A carbon label driven carbon emission reduction grading method

Abstract

The application discloses a carbon label driven carbon emission reduction grading method, relates to the technical field of carbon emission reduction, and solves the technical problem that the prior art fails to effectively connect product data, enterprise operation and industry emission reduction demand; the application comprises the following steps: 1, obtaining a benchmark carbon footprint model for a benchmark product for a to-be-evaluated enterprise, and then calculating the carbon footprint of other products by the difference between the other products and the benchmark product to determine the carbon label of all products; 2, determining the carbon label of the corresponding enterprise; 3, determining the carbon emission reduction demand of different industries; 4, determining the enterprise which needs to carry out carbon emission reduction planning; 5, determining the optimal carbon emission reduction strategy of the enterprise; and 6, grading and controlling the enterprise according to the carbon emission reduction target involved in the optimal carbon emission reduction strategy of the enterprise to determine whether the carbon emission reduction requirement is met; the application realizes effective improvement of carbon footprint evaluation efficiency by introducing the innovative calculation method of the benchmark product combined with the difference threshold.

Description

Technical Field

[0001] This invention relates to the field of carbon emission reduction technology, and more specifically to a carbon tag-driven carbon emission reduction classification method. Background Technology

[0002] Currently, technological pathways in the field of carbon emission reduction mainly revolve around energy substitution, energy efficiency improvement, process innovation, and carbon management. Regarding energy substitution, while renewable energy technologies such as photovoltaics and wind power have been applied on a large scale, their output is intermittent and fluctuating, posing a challenge to the stable operation of the power grid. Hydrogen energy and sustainable aviation fuels, as important clean energy carriers, still require improvement in the economic viability and technological maturity of their entire chain, from preparation and storage to transportation and end-use applications. In terms of industrial process emission reduction, carbon capture, utilization, and storage (CFS) technologies are considered a key path to addressing the deep decarbonization of industries with difficult emission reductions; however, large-scale deployment currently faces core technological bottlenecks such as high energy consumption, high costs, and the security and monitoring reliability of long-term storage.

[0003] While a preliminary methodology for quantifying product carbon footprints has been established at the accounting and management level, significant obstacles remain in practical application. Many existing methods rely on idealized calculation models, assuming companies operate at a uniform level and technological condition. This unrealistic approach leads to discrepancies between carbon footprint calculation results and actual company conditions, and proposed emission reduction pathways often lack feasibility and are difficult to implement. Furthermore, traditional carbon footprint assessments typically require independent and complete lifecycle evaluations for each product, a time-consuming and labor-intensive process. For companies with diverse product portfolios, this results in high computational costs and low efficiency, posing a significant bottleneck for large-scale application. More importantly, existing solutions generally lack a systematic perspective, failing to effectively link product data, company operations, and industry emission reduction needs. This leads to emission reduction strategies often being disconnected from companies' actual capabilities, technological reserves, and the overall direction of industry efforts, making it difficult to form a scientific, tiered management system. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention provides a carbon tag-driven carbon emission reduction classification method. This method solves the problem that the prior art fails to effectively link product data, enterprise operations, and industry emission reduction needs, resulting in emission reduction strategies often being out of touch with the actual capabilities, technological reserves, and overall industry efforts, making it difficult to form a scientific classification and control system.

[0005] A carbon tag-driven carbon emission reduction grading method includes the following steps:

[0006] Step 1: Select a company to be valued, choose a benchmark product for the company to be valued, calculate the carbon footprint of the benchmark product, establish a benchmark carbon footprint model, compare the product to be valued with the benchmark product, obtain the differences, import the differences into the benchmark carbon footprint model to obtain the carbon footprint of the product to be valued, and generate the corresponding carbon label.

[0007] Step 2: Repeat Step 1 until all carbon labels of all products of all companies to be estimated are obtained. Based on the data collected during the calculation of the carbon footprint of all products of the companies to be estimated and other carbon emission-related data, sort out and calculate the carbon footprint of the companies to be estimated, and determine the corresponding carbon label of the companies to be estimated based on the carbon footprint of the companies to be estimated.

[0008] Step 3: Repeat step 2 until all carbon labels of the companies to be evaluated are obtained. Divide all the companies to be evaluated into industries, analyze the distribution of carbon labels of companies in different industries, and then determine the carbon emission reduction needs of the industry based on the characteristics of the industry.

[0009] Step 4: Based on the industry's carbon emission reduction needs and the carbon labels of all companies to be assessed, preliminarily determine the companies with carbon emission reduction plans;

[0010] Step 5: Combine the carbon label of the carbon reduction planning enterprise's products, the carbon reduction demand of the corresponding industry, the optimization cost and optimization return of the available carbon reduction technologies to determine the carbon reduction level of each carbon reduction planning enterprise's products. Based on reinforcement learning, determine the optimal carbon reduction strategy for the carbon reduction planning enterprise at present according to the product carbon reduction level, technology reserves, funds and energy prices.

[0011] Step 6: Classify and manage enterprises according to the carbon reduction targets involved in the enterprise's optimal carbon reduction strategy, and determine whether the carbon reduction requirements are met.

[0012] Further, step 1 includes:

[0013] Step 1.1: Calculate the carbon footprint of the benchmark product and establish a benchmark carbon footprint model based on the calculation results. The benchmark carbon footprint model includes the carbon footprint and total carbon footprint at different life cycle stages. The different life cycle stages include the material preparation stage, production stage, transportation stage, and usage stage.

[0014] Step 1.2: Compare the product to be evaluated with the benchmark product to obtain the differences between the product to be evaluated and the benchmark product at different stages of the product life cycle. The differences include material differences in the material preparation stage, process differences in the production stage, transportation differences in the transportation stage, and energy efficiency differences in the usage stage.

[0015] Step 1.3: Import the differences into the baseline carbon footprint model, call the emission factors in the database to calculate the carbon footprint of the product to be evaluated, and generate the corresponding product carbon label based on the carbon footprint of the product to be evaluated.

