and combustion behavior of the torrefied and raw samples, rather than purely pyrolytic decomposition. Conducting TGA in air provides insight into the thermal stability, ignition, and burnout characteristics of the material under conditions closer to practical combustion applications, which is relevant for co-pelletized sludge–textile cotton fuels. While pyrolysis behavior under inert gas can provide mechanistic information about devolatilization, the use of air better represents real-world combustion conditions, allowing us to assess energy release and combustion kinetics in an oxidative environment like biomass Boiler.

Thermogravimetric analysis was used to evaluate the thermal decomposition behavior of the pellets and their raw feedstocks, since this influences ignition and burnout in combustion applications. The TGA–DTG curves (Figure 4) reveal distinct patterns for cotton waste versus sludge and show how the pellet blends combine these behaviors. Pure TCW (textile cotton waste) undergoes a single, sharp mass-loss event: a major devolatilization peak appears around 300–360 °C on the DTG, corresponding to rapid decomposition of cellulose and other organics, and only a small char residue (~5–10% of mass) remains at 800 °C. This reflects cotton’s high volatile matter and low ash content (as also seen in proximate analysis). By contrast, pure TWS (textile sludge) exhibits a multi-stage thermal degradation. Its DTG profile shows a broad, low-temperature dehydration shoulder, followed by multiple overlapping decomposition peaks, and a long tail of slow weight loss extending up to ~800–900 °C. Even at 900 °C, sludge leaves a large residual mass, on the order of one-third of its initial weight, due to its high ash (mineral) content and thermally recalcitrant components. These baseline observations are consistent with the known behavior of cellulosic vs. sludge materials: cotton (and other textiles) combust mostly in a narrow mid-temperature window, whereas sewage sludges decompose over a wider range and yield considerable char and ash.

All pellet samples show a composite of these characteristics, with the balance governed by the TWS:TCW ratio. Initially, a small weight-loss step below ~150 °C occurs for all pellets (moisture evaporation). Thereafter, a dominant mass-loss peak appears between ~280–380 °C for all pellets, attributable to the pyrolysis and combustion of cotton fibers (cellulose, hemicellulose) and volatile organics. One can see in Figure 4 that Sample-1 (20:80, TCW-rich) has a very pronounced main peak in this range and relatively little residue after this point, whereas Sample-4 (80:20, sludge-rich) shows a more subdued main peak and a higher residue. As the sludge fraction increases (from Sample-1 through Sample-4), two clear trends occur: (1) the main DTG peak becomes progressively smaller and broader—indicating less volatile content and a slower overall combustion rate—and (2) the high-temperature tail (500–800 °C) becomes more prominent, with final residues rising from about 10% (for Sample-1) to roughly 30% (for Sample-4). In Sample-4, for instance, the DTG curve shows secondary shoulders in the 400–600 °C range, likely due to the decomposition of fixed carbon and inorganic-associated compounds in the sludge. These observations mean that cotton-rich pellets ignite and burn out more completely, leaving little char, whereas sludge-rich pellets combust in multiple stages and leave significantly more unburnt residue (ash/char). Sample-1 (20% TWS) achieves the most complete burnout of all pellets, with the smallest char residue at the end of the TGA run, which is in line with its lower ash content and higher fuel value. In contrast, the 80% TWS pellet, while heavier and denser, retains a substantial fraction uncombusted, underscoring the trade-off between adding sludge and maintaining clean combustion.

Thermogravimetric patterns observed here closely match those reported in other studies of textile and sludge materials. Cotton-based biomass typically shows a dominant mid-temperature decomposition peak (around 300 °C) associated with cellulose, and a narrow combustion interval with little char remaining. Our TCW sample’s single sharp DTG peak and low residue are in line with this behavior. Sludge, on the other hand, often exhibits a broad thermal profile with multiple degradation steps due to its mix of organic and inorganic constituents. The extended high-temperature tail and large residue we observed for TWS reflect its high mineral content and the presence of thermally stable compounds, agreeing with prior TGA studies on sewage sludge. When these materials are combined, the literature similarly notes an intermediate behavior: Arjona et al. [31] reported that adding textile waste to sludge produces a pronounced cellulose decomposition peak and reduces the final char fraction, although the sludge component still causes additional weight-loss stages at higher temperatures. In our pellets, the progressive weakening of the main DTG peak and the growth of the high-temperature tail as TWS increases mirror findings that sludge-rich blends burn more slowly and leave more char than biomass-rich ones. For example, co-combustion experiments by Wang et al. [8] showed that introducing textile waste into sludge improved the overall combustibility—increasing volatile release and reducing char residue—which is consistent with our observation that cotton-rich pellets have a more complete burn profile. On the contrary, blends with high sludge content were found to need longer residence times or higher temperatures to achieve complete burnout, supporting our conclusion that the 80:20 pellet would require more aggressive combustion conditions. In practical terms, the cotton-rich pellets (with their sharper combustion and lower residual) should allow for easier ignition and more stable combustion in boilers, whereas sludge-heavy pellets may pose challenges like longer burnout times and greater ash-handling needs. These results underscore the combustion benefits of a TCW-rich formulation: it concentrates the fuel’s volatile content for quick ignition and minimizes residual ash, echoing recent findings that low-ash biomass additions can streamline combustion of waste-derived fuels.

3.4. Torrefaction Performance

Thermal upgrading of the pellets via torrefaction showed significant improvements in fuel quality, especially for the cotton-rich formulation. A series of torrefaction experiments was performed by varying temperature (200–240 °C), residence time (30–90 min), and blend ratio, and the key results are given in

Table 4. Torrefaction results for biofuel pellets (mass yield, energy yield, GCV, proximate values at various conditions).

