Optimizing circular industrial integration for sustainable yarn production with remanufacturing: a carbon policy-based modelling approach Authors: Vishnupriya Kalathil Sahadevan, Abin Thomas, Umakanta Mishra Published: 2026 ABSTRACT AbstractThe rapid expansion of the textile industry presents critical challenges in achieving sustainable consumption and production. In response, this study develops an integrated sustainable production model for the yarn sector. The proposed model incorporates investments in green technology and wastewater purification systems, assessing their influence on both economic performance and environmental sustainability. The average integrated total profit of the system is optimized, and the model’s practicability is demonstrated through a numerical illustration. The findings reveal that upcycling old garments enhances overall profitability by an average of 5.13% under both carbon pricing policies. Furthermore, while the CCT mechanism reduces green investment costs by 3.18%, the CT policy achieves a 2.29% reduction, highlighting the superior emission mitigation efficiency of the CCT approach. The suggested model advances sustainable production-consumption in the yarn industry by repurposing banana waste for yarn, and adopting the extended producer responsibility (EPR) directive, all while maintaining system profitability.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-54680-0.Keywords: Bio-yarn ecosystem, Wastewater treatment, Remanufacturing, Circular interconnectedness, Carbon policiesSubject terms: Engineering, Environmental sciences, Environmental social sciences, Mathematics and computing Introduction The adoption of the Paris Agreement in 2016 marked a turning point in global sustainability efforts, encouraging industries to advance the shift toward sustainable production and consumption (SPC) practices. Among these industries, the textile and fashion sector has experienced rapid growth, recording a 36% increase in economic profits in 2022 alone1. However, despite its economic gains, the textile sector remains a major environmental contributor, significantly impacting global greenhouse gas emissions and contributing to approximately 20% of global wastewater pollution2. A detailed breakdown of the per capita environmental and socioeconomic impact of the global textile and footwear industry across 189 countries in 2015 further highlights the severity of the issue (see Fig. 1). Global apparel production now exceeds 100 billion items each year, making the fashion sector a major environmental contributor. It generates roughly 10% of worldwide greenhouse gas emissions around 1.2 billion tonnes of CO₂ equivalent and is projected to absorb nearly 25% of the global carbon budget by 2050 if current trends remain unchanged3. This alarming environmental footprint underscores the urgent need to adopt circular economy (CE) principles that emphasize resource efficiency, waste valorization, and closed loop operations. Although the CE has gained significant attention in the textile and fashion sectors, its widespread implementation across businesses and supply chains remains limited4. The transition is impeded by critical challenges, including material limitations, resource-intensive processes, dependence on energy and water, and the significant costs associated with sustainability investments. Within this context, the yarn industry being the backbone of textile manufacturing emerges as a critical node due to its significant contribution to carbon emissions and wastewater generation. Therefore, the development of sustainable yarn production models can serve as a strategic entry point for industry wide transformation. Global per capita footprints of 6 metrics across 189 countries from textile and footwear sectors in 20153,5. To mitigate these environmental challenges, researchers have proposed bio-based alternatives and advanced waste treatment methods6. The growing interest in bio-based fibers has directed attention toward agricultural residues as potential raw materials for sustainable textile production. Among these, banana biomass represents a particularly promising yet underutilized resource. Large volumes of banana pseudo-stems are generated after each harvest cycle, most of which are discarded or left to decompose, creating disposal and environmental management challenges. Converting this lignocellulosic waste into usable fiber not only reduces the environmental burden associated with agricultural residues but also creates an additional value stream for farming systems. In contrast to several conventional and synthetic fibers, banana fiber is renewable, biodegradable, and requires relatively low chemical processing, making it suitable for environmentally conscious textile applications. These characteristics position banana fiber as a strategically relevant material for developing circular production systems that integrate waste valorization with sustainable manufacturing. Prior studies, including7 highlighted banana fiber’s potential to revolutionize the traditional textile industry into a sustainable bio-textile sector. While its technical feasibility has been discussed in literature, there is a lack of quantitative production-inventory models that integrate circular strategies and evaluate the implications of environmental regulations. Moreover, current sustainable inventory literature rarely considers the combined effects of remanufacturing, wastewater treatment, green technology investment, and quality-dependent demand under carbon policies. Addressing this research gap, this study develops an integrated circular production inventory model for sustainable yarn manufacturing, utilizing lignocellulosic waste derived from post-harvest banana pseudo-stems as the primary input for yarn production. The model captures circular industrial symbiosis between the banana and yarn industries by valorizing agro-waste and embedding end-to-end sustainability features. It integrates multiple dimensions of circularity, including green technology investments to reduce emissions, wastewater treatment, inspection to ensure product quality, and end-of-life processing, framed as an extended producer responsibility (EPR). These integrated strategies aim to promote resource efficiency, minimize environmental impacts, and enhance the overall sustainability of the production-consumption cycle. The key objectives of the proposed model are shown in Fig. 2. A circular framework for the bio-yarn industry (Figure source: Authors’ own illustration). Market demand is modelled as a function of yarn tensile strength to reflect quality enhancement arising from circular processing. The system is evaluated under four distinct operational scenarios representing carbon tax and carbon cap-and-trade policies, with and without remanufacturing, enabling a comprehensive comparison of economic and environmental outcomes. The major contributions of this study are summarized as follows: Development of an integrated two-layer circular production inventory model for the yarn industry that explicitly links agricultural waste valorization with imperfect manufacturing. Development of an integrated two-layer circular production inventory model for the yarn industry that explicitly links agricultural waste valorization with imperfect manufacturing. Modelling of quality-dependent demand through tensile