[0016] Furthermore, step 1.3 includes: quickly estimating the carbon footprint of the difference point by calling the emission factors in the emission factor database; if the carbon footprint of the difference point reaches a threshold, it is imported into the benchmark carbon footprint model and the carbon footprint of the product to be evaluated is accurately calculated using standard calculation methods; thereby further improving the calculation efficiency. The threshold is set according to industry standards.

[0017] Furthermore, the differential carbon footprint thresholds include multiple thresholds, with different threshold ranges corresponding to different processing methods. Using a% or b% (b>a) of the benchmark product's carbon footprint as the threshold base point, the threshold ranges and corresponding processing methods are set as follows:

[0018] A baseline carbon footprint less than a% is set as the low-impact threshold range. If the carbon footprint difference point falls within this range, it can be ignored or a qualitative note can be made without adjusting the value. A baseline carbon footprint greater than or equal to a% and less than b% is set as the attention threshold range. If the carbon footprint difference point falls within this range, it is not ignored directly, but the quick estimate is used as the final carbon footprint difference point and included in the calculation of the carbon footprint of the product to be evaluated, without further detailed data collection. At the same time, this data is recorded and marked as a quick estimate. A baseline carbon footprint greater than or equal to b% is set as the action threshold. If the carbon footprint difference point falls within this range, the standard calculation method is used: specific data is collected and precise calculation is performed.

[0019] Furthermore, step 2 includes: inputting all activity data, emission factors, and calculation formulas involved in the carbon footprint calculation of all products of the enterprise to be assessed into the carbon footprint database; when calculating the enterprise's carbon footprint, using professional carbon accounting software or an automated calculation template developed based on Excel / database, automatically calling the corresponding data to integrate and calculate the report to obtain the enterprise's carbon footprint according to the organizational boundaries and accounting boundaries; and generating the corresponding enterprise carbon label based on the carbon footprint of the enterprise to be assessed.

[0020] Furthermore, the different industries in step 3 include high-energy-consuming industries, clean energy industries, consumer goods industries, transportation industries, and emerging industries.

[0021] Furthermore, step 4 includes: after determining the carbon emission reduction needs of different industries from step 3, determining the priority of carbon emission reduction needs based on the current technological status and implementation cost estimates of the industries, and then ranking the enterprises for carbon emission reduction based on the enterprise carbon labels obtained in step 2, and selecting enterprises for carbon emission reduction planning from the ranked enterprises to be evaluated. Specifically, the selection method is to take the top n% in the ranking as the carbon emission reduction planning targets.

[0022] Further, step 4 includes:

[0023] Step 4.1: Conduct a technology status assessment. For each target industry, investigate the currently applicable carbon emission reduction technologies, obtain the technology maturity level, current popularity and future improvement potential of each technology, and establish an industry-level carbon emission reduction technology database.

[0024] Step 4.2: Perform cost estimation. For each carbon reduction technology, collect and calculate the initial investment cost, operation and maintenance costs, and financing-related costs required for its implementation. Based on the cost data, quantify the comprehensive implementation cost corresponding to a unit carbon dioxide equivalent (tCO2e) emission reduction.

[0025] Step 4.3: Determine industry emission reduction priorities. Using a multi-criteria decision analysis method, and combining the current technology status information obtained in Step 4.1 and the cost estimation results obtained in Step 4.2, score and rank the urgency of carbon emission reduction needs of each industry to generate an industry carbon emission reduction priority sequence.

[0026] Step 4.4: Implement enterprise ranking. First, based on the carbon emission levels indicated on the carbon labels of enterprises within the same industry, a preliminary ranking is performed, with enterprises having lower carbon emission levels given priority. Then, based on the industry carbon reduction priority sequence and the main carbon emission source categories identified in the enterprise carbon labels, the preliminary ranking is dynamically adjusted to output the final enterprise carbon reduction action priority ranking. The specific formula for this dynamic adjustment is as follows:

[0027]

[0028] Let i be the overall score of the i-th enterprise. Let i be the carbon emission level value of enterprise i. The priority weight of the industry emission reduction demand corresponding to the main source of carbon emissions of the i-th enterprise;

[0029] The calculation process for the priority weight of industry emission reduction demand includes: determining the industry carbon emission reduction demand that is most similar to the main source of carbon emissions in the enterprise's carbon label, determining its corresponding priority, and determining the corresponding weight based on the priority.

[0030] Then, a comprehensive ranking list of all enterprises is generated based on their overall scores, with the ranking from highest to lowest indicating the urgency of emission reduction.

[0031] Step 4.5: Identify the planning targets and extract the top n% of enterprises from the ranking list to form a carbon emission reduction planning target list.

[0032] Furthermore, step 5, determining the product's carbon emission reduction level, includes: calculating a carbon emission reduction potential score for each product using a comprehensive scoring method, as shown in the following formula:

[0033]

[0034] in, The carbon emission reduction potential score for the i-th enterprise's j-th product is given. Let the priority weight of industry emission reduction demand corresponding to the j-th type of product of the i-th enterprise be denoted. Let J be the carbon label value of the j-th type of product of the i-th enterprise. The return-cost ratio for the j-th type of product of the i-th enterprise is given by the following formula:

[0035]

[0036] in, For the optimized return corresponding to the j-th type of product of the i-th enterprise, Let $\frac{i}{j}$ be the optimized cost for the $j$-th type of product of the $i$-th enterprise.

[0037] Based on the emission reduction potential scores of all products, a grading threshold is determined to further determine the carbon emission reduction level of each product.

[0038] Furthermore, step 5 determines the company's optimal carbon reduction strategy, including:

[0039] Set up the state space: Define environmental state variables, including: product carbon emission reduction level, technology reserves, funds, energy prices and other factors;

[0040] Setting up action space: Define the emission reduction actions that enterprises can take, including: investing in new technologies, process optimization, carbon offsetting, and product adjustment;

[0041] Set a reward function: Design it to maximize long-term net benefits or minimize total costs;

[0042] The optimal carbon reduction strategy for enterprises is determined through reinforcement learning based on the set state space, action space, and reward function.