Sample Name	Run	Torrefaction Time (Min)	Ratio TWS:TCW	Torrefaction Temp. (°C)	Mass Yield (%)	GCV (kcal kg−1)	Energy Yield (%)	Volatile Matter (%)	Fixed Carbon (%)	Total Ash (%)
CT1	13	60	Sample-1 (20:80)	200	96	3563	101.3	80.5	6	10.6
CT2	7	90		200	92	3619	98.6	73.6	10.1	11.8
CT3	11	30		220	92	4083	111.2	75.4	9.2	11.8
CT4	2	90		220	58	4362	74.9	65.4	17.4	10.6
CT5	8	60		240	18	4565	24.3	30.1	33.2	32
CT6	3	90		240	16	1220	5.9	18.7	9	68.2
CT7	12	30	Sample-2 (40:60)	200	97	3614	118.3	72.4	4.4	18.1
CT8	5	90		200	97	3241	105.6	74.3	7	14.6
CT17	14	30		220	96	3516	115.3	67.3	10.7	19.2
CT11	10	60		240	39	3581	47	36.7	29	29.3
CT10	6	30		240	49	4158	69.1	39.8	27.3	29.2
CT9	1	60		220	40	4017	54.5	34.2	29.8	33.3
CT12	9	60	Sample-3 (60:40)	200	97	3232	107.2	61.4	12	22.8
CT13	16	30		220	93	3301	105.7	62	10	21.3
CT14	4	90		220	63	3376	73.4	33.9	25	34.8
CT15	17	60		240	64	3452	75.3	34.2	22	40
CT16	15	90		240	55	2909	54.7	35.4	19	42.7

. For the optimal 20:80 TWS:TCW pellet (with 5% starch), mild torrefaction at 200 °C had only a modest effect: after 60–90 min, the mass yield remained high (≈92–96% of the original mass retained) and the GCV increased slightly to ~3560–3620 kcal kg⁻¹ (about 5–7% higher than the untorrefied pellet’s 3377 kcal kg⁻¹). Volatile matter decreased marginally and fixed carbon increased correspondingly under these mild conditions. As torrefaction severity increased, more dramatic changes were observed. At 220 °C (for 60–90 min), the 20:80 pellet underwent substantial devolatilization—mass yield dropped to ~58% at the longer duration—but the GCV was boosted to roughly 4360 kcal kg⁻¹ (18.2 MJ kg⁻¹), a ~30% increase in heating value. The fixed carbon content roughly tripled compared to the raw pellet, indicating a significant concentration of carbon due to the loss of volatiles. At the highest tested severity of 240 °C, energy densification reached a maximum: the 20:80 pellets torrefied for 60 min attained a GCV of ~4565 kcal kg⁻¹ (~19.1 MJ kg⁻¹) and very high fixed carbon content, essentially transforming into a coal-like solid. However, this came at the expense of an extreme mass loss (mass yield <20% at 240 °C/60 min), meaning over 80% of the initial mass was driven off as volatiles and moisture. The ash content in the remaining torrefied product also appeared higher (around 32% in the 240 °C case, versus 10.3% originally) simply because the inorganic portion was left behind in a much-reduced total mass. These results illustrate the classic trade-off in torrefaction: more severe treatment yields a more energy-dense fuel, but with substantially lower solid yield.

Heavier sludge content in the pellets was found to limit the effectiveness of torrefaction upgrading. The 40:60 TWS:TCW pellet (Sample-2) and 60:40 pellet (Sample-3) did show increases in GCV upon torrefaction, but the improvements were smaller relative to the cotton-rich case. For instance, at 220–240 °C the 40:60 pellet’s GCV leveled around 3800–4200 kcal kg⁻¹, and the 60:40 pellet achieved at most ~3450 kcal kg⁻¹, compared to 4360–4565 kcal kg⁻¹ for the 20:80 pellet. Their fixed carbon contents rose only moderately, and because the initial ash content was higher, the resulting torrefied pellets still had ash levels well above 20%. Moreover, the mass yields for these sludge-richer pellets remained relatively high even at 240 °C (e.g., ~39–64%), implying that a larger fraction of non-combustible residue was carried through. In essence, while torrefaction universally raised the energy content and carbon fraction of all pellets, the relative benefit was greatest for the cotton-dominated formulation. The 20:80 pellets responded very well to torrefaction by shedding moisture and volatiles to greatly increase GCV, whereas pellets with higher TWS could not reach similar energy densities because the inorganics and less-volatile matter in sludge capped their improvement. From these results, the optimal torrefaction window for the 20:80 pellet appears to be around 220–240 °C for 1 h: this range produced a substantial jump in GCV (to ~18–19 MJ kg⁻¹) while still retaining 50–60% of the mass, an attractive compromise for fuel upgrading. More severe treatment (240 °C for longer times) gave only marginal GCV gains but incurred excessive mass loss, so it would likely be impractical despite yielding the highest quality solid fuel. In this work, the torrefaction temperature was kept within 200–240 °C (Table 4) based on how the materials actually behaved during the experiments. Although the highest calorific value (4565 kcal kg⁻¹) was reached at 240 °C for 30 min (Sample-1), this condition led to a very low mass yield of only 18%, with a strong drop in energy yield 24.3%. At the same temperature, Sample-2 and Sample-3 showed similar trends, where modest gains in calorific value were accompanied by substantial losses in solid yield. In practical terms, this means that most of the fuel is lost while trying to increase its energy density. Pushing the temperature beyond 240 °C would further intensify devolatilization and move the process closer to pyrolysis, making it inefficient for producing a solid fuel. For this reason, 240 °C was considered the practical upper limit in the present study.

Torrefaction experiments for the optimal 20:80 TWS:TCW blend were conducted only on pellets containing 5 wt.% corn starch, as binder-free pellets of this composition exhibited poor pelletization efficiency and insufficient mechanical integrity during preliminary trials. The addition of starch significantly improved pellet conversion rate (70% to 96%), true density (737 to 1159 kg m⁻³), and compressive strength (2.87 to 4.47 MPa), thereby enabling stable handling and thermochemical treatment. From a compositional perspective, starch (0.9 wt.% ash; 3557 kcal kg⁻¹) constitutes only 5 wt.% of the total blend and exhibits intermediate fuel properties relative to TWS and TCW. A quantitative mass balance calculation indicates that incorporation of starch alters the theoretical ash content by only 0.04 wt.% (8.16 to 8.20 wt.%) and the gross calorific value by approximately 3 kcal kg⁻¹ (3615 to 3612 kcal kg⁻¹) compared with the binder-free 20:80 blend. These minimal differences demonstrate that the thermochemical contribution of starch is negligible, and that torrefaction behavior is primarily governed by the TWS–TCW matrix.

