The product system boundary delineates all processes included within the assessment scope, covering every elementary energy and material flow embedded in individual manufacturing workflows^[33]. The CTG boundary defined in this research encapsulates three interconnected modules for domestic SSSB production: raw material acquisition, battery fabrication, and on-site waste disposal. The raw material acquisition stage is further split into upstream raw material synthesis and cross-regional raw material logistics. Each battery production workflow exhibits distinct power and thermal energy consumption profiles, which necessitates independent boundary segmentation for each fabrication sub-process. All waste streams generated post-manufacturing are treated in compliance with national emission standards or transferred to certified professional waste treatment facilities. System input flows comprise fossil and renewable energy feedstocks plus all battery raw materials, while output streams include particulate pollutants, non-methane hydrocarbons (NMHC), and various solid waste fractions with ecological hazards. The functional unit adopted for all CTG impact calculations is 1 kWh of finished SSSB product, corresponding to the nominal rated energy capacity of cells upon factory delivery^[34].

All inventory and impact datasets deployed in this analysis are sourced from the CLCD and the eFootprint modeling platform. GWP is designated as the primary characterization metric to quantify carbon footprints. A suite of nine complementary environmental impact categories is integrated to deliver a holistic sustainability evaluation, including GWP, primary energy demand (PED), abiotic depletion potential (ADP), water use (WU), acidification potential (AP), eutrophication potential (EP), respirable particulate formation potential (RI), ozone depletion potential (ODP), and photochemical ozone formation potential (POFP). The abbreviation and full name comparison table of environmental indicators and other terms in this article can be found in Supplementary Table 1. Detailed LCIA characterization models, reference characterization factors, equivalent units, and standardized calculation protocols for all nine indicators are documented in Supplementary Table 2 to guarantee full methodological transparency.

The LCI dataset is constructed using 2024 actual operational records obtained from a domestic Chinese SSSB manufacturer, combined with publicly released official industrial statistics in China. The SSSB cell adopts nickel-cobalt manganese oxide, nickel:cobalt:manganese ratio of 8:1:1 (NCM811) as the cathode precursor; lithium aluminum titanium phosphate (LATP) and lithium lanthanum titanium oxide (LLTO) serve as cathode and anode solid electrolytes, respectively. The full background dataset is summarized in Supplementary Table 3. For the liquid electrolyte, lithium hexafluorophosphate (LiPF₆) serves as the main lithium salt, dissolved in a composite solvent system based on propylene carbonate (PC), ethylene carbonate (EC), dimethyl carbonate (DMC) and ethyl methyl carbonate (EMC). Multiple auxiliary carbonates, lithium salt additives, and sulfonate functional agents such as diethyl carbonate (DEC), vinyl carbonate (VC), fluoroethylene carbonate (FEC), and 1,3-propane sultone (PS) are further blended into the formulation, while ethyl acetate and lithium-salt-doped dimethyl carbonate are used as processing purifiers. Aluminum and nickel alloys are selected as conductive materials for electrode tabs. Key observations derived from LCI data collation are outlined below: NCM811 cathode precursor occupies the largest mass fraction among all raw and auxiliary inputs, functioning as the core material that determines the cell’s energy density performance.

Process solvents (purified water and N-methyl-2-pyrrolidone, NMP) represent the second-largest mass input category. Water-based anode slurries consume substantially more purified water than the NMP-dependent oil-based cathode slurry system, marking the primary compositional distinction between cathode and anode slurry preparation workflows.

Graphite serves as the dominant anode precursor, with silicon oxide incorporated as a low-dose additive to elevate specific capacity. This material combination balances cyclic durability and volumetric energy density trade-offs.

Cathode and anode binder formulations display evident compositional divergence. The combined mass loading of styrene butadiene rubber (SBR) and carboxymethyl cellulose (CMC) for anodes far exceeds the polyvinylidene fluoride (PVDF) dosage used for cathodes, reflecting stricter stability requirements for aqueous anode slurries.

Cathode-side solid electrolyte slurry consumption surpasses that of the anode; the mass input of LATP dispersion liquid is markedly higher than that of LLTO dispersion liquid, a structural design optimized to stabilize cathode-electrolyte interfacial contact.

