Closed-Loop Supply Chain Recycling Mode Selection for New Energy Vehicle Power Battery Considering Traceability Factor

1. Introduction and Research Background

The global energy structure transition and the pursuit of carbon neutrality have driven explosive growth in the new energy vehicle (NEV) industry. China has with strategic policy frameworks systematically deployed clean transportation transitions, including the comprehensive promotion of NEVs and the acceleration of traditional fuel vehicle replacement. By 2023, the global stock of new energy vehicles has surpassed 40 million units, with China contributing 60 percent of this market share. However, the “heart” of new energy vehicles—the power battery—has an average service life of only 5 to 8 years. An impending wave of retired power batteries is creating new environmental and resource challenges. Data indicates that China will generate an estimated 780,000 tonnes of retired EV batteries in 2025, exceeding 3 million tons by 2030. If mishandled, this volume will not only waste strategic resources such as cobalt, nickel, and lithium but also risk irreversible damage to ecosystems due to electrolyte leakage and heavy metal contamination.

Current global proven lithium reserves can only satisfy demand for the next 20 years, notwithstanding EV batteries contain recyclable metal rates up to 90 percent. At the same time, international regulatory frameworks such as the EU New Battery Regulation and China’s 14th Five-Year Plan for circular economy development have elevated EV battery recycling from a purely technical issue to a strategic high ground in global industrial competition. Within this scenario, the recycling segments of the EV battery closed-loop supply chain manifest challenges of information asymmetry, low efficiency, and traceability difficulties. Consequently, guaranteeing EV battery traceability and optimizing terminal recycling governance have become urgent issues. Blockchain technology—as a decentralized distributed ledger with immutability, transparency, and trustlessness characteristics—has migrated from the cryptocurrency field to diversified scenarios. Research on blockchain applications has concentrated mainly on traceability in food, agriculture, and pharmaceutical products. However, within the EV battery closed-loop supply chain domain, deploying blockchain-powered traceability can realize full-process tracking of waste sources, destinations, and treatment procedures, thereby elevating resource utilization efficiency and environmental protection.

In the forward supply chain, blockchain-enabled traceability can create transparent battery profiles through encrypted full-chain data storage for raw material tracing, standardized manufacturing monitoring, and after-sales dynamic tracking. Consumers gain real-time insight into EV battery conditions, which elevates purchase intention. Empirical studies indicate that consumers are willing to pay premiums for traceable products; moreover, the more process information about product stewardship obtained through blockchain, the stronger the willingness to purchase. In the reverse supply chain, traceability can ensure retired EV batteries are formally recycled and processed, meeting demands from environmentally-conscious segments while smart contracts trigger end-of-life alerts and prompt consumers toward recycling participation.

As policy frameworks reinforce battery supervision, the Chinese Ministry of Industry and Information Technology issued the Interim Provisions on Traceability Management for Recycling and Utilization of New Energy Vehicle Power Batteries in 2018, mandating the construction of full-life-cycle traceability supervision platforms for power batteries. As policy responses, battery enterprises launched blockchain-based traceability systems for cathode material provenance and full-life-cycle intelligent management in 2019 and 2023 respectively. Many EV battery manufacturers have since deployed battery traceability for recycling purposes, causing increasingly complex competitive and cooperative relationships among enterprises. On one hand, enterprises establish partnerships to jointly build traceability systems and participate in recycling operations; on the other hand, there is simultaneous competition from other recyclers seeking recycled resources. Collaboration enables shared resources, technologies, and market intelligence to improve overall recycling efficiency and economy. Competition encourages continual innovation and optimization in recycling technologies, improving service quality to attract more consumers and partners. This dynamic of concurrent competition and cooperation has become customary in the recycling sector and is crucial for the sustainable development of the whole industry.

This evolving competitive-cooperative structure has transformed cooperation patterns among EV battery-related enterprises. Which cooperation mode proves most effective requires context-specific investigation. There are two principal concerns: one focuses on upstream design by introducing traceability systems to record full-life-cycle information, providing data support for eco-design; the other addresses end-of-pipe governance by determining how multiple parties should cooperate within traceability-driven systems to maximize recycling volume and benefits. The former is a design problem, while the latter is a supply chain management problem. This thesis addresses the second problem—constructing cooperation models among battery cell makers, vehicle manufacturers, and third-party recyclers, and comparing supply-chain efficiency and profitability under different cooperation-competition structures to identify optimal recycling modes.

The specific research questions include: (1) under distinct recycling modes, how do chain members determine optimal pricing and traceability decisions? (2) how do traceability cost coefficients, cross-channel competition coefficients, and consumer traceability preferences influence key decision variables? (3) under profit-maximization and collection-volume-maximization objectives, which cooperation mode is optimal for chain members?

2. Literature Review

2.1 Closed-Loop Supply Chain Studies for EV Batteries

The closed-loop supply chain concept was introduced by Stock and subsequently refined by Guide and Van Wassenhove and Fleischmann et al., who developed theoretical frameworks and established quantitative research foundations. Since then, closed-loop supply chain research has delved into issues surrounding EV battery ecosystems, focusing on government subsidy policies, carbon trading policies, and echelon utilization.

Government subsidies constitute a robust policy instrument for promoting EV battery closed-loop supply chains. Studies have systematically compared scenarios of no subsidy versus subsidizing different chain members, analyzed how subsidy objects influence pricing and profitability, and investigated how direct subsidies lower recycling costs while indirect subsidies, such as tax incentives, support longer-term incentives. Further research has compared quantity-based subsidies and capacity-based subsidies—discovering that quantity-based subsidies effectively improve recycling rates while capacity-based subsidies may even demonstrate counterproductive effects. Additional findings suggest that subsidy policies, dual-credit policies, and charging infrastructure promote NEV diffusion, while restrictions on fuel vehicles exhibit an inverted-U relationship. Consumer green preferences contribute to improving social welfare under well-designed subsidies.

Carbon trading policy has also emerged as a critical pathway towards low-carbon transformation in the EV battery closed-loop supply chain. Studies have found that under cap-and-trade regimes, positive market outcomes can arise in both forward sales and remanufacturing markets. Researchers explored decentralized models, discovering that consumers’ willingness to pay for remanufactured products simultaneously boosts manufacturer and retailer profits, whereas carbon quotas produce threshold effects: at lower values they increase manufacturer profits, but once a certain level is passed, these quotas restrict profit enhancement. Carbon emission reduction of products increases output levels. Considering consumer preferences, investment difference factors constitute a key determinant in selecting emission-reduction modes.