Because binder-free 20:80 pellets were not subjected to torrefaction due to mechanical instability, the reported performance corresponds to a binder-assisted system. Direct experimental comparison with a starch-free torrefied counterpart was beyond the scope of the present study and represents a limitation.

The changes in elemental composition due to torrefaction were quantified for the 20:80 pellet, as shown in

Table 5. Ultimate analysis and GCV of the 20:80 pellet before and after torrefaction under selected conditions.

Sample Name	Sample Recipe	Carbon (%)	Nitrogen (%)	Sulfur (%)	Hydrogen (%)	Oxygen (%)	GCV (kcal kg−1)	Mass Yield (%)
Sample-1 (Untorrefied Pellets)	(20:80 with 5% starch)	39.39	0.48	0.18	6.03	53.92	3377	100
Sample-CT3 (Torrefied Pellets)	20:80 | 220 °C | 30 min	47.3	0.18	0.31	5.22	46.53	4083	92
Sample-CT4 (Torrefied Pellets)	20:80 | 220 °C | 90 min	49.21	1.16	0.25	4.78	44.6	4362	58
Sample-CT5 (Torrefied Pellets)	20:80 | 240 °C | 60 min	51.52	1.29	0.36	3.76	43.06	4565	18

. Carbon content increased from 39.39% in the raw pellet to 47–52% after torrefaction at 220–240 °C, while oxygen content dropped from ~53.9% to about 44–46%. Hydrogen content also decreased (from 6.03% to ~3.8–5.2%), which is expected as dehydration and devolatilization remove hydrogen-rich compounds. Nitrogen and sulfur percentages showed minor changes (e.g., N rose from 0.48% to ~1.16% at 220 °C/90 min, likely concentrated by mass loss; S increased slightly from 0.18% to ~0.25–0.36%). These ultimate analysis shifts confirm that torrefaction drives off oxygenated volatiles (CO₂, CO, H₂O, etc.), enriching the solid with carbon. The net result is a higher-energy, more carbonaceous fuel, which is reflected in the GCV climbing from ~3377 kcal kg⁻¹ (untreated) to as high as 4565 kcal kg⁻¹ after torrefaction (Table 5). The torrefied pellets thus approximate the elemental makeup of low-rank coals, with carbon contents exceeding 50% and O/C ratios greatly reduced. This is also evident in their TGA behavior: thermogravimetric curves for torrefied vs. raw pellets (Figure 5) show that torrefaction shifts the decomposition to higher temperatures and reduces the intensity of the main volatile combustion peak. For example, the DTG of the raw 20:80 pellet has a strong peak around 320 °C, whereas after torrefaction at 220–240 °C the peaks become smaller and occur at slightly higher temperatures (indicating more stability). The torrefied samples also leave a higher char residue in TGA, consistent with their increased fixed carbon content. Among the tested torrefaction conditions, 220 °C for 90 min (Sample CT4) appeared to offer the best balance: it yielded a carbon-rich pellet (49.2% C, 0.25% S) with a high GCV (~4362 kcal kg⁻¹) while retaining about 58% of the mass. The more severe 240 °C for 60 min condition (Sample CT5) pushed the GCV slightly higher (4565 kcal kg⁻¹) but left only 18% of the mass—such a low mass yield may not be economical despite the fuel quality. These observations underscore that an intermediate torrefaction severity is optimal for maximizing energy yield from cotton-rich pellets.

The torrefaction trends observed in this study align with the general understanding of biomass torrefaction and are supported by recent research. There is a well-known trade-off between mass yield and energy densification during torrefaction. As temperature and time increase, more volatiles are released (reducing mass), but the remaining solid becomes richer in carbon and energy content. Our findings exemplify this: higher torrefaction temperature was the dominant driver for increasing GCV and fixed carbon, at the cost of mass loss, whereas milder conditions preserved more mass with smaller energy gains. In the present response surface framework, the combined thermal severity effect is inherently captured through the interaction (AC) and quadratic terms of retention time and temperature, thereby allowing independent as well as coupled evaluation of these parameters. Our findings exemplify this severity trade-off: increasing torrefaction severity (higher temperature and/or longer residence time) increased GCV and fixed carbon, but reduced mass yield, whereas lower-severity conditions preserved more mass with smaller gains in energy content (Figure 5 and Figure 6). This trend is consistent with our RSM–ANOVA results (

Table 6. Analysis of variance (ANOVA) for the quadratic model of mass yield (Ymass).

Factor			Coefficient	F-Value	p-Value and Remarks
A—Torrefaction Retention Time			−0.40524	10.25	0.0150	Significant
B—Ratio (TWS:TCW)			−7.60482	3.31	0.1115	–
C—Torrefaction Temperature			−4.55956	45.73	0.0001	Highly Significant
A × B Interaction			+0.004673	0.28	0.6132	–
A × C Interaction			−0.008260	0.77	0.4103	–
B × C Interaction			+0.025769	4.04	0.0483	Significant
A2—Quadratic (Torrefaction Retention Time)			+0.013129	2.93	0.1309	–
B2—Quadratic (Ratio TWS:TCW)			+0.025394	1.97	0.2037	–
C2—Quadratic (Torrefaction Temperature)			+0.006053	0.12	0.7420	–
Intercept (Constant)			+876.55008	–	–	–
Model (Overall)			–	8.74	0.0046	Model Significant
Fit Statistics	R2	Adjusted R2	Predicted R2	Adeq Precision	Mean	Std. Dev.	C.V. %
	0.9182	0.8131	0.5250	9.5033	68.12	12.70	18.64

), which show torrefaction temperature as the dominant factor across key responses (e.g., mass yield: F = 45.73, p = 0.0001; energy yield: F = 43.60, p = 0.0003; volatile matter: F = 28.46, p = 0.0001), with residence time generally secondary but still significant for several outcomes (mass yield: F = 10.25, p = 0.0150; energy yield: F = 16.19, p = 0.0050; volatile matter: F = 5.38, p = 0.0355), while blend ratio mainly modulates the magnitude of change and is often non-significant. Castells et al. [30] Recently reported that in sewage sludge torrefaction, temperature had a much larger influence on energy yield and fuel properties than the sludge blend ratio or time, which is in line with our observation that all pellets benefited from higher temperatures, especially the cotton-rich ones. Furthermore, the suggestion that an intermediate torrefaction severity (~230 °C, ~1 h) is optimal for lignocellulosic biomass is echoed in the literature. Studies on wood and agricultural residues note that going beyond about 250 °C yields diminishing returns in GCV while sharply cutting mass yield [32,33]. Our result that 220–240 °C gave a large GCV boost (to ~18–19 MJ kg⁻¹) while retaining roughly half the mass is consistent with those reports. It was also observed that the quadratic model could not adequately fit GCV in our optimization (discussed below), which is unsurprising since the energy content does not vary linearly or quadratically across wide torrefaction ranges—others have noted non-linear plateaus or threshold effects in GCV beyond certain temperatures.