strength to capture market response to circular production. Modelling of quality-dependent demand through tensile strength to capture market response to circular production. The economic and environmental implications of carbon tax and carbon cap-and-trade policies are comparatively examined under remanufacturing and non-remanufacturing operational settings. The economic and environmental implications of carbon tax and carbon cap-and-trade policies are comparatively examined under remanufacturing and non-remanufacturing operational settings. Development of a single analytical framework that simultaneously captures remanufacturing operations, wastewater treatment, inspection-driven quality control, and investments aimed at emission reduction. Development of a single analytical framework that simultaneously captures remanufacturing operations, wastewater treatment, inspection-driven quality control, and investments aimed at emission reduction. Numerical experiments and sensitivity analyses are conducted to quantify profit–emission trade-offs and to derive operational and policy-oriented insights for sustainable textile supply chains. Numerical experiments and sensitivity analyses are conducted to quantify profit–emission trade-offs and to derive operational and policy-oriented insights for sustainable textile supply chains. In contrast to prior green inventory studies in the textile sector, the proposed model demonstrates how the inclusion of remanufacturing reshapes the comparative performance of carbon tax and cap-and-trade policies when demand responds to product quality and wastewater treatment is explicitly modelled. The findings offer practical implications for textile manufacturers seeking sustainable alternatives, for environmental policymakers designing effective carbon regulation frameworks, and for practitioners aiming to implement circular economy principles across industrial supply chains. Literature survey Dependent demand The modern business environment is marked by heightened competition, where customer preferences significantly impact demand patterns. In8, Feng et al. proposed an EOQ model with a multivariate demand function that explicitly incorporates freshness and stock display, offering a more realistic representation of consumer behaviour for perishable goods such as fruits, vegetables, and baked items. To enhance realism in inventory modelling, Pervin et al.9 developed a time-dependent demand methodology that addresses the unique challenges posed by deteriorating items. Ruidas et al.10 developed an imperfect production model under different carbon policies, positing a linear demand-price inverse relationship where demand declines as the selling price rises. Pando et al.11 extended the analysis by developing an inventory model that incorporates demand rate as a function of both sales price and inventory level. Khan et al.12 proposed a demand model that multiplicatively integrates logit price sensitivity and power-function time dynamics, complemented by an advanced payment approach. Recognizing the profound impact of time on market demand for perishable goods,13 designed a sustainable model exhibiting power-law demand dynamics. Although demand dependent inventory models have evolved significantly, most studies focus on pricing, deterioration, or payment mechanisms without incorporating carbon regulatory constraints or remanufacturing integration within a circular production framework. Carbon regulations Establishing carbon emissions limits and regulatory frameworks, incorporating taxation, trading, and offset schemes, constitutes a key governance strategy for promoting sustainable environmental development. Toptal et al.14 conducted a study analysing a retailer’s inventory and carbon reduction investment decisions under three carbon regulation policies cap, tax, and cap and trade by extending the EOQ model, demonstrating that green investments can not only ensure compliance but also lower both emissions and costs through analytical and numerical comparisons. Huang et al.15 evaluate the effectiveness of three carbon policies and discover that firms prefer carbon taxation when accompanied by efficient GT adoption. Within the context of time-varying demand and carbon taxation regulations,16 developed a mathematical model to investigate the impact of carbon emission taxes on system-wide optimal costs16. Ghosh17 presents a comparative analysis of CT and CCT policies in a vendor–buyer supply chain, highlighting the benefits of a hybrid approach in achieving cost savings and emissions reductions. Hasan et al.18 explored the effects of CT, CCT, and strict carbon limits on emission mitigation and profit maximization, concluding that CT policy has a statistically significant influence on total profit19. Manna et al.20 developed an analytical framework to examine the effects of CT and CCT policies on production planning for perishable items, highlighting the critical role of carbon taxation in reducing emissions and enhancing profitability. Sebatjane21 explored the intersection of economic and environmental objectives, assess the performance of CT, CCT, and cap-and-offset policies in an integrated production model, and demonstrate the potential of these policies to reduce carbon emissions while maintaining profitability. Furthermore, a multi-stage examination of carbon emissions from production to refurbishing is undertaken to assess the effectiveness of CT and CCT policies. Findings suggest that CCT policies provide a more favourable economic and environmental trade-off22. While several studies compare CT and CCT policies, these analyses are typically confined to conventional production systems and rarely consider circular remanufacturing dynamics, particularly within dependent demand environments. Remanufacturing strategies Öztürk23 formulated a mathematical model to investigate production planning under imperfect production conditions, incorporating the rework of a specified fraction of defective items at the end of the production process. Ullah et al.24 present a stochastic three-echelon closed-loop model to assess remanufacturing strategies. Sarkar & Bhuniya.25 introduced a closed-loop production model, integrating remanufacturing and virgin material sourcing, to examine the economic and environmental benefits of the circular production systems. Sahadevan & Mishra 26 suggested that a CE model integrating rework, recycling, and reuse boosts profitability while minimizing waste and environmental impact. Salas-Navarro et al.27 presented a comprehensive three-layer framework to analyse the impact of remanufacturing on replenishment strategies, production processes, and profit performance within an imperfect production environment. Existing remanufacturing models predominantly address production imperfections or closed-loop design but seldom evaluate performance under alternative carbon regulatory regimes, limiting policy-oriented insights. Existing remanufacturing models predominantly address production imperfections or closed-loop design but seldom evaluate performance under alternative carbon regulatory regimes, limiting policy-oriented insights. Sustainable yarn manufacturing The yarn manufacturing process is characterized by high energy consumption and significant waste generation, underscoring the necessity for sustainable practices. Nayak et al.28 explored the