Forecasting Lithium Demand for Electric Ship Batteries in China’s Inland Shipping Under Decarbonization Scenarios

Abstract

As China advances toward its 2060 carbon neutrality goal, the electrification of inland waterway shipping has emerged as a strategic pathway for reducing emissions. This study constructs a 2025–2060 dynamic material flow analysis framework that integrates three core dimensions: (1) all-electric ships (AES) diffusion, estimated via a GDP-elasticity model and carbon emission accounting; (2) battery technology evolution, including lithium iron phosphate and solid-state batteries; and (3) recycling system improvements, incorporating direct recycling, cascade utilization, and metallurgical processes. The research sets up three AES penetration scenarios, two battery technologies, and three recycling technology improvement scenarios, resulting in seven combination scenarios for analysis. Through multi-scenario simulations, it reveals synergistic pathways for resource security and decarbonization goals. Key findings include that to meet carbon reduction targets, AES penetration in inland shipping must reach 25.36% by 2060, corresponding to cumulative new ship constructions of 51.5–79.9k units, with total lithium demand ranging from 49.1–95.9 kt, and recycling potential reaching 5.4–25.2 kt. Results also reveal that under current allocation assumptions, the AES sector may face lithium shortages between 2047 and 2057 unless recycling rates improve or electrification pathways are optimized. The work innovatively links battery tech dynamics and recycling optimization for China’s inland shipping and provides actionable guidance for balancing decarbonization and lithium resource security.

1. Introduction

Against the backdrop of global efforts to address climate change, the decarbonization of the maritime sector has become a vital pathway toward sustainable development. The International Maritime Organization (IMO) has set a target of achieving net-zero greenhouse gas emissions for international shipping by around 2050 [1], while the inland shipping system, as a crucial corridor extending the hinterland of seaports, directly impacts the emission reduction efficiency of coastal port clusters. In China, this transition has gained significant momentum with the introduction of the national “dual carbon” goals—peaking carbon emissions and achieving carbon neutrality [2]. As a result, the inland waterway transport sector is accelerating its shift toward clean energy alternatives. Among various technological pathways, all-electric ships (AESs) [3] have attracted considerable attention as a key low-carbon solution, owing to their zero-emission operation, low noise levels, and high energy efficiency [4].

However, the large-scale deployment of AESs also introduces a series of challenges, particularly regarding the sustainability of battery supply chains and the security of lithium (Li) resources. Lithium, as a critical raw material in electric propulsion systems, is increasingly in demand across multiple sectors—including electric vehicles and stationary energy storage—raising concerns about future resource competition [5,6,7]. Moreover, uncertainties in battery technology evolution and the pace of recycling infrastructure development further exacerbate concerns over the long-term availability and circularity of lithium—including in maritime applications. While ocean-going vessels currently rely more on alternative low-carbon fuels (e.g., ammonia, methanol) due to long-haul requirements, short-haul coastal shipping and port operation vessels are exploring electrification, creating potential overlaps in lithium demand [8,9]. Given these challenges, a comprehensive understanding of lithium flows, from initial extraction to end-of-life treatment, is necessary to ensure that the electrification of inland shipping remains both environmentally and resource sustainable.

Existing studies have increasingly focused on the material implications of energy transition technologies, particularly in relation to the rapid adoption of electric vehicles (EVs) and lithium-ion batteries (LIBs) [10,11,12,13]. Dynamic Material Flow Analysis (MFA) has been widely used to project long-term trends in critical material demand, incorporating product lifecycles, battery chemistry transitions, and recycling strategies [14,15,16]. For instance, Xu et al [10] developed a comprehensive framework to forecast the future material demand for lithium-based batteries in the automotive sector. Their study integrated vehicle stock modeling with battery chemistry evolution, highlighting that lithium demand will increase substantially even under moderate EV adoption scenarios. Maisel et al [11] further expanded on this by estimating both future raw material requirements and recycling potential for lithium-ion batteries in electric vehicles, emphasizing the critical role of closed-loop recycling systems in mitigating raw material shortages. Lu et al [12] focused specifically on China’s transport decarbonization pathways and projected the demand for key battery minerals under a carbon neutrality scenario. Their results showed that achieving deep decarbonization in the road transport sector would significantly strain domestic critical mineral supply chains, particularly for lithium, cobalt, and nickel. Shafique et al [15] conducted a dynamic material flow analysis for end-of-life LIBs in China and the U.S., identifying substantial gaps in recycling rates and the temporal mismatch between battery retirement and material recovery, especially when second-life applications delay recycling. Dong et al [16] introduced a novel MFA model that incorporates vehicle segment heterogeneity into projections of critical material demand and recycling in China’s passenger EV sector. Their findings revealed that vehicle type has a significant impact on the timing and magnitude of lithium demand and potential recovery, suggesting that policy and recycling strategies must consider fleet composition and replacement cycles to be effective.