Liquid electrolyte is maintained at a moderate loading level, retaining the ion transport advantages of hybrid solid-liquid architectures while distinguishing SSSBs from conventional LIBs with excessive liquid electrolyte filling volumes.

A carbon footprint sensitivity assessment was conducted to identify dominant contributing factors and establish a clear priority hierarchy for industrial carbon abatement. The sensitivity coefficient is defined as the percentage variation in overall stage carbon footprint induced by a ±20% fluctuation of a single parameter. Larger sensitivity coefficient magnitudes indicate that minor deviations in the corresponding sub-process or raw material trigger substantial shifts in total greenhouse gas emissions, marking high-priority optimization targets.

The sensitivity evaluation framework covered three core production phases: raw material procurement, cell manufacturing, and waste remediation. For the fabrication stage, three energy fluctuation scenarios were differentiated to decouple the respective impacts of electricity and natural gas: synchronous variation of total energy inputs, independent fluctuation of electricity consumption only, and independent fluctuation of natural gas consumption only. This tripartite classification quantifies the differentiated carbon-footprint responses driven by the power grid mix and thermal process energy loads. The sensitivity coefficient calculation formula is defined as Equation (1):

(1) $$ \begin{equation} \begin{aligned} \mathrm{SC}_{i,j} = (\mathrm{C}_{i,j} × Δ)/\mathrm{C}_{i,total} × 100\% \end{aligned} \end{equation} $$

where SC_i,j denotes the sensitivity coefficient (%) corresponding to the j-th sub-process or raw material within the i-th production stage; C_i,j refers to the carbon footprint (kg CO₂ eq) generated by the j-th sub-stage or raw material; Δ represents the standardized parameter fluctuation amplitude, fixed at 0.2 (±20%) across all model calculations; C_i,total equals the aggregated total carbon footprint of the entire i-th production stage.

Notably, SC_i,j maintains a strict linear proportional relationship with the fluctuation amplitude Δ. If alternative fluctuation ranges of ±10% or ±30% are adopted, all sensitivity coefficients will scale proportionally by factors of 0.5 and 1.5, respectively, without altering the relative ranking of high-impact parameters or carbon reduction priority sequences. Supplementary multi-scenario sensitivity simulations therefore deliver no incremental insight for hotspot identification, rendering a unified ±20% fluctuation bandwidth sufficient to reliably differentiate high-magnitude impact factors from negligible contributors. All sensitivity computations are executed on the eFootprint platform integrated with the CLCD database, covering every key raw material and manufacturing sub-process across the full CTG boundary.

The raw material acquisition module is subdivided into upstream raw material synthesis and inter-factory raw material transportation. A standalone CTG inventory model is constructed for every input material to quantify embedded environmental burdens. All material-specific inventory datasets are extracted from domestic LIB manufacturers’ environmental impact assessment reports, with detailed LCI records cataloged in Supplementary Tables 4-22. The transportation sub-module accounts for all emissions generated during delivery from upstream chemical plants to the SSSB manufacturing facility. Logistics energy consumption and associated carbon footprints are quantified based on the net mass of each transported raw material^[31]. A medium-distance transport radius of 200 km is adopted, with 18-ton heavy-duty diesel trucks specified as the exclusive delivery vehicle.

The fabrication module is developed by integrating standardized industrial SSSB production workflows with peer-reviewed, authoritative energy consumption benchmarks documented for full-scale solid-state battery production lines^[35]. The complete manufacturing value chain covers cathode/anode electrode fabrication, cell assembly, formation cycling, and static aging. Electrode production encompasses batch material homogenization, precursor coating and drying, calendering, slitting, and multi-stage vacuum drying. Cell assembly workflows include electrode blank cutting, layer stacking, tab welding, aluminum plastic packaging, electrolyte injection, and hermetic sealing. Post-assembly formation and aging procedures consist of electrochemical formation cycles, long-duration static maturation, and final end-of-line performance testing. Electricity and natural gas represent the two primary energy carriers consumed throughout all fabrication sub-processes^[36].