Echelon utilization represents a significant outlet for retired EV batteries. Research has moved from subsidy-mode design to game-theoretic strategy analysis. Studies have proposed four distinct subsidy models and analyzed the effects of subsidy objects, amounts, and scale economies on downstream recycling and echelon utilization profits. Subsidizing echelon utilizers has been shown to raise profits for both recyclers and echelon utilizers. Government interventions significantly improve recycling rates, with reward-penalty mechanisms proving more efficient than subsidies alone. Fairness concerns influence pricing; manufacturers, under fairness concerns, may increase retail prices for new batteries while increasing retired battery collection rates. Price sensitivities among echelon utilizers further determine effort utility and utilization rates. Approaches comparing scenarios of no subsidies, production R&D subsidies, and recycling subsidies demonstrate that government recycling subsidies can increase recycling rates while production R&D subsidies contribute more to overall supply chain profitability.

2.2 Recycling Mode Studies for EV Batteries

Effective EV battery closed-loop supply chain operation requires rational and efficient recycling configurations. Research has mainly addressed single-entity recycling and multi-agent mixed recycling.

In single-entity systems, pioneering studies built manufacturer-collection, retailer-collection, and third-party-collection model structures, finding that retailer-managed collection maximizes supply chain profit—a foundational benchmark. Later extensions combined sales channel structures to produce five recycling modes and compared optimal pricing decisions. Contextual refinement incorporated initial investment coefficients to quantify differentiated channel operating costs. With increasingly complex policy environments, researchers investigated external factors affecting single-entity channel selection, finding that retailer recycling dominates when government subsidies are introduced. Deeper study of subsidy intensity showed that high subsidy environments make manufacturers more inclined toward direct collection. Subsidy threshold effects influence selection between retailer and third-party collection. Carbon trading combined with advertising levels shapes recycling channel choices through joint effects. Patent licensing strategies, under manufacturer leadership, favor retailer recycling and remanufacturing with fixed royalty; outsourcing versus authorization remanufacturing modes also determine channel attractiveness.

Multi-agent hybrid recycling has become a salient research subject as EV battery recycling industries evolve. Early analysis focused on channel competition in dual-channel direct-indirect settings, showing that the degree of competition significantly impacts channel selection. Quantitative analysis of market competition found that retailer-led recycling gains advantages from synergy effects under intense competition, whereas manufacturers prefer in-house collection under lower competition intensity. Cooperative versus non-cooperative models demonstrated that chain-member coordination achieves Pareto improvements. Conceptual frameworks distinguishing autonomous, outsourced, and joint recycling were developed with graphical system models. Product-quality-difference extensions revealed that manufacturer-retailer combined channels generate the highest recycling quantity and profit. Later research introduced triple-party joint recycling with third-party enterprises, establishing that tripartite recycling achieves maximum quantity and profit. Dual-channel considerations under fairness concerns showed that carbon emission reductions and retail sales effort respond to fairness coefficient variations. Online-offline dual-channel retail under fairness-concern scenarios determined that fairness coefficients correlate positively with retail prices; when fairness is low, bank financing is preferable; otherwise, third-party platforms outperform. Emission-reduction technology innovation investment affects dual-channel pricing decisions. Carbon-regulation analyses comparing three recycling channels found that joint recycling yields superior profit and carbon-reduction outcomes when recycling-competition coefficients are moderate. Government subsidy strategies were then incorporated, comparing two subsidy mechanisms under dual goals of overall profit and unit subsidy efficiency.

2.3 Traceability-Conscious EV Battery Recycling Research

Increasing environmental concern and consumer focus on quality and security in NEVs have catalyzed traceability-conscious recycling modes for EV batteries. These approaches integrate traceability systems into existing structures, refine information flow and trust mechanisms among actors, and decisively affect responsibility assignment, quality control, and optimization decision-making.

Technical foundations were first established by developing supply-chain traceability platforms through blockchain and IoT integration. Blockchain-based echelon utilization schemes for EV batteries soon followed, emerging as information storage chains with consensus mechanisms for managing full-life-cycle data. Methods for energy-storage monitoring transactions, block-based accounting, and traceability were designed subsequently. These efforts underpinned traceability applications in closed-loop supply chains. Conceptual frameworks were formulated through five-layer blockchain architecture models specific to NEV recycling, combining data, network, consensus, contract, and application layers. The focus has since shifted toward comparing profit, recycling quantities, and traceability levels as bases for recycling-mode selection. Studies between manufacturer and platform recycling found manufacturer-controlled recovery economically superior. When considering product greenness, retailer recycling provided superior overall supply chain profits, irrespective of traceability deployment. Quantifying traceability levels as endogenous decision variables revealed that manufacturer-led recycling achieves the highest traceability levels. Comparative analyses of recycling efficiencies demonstrated that post-blockchain recycling efficiency exceeds traditional modes, with joint recycling and third-party recycling yielding the highest performance. Double-channel recycling outperforms single-channel configurations.

2.4 Research Gap and Contributions

Synthesizing prior literature, the following observations emerge. (1) EV-battery closed-loop supply chain scholarship has principally addressed carbon cap-and-trade, subsidies, and echelon utilization through policy transmission and market regulation interactions—these provide a robust theoretical foundation for recycling practice. (2) Recycling mode research based on entity differences has bifurcated into single-subject and multi-agent mixed streams. Nevertheless, information asymmetry and operational redundancy continue to characterize real practice, requiring traceability integration. (3) Traceability-conscious recycling mode research has evolved from technical inception to architectural mode building and multidimensional comparative analysis, yet most studies assume single-recycler recycling or symmetrical duopoly competition. They rarely model heterogeneous alliances that internally cooperate while externally competing against an independent recycler—the primary pattern observed in the contemporary Chinese EV battery recycling market.

This dissertation introduces blockchain traceability to the circular EV battery supply chain and examines hybrid recycling modes formed by cooperation and competition among participating entities. Three hybrid modes are modelled and compared. This extends traceability theory to EV battery end-of-life governance, and simultaneously supplements the recycling-mode selection literature with explicit competitor-cooperator structures. It also creates practical decision support for ecosystem actors.