Although these studies have provided valuable insights into lithium dynamics, they are primarily focused on road transport or global-level projections. However, there is currently a lack of systematic research to predict the Li resources required for the decarbonization of the inland shipping sector, which presents unique patterns of electrification and resource use. In particular, research is lacking on how the interplay between carbon reduction targets, battery technology development, and recycling improvements affects lithium demand in AESs. To address these gaps, this study aims to construct a multidimensional scenario analysis framework to quantify the demand for lithium resources and recycling potential of AESs in China under the 2060 carbon neutrality goal. The core research topics include the following: (1) requirements for the penetration rate of AES in inland waterway freight under China’s decarbonization goals; (2) the scale and temporal distribution characteristics of lithium resource demand for inland all-electric ships under different decarbonization pathways’ and (3) the synergistic impact of battery technology evolution and advancements in recycling on the balance of lithium supply and demand.

The innovation of the paper lies in providing the first in-depth quantitative analysis of lithium resource demand specific to China’s inland electric shipping sector, addressing a critical knowledge gap in low-carbon transition research, and offering actionable guidance for policymakers on how to manage lithium supply risks through coordinated development of electrification roadmaps, recycling infrastructure, and battery technology strategies. Also, as inland AESs share key characteristics with short-haul marine electric vessels, such as fixed routes and frequent docking, which facilitate battery charging and recycling [17], the findings help to support not only inland shipping planning but also provide a reference for maritime stakeholders exploring electric propulsion in coastal and port operations, contributing to a coordinated decarbonization strategy across the broader maritime transport network.

The rest of the paper is organized as follows: Section 2 elaborates on the research methodology; Section 3 outlines the dimensions of scenario design and key parameter assumptions; Section 4 presents the results of new-built volumes for AESs annually, battery demand forecasts, Li resource demand predictions and recycling potential assessments; and Section 5 summarizes the core conclusions and discusses the policy implications, limitations, and directions for future research of the findings.

2. Methods

2.1. Research Framework

This study establishes a comprehensive modeling framework to estimate the lithium resource demand for inland waterway AESs in China under the 2060 carbon neutrality target. An overview of the framework is illustrated in Figure 1. The analysis begins with estimating the market penetration of AESs in the target year, based on a carbon emission accounting model. This estimation forms the basis for determining the expected scale of new vessel construction. Subsequently, a dynamic MFA model is developed to project lithium demand and recovery potential across multiple future scenarios.

This research specifically focuses on China’s inland cargo fleet. Due to the nature of inland shipping, these vessels are predominantly powered by full-electric propulsion systems [18]; thus, hybrid vessels are not considered in this study. Lithium is the key metal addressed, primarily due to regulatory constraints: ternary lithium batteries are not permitted for maritime use in China. As a result, newly constructed inland electric vessels currently adopt lithium iron phosphate (LFP) batteries [19]. Even with anticipated advances in battery technologies, such as the adoption of solid-state batteries (SSBs), lithium remains the primary metallic component, while critical metals like cobalt and nickel are largely excluded [20,21]. This study incorporates historical data from 2020 to 2024 which can be found in Supplementary Materials, and conducts future projections for the period from 2025 to 2060.

2.2. Calculation of the Penetration of AESs

In the first part of the model, we estimate the penetration rate that AES should achieve in the fleet by 2060 under the emission reduction target. This estimation is based on projected freight turnover, fuel consumption rates, and the carbon intensity of different fuel types. A profile of China’s inland waterway cargo fleet is set up, encompassing two propulsion modes (fuel-based and electric) and seven fuel types. The fuel-based vessels operate on diesel, LNG, ammonia, methanol, hydrogen, and biofuels, while electric vessels are powered entirely by electricity. As discussed in Section 1, all electric vessels in this study are assumed to utilize lithium-based batteries.

Given that AESs emit zero carbon during operation [22], all carbon emissions in inland waterway transport are attributed to fuel-powered ships. Regarding decarbonization targets in the maritime sector, the International Maritime Organization [1] announced in July 2023 that international shipping aims to achieve net-zero greenhouse gas emissions by or around 2050. In alignment with the global climate agenda, China has pledged to reach carbon neutrality before 2060, supporting the Paris Agreement’s goal of limiting global warming to well below 2 °C, preferably to 1.5 °C. However, for China’s inland waterway sector, there is currently no unified national target or standardized planning mechanism for carbon emissions reduction. Therefore, based on relevant existing studies [23,24], this study sets a sectoral carbon intensity reduction target: reducing the carbon emission intensity of inland vessels from 5.832 gCO₂/t·km in 2025 to 2.435 gCO₂/t·km by 2060.

Consequently, as an initial step, we assume (1) a standardized vessel type across the fleet, based on the Yangtze River standard ship configuration, and (2) uniform annual mileage across all vessels.