This module exclusively models the treatment of hazardous byproducts and industrial wastewater generated on-site prior to finished cell delivery, excluding post-service battery recycling. Environmental burden calculations for waste streams leverage the embedded waste impact calculation function within eFootprint and the CLCD inventory framework. The waste remediation module was split into two core subsections: hazardous waste secure disposal and municipal wastewater treatment. All hazardous solid waste was transported to certified facilities for stabilized secure landfilling, while factory wastewater was routed to municipal sewage treatment plants equipped with modified Bardenpho biochemical tanks coupled with membrane bioreactor units. Environmental emissions from heavy-duty truck transportation of hazardous waste from the production plant to disposal sites were fully incorporated within the module’s accounting boundary.

Carbon footprint quantification for all upstream raw material inputs was conducted in line with the accounting framework illustrated in Section "DATA AND METHODS", with aggregated results visualized in Figure 2.

From a holistic perspective, the total embodied carbon footprint of all raw materials required to produce 1 kWh of SSSBs equals 126.55 kg CO₂ eq, accompanied by substantial multi-category ecological burdens. Environmental pressures are overwhelmingly concentrated in three core cell components: cathodes, anodes, and current collectors, with component-specific carbon distribution characteristics detailed as follows.

Cathode materials constitute the dominant emission hotspot within the raw material phase, with greenhouse gas loads primarily originating from NCM811 precursor synthesis. NCM811 precursor generates 105.87 kg CO₂ eq, accounting for over 90% of the total cathode carbon inventory and representing the single largest emission contributor across all raw materials. All power consumption datasets for precursor co-precipitation and high-temperature calcination of NCM811 adopt China’s national grid emission factor embedded within the CLCD database, which accurately reflects the domestic power generation mix for carbon footprint modeling. Corresponding supplementary environmental indicators remain at elevated levels, including a PED of 1,402.63 MJ, AP of 1.82 kg SO₂ eq, and RI of 0.27 kg PM_2.5 eq, further verifying the dominant environmental footprint of NCM811 production.

Auxiliary cathode materials deliver marginal carbon contributions: NMP solvents generate 1.37 kg CO₂ eq, while carbon nanotube (CNT) conductive agents contribute 0.85 kg CO₂ eq; collectively, these two inputs account for merely 2.1% of total cathode emissions. Minor carbon loads are recorded for cathode binders (0.0188 kg CO₂ eq), acetylene black conductive additives (0.0307 kg CO₂ eq), and LATP solid cathode electrolyte (0.2631 kg CO₂ eq), with negligible corresponding values across all supplementary impact categories.

Anode-related carbon emissions are primarily driven by graphite and CMC binder production. Graphite generates a carbon footprint of 2.53 kg CO₂ eq, while CMC contributes 2.16 kg CO₂ eq; together these two materials account for more than 70% of the full anode carbon inventory. The graphitization stage of natural/synthetic graphite constitutes the major energy-intensive procedure underlying its elevated embodied carbon, accompanied by a PED of 70.35 MJ and RI of 0.0049 kg PM_2.5 eq, consistent with documented graphite manufacturing characteristics. Silicon oxide and anode CNTs generate secondary carbon burdens of 0.48 and 0.85 kg CO₂ eq, respectively. By contrast, SBR, anode acetylene black, and aqueous process purified water impose nearly undetectable carbon loads, with water recording negligible GWP. All auxiliary environmental indicators for anode materials mirror the uneven carbon distribution pattern dominated by graphite and CMC.

Copper and aluminum foil manufacturing constitute the core emission sources for current collectors. Electrolytic copper foil carries an embodied carbon footprint of 3.86 kg CO₂ eq, approximately 2.5 times the 1.54 kg CO₂ eq recorded for aluminum foil, rendering copper foil the primary environmental burden within this component group. This disparity stems from high power demand and pollutant discharge during copper smelting and electrolytic foil fabrication. Copper foil also exhibits extreme water intensity, with total WU reaching 1,917.10 kg, far exceeding all other raw material inputs, a direct consequence of repeated electroplating and high-purity rinsing steps required for domestic electrolytic copper production. Additional metrics including AP, EP, and RI all register higher values for copper foil relative to aluminum foil, maintaining consistent distribution trends with carbon footprint outcomes.

