Data center green electricity consumption index evaluation method based on carbon emission measurement and calculation

Abstract

The invention discloses a data center green electricity consumption index evaluation method based on carbon emission measurement and calculation, and the method comprises the steps: carrying out the accounting of the carbon emission of a diesel generator unit through combining the total consumption amount of fuel, the lower heating value and the carbon oxidation rate, and building a diesel generator unit carbon emission accounting model; a carbon emission source of the data center is considered, and the carbon emission of the data center is calculated by using a "producer + apportionment" mixed accounting mode; and based on the link mechanism of the green certificate and the carbon trading market, according to the carbon emission accounting result of the diesel generator set and the carbon emission accounting result of the data center, obtaining an evaluation result of the green electricity consumption index of the data center. According to the method, a theoretical basis is provided for data center green power consumption index evaluation, green development of the data center is promoted, carbon emission target realization is promoted, and powerful support is provided for realizing economic and environmental benefit win-win in data center energy transformation and sustainable development by optimizing a carbon emission accounting method and a green power consumption index evaluation system.

Description

Technical Field

[0001] The present invention belongs to the field of power systems, relates to green electricity trading and green certificate mechanism technology, and specifically to a method for evaluating green electricity consumption indicators of data centers based on carbon emission measurement. Background Art

[0002] Today, digital industries are rapidly developing. As key nodes and hubs for information exchange, data centers' environmental emissions and energy consumption have naturally become a global concern. The continuous increase in computing power and the number of racks in data centers has led to a surge in energy consumption, costs, and carbon emissions. In particular, the growing application of large language models in AI, such as ChatGPT and Sora, has led to extremely high energy costs for data centers. To ensure a sufficient level of computing power while achieving low-carbon development, we must rely on renewable energy. One approach is to source computing power from green electricity, a "local power generation, local computing" model. Alternatively, we can source computing power from green electricity.

[0003] Currently, the standard system for calculating data center carbon emissions is incomplete. The industry has yet to propose quantitative indicators for reducing data center carbon emissions. Information on data center carbon emissions is also very limited, and regulatory and assessment mechanisms must also be improved. Existing standards primarily specify the calculation and testing methods for energy efficiency metrics such as PUE, but fail to fully reflect the carbon emissions levels of data centers. Summary of the Invention

[0004] Purpose of the invention: In order to overcome the deficiencies in the existing technology, a method for evaluating green electricity consumption indicators of data centers based on carbon emission measurement is provided.

[0005] Technical Solution: To achieve the above objectives, the present invention provides a method for evaluating green electricity consumption indicators of data centers based on carbon emission measurement, comprising the following steps:

[0006] S1: Calculate the carbon emissions of diesel-generator units based on the total fuel consumption, low calorific value, and carbon oxidation rate, and establish a carbon emissions calculation model for diesel-generator units;

[0007] S2: Considering the sources of carbon emissions from data centers, a hybrid accounting model of "producer + apportionment" is used to calculate the carbon emissions of data centers.

[0008] S3: Based on the linkage mechanism between green certificates and carbon trading markets, and according to the carbon emission accounting results of diesel generator sets and data centers, the evaluation results of the green electricity consumption indicators of data centers are obtained.

[0009] Furthermore, in step S1, the emission sources of the diesel-generator set are divided into direct emissions and indirect emissions, and the carbon emissions of the diesel-generator set are calculated by quantifying the carbon dioxide emissions generated by diesel combustion.

[0010] Furthermore, the calculation of the direct emissions portion in the carbon emissions accounting of the diesel generator set in step S1 includes:

[0011] Direct emissions calculation formula:

[0012] E 燃烧 =∑(AD i ×EF i )

[0013] Among them, AD i Activity data refers to diesel consumption (in tons). Daily measured data should be used first. If it is not available, monthly statistics should be used. i Expressed as an emission factor, the calculation formula is:

[0014]

[0015] Among them, CC i Indicates the carbon content per unit calorific value, which refers to the carbon content of diesel. If there is no measured data, refer to the default value (e.g. 0.0201tC / GJ, which needs to be confirmed according to the latest guidelines); OF i Indicates the carbon oxidation rate, which refers to the carbon oxidation rate of diesel combustion, usually 99%; NCV i Indicates the low calorific value, which refers to the low calorific value of diesel. The default value is 42.652GJ / t.

[0016] Furthermore, the calculation of the indirect emissions in the carbon emissions accounting of the diesel generator set in step S1 includes:

[0017] Indirect emissions calculation formula:

[0018] E 电 =∑(AD 电 ×EF 电 )

[0019] Among them, AD 电 Indicates the amount of electricity purchased (in MWh); EF 电 The emission factor is expressed as the national average emission factor of the power grid (the default value in 2023 is 0.6101tCO2 / MWh) or the latest published value.