The total annual CO₂ emissions in 2060 are calculated as the product of freight turnover, fuel consumption rates, and the carbon emission intensity of each fuel type. Notably, ship energy intensity is defined as the product of fuel consumption and carbon intensity. Given that all actual CO₂ emissions come from fuel-powered ships, where its proportion within the fleet is 1 minus the AES penetration rate, the penetration rate of electric vessels is then derived by comparing the allowable emissions under the decarbonization target to the projected emissions of a fleet partially composed of AESs.

$$\begin{array}{c}{\mathrm{CO}}_2={\displaystyle \sum _{\mathrm{i}}}{\mathrm{IWFT}}_{\mathrm{i}}\times {\mathrm{FCR}}_{\mathrm{i}}\times {\mathrm{FCEI}}_{\mathrm{i}}\end{array}$$

(1)

$${\mathrm{EI}}_{\mathrm{i}}={\mathrm{FCR}}_{\mathrm{i}}\times {\mathrm{FCEI}}_{\mathrm{i}}$$

(2)

$$\mathrm{p}=1-{\mathrm{TCO}}_2/({\displaystyle \sum _{\mathrm{i}}}{\mathrm{EI}}_{\mathrm{i}}\times {\mathrm{IWFT}}_{\mathrm{i}})$$

(3)

where ${\mathrm{IWFT}}_{\mathrm{i}}$ is the freight turnover using fuel type i, ${\mathrm{FCR}}_{\mathrm{i}}$ is the fuel consumption rate of ships using fuel type i, ${\mathrm{FCEI}}_{\mathrm{i}}$ is the CO₂ emissions per unit of fuel type i consumed, ${\mathrm{EI}}_{\mathrm{i}}$ is the energy intensity of ships using fuel type i, TCO₂ is the target total CO₂ emissions in 2060, and p is the penetration rate of AES.

2.3. Construction of Dynamic MFA Model

In alignment with China’s 2060 decarbonization target, we developed a dynamic MFA model to assess the full life-cycle of lithium use in marine battery systems. This model tracks lithium flows from battery manufacturing to operation, retirement, and recycling. As illustrated in Figure 1, the framework consists of three key layers: the AES fleet layer, the battery layer, and the Li resource layer. The model integrates three primary dimensions: the penetration rate of AES in the inland fleet, technological and market evolution of battery systems, and advancements in recycling practices. This approach enables us to quantify both the future lithium demand for AES deployment and the potential recovery of lithium through end-of-life battery treatment.

To provide a clearer visualization of Li resource flows within marine battery systems, this study outlines the full life cycle of Li in AES batteries, as illustrated in Figure 2. The life cycle consists of three main stages: (1) battery production, (2) battery operation, and (3) battery retirement and recycling.

In the production stage, lithium is refined into compounds such as lithium carbonate and then processed with other materials to manufacture battery chemistries, including LFP and SSBs. These batteries are installed in new-built AESs, where they remain in use until degrading to below 80% of their original capacity—the threshold at which they are deemed unsuitable for primary propulsion and transition to the retirement and recycling stage [25,26]. During this stage, spent battery packs from decommissioned AESs are collected and disassembled into individual cells for further treatment, with end-of-life management primarily following three pathways [27]: direct recycling, which involves dismantling cells and reassembling them into new ones by supplementing lithium carbonate [28,29]; cascade utilization, a stepwise repurposing strategy where batteries with sufficient residual State of Health (SOH) are redeployed from high-demand scenarios to less demanding applications [30,31]; and metallurgical recycling, which consists of shredding cells, recovering materials via multi-stage mechanical separation, and subsequent pyrometallurgical or hydrometallurgical processes [32,33,34]—notably, batteries from secondary applications eventually enter this phase.

2.3.1. AES Layer

The AES layer simulates the development of the inland AES fleet from 2025 to 2060. Since freight demand in a given country or region is closely correlated with the level of economic development [35,36], this study adopts an elasticity model based on China’s GDP growth to estimate the expansion of the inland shipping fleet. The calculation formulas are as follows:

$${\mathrm{GR\_IWF}}_{\mathrm{t}} ={ \mathrm{E}}_{\mathrm{t}}\times {\mathrm{GR\_GDP}}_{\mathrm{t}}$$

(4)

$${\mathrm{IWF}}_{\mathrm{t}}={\mathrm{IWF}}_{\mathrm{t}-1}\times (1+{\mathrm{GR\_IWF}}_{\mathrm{t}})$$

(5)

where ${\mathrm{GR\_IWF}}_{\mathrm{t}}$ is the growth rate of inland waterway freight volume in year t, ${\mathrm{GR\_GDP}}_{\mathrm{t}}$ is the GDP growth rate in year t, ${\mathrm{E}}_{\mathrm{t}}$ is the Elasticity coefficient between freight demand and GDP in year t, and ${\mathrm{IWF}}_{\mathrm{t}}$ is the inland waterway freight volume in year t.

The size of the river fleet is related to freight demand, average deadweight tons, and a discount factor, the value of which—based on historical data—has been estimated to be about 29:

$${\mathrm{S}}_{\mathrm{t}} = {\mathrm{IWF}}_{\mathrm{t}}/\mathrm{A}{\mathrm{DWT}}_{\mathrm{t}}/\mathrm{k}$$

(6)

where ${\mathrm{S}}_{\mathrm{t}}$ denotes the size of the river fleet in year t; $\mathrm{A}{\mathrm{DWT}}_{\mathrm{t}}$ denotes the average deadweight tons in year t; and $\mathrm{k}$ denotes the discount factor between freight demand and total deadweight tons of inland ships.