3. Theoretical Framework

3.1 Definition of Traceability Factors in the EV Battery Context

Traceability emerged in the 1990s as European responses to food safety crises established information chains spanning complete production-to-consumption routes. These mechanisms enable reverse tracing to raw material origins and production units in the event of quality anomalies, pinpointing responsible parties. The application scope has since extended into pharmaceutical monitoring, textile life cycle management, and waste reverse supply chains. For EV battery systems, traceability encompasses information acquisition throughout raw material procurement, cell production, vehicle integration, usage through maintenance and repair, and final recycling and dismantling.

In this dissertation, traceability factors are the technical attributes and behavioral parameter sets embedded in the EV-battery closed-loop supply chain to realize full-life-cycle traceability of the battery. They encompass: (1) a technical attribute dimension, referring to the system architecture that the cell manufacturer builds based on blockchain, and in which all supply chain members join to upload node information; and (2) a behavioral parameter dimension, comprising key mathematical model parameters: a traceability cost coefficient representing system investment; a consumer traceability-preference coefficient embodying willingness to pay for traceable products; and traceability levels as endogenous decision variables.

3.2 Theoretical Underpinnings

Extended Producer Responsibility. Originating in Europe during the 1990s, extended producer responsibility identifies producers as principal environmental stewards across the complete horizon from product design to end-of-life processing. In the EV-battery closed-loop supply chain, cell manufacturers bear such responsibility. This motivates them to embed traceability into design, establish blockchain tracing systems, and execute reverse channels either alone or through alliances with vehicle manufacturers or third-party recyclers.

Circular Economy Theory. The core 3R principles—Reduce, Reuse, and Recycle—structure effective resource efficiency. The constructed closed-loop supply chain embodies resource circularity: full-chain traceability systems track production, minimize input consumption through remanufacturing design, prolong battery lifecycles, and feed recovered metals into closed-loop production. These principles are operationalized in collection network optimization.

Blockchain Theory. Blockchain is a distributed database combining decentralized consensus and cryptographic immutability. Its architecture—distributed ledger, consensus mechanism, smart contract, encryption—forms the basis for multi-party collaborative traceability platforms. All nodes within the EV battery closed-loop supply chain upload their compliance and product-related information, forming a cross-organizational “trusted data pool” that supports precise pricing, dynamic monitoring, and terminal recovery.

Game Theory. Stackelberg games characterize leader-follower hierarchies observed in the focal chain. Cell manufacturers lead as Stackelberg leaders, while vehicle manufacturers and third-party recyclers follow. Each decision-maker is rational, information is complete, and backward induction generates subgame-perfect equilibria.

4. Model Design and Assumptions

4.1 Problem Statement and Parameter Settings

This study investigates a closed-loop supply chain consisting of an EV battery cell manufacturer, a vehicle manufacturer, and a third-party recycler. The cell manufacturer is the Stackelberg leader, manufactures batteries, and constructs a blockchain-based traceability system. The vehicle manufacturer is responsible for battery sales. After consumers have utilized their batteries, recyclers collect them and transfer them to the cell manufacturer for dismantling and remanufacturing. Based on real-world alliance configuration, three hybrid recycling modes are constructed:

Mode 1 (MR–T): the cell manufacturer and the vehicle manufacturer operate as a joint recycling coalition, categorized as “cell manufacturer–vehicle manufacturer joint recycling with a competing independent third-party recycler.”

Mode 2 (MT–R): the cell manufacturer and the third-party recycler form the coalition, categorized as “cell manufacturer–third-party joint recycling with a competing independent vehicle manufacturer.”

Mode 3 (M–RT): the vehicle manufacturer and the third-party recycler operate jointly while the cell manufacturer collects independently, categorized as “independent cell manufacturer with vehicle manufacturer–third-party coalition recycling.”

The variables and symbols used in modeling are listed below in Table 1.

Table 1: Model symbols and definitions
Symbol Definition
\( W \) Unit wholesale price of EV battery
\( P \) Unit retail price of EV battery
\( P_i \) Unit collection price of channel \( i,\; i \in \{MR,T,MT,R,M,RT\} \)
\( \gamma_i \) Traceability level of channel \( i,\; i \in \{MR,T,MT,R,M,RT\} \)
\( b \) Unit transfer price
\( C_n \) Unit production cost for new batteries, \( C_n>0 \)
\( C_r \) Unit production cost for remanufactured batteries, \( C_r>0 \)
\( \delta_1 \) Consumer sensitivity coefficient for traceability level, \( 0<\delta_1<1 \)
\( \delta_2 \) Collection-volume sensitivity coefficient to traceability level, \( 0<\delta_2<1 \)
\( A_1 \) Market size of new EV battery demand, \( A_1\gg0 \)
\( A_2 \) Voluntary consumer return volume without monetary incentives
\( k \) Traceability cost coefficient, \( k>0 \)
\( K \) Fixed cost of traceability system
\( \theta \) Recycling channel competition coefficient

Additional superscripts or subscripts identify channel-specific variables that correspond to each recycling configuration. In mode MR–T, \( P_{MR} \) and \( \gamma_{MR} \) denote the price and traceability level practiced by the cell-manufacturer–vehicle-manufacturer coalition, while \( P_T \) and \( \gamma_T \) denote the third-party recycler’s values. In mode MT–R, \( P_{MT} \) and \( P_R \) refer to values for the coalition and the vehicle manufacturer. In mode M–RT, \( P_M \) and \( P_{RT} \) refer to the independent cell manufacturer and the coalition, respectively.

