Method and system for quantizing and dynamically optimizing full-life-cycle carbon footprint of transformer substation

Abstract

The invention provides a transformer substation full life cycle carbon footprint quantification and dynamic optimization method and system, and relates to the technical field of power system low carbon. The method comprises the following steps: acquiring construction period data of a plurality of transformer substations, inputting the construction period data into a carbon efficiency evaluation model, respectively acquiring a carbon efficiency evaluation result of each transformer substation, and formulating a corresponding investment structure adjustment scheme for each transformer substation according to the corresponding carbon efficiency evaluation result; the carbon efficiency evaluation model is constructed based on a data envelope analysis model and a Tobit regression model; collecting operation period data of each transformer substation, and inputting the operation period data into the dynamic attribution model to obtain marginal contribution degree data; the dynamic attribution model is constructed based on an LMDI decomposition algorithm; and performing clustering analysis on the carbon efficiency evaluation result and the marginal contribution degree data of each transformer substation by adopting a K-means clustering algorithm, performing carbon efficiency comparison on the same type of transformer substations, and generating a carbon emission reduction improvement scheme of each transformer substation. According to the invention, a carbon emission reduction improvement scheme conforming to the actual condition of each transformer substation can be generated.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of low-carbonization of power systems, and particularly relates to a method and system for quantifying and dynamically optimizing carbon footprint of a substation in a whole life cycle. BACKGROUND

[0002] Under the background of the global energy system accelerating towards low-carbonization and intelligentization, the carbon footprint management of the power system, as the core bearing field of carbon neutralization, is extending from a single link to the whole life cycle. As a key node of the power grid, the carbon emission of the substation involves multiple stages such as civil construction, high-load operation, sulfur hexafluoride management, equipment retirement and recycling, and is affected by multiple dynamic factors such as regional energy structure, power grid technology iteration, and new energy volatility grid connection. In recent years, international research trends show that carbon emission accounting has shifted from the traditional “operation stage dominated” to “full chain integration”. For example, the European Union's “Product Environmental Footprint” guidelines clearly propose that infrastructure projects need to cover the complete life cycle “from cradle to regeneration”. At the same time, with the popularization of digital twinning and Internet of Things technology, dynamic tracking and real-time optimization of carbon emissions become possible, and the academic community begins to explore the combination of equipment energy efficiency degradation models, power grid loss time sequence characteristics and carbon flow analysis to achieve fine spatio-temporal analysis of carbon footprint. The existing carbon footprint research presents three aspects of deepening trends: first, from “single-stage accounting” to “full life cycle integration”; second, from “static list analysis” to “dynamic simulation prediction”, such as combining Monte Carlo simulation to quantify the nonlinear impact of equipment aging on carbon emissions; third, from “isolated carbon accounting” to “energy-carbon collaborative optimization”, by coupling data envelopment analysis, decomposition model and other methods to tap the potential of carbon efficiency improvement.

[0003] At present, the research on the carbon footprint of substations still has some defects. The existing research focuses on the static accounting of direct emissions in the operation stage, lacks an integrated framework for the systematic quantification of construction period implicit carbon, operation period carbon absorption and scrap period recycling carbon, and traditional methods cannot couple dynamic factors such as equipment technology evolution, energy structure change and operation efficiency degradation. For example, the current mainstream Life Cycle Assessment (LCA) model often uses a fixed emission factor library, ignoring the influence of regional power grid carbon intensity spatial and temporal differences on network loss carbon emission calculation. Efficiency analysis is mostly based on the CCR model of cross-sectional data, which cannot reveal the longitudinal evolution law of the whole life cycle efficiency of the substation, leading to the disconnection between carbon emission reduction path and actual technology development. Under this background, it is urgent to build a method system covering the whole chain of “construction-operation-retirement”, integrating dynamic efficiency evaluation and continuous optimization feedback, to support the accurate management and step-by-step improvement of the carbon footprint of substations. This is not only a technical necessity for the low-carbon transformation of the power grid, but also a key path to realize the collaborative carbon reduction of the “source-grid-load-storage” of the new power system. SUMMARY

[0004] The embodiment of the present application provides a substation full life cycle carbon footprint quantification and dynamic optimization method and system to solve the problem of substation full life cycle dynamic optimization.

[0005] In the first aspect, the embodiment of the present application provides a substation full life cycle carbon footprint quantification and dynamic optimization method, comprising: Construction period data of a plurality of substations are acquired, the construction period data is input into a carbon efficiency evaluation model, carbon efficiency evaluation results of the substations are acquired respectively, and corresponding investment structure adjustment schemes are formulated for the substations according to the corresponding carbon efficiency evaluation results; wherein the carbon efficiency evaluation model is constructed based on a data envelopment analysis model and a Tobit regression model; Operation period data of the substations are collected, the operation period data is input into a dynamic attribution model, and corresponding marginal contribution degree data of the substations are acquired; wherein the operation period data includes carbon dioxide emissions of sulfur hexafluoride in a use process and carbon dioxide emissions caused by power grid loss; the marginal contribution degree data includes marginal contribution degrees of an energy structure, an energy efficiency, a load size, a technical path and equipment energy efficiency to carbon emission reduction of the substations in the operation period; and the dynamic attribution model is constructed based on a Log Mean Divisia Index (LMDI) decomposition algorithm; The carbon efficiency evaluation results and the marginal contribution degree data of the substations are analyzed by using a K-means clustering algorithm, carbon efficiency comparisons are performed on substations of the same type, and carbon emission reduction improvement schemes of the substations are generated.

[0006] In the second aspect, the embodiment of the present application provides a substation full life cycle carbon footprint quantification and dynamic optimization system, characterized in that the system comprises Internet of Things sensing devices, a Supervisory Control And Data Acquisition (SCADA) system and a substation carbon efficiency intelligent management device; wherein the substation carbon efficiency intelligent management device is in communication connection with the Internet of Things sensing devices and the SCADA system respectively; The SCADA system is used for collecting data in construction, operation and scrapping stages of a substation; The substation carbon efficiency intelligent management device is used for executing the method in the first aspect or any possible implementation manner of the first aspect.

[0007] In the third aspect, the embodiment of the present application provides an electronic device comprising a memory and a processor, the memory stores a computer program, and the processor implements the method in the first aspect or any possible implementation manner of the first aspect when executing the computer program.

[0008] In this embodiment of the invention, by acquiring construction period data from multiple substations and inputting it into a carbon efficiency evaluation model constructed based on a data envelopment analysis model and a Tobit regression model, the carbon efficiency level of each substation during its construction phase can be accurately assessed and key influencing factors can be derived. This allows for targeted investment structure adjustment plans, addressing the difficulty of quantifying carbon efficiency and attribution during the construction phase using traditional methods. By collecting operation period data and inputting it into a dynamic attribution model based on the LMDI decomposition algorithm, the marginal contribution of five dimensions, including energy structure and energy efficiency, to carbon emission reduction during the operation period can be clearly analyzed, enabling precise tracing of carbon emission drivers during the operation period. Through K-means clustering analysis of the carbon efficiency evaluation results and marginal contribution data, substations can be divided into groups with similar characteristics and their carbon efficiency can be compared, avoiding a "one-size-fits-all" optimization approach. Ultimately, this generates carbon emission reduction improvement plans that conform to the actual conditions of each substation, forming a carbon footprint quantification and optimization system covering the "construction-operation" phase. Attached Figure Description

[0009] Figure 1 This is a flowchart illustrating the implementation of a method for quantifying and dynamically optimizing the carbon footprint of a substation throughout its entire lifecycle, provided in an embodiment of the present invention. Figure 2 This is a flowchart illustrating the implementation of a method for quantifying and dynamically optimizing the carbon footprint of a substation throughout its entire lifecycle, provided in another embodiment of the present invention. Figure 3 This is a carbon footprint calculation list for the substation construction phase, as described in one embodiment of the present invention. Figure 4 This is an architectural diagram of a carbon efficiency intelligent management device in one embodiment of the present invention; Figure 5 This is a schematic diagram of the structure of a substation full life cycle carbon footprint quantification and dynamic optimization system provided in an embodiment of the present invention; Figure 6 This is a schematic diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0010] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0011] Figure 1 This is a flowchart illustrating the implementation of the substation full lifecycle carbon footprint quantification and dynamic optimization method provided in this embodiment of the invention. Figure 1 As shown, it includes the following steps: S101. Obtain construction period data for multiple substations, input the construction period data into the carbon efficiency evaluation model, obtain the carbon efficiency evaluation results for each substation, and formulate corresponding investment structure adjustment plans for each substation based on the corresponding carbon efficiency evaluation results; wherein, the carbon efficiency evaluation model is constructed based on the data envelopment analysis model and the Tobit regression model.