The number of AESs in the inland fleet is determined by multiplying the total fleet size by the AES penetration rate. Each year, the number of new AESs is calculated by subtracting the previous year’s number from the current year’s number, and then adding the number of vessels decommissioned. The average lifespan of vessels is generally assumed to range between 20 and 25 years, with container ships typically having a shorter service life before decommissioning [37]. The formula for calculating the number of new electric vessels to be built each year is as follows:

$${\mathrm{ES}}_{\mathrm{t}} = {\mathrm{S}}_{\mathrm{t}}\times {\mathrm{p}}_{\mathrm{t}}$$

(7)

$${\mathrm{NES}}_{\mathrm{t}}={ \mathrm{ES}}_{\mathrm{t}}-{\mathrm{ES}}_{\mathrm{t}-1}+{\mathrm{D}}_{\mathrm{t}}$$

(8)

where ${\mathrm{ES}}_{\mathrm{t}}$ denotes the number of AESs in the fleet in year t, ${\mathrm{S}}_{\mathrm{t}}$ denotes the total number of the inland fleet in year t, ${\mathrm{p}}_{\mathrm{t}}$ denotes the penetration of AES in the inland waterway fleet in year t; ${\mathrm{NES}}_{\mathrm{t}}$ denotes the number of newly built electric vessels in year t; and ${\mathrm{D}}_{\mathrm{t}}$ denotes the number of electric ships scrapped in year t.

2.3.2. Battery Layer

The demand for marine batteries is primarily determined by the number of newly built AESs each year. By multiplying the annual number of new AESs by the average battery capacity per vessel, we obtain the inflow of batteries into the operation stage. Considering that the service life of traction batteries is generally projected to range between 15 and 20 years [38,39], and that battery swapping is widely adopted in China’s inland electric shipping sector, this study assumes that batteries are not replaced during a ship’s operational lifetime. Thus, the lifespan of marine batteries is set equal to the ship’s decommissioning age—20 years [10,16]. Consequently, the outflow of batteries from the operation stage is determined by the annual number of vessel retirements.

In parallel, based on the current battery market composition and anticipated advancements in electrochemical technologies, this study considers two types of battery chemistries: LFP, which is currently used in all AESs and SSBs (i.e., Li-S, Li-Air and All-solid-state batteries (ASSBs)) [21,40], which may enter the market in the future. It is important to note that SSBs are still under laboratory development and are not yet ready for large-scale commercial deployment. Therefore, this study incorporates multiple scenarios reflecting varying levels of technological advancement and uses their average performance to represent the overall battery technology trajectory.

$${\mathit{MBIF}}_{t,b}={\mathit{NES}}_t\times {\mathit{MBC}}_{t,b}\times {\mathit{Pmbi}}_{t,b}$$

(9)

$${\mathit{MBOF}}_{t,b}={\displaystyle \sum _T}{\mathit{MBIF}}_{t-T,b}$$

(10)

$${\mathit{MBIS}}_{t,b}={\sum }_{n=0}^{20}\left({\mathit{MBIF}}_{t-n,b}-{\mathit{MBOF}}_{t-n,b}\right)$$

(11)

where ${\mathit{MBIF}}_{t,b}$ is the maritime battery inflow volume of b-type batteries in year t, ${\mathit{MBOF}}_{t,b}$ is the maritime battery outflow volume of b-type batteries in year t, ${\mathit{MBIS}}_{t,b}$ is the maritime battery retained in the use phase of b-type batteries in year t, ${\mathit{MBC}}_{t,b}$ is the capacity of b-type batteries in year t, and ${\mathit{Pmbi}}_{t,b}$ is the market share of b-type batteries in year t.

2.3.3. Lithium Layer

Based on the lithium demand parameters of different battery types, this section further estimates the lithium resource inflow, outflow, and stock within the use phase.

$${\mathit{RIF}}_t={\displaystyle \sum _b}{\mathit{MBIF}}_{t,b}\ast {\mathit{RDP}}_{t,b}$$

(12)

$${\mathit{ROF}}_t={\displaystyle \sum _b}{\mathit{MBOF}}_{t,b}\ast {\mathit{RDP}}_{t,b}$$

(13)

$${\mathit{RIS}}_t={\displaystyle \sum _b}{\mathit{MBIS}}_{t,b}\ast {\mathit{RDP}}_{t,b}$$

(14)

where ${\mathit{RIF}}_t$ is the Li resource inflow in year t, ${\mathit{ROF}}_t$ is the Li resource outflow in year t, ${\mathit{RIS}}_t$ is the Li resource retained in the use phase in year t, and ${\mathit{RDP}}_{t,b}$ is the amount of Li resources required to generate 1 kWh of electricity in battery type b in year t.