Separators represent one of the lowest-impact components, with a total carbon footprint of only 0.174 kg CO₂ eq, paired with minimal PED (2.96 MJ) and RI (0.00013 kg PM_2.5 eq) values, owing to low energy consumption during polymer film extrusion; their overall environmental contribution can be largely disregarded in full-system assessments.

Liquid electrolyte generates a carbon burden of 1.84 kg CO₂ eq, driven by multi-step synthesis of LiPF₆ and mixed carbonate solvents, accompanied by PED of 31.05 MJ, RI of 0.0045 kg PM_2.5 eq, and EP of 0.0008 kg PO₄^3--eq, indicating the concurrent risk of particulate and water pollution.

Tab conductive materials record a modest carbon footprint of 0.53 kg CO₂ eq, with medium-to-low supplementary indicator values, including PED (6.80 MJ) and AP (0.0008 kg SO₂ eq), serving as low-impact auxiliary structural components.

Aluminum plastic packaging film forms the sole constituent of cell shells, generating 3.44 kg CO₂ eq from polymer compounding and thermoforming workflows, alongside PED of 39.71 MJ, AP of 0.0056 kg SO₂ eq and RI of 0.0029 kg PM_2.5 eq, representing a non-negligible secondary emission source.

Overall carbon distribution across raw material inputs demonstrates a severe imbalance. Distinct from lithium metal anode ASSBs, SSSBs adopt NCM811 ternary cathodes that account for 81.1% of total CTG carbon emissions, emerging as the irreplaceable core hotspot for decarbonization, aligning with established LIB LCA conclusions highlighting cathode production as primary climate burden^[37]. All supplementary environmental indicators replicate this uneven distribution, with peak values concentrated in NCM811 synthesis while separators and tabs deliver trivial ecological loads. Disaggregated component carbon inventories confirm that high-energy mineral raw materials (NCM, graphite, copper foil) dominate greenhouse gas emissions, whereas thin-film solid electrolytes LATP and LLTO impose minimal environmental pressure due to low mass loading and mild processing energy demand. These findings provide targeted guidance for low-carbon material formulation and upstream supply chain optimization of SSSBs.

A cross-study comparative analysis is conducted to benchmark the carbon footprint magnitude against published domestic LCA literature, as presented in Supplementary Figure 1. Two representative datasets are selected for horizontal comparison: regional NCM liquid LIB CTG averages documented by Kallitsis et al.^[38], capturing supply chain and grid-induced regional disparities; and commercial NCM811, NCM111 and LFP cell baseline datasets compiled by Lai et al.^[39], representing mainstream industrial liquid battery technical routes.

The aggregate CTG carbon value derived in this study falls within the documented range of domestic NCM ternary LIBs, positioned at the upper bound of established intervals. Numerical discrepancies originate from three distinct drivers: first, this work utilizes the localized CLCD database calibrated to China’s industrial conditions instead of European foreign background inventories; second, semi-solid-state cell designs require elevated raw material mass input per unit kWh relative to mature liquid counterparts, which can be found in the Supplementary Table 23, a byproduct of ongoing industrialization and marginally lower volumetric energy density; third, coating, drying and dry-room energy parameters are calibrated against peer-reviewed solid-state manufacturing benchmarks that diverge from conventional liquid cell production lines. Despite quantitative deviations stemming from database and compositional differences, the universal conclusion that ternary cathode materials constitute the dominant carbon source remains fully consistent across all referenced scholarly work.

Logistics emissions are quantified based on the 200 km medium haul distance and 18-ton diesel heavy-duty truck parameters defined in Section "DATA AND METHODS", with multi-indicator outcomes plotted in Figure 3.

The aggregate carbon footprint generated by all raw material transportation totals 0.1358 kg CO₂ eq, with complete supplementary indicator datasets tabulated in the Supplementary Materials. Logistics burdens also display highly uneven component-wise distribution: cathode material delivery dominates at 59.2% of total transport emissions (of which NCM811 precursor accounts for 32.4% of total transportation emissions), followed by anode raw material logistics at 23.7% (of which graphite precursor accounts for 12.9% of total transportation emissions). Current collectors, electrolytes and cell shells represent secondary contributors with respective shares of 4.9%, 6.3% and 5.4%. Separator and tab transportation each contribute less than 0.3% of logistics carbon loads, rendering their environmental impacts negligible.