Another important aspect is how torrefaction affected ash and inorganic matter in the pellets. Interestingly, we found that for the cotton-rich 20:80 pellet, the measured ash content (10.3%) after torrefaction increased slightly at moderate conditions (e.g., from 10.6% to 11.8% at 220 °C, see Table 4), even though at very severe conditions it increased due to mass concentration. A plausible explanation is that mild torrefaction can volatilize or redistribute some inorganic constituents (such as loss of chlorine, or conversion of certain minerals to gaseous species). Prior studies have observed small reductions in ash or mineral content for biomass upon torrefaction, as some alkali chlorides or organo-mineral compounds may decompose or sublimate [34,35]. Our observation that higher torrefaction temperature and longer time tended to increased ash in the cotton-rich pellets is in line with those findings. This suggests that when the initial mineral content is large (as in 60–80% TWS samples), the fraction of inorganics that can be driven off is negligible compared to the total. Other studies have similarly shown that feedstocks with elevated ash, such as sewage sludge, display this trend during torrefaction, where mineral nutrients like phosphorus and potassium become more concentrated while volatile elements like nitrogen and sulfur diminish, leading to a relative increase in ash content [36,37]. In our case, the 60:20 pellet started with ~20.5% ash and increased around 21–34% at 220 °C after torrefaction (Table 4), indicating no significant improvement. The primary way to manage ash in such cases is still through blend selection (adding low-ash biomass) rather than thermal treatment. Nonetheless, for the cotton-rich blend, torrefaction’s slight ash-reducing effect (along with its major moisture/volatile reduction) could be beneficial for combustion—e.g., lowering slagging/fouling propensities a bit—but it should be noted that composition (blend ratio) remains the dominant factor determining absolute ash content after torrefaction.

Torrefied sample CT6, a 20:80 blend rich in TCW with very high volatile matter content (73.1%), exhibited an unusually low gross calorific value (GCV) of 1220 kcal kg⁻¹ under severe torrefaction conditions (240 °C for 90 min). This drastic reduction in energy content resulted from over-severe torrefaction, during which almost all combustible carbon and hydrogen were volatilized, leaving a residual solid that is largely non-combustible and oxygen-rich. The mass yield dropped to 16%, indicating that most of the original pellet mass was lost. The extensive devolatilization and strong concentration of inorganic components increased the ash content to 68.2%, further diluting the fraction of material contributing to heating value. In contrast, blends with higher TWS ratios (40:60 and 60:80) were better able to withstand high temperature and long residence time, maintaining higher energy density. The extreme sensitivity of the TCW-dominated 20:80 blend explains the unusually low GCV, which, while deviating from the gradual trends observed in other samples (e.g., CT5: GCV 4565 kcal kg⁻¹, mass yield 18%), accurately reflects the physical limits of torrefaction at such severe conditions.

It is important to note that torrefaction in this study was conducted in an oxygen-limited furnace without continuous inert-gas purging. Therefore, although flaming combustion was avoided, minor oxidation may have occurred during heating, particularly at higher temperatures. Such limited oxidation could slightly influence the measured mass yield (additional mass loss) and elemental composition (e.g., lower carbon retention or higher oxygen content relative to fully inert torrefaction). However, since identical atmospheric conditions were applied consistently to all experimental runs, the comparative trends across torrefaction temperature, residence time, and blend ratio, as well as the RSM-based optimization conclusions, remain valid.

The statistical significance of the observed trends and the quantitative effects of process variables were further evaluated using analysis of variance (ANOVA) within the response surface modeling framework, as discussed in Section 3.5.

3.5. Response Surface Modeling and Optimization

3.5.1. Effect of Process Variables on Mass Yield

This study performed the mass yield optimization to understand how torrefaction conditions (temperature, retention time, and TWS:TCW ratio) influence solid recovery during the conversion of textile and sludge waste into biofuel pellets. The purpose of this analysis was to determine the balance point between mass conservation and energy densification—in other words, to identify when material loss becomes beneficial because it leads to higher fuel quality. The statistical model was applied to evaluate how each variable contributes and interacts under controlled thermal treatment. The ANOVA results (Table 6) confirm that torrefaction temperature (C) and retention time (A) are the main governing parameters for mass yield, with both exhibiting negative coefficients (C: −4.55956; A: −0.40524) and statistically significant effects (C: F = 45.73, p = 0.0001; A: F = 10.25, p = 0.0150). In addition, the BC interaction is significant (F = 4.04, p = 0.0483), indicating that blend ratio modifies the temperature effect on yield. As temperature increased from 200 °C to 240 °C and residence time from 30 to 90 min, yield progressively declined, confirming that higher thermal severity enhances devolatilization and moisture removal, resulting in reduced solid mass. The blend ratio (TWS:TCW) had a smaller influence individually, but its interaction with temperature (BC) was significant, implying that the composition of feedstock determines how strongly the material responds to heat. The model showed excellent reliability (R² = 0.918, Adj R² = 0.813, Adeq. Precision = 9.5) and strong agreement between predicted and experimental data, proving its predictive strength. These results demonstrate that torrefaction temperature has the greatest effect on yield reduction, while moderate conditions—around 220 °C for 60–70 min—offer an optimal balance, maintaining sufficient mass while achieving desirable thermal upgrading. In essence, the optimization confirmed that controlled torrefaction transforms waste biomass into a denser, more stable biofuel while avoiding excessive mass loss, making it efficient for sustainable energy generation. The final quadratic regression equations in terms of coded factors for the key responses are given below:

Ymass = 54.75 − 13.81A + 7.53B − 27.22C + 2.80AB − 4.96AC + 10.31BC + 11.82A² + 10.16B² + 2.42C²

The magnitude and sign of the coefficients indicate that torrefaction temperature (C) exerts the strongest negative linear effect on mass yield within the studied range.