Lithium recovery originates from two sources: (1) batteries that are retired and directly collected for recycling, and (2) batteries from scrapped vessels that were not recycled in earlier years but are recovered later in conjunction with ship dismantling. The annual primary lithium demand is calculated as the difference between the lithium required for batteries installed in newly built AESs and the amount of lithium recovered from decommissioned batteries. Equations present the calculations for lithium resource recovery potential, annual primary resource demand, and cumulative demand throughout the study period.

$${\mathit{RP}}_t=\left({\mathit{ROF}}_t\times {\mathit{Rrmb}}_t+{\mathit{ROF}}_{t-T^{\prime }}\times {\mathit{Rrmb}}_{t-T^{\prime }}\times {\displaystyle \frac{{\mathit{Pcu}}_{t-T^{\prime }}}{{\mathit{Pdr}}_t+{\mathit{Pmr}}_t}}\right)\times \left({\mathit{Pdr}}_t\times {\mathit{Rdr}}_t+{\mathit{Pmr}}_t\times {\mathit{Rmr}}_t\right)$$

(15)

$${\mathit{RPD}}_t={\mathit{RIF}}_t-{\mathit{RP}}_t$$

(16)

$$\mathit{CR}={\displaystyle \sum _t}\left({\mathit{RIF}}_t-{\mathit{RP}}_t\right)$$

(17)

where ${\mathit{Rrmb}}_t$ denotes the recovery rate of maritime batteries in year t, ${\mathit{Pdr}}_t$ denotes the proportion of direct recycling in year t, ${\mathit{Pcu}}_t$ denotes the cascade utilization year t, ${\mathit{Pmr}}_t$ denotes the metallurgical recycling in year t, ${\mathit{RP}}_t$ denotes the lithium resource recovery potential in year t, ${\mathit{RPD}}_t$ denotes the primary lithium resource demand in year t, and CR denotes the cumulative lithium resource demand.

3. Scenarios and Assumptions

3.1. Scenario Settings

To deepen the study of the dynamic material flow of lithium resources in AES, this research conducts a scenario analysis across three key dimensions: (1) market penetration of AESs, (2) battery technology and market evolution, and (3) improvements in recycling technologies. For the market penetration of AES, three scenarios are defined: E-LSD (Low-Speed diffusion of AESs), E-MSD (Medium-Speed diffusion of AESs), and E-HSD (High-Speed diffusion of AESs). Two scenarios are considered for battery technology evolution: B-LFP (continuity with the current market dominated by LFP batteries), and B-SSB (transition towards high-consumption, high-tech batteries, gradually transitioning to an SSB-dominated market). Considering China’s requirements for efficient recycling of critical resources and the continuous evolution of recycling technologies [41], three scenarios for improvements in recycling technology are also defined: R-LSI (progress at a low speed under the current recycling paradigm), R-MSI (steady improvement with an increase in direct recycling and cascade utilization), and R-HSI (breakthrough-driven recycling enhancement scenario, e.g., cathode regeneration without full crushing). Detailed assumptions and parameter settings for each scenario are provided in subsequent sections.

Based on these combinations of dimensions, seven distinct scenarios are set, as shown in Figure 3. The BMS scenario represents the future lithium resource demand under current technological and market development trends. The MID and MAD scenarios are used to compare the upper and lower limits of future lithium resource demand, while the MIR and MAR scenarios indicate the upper and lower limits of lithium resource recycling potential. The BML and BMH scenarios are designed to compare the impacts of different recycling improvements on resource demand.

3.2. Parameters Related to Vessel

The parameters related to the vessel include two aspects: the parameters related to ship emissions and those related to the number of newly built AESs.

The ship emissions-related parameters mainly include inland river cargo turnover, the proportion of different types of fuel-powered ships, fuel consumption rates, and the carbon emission intensity of fuels. The inland river cargo turnover and the proportion of different fuel-powered ships in 2060 are set based on the Study on Medium- and Long-Term Low-Carbon Development Pathways for China’s Inland Waterway Shipping (CWTRI, 2023), and the assumptions for FCEI follow the Intergovernmental Panel on Climate Change (IPCC) Guidelines for National Greenhouse Gas Inventories [42]. The assumptions for FCR mainly refer to World Energy Outlook 2022 [43] and the study by Zincir et al. [44].

The parameters related to the number of newly built AESs mainly include GDP growth rate, elasticity coefficients, and AES market penetration. In this study, the GDP growth rate for China from 2025 to 2060 is primarily referenced from the setting of Pan [12]. The assumptions for elasticity coefficients are based on Liu’s estimation [45] and adjusted according to historical inland river cargo volume data. For AES market penetration, the study sets three scenarios for the increase in AES penetration rates from 2025 to 2060, based on the target AES penetration rate calculated in the first section of the model. Each scenario ensures the target penetration rate is achieved by 2060. In the E-LSD scenario, the AES penetration rate grows exponentially, with the growth rate accelerating over time. In the E-MSD scenario, the penetration rate increases at a steady pace, while in the E-HSD scenario, the penetration rate grows logarithmically, with a decelerating growth rate [13]. Detailed settings are provided in Supplementary Materials

3.3. Parameters Related to Battery

The parameters related to marine batteries also include two aspects: those related to battery usage and those related to battery recycling.