[0043] The beneficial effects of this invention include:

[0044] First, by introducing an innovative calculation method that combines benchmark products with difference thresholds, the efficiency of carbon footprint assessment has been effectively improved. This avoids repetitive and tedious calculations for massive amounts of products, and enables the rapid generation of product carbon labels and the aggregation of enterprise carbon labels while ensuring data accuracy and comparability. This provides an efficient data foundation for subsequent analysis, making large-scale, industry-wide carbon footprint assessment possible.

[0045] Secondly, this invention constructs a closed-loop decision-making mechanism that progresses from micro-level products to enterprise-level carbon emission reduction needs, and then from macro-level industry carbon emission reduction needs to determine the priority of micro-level product adjustments, feeding back to the macro level. It considers both the overall emission reduction needs at the industry level and deeply integrates with the specific product structure, financial situation, technological capabilities, and even external energy prices of enterprises. Furthermore, it utilizes reinforcement learning algorithms to find the optimal solution under multiple constraints, thereby tailoring an optimal emission reduction strategy for each enterprise that is both ambitious and realistic, greatly improving the acceptability and executability of the strategy.

[0046] Finally, this invention establishes a dynamic and scientific management system by linking the final carbon emission reduction target with the hierarchical management of enterprises. This not only ensures that emission reduction pressure is accurately transmitted to the enterprises that need to take the most measures, but also gives compliant enterprises greater room for development, thereby systematically promoting the entire industry to transform towards green and low-carbon practices in the most cost-effective way, and contributing practical value to achieving broader climate goals. Attached Figure Description

[0047] Figure 1 This is a flowchart of a carbon tag-driven carbon emission reduction classification method according to an embodiment of this application. Detailed Implementation

[0048] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0049] Example 1

[0050] The following is in conjunction with the appendix Figure 1 Specific embodiments of the present invention will be described in detail;

[0051] A carbon tag-driven carbon emission reduction classification method includes the following steps:

[0052] Step 1: Select a company to be valued, choose a benchmark product for the company to be valued, calculate the carbon footprint of the benchmark product, establish a benchmark carbon footprint model, compare the product to be valued with the benchmark product, obtain the differences, import the differences into the benchmark carbon footprint model to obtain the carbon footprint of the product to be valued, and generate the corresponding carbon label.

[0053] The specific methods for determining the enterprises to be assessed include: setting carbon emission level values ​​based on national carbon emission standards, and identifying enterprises whose carbon emission levels are greater than or equal to the carbon emission level thresholds as enterprises to be assessed that need to undergo carbon emission reduction optimization;

[0054] Step 2: Repeat Step 1 until all carbon labels of all products of all companies to be estimated are obtained. Based on the data collected during the calculation of the carbon footprint of all products of the companies to be estimated and other carbon emission-related data, sort out and calculate the carbon footprint of the companies to be estimated, and determine the corresponding carbon label of the companies to be estimated based on the carbon footprint of the companies to be estimated.

[0055] Step 3: Repeat step 2 until all carbon labels of the companies to be evaluated are obtained. Divide all the companies to be evaluated into industries, analyze the distribution of carbon labels of companies in different industries, and then determine the carbon emission reduction needs of the industry based on the characteristics of the industry.

[0056] Step 4: Based on the industry's carbon emission reduction needs and the carbon labels of all companies to be assessed, preliminarily determine the companies with carbon emission reduction plans;

[0057] Step 5: Combine the carbon label of the carbon reduction planning enterprise's products, the carbon reduction demand of the corresponding industry, the optimization cost and optimization return of the available carbon reduction technologies to determine the carbon reduction level of each carbon reduction planning enterprise's products. Based on reinforcement learning, determine the optimal carbon reduction strategy for the carbon reduction planning enterprise at present according to the product carbon reduction level, technology reserves, funds and energy prices.

[0058] Step 6: Classify and manage enterprises according to the carbon reduction targets involved in the enterprise's optimal carbon reduction strategy, and determine whether the carbon reduction requirements are met.

[0059] In another embodiment, step 1 includes:

[0060] Step 1.1: The benchmark product is the product with the largest carbon footprint among the enterprises to be evaluated, based on the available carbon footprint data and the enterprise's production and sales data. For the benchmark product, a detailed product carbon footprint calculation is performed according to the international standard "PAS2050-2011 Life Cycle Greenhouse Gas Emission Assessment Specification" to obtain the benchmark carbon footprint model. The benchmark carbon footprint model includes the carbon footprint and total carbon footprint at different life cycle stages. The different life cycle stages include the material preparation stage, production stage, transportation stage, and usage stage.

[0061] Step 1.2: Compare the product to be evaluated with the benchmark product to obtain the differences between the product to be evaluated and the benchmark product at different stages of the product life cycle. The differences include material differences in the material preparation stage, process differences in the production stage, transportation differences in the transportation stage, and energy efficiency differences in the usage stage.

[0062] Specifically, taking manufacturing as an example, material differences may involve adding a component, reducing weight, or replacing materials; process differences may involve adding or removing a processing step; transportation differences may involve changes in transportation distance and method due to different sales destinations; and energy efficiency differences may involve changes in energy consumption during the product's usage phase.

[0063] Taking the service industry as an example, differences in materials may be due to the need for more refined services, leading to additional consumables; differences in processes may be due to the need for additional value-added services, leading to the use of certain tools or electronic devices; and differences in transportation may be due to the handling of additional materials required for serving customers or services.

[0064] Step 1.3: Import the differences into the baseline carbon footprint model, call the emission factors in the database to calculate the carbon footprint of the product to be evaluated, and generate the corresponding product carbon label based on the carbon footprint of the product to be evaluated.

[0065] In another embodiment, step 1.3 includes: calling emission factors from the emission factor database to quickly estimate the carbon footprint of the difference point; if the carbon footprint of the difference point reaches a threshold, it is imported into the benchmark carbon footprint model and the carbon footprint of the product to be evaluated is accurately calculated using standard calculation methods; thereby further improving the calculation efficiency.

[0066] Specifically, the emission factor database identifies the corresponding emission factors for various energy sources and materials that have been pre-collected and pre-treated.