3.5.2. Effect of Process Variables on Gross Calorific Value

The optimization analysis for GCV indicates that the constructed response-surface model is not statistically meaningful. As shown in

Table 7. Analysis of variance (ANOVA) for the quadratic model of gross calorific value (GCV).

Factor			Coefficient	F-Value	p-Value and Remarks
Intercept (Constant)			3518.18	–		–
Model (Overall)			–	0.00	–	–
Fit Statistics	R2	Adjusted R2	Predicted R2	Adeq Precision	Mean	Std. Dev.	C.V. %
	0.0000	0.0000	−0.1289	NA	3518.18	737.12	20.95

, the model collapses to a constant term (intercept = 3518.18 kcal kg⁻¹), which is simply the overall mean of the data. For GCV, only the intercept term remained significant in the reduced quadratic model, indicating limited sensitivity of calorific value to factor variations within the selected experimental range. The model F-value is effectively zero and no factor contributes significantly, reflected in R² = 0.0000, adjusted R² = 0.0000, and a negative predicted R² (−0.1289). The relatively high standard deviation (737.12 kcal kg⁻¹) and C.V. of 20.95% further point to large unexplained variability. The predicted-versus-actual plot shows points scattered around a horizontal band rather than along the 1:1 line, and the normal residual plot reveals clear departures from ideal behavior. Together, these diagnostics confirm that, within the studied design space, GCV cannot be reliably described or optimized by the chosen quadratic model and should therefore be interpreted descriptively rather than used for numerical optimization as shown in Figure 7.

3.5.3. Effect of Process Variables on Energy Yield

The optimization of energy yield demonstrated that the quadratic response-surface model accurately describes the torrefaction process. Torrefaction temperature was identified as the most influential factor, followed by retention time, both contributing positively to energy yield as shown in Figure 8. The TWS:TCW ratio had a minor and statistically insignificant effect, acting mainly as a secondary modifier. Statistical validation confirmed that the model aligns well with experimental data, as shown by the close agreement between predicted and actual values and the normal distribution of residuals as shown in

Table 8. Analysis of variance (ANOVA) for the quadratic model of energy yield (Yenergy).

Factor			Coefficient	F-Value	p-Value and Remarks
A—Torrefaction Retention Time			+0.351633	16.19	0.0050	Significant
B—Ratio (TWS:TCW)			−7.59484	2.99	0.1273	–
C—Torrefaction Temperature			+1.59121	43.60	0.0003	Highly significant
A × B Interaction			+0.006802	0.5013	0.5018	–
A × C Interaction			−0.013246	1.67	0.2376	–
B × C Interaction			+0.028231	4.10	0.0824	–
A2—Quadratic (Torrefaction Retention Time)			+0.013847	2.75	0.1410	–
B2—Quadratic (Ratio TWS:TCW)			+0.017059	0.7502	0.4151	–
C2—Quadratic (Torrefaction Temperature)			−0.007661	0.1590	0.7020	–
Intercept (Constant)			+207.44942	–	–	–
Model (Overall)			–	9.87	0.0032	Model significant
Fit Statistics	R2	Adjusted R2	Predicted R2	Adeq Precision	Mean	Std. Dev.	C.V. %
	0.9270	0.8330	0.5920	10.5660	78.74	13.80	17.53

. The three-dimensional response surfaces further illustrate that higher torrefaction temperatures and longer retention times enhance energy yield, highlighting that optimized thermal conditions are critical for improving the energy density and quality of the resulting pellets. The final quadratic regression equations in terms of coded factors for the key responses are given below:

Yenergy = 71.09 − 18.87A + 7.78B − 28.90C + 4.08AB − 7.95AC + 11.29BC + 12.46A² + 6.82B² − 3.06C²

Similar to mass yield, torrefaction temperature (C) exhibited the highest linear coefficient magnitude, confirming its dominant influence on overall energy recovery.

3.5.4. Effect of Process Variables on Volatile Matter

The optimization of volatile matter shows that the quadratic model is statistically acceptable and captures the main trends of the torrefaction process. The overall model is significant (F = 5.47, p = 0.0178) with R² = 0.8754 and adjusted R² = 0.7153, indicating that most of the variation in volatile matter (mean ≈ 52.7%, C.V. ≈ 20%) is explained by the selected factors. Torrefaction temperature (C) exhibited the highest F-value among the linear terms, confirming its dominant influence is the dominant term. It had a negative coefficient (−1.28) and a highly significant effect (F = 28.46, p = 0.0001), while retention time (A) also reduces volatile matter and is statistically significant (F = 5.38, p = 0.0355) as shown in

Table 9. Analysis of variance (ANOVA) for the quadratic model of volatile matter (VM).

Factor			Coefficient	F-Value	p-Value and Remarks
A—Torrefaction Retention Time			−0.134801	5.38	0.0355	Significant
B—Ratio (TWS:TCW)			−4.94670	2.01	0.1995	–
C—Torrefaction Temperature			−1.27532	28.46	0.0001	Highly significant
A × B Interaction			+0.000820	0.0139	0.9161	–
A × C Interaction			−0.004977	0.3853	0.5534	–
B × C Interaction			+0.017025	2.04	0.1620	–
A2—Quadratic (Torrefaction Retention Time)			−0.007614	1.53	0.2486	–
B2—Quadratic (Ratio TWS:TCW)			+0.0112849	0.5376	0.4872	–
C2—Quadratic (Torrefaction Temperature)			−0.000045	0.0001	0.9977	–
Intercept (Constant)			+403.35027	–	–	Constant term
Model (Overall)			–	5.47	0.0178	Model significant
Fit Statistics	R2	Adjusted R2	Predicted R2	Adeq Precision	Mean	Std. Dev.	C.V. %
	0.8754	0.7153	0.2636	6.7781	52.66	10.79	20.49

. The TWS:TCW ratio (B) and all interaction and quadratic terms remain non-significant, indicating only minor contributions. The normal plot of residuals shows points close to the straight line, and the predicted-versus-actual plot follows the 1:1 line, confirming reasonable agreement between model and experiments. Response surfaces as shown in Figure 9 further illustrate that volatile matter decreases with increasing temperature and, to a lesser extent, with longer residence time, especially for cotton-rich blends, which is consistent with progressive devolatilization and upgrading of the pellets during torrefaction.