The battery usage-related parameters mainly include the per-ship battery capacity, the battery installation ratio, and the lithium resource demand parameters (RDP).

The per-ship battery capacity is referenced from General Energies [46]. For the average deadweight tonnage, using the PM120 inland waterway pure electric container ship as the reference vessel type, the PM120 inland pure electric container ship in the baseline scenario carries four containerized storage batteries, each with a capacity of approximately 2000 kWh. This study assumes that in the B-LFP scenario, the battery capacity per ship increases from 8000 kWh in 2025 to 9600 kWh by 2060. In the B-SSB scenario, the battery capacity increases from 8000 kWh to 11,200 kWh.

Currently, the types of batteries installed on China’s inland AES are predominantly LFP. The advantages of LFP include lower production costs due to the abundance of precursor materials, better thermal stability leading to improved safety, and a longer cycle life [28]. SSBs are still in the early development phase [47]. Specifically, Li-S and Li-Air batteries, which have been identified as promising successors to LIBs [10,21]. Although these technologies are still in the early stages of development, they face significant hurdles such as limited cycle life and safety concerns [48], Li-S batteries may offer up to twice, and Li-Air batteries up to three times, the specific energy of current LIBs. These improvements could contribute to cost reductions and extended operational ranges in electric marine applications.

This study assumes that in the B-LFP scenario, the battery market will remain dominated by LFP, with SSB entering the market in 2035 and reaching 45% market penetration in the marine battery market by 2060. In the B-SSB scenario, SSB enters the market in 2030 and surpasses LFP in market share by 2045, becoming the dominant power battery technology. By 2060, its market share will reach 70%. [10,16]. The specific settings are shown in Figure 4 and detailed data is provided in Supplementary Materials.

RDP for lithium comes from the Ministry of Industry and Information Technology (MIIT, 2023) and the study by Sun et al. [49,50]. On this basis, following the experimental setup described by Lu et al. [12], it is assumed that the RDP for LFP decreases linearly from 2020 to 85% of the 2020 level in 2030, and the RDP for SSBs decreases linearly from 2030 to 2060. The specific parameters are shown in

Table 1. RDP for LFP and SSB. Unit: kg/kWh.

Year		SSB
	LFP	Li-Sulfur	Li-Air	ASSB
2020	0.09	-	-	-
2030	0.08	0.28	0.26	0.18
2040	0.08	0.26	0.24	0.16
2050	0.08	0.24	0.22	0.14
2060	0.08	0.22	0.20	0.12

.

The battery recycling-related parameters include the battery recycling rate, the proportion of different recycling methods, and the lithium recovery rate. Based on the configuration proposed by Dong et al. [16], in this study, under the R-LSI scenario, China’s end-of-life battery recycling rate increases from 40% in 2025 to 60% by 2060, with metallurgical recycling being the predominant method. In the R-MSI scenario, the recycling rate increases from 50% to 80%, with a moderate rise in the shares of cascade utilization and direct recycling. In the R-HSI scenario, the recycling rate grows from 60% to 95%, with a significant increase in the shares of both cascade utilization and direct recycling. Specific settings are illustrated in Figure 5 and detailed data is provided in Supplementary Materials. The lifespan of batteries in cascade utilization is assumed to be five years [51], starting after their retirement from primary use in AES propulsion. Regarding lithium recovery efficiency, due to the low recovery rate of lithium in pyrometallurgical processes [32], this study assumes that hydrometallurgical recycling will be the dominant method for lithium recovery. Considering the optimization of recycling infrastructure, advances in direct recycling technologies (e.g., cathode regeneration with lithium carbonate supplementation), and policy-driven upgrading, the lithium recovery efficiency in metallurgical recycling is assumed to increase slightly from 90% in 2025 to 95% in 2060 [33], while the recovery rate through direct recycling increases from 90% to 98% [52,53].

4. Results

4.1. AES

With continued socioeconomic development and rising household income levels, the demand for inland waterway freight transport in China is expected to increase in the future. In our study, the size of the inland freight fleet continues to grow, along with a steady rise in the penetration of AESs within the fleet. Under the carbon reduction target for inland waterway freight transport, the model estimates that the penetration rate of AESs in the inland freight fleet should reach 25.36% by 2060. Figure 6 illustrates the annual number of new-built AESs and the cumulative number over the study period under three different AES penetration scenarios, based on our model and assumptions. All three scenarios achieve a growth in AES penetration from 0.05% in 2025 to 25.36% by 2060.

In the E-LSD scenario, the annual number of new AES increases slowly until 2045. To meet the 2060 penetration target, new shipbuilding surges after 2045, reaching 3.26k units by 2060. In the E-MSD scenario, the number of new AES increases steadily year by year, maintaining a relatively constant growth rate, and reaches 2.96k units in 2060. In contrast, the E-HSD scenario features rapid growth in new AES at the beginning of the period, peaking at 2.69k units in 2036. Afterward, the build rate slows, then accelerates again after 2045 to reach a second peak of 3.35k units by 2052.