[0012] The execution subject of each embodiment of this application can be a server, processor, microprocessor, or other device with data processing capabilities. In actual implementation, the specific implementation method of the execution subject can be selected according to actual needs. This embodiment does not impose any special restrictions on this, as long as it is a device with data processing capabilities.

[0013] The purpose of inputting construction period data into the carbon efficiency evaluation model is to screen out comparable substations.

[0014] In one possible implementation, the data envelopment analysis model is the SBM-DEA model; construction period data is input into the carbon efficiency evaluation model to obtain the carbon efficiency evaluation results for each substation, including: The construction period data is input into the carbon efficiency evaluation model, and the SBM-DEA model is used to evaluate the carbon efficiency of the substation during the construction period and calculate the static carbon efficiency value of each substation. Based on static carbon efficiency values, the Tobit regression model is used to analyze the impact of carbon emission reduction special budget ratio, R&D investment as a percentage of revenue, policy compliance investment ratio, digital substation penetration rate, and low-carbon technology investment concentration on carbon efficiency, generating regression coefficients and significance levels.

[0015] This implementation method employs the SBM-DEA model, inputting construction period data and considering input-output slack variables to accurately calculate the static carbon efficiency value. This model better reflects resource input redundancy and carbon emission control than the traditional DEA model. Based on the static carbon efficiency value, a Tobit regression model is used to analyze the impact of five types of factors (such as the proportion of special carbon emission reduction budget) on carbon efficiency, outputting regression coefficients and significance levels. This quantifies the marginal contribution and significance of each factor, addressing the problem that traditional methods cannot identify key driving factors and providing data support for investment structure adjustment.

[0016] In one possible implementation, the independent variables of the Tobit regression model include: the proportion of special budget for carbon emission reduction, the proportion of R&D investment to revenue, the proportion of policy compliance investment, the penetration rate of digital substations, and the concentration of low-carbon technology investment. Among them, the carbon emission reduction special budget ratio is the percentage of the power grid company's investment in carbon emission reduction technology as a percentage of its total annual investment; the R&D investment as a percentage of revenue is the percentage of the power grid company's R&D investment in substation-related technologies as a percentage of its total revenue; the policy compliance investment ratio is the percentage of the power grid company's additional investment due to environmental policies as a percentage of its total investment; the digital substation penetration rate is the percentage of digital substations already in operation as a percentage of all substations; and the low-carbon technology investment concentration is the multiple of the company's investment in low-carbon technologies for substations as a percentage of the industry average.

[0017] This implementation defines five independent variables for the Tobit regression model: the proportion of special carbon reduction budget refers to the proportion of investment in carbon reduction technologies; the proportion of R&D investment to revenue refers to substation technology R&D; the proportion of policy compliance investment relates to investment driven by environmental policies; the penetration rate of digital substations reflects the degree of intelligent application; and the concentration of low-carbon technology investment reflects the industry's leading level. These definitions ensure consistent data collection standards, avoid regression bias, and lay the foundation for quantifying the marginal contribution of each factor to carbon efficiency during the construction period.

[0018] In one possible implementation, a corresponding investment structure adjustment plan is formulated for each substation based on the corresponding carbon efficiency evaluation results, including: Based on the regression coefficients output by the Tobit regression model, the marginal contribution of each independent variable to carbon efficiency is analyzed. Develop differentiated investment adjustment strategies based on factors such as the proportion of carbon emission reduction budget, the proportion of R&D investment to revenue, the proportion of policy compliance investment, the penetration rate of digital substations, and the concentration of low-carbon technology investment. These investment adjustment strategies include one or more of the following: increasing investment in low-carbon technologies, increasing the proportion of R&D investment, and optimizing the allocation of policy compliance investment.

[0019] In this approach, the regression coefficients output by the Tobit regression model can be used to intuitively analyze the marginal contribution of individual variables such as the proportion of carbon emission reduction budget and the proportion of R&D investment to revenue to carbon efficiency. Based on this, differentiated investment adjustment strategies can be formulated for five types of factors, such as increasing investment in low-carbon technologies, increasing the proportion of R&D investment, or optimizing the allocation of policy-compliant investments. This avoids the blindness of traditional investment adjustments and directs investment resources toward areas that contribute more significantly to improving carbon efficiency. It ensures that every investment can be efficiently converted into an improvement in carbon efficiency during the construction period, while also aligning with the power grid company's "carbon peaking and carbon neutrality" strategic plan, and promoting the coordinated advancement of carbon efficiency optimization and strategic goals during the construction phase.

[0020] S102, collect operational data from each substation, input the operational data into the dynamic attribution model, and obtain the marginal contribution data corresponding to each substation; among which, the operational data includes: carbon dioxide emissions from sulfur hexafluoride during use and carbon dioxide emissions caused by grid losses; the marginal contribution data includes the marginal contribution of energy structure, energy efficiency, load scale, technology path, and equipment energy efficiency to carbon emission reduction during the substation's operation period; the dynamic attribution model is constructed based on the LMDI decomposition algorithm.

[0021] In one possible implementation, operational data is input into a dynamic attribution model to obtain the marginal contribution data for each substation, including: The operational data is divided into base period data and reporting period data to obtain data on the energy structure, energy efficiency, load scale, technology path and equipment energy efficiency indicators of each substation in the base period and reporting period. The additive form of the LMDI decomposition algorithm is used to decompose carbon emission changes into basic carbon emission factor effects, clean energy structure effects, grid loss efficiency effects, load intensity effects, management effects, and equipment loss effects.

[0022] This implementation divides operational period data into base period and reporting period data. By comparing the two periods, the range of carbon emission changes during the operational period can be clearly defined. Simultaneously, it acquires data on five categories of indicators, including energy structure and energy efficiency, for both the base and reporting periods, providing complete time-series data support for subsequent carbon emission change decomposition. The LMDI decomposition algorithm, using an additive approach, decomposes carbon emission changes into six types of effects, including the basic carbon emission factor effect and the clean energy structure effect. This allows for a direct quantification of the absolute contribution of each effect to carbon emission changes, addressing the problem that traditional methods cannot accurately analyze the intrinsic mechanisms of carbon emission changes during the operational period. This enables targeted identification of key improvement directions for carbon emission reduction during the operational period.

[0023] S103. The K-means clustering algorithm is used to perform cluster analysis on the carbon efficiency evaluation results and marginal contribution data of each substation, compare the carbon efficiency of similar substations, and generate carbon emission reduction improvement schemes for each substation.

[0024] In one possible implementation, the K-means clustering algorithm is used to perform cluster analysis on the carbon efficiency evaluation results and marginal contribution data of each substation, including: Obtain the carbon efficiency evaluation results and marginal contribution data of each substation to obtain initial clustering data; The initial clustering data is standardized, and the elbow rule and silhouette coefficient are used to determine the optimal number of clusters. The K-means clustering algorithm is then executed to generate the clustering results.

[0025] The initial clustering data includes: static carbon efficiency values ​​during the construction period calculated based on the SBM-DEA model, and the influence coefficients of basic carbon emission factors, clean energy structure effects, grid loss efficiency effects, load intensity effects, management effects, and equipment loss effects during the operation period quantified based on the LMDI model. The data is standardized, and the elbow rule and silhouette coefficient are used to determine the optimal number of clusters. The K-means clustering algorithm is then executed to generate the clustering results.

[0026] In this implementation, the clustering data sources simultaneously cover the core carbon efficiency characteristics of both the construction and operation phases. The static carbon efficiency value during the construction phase reflects the overall carbon efficiency level, while the influence coefficients such as the basic carbon emission factor and the clean energy structure effect during the operation phase reflect the carbon emission-driven characteristics of the operation phase. This ensures that the clustering analysis comprehensively covers the key carbon efficiency dimensions throughout the substation's entire lifecycle. Data standardization eliminates the impact of differences in the units of measurement of different indicators, avoiding clustering bias caused by different indicator units. The elbow rule is used to analyze the total squared error inflection point and the silhouette coefficient to assess intra-cluster tightness and inter-cluster separation, scientifically determining the optimal number of clusters and avoiding inaccurate classification caused by subjectively setting the number of clusters. Finally, the clustering results generated by the K-means clustering algorithm can divide substations into groups with similar carbon efficiency characteristics, providing a reasonable classification basis for subsequent comparisons of carbon efficiency in similar substations and the formulation of differentiated improvement plans.

[0027] In one possible implementation, carbon efficiency comparisons are performed on similar substations to generate carbon reduction improvement plans for each substation, including: For substations with high carbon efficiency, promote the standardized application of modular construction technology and carbon asset development, and integrate carbon capture technology with carbon quota trading. For substations with medium carbon efficiency, implement a clean energy penetration rate improvement plan, deploy intelligent reactive power compensation devices and tiered electricity price incentive mechanisms; For substations with low carbon efficiency, implement a system to eliminate high-energy-consuming equipment, force the replacement of inefficient transformers, and apply dynamic power flow optimization algorithms to reduce the load peak-valley difference.