3.5.5. Effect of Process Variables on Fixed Carbon

The optimization of fixed carbon shows a weaker model compared with the other responses, but it still provides some useful insight. The quadratic model is statistically significant at the 5% level (overall F ≈ 3.7, p = 0.0399), yet the fit is modest (R² ≈ 0.46, adjusted R² ≈ 0.34) and the C.V. is relatively high, indicating considerable scatter in the data as shown in

Table 10. Analysis of variance (ANOVA) for the quadratic model of fixed carbon (FC).

Factor			Coefficient	F-Value	p-Value and Remarks
A—Torrefaction Retention Time			+0.033183	0.1812	0.6773	–
B—Ratio (TWS:TCW)			+0.057614	0.2406	0.6320	–
C—Torrefaction Temperature			+0.376117	10.33	0.0068	Significant
Intercept (Constant)			−70.93919	–	–	–
Model (Overall)			–	3.70	0.0399	Model Significant
Fit Statistics	R2	Adjusted R2	Predicted R2	Adeq Precision	Mean	Std. Dev.	C.V. %
	0.4608	0.3364	0.0842	4.9077	16.54	7.71	46.60

. Among the factors, only torrefaction temperature has a significant positive effect on fixed carbon, while retention time and the TWS:TCW ratio remain statistically non-significant. The normal residual plot shows an approximately linear trend, which supports the basic assumptions of the regression, but the predicted-versus-actual plot displays a wide spread of points around the 1:1 line, consistent with the moderate R². The response surface is almost flat in the time–ratio plane and increases mainly along the temperature axis, confirming that, within the studied range, fixed carbon is controlled primarily by temperature and is only weakly sensitive to residence time or blend composition as shown in Figure 10.

3.5.6. Effect of Process Variables on Ash Content

The optimization results for ash content show that the quadratic model is statistically acceptable and captures the main trend in the data (overall F = 5.34, p = 0.0103; R² ≈ 0.76) as shown in

Table 11. Analysis of variance (ANOVA) for the quadratic model of ash content.

Factor			Coefficient	F-Value	p-Value and Remarks
A—Torrefaction Retention Time			−2.48311	4.30	0.0650	Marginally significant
B—Ratio (TWS:TCW)			+2.99637	1.54	0.2427	–
C—Torrefaction Temperature			−0.22055	13.21	0.0046	Highly Significant
A × B Interaction			−0.003981	0.39	0.5446	–
A × C Interaction			+0.012899	3.80	0.0798	–
B × C Interaction			−0.011692	1.63	0.2306	–
Intercept (Constant)			−51.27130	–	–	–
Model (Overall)			–	5.34	0.0103	Model Significant
Fit Statistics	R2	Adjusted R2	Predicted R2	Adeq Precision	Mean	Std. Dev.	C.V. %
	0.7620	0.6192	0.1479	7.8689	26.55	9.19	34.62

. Torrefaction temperature is the key factor: it has a negative coefficient and is highly significant, so ash content decreases as temperature increases within the studied range. Retention time also has a negative effect and is marginally significant, indicating a further reduction in ash at longer residence times, although with weaker influence than temperature. The TWS:TCW ratio has a positive but non-significant coefficient, consistent with higher sludge fractions tending to raise ash, but composition alone does not dominate the response. Interaction terms contribute little. The relatively high C.V. reflects some experimental scatter, which is also visible as a few points deviating from the 1:1 line in the predicted-versus-actual plot, although the residuals are mostly aligned along the normal probability line as shown in Figure 11. The response surfaces confirm these findings: ash decreases along the temperature and time axes and increases slightly with higher TWS proportion, indicating that higher torrefaction severity applied to cotton-rich blends is favorable for lowering the ash burden of the pellets.

3.5.7. Multi-Objective Numerical Optimization

To achieve a balanced improvement in fuel properties while maintaining practical solid yield, multi-objective numerical optimization was performed using the desirability function approach in Design-Expert^®. The optimization criteria were defined to maximize gross calorific value (GCV), energy yield, and fixed carbon; minimize ash and volatile matter; and maintain mass yield within the studied experimental range. The independent variables (torrefaction retention time, TWS:TCW ratio, and torrefaction temperature) were constrained within their respective experimental limits.

A total of 56 feasible solutions satisfying the specified constraints were generated. The optimal solution (Solution 1), corresponding to the highest composite desirability value (0.701), was selected as the best compromise among the competing objectives. The predicted optimal process conditions were 30 min torrefaction retention time, TWS:TCW ratio of 60, and torrefaction temperature of 240 °C. The corresponding predicted responses are summarized in

Table 12. Optimal process parameters and predicted responses obtained using multi-objective desirability function.

Parameter	Optimal Value
Torrefaction retention time (min)	30.000
TWS:TCW Ratio	60.000
Torrefaction temperature (°C)	240.000
Mass Yield (%)	85.712
Gross Calorific Value (kcal kg−1)	3518.176
Energy Yield (%)	100.211
Volatile Matter (%)	52.153
Fixed Carbon (%)	23.781
Ash (%)	24.290
Composite Desirability	0.701

. Under these optimal conditions, the model predicted a mass yield of 85.712%, GCV of 3518.176 kcal kg⁻¹, energy yield of 100.211%, volatile matter of 52.153%, fixed carbon of 23.781%, and ash content of 24.290%. The composite desirability value indicates a satisfactory overall balance between energy densification and solid mass preservation.