Under all three scenarios, the total number of inland AESs in China increases continuously, with a marked inflection point occurring around 2045. This shift is primarily attributed to the retirement of AESs built in the early phase, which begin to exit the fleet after a 20-year service life, thereby necessitating new shipbuilding to maintain the AES penetration level. Over the entire period, the cumulative number of AES built reaches 51.47k, 60.98k, and 79.91k in the E-LSD, E-MSD, and E-HSD scenarios, respectively.

4.2. Battery

Figure 7 illustrates the projected demand for total battery capacity under different scenarios for the inland waterway freight fleet. Variations in scenario settings result in significantly different trajectories of battery demand. The total future demand for battery capacity depends on the penetration trajectory of AES, the battery capacity per vessel, and battery lifespan. In general, the demand pattern for total battery capacity closely follows the annual number of newly built AES.

In the baseline scenario, the annual AES penetration increases at a relatively stable rate. The battery market follows the current trend, where LFP batteries dominate marine applications. Li-S and Li-Air batteries are introduced after 2035, while ASSBs do not enter the market until 2050. Even by 2060, LFP remains the dominant battery type for marine use. Under this scenario, the total battery capacity demand reaches 586.2 GWh, with LFP, Li-S, Li-Air, and ASSB accounting for 67.5%, 12.0%, 15.0%, and 5.4% of the cumulative battery capacity, respectively.

In the high-demand scenario, AES penetration follows a logarithmic growth pattern. The battery market is developing rapidly, with SSB technologies making significant progress. Li-S and Li-Air batteries enter the marine battery market as early as 2030 and jointly achieve a market share of 50% by 2038. ASSBs enter the market in 2040 and reach over 30% market share by 2060. The total battery capacity demand in this scenario is 677.7 GWh, with LFP, Li-S, Li-Air, and ASSB contributing 44.2%, 19.0%, 19.1%, and 17.7% of the total battery capacity, respectively.

In the low-demand scenario, AES penetration increases slowly before 2045 and accelerates sharply in the latter half of the period to meet the 2060 decarbonization target—resulting in an overall exponential growth pattern. The battery market continues to follow current trends, with limited adoption of next-generation batteries in marine applications. The total battery capacity demand in this scenario is 402.4 GWh, with LFP, Li-S, Li-Air, and ASSB accounting for 61.1%, 13.5%, 17.3%, and 8.1%, respectively.

4.3. Lithium

The demand for lithium resources primarily depends on the battery chemistries adopted. Figure 8 presents projections of future demand for critical materials under different scenarios. Driven by the expansion of the AES market, the total demand for lithium resources is expected to grow steadily over time.

In the baseline scenarios, lithium demand increases at a relatively stable pace, reaching 2.8 kt by 2060, with a cumulative demand of 69.5 kt from 2025 to 2060. A notable acceleration in demand occurs after 2045, primarily due to the pressure on shipbuilding from fleet retirements. In the MAD and MAR scenarios, where battery chemistry shifts toward higher energy density and more lithium-intensive SSBs, lithium demand peaks at 4.2 kt in 2052 before declining slightly toward 2060. The total cumulative lithium demand in this case reaches 95.9 kt, indicating significantly higher pressure on lithium supply chains. In contrast, the MID and MIR scenarios assume continued dominance of LFP batteries, which are characterized by lower energy density but higher thermal stability and safety. Lithium demand remains relatively low before 2045, then increases rapidly to 3.9 kt by 2060. The total cumulative demand in these scenarios is approximately 49.1 kt.

Figure 9 and Figure 10 illustrate the projected recycling potential of lithium resources and the corresponding reduction in virgin lithium demand under various scenarios.

In the BMS scenario, the amount of recyclable lithium is expected to reach 1.5 kt by 2060, resulting in a total reduction of 12.4 kt in cumulative lithium demand. Under the BMH and BML scenarios, which account for varying levels of improvement in recycling technologies, the recoverable lithium by 2060 is projected to reach 1.8 kt and 1.1 kt, respectively, leading to cumulative reductions of 14.7 kt and 9.3 kt in lithium demand. Notably, the total lithium recovered in the BMH scenario is approximately 25.8% lower than in the BMS scenario. This is primarily attributed to the time-lag effect introduced by second-life battery utilization, which delays the availability of recyclable lithium.

The MAR and MIR scenarios represent the upper and lower bounds of lithium recycling potential. In the MAR scenario, cumulative recycled lithium could reach 25.2 kt, while in the MIR scenario, the lowest estimated cumulative recovery is 5.4 kt. These variations highlight the substantial uncertainty associated with future recycling potential for critical materials, largely driven by fluctuations in resource inflows and technological advancements.