[0028] This approach involves developing differentiated carbon reduction improvement plans for substations with different carbon efficiency clusters. For high-carbon-efficiency substations, promoting the standardized application of modular construction processes can further consolidate their carbon efficiency advantages during the construction phase. Developing carbon assets and integrating carbon capture technology and carbon quota trading can unlock their low-carbon value potential. For medium-carbon-efficiency substations, implementing a clean energy penetration rate improvement plan can directly reduce fossil fuel dependence during operation. Deploying intelligent reactive power compensation devices and tiered electricity pricing incentive mechanisms can optimize energy utilization efficiency and load regulation, promoting a leap in carbon efficiency from intermediate to high levels. For low-carbon-efficiency substations, implementing a system for eliminating high-energy-consuming equipment and mandating the replacement of inefficient transformers can quickly reduce carbon emissions from high-energy-consuming links. Applying dynamic power flow optimization algorithms to compress load peak-valley differences can reduce energy waste during operation, effectively improving the current low-carbon-efficiency status, avoiding the drawbacks of "one-size-fits-all" optimization, and promoting a tiered improvement in the overall carbon efficiency level of substations.

[0029] In this embodiment, by acquiring construction-phase data from multiple substations and inputting it into a carbon efficiency evaluation model constructed based on a data envelopment analysis (DEA) model and a Tobit regression model, the carbon efficiency level of each substation during its construction phase can be accurately assessed and key influencing factors can be derived. This allows for targeted investment structure adjustment plans, addressing the difficulty of quantifying carbon efficiency and attribution during the construction phase using traditional methods. By collecting operational-phase data and inputting it into a dynamic attribution model based on the LMDI decomposition algorithm, the marginal contribution of five dimensions, including energy structure and energy efficiency, to carbon emission reduction during the operational phase can be clearly analyzed, enabling precise tracing of carbon emission drivers during the operational phase. Furthermore, by using a K-means clustering algorithm to cluster the carbon efficiency evaluation results and marginal contribution data, substations can be divided into groups with similar characteristics and their carbon efficiency can be compared, avoiding a "one-size-fits-all" optimization approach. Ultimately, this generates carbon emission reduction improvement plans that conform to the actual conditions of each substation, forming a carbon footprint quantification and optimization system covering the "construction-operation" phase.

[0030] In one possible implementation, the input indicators of the SBM-DEA model include resource consumption inputs and technology and management inputs; among them, resource consumption inputs include the cost of dismantling and disposing of existing facilities, the cost of transitional power supply facilities, the deduction of recyclable resources, and the cost of construction resources; technology and management inputs include the cost of technology for reusing and upgrading existing facilities, the cost of construction risk management, the cost of community coordination, the cost of clean technology, the cost of environmental management, and the cost of personnel training. The output indicators of the SBM-DEA model include expected outputs and unexpected outputs; among them, expected outputs include power supply availability, transformer load rate, comprehensive line loss rate, voltage qualification rate and harmonic distortion rate; unexpected outputs include total carbon emissions during the construction period and the amount of construction waste generated.

[0031] In this embodiment, the method explicitly defines the input indicators of the SBM-DEA model as encompassing resource consumption and technology / management categories. Resource consumption inputs cover core resource consumption dimensions during the construction phase, such as the cost of dismantling and handling existing facilities and the investment in transitional power supply facilities. Technology / management inputs include technical and management elements affecting carbon efficiency, such as investments in the reuse and renovation of existing facilities and clean technology investments. This ensures that the input indicators comprehensively cover resource and technology management scenarios related to carbon efficiency during the construction phase. Output indicators distinguish between expected and unexpected outputs. Expected outputs include indicators reflecting the functional performance of substations, such as power availability and transformer load rate. Unexpected outputs include environmental impact indicators such as total carbon emissions and construction waste generation during the construction period. This achieves full-dimensional coverage of "resource input - functional output - environmental impact" in the carbon efficiency evaluation during the construction phase, avoiding the bias in carbon efficiency evaluation caused by the one-sidedness of traditional indicator systems and improving the accuracy and relevance of the SBM-DEA model's evaluation results.

[0032] Based on the above embodiments, this invention aims to provide a method and system for quantifying and dynamically optimizing the carbon footprint of substations throughout their entire lifecycle. Its core lies in constructing a multi-stage coupled analysis framework covering the construction, operation, and decommissioning phases. For carbon emissions during the construction phase, a carbon accounting model is established based on process analysis, encompassing building material production, construction energy consumption, and equipment transportation. This model integrates the SBM data envelopment model and Tobit regression technology. By comparing the resource input efficiency of substations of the same specifications horizontally and analyzing the efficiency differences across phases vertically, the contribution of technological advancements and management optimization to carbon efficiency improvement is separated. During the operation phase, grid loss carbon emissions and sulfur hexafluoride (SF6) emissions are incorporated into the carbon emission accounting system. A five-dimensional driving analysis framework of "energy structure - energy efficiency level - load fluctuation - technology path - equipment energy efficiency" is constructed using the LMDI decomposition model to quantify the marginal impact of factors such as increased renewable energy penetration, accelerated equipment aging and wear, and adjustments to reactive power compensation strategies on carbon emissions. Furthermore, based on the K-means clustering algorithm, the static carbon efficiency value calculated by the SBM model, and the impact coefficients of basic carbon emission factors, clean energy structure effects, network loss efficiency effects, load intensity effects, management effects, and equipment losses on carbon emissions during operation, quantified by the LMDI model, are used to dynamically classify and profile substations. This is combined with industry benchmarks to generate differentiated optimization schemes. Simultaneously, a closed-loop feedback mechanism of "monitoring-evaluation-optimization-verification" is designed. By collecting real-time equipment status, load curves, and recycled material recovery data, the cluster centers and optimization thresholds are periodically updated, driving the adaptive iteration of carbon efficiency management strategies. The system integrates BIM engineering data, SCADA operating parameters, and IoT monitoring terminals to achieve visualized tracking and intelligent decision support for the entire carbon flow chain. Ultimately, this forms a standardized and scalable substation carbon efficiency improvement path library, providing power grid companies with technical tools to accurately formulate "one-stop-one-policy" carbon reduction solutions.

[0033] To achieve the quantification and dynamic optimization of the carbon footprint of substations throughout their entire lifecycle, this invention is divided into the following objectives: 1) Constructing a phased quantification model of the carbon footprint of substations covering the entire lifecycle of construction, operation, and decommissioning, solving the problem of phase fragmentation in traditional methods; 2) Achieving accurate analysis of dynamic driving factors of carbon emissions, supporting targeted decision-making for carbon efficiency optimization; 3) Establishing a differentiated carbon management and control system based on dynamic clustering and closed-loop feedback, promoting the step-by-step improvement of the carbon efficiency level of substations.

[0034] The above embodiments provide an overview of the method for quantifying and dynamically optimizing the carbon footprint of a substation throughout its entire lifecycle. The following sections will provide a detailed description of the solutions provided in this application, taking into account equipment parameters and multi-source heterogeneous data from the construction and operation phases during specific implementation processes.

[0035] Figure 2 This diagram illustrates the framework of a substation lifecycle carbon footprint quantification and dynamic optimization method according to another embodiment of the present invention.Figure 2 As shown, the process includes several steps such as phased modeling of the entire life cycle, dynamic driving factor analysis, and closed-loop optimization to improve the carbon efficiency of substations.

[0036] Please refer to Figure 2 This includes steps S1-S7: S1. Select comparable substations and define the research period. Based on this, conduct data collection and preprocessing, specifically including S11-S13: S11. Select comparable substations based on their voltage level, main transformer capacity, and equipment type.

[0037] Optionally, the selected voltage levels cover medium-voltage transmission and distribution voltage levels as well as low-voltage distribution voltage levels, specifically including: 220kV: Suitable for connecting transmission and distribution networks, serving as an important intermediate link in the power system; 110kV: Used for connecting the transmission network and the distribution network, while also providing power to medium-capacity users; 66kV: Used in some areas, mainly for urban power distribution networks; 35kV: Suitable for power distribution networks in large cities, large enterprises, and rural areas; 10kV: Widely used in urban and rural power distribution networks.

[0038] The main transformer capacity of a substation refers to the total capacity of the main transformers in the substation, and it is a key indicator for measuring the substation's power supply capability. Main transformer capacity includes the capacity of a single main transformer and the total capacity of multiple main transformers. Substations of different voltage levels typically have different main transformer capacities.