[0020] Furthermore, the carbon emission sources of the data center in step S2 are divided into direct emissions, indirect emissions and value chain emissions. Direct emissions include fuel combustion and refrigerant leakage, indirect emissions include purchased electricity and heat, and renewable energy offsets. Value chain emissions are divided into upstream and downstream. The upstream includes implicit carbon emissions from server manufacturing and building material production, while the downstream is the energy consumption of terminal devices generated by users using cloud services (which needs to be shared according to the service ratio).

[0021] Furthermore, the calculation of the direct emissions portion in the data center carbon emissions accounting in step S2 includes:

[0022] Direct emissions calculation formula:

[0023] Diesel generator emissions (same calculation method as diesel generator sets):

[0024]

[0025] Refrigerant leakage and emission (taking HFC134a as an example):

[0026] E 制冷 = Leakage volume (kg) × GWP HFC-134a (GWP = 1430 CO2-eq / kg)

[0027] The direct carbon emissions are as follows:

[0028] E 范围1 =E 柴油 +E 制冷剂 .

[0029] Furthermore, the calculation of the indirect emissions portion in the data center carbon emissions accounting in step S2 includes:

[0030] Indirect emissions calculation formula:

[0031] Electricity emissions (core source):

[0032] Q 总 =∑(IT 耗电 ×PUE)

[0033] Among them, PUE stands for power usage efficiency, which reflects the energy efficiency of the data center. Used to amplify auxiliary power consumption outside IT equipment (such as cooling systems);

[0034] The indirect carbon emissions are as follows:

[0035] E 范围2 =(AD 总电 -AD 绿电 )×EF 电网 ×PUE.

[0036] Furthermore, the calculation of the value chain emissions portion in the data center carbon emissions accounting in step S2 includes:

[0037] Carbon emissions throughout the server life cycle:

[0038]

[0039] User-side emission allocation:

[0040]

[0041] Among them, Q 总 represents the total annual electricity consumption of the data center (in MWh); P IT Indicates the power of IT equipment (in MW); P 辅助 Indicates the power of auxiliary equipment such as cooling / power supply (in MW) and meets P 辅助 =P IT ×(PUE-1); η represents the user computing power usage ratio (%), which defaults to 100% (all computing power is used for external services); EF 用户电网 Indicates the grid emission factor of the user's area (unit: t CO2 / MWh);

[0042] The carbon emissions from the value chain are as follows:

[0043]

[0044] Furthermore, the linking mechanism between green certificates and the carbon trading market in step S3 includes:

[0045] How to link the new green certificate with indirect emissions:

[0046] For every additional green certificates purchased by a data center, the amount of electricity purchased by the enterprise will be deducted according to the amount of green certificates, with the upper limit being the total electricity consumption of the enterprise. The formula for calculating indirect emissions from purchased electricity of a data center is:

[0047] E=(Q t -Q g )F f

[0048] Where E represents the indirect emissions of the emitting enterprises; Q t represents the total amount of electricity purchased by the emission-collecting enterprises; Q g Indicates the amount of electricity corresponding to the new green certificate; F f Represents the average carbon emission factor of thermal power generation. The relevant factors can be updated annually;

[0049] Proportional constraints on the link between new green certificates and the carbon market:

[0050] Assume that the electricity emission factor is in represents the total emissions from the power industry, Q' t Let θ represent the total power supply of the power industry and θ represent the proportion of renewable energy generation. Then the necessary and sufficient conditions for achieving emission reduction are:

[0051]

[0052] Because Q' t =Q' f +Q' g ,Q' f Indicates the thermal power supply, Q' g Indicates the green electricity supply, then:

[0053]

[0054] Derivatives of both sides with respect to θ:

[0055]

[0056] Solve the differential equation:

[0057]

[0058] We can get:

[0059]

[0060] From the above calculations, it can be concluded that if and only if Only when , can real emission reduction be achieved, where c is a constant.

[0061] Beneficial effects: Compared with the prior art, the present invention has the following advantages:

[0062] 1. The present invention develops a carbon emission accounting model for diesel-generator units, which comprehensively solves the problems of traditional calculation methods, such as the vague definition of emission sources, the lack of reasonable basis for parameter selection, and actual operation instructions. Combining the total fuel consumption, low heating value and carbon oxidation rate, the calculation system is highly feasible and accurate, and can effectively quantify the contribution of carbon emissions and quantify the emissions of the backup power system.

[0063] 2. The present invention uses a "producer + sharing" hybrid accounting model, which overcomes the difficulties of the traditional responsibility allocation method and is more applicable in data center scenarios.