This skewed distribution correlates directly with raw material mass density and truck exhaust characteristics: increased cargo weight generates amplified diesel combustion emissions during highway transit. NCM811 cathode precursor shipments constitute the single largest logistics emission subcategory. All eight supplementary environmental indicators replicate this cathode-anode dominated distribution pattern. Nevertheless, the full transportation-stage carbon footprint is only a quarter of the emissions generated during tab raw material synthesis. Economies of scale within bulk logistics operations minimize the relative environmental pressure of transportation, consistent with prior LCA research outcomes.

Full-process power and thermal energy consumption along with the corresponding multi-category environmental burdens are systematically quantified in accordance with standardized manufacturing workflows and energy benchmarks outlined in Section "DATA AND METHODS". Process energy profiles and associated impact distributions are visualized in Figure 4 and Figure 5 respectively. The complete fabrication workflow encompasses cathode/anode electrode fabrication, cell assembly, formation cycling and auxiliary supporting operations, with two primary energy carriers (grid electricity and natural gas) deployed across all sub-processes exhibiting drastically divergent consumption intensities.

From an electricity consumption perspective, electrochemical formation registers the highest unit power demand at 5.19 kWh per kWh finished cell, substantially exceeding all alternative fabrication steps. Washing workflows, material handling, tab welding, packaging, and electrolyte filling also consume substantial volumes of grid electricity. Conversely, batch mixing, knife slitting and calendering procedures record minimal power draw, delivering marginal contributions to total electrical demand.

Natural gas is exclusively deployed for temperature-controlled manufacturing units. Dry rooms constitute the largest natural gas consumer at 4.73 kWh per kWh cell, followed by cathode and anode coating-drying processes each consuming 2.85 kWh thermal energy. Vacuum drying and static aging also require moderate natural gas input, while nearly all remaining sub-processes operate without thermal gas demand. Overall energy allocation aligns precisely with process functional requirements, establishing a consumption pattern where electricity serves as the universal primary energy source and natural gas is restricted to heating and humidity-control workflows.

Manufacturing-stage environmental burdens are co-driven by electricity and natural gas combustion, with divergent contribution ratios across impact categories. Focusing on GWP as the core climate metric, electricity consumption generates 6.7394 kg CO₂ eq while natural gas contributes merely 0.09264 kg CO₂ eq; grid power therefore accounts for over 98% of total fabrication greenhouse gas emissions and functions as the definitive dominant emission driver. Uncertainty bounds for electricity-derived carbon footprints span 6.0698-7.4152 kg CO₂ eq, with natural gas ranges limited to 0.08344-0.10192 kg CO₂ eq; full-process uncertainty remains constrained within reasonable limits without altering electricity’s dominant emission role.

For resource-oriented indicators PED and ADP, contrasting trends emerge: natural gas combustion delivers a PED of 108.6204 MJ, exceeding the 81.74 MJ associated with grid electricity, while natural gas ADP (0.00002 kg antimony eq) also surpasses electricity’s 0.000008 kg antimony eq, attributed to fossil resource extraction burdens embedded within natural gas supply chains.

For all residual impact metrics including WU, AP, EP, RI, ODP and POFP, electricity consumption dominates total loads. Grid-powered workflows generate water use of 2.77185 kg, vastly outstripping natural gas’s 0.090324 kg water demand. Additional electricity-derived values include AP (0.003 kg SO₂ eq), EP (0.0006 kg PO₄^3--eq), RI (0.0008 kg PM_2.5 eq) and POFP (0.0007 kg NMVOC eq), all substantially elevated relative to natural gas equivalents. ODP emissions are almost exclusively linked to grid electricity generation, with natural gas contributions negligible.

Electricity’s overwhelming dominance across most environmental indicators delineates clear low-carbon optimization pathways for cell fabrication. Efficiency upgrades for high-power units including formation and washing, paired with grid decarbonization via renewable electricity procurement, can drastically mitigate manufacturing-stage ecological footprints. Meanwhile, natural gas’s disproportionate PED and ADP contributions indicate that thermal process electrification via heat pump substitution merits integration into future low-carbon manufacturing design frameworks.