[0039] Substation equipment types include air-insulated substations (AIS) and gas-insulated substations (GIS). AIS substations are often built in suburban or rural areas where land costs are low and equipment reliability requirements are not extreme. GIS substations are often built in urban centers, industrial areas, and other scenarios where land is scarce, environmental conditions are complex, and power supply reliability requirements are high.

[0040] S12. Determine the research period. Clearly define the start and end years for data collection, identify key milestones in the substation's construction and operation phases, providing a clear timeframe for subsequent data collection and ensuring the timeliness and accuracy of the data.

[0041] S13. Through IoT sensing devices and SCADA systems, collect multi-source heterogeneous data in real time during the construction and operation phases of substations, including key indicators such as building material usage, construction energy consumption, and equipment operating parameters. Clean, normalize, and standardize the data to establish a unified data warehouse. Specific requirements include: (1) The substation data collected are collected separately according to the construction stage, operation stage and decommissioning stage. For the construction phase, the required data includes the human, material, and machinery resources consumed during the construction of substations of the same specifications, as well as the corresponding carbon emission factors; data used to calculate the substation's working performance indicators and clean efficiency indicators; carbon emission data and construction waste generation data during the substation construction period.

[0042] For the operational phase, the required data includes the carbon emissions generated by the same substation over a continuous time period, as well as the data needed to calculate five indicators: the proportion of clean energy, the overall network loss rate, the equivalent full load hours, the sulfur hexafluoride recovery rate, and the equipment loss rate.

[0043] The end-of-life assessment primarily tracks the types and quantities of recyclable materials after equipment decommissioning, along with their carbon emission factors from recycling, to evaluate the resource recycling benefits during the end-of-life phase. Standardized measurement methods are used for data at each stage to ensure the accuracy and comparability of the full life-cycle carbon efficiency assessment.

[0044] (2) Data completeness: Priority should be given to samples with complete engineering archives and continuous operation monitoring.

[0045] (3) Exclude abnormal samples: remove substations whose data is interrupted due to natural disasters or major renovations, exclude special function stations, and focus on conventional substations.

[0046] S2. Establish a carbon emission measurement system covering all stages and compile it into carbon emission datasets by stage and year; the carbon emission calculation framework for each part of the construction, operation and decommissioning stages is as follows, specifically including S21-S24: S21: During the construction phase, the study comprehensively considers the project construction process when establishing a carbon footprint measurement model for the physical phase of substations to ensure the model's accuracy and practicality. The study divides carbon emissions during the physical phase of power transmission and transformation projects into three aspects: labor, materials, and machinery. Therefore:

[0047] In the formula, This refers to the total carbon emissions generated during the physicalization phase of power transmission and transformation projects. Carbon emissions generated by workers at the construction site during the materialization phase; Carbon emissions during the production and processing of building materials; This refers to the carbon emissions generated by machinery and equipment during the transportation and construction processes in the physicalization stage.

[0048] Among them, man-made carbon emissions Carbon emissions from materials Mechanical carbon emissions as follows:

[0049] In the formula, For the first The amount of work involved in each construction technique; Unit project quantity The number of man-days consumed by the construction process; It is an artificial carbon emission factor.

[0050]

[0051] In the formula, For the first Consumption of major building materials; For the first Carbon emission factors for various building materials. Considering the losses incurred during the production, transportation, storage, and construction of building materials due to cutting errors, breakage, and disposal, and combining material loss rate data provided in power industry standards and specifications, the carbon emission factors are adjusted, and the following formula is used to calculate the carbon emission factors for some power transmission and transformation engineering materials:

[0052] In the formula, The carbon emission factor in the "Standard for Calculating Carbon Emissions in Buildings"; This refers to the building material loss rate.

[0053]

[0054] In the formula, For the first The amount of work involved in each construction technique; Construction techniques for completing a unit of work The number of machine shifts consumed; The amount of energy consumed per unit machine shift; Carbon emission factors of energy used in machinery.

[0055] The carbon footprint of a substation during construction involves six processes, including foundation engineering, grounding engineering, and tower engineering, totaling 87 procedures, covering the entire physicalization phase. Based on process life cycle theory, this study establishes a carbon footprint assessment list for the project from three aspects of the physicalization phase: "building material production," "building material transportation," and "construction." Building materials are categorized into seven types, including wood, bamboo, and metal; personnel are divided into skilled workers and general workers; and machinery is divided into 14 categories. Detailed carbon footprint information is as follows: Figure 3 As shown.

[0056] S22: The calculation basis for the carbon emissions of substations during operation and maintenance is as follows: This part of the carbon emissions is the sum of sulfur hexafluoride emissions generated during the maintenance of sulfur hexafluoride equipment and carbon dioxide emissions generated in the power production process corresponding to power transmission and distribution losses.

[0057]

[0058] In the formula, Total carbon emissions from power grid company equipment / (CO2e); The amount of sulfur hexafluoride (SF6) emitted during the maintenance and decommissioning of equipment using SF6 is expressed as (CO2e). This represents the total carbon dioxide emissions caused by power transmission and distribution losses, expressed as CO2e.

[0059] 1) Emissions from maintenance and decommissioning. The difference between the actual amount of sulfur hexafluoride (SF6) recovered during maintenance and decommissioning of SF6 equipment used by power grid companies and the amount indicated on the equipment nameplate is used as the basis for calculating SF6 emissions, as detailed below:

[0060] In the formula, The sulfur hexafluoride capacity / kg displayed on the nameplate of decommissioned and scrapped equipment; The actual amount of sulfur hexafluoride recovered from decommissioned and scrapped equipment (kg); The sulfur hexafluoride capacity / kg displayed on the nameplate of the equipment under maintenance; The actual amount of sulfur hexafluoride recovered from the equipment under maintenance (kg); This is the sulfur hexafluoride conversion coefficient, typically taken as 23.9.

[0061] 2) Emissions from transmission and distribution losses. The main source of carbon dioxide emissions from power grid equipment is transmission line losses, calculated as follows:

[0062] In the formula, Power supply / (MW·h); Equipment line loss rate / % The average emission factor of the regional power grid is CO2e / (MW·h); This is the carbon dioxide conversion coefficient, which is generally taken as 1.

[0063] S23. Decommissioning and Scrapping Emissions. For decommissioning equipment in power grid companies, there are generally three methods of handling it: scrapping, reuse, and recycling. Here, recycling is the primary method, and resource losses during the decommissioning and scrapping phase are not considered. The specific formula is as follows:

[0064] In the formula, Recyclable materials in the equipment Carbon emissions / t; For recycling materials Carbon emission factor / (CO2e / t); This refers to the number of different categories of recyclable materials in the equipment.

[0065] S3. This application uses the SBM-DEA model to systematically evaluate the carbon emission efficiency of substations during the construction period. First, based on the characteristics of substation construction, an evaluation index system is constructed that includes resource input, capital input, and carbon emission output. Resource input covers the consumption of major building materials such as steel and concrete; capital input considers technical and management costs; and carbon emission output involves direct and indirect emissions generated during construction. Second, the efficiency value of each decision unit (DMU) is calculated using the model, and slack variables are introduced to analyze specific improvement directions. The environmental performance of different substations is evaluated from a static perspective, and a horizontal comparative analysis is conducted to examine the environmental performance of each substation during the construction phase, specifically including S31-S34: S31. When using the SBM model to measure the carbon emission performance during the construction period of a substation, the input indicators involved include resource consumption input and technology and management input. These two types of indicators specifically include: (1) Resource consumption input Costs of dismantling and disposing of existing facilities: The indicator includes two levels of cost components. The first is the cost of dismantling the original power facilities, including direct construction costs such as equipment dismantling and structural demolition. The second is the cost of waste disposal, which covers environmental compliance expenditures such as construction waste removal and professional treatment of hazardous waste.

[0066] Investment in transitional power supply facilities: This includes the cost of leasing or purchasing equipment such as temporary substations and mobile generator sets, as well as related installation, commissioning, operation, maintenance, and dismantling costs.

[0067] Recyclable resource deduction: The indicator calculates the residual value of steel, copper and other metal materials and transformer oil recycled during the construction process, including the discounted value of waste equipment and materials collected on-site after professional evaluation.

[0068] Construction Resource Input: This patent considers the construction of substations involving six processes, including foundation engineering, grounding engineering, and tower engineering, totaling 87 procedures. Based on process life cycle theory, the study summarizes the resource input during the construction process from three aspects: "building material production," "building material transportation," and "construction." Building materials are categorized into seven types, such as wood, bamboo, and metal; personnel are divided into skilled workers and general workers; and machinery is divided into 14 categories. A detailed resource input list is as follows: Figure 3 As shown, the input of construction resources is the sum of the products of the amount of various resources used and the unit cost of use.