[0067] Specifically, the rapid estimation refers to direct calculation based on the difference point values ​​and corresponding emission factors.

[0068] For example, using product A as the benchmark product with a total life-cycle carbon footprint of 100 kg CO2e, and product B as the product to be evaluated; the differences between the two are as follows:

[0069] Difference 1: Product B is 0.5 kg heavier than Product A, which means an increase of 0.5 kg of steel. Using the steel emission factor (2.5 kg CO2e / kg) from the database, the increment is calculated as: 0.5 kg * 2.5 kg CO2e / kg = 1.25 kg CO2e.

[0070] Difference 2: Product B has improved energy efficiency compared to Product A, reducing annual power consumption by 10 kWh; using the grid power emission factor (0.5 kg CO2e / kWh), the reduction is calculated as: -10 kWh * 0.5 kg CO2e / kWh = -5 kg ​​CO2e.

[0071] The carbon footprint of new product B = 100 + 1.25 - 5 = 96.25 kg CO2e.

[0072] This method avoids repeatedly calculating all unchanged parts, reduces the amount of computation, and is more conducive to application and promotion.

[0073] In another embodiment, the differential point carbon footprint threshold includes multiple thresholds, with different threshold ranges corresponding to different processing methods. This ensures that resources are invested in the areas with the greatest impact while omitting unnecessary calculations for minor changes, and also provides a quantitative consideration for changes with moderate impact, rather than simply discarding information.

[0074] Based on the international standard PAS 2050-2011, "Life Cycle Greenhouse Gas Emissions Assessment Specification for Goods and Services," the threshold base point is set at 1% or 5% of the carbon footprint of the benchmark product. The threshold range and corresponding processing methods are as follows:

[0075] A baseline carbon footprint of less than 1% is set as the low-impact threshold range. If the carbon footprint difference falls within this range, it can be ignored or only qualitatively noted without adjusting the value. A baseline carbon footprint greater than or equal to 1% and less than 5% is set as the attention threshold range. If the carbon footprint difference falls within this range, it is not ignored but rather the rapid estimate is used as the final carbon footprint difference to calculate the carbon footprint of the product to be evaluated, without further detailed data collection. This data is recorded as a rapid estimate. A baseline carbon footprint greater than or equal to 5% is set as the action threshold. If the carbon footprint difference falls within this range, the standard calculation method is used: specific data is collected and precise calculations are performed.

[0076] Specifically, if the carbon footprint of benchmark product A is 100 kg CO2e, then 1% of the benchmark product's carbon footprint is equivalent to 1 kg CO2e, and 5% is equivalent to 5 kg CO2e. If product B replaces a small screw compared to product A, the calculated change is only 0.2 kg CO2e. Since 0.2 < 1, meaning it falls within the low impact threshold range, the difference is insignificant and can be ignored.

[0077] If the increment of product B compared to product A is 1.25 kg CO2e, where 1 < 1.25 < 5, indicating a certain difference but not a significant difference, the carbon footprint of the difference point obtained from the preliminary estimate can be used as the final carbon footprint of the difference point to calculate the carbon footprint of the product to be evaluated.

[0078] If the increase in product B compared to product A is 6 kg (CO2e): 6 > 5, and the difference is significant, then the difference needs to be imported into the benchmark carbon footprint model and the carbon footprint of the product to be evaluated needs to be accurately calculated using standard calculation methods.

[0079] Setting multiple thresholds ensures that computing resources are allocated to the areas with the greatest impact, while also quantifying changes with moderate impact, rather than simply discarding information.

[0080] In another embodiment, step 2 includes: inputting all activity data, emission factors, and calculation formulas involved in the carbon footprint calculation of all products of the enterprise to be estimated into the carbon footprint database; when calculating the enterprise's carbon footprint, using professional carbon accounting software or an automated calculation template developed based on Excel / database, automatically calling the corresponding data to integrate and calculate the report to obtain the enterprise's carbon footprint according to the organizational boundaries and accounting boundaries; and generating the corresponding enterprise carbon label based on the carbon footprint of the enterprise to be estimated.

[0081] The specific integration and calculation process includes:

[0082] Integration Scope 1 includes: reviewing the fuel purchase records and inventory records of the entire enterprise's finance or energy departments, and integrating and calculating the fuel combustion, industrial production processes, and fugitive emissions of the enterprise's own facilities; for example, how many liters of diesel and how many cubic meters of natural gas the enterprise consumed throughout the year;

[0083] The integrated formula is: Σ(fuel or material consumption × corresponding emission factor)

[0084] Integration Scope 2 includes: obtaining data on the company's total annual electricity bill (kWh) and total steam / heat purchase invoices to determine the company's consumption of purchased electricity, heat, and steam.

[0085] The integration formula is: Σ(purchased electricity / heat × corresponding emission factor)

[0086] Integration Scope 3 includes: acquiring emissions from upstream and downstream sources, such as raw material procurement, outsourced processing, employee travel, and waste disposal. Regarding carbon emissions from upstream raw materials, the upstream emissions portion of the raw material emissions from all product carbon footprints is retrieved from the carbon footprint database, and then proportionally amplified and adjusted according to the company's total annual procurement amount or volume. For example, the expenditure method can be used: the company's total annual expenditure in a certain category (such as raw materials or travel) × the industry's average carbon emission factor (yuan / ton CO2e). Other categories, such as employee commuting and waste disposal, are then aggregated from the corresponding data sources.

[0087] Finally, the corporate carbon label includes the carbon emission level and the main sources of the company's carbon emissions. The carbon emission level is usually expressed as a number or grade, such as AJ level, where A level indicates low carbon and J level indicates high carbon, and so on.