[0064] 3. This invention provides a theoretical basis for the evaluation of green electricity consumption indicators in data centers, promotes the green development of data centers and promotes the realization of carbon emission targets. By optimizing the carbon emission accounting method and the green electricity consumption indicator evaluation system, it provides strong support for achieving a win-win economic and environmental benefit in the energy transformation and sustainable development of data centers. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is the indirect emission accounting flow chart;

[0066] Figure 2 It is the energy consumption structure diagram of the data center;

[0067] Figure 3 It is a schematic diagram of the connection and deduction mechanism between electricity green equity certificates and the carbon market. DETAILED DESCRIPTION

[0068] The present invention is further illustrated below with reference to the accompanying drawings and specific embodiments. It should be understood that these embodiments are only used to illustrate the present invention and are not used to limit the scope of the present invention. After reading the present invention, modifications of various equivalent forms of the present invention made by those skilled in the art all fall within the scope defined by the claims attached to this application.

[0069] Example 1:

[0070] This embodiment provides a method for evaluating green electricity consumption indicators of a data center based on carbon emission measurement, including the following steps:

[0071] S1: Calculate the carbon emissions of diesel-generator units based on the total fuel consumption, low calorific value, and carbon oxidation rate, and establish a carbon emissions calculation model for diesel-generator units;

[0072] S2: Considering the sources of carbon emissions from data centers, a hybrid accounting model of "producer + apportionment" is used to calculate the carbon emissions of data centers.

[0073] S3: Based on the linkage mechanism between green certificates and carbon trading markets, and according to the carbon emission accounting results of diesel generator sets and data centers, the evaluation results of the green electricity consumption indicators of data centers are obtained.

[0074] In step S1:

[0075] The carbon accounting method for diesel-generating units (diesel generator sets) is mainly based on the "Guidelines for Corporate Greenhouse Gas Emissions Accounting and Reporting for Power Generation Facilities" and related technical specifications. Its core is to calculate by quantifying the carbon dioxide emissions generated by diesel combustion.

[0076] The emission sources of diesel-generator units are divided into direct emissions and indirect emissions. The carbon emissions of diesel-generator units are calculated by quantifying the carbon dioxide emissions generated by diesel combustion.

[0077] The calculation of direct emissions in the carbon emissions accounting of diesel generator sets includes:

[0078] Direct emissions calculation formula:

[0079] E 燃烧 =∑(AD i ×EFi )

[0080] Among them, AD i Activity data refers to diesel consumption (in tons). Daily measured data should be used first. If it is not available, monthly statistics should be used. i Expressed as an emission factor, the calculation formula is:

[0081]

[0082] Among them, CC i Indicates the carbon content per unit calorific value, which refers to the carbon content of diesel. If there is no measured data, refer to the default value (e.g. 0.0201tC / GJ, which needs to be confirmed according to the latest guidelines); OF i NCV represents the carbon oxidation rate, which refers to the carbon oxidation rate of diesel combustion. In this embodiment, it is 99%; i Indicates the low calorific value, which refers to the low calorific value of diesel. The default value is 42.652GJ / t.

[0083] The calculation of indirect emissions in the carbon emissions accounting of diesel generator sets includes:

[0084] Indirect emissions calculation formula:

[0085] E 电 =∑AD 电 ×EF 电 )

[0086] Among them, AD 电 Indicates the amount of electricity purchased (in MWh); EF 电 The emission factor is expressed as the national average emission factor of the power grid (the default value in 2023 is 0.6101tCO2 / MWh) or the latest published value.

[0087] In step S2:

[0088] 1. Accounting scope and emission source classification

[0089] As a high-energy-consuming facility, data centers emit carbon mainly from electricity consumption and backup generators (such as diesel generators). Accurately calculating the carbon emissions of data centers requires considering the sources of carbon emissions from data centers. Data center carbon emission responsibility accounting is not dominated by a single method, but is based on the responsibility allocation method, combined with the producer responsibility method to deal with directly controllable emissions, and introduces consumer responsibility elements in specific scenarios. The energy consumption structure of a typical data center is as follows: Figure 2 shown.

[0090] Table 1 shows the scope and emission sources for data center accounting, primarily categorized as direct, indirect, and value chain emissions. Direct emissions include fuel combustion and refrigerant leakage, while indirect emissions include purchased electricity and heat, as well as renewable energy offsets. Value chain emissions are further divided into upstream and downstream. Upstream emissions include embodied carbon emissions from server manufacturing and building material production, while downstream emissions include energy consumption from end-user devices generated by user cloud services (which is allocated based on the service ratio).

[0091] Given the complexity of the data center supply chain, with reference to the "economic activity proportion allocation method", server manufacturing carbon emissions are allocated according to the equipment life, and user-side emissions are allocated according to the amount of computing power used, which is a "producer + allocation" hybrid model. Direct emissions are borne entirely by the data center operator (producer responsibility). Indirect emissions are borne by the operator, but green electricity procurement can partially transfer responsibility (such as offsetting through green certificates). In the value chain emissions, server manufacturing carbon emissions are allocated to the data center according to the equipment life (producer allocation), and user terminal energy consumption is allocated to the user according to the amount of computing power used (consumer allocation).