4.2 Model Assumptions

The principal assumptions are that: (1) retired EV batteries are directed only to component recovery for remanufacturing, not to echelon use applications; (2) the three enterprises are independent, fully rational, risk-neutral self-profit optimizers, with the cell manufacturer as the dominant mover in pricing and traceability decisions; (3) remanufactured batteries achieve quality parity with new batteries, so consumers make no distinction between them and retail prices coincide; demand for EV batteries is stable in a complete sales cycle, so product prices remain constant; (4) wholesale and retail margins are positive: \( P>W>0 \); (5) production costs satisfy \( C_n>C_r+b>0 \), ensuring the cell manufacturer engages in remanufacturing, and transfer margins also satisfy \( b>P_i \); (7) traceable information improves consumer trust, generating market demand \( Q=A_1-P+\delta_1 ( \gamma_m+\gamma_f ) \), where \( \gamma_m \) and \( \gamma_f \) represent leader and follower levels of traceability; (8) the consumer-return function to each channel i is \( Q_{i1}=A_2+P_{i1}-\theta P_{i2}+\frac{\delta_2}{2}(\gamma_m+\gamma_f) \), capturing own-price positive effects, cross-price negative effects, traceability-driven convenience, and voluntary returns; (9) traceability-related investment costs follow a quadratic function \( C(\gamma_i)=\frac{1}{2}k\gamma_i^2 \), and fixed platform access costs are \( K \).

5. Mathematical Modelling and Equilibrium Solutions

5.1 Model of Cell Manufacturer–Vehicle Manufacturer Joint Recycling

In the MR–T configuration, the coalition produces, sells, and collects the EV battery, whereas the third-party recycler collects and transfers retired units to the coalition. Profit functions for the coalition \( \Pi_{MR} \) and the third-party company \( \Pi_T \) are expressed as follows:

$$ \Pi_{MR} = (P-C_n)\bigl[A_1-P+\delta_1(\gamma_{MR}+\gamma_T)\bigr] + (C_n-C_r-b)\Bigl[A_2+P_T-\theta P_{MR}+\frac{\delta_2}{2}(\gamma_{MR}+\gamma_T)\Bigr] + (C_n-C_r-P_{MR})\Bigl[A_2+P_{MR}-\theta P_T+\frac{\delta_2}{2}(\gamma_{MR}+\gamma_T)\Bigr] – \frac{k}{2}\gamma_{MR}^2 – K $$

$$ \Pi_T = (b-P_T)\Bigl[A_2+P_T-\theta P_{MR}+\frac{\delta_2}{2}(\gamma_{MR}+\gamma_T)\Bigr] – \frac{k}{2}\gamma_T^2 $$

Since \( \Pi_T(P_T,\gamma_T) \) is concave, the third-party recycler’s reaction functions are obtained by first-order optimality. Substituting these into \( \Pi_{MR} \) and solving the resulting system yields the equilibrium solutions of Proposition 1.

Proposition 1. In Mode MR–T, the equilibrium retail prices, collection prices, wholesale conditions, and traceability levels are:

$$ P = \frac{M_7+(A_2+\Delta S)\bigl[2(1+\theta)\delta_1\delta_2^2-64\delta_1\delta_2 k\bigr]+8C_n\delta_1^2 S(\delta_2^2+16k)}{2B+2\delta_2^4(1+\theta)+16\delta_2^2 k(5-\theta)} $$

$$ b = \frac{M_6-M_3+4A_2\delta_1^2\delta_2^2+M_4-(1+\theta)\delta_2^4+8(2A_2-3\Delta)\delta_2^2 k}{B+\delta_2^4(1+\theta)+8\delta_2^2 k(5-\theta)} $$

$$ \gamma_{MR} = \frac{(A_1-C_n)M_1+M_5+(2A_2S+2S^2\Delta)\delta_1^2\delta_2-4k\theta^2\Delta}{B+\delta_2^4 Q+8\delta_2^2 k(5-\theta)} $$

$$ P_{MR} = \frac{-\Delta C+B+A_2 M_2-(A_1-C_n)\delta_1\delta_2^2-16(A_2-C_n)\delta_1\delta_2 k}{B+\delta_2^4(1+\theta)+8\delta_2^2 k(5-\theta)} $$

$$ P_T = \frac{M_8+\Delta\delta_2^4(1+\theta)-48\delta_1^2 k A_2-4\delta_2^2 k(3A_2-7\Delta)+M_9+M_{10}}{2B+2\delta_2^4(1+\theta)+16\delta_2^2 k(5-\theta)} $$

$$ \gamma_T = \frac{-\delta_2\bigl[M_{11}+\delta_2^2 A_2(1+\theta)-\Delta\delta_1^2(1+4\theta^2)+\Delta\delta_2^2(1-\theta^2)\bigr]}{B+\delta_2^4(1+\theta)+8\delta_2^2 k(5-\theta)} $$

where \( B=2S[16\delta_1^2 k-32k^2+\delta_1^2\delta_2^2] \), \( Q=1+\theta \), \( S=1-\theta \), \( \Delta=C_n-C_r \), and \( M_1,\ldots,M_{11} \) represent polynomial groupings. These equilibrium solutions characterize the full pricing and traceability equilibrium under hybrid competition with a third-party recycler.

5.2 Model of Cell Manufacturer–Third-Party Joint Recycling

In Mode MT–R, the cell manufacturer and third-party recycler form the coalition, while the vehicle manufacturer separately collects and sells. The profit functions are:

$$ \Pi_{MT} = (W-C_n)\bigl[A_1-P+\delta_1(\gamma_{MT}+\gamma_R)\bigr] + (C_n-C_r-b)\Bigl[A_2+P_R-\theta P_{MT}+\frac{\delta_2}{2}(\gamma_{MT}+\gamma_R)\Bigr] + (C_n-C_r-P_{MT})\Bigl[A_2+P_{MT}-\theta P_R+\frac{\delta_2}{2}(\gamma_{MT}+\gamma_R)\Bigr] – \frac{k}{2}\gamma_{MT}^2 – K $$

$$ \Pi_R = (P-W)\bigl[A_1-P+\delta_1(\gamma_{MT}+\gamma_R)\bigr] + (b-P_R)\Bigl[A_2+P_R-\theta P_{MT}+\frac{\delta_2}{2}(\gamma_{MT}+\gamma_R)\Bigr] – \frac{k}{2}\gamma_R^2 $$

The vehicle manufacturer determines retail price, collection price, and traceability level in the second stage. Embedding these reaction functions into the coalition profit and optimizing yields equilibrium decisions characterized in Proposition 2.