Environmental burdens originating from on-site waste treatment post-fabrication are quantified following the accounting protocols defined in Section "DATA AND METHODS", with segmented loads split across three sub-units: municipal wastewater treatment, secure hazardous waste landfilling, and hazardous waste highway transportation; differentiated category contributions are displayed in Figure 6.

The carbon footprint distribution across waste remediation workflows exhibits an extreme imbalance. Wastewater treatment generates 0.01015 kg CO₂ eq, hazardous waste disposal produces 0.04105 kg CO₂ eq, and hazardous waste transportation accounts for 0.06520 kg CO₂ eq. Heavy-duty diesel transit constitutes the primary carbon hotspot, responsible for 56.0% of total waste-stage greenhouse gas emissions, with hazardous waste secure landfilling ranking second at 35.3%, and wastewater treatment delivering the minimum share of merely 8.7%.

Divergent dominant contributors emerge across supplementary environmental indicators. Hazardous waste landfilling dominates PED (0.9512 MJ) and WU (0.2720 kg), driven by energy and clean water consumption during stabilization pretreatment. Hazardous diesel truck transportation dominates all remaining atmospheric impact categories, including ADP (0.0000002 kg antimony eq), AP (0.00132 kg SO₂ eq), EP (0.000237 kg PO₄^3--eq), RI (0.000249 kg PM_2.5 eq), ODP (0.000000006 kg CFC-11 eq) and POFP (0.000398 kg NMVOC eq), directly stemming from sulfur oxide, nitrogen oxide, particulate and volatile organic compound exhaust streams. Wastewater treatment maintains minimal values across all impact categories, exerting trivial overall environmental pressure.

Collectively, waste-stage environmental burdens demonstrate multi-source differentiation characteristics. Diesel-powered hazardous waste transit acts as the primary source of carbon emissions and most air pollutants, while hazardous waste landfilling dominates energy and water consumption metrics. Wastewater treatment delivers negligible ecological loads. These segmented distribution patterns establish targeted mitigation routes for factory waste management: optimizing hazardous waste logistics modes and improving transit efficiency delivers the most substantial reductions to aggregate waste-phase environmental footprints.

Raw material procurement is segmented into upstream raw material synthesis and cross-regional logistics subsections, with carbon footprint adopted as the core evaluation metric. Sensitivity outcomes replicate the uneven emission distribution documented in Section "Environmental impacts of the raw material acquisition stage", with cathode raw materials delivering dominant coefficients and anode/auxiliary inputs registering secondary influence (refer to Supplementary Figures 2-4).

Within raw material synthesis workflows, NCM811 cathode precursor exhibits the maximum sensitivity coefficient of 16.73%, vastly exceeding all alternative inputs and representing the highest-priority decarbonization target across upstream supply chains. Auxiliary cathode materials including single-walled CNTs, polymeric binders and cathode acetylene black all record sensitivity coefficients below 0.14%, exerting negligible influence on aggregate carbon inventories.

For anode raw materials, graphite (0.40%) and CMC binder (0.34%) constitute secondary sensitive parameters, while silicon oxide, conductive agents and residual anode inputs register coefficients below 0.1%. Among miscellaneous raw materials, electrolytic copper foil (0.61%) and aluminum plastic shell film (0.54%) display relatively elevated sensitivity magnitudes, whereas separators, electrolyte salts and tab materials all deliver coefficients lower than 0.30%.

Within raw material logistics sensitivity evaluation, cathode precursor transit remains the dominant contributor, with NCM811 transportation recording a coefficient of 6.48%, followed by cathode CNT logistics (2.89%) and NMP solvent delivery (1.80%). Graphite anode transportation represents the primary sensitive anode logistics parameter at 2.59%, with all remaining anode material transit coefficients falling below 1.00%. Shell film logistics (1.07%) and electrolyte transportation (1.26%) register mild secondary sensitivity, while separator and tab delivery coefficients drop below 0.10, carrying minimal analytical significance [Supplementary Figures 5-7].