[0088] In another embodiment, in step 3, different industries are classified according to China's industry classification standard, namely the "National Industrial Classification of Economic Activities" (GB / T 4754-2017). This standard systematically classifies and codes all economic activities in society. Based on the carbon emissions and characteristics of different industries, the general situation of different industries is classified into high-energy-consuming industries, clean energy industries, consumer goods industries, transportation industries, and emerging industries. Specifically:

[0089] 1. High-energy-consuming industries refer to industries with high energy consumption and carbon emissions, mainly involving mining, raw material processing, and heavy industry. For example, ferrous metal smelting and rolling processing, and chemical raw materials and chemical products manufacturing. The corresponding carbon label may tend to disclose: energy structure, such as the proportion of coal power or green energy, emission reduction of CCUS projects, such as annual CO2 sequestration of tens of thousands of tons, process substitution progress, such as hydrogen reduction ratio.

[0090] The corresponding carbon emission reduction needs may involve: technological innovation, such as promoting hydrogen metallurgy and CCUS (carbon capture and storage) technology, with the goal of reducing carbon emissions per ton of steel per year; and energy substitution, such as increasing the proportion of green electricity use, building wind and solar power plants or participating in green electricity trading.

[0091] 2. The clean energy industry refers to industries engaged in the production and supply of renewable energy, nuclear energy, and clean energy, with low or zero carbon emissions. For example, solar power generation and wind power generation. The corresponding carbon label may tend to focus on: quantifying energy payback period and material recycling rate, such as the increased recycling rate of glass and aluminum frames, and emission reductions brought about by technological iteration, such as the annual CO2 reduction of 1 million tons for GW modules.

[0092] The corresponding carbon emission reduction needs may involve: decarbonization of the entire industrial chain, such as reducing the power consumption of polysilicon production and increasing the coverage of the module recycling system; and technological upgrading.

[0093] 3. The consumer goods industry refers to industries that directly provide goods and services to consumers, involving the manufacturing and sale of daily necessities. For example, the food manufacturing and textile industries may have carbon labeling practices that focus on disclosing transportation carbon emissions, the carbon footprint of packaging materials, and the impact of consumer behavior.

[0094] Corresponding carbon emission reduction needs may involve: supply chain optimization, such as shortening the transportation range of dairy cold chain or adopting bio-based packaging; and consumer guidance: promoting low-carbon consumption behaviors, such as encouraging the recycling of old clothes and reducing food waste.

[0095] 4. The transportation industry refers to industries engaged in freight and passenger transport, warehousing, and postal services. For example, the carbon labeling of air transport and water transport may focus on: fuel type, operational efficiency improvements, such as emission reduction ratio per unit of freight turnover, and progress in alternative fuel research and development.

[0096] The corresponding carbon emission reduction needs may involve: fuel transition, such as increasing the proportion of sustainable aviation fuel blending in the aviation industry and promoting green methanol and ammonia fuels in the shipping industry; and operational optimization, such as reducing ship speeds to lower fuel consumption and optimizing flight routes.

[0097] 5. Emerging industries refer to high-tech industries and modern service industries centered on new technologies and new business models, possessing innovation and growth potential. For example, the internet and related services, computer, communication and other electronic equipment manufacturing industries may have carbon labeling tendencies based on: raw material sources such as aluminum smelted using green electricity, product energy efficiency ratings such as ENERGY STAR certification, and contributions to the circular economy such as the rate of refurbishment and reuse of old equipment.

[0098] The corresponding carbon emission reduction needs may involve: material innovation, such as developing rare earth-free permanent magnets and biodegradable electronic packaging materials to reduce dependence on high-carbon raw materials; and energy efficiency improvement, such as reducing the power consumption ratio of data centers and improving the energy efficiency ratio of AI chip computing power.

[0099] In another embodiment, step 4 includes: after determining the carbon emission reduction needs of different industries from step 3, determining the priority of carbon emission reduction needs by combining the current technological status of the industries and the estimated implementation costs, and then ranking the enterprises for carbon emission reduction by combining the enterprise carbon labels obtained in step 2, and taking the top n% in the ranking as the carbon emission reduction planning targets.

[0100] Specifically, the technology status assessment involves surveying available carbon reduction technologies for each industry and evaluating their technological maturity, adoption rate, and potential for improvement. For example, renewable energy technologies are relatively mature in the power industry, while CCUS is still in the demonstration stage in the industrial sector. Barriers to technology diffusion are analyzed, such as infrastructure requirements and the difficulty of technology transfer.

[0101] Among them, technology maturity is assessed through the Technology Readiness Level (TRL) and is divided into 1-9 levels;

[0102] The dimensions for assessing market penetration include market penetration indicators, regional distribution analysis, and industry chain maturity.

[0103] The evaluation dimensions for improvement potential include performance enhancement potential, cost reduction potential, iteration and integration potential, and resource and environmental constraints;

[0104] Cost estimation: Calculate the cost of implementing emission reduction technologies, including initial investment, operation and maintenance costs, financing costs, etc. Use cost-benefit analysis (CBA) or levelized cost of emissions reduction (LCOA) to quantify the cost of reducing CO2 emissions per ton. Consider external costs, such as environmental benefits and social costs, and carbon pricing (carbon tax or carbon trading price) can also be used for calculation.

[0105] The estimation process includes: listing all cost items for initial investment, operation and maintenance, and financing; assigning values ​​to each cost through supplier quotations, historical data, industry databases, etc.; building a life cycle cost model in a spreadsheet, incorporating time factors and discount rates to derive LCOA, NPV, payback period, etc.; and testing the impact of changes in key variables (such as energy prices, carbon prices, and interest rates) on the results.

[0106] Determine industry emission reduction priorities: Using a multi-criteria decision analysis method, combined with information on the current state of technology and cost estimation results, score and rank the urgency of carbon emission reduction needs of each industry to generate an industry carbon emission reduction priority sequence;

[0107] The enterprise ranking process begins by initially ranking enterprises within the same industry based on their carbon emission levels as indicated on their carbon labels, prioritizing enterprises with lower carbon emission levels. Subsequently, the initial ranking is dynamically adjusted based on the industry's carbon reduction priority sequence and the main carbon emission source categories identified on the enterprises' carbon labels. The final ranking of enterprise carbon reduction actions is then output, with the specific formula for this dynamic adjustment being:

[0108]

[0109] Let i be the overall score of the i-th enterprise. Let i be the carbon emission level value of enterprise i. The priority weight of industry emission reduction needs corresponding to the main carbon emission sources of the i-th enterprise; the carbon emission level value is determined according to the carbon emission level, as described in this embodiment. and It is an integer between 1 and 10.