integration of lean methodology with the 3R framework to promote circular business models, specifically textile reuse, exchange, and rental, for effective upcycling of textile waste. Faysal et al.29 investigated the potential for closed-loop production in the apparel industry by developing a novel process to convert leftover fabric into recycled yarn, thereby achieving zero waste. Exploring food waste’s textile applications,30 underscores the importance of interdisciplinary approaches across industries for establishing a circular economy30. Furthermore, Provin et al.7 examined the potential of banana fibers, derived from agricultural waste, as an eco-friendly substitute in textile production, contributing to a more sustainable bio-textile sector. A. Khan et al.31 investigated the development of bio-yarns by blending banana fiber with cotton and regenerated cellulosic fibers, exploring varying proportions to achieve optimal sustainability. Despite increasing attention to bio-based and circular textile production, quantitative carbon regulated inventory optimization models tailored to sustainable yarn manufacturing remain scarce. Synthesis of literature and research gap The existing body of literature demonstrates significant advancements in individual domains of sustainable inventory and production modelling. Table 1, provides a comprehensive overview of the conceptual connections between our study and the preceding research. Studies on demand-dependent inventory systems have incorporated pricing strategies, deterioration effects, and time-sensitive demand dynamics, thereby improving the realism of market behaviour representation. However, these models rarely consider the influence of environmental regulations or the integration of circular practices such as remanufacturing within the demand structure. While, research on carbon regulatory frameworks, including carbon tax and cap-and-trade policies, has provided valuable insights into emission control and cost optimization. Nevertheless, these studies are predominantly confined to conventional production systems and often overlook the operational complexities introduced by circular economy strategies, particularly in multi-stage production environments. In parallel, remanufacturing and closed-loop supply chain models have been widely explored to enhance resource efficiency and reduce waste. While these models capture production imperfections and recovery processes, they typically do not evaluate the impact of alternative carbon regulatory policies, thereby limiting their applicability for policy-driven decision making. Furthermore, recent studies on sustainable yarn manufacturing and bio-based fibers emphasize the environmental benefits of alternative raw materials and circular textile practices. Despite this progress, there remains a lack of quantitative analytical models that integrate bio-based material utilization with carbon regulation, remanufacturing, and demand-dependent behaviour within a unified framework. Table 1The suggested model’s comparison with the current body of work.ReferencesNLDDGTRMCPEmission sourcesBYSCMCTCCT152✕P✕PPProduction,transportation,and storage✕23NA✕✕P✕✕✕102Selling price✕✕PP✕✕172✕✕✕PPProduction and storage✕243✕PP✕✕✕✕112Selling price and stock quantity✕✕✕✕✕✕31✕Selling price✕✕✕✕✕✕202Selling price, warranty period✕✕P✕Production✕273Run-time production✕P✕✕✕✕31✕✕✕✕✕✕✕P223✕✕PPPProduction,storage✕This paper2Tensile strengthPPPPProduction,transportation,remanufacturing and storagePNBCT: Carbon tax; RM: Remanufacturing; P: Available; DD: Dependent Demand; ✕: Not available; CP: Carbon policy; NL: Number of layers; DD: Dependent Demand; PR: Parameters; NA: Not Applicable; CCT: Carbon cap and trade; BYSCM: Banana yarn supply chain management. The suggested model’s comparison with the current body of work. Production, transportation, and storage Production, storage Production, transportation, remanufacturing and storage CT: Carbon tax; RM: Remanufacturing; P: Available; DD: Dependent Demand; ✕: Not available; CP: Carbon policy; NL: Number of layers; DD: Dependent Demand; PR: Parameters; NA: Not Applicable; CCT: Carbon cap and trade; BYSCM: Banana yarn supply chain management. Therefore, a critical research gap exists at the intersection of these domains. To the best of the authors’ knowledge, no prior study has developed an integrated production inventory model that simultaneously incorporates demand dependent behaviour, remanufacturing processes, green investment strategies, wastewater treatment, and carbon regulatory mechanisms within the context of sustainable yarn manufacturing. Motivated by this gap, the present study proposes a two-layer circular production inventory model that captures the interaction between agricultural waste valorization and yarn production under carbon tax and cap-and-trade policies. The proposed framework captures the interaction between agricultural waste utilization and yarn production, offering a more comprehensive perspective on the economic and environmental performance of sustainable industrial systems. Proposed industrial model A circular sustainable production model for the yarn industry consisting of a manufacturer and a retailer is proposed to align with the UN’s sustainable development goals. The manufacturer establishes a supply chain partnership with farmers to source banana pseudo stem, which is then processed into high-quality yarns for textile applications. The production process commences with the harvesting of banana pseudo stems, followed by sheath separation. The extracted sheaths undergo water retting, involving immersion in water to facilitate fiber separation. Subsequent processing stages include water removal, fiber separation, washing, and drying. The dried fibers are then spun into yarn, which is supplied to retailers for the fashion industry. The manufacturer has invested in GT and WWT to align with sustainability benchmarks, thereby reducing environmental impact. Utilizing agro-industrial waste from banana production for yarn manufacturing creates a sustainable inter-industrial ecosystem between the banana and yarn industries. Coupled with the manufacturer’s extended producer responsibility directive for upcycling old fabric, it offers a new perspective for the yarn sector to achieve an SPC system. The proposed sustainable banana yarn supply chain system is depicted in Fig. 3. Circular sustainable supply chain for the banana yarn industry (Figure source: Authors’ own illustration). Nomenclature and assumptions This section provides an overview of the notations and key assumptions that underpin the development of the proposed model, establishing the foundational framework for the subsequent analysis. Nomenclature Mathematical notations Regular demand rate (units/year) η Collection rate of used clothes (units/cycle) Yarn production rate (units/year) β Fraction of recyclable clothes (units/cycle) M Number of shipments Purchasing cost at the manufacturer ($/unit) Y Ordering quantity (units/year) CO2 emitted during yarn production (ton/unit) Holding cost at yarn manufacturer ($/unit/year) CO2 emitted during warehousing at the manufacturer (ton/unit/year) Holding cost at yarn retailer ($/unit/year) Emission cost for warehousing unit item at the retailer (ton/unit/year) Manufacturer’s setup cost ($/setup) Tax imposed on unit carbon emitted ($/ton) Retailer’s ordering cost ($/order) Buying/selling price per unit carbon ($/ton) σ Efficiency of WWT Carbon allowance (ton/year) δ Conversion factor ϕ Fraction of carbon emitted after GI (ton) x Transportation cost ($/km) ψ Efficiency of GT implemented q Fuel price ($/litre) g Capital investment for GT ($/unit item) θ Fuel consumption rate Regular rate of wastewater discharge Distance between the manufacturer and the retailer (km) CO2 emitted during the recycling process (ton/unit) Indirect emission (ton) μ Capital investment for WWT ($) Direct emission (ton) Cost incurred on wastewater discharge ($) w Weight of shipped item Selling price of yarn at the manufacturer ($) management Abbreviations CECircular Economy AITP Average integrated total profitPRParametersSPCSustainable production consumptionRMRemanufacturingEPRExtended producer responsibility Assumptions The mathematical model is built under the following assumptions: An integrated manufacturer-retailer model for the banana yarn industry focusing on emission reduction, waste management, profit maximization, and remanufacturing under CT and CCT policies is proposed in this study. The yarn manufacturer initiates the collection of used clothes from the end users and remanufactures them into yarn to establish an extended producer responsibility. An integrated manufacturer-retailer model for the banana yarn industry focusing on emission reduction, waste management, profit maximization, and remanufacturing under CT and CCT policies is proposed in this study. The yarn manufacturer initiates the collection of used clothes from the end users and remanufactures them into yarn to establish an extended producer responsibility. The AITP of the system is maximized with and without remanufacturing under CT and CCT policy. The AITP of the system is maximized with and without remanufacturing under CT and CCT policy. The demand rate D is considered as a non-linear function of the tensile strength of the yarn. The non-linear demand is , where ρ denotes the tensile strength of yarn with consumer sensitivity . Also, the selling price of yarn depends on the market’s sensitivity towards the tensile strength, and the selling price is, , where denotes the consumer’s sensitivity towards the tensile strength and denotes the regular selling price of banana yarn. In the proposed model, market demand is assumed to increase with yarn tensile strength, reflecting the positive influence of product quality on buyer preference. An exponential functional form is initially adopted to capture this monotonic relationship. However, to ensure analytical tractability and to avoid unbounded demand growth, the exponential term is approximated using a Taylor series expansion, and only the first two terms are retained. This truncation yields a bounded and smooth demand representation within the feasible operating range of tensile strength considered in the model. Consequently, the adopted formulation captures the incremental effect of quality improvement on demand without implying unlimited market expansion. While alternative forms such as logistic or price quality coupled demand functions may capture saturation effects more explicitly, the present approximation provides a reasonable and controlled representation for analytical optimization and is treated as a modelling assumption. The demand rate D is considered as a non-linear function of the tensile strength of the yarn. The non-linear demand is , where ρ denotes the tensile strength of yarn with consumer sensitivity . Also, the selling price of yarn depends on the market’s sensitivity towards the tensile strength, and the selling price is, , where denotes the consumer’s sensitivity towards the tensile strength and denotes the regular selling price of banana yarn. In the proposed model, market demand is assumed to increase with yarn tensile strength, reflecting the positive influence of product quality on buyer preference. An exponential functional form is initially adopted to capture this monotonic relationship. However, to ensure analytical tractability and to avoid unbounded demand growth, the exponential term is approximated using a Taylor series expansion, and only the first two terms are retained. This truncation yields a bounded and smooth demand representation within the feasible operating range of tensile strength considered in the model. Consequently, the adopted formulation captures the incremental effect of quality improvement on demand without implying unlimited market expansion. While alternative forms such as logistic or price quality coupled demand functions may capture saturation effects more explicitly, the present approximation provides a reasonable and controlled representation for analytical optimization and is treated as a modelling assumption. The banana yarn industry generates wastewater with a high organic load, necessitating efficient treatment methods for both environmental sustainability and resource recovery. Flocculation serves as a multifunctional technique for simultaneous wastewater purification, carbon capture, and hydrogen production6. The manufacturer invested capital μ for treating wastewater, which is discharged initially at a rate . The reduction volume in discharge after the investment is , where σ is the efficiency of technology and the newly reduced discharge rate of wastewater is, 32. The working mechanism of the flocculation method is shown in Fig. 8 in Appendix 1 of supplementary file. The banana yarn industry generates wastewater with a high organic load, necessitating efficient treatment methods for both environmental sustainability and resource recovery. Flocculation serves as a multifunctional technique for simultaneous wastewater purification, carbon capture, and hydrogen production6. The manufacturer invested capital μ for treating wastewater, which is discharged initially at a rate . The reduction volume in discharge after the investment is , where σ is the efficiency of technology and the newly reduced discharge rate of wastewater is, 32. The working mechanism of the flocculation method is shown in Fig. 8 in Appendix 1 of supplementary file. Let g be the capital invested by the banana yarn manufacturer in green technology (GT) to reduce emissions. If ψ denotes the efficiency and ϕ denote the fraction of emissions reduced after investing in technology, then the amount of emissions reduced is . Here H(g) is a continuously differential function with H(0) = 0 and 33. Let g be the capital invested by the banana yarn manufacturer in green technology (GT) to reduce emissions. If ψ denotes the efficiency and ϕ denote the fraction of emissions reduced after investing in technology, then the amount of emissions reduced is . Here H(g) is a continuously differential function with H(0) = 0 and 33. Mathematical model In the manufacturer-retailer yarn supply chain model, the retailer has the annual demand D and the retailer initiate the process by asking the manufacturer to supply MY units of yarn. To respond to the order, the manufacturer produces high-quality banana yarn and dispatches the finished items in M equal-sized shipments. The production takes place at a constant rate of until the demand is met over a production time . During this phase, inventory accumulates at the manufacturer. After the production ends, the inventory at the manufacturer depletes at a constant demand rate D as items are