Proposition 2. In Mode MT–R, the unique equilibrium wholesale price, retail prices, collection prices, transfer price, and traceability levels satisfy the system of first-order equations derived from the coalition’s Hessian matrix. The explicit expressions are of the following functional form:

$$ P = \frac{N_1+128S\delta_1^2 k(A_1+2C_n)+16A_1\delta_1^2\delta_2^2 Q-N_2}{4S(8\delta_1^2\delta_2^2+192\delta_1^2 k-256k^2)+8\delta_2^4 Q+32\delta_2^2 k(5-\theta)} $$

$$ W = \frac{N_3+64\delta_1^2 k S(A_1+2C_n)+8\delta_1^2\delta_2^2 Q+(2\delta_1\delta_2^2-16\delta_1^2\delta_2)(A_2+A_2\theta+\Delta+\Delta\theta^2)}{S(8\delta_1^2\delta_2^2+192\delta_1^2 k-256k^2)+2\delta_2^4 Q+16\delta_2^2 k(5-\theta)} $$

$$ b = \frac{N_4+\delta_2^2 Q(\delta_2^2+2A_2\delta_1^2)+16k(A_2-\Delta S)(4k-3\delta_1^2)-N_5+2\Delta\delta_1^2\delta_2^2(3+2\theta-\theta^2)}{-32k S(3\delta_1^2-4k)-(4\delta_1^2\delta_2^2+\delta_2^4)Q-8\delta_2^2 k(5-\theta)} $$

$$ P_{MT} = \frac{N_6+(A_1-C_n)\delta_1\delta_2 k-4\delta_2^2 k(A_2-9\Delta)-N_7-16\Delta k\theta(3\delta_1^2-4k)}{-32k S(3\delta_1^2-4k)-(4\delta_1^2\delta_2^2+\delta_2^4)Q-8\delta_2^2 k(5-\theta)} $$

$$ P_R = \frac{N_8+16k A_2(3-\theta)(4k-3\delta_1^2)+N_9+2\Delta Q(\delta_2^4-5\delta_1^2\delta_2^2-32k^2 S)}{-128k S(3\delta_1^2-4k)-4(4\delta_1^2\delta_2^2+\delta_2^4)Q-32\delta_2^2 k(5-\theta)} $$

$$ \gamma_{MT} = \frac{32(A_1-C_n)\delta_1 k+N_{10}+\delta_2(A_2+\Delta)(48k-\delta_1^2)+16\delta_2 k\theta(A_2-2\Delta-\Delta\theta)}{-32k S(3\delta_1^2-4k)-(4\delta_1^2\delta_2^2+\delta_2^4)Q-8\delta_2^2 k(5-\theta)} $$

$$ \gamma_R = \frac{4\delta_1(A_1-C_n)(8k S-\delta_2^2 Q)+2\delta_2(A_2-\Delta S)\bigl[8k S+(8\delta_1^2+\delta_2^2)Q\bigr]}{32k S(3\delta_1^2-4k)+(4\delta_1^2\delta_2^2+\delta_2^4)Q+8\delta_2^2 k(5-\theta)} $$

where \( N_1,N_2,\ldots,N_{10} \) are polynomial groupings arising from the first-order conditions and matrical inversion.

5.3 Model of Vehicle Manufacturer–Third-Party Joint Recycling

In Mode M–RT, the vehicle manufacturer and third-party recycler operate the coalition, while the cell manufacturer independently collects retired EV batteries from consumers. The profit functions are:

$$ \Pi_M = (W-C_n)\bigl[A_1-P+\delta_1(\gamma_M+\gamma_{RT})\bigr] + (C_n-C_r-b)\Bigl[A_2+P_{RT}-\theta P_M+\frac{\delta_2}{2}(\gamma_M+\gamma_{RT})\Bigr] + (C_n-C_r-P_M)\Bigl[A_2+P_M-\theta P_{RT}+\frac{\delta_2}{2}(\gamma_M+\gamma_{RT})\Bigr] – \frac{k}{2}\gamma_M^2 – K $$

$$ \Pi_{RT} = (P-W)\bigl[A_1-P+\delta_1(\gamma_M+\gamma_{RT})\bigr] + (b-P_{RT})\Bigl[A_2+P_{RT}-\theta P_M+\frac{\delta_2}{2}(\gamma_M+\gamma_{RT})\Bigr] – \frac{k}{2}\gamma_{RT}^2 $$

Proposition 3 follows from the same procedure: the coalition determines its retail and collection prices and traceability level, and the cell manufacturer leads the wholesale, transfer, collection, and traceability decisions. The equilibrium is structurally analogous to Proposition 2 since the coalition always performs downstream retail service and the chain leader decides transfer terms, but the allocation of remanufacturing savings and responsibilities varies.

Proposition 3. In Mode M–RT, the optimal decisions are:

$$ P = \frac{4\delta_1(A_1-C_n)\bigl[8k S-\delta_2^2(1+\theta)\bigr]+2\delta_2(A_2-\Delta S)\bigl[8k S+(8\delta_1^2+\delta_2^2)(1+\theta)\bigr]}{32k S(3\delta_1^2-4k)+(4\delta_1^2\delta_2^2+\delta_2^4)(1+\theta)+8\delta_2^2 k(5-\theta)} $$

$$ \gamma_M = \frac{32(A_1-C_n)\delta_1 k+N_{10}+\delta_2(A_2+\Delta)(48k-\delta_1^2)+16\delta_2 k\theta(A_2-2\Delta-\Delta\theta)}{-32k S(3\delta_1^2-4k)-(4\delta_1^2\delta_2^2+\delta_2^4)Q-8\delta_2^2 k(5-\theta)} $$

$$ \gamma_{RT} = \frac{4\delta_1(A_1-C_n)(8k S-\delta_2^2 Q)+2\delta_2(A_2-\Delta S)\bigl[8k S+(8\delta_1^2+\delta_2^2)Q\bigr]}{32k S(3\delta_1^2-4k)+(4\delta_1^2\delta_2^2+\delta_2^4)Q+8\delta_2^2 k(5-\theta)} $$

$$ P_M = \frac{N_6+(A_1-C_n)\delta_1\delta_2 k-4\delta_2^2 k(A_2-9\Delta)-N_7-16\Delta k\theta(3\delta_1^2-4k)}{-32k S(3\delta_1^2-4k)-(4\delta_1^2\delta_2^2+\delta_2^4)Q-8\delta_2^2 k(5-\theta)} $$

$$ P_{RT} = \frac{N_8+16k A_2(3-\theta)(4k-3\delta_1^2)+N_9+2\Delta Q(\delta_2^4-5\delta_1^2\delta_2^2-32k^2 S)}{-128k S(3\delta_1^2-4k)-4(4\delta_1^2\delta_2^2+\delta_2^4)Q-32\delta_2^2 k(5-\theta)} $$

6. Model Properties and Proposition Analysis

Proposition 4. In all three hybrid recycling modes, collection prices and traceability levels are negatively correlated for both coalition channels and independent third-party channels.