[0110] The calculation process for the priority weight of industry emission reduction demand includes: determining the industry carbon emission reduction demand that is most similar to the main source of carbon emissions in the enterprise's carbon label, determining its corresponding priority, and determining the corresponding weight according to the priority. For example, the weight range for high priority is 8-10, the weight range for general priority is 4-7, and the weight range for low priority is 1-3. The specific value is determined according to the granularity of the corresponding priority division.

[0111] Then, a comprehensive ranking list of all enterprises is generated based on their overall scores, with the ranking from highest to lowest indicating the urgency of emission reduction.

[0112] Choosing a threshold n%: The value of n is determined based on resource constraints (such as funding and regulatory capacity). For example, n=10% means selecting the top 10% of companies as the focus. The value of n can be determined through stakeholder negotiation or optimization model.

[0113] Identify planning targets: Extract the top n% of enterprises from the ranking list to form a carbon emission reduction planning target list.

[0114] In another embodiment, step 5 includes:

[0115] Step 5.1: Determine the product's carbon reduction level based on industry carbon reduction needs, the company's product carbon labeling, optimization costs, and returns.

[0116] Input data includes: industry carbon emission reduction needs, product carbon labeling, optimization costs, and optimization returns.

[0117] The industry's carbon emission reduction needs include industry emission reduction targets and priorities; optimization costs are the costs required to reduce the carbon footprint of products, including technological transformation costs, material substitution costs, certification costs, etc., calculated using life cycle cost analysis (LCCA); optimization returns include the benefits brought by reducing the carbon footprint of products, such as energy savings, carbon credit income, and improved market competitiveness, quantified as financial returns or non-financial benefits.

[0118] Determining the carbon emission reduction level of a product includes: calculating a carbon emission reduction potential score for each product using a comprehensive scoring method, as shown in the following formula:

[0119]

[0120] in, The carbon emission reduction potential score for the i-th enterprise's j-th product is given. Let the priority weight of industry emission reduction demand corresponding to the j-th type of product of the i-th enterprise be denoted. Let J be the carbon label value of the j-th type of product of the i-th enterprise. The return-cost ratio for the j-th type of product of the i-th enterprise is given by the following formula:

[0121]

[0122] in, For the optimized return corresponding to the j-th type of product of the i-th enterprise, Let $\frac{i}{j}$ be the optimized cost for the $j$-th type of product of the $i$-th enterprise.

[0123] Based on the emission reduction potential scores of all products, a grading threshold is determined to further determine the carbon emission reduction level of each product.

[0124] Products are categorized into different levels based on their scores (e.g., levels 1-5, with level 1 indicating the highest priority for emission reduction). Higher-level products indicate greater urgency and economic viability in emission reduction.

[0125] Output a list of product carbon emission reduction levels to guide internal emission reduction decisions within the company.

[0126] Step 5.2: Based on reinforcement learning, determine the company's optimal carbon reduction strategy according to factors such as product carbon reduction level, technology reserves, funding, and energy prices. Specifically:

[0127] State Space Setup: Define environmental state variables, including: product carbon emission reduction level, technology reserves, capital, energy prices, and other factors. Among these, the product carbon emission reduction level represents internal emission reduction needs; technology reserves refer to the company's existing technological capabilities, such as energy efficiency technologies and renewable energy equipment, expressed numerically or in levels; capital refers to the company's available funds, including cash flow and financing capabilities; energy prices include electricity and fossil fuel prices, affecting operating costs; other factors include market demand and competitor actions.

[0128] Setting up action space: Define the emission reduction actions that enterprises can take, including: investing in new technologies: purchasing high-efficiency equipment, introducing clean energy and installing solar panels; process optimization: improving production processes and reducing waste; carbon offsetting: purchasing carbon credits and participating in carbon trading; product adjustment: switching to low-carbon raw materials and developing low-carbon products.

[0129] Set a reward function: Design it to maximize long-term net profit or minimize total cost. The reward function is:

[0130] Reward = (Carbon emission reduction benefits + Cost savings + Policy incentives) - (Investment costs + Operating costs)

[0131] Carbon emission reduction benefits include carbon trading revenue and avoided carbon taxes; cost savings include savings from improved energy efficiency; and policy incentives include subsidies or tax breaks.

[0132] Consider a time discount factor to reflect the present value of future returns.

[0133] Reinforcement learning process:

[0134] The Deep Q-Network (DQN), which is suitable for continuous states and action spaces, is selected. The training process includes:

[0135] Initialization: Random strategy or based on historical data.

[0136] Interaction: Simulate the interaction between the firm and its environment (policy, market), take action, and observe new states and rewards.

[0137] Update: Learn through trial and error, adjusting the strategy to maximize cumulative rewards. Training data can be derived from historical cases or generated through simulation.

[0138] Convergence: Stop when the policy is stable or the maximum number of iterations is reached.

[0139] Outputting the optimal strategy: The reinforcement learning model recommends a series of actions, such as investing in energy efficiency technologies in the first year and participating in carbon trading in the second year, taking into account dynamic factors such as policy changes and energy price fluctuations. The specific strategy output is a timetable and corresponding resource allocation plan.

[0140] In another embodiment, step 6 includes: setting carbon reduction targets for different periods based on the enterprise's optimal carbon reduction strategy, classifying and managing the enterprise according to the carbon reduction targets, and determining whether the enterprise meets the carbon reduction requirements.

[0141] Enterprises develop detailed plans based on recommended strategies and set KPIs (such as emission reductions and costs); they continuously monitor environmental changes and use reinforcement learning models to adjust strategies, achieving adaptive management. Through industry demand analysis and techno-economic assessment, key enterprises for carbon emission reduction are systematically identified, ensuring a focus on companies with high current emission reduction potential and low costs.