[0069] (2) Technology and management inputs Investment in technology for reusing and upgrading existing facilities: Assess the cost difference between upgrading existing equipment (such as extending the life of transformers) and purchasing new equipment; Construction risk management costs: Investment in special protective measures such as live-line work and construction near operating equipment; Community coordination costs: Hidden costs such as coordination and compensation for residents caused by the impact of reconstruction on the surrounding power supply.

[0070] Clean technology investment: Additional investment required to adopt low-carbon building materials and energy-saving construction equipment, including prefabricated components, electric construction machinery, etc.

[0071] Environmental management costs: investment in environmental monitoring systems, dust control equipment, and wastewater treatment facilities.

[0072] Personnel training investment: Regularly conducting environmental protection construction training and skills assessments requires dedicated funding, including training venues, instructor fees, and assessment and certification.

[0073] S32. Evaluating the input-output efficiency of a substation during its construction period based on its operational performance and cleanliness, the output indicators of the SBM model include expected and unexpected outputs, specifically S321-S322: S321. Expected output indicators include: (1) Power supply availability: measures the ability of a substation to continuously supply power to users.

[0074]

[0075] (2) Transformer load factor: The ratio of actual load to rated capacity, reflecting the equipment utilization rate.

[0076]

[0077] (3) Overall line loss rate: The ratio of the difference between input power and output power, reflecting the power grid transmission efficiency.

[0078]

[0079] (4) Voltage pass rate: the percentage of time the voltage is within the allowable deviation range.

[0080]

[0081] (5) Harmonic distortion rate (THD, %): The proportion of harmonic components to the total voltage / current, reflecting the purity of the waveform.

[0082]

[0083] S322. Undesirable output indicators include: (1) Total carbon emissions during the construction period: Greenhouse gas emissions during the entire process of building material production, transportation and construction.

[0084] (2) Construction waste generation (tons): The total amount of non-recyclable waste generated during construction, calculated as follows: Construction waste generation = Total building material usage × Waste rate + Construction loss amount S33. Given that the outputs during the substation construction period include not only the conventional power facility construction results, but also undesirable output factors that negatively impact the environment, such as carbon emissions and construction waste generation, an efficiency evaluation based on a slack variable-based SBM model is adopted. The specific steps are as follows: (1) Treat each substation as a decision-making unit, and denote the k-th substation as DMUk. Each DMU has m input options, denoted as... , recorded as Actual expected output For actual undesired outputs, there are respectively indivual.

[0085] Specifically, use To describe a specific DMU0:

[0086]

[0087]

[0088] in, This indicates excessive input. This indicates insufficient expected output. This indicates an excess of undesirable output, collectively referred to as the slack vector of input and output.

[0089] (2) with Indicators for building SBM models:

[0090]

[0091]

[0092]

[0093]

[0094]

[0095] In the formula, This indicates the distance between the input-output performance of different substations during the construction period and the cutting edge. The smaller the value, the higher the input-output efficiency and the higher the carbon efficiency of the DMU. Indicates the quantity of input indicators. This indicates the amount of redundant input that can be reduced; , It is the adjustment factor used in the Charnes-Cooper conversion. The value reflects the degree to which output slack affects overall efficiency. , Let these represent the slack variables for expected and unexpected outputs, respectively. This indicates the DMU being evaluated. The actual value of the crop output; This represents the weights of different DMUs in constructing the frontier, and the input of the target decision unit. It is represented as the weighted combination of inputs from all other decision-making units plus any possible input redundancy. The output of the target unit is represented as the weighted combination of outputs from other units minus or plus any under- or over-output portions.

[0096] (3) A feasible solution to the original model can be obtained. .

[0097] like If so, the DMU is called SBM-efficient. , , This indicates that the substation achieved optimal efficiency during the construction phase, with no redundancy in resource input, proper environmental emission control, and fully compliant output. It can serve as an industry benchmark, and its construction plan is worth promoting and learning from. Maintenance measures include continuous monitoring of key indicators and regular updates to technical standards.

[0098] like If the input and output of a DMU are not sufficiently efficient, it is called an SBM-inefficient DMU, ​​indicating that there is an efficiency loss during construction, potentially leading to resource waste or excessive emissions. Optimization measures include analyzing slack variables to pinpoint specific weaknesses, such as reducing redundant inputs through refined material management, adopting low-carbon construction techniques to reduce undesirable outputs, and referencing best practices from high-efficiency substations for improvement. For an SBM-inefficient decision-making unit, its inputs and outputs satisfy the following:

[0099]

[0100]

[0101] (4) Since specific values ​​of over-input and under-output have been obtained, the comparable decision unit can be optimized, which is called SBM projection:

[0102]

[0103]

[0104] The core function of this step is to provide a precise optimization path for substation construction by quantitatively analyzing the specific values ​​of input redundancy and output insufficiency.

[0105] S4. For the construction period, based on the efficiency calculated using the SBM model, the Tobit model is used to quantitatively analyze the input-output efficiency of substations at different construction periods. This is particularly suitable for situations where efficiency data is truncated. During implementation, it is necessary to ensure that the substation samples from the same year are fully representative, and to strictly match samples from different construction periods to guarantee comparability in key characteristics such as technical parameters and construction scale. Ultimately, this achieves an accurate longitudinal comparative evaluation of substation construction efficiency. This specifically includes S41-S44: S41. Select five indicators related to carbon emission reduction efficiency during the substation construction period, and analyze their correlation coefficients with the substation construction carbon efficiency calculated by the SBM model. The selected indicators include: (1) Percentage of Carbon Emission Reduction Budget: This refers to the percentage of the power grid company's total annual investment specifically allocated to carbon emission reduction technologies, including SF6 substitution, carbon capture, and high-efficiency cooling systems. A higher percentage indicates that the company is more proactive in adopting low-carbon technologies, directly impacting the carbon emission intensity of substations. The calculation method is as follows:

[0106] (2) R&D investment as a percentage of revenue: This refers to the percentage of the power grid company's total revenue invested in R&D related technologies for substations, such as high-efficiency transformers, intelligent monitoring, and low-carbon materials. A higher percentage indicates that the company places greater emphasis on technological innovation, which will improve substation efficiency and reduce carbon emissions in the long run. The calculation method is as follows:

[0107] (3) Investment Ratio Due to Policy Compliance: This refers to the percentage of total investment that the power grid company incurs due to environmental / energy efficiency policies (such as carbon tax and green building standards). This ratio reflects the company's sensitivity to policies and directly affects the degree of decarbonization of substations. The calculation method is as follows:

[0108] (4) Digital substation penetration rate: The proportion of digital / intelligent substations already in operation out of all substations. The calculation method is as follows:

[0109] (5) Low-carbon technology investment concentration: The multiple of the company's investment in low-carbon technologies for substations (such as SF6 replacement, energy storage integration, and waste heat recovery) relative to the industry average. The calculation method is as follows:

[0110] S42. The Tobit regression model uses the construction efficiency of substations of consistent size in different years, calculated by the SBM model in step S3, as the dependent variable. The data basis for this efficiency index is the longitudinal comparability design of substations of the same specification considering the time span. Its efficiency value is distributed between 0 and 1, where an efficiency value of 1 indicates the Pareto optimal state, that is, the resource input and clean output have reached the frontier; while an efficiency value of 0 indicates that there are significant negative environmental externalities in the construction process of the substation.

[0111] S43. Perform Tobit regression. Input data includes standardized substation construction efficiency values ​​generated in stage S3, strategic variable sets for each substation, and control variables. Output data includes the Tobit regression coefficient matrix, significance level, cutoff standard deviation, likelihood ratio test results, and the average marginal effect of observed efficiency values. The final result is a report on the weights of key factors affecting substation efficiency and a list of strategic optimization recommendations. Specific steps S431-S432 are as follows: S431, Joint Modeling (1) The formula for latent variables is as follows:

[0112] In the formula, For the parameter to be estimated, , , , , The independent variables, i.e., the explanatory variables, represent the impact coefficients of the proportion of carbon emission reduction special budget investment, the proportion of R&D investment to revenue, the proportion of government compliance investment, the penetration rate of digital substations, and the concentration of low-carbon technology investment on the commissioning efficiency of substations during the construction period, respectively. The dependent variable, i.e., the explained variable, represents the efficiency value of substation construction, which may be less than 0 or greater than 1. It is the core assumption of the model, representing the error term. Follows a mean of 0 and a variance of It follows a normal distribution.