The individual and combined desirability values for the selected optimal solution are illustrated in Figure 12, demonstrating the contribution of each response to the overall optimization outcome. The results confirm that moderate retention time combined with higher torrefaction temperature and increased cotton waste proportion provides improved fuel characteristics without excessive degradation of pellet yield.

4. Discussion

This study shows that blending textile cotton waste (TCW) with textile wastewater sludge (TWS) and a small starch addition shifts fuel quality in predictable ways grounded in current evidence. Cotton-rich fractions carry higher volatile matter and fewer ash-forming minerals, which aligns with thermogravimetric work on cotton and mixed textiles showing a sharp devolatilization band and limited residue, whereas sludge exhibits multi-stage weight loss and large residues due to bound oxygen and mineral matter [31]. In consequence, the observation that increasing TWS depresses heating value and raises ash is consistent with co-combustion and emissions studies reporting the ash- and nitrogen–sulfur-rich character of sludge streams [31,38]. The near-linear decline in energy content with higher TWS recorded in this study is therefore expected because the combustible carbon share falls and the mineral fraction rises, trends widely reported for sludge-containing fuels [39,40]. In this study’s TGA/DTG are in line with the broader textile–sludge literature. Cotton-based materials typically show dominant mid-temperature peaks linked to cellulose degradation and a narrow combustion window; TWS adds high-temperature shoulders and larger residues because of inorganic load and more oxygenated organics [41,42]. The progressive weakening of the main DTG peak and growth of the high-temperature tail as TWS rises mirrors evidence that sludge-rich fuels burn slower and leave more residue, necessitating longer residence or higher setpoints to reach complete burnout [38,43]. In practical terms, the narrower temperature window and lower residues observed for cotton-rich pellets in this study translate into simpler control of ignition and burnout, which echoes recent combustion optimization results for co-firing low-ash biomass with ash-bearing wastes [39,44]. The mechanical outcomes recorded in this study—especially the strength and ductility gains with 5% starch at a 20:80 TWS:TCW ratio—are aligned with two independent strands of evidence. First, pellet mechanics depend strongly on interparticle bonding, pore closure, and internal architecture; recent large-sample tests link higher durability and controlled failure to adequate binder-enabled bonding and optimized pore networks [45,46]. Second, starch-type binders improve compressive performance and handling when moisture and binder levels are tuned, with post-treatment not eliminating the benefit altogether [18,47]. The observation in this study that pellets become denser as TWS content increases, while peak force and breaking strain decrease, is consistent with the idea that densification alone does not guarantee strong inter-particle bonding. Ash-rich matrices tend to form weaker interfacial connections and stress concentrations, whereas cotton-rich matrices with lower ash content allow starch to gelatinize and bridge pores, increasing the usable strain before failure [46,48]. The improved mechanical strength of sludge-containing pellets is likely due to the combined effect of starch acting as an organic binder and the presence of inorganic minerals in the sludge, which may enhance inter-particle cohesion during compression. We note that microstructural characterization (e.g., SEM) was not performed, and this interpretation is offered as a plausible explanation based on similar mechanisms reported in the literature. These patterns help explain why this study’s TCW-rich, starch-assisted formulation achieved the highest mechanical performance even though true density was not maximal.

From a fuel-quality perspective, the study’s recommendation of a TCW-rich blend with modest starch is reasonable given recent market-wide assessments. Retail pellet surveys and standardized tests emphasize that high energy content and low ash reduce both emissions and fouling while improved mechanical durability limits fines formation during transport and feeding [7,45]. The cotton-rich formulation in this study delivered these attributes simultaneously, which is expected to reduce breakage and dusting in hoppers and screws, an inference supported by reported correlations between microstructure and durability quantified using 3D imaging [46].The torrefaction findings in this study—rising fixed carbon and heating value with severity, coupled with declining mass yield—match what recent optimization studies describe as the core energy–mass trade-off. Response surface and design-of-experiments work shows temperature as the dominant driver of energy densification, with time exerting a secondary but clear effect; composition often modulates the slope but rarely overtakes temperature in explanatory power [49,50]. The indication in this study that 220–240 °C with moderate to longer holds gives the best balance for cotton-rich material is consistent with optimization windows reported for lignocellulosics where moderate severity achieves appreciable energy upgrading without excessive solids loss [32,33]. That the gross calorific value could not be captured by a simple quadratic response surface in this study is also unsurprising; energy response often becomes non-quadratic near transition zones where devolatilization pathways and hemicellulose–cellulose interactions change rapidly, a behavior documented in recent parametric studies [33,49]. With respect to ash behavior under torrefaction, this study’s observation that higher temperature and, to a lesser extent, longer time reduce measured ash for cotton-rich pellets accords with work showing volatilization of some inorganics and redistribution of mineral phases during mild-to-moderate heat treatment, though composition remains the primary determinant of absolute ash [30,51]. In sludge-richer blends, however, the same torrefaction schedules yield smaller net benefits because inorganic load dominates, agreeing with reports that mineralogy and ash chemistry inherited from wastewater treatment set a floor under achievable ash reduction [40,52]. This study’s finding that ash continues to track the sludge share even after torrefaction emphasizes that process windows should be tuned to the cotton fraction if the aim is to ease ash handling and minimize slagging propensity. In particular, the ash content of Sample-1 (10.3%) and Sample-2 (13.2%) is substantially higher than EN ISO 17225 limits for graded wood pellets (<3 wt.% for most classes), indicating that these pellets are not intended for premium wood-pellet markets.