Cell fabrication encompasses cathode/anode electrode production, cell assembly, formation cycling and miscellaneous auxiliary workflows, with grid electricity functioning as the primary energy carrier and natural gas as secondary thermal input. Sensitivity outcomes confirm that electricity parameter fluctuation drives nearly all carbon footprint variance, while natural gas consumption delivers marginal influence, with core sensitive sub-processes concentrated within high-power fabrication units [Supplementary Figures 8-12].

Identical sensitivity magnitudes are recorded for cathode and anode electrode workflows. Coating and drying units deliver the highest aggregate sensitivity of 0.54%, decomposed into an electricity-derived component of 0.47% and a natural gas contribution of 0.07%. Calendering processes register a sensitivity coefficient of 0.33%, while batch mixing records 0.16%; both operations rely exclusively on grid electricity. Slitting and vacuum drying deliver coefficients below 0.03%, with vacuum drying’s natural gas-associated sensitivity limited to merely 0.01%, functionally negligible.

Within cell assembly workflows, washing operations exhibit an aggregate sensitivity coefficient of 1.887611%, with welding, packaging and electrolyte filling-sealing recording 1.51%; both units operate solely via grid power. Electrode cutting and stacking register coefficients below 0.10%, exerting trivial influence on total fabrication carbon footprints.

For formation and aging sub-processes, electrochemical formation delivers the single highest sensitivity coefficient across all manufacturing workflows at 9.42%, driven entirely by grid electricity draw. End-of-line performance testing records a sensitivity of 0.942% (pure electricity consumption), while static aging’s coefficient stands at only 0.004%, originating from minor natural gas heating demand with negligible systemic impact.

For miscellaneous auxiliary units, material handling registers a sensitivity of 1.89% (100% electricity-dependent), while dry rooms record 1.65%, split into 1.54% electricity and 0.11% natural gas contributions; natural gas accounts for merely 6.70% of dry room total sensitivity, reaffirming electricity’s dominant role.

Collectively, over 98.00% of fabrication-stage carbon footprint sensitivity originates from grid electricity variation, with natural gas demand fluctuations generating minimal systemic shifts. Electrochemical formation, industrial washing, material handling, and humidity-controlled dry rooms constitute the four core electricity-intensive sensitive workflows, validating the prior conclusion that grid power consumption acts as the definitive manufacturing emission hotspot.

Waste remediation carbon footprint sensitivity displays clear hierarchical differentiation: hazardous waste highway transportation ranks first, secure hazardous waste landfilling second, and municipal wastewater treatment delivers the weakest parametric response [Supplementary Figure 13]. Heavy-duty diesel truck transit of hazardous waste records a sensitivity coefficient of 11.20%, representing the most impactful single parameter within the waste phase. Hazardous waste secure disposal follows at 7.05%, while full wastewater treatment workflows register the minimum coefficient of 1.74%. Sensitivity distribution across all supplementary environmental indicators mirrors the carbon footprint ranking hierarchy, confirming diesel hazardous waste transit as the primary driver for waste-stage greenhouse gas emissions.

Integrating sensitivity outcomes across all three CTG phases, the four highest-priority decarbonization parameters ranked by sensitivity coefficient magnitude are listed sequentially: NCM811 raw material synthesis (16.73%), heavy-duty diesel hazardous waste transportation (11.20%), electricity consumption during cell formation (9.42%), and NCM811 raw material logistics (6.48%). All residual raw material, manufacturing and thermal gas parameters record sensitivity coefficients below 2%, delivering marginal systemic carbon impacts.

These high-sensitivity hotspots demonstrate perfect alignment with the segmented carbon footprint accounting results outlined in Section "Accounting of environmental indicators for SSSBs at the CTG stage", verifying the robustness and internal consistency of the full LCA modeling framework. The ranked sensitivity hierarchy delivers unambiguous actionable guidance for industrial low-carbon retrofits: prioritizing decarbonization of NCM811 upstream manufacturing, efficiency upgrades for high-power formation equipment, and low-emission hazardous waste logistics delivers the maximum quantifiable greenhouse gas mitigation gains for SSSB production systems.