In Model MR–T:

$$ \frac{\partial P_T}{\partial \gamma_{MR}} = \frac{-2k\delta_2}{8k-\delta_2^2}<0, \qquad \frac{\partial P_T}{\partial \gamma_T} = -\frac{\delta_2}{2}<0 $$

From the perspective of consumer responses, traceability levels serve as a differentiating attribute. Rational, environmentally-conscious consumers who value verifiability are willing to accept higher prices when traceability information is rich. But the extra premium consumers accept reaches a ceiling, whereas increased traceability requires costly investment in information architecture and auditing. Under the constraint of positive margins, enterprises that invest in traceability must be cautious not to reduce collection-price competitiveness. Consequently, increases in traceability levels are accompanied by downward pressure on collection prices, keeping total recovery volume economically viable.

Proposition 5. In Mode MR–T, when the channel competition coefficient is comparatively small (below a threshold), remanufacturing cost savings and the coalition’s traceability level are positively correlated:

$$ \frac{\partial \gamma_{MR}}{\partial \Delta} = \frac{-4\theta^2+2S^2\delta_1\delta_2-4\delta_2 k(2\theta-3)}{B+\delta_2^4(1+\theta)+8\delta_2^2 k(5-\theta)}>0 \quad \text{for } 1<\theta<1.5 $$

This arises because milder competitive pressure permits enterprises to channel cost savings from remanufacturing into traceability investments as strategic differentiation. When competition becomes excessive, however, the need to keep short-term product prices attractive drains the surplus available for traceability enhancement.

Proposition 6. In Mode MR–T, the transfer price \( b \) is negatively related to the cross-channel competition coefficient. As recycling competition intensifies among collectors, the EV battery producer—who is the sole downstream recipient for remanufacturing—gains bargaining dominance. It can lower the transfer price to squeeze collection costs, and the fragmented collectors accept the reduction because alternative buyers are essentially unavailable. Higher \( \theta \) therefore increases the leader’s monopsony power over secondary EV battery material flows.

Proposition 7. The coalition traceability level and the channel competition coefficient are positively related in Mode MR–T:

$$ \frac{\partial \gamma_{MR}}{\partial \theta} = \frac{(A_1-C_n)f_1+f_2-8k\theta\Delta}{B+\delta_2^4 S+8\delta_2^2 k(5-\theta)} – \frac{\bigl[(A_1-C_n)M_1+M_5+f_4\bigr]f_3}{{\bigl[B+\delta_2^4 S+8\delta_2^2 k(5-\theta)\bigr]}^2} >0 $$

as numerically validated under the admissible parameter ranges. The intuition is that under intensifying horizontal challenge from the external third-party recycler, the leading coalition cannot fight a pure price war without eroding profits. Instead, it raises EV battery traceability levels, creating a verifiability barrier that its competitor finds more difficult to replicate. This strategy positions the leader’s outlets as trustworthy and ESG-compliant in consumers’ eyes.

Proposition 8. In Mode MR–T, the coalition’s payoff-dominant symmetric equilibrium collection price is below the third-party recycler’s:

$$ P_T – P_{MR} > 0 $$

The coalition bears fixed traceability system costs and conducts recycling under the producer-responsibility obligation. Their greater efficiency and stronger market dominance allow them to sustain lower buy-back prices. Followers, lacking direct control over downstream remanufacturing valuation, must post a premium collection price to attract sufficient EV battery return volume.

Proposition 9. When channel competition intensity is high, the coalition traceability level exceeds that of the independent recycler; when competition intensity is low, the independent follower provides more traceability. For mode MR–T:

$$ \gamma_{MR} – \gamma_T = \frac{(-2\delta_2^3+32\delta_2 k)[A_2(1+\theta)+\Delta S]+S(A_1-C_n)(64\delta_1 k-12\delta_1\delta_2^2)+16\delta_1^2\delta_2[A_1 S+\Delta(S+\theta^2)]}{2B+2\delta_2^4(1+\theta)+16\delta_2^2 k(5-\theta)} $$

For \( 1<\theta<5 \), the numerator and denominator are both positive and the leading coalition has the higher traceability level. When \( \theta<1 \), the sign reverses. The leader can exploit intrinsic superiority in EV battery information architecture, since it created and controls the primary traceability records.

Similar correlations hold for Mode MT–R and Mode M–RT: retail price increases with traceability level since consumers value transparent provenance; collection prices decline with traceability due to cost pass-through; coalition collection prices are lower than those of independent collectors; and the chain-leader traceability level is higher than the follower’s when competition is intense.

Proposition 10. In Mode MT–R retail price and traceability levels co-move positively:

$$ \frac{\partial P}{\partial \gamma_{MT}} = \frac{8k\delta_1}{8k-\delta_2^2-4\delta_1^2} > 0 $$

Consumers of NEVs with environmentally preferable, fully traceable EV battery provenance exhibit higher willingness to pay for traceable products. High levels of traceability signal quality and security, while the system’s cost structure contributes to a price premium reflected in the retail market.

Proposition 11. In Mode MT–R and Mode M–RT, analogous channel price comparisons hold:

$$ P_{MT} < P_R \quad\text{and}\quad P_M < P_{RT} $$

The endogenous decision rights for system architecture and remanufacturing margins let the chain leader exercise restraint in collection pricing, because it also captures remanufacturing savings downstream.

7. Numerical Analysis

7.1 Parameter Calibration

Based on published industry reports, financial statements, and modeling precedents in closed-loop supply chain and traceability research, I calibrate the base parameter set as shown in Table 2.