The emissions-relevance of the nitrogen and sulfur trends seen in this study is supported by recent co-combustion literature. Studies on sludge co-firing and oxygen-enriched combustion show that higher sludge fractions typically shift NO_x and SO₂ upward unless countermeasures are used, reflecting the larger inorganic-nitrogen and sulfur pools in sludge and the catalytic roles of mineral species [44,53]. Conversely, introducing low-nitrogen, low-sulfur biomass streams can dilute precursor pools and change reaction environments, with some studies reporting reductions in acid gas emissions at modest biomass shares [38,39]. These findings substantiate this study’s inference that cotton-rich pellets at appropriate torrefaction settings would simplify compliance with emission constraints relative to sludge-rich formulations. On the processing and plant-operation side, the density–strength divergence documented in this study has pragmatic implications for storage and feeding. Evidence from mechanical characterization and imaging studies indicates that optimizing internal architecture and binder chemistry matters more than maximizing bulk or true density for maintaining integrity through shocks and shear in conveyors and screws [45,46]. This study’s force–crush trajectories for the cotton-rich, binder-assisted pellets echo those conclusions, suggesting fewer line stoppages from fines generation and smoother mass flow under industrial conditions compared to more mineral-laden, denser pellets whose interfaces break sooner. The broader value proposition—converting two waste streams into a tractable, solid fuel—fits with recent pathways for textile and sludge valorization. Cotton-heavy textile wastes are particularly amenable to thermochemical upgrading with predictable TGA features and manageable ash, while sludge benefits more from targeted treatments and careful co-formulation [41,54]. Where sludge must be incorporated, torrefaction windows similar to those explored in this study can be used to improve handling and partial energy densification, with complementary strategies such as mineral management or co-firing adopted to meet emissions and slagging limits [30,55]. The superiority of a TCW-rich blend with modest starch for energy content, ash management, and mechanical reliability, and the existence of a moderate-severity torrefaction window that strengthens those advantages.

Industrial fluidized bed combustion (FBC) boilers, widely used in textile factories for low-pressure steam, can handle fuels with ash contents up to 62 wt.% [56]. However, high ash content may still be acceptable for industrial applications, where combustion systems are more robust than domestic ones [56]. The TCW-rich pellets, particularly Sample-1 and Sample-2 with ash below 15 wt.% and high volatile matter, ignite easily and burn efficiently, making them well-suited for modern biomass boilers with standard ash-handling systems. Sulfur and nitrogen contents in these cotton-rich pellets are low (≤0.37 wt.% S, ≤1.35 wt.% N), minimizing the risk of corrosion from acidic flue gases. In contrast, sludge-heavy pellets (Sample-3 and Sample-4) have ash contents above 20 wt.%, lower calorific values, and higher oxygen, leading to longer burnout times and increased ash handling, making them challenging for standard industrial boilers. Other factors, such as moisture content, pellet durability, and potential slagging from mineral components in the sludge, should also be considered when scaling up for industrial applications.

5. Conclusions and Recommendations

This study demonstrates that co-pelletization of textile wastewater sludge (TWS) and cotton waste (TCW) can produce stable biofuel pellets with practical mechanical and thermal properties. The 20:80 TWS:TCW blend with 5% starch (Sample-1) exhibited the best performance, achieving high pellet conversion (96%), true density (1159 kg m⁻³), and compressive strength (14.9 N/mm²), along with favorable fuel characteristics, including a gross calorific value of 3377 kcal kg⁻¹, moderate volatile matter (73.1%), and controlled ash content (10.3%). Sample-2 (40:60 TWS:TCW) also showed acceptable performance, with a pellet conversion rate of 81%, true density of 1209 kg m⁻³, compressive strength of 12.1 N/mm², and a calorific value of 2968 kcal kg⁻¹, making it suitable for industrial applications where slightly lower energy density is acceptable.

Mild torrefaction of Sample-1 at 200–220 °C further improved energy density, with gross calorific values of 3563–4083 kcal kg⁻¹ and energy yields of 98–111%, while severe torrefaction at 240 °C caused mass and energy losses and increased ash. Sludge-rich blends at 240 °C showed higher calorific values due to increased carbon content, though mechanical properties and mass yield were compromised.

Overall, feedstock ratio, starch addition, and torrefaction conditions critically influence pellet quality. The 20:80 TWS:TCW blend with starch is the optimal choice, while the 40:60 blend is also acceptable for industrial combustion, providing a balance between mechanical durability, pelletization efficiency, and thermal energy output.

This study addresses a clear industrial and environmental need: textile manufacturing produces two difficult wastes—cotton-rich production offcuts and wastewater-treatment sludge—that are costly to manage and carbon-intensive to dispose-off. Whether these streams can be co-converted into a consistent, specification-ready solid biofuel using equipment and skills already common in textile plants and allied energy facilities. The approach favored practicality: pelletization with a simple, food-grade binder; mild thermal conditioning to stabilize fuel quality; and routine characterization that any regional laboratory can replicate. The rationale was threefold: diverting waste from landfilling or uncontrolled burning to lower fees and support compliance; turning residues into a usable energy carrier to strengthen on-site energy security and hedge against fuel price volatility; and aligning fuel preparation with circular-economy goals without demanding boiler overhauls or exotic reagents. The research was designed to yield operational rules of thumb—how to select and condition feedstock, press pellets, and apply light thermal treatment—so textiles can standardize fuel preparation and downstream handling while meeting quality, safety, and air-quality expectations. Co-pelletization of cotton-leaning residues with a modest, food-grade binder should be prioritized, paired with disciplined moisture control before pressing and simple in-house quality checks (proximate analysis, periodic ultimate analysis, and mechanical durability). Mild torrefaction is best used tactically to meet seasonal or site-specific targets, with clear setpoints and batch records to track stability. Combustion readiness requires attention to air staging and residence time, routine ash monitoring, and good housekeeping to limit fines along conveyors and feeders. Medium-term work should validate storage behavior across seasons, confirm feeding stability on existing boiler lines, and document emissions under representative loads to codify specifications and operating envelopes. A key limitation is the elevated ash content associated with the sludge fraction, which may require ash management and slagging/fouling considerations; therefore, the intended niche is industrial/co-firing applications rather than premium low-ash pellet markets. Strategic next steps include nitrogen and phosphorus concentration in ash for land application, heavy metal ash analysis, exhaust gas emission characteristic, life-cycle and cost benefits assessments (CBA), and alignment with fuel standards. Simple digital dashboards for moisture and density, coupled with periodic data reviews, can tighten process windows without added cost. Taken together, these measures move the pathway from bench to routine practice—turning two liabilities into a dependable, cleaner solid fuel that reduces disposal burdens, stabilizes energy costs, and meets evolving regulatory expectations—while converting hazardous textile wastewater sludge (TWS) and cotton waste (TCW) into a useful renewable energy resource that advances resource recovery and minimizes waste.