shipped to the retailer. The non-production period , during which only shipments occur, is given by . During , the manufacturer completes deliveries to the retailer in multiple equal-sized batches. The inventory pattern for the yarn manufacturer-retailer is depicted in Fig. 4. Inventory pattern of the yarn manufacturer and the retailer. Manufacturer’s outlook The average inventory for the manufacturer is (see Fig. 4). Holding cost The average holding cost incurred by the manufacturer is, (see Fig. 4). Setup cost The setup cost per unit time at the yarn manufacturer is, . Production cost The production cost at the yarn manufacturer is the aggregate sum of labour, chemical, machinery, and purchasing costs. If L denotes the labour cost for each unit, I denotes the chemical cost for each unit, H denotes the machinery cost for each unit and denotes the purchasing cost for each unit, then the total purchasing cost per cycle without considering remanufacturing is, On considering the remanufacturing process, the reduced production cost per unit time is, where denotes the fraction of recycled items that can be used for the production of yarn. Energy cost Let the amount of electricity required for yarn production be , and δ be the conversion factor, then is the consumed energy. If denotes the cost incurred on per unit energy consumed, then the total energy cost per unit time is, . Transportation cost Let the price of fuel consumed at a rate θ be q. If denotes the distance between the manufacturer and the retailer, then the transportation cost incurred by the manufacturer per unit time is, where x denotes the fixed transportation cost (see Fig. 3). Remanufacturing cost Let η denotes the collection rate of used clothes and β denotes the fraction of recyclable clothes. If the cost incurred on remanufacturing each unit of used cloth is (sum of incentive cost for returned clothes and the recycling cost ), then the total cost incurred by the manufacturer on remanufacturing old clothes per unit of time is, (see Fig. 3). Emission cost The manufacturer produces emissions during the production, warehousing, transportation, and remanufacturing of old clothes into yarn. Here, and denote the carbon emission components associated with production, warehousing, and remanufacturing activities, respectively, measured in tons of CO₂-equivalent per unit of activity. Then, the amount of carbon emitted during the production per unit time is . The amount of carbon emitted during warehousing is . The amount of carbon emitted during transportation per unit time is . The amount of carbon emitted during remanufacturing per unit time is . The total amount of carbon emitted from the yarn manufacturing sector without remanufacturing is, . Total amount of carbon emitted from the yarn manufacturing sector with remanufacturing is, The yarn manufacturer invested capital g in GT to reduce emissions. If ψ denotes the efficiency and ϕ denote the fraction of emissions reduced after investing in technology, then the reduction of emissions from the yarn manufacturing sector without remanufacturing is, and with remanufacturing is,. Carbon tax policy Emission cost incurred by the manufacturer under CT policy without remanufacturing is, . Emission cost incurred by the manufacturer under CT policy with remanufacturing is, . Carbon cap and trade policy Emission cost incurred by the manufacturer under CCT policy with carbon allowance , without considering remanufacturing is, . Emission cost incurred by the manufacturer under CCT policy with carbon allowance , considering remanufacturing is, Wastewater treatment cost Without investment, the cost of managing wastewater is and with investment, the discharge cost of wastewater gets reduced to (see Fig. 8 in Appendix 1 of Supplementary file). Revenue component of yarn manufacturer The yarn manufacturer sells each unit at a price , and the total revenue generated by the manufacturer in unit time is, . Retailer’s outlook The average holding cost at the retailer is . Ordering cost per order per unit time is . Emission cost incurred on holding yarn per unit time is . The total cost at the retailer is, The average integrated total profit (AITP) of the system is defined as the difference between the total revenue and the total cost incurred across all activities. The AITP of the yarn supply chain under CT policy without remanufacturing is, The AITP of the yarn supply chain under CT policy with remanufacturing is, The AITP of the yarn supply chain under CCT policy without remanufacturing is, The AITP of the yarn supply chain under CCT policy with remanufacturing is, The expanded AITP is given in Appendix 2. To find the optimal solution, the exponential terms are approximated using the Taylor series expansion of , and is approximated as . Theorem 1 AITP of the system under CT policy without remanufacturing is concave with respect to the decision variables and for any positive discrete values of ρ. Proof See Appendix 3. Theorem 2 AITP of the system under CT policy with remanufacturing is concave with respect to the decision variables Y, g and for any positive discrete values of ρ. Proof Similar proof as Theorem 1 (See Appendix 3). Theorem 3 AITP of the system under CCT policy without remanufacturing is concave with respect to the decision variables Y, g and for any positive discrete values of ρ. Proof Similar proof as Theorem 1 (See Appendix 3). Theorem 4 AITP of the system under CCT policy with remanufacturing is concave with respect to the decision variables Y, g and for any positive discrete values ρ. Proof Similar proof as Theorem 1 (See Appendix 3). Numerical analysis The sustainable production consumption model designed for the banana yarn industry with extended producer responsibility, green investment and wastewater purification is given in Sect. 3.1 (see Fig. 3). The model’s applicability is substantiated through numerical examples. AITP is calculated for four distinct cases, and the concavity of AITP with respect to the decision variables is verified mathematically and numerically. The parameter values employed to validate the proposed model are approximate estimates based on those found in the existing literature17,31,34 and are provided in Appendix 4. The parameters related to holding, setup, production, ordering, and emission costs are approximate values to the data values reported in Ghosh et al. (2020)17. The demand parameters are similar to those used in Rabta (2020)35. The transportation costs are estimated based on distances and logistics structures according to current market rates and are comparable to the values reported in Mashud et al. (2021)33. The parameter values associated with energy cost and remanufacturing cost are similar to those reported in Khan et al. (2024)31. The wastewater treatment parameter values are similar to those reported in Thomas and Mishra (2023)34. The solution is found using Mathematica 9.0 software with a system of specifications 12th Gen Intel(R) Core (TM) i5–1235U with 1.30 GHz processor. The first-order necessary conditions of the approximated AITP function are solved numerically using Mathematica’s “FindRoot” command, which applies a Newton-type iterative method with symbolically computed Jacobians. This procedure determines stationary points of the nonlinear system corresponding to , and . The search is performed within the feasible region defined