[0113] (2) The formula for the observed variables is as follows:

[0114] S432, Maximum Likelihood Estimation After completing the joint modeling of the Tobit model, parameter estimation and inference can be directly achieved using statistical software. The input consists of a complete dataset containing dependent, independent, and control variables, with explicit specification of the cutoff boundary for the dependent variable. The software's built-in Tobit module performs maximum likelihood estimation, automatically optimizing the joint likelihood function to output the regression coefficients, standard deviations, and significance levels of the latent variable model. Furthermore, marginal effects calculation converts the impact of latent variables into quantified effects of actual observed efficiency values, ultimately outputting scientifically reliable regression results to support causal effect analysis and policy optimization of strategic variables.

[0115] S44. The five coefficients obtained from the regression reflect the marginal contributions of five types of factors to the efficiency of the substation during the construction period: (1) Coefficient of carbon emission reduction budget: reflects the effect of the company's targeted investment in low-carbon technologies, such as SF6 substitution and energy storage systems, on efficiency improvement. If the coefficient is positive and significant, it means that every 1% increase in carbon emission reduction budget can drive the efficiency value to rise by the corresponding coefficient value, indicating that the company's special investment is effectively transformed into clean energy improvement.

[0116] (2) R&D investment as a percentage of revenue coefficient: measures the long-term driving effect of technological R&D on efficiency. A positive coefficient indicates that every 1% increase in R&D investment can improve efficiency through innovative technologies, such as high-efficiency transformer design and modular construction processes.

[0117] (3) Coefficient of the proportion of policy-compliant investment: This reflects the forced efficiency improvement effect of policy pressure. If the coefficient is positive, it means that the additional investment to meet environmental protection policies, such as carbon tax and green building standards, can bring efficiency improvement for every 1% of the proportion, thus verifying the effectiveness of the policy-driven mechanism.

[0118] (4) Coefficient of digital substation penetration rate: Quantifying the energy efficiency optimization effect of intelligent technology. A positive coefficient indicates that for every 1% increase in the proportion of digital substations, energy loss is reduced through technologies such as real-time monitoring and dynamic voltage regulation, thereby improving the efficiency value, reflecting the actual benefits of technology implementation.

[0119] (5) Coefficient of low-carbon technology investment concentration: to assess the impact of the company’s leading position in low-carbon technology. If the coefficient is positive, it means that for every 1 times higher than the industry average in low-carbon investment, the efficiency value will increase by the corresponding coefficient value, indicating that the technological advantage can be directly transformed into efficiency competitiveness.

[0120] S45. Based on the analysis of different Tobit regression coefficients, identify the input factors that have the most significant impact on the carbon efficiency changes during the substation operation period, and in conjunction with the power grid company's "carbon peaking and carbon neutrality" strategic plan, formulate differentiated investment structure adjustment schemes.

[0121] S5. Based on the LMDI decomposition algorithm, a dynamic attribution model is constructed to decompose carbon emission changes into investment structure effects, technology intensity effects, and scale effects. The Divisia index is used to decouple the marginal contribution of five dimensions—energy structure, energy efficiency, load scale, technology path, and equipment energy efficiency—to substation carbon emission reduction during operation, revealing the dynamic correlation mechanism between each factor and the emission reduction effect. Specifically, this includes S51-S54: S51. Determine the base period and reporting period, define the time span of the analysis, and obtain data on energy structure, energy efficiency, load scale, technology pathways, equipment energy efficiency indicators, and carbon emissions for both the base period and the reporting period. The base period is the starting point of the analysis, used to determine the variable values ​​of the initial state; the reporting period is the ending point of the analysis, used to determine the variable values ​​of the final state.

[0122] The energy structure is quantified by the proportion of clean energy supply, which refers to the proportion of renewable energy in the electricity purchased by the substation during the operation period. The calculation formula is as follows:

[0123] Energy efficiency is quantified by the overall network loss rate, which refers to the proportion of power transmission loss to total power supply during operation, reflecting the level of energy efficiency. The calculation formula is as follows:

[0124] Load capacity is quantified by equivalent full-load hours, which refers to the ratio of actual power supply to rated capacity, reflecting load capacity and continuity. The calculation formula is as follows:

[0125] The technical approach is quantified by the sulfur hexafluoride (SF6) recovery rate, which refers to the proportion of SF6 recovered to the total amount of make-up gas. This reflects the level of application of emission reduction technologies, and the calculation formula is as follows:

[0126] Equipment energy efficiency is quantified by equipment loss rate, referring to the energy consumption efficiency of equipment inside a substation, which affects total electricity consumption. The calculation formula is as follows:

[0127] S52. Data preprocessing: Clean and standardize the collected data on energy structure, energy efficiency, load scale, technology path indicators, and equipment energy efficiency data of each substation in the base period and reporting period to ensure the accuracy and consistency of the data.

[0128] S53. Constructing the LMDI model. The LMDI model has additive and multiplicative forms. The additive form is suitable for decomposing absolute changes in the total amount, directly quantifying the absolute contribution of each factor to the change. The multiplicative form is suitable for decomposing proportional changes, focusing on the relative contribution of each factor to the total rate of change. In the scenario discussed in this patent, the additive model can clearly demonstrate the specific contribution of different indicators to the carbon reduction capacity of the substation during operation. Therefore, the additive form of the LMDI model is chosen for analysis, and the formula is as follows:

[0129] In the formula, For the first A substation Carbon emissions during the period, and its relationship with The factors involved during the period include the proportion of clean energy during substation operation. Overall network loss rate Equivalent full load hours Sulfur hexafluoride recovery rate and equipment loss rate Related, with As a substitute for basic carbon emission factors ,by As a substitute for clean energy structural effect factors ,by As a substitute for network loss efficiency factor ,by As a substitute for load intensity effect factor ,by As a substitute for management effect factors ,by As a substitute for equipment loss factor .

[0130] The total carbon emission effect over the years As shown in the following formula, this formula is used to reflect the correlation between changes in carbon emissions and changes in various influencing factors during the operation of a substation:

[0131] The changes in each secondary indicator in the additive form are calculated as follows:

[0132]

[0133]

[0134]

[0135]

[0136]

[0137] In the formula, The coefficient representing the impact of the baseline carbon emission factor on the carbon emission reduction achieved by the substation operation during the analysis period. yes The assessment of substation carbon emissions during the period. This is the carbon emissions in the base period. and They are Base carbon emission coefficients for the period and the base period; The coefficient representing the impact of the clean energy structure effect on the carbon emission reduction achieved by the substation operation during the analysis period. and These represent the clean energy structure effect factors for the analysis period and the base period, respectively; The coefficient representing the impact of network loss efficiency on the carbon emission reduction achieved by substation operation during the analysis period. , These represent the network loss efficiency effect factors for the analysis period and the base period, respectively; The coefficient representing the impact of load intensity effect on the carbon emission reduction achieved by substation operation during the analysis period. , These represent the load intensity effect factors for the analysis period and the base period, respectively; The coefficient representing the impact of management effects on the carbon emission reductions achieved by substation operation during the analysis period. , These represent the management effect factors for the analysis period and the base period, respectively; The coefficient representing the impact of equipment loss rate on the carbon emission reduction achieved by substation operation during the analysis period. and These represent the equipment loss factors for the analysis period and the base period, respectively.

[0138] S54. Quantifying Contribution. The contribution of each factor to the change in carbon emissions is calculated using the LMDI formula, and the analysis is as follows: Basic carbon emission factor coefficient This coefficient reflects changes in baseline emission intensity resulting from technological improvements or policy adjustments. A negative value indicates a decrease in emission intensity, and managers should promote proven low-carbon technologies and optimize operating parameters; a positive value indicates a need to investigate equipment aging or operational issues, and update technical standards if necessary.

[0139] Clean energy structure effect coefficient The larger the negative value of the coefficient, the more significant the emission reduction contribution from increasing the proportion of clean energy. Regulators should prioritize negotiating with the power grid to increase the procurement of green electricity and establish a real-time monitoring and incentive mechanism for the proportion of clean energy.

[0140] Network loss efficiency effect coefficient Negative values ​​indicate that reduced network losses lead to emission reduction benefits. Key measures include upgrading reactive power compensation devices, optimizing grid connection schemes, strengthening line inspection and maintenance, and, when necessary, introducing intelligent dispatching systems to reduce transmission losses.

[0141] Load intensity effect coefficient The current situation indicates that increased load is offsetting the effects of emissions reductions. It is necessary to optimize operational strategies, configure energy storage systems to participate in demand response, implement peak-valley pricing to guide demand, and balance load with clean energy generation curves.

[0142] Management effect coefficient This primarily reflects the effectiveness of SF6 and other gas management. Negative values ​​require maintaining existing measures, while positive values ​​necessitate increased investment in recovery equipment, improved leak detection procedures, operational training, and exploration of environmentally friendly alternative gas application solutions.