[0142] Specifically, examples of hierarchical management for enterprises in different industries are shown in Table 1:

[0143] Table 1. Example 1 of Hierarchical Management

[0144]

[0145] Among them, Company A, as a steel company in the ferrous metal smelting and rolling processing industry, which is a high-energy-consuming industry, adopted the blast furnace gas recovery system upgrade in the first year and set a carbon emission reduction target of 50,600 tons of CO2e / year. In the second year, it adopted the intelligent scheduling optimization strategy and set a carbon emission reduction target of 32,500 tons of CO2e / year. In the third year, it adopted carbon market quota trading and set a carbon emission reduction target of 23,100 tons of CO2e / year.

[0146] Company B, an air transport company in the transportation industry, adopted an optimized flight operation plan in the first year, setting a carbon emission reduction target of 18,800 tons of CO2e / year. In the second year, it adopted the promotion of sustainable aviation fuel, setting a carbon emission reduction target of 35,200 tons of CO2e / year. In the third year, it adopted more efficient engines, setting a carbon emission reduction target of 40,400 tons of CO2e / year.

[0147] Company C, a solar power generation company in the clean energy industry, upgraded its photovoltaic cell technology in the first year and set a carbon emission reduction target of 34,200 tons of CO2e / year. In the second year, it optimized operation and maintenance management and upgraded equipment and set a carbon emission reduction target of 22,100 tons of CO2e / year. In the third year, it increased its installed capacity and set a carbon emission reduction target of 16,400 tons of CO2e / year.

[0148] The carbon reduction levels are adjusted in real time based on the relative performance of each company's carbon reduction targets for each year. In the first year, Company A's carbon reduction target is among the top 30% of all carbon reduction targets for that year, so it is managed as Level 1. Company C's carbon reduction target is between 30% and 70% of all carbon reduction targets for that year, so it is managed as Level 2. Company B's carbon reduction target is among the bottom 30% of all carbon reduction targets for that year, so it is managed as Level 3. Similarly, in the second year, Company B is managed as Level 1, Company A as Level 2, and Company C as Level 3. In the third year, Company B is managed as Level 1, Company A as Level 2, and Company C as Level 3.

[0149] In another embodiment, the difference from the previous embodiment lies in the different grading method for the carbon emission reduction targets of enterprises, as shown in Table 2.

[0150] Table 2 Example 2 of Hierarchical Management

[0151]

[0152] The real-time adjustment classification is determined based on the threshold range of each company's carbon emission reduction target for each year. In the first year, company A's carbon emission reduction target is in the level 1 range [4-6], so it is managed as level 1; company B's carbon emission reduction target is in the level 3 range [0-2), so it is managed as level 3; and company C is in the level 2 range [2-4), so it is managed as level 2. Similarly, in the second year, company A, company B, and company C are all managed as level 2. In the third year, company A is managed as level 2, company B is managed as level 1, and company C is managed as level 3.

[0153] The two embodiments above mainly illustrate a dynamic classification approach. The specific classification method and classification parameter values ​​can be determined according to the regulatory capacity, the number of enterprises participating in the classification, and the distribution of carbon emission reduction targets.

[0154] The embodiments described above merely illustrate specific implementation methods of this application, and while the descriptions are detailed and specific, they should not be construed as limiting the scope of protection of this application. It should be noted that those skilled in the art can make various modifications and improvements without departing from the concept of the technical solution of this application, and these modifications and improvements all fall within the scope of protection of this application.

Claims

1. A carbon tag-driven carbon emission reduction classification method, characterized in that, Includes the following steps: Step 1: Select a company to be valued, choose a benchmark product for the company to be valued, calculate the carbon footprint of the benchmark product, establish a benchmark carbon footprint model, compare the product to be valued with the benchmark product, obtain the differences, import the differences into the benchmark carbon footprint model to obtain the carbon footprint of the product to be valued, and generate the corresponding carbon label. Step 2: Repeat Step 1 until all carbon labels of all products of all companies to be estimated are obtained. Based on the data collected during the calculation of the carbon footprint of all products of the companies to be estimated and other carbon emission-related data, sort out and calculate the carbon footprint of the companies to be estimated, and determine the corresponding carbon label of the companies to be estimated based on the carbon footprint of the companies to be estimated. Step 3: Repeat step 2 until all carbon labels of the companies to be evaluated are obtained. Divide all the companies to be evaluated into industries, analyze the distribution of carbon labels of companies in different industries, and then determine the carbon emission reduction needs of the industry based on the characteristics of the industry. Step 4: Based on the industry's carbon emission reduction needs and the carbon labels of all companies to be assessed, preliminarily determine the companies with carbon emission reduction plans; Step 5: Comprehensively determine the carbon emission reduction level of each carbon emission reduction planning enterprise's product carbon label, the carbon emission reduction demand of the corresponding industry, the optimization cost and optimization return of the carbon emission reduction technology. Based on reinforcement learning, determine the optimal carbon emission reduction strategy of the carbon emission reduction planning enterprise at present according to the product carbon emission reduction level, technology reserves, funds and energy prices. Step 6: Classify and manage enterprises according to the carbon reduction targets involved in the enterprise's optimal carbon reduction strategy, and determine whether the carbon reduction requirements are met; Step 1 includes: Step 1.1: Calculate the carbon footprint of the benchmark product and establish a benchmark carbon footprint model based on the calculation results. The benchmark carbon footprint model includes the carbon footprint and total carbon footprint at different life cycle stages. The different life cycle stages include the material preparation stage, production stage, transportation stage, and usage stage. Step 1.2: Compare the product to be evaluated with the benchmark product to obtain the differences between the product to be evaluated and the benchmark product at different stages of the product life cycle. The differences include material differences in the material preparation stage, process differences in the production stage, transportation differences in the transportation stage, and energy efficiency differences in the usage stage. Step 1.3: Import the differences into the baseline carbon footprint model, call the emission factors in the database to calculate the carbon footprint of the product to be evaluated, and generate the corresponding product carbon label based on the carbon footprint of the product to be evaluated. Step 5, determining the product's carbon reduction level, includes: calculating the carbon reduction potential score for each product using a comprehensive scoring method, as shown in the following formula: ； in, The carbon emission reduction potential score for the i-th enterprise's j-th product is given. Let the priority weight of industry emission reduction demand corresponding to the j-th type of product of the i-th enterprise be denoted. Let J be the carbon label value of the j-th type of product of the i-th enterprise. The return-cost ratio for the j-th type of product of the i-th enterprise is given by the following formula: ； in, For the optimized return corresponding to the j-th type of product of the i-th enterprise, Let $\frac{i}{j}$ be the optimized cost for the $j$-th type of product of the $i$-th enterprise. Based on the emission reduction potential scores of all products, a grading threshold is determined to further determine the carbon emission reduction level of each product. Step 5 determines the company's optimal carbon reduction strategy, including: Set up the state space: Define environmental state variables, including: product carbon emission reduction level, technology reserves, funds, energy prices and other factors; Setting action space: Define the emission reduction actions that enterprises take, including: investing in new technologies, process optimization, carbon offsetting, and product adjustment; Set a reward function: Design it to maximize long-term net benefits or minimize total costs; The optimal carbon reduction strategy for enterprises is determined through reinforcement learning based on the set state space, action space, and reward function.