by the operational constraints of the model, including non-negativity and bounded economic parameters. Appropriate initial values are selected based on realistic parameter magnitudes to ensure numerical convergence. Since Newton’s method is locally convergent, the procedure yields a locally optimal stationary point. The nature of the stationary point is verified by evaluating the Hessian matrix at the solution and confirming the second-order conditions. The objective of this study is to obtain economically meaningful local optimal solutions rather than to establish global optimality. The optimization algorithm developed to compute the optimal values of decision variables and the AITP of the system for all four cases is given in Fig. 5. The algorithm used in Mathematica 9.0 software. The optimal solution table for the system under CT and CCT policy with and without considering remanufacturing is illustrated in Tables 2 and 3. Table 2The optimal solution table for CT and CCT without remanufacturing.CT without remanufacturingCCT without remanufacturingρYgemiρYgemi0331.3732.506416,254.942522.680331.6372.186416,571.582398.621331.7762.504416,300.212524.111332.0392.183416,617.052400.642332.1792.502416,344.742525.522332.4412.181416,661.762402.063332.5822.500416,388.522526.943332.8442.179416,705.732403.474332.9862.497416,431.562528.954333.2472.176416,748.962405.495333.3912.495416,473.852530.365333.6502.174416,791.442406.89⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ 56354.8352.374417,643.212606.4656355.0362.054417,970.202484.6957355.2722.372417,646.792609.6557355.4732.051417,973.952486.5358355.7102.370417,649.632610.8358355.9092.049417,976.962487.7259356.1492.367417,651.722612.6459356.3472.047417,979.232488.9260356.5892.365417,653.072613.8160356.7852.044417,980.752490.7461357.0292.362417,653.672615.6261357.2242.042417,981.532491.9262357.4702.360417,653.532616.7862357.6642.039417,981.572493.7463357.9112.357417,652.652619.8563358.1042.037417,980.862494.92 The optimal solution table for CT and CCT without remanufacturing. Table 3The optimal solution table for CT and CCT with remanufacturing.CT with remanufacturingCCT with remanufacturingρYgemiρYgemi0331.3872.519436,192.532527.310331.6372.185436,521.582399.221331.7902.516436,257.732529.351332.0392.183436,586.992400.642332.1932.514436,322.182530.782332.4412.181436,651.662402.063332.5962.512436,385.902532.203332.8442.178436,715.582404.084333.0012.510436,448.862533.634333.2472.176436,778.762405.495333.4052.508436,511.072529.035333.6502.174436,841.192406.89⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ ⋮ 84367.3672.318439,074.052653.8984367.5201.984439,419.972525.6285367.8252.315439,076.742655.6085367.9771.982439,422.852526.6986368.2842.313439,078.692656.6686368.4341.979439,424.992528.4187368.7432.310439,079.892658.3787368.8931.977439,426.382529.4888369.2042.308439,080.362659.4188369.3521.974439,427.032531.1989369.6652.305439,080.082661.1189369.8121.972439,426.942532.24 The optimal solution table for CT and CCT with remanufacturing. For all the four cases, the concavity nature of the AITP with respect to the tensile strength and green investment is given in Fig. 6. The concave nature of AITP w.r.t ρ and g under CT without remanufacturing is given in 7(a), under CT with remanufacturing is given in 7(b), under CCT without remanufacturing is given in 7(c), under CCT with remanufacturing is given in 7(d). Comparative analysis A comparative analysis is done to identify the carbon policy that maximizes system profitability while minimizing emissions and to evaluate the efficacy of remanufacturing strategy implementation in achieving the objectives (see Tables 4, 5, 6 and 7). Table 4The optimal values under CT with and without remanufacturing.RemanufacturingρYgAITPemiWithout61357.0292.362417,653.672615.62With88369.2042.308439,080.362659.41 The optimal values under CT with and without remanufacturing. Table 5The optimal values under CCT with and without remanufacturing.RemanufacturingρYgAITPemiWithout62357.6642.039417,981.572493.74With88369.3521.975439,427.042530.53 The optimal values under CCT with and without remanufacturing. Table 6The optimal values under CT and CCT policy without remanufacturing.Carbon policyρYgAITPemiCT61357.0292.362417,653.672615.62CCT62357.6642.039417,981.572493.74 The optimal values under CT and CCT policy without remanufacturing. Table 7The optimal values under CT and CCT policy with remanufacturing.Carbon policyρYgAITPemiCT88369.2042.308439,080.362659.41CCT88369.3521.975439,427.042530.53 The optimal values under CT and CCT policy with remanufacturing. The comparative analysis results demonstrate that the CCT policy consistently dominate the CT policy in terms of effectiveness, irrespective of the inclusion of remanufacturing. Additionally, integrating remanufacturing strategies under both carbon policies significantly enhances system outcomes. Consequently, the system achieves optimal AITP and ensures effective waste management under the CCT policy with remanufacturing. Therefore, adopting a remanufacturing strategy is advantageous for the banana yarn industry. Sensitivity analysis A sensitivity analysis is carried out to investigate the responsiveness of the system’s decision factors to variations in individual parameters. By analysing the sensitivity table, the manager can identify key parameters and make informed decisions to maintain profitability while adhering to sustainable standards. The outcomes of the sensitivity analysis are given in Tables 8 and 9. Table 8Sensitive analysis without considering remanufacturing.Carbon tax without remanufacturingCarbon cap and trade without remanufacturingPRValuesρYgemiPRValuesρYgemiM1661410.5002.373417,942.122620.28M1662411.0072.051418,269.562497.281861381.4442.365417,800.762616.891862382.0272.040418,128.332496.162061357.0292.362417,653.672615.622062357.6642.039417,981.572493.742261336.1742.363417,503.882615.582262336.8442.040417,832.262493.682461318.1182.366417,353.292617.032462318.8122.043417,682.252494.990.161356.9832.312417,706.032595.060.162357.6181.989418,031.292475.850.261357.0062.337417,679.772605.420.262357.6412.015418,006.352484.240.361357.0292.362417,653.672615.620.362357.6642.039417,981.572493.740.461357.0512.387417,627.732625.660.462357.6862.064417,956.952502.450.561357.0732.411417,601.942636.180.562357.7082.089417,932.472511.00161380.2221.224418,732.572241.22162379.9610.911419,008.852166.64261366.3851.875418,144.952434.00262366.6291.556418,448.372334.33361357.0292.362417,653.672615.62362357.6642.039417,981.572493.74461349.8812.755417,222.182789.38462350.8532.431417,573.082645.65561344.0283.084416,832.062958.41562345.3062.760417,204.982793.68161363.8572.366418,014.082617.35162364.5722.044418,342.652494.65261360.3802.364417,833.022616.34262361.0542.041418,161.242494.38361357.0292.362417,653.672615.62362357.6642.039417,981.572493.74461353.7952.360417,475.972615.16462354.3942.038417,803.562492.72561350.6722.359417,299.862614.31562351.2372.036417,627.162492.580.7061357.2122.029417,994.152487.220.662357.8481.654418,355.722364.960.7561357.1202.201417,816.492664.880.6562357.7561.854418,163.922429.300.8061357.0292.362417,653.672615.620.7062357.6642.039417,981.572493.740.8561356.9382.514417,493.212679.270.7562357.5722.212417,807.372557.680.9061356.8482.656417,335.612836.870.862357.4812.373417,640.272622.16ϕ0.7061356.3202.028417,354.373127.11ϕ0.7062357.0371.705417,744.102943.670.7561356.6702.201417,500.462870.100.7562357.3461.878417,859.282718.440.8061357.0292.362417,653.672615.620.8062357.6642.040417,981.572493.100.8561357.3982.514417,813.122359.600.8562357.9892.191418,110.112268.510.9061357.7782.658417,978.122103.120.9062358.3222.334418,244.192043.47ψ0.261354.9431.257416,764.764206.49ψ0.262355.5710.611417,297.744086.090.361356.2952.191417,232.323145.370.362356.9271.760417,628.463024.200.461357.0292.362417,653.672615.620.462357.6642.040417,981.572493.100.561357.4902.340417,988.692294.850.562358.1262.078418,275.722174.970.661357.8062.251418,255.372084.980.662358.4442.036418,515.191962.4070061457.0831.321412,559.212266.7470062457.7601.001421,850.322187.2980061389.1852.000419,216.432477.2980062389.8231.678419,529.602372.2290061357.0292.362417,653.672615.6290062357.6642.040417,981.572493.10100061357.6942.596416,466.852894.36100062338.3282.273416,846.542581.33110061324.6312.762416,507.872715.80110062325.2652.438415,970.052649.16 