[0143] Equipment loss rate coefficient Negative values ​​indicate the effectiveness of energy efficiency improvements. Priority should be given to upgrading high-energy-consuming equipment, establishing sub-metering systems, developing energy efficiency improvement plans, and regularly evaluating the operating efficiency of key equipment such as transformers.

[0144] S6. K-means clustering analysis of substation carbon efficiency based on full life cycle characteristics: By integrating multi-dimensional carbon efficiency characteristics of substations throughout their entire life cycle, a scientific classification basis for differentiated carbon management strategies is constructed. Based on the K-means clustering algorithm and combined with the dynamic emission characteristics during the construction and operation phases, the heterogeneity of carbon efficiency performance of different substations is revealed, providing data-driven decision support for power grid companies to accurately formulate "one policy per substation" low-carbon optimization solutions. Specifically, this includes S61-S64: S61. Data processing includes two steps: data filtering and standardization. First, valid data is filtered and outliers are processed. Then, the data is standardized and transformed to ensure that the format is standardized and usable.

[0145] S611. Data Preparation and Feature Selection: The selected clustering data comes from carbon emission-related data across different lifecycles of the substation, including: The selected data from the substation construction period refers to the static carbon efficiency value (0-1 range) calculated based on the SBM model, which characterizes the overall efficiency level of resource input and carbon emission output.

[0146] The selected data from the substation operation period include the basic carbon emission factor quantified based on the LMDI model, the clean energy structure effect, the network loss efficiency effect, the load intensity effect, the management effect, and the impact coefficient of equipment loss on carbon emissions during the operation period.

[0147] S612. Data standardization: Perform Z-score standardization on continuous variables to eliminate differences in units. Differences in units among different indicators require a unified scale.

[0148] S62. Determine the number of clusters (K value). A systematic approach is used to determine the optimal number of clusters for the carbon efficiency characteristics of substations. First, based on the elbow rule, the total squared error variation curve of substation clustering under different K values ​​is analyzed. By identifying the inflection point of the error decrease rate, the potential number of clusters is initially determined. At the same time, combined with the silhouette coefficient evaluation method, the intra-cluster compactness and inter-cluster separation of substations in the carbon efficiency characteristic space under each K value are quantitatively calculated. Finally, the K value that maximizes the silhouette coefficient is selected as the optimal number of clusters, ensuring that the clustering results can accurately reflect the differences in carbon efficiency of substations.

[0149] S63. Perform K-means clustering. First, initialize cluster centers using the K-means optimization algorithm. Maximizing the initial centroid spacing effectively avoids the local optimum trap caused by random initialization, improving clustering convergence efficiency. Then, proceed to the iterative optimization phase. Calculate the Euclidean distance between each substation and the cluster center based on the standardized feature vectors, assigning samples to the nearest neighbor clusters. After assignment, recalculate the arithmetic mean of the feature vectors of each cluster as the new centroid. The iterative process continues until a preset termination condition is met, i.e., the change in centroid position is less than a threshold or the maximum number of iterations is reached, thus ensuring a balance between algorithm convergence and computational efficiency. By dynamically adjusting the centroid positions, a substation cluster with high cohesion and low coupling is ultimately formed.

[0150] S64. Evaluation and Interpretation of Clustering Results S641. Feature Statistics and Pattern Recognition: Calculate the mean and distribution of static carbon efficiency during the construction period, LMDI driving coefficient during the operation period, and regenerated carbon offset rate during the decommissioning period for each cluster, and identify the core feature combinations of high / medium / low carbon efficiency clusters.

[0151] S642. Dynamic optimization strategies and management rules are generated for different cluster combinations. For high-carbon-efficiency clusters, the focus is on promoting the standardized application of modular construction processes and carbon asset development. A carbon-efficiency-cost synergistic optimization model is built based on a digital twin platform to explore the integrated implementation path of carbon capture technology and carbon quota trading. For medium-carbon-efficiency clusters, a clean energy penetration rate improvement plan is implemented. Combined with the deployment of intelligent reactive power compensation devices and a tiered electricity price incentive mechanism, indicators such as photovoltaic coverage and energy storage ratio are incorporated into the power grid planning constraints. For low-carbon-efficiency clusters, the elimination system for high-energy-consuming equipment is strictly enforced. Inefficient transformers are forcibly replaced according to the "Energy Efficiency Limits for Power Transformers". Dynamic power flow optimization algorithms are applied simultaneously to compress the load peak-valley difference and establish a demand response resource pool to achieve minute-level flexible regulation.

[0152] S7. Construct a closed-loop management system of "monitoring-evaluation-optimization-verification" to provide power grid companies with a complete substation full life cycle carbon emission control solution. This system specifically includes: S71. Carbon Efficiency Dynamic Monitoring and Data Fusion: Establish a three-tiered IoT sensing network architecture to achieve comprehensive collection and fusion of substation carbon efficiency data. The sensing layer constructs a three-dimensional monitoring network covering equipment energy consumption, greenhouse gas emissions, and environmental parameters through devices such as smart meters, gas sensors, and environmental monitoring terminals. The transmission layer adopts 5G power slicing networks and edge computing technology to ensure real-time data transmission and local preprocessing. The fusion layer integrates BIM asset information, SCADA operation data, and external meteorological data through a data platform to form a spatiotemporally correlated multidimensional data matrix, providing complete data support for carbon efficiency assessment.

[0153] S72. Adaptive Carbon Efficiency Assessment and Cluster Update: During the construction phase, the SBM model is used to assess the substation's commissioning performance, combined with the Tobit model to identify key carbon-saving factors; during the operation phase, the LMDI model is used to analyze carbon emission drivers. Dynamic cluster analysis is achieved through the K-means algorithm, with periodic updates to the cluster baseline value and setting inter-cluster difference thresholds to ensure that the classification results always reflect the latest carbon efficiency characteristics.

[0154] S73. Differentiated Optimization Strategy Generation and Implementation: Based on clustering results, a strategy engine is built to formulate targeted optimization measures for substation clusters with different carbon efficiency levels. Through the "feature-measure-parameter" mapping rule, intelligent matching from carbon efficiency characteristics to specific optimization schemes is achieved, providing clear operational guidance for power grid companies.

[0155] S74. Effect Verification and Model Iteration: Establish a comprehensive post-evaluation system to verify the actual effects of various optimization measures through quantitative analysis. Feed the verification results back to the evaluation model to continuously optimize parameter settings and decision rules, forming a virtuous cycle of knowledge accumulation.

[0156] S75. Closed-Loop Operation and Tiered Improvement: Construct a self-evolving management system that updates the station's carbon efficiency profile annually and optimizes strategies quarterly. By setting tiered targets for carbon efficiency improvement and collaborating with the power grid planning department to develop a carbon budget allocation plan, the system ensures continuous improvement in carbon efficiency management.

[0157] S8: Design a smart carbon efficiency management device for the entire lifecycle of a substation. This device, through an innovative "cloud-edge-device" collaborative architecture, transforms the aforementioned methodology into an engineering-applicable hardware system, specifically including: S81: The substation carbon efficiency intelligent management device adopts a three-tier "cloud-edge-device" architecture, consisting of three core modules: a data acquisition layer, an intelligent analysis layer, and a decision execution layer. Data interaction and command transmission between these modules are achieved through standardized interfaces. This device can be deployed on the existing energy management system platform of the power grid company, achieving seamless integration with systems such as SCADA and BIM.

[0158] S82: The core modules of this device include a data acquisition layer, an intelligent analysis layer, and a decision execution layer. The core functions of each module are shown in Table 1. Table 1. Functional Module Configuration Table of Substation Full Life Cycle Carbon Efficiency Intelligent Management Device

[0159] S83: This device interfaces with the power grid company's existing energy management system, equipment management platform, and carbon asset management system to achieve automatic data flow and closed-loop execution of optimization strategies for substation carbon efficiency. After deployment, the device can monitor the carbon efficiency status of each substation in real time, intelligently push clean energy transformation plans, and dynamically evaluate the improvement effects, ultimately forming a continuous improvement mechanism of "monitoring-evaluation-optimization-verification," systematically improving the clean operation level and resource utilization efficiency of substations. By incorporating carbon efficiency indicators into the power grid planning and assessment system, it promotes the transformation of substations from passive emission reduction to proactive carbon management.

[0160] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0161] The following are device embodiments of the present invention. For details not described in detail, please refer to the corresponding method embodiments described above.