2. The carbon tag-driven carbon emission reduction classification method according to claim 1, characterized in that, Step 1.3 includes: calling emission factors from the emission factor database to quickly estimate the carbon footprint of the difference point; if the carbon footprint of the difference point reaches the threshold, it is imported into the benchmark carbon footprint model and the carbon footprint of the product to be evaluated is calculated using standard calculation methods.

3. The carbon tag-driven carbon emission reduction classification method according to claim 2, characterized in that, The differential carbon footprint thresholds include multiple types, with different threshold ranges corresponding to different processing methods. Using a% or b% (b>a) of the benchmark product's carbon footprint as the threshold base point, the threshold ranges and corresponding processing methods are set as follows: A baseline carbon footprint less than a% is set as the low-impact threshold range. If the carbon footprint of the difference point falls within this range, it is either ignored or qualitatively noted without adjusting the value. A baseline carbon footprint greater than or equal to a% and less than b% is set as the attention threshold range. If the carbon footprint of the difference point falls within this range, it is not ignored directly, but the rapid estimate is used as the final carbon footprint of the difference point and included in the calculation to obtain the carbon footprint of the product to be evaluated. At the same time, this data is recorded and marked as a rapid estimate. A baseline carbon footprint greater than or equal to b% is set as the action threshold. If the carbon footprint of the difference point falls within this range, the standard calculation method is used: data is collected and calculated.

4. The carbon tag-driven carbon emission reduction classification method according to claim 1, characterized in that, Step 2 includes: entering the activity data, emission factors, and calculation formulas involved in the carbon footprint calculation of all products of the enterprise to be estimated into the carbon footprint database; when calculating the enterprise's carbon footprint, according to the organizational boundaries and accounting boundaries, calling the corresponding data to integrate and calculate to generate a report to obtain the enterprise's carbon footprint; and generating the corresponding enterprise carbon label based on the carbon footprint of the enterprise to be estimated.

5. The carbon tag-driven carbon emission reduction classification method according to claim 1, characterized in that, The different industries mentioned in step 3 are high-energy-consuming industries, clean energy industries, consumer goods industries, transportation industries, and emerging industries.

6. The carbon tag-driven carbon emission reduction classification method according to claim 1, characterized in that, Step 4 includes: after determining the carbon emission reduction needs of different industries from step 3, determining the priority of carbon emission reduction needs by combining the current technological status of the industries and the estimated implementation costs, and then ranking the enterprises to be evaluated for carbon emission reduction by combining the enterprise carbon labels obtained in step 2, and selecting enterprises to plan for carbon emission reduction from the ranked enterprises to be evaluated.

7. The carbon tag-driven carbon emission reduction classification method according to claim 1, characterized in that, Step 4 includes: Step 4.1: Conduct a technology status assessment. For each target industry, investigate the carbon emission reduction technologies currently in use, obtain the technology maturity level, current popularity and future improvement potential of each technology, and establish an industry-level carbon emission reduction technology database. Step 4.2: Perform cost estimation. For each carbon reduction technology, collect and calculate the initial investment cost, operation and maintenance costs, and financing-related costs required for its implementation. Based on the cost data, quantify the comprehensive implementation cost corresponding to a unit carbon dioxide equivalent (tCO2e) emission reduction. Step 4.3: Determine industry emission reduction priorities. Using a multi-criteria decision analysis method, and combining the current technology status information obtained in Step 4.1 and the cost estimation results obtained in Step 4.2, score and rank the urgency of carbon emission reduction needs of each industry to generate an industry carbon emission reduction priority sequence. Step 4.4: Implement enterprise ranking. First, based on the carbon emission levels indicated on the carbon labels of the enterprises to be evaluated within the same industry, a preliminary ranking of the enterprises to be evaluated is performed, with enterprises having lower carbon emission levels given priority. Subsequently, according to the industry carbon emission reduction priority sequence and in conjunction with the carbon emission source categories identified on the carbon labels of the enterprises to be evaluated, the preliminary ranking results are dynamically adjusted to output the final priority ranking of the carbon emission reduction actions of the enterprises to be evaluated. The specific formula for the dynamic adjustment is as follows: ； Let i be the overall score of the i-th enterprise. Let i be the carbon emission level value for enterprise i. The priority weight of industry emission reduction demand corresponding to the carbon emission source of the i-th enterprise; The calculation process for the priority weight of industry emission reduction demand includes: determining the industry carbon emission reduction demand that is most similar to the carbon emission source in the carbon label of the company to be evaluated, determining its corresponding priority, and determining the corresponding weight based on the priority. A comprehensive ranking list of all companies to be evaluated is generated based on their overall scores, with the ranking from highest to lowest indicating the urgency of emission reduction. Step 4.5: Identify the planning targets, extract enterprises from the ranking list, and form a list of carbon emission reduction planning targets.

Images