Sensitive analysis without considering remanufacturing. Table 9Sensitive analysis considering remanufacturing.Carbon tax with remanufacturingCarbon cap and trade with remanufacturingPRValuesρYgemiPRValuesρYgemiM1688424.1232.326439,005.402999.26M1688424.0851.992439,686.222536.831888394.2962.314438,880.602994.461888394.3651.980439,560.302533.162088369.2042.308439,080.362659.412088369.3521.975439,427.042530.532288347.7522.306438,610.882991.022288347.9591.972439,289.802530.632488329.1642.307438,471.682991.482488329.4161.974439,150.702530.430.188329.1642.257438,805.002968.050.188369.3131.923439,478.842512.650.288369.1852.283438,776.412979.680.288369.3321.949439,452.862521.680.388369.2042.308439,080.362659.410.388369.3521.975439,427.042530.530.488369.2232.334438,719.733003.080.488369.3711.999439,401.392540.520.588369.2412.359438,691.633014.870.588369.3902.025439,375.902549.04188393.5021.198440,150.432291.01188392.8020.867440,451.092208.29288379.0251.832439,568.252480.74288378.7861.498439,892.642374.90388369.2042.308439,080.362659.41388369.3521.975439,427.042530.53488361.6992.693438,651.412830.27488362.1932.360439,019.472680.62588355.5613.017438,263.422995.67588356.3712.685438,652.222825.71188376.5962.311439,453.222661.09188376.8031.978439,800.072532.08288372.8292.309439,265.872660.43288373.0061.976439,612.622531.49388369.2042.308439,080.362659.41388369.3521.975439,427.042530.53488365.7112.307438,896.642658.68488365.8331.973439,243.252530.50588362.3432.305438,714.632658.86588362.4401.972439,061.192530.500.7088369.3521.975439,426.692530.530.6088369.5021.589439,806.962402.590.7588369.2782.147439,249.892595.020.6588369.4271.789439,612.162466.840.8088369.2042.308439,080.362659.410.7088369.3521.975439,427.042530.530.8588369.1312.459438,917.232723.950.7588369.2782.150439,250.262593.060.9088369.0582.602438,759.742788.050.8088369.2042.308439,080.762659.41ϕ0.7088368.6321.974438,784.703172.84ϕ0.7088369.2042.308439,142.262597.870.7588368.9142.147438,928.892915.850.7588369.0941.813439,307.672755.550.8088369.2042.308439,080.362659.410.8088369.3521.975439,427.042530.530.8588369.5022.460439,238.232402.440.8588369.6162.126439,551.772248.580.9088369.8082.603439,401.772145.670.9088369.8862.260439,684.242087.49ψ0.288367.4861.149438,204.154291.31ψ0.288367.6110.482438,768.714162.590.388368.5992.118438,660.093203.430.388368.7391.673439,079.403075.040.488369.2042.308439,080.362659.410.488369.3521.975439,427.042530.530.588369.5842.293439,416.812332.910.588369.7372.026439,719.912204.520.688369.8452.215439,685.532115.200.688370.0021.992439,959.581987.1370088487.1231.152443,301.162278.4170088486.6530.818443,600.282197.9980088405.5911.917440,750.502510.7180088405.5791.584441,078.592400.3690088369.2042.308439,080.362659.4190088369.3521.975439,427.042530.53100088347.8292.556437,864.662766.54100088348.0582.223438,224.732624.37110088333.5702.730436,929.542848.51110088333.8462.397437,299.852696.23η0.1588369.1982.301426,685.282655.86η0.1588369.3461.967427,031.572528.030.2588369.2012.305432,882.822657.310.2588369.3491.971433,229.302529.280.3588369.2042.308439,080.362659.410.3588369.3521.975439,427.042530.530.4588369.2072.312445,277.912660.860.4588369.3551.978445,624.782532.430.5588369.2102.316451,475.462662.300.5588369.3581.982451,822.532533.68β0.388369.2002.302430,403.812657.65β0.388369.3481.969430,750.212529.040.488369.2022.306434,742.082657.880.488369.3501.972435,088.622529.790.588369.2042.308439,080.362659.410.588369.3521.975439,427.042530.530.688369.2062.311443,418.652660.290.688369.3541.978443,765.462531.280.788369.2082.314447,756.932661.170.788369.3561.980448,103.882532.67 Sensitive analysis considering remanufacturing. Effect of emission parameters and shipment parameter on AITP The variations in emissions across different supply chain activities significantly impact overall system performance. As , and increase, they contribute to financial losses and elevate the overall emission burden. Notably, emissions from the manufacturer’s warehousing () negatively affect green investment (GI), as higher emissions necessitate increased investment in sustainable practices, thereby placing additional financial pressure on the manufacturer. Also, an increase in the number of shipments prolongs material holding times, leading to higher emissions. Therefore, it is crucial to optimize the number of shipments while managing the emission costs associated with warehousing to ensure balanced profitability. Effect of tax imposition on AITP The imposition of a carbon tax directly impacts the system’s profitability, wherein an increase in the tax levied on each unit of carbon leads to a corresponding increase in emission costs, ultimately resulting in a decrease in AITP. Furthermore, the analysis reveals that fluctuations in the carbon tax also have a significant effect on green investment, underscoring the need for policymakers to strike a balance between environmental sustainability and economic viability. To mitigate the adverse effects of emissions on the system’s profitability, it is essential to implement sustainable measures that limit emissions, thereby minimizing the economic burden of environmental degradation. Effect of efficiency parameter in green technology The analysis of emission reduction factors reveals a direct relationship between the fraction of emission reduced and the system’s economic performance, where an increase in the former leads to an increase in AITP and a decrease in emission costs. Additionally, the data presented in Tables 8 and 9 highlight the critical role of technological efficiency in shaping system outcomes, as an increase in efficiency is accompanied by an increase in AITP, a decrease in emission costs, and an increase in green investment. To achieve optimal system performance, it is essential to make strategic decisions that integrate investment costs, technological efficiency, and emission reduction targets, ensuring a balanced and sustainable approach to emission reduction. Effect of β and η on AITP and emi The implementation of remanufacturing strategies is significantly influenced by the proportion of recyclable garments within the collected used clothing. An increase in the collection rate of old clothes corresponds to a higher fraction of recyclable materials. As demonstrated in Table 9, a greater fraction of recyclable clothing positively impacts the AITP. Furthermore, the effectiveness of remanufacturing is more pronounced