The MAD and MID scenarios represent the upper and lower bounds of lithium demand after considering recycling. In the MAD scenario, the combination of high-intensity battery chemistries and low recycling efficiency leads to a rapid increase in lithium demand, peaking at 3.6 kt in 2050. This is followed by a sharp decline, significantly alleviating supply pressure after 2055. Conversely, in the MID scenario—characterized by low-intensity battery chemistries and high recycling efficiency—the initial demand for AES is relatively low, with production concentrated after 2045. As a result, lithium demand remains modest before 2053. However, the limited volume of recycled lithium in the later stages becomes insufficient to meet the needs of new ship construction. Consequently, lithium demand reaches 3.3 kt by 2060, still posing considerable pressure on supply.

According to data from the U.S. Geological Survey [54], global lithium production in 2022 was approximately 130 kt, with China accounting for 55.2% of global consumption, and 80% of that consumption being used in the battery sector. Considering the surging demand for lithium batteries driven by the rapid expansion of the electric vehicle market, as well as national resource prioritization toward the EV sector [12,16], this study assumes that the share allocated to the AES market is 2 kt. Figure 10 further illustrates the years and magnitudes in which lithium demand exceeds the allocated supply for the inland electric shipping sector under different scenarios. In the BMS scenario, resource supply pressure is expected between 2047 and 2057, resulting in a cumulative lithium shortage of approximately 1.3 kt. By comparing the post-recycling lithium demand across all scenarios with the annual lithium supply allocated to the inland AES market, it is evident that the BMH scenario faces the least supply pressure, with a total shortage of only 0.5 kt concentrated between 2048 and 2052. In contrast, under the low-recycling BML scenario, the shortage reaches 4.2 kt, indicating that improving the collection rate of retired batteries and enhancing recycling technologies can significantly alleviate lithium supply pressure.

Simultaneously, a well-planned electrification pathway for inland shipping can also substantially ease supply constraints. In the high-penetration MAD and MAR scenarios, cumulative shortages of 20.9 kt and 17.6 kt, respectively, are observed. While the MID and MIR scenarios—characterized by slower AES deployment—exhibit lower overall lithium demand, they still experience persistent supply shortages from 2052 to 2060, with total deficits of 7.5 kt and 9.0 kt, respectively.

5. Conclusions and Discussion

The study forecasts long-term lithium demand and recycling potential for China’s inland AES through a dynamic MFA framework integrating fleet expansion, battery chemistry transitions, and recycling improvements across seven scenarios (2025–2060). Our results indicate that under China’s 2060 decarbonization target—reducing the carbon emission intensity of inland vessels from 5.832 gCO₂/t·km in 2025 to 2.435 gCO₂/t·km by 2060—AES penetration must reach 25.36%, requiring 51.5–79.9k cumulative new-builds. Battery demand ranges 402.4–677.7 GWh, driving lithium consumption of 49.1–95.9 kt, offset by 5.4–25.2 kt recycled lithium by 2060. High-intensity battery scenarios with slow recycling (e.g., MAD) create severe supply pressures, while low-intensity batteries plus robust recycling (e.g., MID/BMH) mitigate risks despite potential late-stage shortages. Recycling proves critical for closing supply gaps, though second-life battery applications cause recovery time-lags, reducing short-term effectiveness despite long-term benefits.

Moreover, this study highlights the importance of coordinated policy design. The allocation of limited domestic lithium resources to the AES sector must be balanced with rising demand from electric vehicles and stationary storage. It seems that policy intervention is necessary: In the short term (2025–2035), recycling infrastructure should be initiated, with battery recycling rates increased to over 70% by 2035; in the medium term (2035–2045), efforts should focus on transitioning battery technology; in the long term (post-2045), international cooperation should be deepened, including signing supply agreements with lithium resource countries, while supporting research and development of alternative materials. In conclusion, achieving a sustainable and low-carbon transition for inland waterway shipping in China is technically feasible, but requires synchronized advancements in AES deployment, next-generation battery technologies, and robust lithium recycling systems, which are critical to ensuring the resilience of China’s inland shipping and lithium supply chains.

This research is not without limitations. Forecasting long-term dynamics involves uncertainties in economic growth rates, technological innovation timelines, and market behavior. Additionally, this study does not fully capture the dynamic competition from other battery markets or environmental impacts beyond carbon metrics. Future studies could incorporate agent-based modeling, cross-sectoral lithium demand analysis, and international trade dependencies to enrich scenario robustness. Moreover, the research focus is primarily on inland waterways, while future research can be further expanded to the field of marine transportation. Marine electric applications—such as coastal ferries, port tugs, and short-haul cargo ships—will likely face intensified constraints due to their higher energy density requirements (e.g., for extended offshore operations) and harsher operating environments (e.g., saltwater corrosion) [55]. Unlike inland vessels, marine batteries require specialized adaptations, including anti-corrosion casings, thermal management systems resilient to temperature fluctuations, and compatibility with seawater cooling infrastructure [56]. Future work should assess how these modifications affect lithium resource demand intensity and recyclability. Also, integrating hybrid decarbonization pathways into resource demand models, such as electric propulsion combined with green hydrogen or ammonia for long-haul segments, would provide a more holistic view of maritime sector sustainability.