Table 2: Baseline parameter calibration for numerical analysis
Parameter Value Source/Rationale
\( A_1 \) 350,000 units EV battery demand scale from published market research
\( A_2 \) 10,000 units Voluntary consumer returns, baseline calibration
\( C_n \) 28,590 CNY Average Li-ion battery pack production cost reported by major manufacturer
\( C_r \) 14,000 CNY About 50 percent saving through metals recovery and closed-loop materials
\( K \) 100,000 CNY Annual traceability cloud platform subscription fee
\( k \) 10 Baseline traceability elasticity coefficient
\( \delta_1 \) 0.4 Consumer sensitivity to traceability in demand
\( \delta_2 \) 0.4 Collection volume sensitivity to traceability
\( \theta \) 0.46 Baseline channel competition coefficient

7.2 Sensitivity Analysis on Traceability Cost Coefficient \( k \)

I performed sensitivity analysis on the traceability cost coefficient \( k \) at levels \(\{5,10,15,20,25\}\); results are shown in Table 3.

Table 3: Sensitivity analysis with respect to traceability cost coefficient \( k \)
k \( k=5 \) \( k=10 \) \( k=15 \) \( k=20 \) \( k=25 \)
\( Q_0 \) (MR–T) 7407 7465 7483 7493 7498
\( Q_0 \) (MT–R) 15431 15346 15319 15306 15298
\( Q_0 \) (M–RT) 7401 7462 7482 7492 7497
\( \Pi_i \) (MR–T) 269,761,589 267,955,380 267,368,450 267,077,745 266,904,196
\( \Pi_i \) (MT–R) 269,769,198 267,805,742 267,179,357 266,871,217 266,687,924
\( \Pi_i \) (M–RT) 269,993,633 268,055,521 267,431,861 267,124,072 266,940,671
\( \gamma \) (MR–T) 1104 542 359 269 215
\( \gamma \) (MT–R) 1128 548 362 270 215
\( \gamma \) (M–RT) 1128 548 362 270 215

These results confirm that the traceability cost coefficient exerts a globally suppressing effect on the economic performance, collection efficiency, and traceability capability of the supply chain. Higher \( k \) discourages supply-chain members from investing in information transparency. Moreover, when \( k \) increases—most noticeably between 5 and 15—profit decreases nonlinearly: the initial damage is large, but thereafter the curve flattens because participants adapt their operational processes to partially absorb rising costs.

7.3 Sensitivity Analysis on Consumer Traceability Preferences

I investigated sensitivity to the demand-side traceability preference coefficient \( \delta_1 \), and the collection-side sensitivity coefficient \( \delta_2 \), respectively. Tables 4 and 5 present the results.

Table 4: Sensitivity analysis with respect to \( \delta_1 \)
\( \delta_1 \) 0.3 0.4 0.5 0.6 0.7
\( Q_0 \) (MR–T) 6875 6872 6869 6865 6861
\( Q_0 \) (MT–R) 15346 15350 15356 15362 15370
\( Q_0 \) (M–RT) 6872 6867 6862 6856 6849
\( \Pi_i \) (MR–T) 270,529,989 270,675,806 270,833,557 271,004,576 271,190,423
\( \Pi_i \) (MT–R) 270,406,097 270,614,286 270,847,343 271,108,966 271,403,745
\( \Pi_i \) (M–RT) 270,651,342 270,850,608 271,070,331 271,313,513 271,583,819
Table 5: Sensitivity analysis with respect to \( \delta_2 \)
\( \delta_2 \) 0.3 0.4 0.5 0.6 0.7
\( Q_0 \) (MR–T) 6869 6774 6652 6499 6314
\( Q_0 \) (MT–R) 15356 15481 15644 15848 16094
\( Q_0 \) (M–RT) 6862 6764 6636 6478 6285
\( \Pi_i \) (MR–T) 270,833,557 273,419,214 276,759,430 280,909,387 285,940,733
\( \Pi_i \) (MT–R) 270,847,343 273,564,693 277,063,317 281,401,850 286,656,486
\( \Pi_i \) (M–RT) 271,070,331 273,754,245 277,206,840 281,485,317 286,663,944

Across all hybrid recycling modes, higher consumer traceability preferences expand EV battery recycling volume and raise supply-chain profit. This is because an environmentally-sensitive customer segment actively seeks verifiable recovery pathways; when traceability is visible and management is auditable, customer trust in the disposal/revitalization process grows, which motivates participation. The effect is practically linear over the considered interval. In particular, collection quantity responds strongly to increases in the collection-side parameter \( \delta_2 \), confirming that EV battery consumers respond to transparent recycling information; the feedback mechanism runs from consumer preference to enhanced traceability investment to higher return volume, and finally to improved profitability.

7.4 Sensitivity Analysis on Channel Competition Coefficient \( \theta \)

I varied the cross-channel competition coefficient across the feasible range \( \theta \in \{0.2,0.3,0.4,0.5,0.6\} \). The results are reported in Table 6.

Table 6: Sensitivity analysis with respect to \( \theta \)
\( \theta \) 0.2 0.3 0.4 0.5 0.6
\( Q_0 \) (MR–T) 9766 13207 15502 4443 negative
\( Q_0 \) (MT–R) 15797 16479 17095 15049 14240
\( Q_0 \) (M–RT) 9761 13204 15500 4435 negative
\( \Pi_i \) (MR–T) 256,414,585 259,405,898 265,243,618 275,800,952 294,846,285
\( \Pi_i \) (MT–R) 256,314,968 259,341,043 265,223,340 275,841,905 294,980,015
\( \Pi_i \) (M–RT) 256,563,089 259,581,782 265,454,197 276,058,423 295,173,237

Collection volumes rise then fall with respect to \( \theta \), while profit shows a steadily increasing trajectory across the tested domain. This non-linear pattern indicates that moderate cross-channel rivalry energizes the market: recyclers sharpen service levels, increase convenience for the EV battery owners, and expand outreach. Excessive competition, however, produces diminishing returns and loss of coordination. In very competitive settings, the distinction among hybrid recycling modes narrows considerably. In extreme situations with \( \theta>0.55 \), some modes generate negative collection flow for the coalition mode, implying market saturation and channel cannibalization. A practically important reminder is that inter-channel rivalry cannot replace a well-governed reverse-logistics architecture.

8. Recycling Mode Comparison under Alternative Objectives

8.1 Profit-Maximization Comparison

Table 7 compares total supply chain profit under the three modes over a range of \( k \).