[0162] Figure 5 The diagram illustrates the structure of the substation full life-cycle carbon footprint quantification and dynamic optimization system provided in an embodiment of the present invention. For ease of explanation, only the parts relevant to the embodiment of the present invention are shown, and are described in detail below: like Figure 5As shown, the substation full life cycle carbon footprint quantification and dynamic optimization system 5 includes: IoT sensing device 501, SCADA system 502 and substation carbon efficiency intelligent management device 503; wherein, the substation carbon efficiency intelligent management device 503 is communicatively connected to IoT sensing device 501 and SCADA system 502 respectively. The SCADA system 502 is used to centrally collect data during the construction, operation, and decommissioning phases of substations. The substation carbon efficiency intelligent management device 503 is used to perform the method provided in any of the foregoing embodiments.

[0163] Figure 6 This is a schematic diagram of an electronic device provided in an embodiment of the present invention. For example... Figure 6 As shown, the electronic device 6 of this embodiment includes a processor 60 and a memory 61. The memory 61 stores a computer program 62. When the processor 60 executes the computer program 62, it implements the steps in the various method embodiments described above. Alternatively, when the processor 60 executes the computer program 62, it implements the functions of each module / unit in the various device embodiments described above.

[0164] For example, computer program 62 may be divided into one or more modules / units, which are stored in memory 61 and executed by processor 60 to complete the present invention. The one or more modules / units may be a series of computer program instruction segments capable of performing a specific function, which describe the execution process of computer program 62 in electronic device 6.

[0165] Electronic device 6 may include, but is not limited to, processor 60 and memory 61. Those skilled in the art will understand that... Figure 6 This is merely an example of electronic device 6 and does not constitute a limitation on electronic device 6. It may include more or fewer components than shown, or combine certain components, or different components. For example, electronic device 6 may also include input / output devices, network access devices, buses, etc.

[0166] For the sake of simplicity and clarity, only the above-described functional modules / units are used as examples. In practical applications, the functions described above can be assigned to different functional modules / units as needed. These modules / units can be implemented in hardware, software, or a combination of both.

[0167] In the above embodiments, the descriptions of each embodiment have their own emphasis. Parts not detailed or described in a particular embodiment can be referred to in the relevant descriptions of other embodiments. Unless otherwise specified or in conflict with logic, the terminology and / or descriptions between different embodiments are consistent and can be referenced interchangeably. Technical features in different embodiments can be combined to form new embodiments based on their inherent logical relationships.

[0168] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A method for quantifying and dynamically optimizing the carbon footprint of a substation throughout its entire life cycle, characterized in that, include: Construction period data of multiple substations are obtained, and the construction period data are input into a carbon efficiency evaluation model to obtain the carbon efficiency evaluation results of each substation. Based on the corresponding carbon efficiency evaluation results, an investment structure adjustment plan is formulated for each substation. The carbon efficiency evaluation model is constructed based on a data envelopment analysis model and a Tobit regression model. Operational data from each substation is collected and input into a dynamic attribution model to obtain marginal contribution data for each substation. The operational data includes carbon dioxide emissions from sulfur hexafluoride (SF6) and carbon dioxide emissions caused by grid losses during operation. The marginal contribution data includes the marginal contribution of energy structure, energy efficiency, load scale, technology path, and equipment energy efficiency to carbon emission reduction during substation operation. The dynamic attribution model is constructed based on the logarithmic mean divisor exponent (LMDI) decomposition algorithm. The K-means clustering algorithm was used to perform cluster analysis on the carbon efficiency evaluation results and marginal contribution data of each substation, and the carbon efficiency of similar substations was compared to generate carbon emission reduction improvement schemes for each substation.

2. The method according to claim 1, characterized in that, The data envelopment analysis model is the SBM-DEA model; the step of inputting the construction period data into the carbon efficiency evaluation model to obtain the carbon efficiency evaluation results for each substation includes: The construction period data is input into the carbon efficiency evaluation model, and the SBM-DEA model is used to evaluate the carbon efficiency of the substation during the construction period and calculate the static carbon efficiency value of each substation. Based on the static carbon efficiency value, the Tobit regression model is used to analyze the impact of carbon emission reduction special budget ratio, R&D investment ratio, policy compliance investment ratio, digital substation penetration rate and low-carbon technology investment concentration on carbon efficiency, and to generate regression coefficients and significance levels.

3. The method according to claim 2, characterized in that, The input indicators of the SBM-DEA model include resource consumption inputs and technology and management inputs; wherein, the resource consumption inputs include the cost of dismantling and disposing of existing facilities, the cost of transitional power supply facilities, the deduction of recyclable resources, and the cost of construction resources; the technology and management inputs include the cost of technology for reusing and upgrading existing facilities, the cost of construction risk management, the cost of community coordination, the cost of clean technology, the cost of environmental management, and the cost of personnel training. The output indicators of the SBM-DEA model include expected outputs and unexpected outputs; wherein, the expected outputs include power supply availability, transformer load rate, comprehensive line loss rate, voltage qualification rate and harmonic distortion rate; and the unexpected outputs include total carbon emissions during the construction period and the amount of construction waste generated.

4. The method according to claim 2, characterized in that, The independent variables of the Tobit regression model include: the proportion of special budget for carbon emission reduction, the proportion of R&D investment to revenue, the proportion of policy compliance investment, the penetration rate of digital substations, and the concentration of low-carbon technology investment. The carbon emission reduction budget percentage refers to the percentage of the power grid company's investment specifically for carbon emission reduction technologies in its total annual investment; the R&D investment as a percentage of revenue refers to the percentage of the power grid company's R&D investment in substation-related technologies in its total revenue; the policy compliance investment percentage refers to the percentage of the power grid company's additional investment due to environmental policies in its total investment; the digital substation penetration rate is the percentage of operational digital substations in all substations; and the low-carbon technology investment concentration is the multiple of the company's investment in low-carbon technologies for substations relative to the industry average.

5. The method according to claim 4, characterized in that, The aforementioned plan, based on the corresponding carbon efficiency evaluation results, outlines a corresponding investment structure adjustment scheme for each substation, including: Based on the regression coefficients output by the Tobit regression model, the marginal contribution of each independent variable to carbon efficiency is analyzed. Develop differentiated investment adjustment strategies based on factors such as the proportion of carbon emission reduction budget, the proportion of R&D investment to revenue, the proportion of policy compliance investment, the penetration rate of digital substations, and the concentration of low-carbon technology investment. These investment adjustment strategies may include one or more of the following: increasing investment in low-carbon technologies, raising the proportion of R&D investment, and optimizing the allocation of policy compliance investment.

6. The method according to claim 1, characterized in that, The step of inputting the operational data into the dynamic attribution model to obtain the marginal contribution data corresponding to each substation includes: The operational period data is divided into base period data and reporting period data to obtain energy structure, energy efficiency, load scale, technology path and equipment energy efficiency index data for each substation in the base period and reporting period; The additive form of the LMDI decomposition algorithm is used to decompose carbon emission changes into basic carbon emission factor effects, clean energy structure effects, grid loss efficiency effects, load intensity effects, management effects, and equipment loss effects.

7. The method according to claim 1, characterized in that, The K-means clustering algorithm is used to perform cluster analysis on the carbon efficiency evaluation results and marginal contribution data of each substation, including: Obtain the carbon efficiency evaluation results and marginal contribution data of each substation to obtain initial clustering data; The initial clustering data is standardized, and the elbow rule and silhouette coefficient are used to determine the optimal number of clusters. The K-means clustering algorithm is then executed to generate the clustering results.

8. The method according to claim 7, characterized in that, The initial clustering data includes: static carbon efficiency values ​​during the construction period calculated based on the SBM-DEA model, and the influence coefficients of basic carbon emission factors, clean energy structure effects, network loss efficiency effects, load intensity effects, management effects, and equipment loss effects during the operation period quantified based on the LMDI model.

9. The method according to claim 8, characterized in that, The process of comparing the carbon efficiency of similar substations and generating carbon reduction improvement plans for each substation includes: For substations with high carbon efficiency, promote the standardized application of modular construction technology and carbon asset development, and integrate carbon capture technology with carbon quota trading. For substations with medium carbon efficiency, implement a clean energy penetration rate improvement plan, deploy intelligent reactive power compensation devices and tiered electricity price incentive mechanisms; For substations with low carbon efficiency, implement a system to eliminate high-energy-consuming equipment, force the replacement of inefficient transformers, and apply dynamic power flow optimization algorithms to reduce the load peak-valley difference.

10. A system for quantifying and dynamically optimizing the carbon footprint of a substation throughout its entire life cycle, characterized in that, include: The system includes an Internet of Things (IoT) sensor device, a data acquisition, monitoring, and control (SCADA) system, and a substation carbon efficiency intelligent management device; wherein the substation carbon efficiency intelligent management device is communicatively connected to the IoT sensor device and the SCADA system, respectively. The SCADA system is used to centrally collect data during the construction, operation, and decommissioning phases of substations. The substation carbon efficiency intelligent management device is used to perform the method as described in any one of claims 1 to 9.

Images