under the CCT policy compared to the CT policy. This is evident from the superior outcomes observed in green investment costs, emission costs, and AITP under the CCT framework. Figure 7 presents the variation in AITP and emission cost in response to changes in the collection rate of used clothes and the fraction of recyclable clothes under carbon tax (CT) and carbon cap-and-trade (CCT) policies with remanufacturing. The impact of the fraction of recyclable clothes β and the collection rate of used clothes η in the average integrated total profit and emission cost. Managerial insights The sensitivity analysis provides clear quantitative priorities for managerial decision-making. First, adopting remanufacturing increases AITP by approximately 5% under both CT and CCT policies, establishing it as a structurally significant profit driver. Among remanufacturing parameters, η exhibits the strongest impact, with a 5.8% increase in AITP as it rises from 0.15 to 0.55, followed by β which generates a 4.0% profit improvement over its range; notably, both effects are achieved with only marginal increases in emission cost (below 0.25%). In contrast, emission-related parameters primarily exert downward pressure on profitability: increases in the manufacturer’s production emission parameter () reduce AITP by only 0.02–0.03%, while the retailer’s warehousing emission parameter () causes a moderate 0.17% decline. The most influential emission driver is the manufacturer’s warehousing emission parameter (), which leads to a 0.41–0.45% reduction in AITP and a substantial 28–31% increase in emission cost, identifying upstream storage operations as the most critical emission control point. Across all scenarios, the CCT framework consistently yields marginally higher profitability and lower emission costs than CT. Overall, the results indicate that profitability improvements are predominantly driven by strengthening remanufacturing efficiency and system adoption decisions, whereas emission intensity particularly at the manufacturer’s warehousing stage constitutes the primary regulatory cost risk under environmental policies. Discussion and findings This study utilizes banana pseudo-stems as a raw material for yarn production, illustrating inter-industrial connectedness, a concept in line with36 on integrating meat and leather industries to harness refused hides for sustainable leather production. The findings indicate that the cap-and-trade (CCT) policy outperforms the carbon tax (CT) mechanism due to its dual incentive structure, whereby firms reduce compliance costs through emission mitigation while simultaneously generating revenue from trading surplus allowances. This flexibility strengthens profitability, particularly in supply chains with established remanufacturing and reverse logistics systems. The results of this study are consonant with the findings of22, suggesting that CCT policies offer a more advantageous economic-environmental trade-off. Sensitivity analysis demonstrates that as emission reduction efficiency and customer return rates increase, the economic advantage of CCT becomes more pronounced, since remanufacturing decreases reliance on virgin raw materials, lowers production costs, and reduces associated emissions. In contrast, under a CT regime, emission reductions merely reduce tax liability without creating additional financial leverage, resulting in a linear cost burden that limits profit expansion. The model also highlights the critical role of end customer-initiated returns, revealing that higher collection rates significantly enhance profitability because remanufactured units retain greater economic value compared to salvage operations. These results suggest that firms in circular agro-based industries should strategically strengthen reverse logistics and emission reduction investments to maximize gains under tradable permit systems, while policymakers should prioritize cap-and-trade mechanisms in sectors with strong remanufacturing potential to better align environmental regulation with economic sustainability. Conclusions This study develops an integrated sustainable production model for the banana yarn industry, encompassing a manufacturer–retailer framework that incorporates incentive-based end-of-life processing for discarded textiles. The proposed supply chain model embeds green technology investments for emission mitigation and wastewater treatment systems for textile effluent purification. The model is examined under four distinct policy scenarios: (1) CT with remanufacturing, (2) CT without remanufacturing, (3) CCT with remanufacturing, and (4) CCT without remanufacturing. For each scenario, the optimal system performance and corresponding decision variables were determined. Comparative analysis reveals that adopting remanufacturing strategies under the CCT policy results in higher profitability with lower emissions and reduced investment requirements. The sensitivity analysis further provides managerial insights into how variations in key parameters influence system performance, assisting the banana yarn industry in designing sustainable and economically viable operational policies. Notably, the model demonstrates that fulfilling the Extended Producer Responsibility (EPR) obligation leads to a 5.13% increase in the AITP, emphasizing the economic and environmental advantages of remanufacturing. Overall, the findings underscore that strategic investments in green technology, wastewater treatment, and remanufacturing practices can enhance profitability while advancing sustainability objectives. This research contributes to the existing literature by introducing a Sustainable Production–Consumption (SPC) model for the yarn industry through the establishment of circular industrial connectedness between the banana and yarn sectors. Future research directions include examining the impact of additional factors such as consumer behaviour, market competition, and developing more advanced optimization techniques to support sustainable production decision-making. Supplementary Information Supplementary Information. Author contributions Vishnupriya Kalathil Sahadevan conceptualized and designed the study and prepared the original manuscript draft. Abin Thomas conducted the validation and performed the visualization. Umakanta Mishra reviewed and edited the final manuscript draft and contributed to the visualization. All authors reviewed and approved the final version of the manuscript. Funding Open access funding provided by Vellore Institute of Technology. Data availability The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request. Code availability The computational analysis in this study was performed using Wolfram Mathematica 9.0. Available at: https://www.wolfram.com/mathematica/. No custom code is required to reproduce the results, as all mathematical formulations and solution procedures are fully described in the manuscript. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Information. Data Availability Statement The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request. The computational analysis in this study was performed using Wolfram Mathematica 9.0. Available at: https://www.wolfram.com/mathematica/. No custom code is required to reproduce the results, as all mathematical formulations and solution procedures are fully described in the manuscript.