Table 7: Supply chain profit comparison under different \( k \) values
range of k Profit ranking Optimal mode
\( 1<k\le 2.2="" \)

MT–R > M–RT > MR–T Cell manufacturer–third-party joint recycling
\( 2.2<k\le 5="" \)

M–RT > MT–R > MR–T Vehicle manufacturer–third-party joint recycling
\( k>5 \) M–RT > MR–T > MT–R Vehicle manufacturer–third-party joint recycling

At low values of the traceability cost coefficient, the cell manufacturer–third-party alliance is the most profitable. In this regime, the third-party enterprise offers economies of scale and connectivity into diverse collection networks; these synergies are not yet eaten up by traceability expenses. As \( k \) increases, the chain leader invests less in vertical traceability control and delegates more end-customer EV battery collection to the vehicle manufacturer–third-party coalition. Since the cell manufacturer still sells its wholesale EV batteries to that coalition at arm’s length and determines the transfer price, it avoids heavy cost-bearing for its own collection network. At high \( k \), the cell manufacturer–third-party alliance loses the cost advantage because third-party collection now requires heavier traceability coordination, while the vehicle manufacturer is uniquely positioned to deliver collection through the sales and service network at lower additional cost; consequently the M–RT mode dominates.

8.2 Collection-Volume Maximum Comparison

For recovering maximum EV battery volume, the mode MT–R outperforms both competing modes throughout the feasible parameter domain. This is demonstrated in Table 8 for selected \( k \) values.

Table 8: Collection volume comparison under different \( k \) values
\( k \) MR–T MT–R M–RT
5 7407 15431 7401
10 7465 15346 7462
15 7483 15319 7482
20 7493 15306 7492
25 7498 15298 7497

The interaction between the EV battery manufacturer’s brand credibility and its closed-loop stewardship claims, combined with the third-party recycler’s logistical coverage and handling competence, produces the highest return rate of retired EV battery. The public trusts the producer’s take-back obligation, while the logistics and sorting efficiency of a professional recycler raises the convenience for the consumer. In contrast, vehicle manufacturer-centric coalitions are hampered by narrower returns acceptance points, since their retail networks remain primarily optimized for new vehicle sales rather than battery recovery. Thus, the production-linked third-party alliance most successfully aligns the environmental, informational and convenience attributes in a single recovery route, making the MT–R mode the best from the standpoint of circularity.

9. Discussion and Managerial Implications

This analysis arrives at four management recommendations. First, EV battery producers should strengthen the integration of traceability technology with market-enhancing mechanisms to amplify the demand-side effect of traceability on EV battery recovery. By chaining provenance and status information on the ledger, consumers can access raw-material information, supplier qualifications, manufacturing history, health status, and any maintenance actions at any time. Providing an “all-you-can-know” traceability experience expands new NEV demand while simultaneously extending reverse collection functions such as life-remaining alerts and pick-up scheduling, which fortifies consumer engagement in returning retired batteries and promotes “everything recoverable when due”. High volumes in turn lower unit traceability expenditure and strengthen economies of scale in dismantling.

Second, vehicle manufacturers should take part actively in traceability-driven recycling modes. They should build comprehensive provenance records together with cell producers and operate the end-of-life collection. Since vehicle makers interface directly with the consumer, they can deploy usable customer-education initiatives: official website content, social campaigns, mobile application notifications, showroom displays, and integrated dealer programs. Member reward structures that accumulate points for owning traceable vehicles and participating in recycling campaigns increase the return rate of retired EV battery while enlarging their customer base of sustainability-oriented consumers.

Third, third-party recyclers should unlock their natural advantage of multi-brand adaptability and professional pretreatments. By cooperating with several upstream EV battery producers and vehicle manufacturers, they can coordinate their proprietary collection points with joint recycling hubs and shared reverse logistics, while their participation in tracing records improves process transparency for downstream governments and consumers. Third-party recyclers that actively upload EV battery information into collaborative ledgers will improve reputation, attract upstream contracts, and sustain economies of scale.

Fourth, policy makers should anchor governance on traceability: complete recycling standards and rules, establish responsibility for information disclosure, decrease the cost burden on early-stage blockchain adoption through tax preferences, improve the national traceability platform, and promote public communication campaigns to build traceability preference. Because the traceability cost coefficient is one of the most potent negative levers of supply-chain EV battery recovery volume, government-sponsored infrastructure—for instance, providing unified interfaces, open standards, and subsidized platform subscriptions—can reduce the marginal cost of credible data recording for all participants and shift recycling mode profitability comparisons accordingly.

10. Conclusions and Limitations

In summary, this research develops profit-oriented equilibrium decisions for a traceable EV battery closed-loop supply chain under three hybrid recycling configurations. The main conclusions include the following: (1) the traceability cost coefficient is a core regulating variable in recycling-mode optimization, with a globally negative effect on economic performance, collection volume, and achievable traceability intensity; (2) channel competition and consumer traceability preference jointly drive recovery efficiency—channel rivalry acts as an inverted-U force on collection, whereas greater traceability preference induces a positive linear feedback loop; (3) under the profit-maximization objective, the cell manufacturer–third-party recycler cooperative mode is optimal for small traceability costs, while after crossing a threshold \( k \) value, the vehicle manufacturer–third-party recycler cooperation mode becomes the most profit-efficient choice; (4) under the collection-volume-maximization objective, the alliance between the EV battery producer and the third-party recycler dominates over all tested traceability cost levels—making it the best choice for maximizing recovered volumes of retired EV battery.

This thesis has three limitations. First, data acquisition constraints forced the numerical segment to rely on public financial statements and industry reports; no exclusive enterprise-level operational data were available for a proprietary case validation. Second, the model restricts the closed-loop network to three entities—the cell maker, vehicle maker, and the third-party recycler—whereas actual EV battery ecosystems are embedded in a richer environment that contains echelon utilizers, governmental agents, raw-material suppliers and logistics providers. Third, factors not considered include behavioral biases, the firms’ fairness preferences, heterogenous battery chemistries, and inter-temporal dynamics. Subsequent investigations may thus extend the actor set, incorporate information asymmetry and fairness parameters, integrate carbon-trading constraints, and empirically verify sensitivity magnitudes with company-level data from EV battery producers, vehicle manufacturers, and independent recyclers.

Scroll to Top