[0092] Table 1 Data center accounting scope and emission source classification

[0093]

[0094]

[0095] 2. Specific methods for calculating data center carbon emissions

[0096] The calculation of direct emissions in data center carbon emissions accounting includes:

[0097] Direct emissions calculation formula:

[0098] Diesel generator emissions (same calculation method as diesel generator sets):

[0099]

[0100] Refrigerant leakage and emission (this embodiment takes HFC134a as an example):

[0101] E 制冷 = Leakage volume (kg) × GWP HFC-134a (GWP = 1430 CO2-eq / kg)

[0102] The carbon emissions from direct emissions (scope 1) are as follows:

[0103] E 范围1 =E 柴油 +E 制冷剂

[0104] The calculation of indirect emissions in data center carbon emissions accounting includes:

[0105] Indirect emissions calculation formula:

[0106] Electricity emissions (core source):

[0107] Q 总 =∑(IT 耗电 ×PUE)

[0108] Among them, PUE stands for power usage efficiency, which reflects the energy efficiency of the data center. Used to amplify auxiliary power consumption outside IT equipment (such as cooling systems);

[0109] The carbon emissions of indirect emissions (scope 2) are as follows:

[0110] E 范围2 =(AD 总电 -AD 绿电 )×EF 电网 ×PUE

[0111] The calculation of the value chain emissions in data center carbon emissions accounting includes:

[0112] Carbon emissions throughout the server life cycle:

[0113]

[0114] User-side emission allocation:

[0115]

[0116] Among them, Q 总 represents the total annual electricity consumption of the data center (in MWh); P IT Indicates the power of IT equipment (in MW); P 辅助 Indicates the power of auxiliary equipment such as cooling / power supply (in MW) and meets P 辅助 =P IT ×(PUE-1); η represents the user computing power usage ratio (%), which defaults to 100% (all computing power is used for external services); EF 用户电网 Indicates the grid emission factor of the user's area (unit: t CO2 / MWh);

[0117] The carbon emissions of the value chain (scope 3) are as follows:

[0118]

[0119] In step S3:

[0120] The new green certificate is only related to the company's indirect emissions, and has no direct relationship with the company's carbon quota. If the green power project successfully applies for CCER, it will be directly linked to the company's carbon quota. According to the current carbon market requirements, only 5% of the proportion can be offset by CCER. The connection and deduction mechanism between the power green equity certificate and the carbon market is as follows. Figure 3 shown.

[0121] The linking mechanism between green certificates and the carbon trading market includes:

[0122] 1. Linking method between new green certificates and indirect emissions:

[0123] Essentially, the new green certificate is an alternative form of electricity consumption. The data center is only related to the indirect carbon emissions of the enterprise and has nothing to do with the direct emissions of the enterprise. The calculation stipulates that all electricity for the data center is taken from coal-fired units. For each additional green certificate purchased by the data center, the amount of electricity purchased by the enterprise is deducted by the amount of green certificate electricity, with the upper limit being the total electricity consumption of the enterprise. The formula for calculating indirect emissions from purchased electricity for the data center is:

[0124] E=(Q t -Q g )F f

[0125] Where E represents the indirect emissions of the emitting enterprises; Q t represents the total amount of electricity purchased by the emission-collecting enterprises; Q g Indicates the amount of electricity corresponding to the new green certificate; F f Represents the average carbon emission factor of thermal power generation. The relevant factors can be updated annually;

[0126] Since green certificates can be issued for a lifetime, with no time limit, as renewable energy generation costs decrease, the additionality of green electricity costs can be questioned. To promote proactive emissions reductions at the consumer end, the use of green certificates in the carbon market must be time-limited. Furthermore, to avoid double counting, green certificates used in indirect emissions accounting in the carbon market must be cancelled on the corresponding green certificate management platform. Different renewable energy power projects should have different, more scientific time limits. Numerous studies have shown that, after an average of six years, the levelized cost of electricity (LCO) for renewable energy power projects will be roughly equivalent to that of thermal power. Therefore, to further ensure relative scientificity and fairness, certain requirements are being established to improve the connection between renewable energy power projects and the carbon market.

[0127] First of all, for some existing onshore wind power and centralized photovoltaic power stations, the time of issuance of the green certificate is subtracted from the starting time of project card establishment during verification, and only green certificates issued within 6 years can be used.

[0128] Secondly, apart from onshore wind power and centralized photovoltaic projects, there is a gap in the green certificate policy for other green energy projects. Such projects can use the newly issued green certificates in Document No. 1044 for compensation. However, they can only use green certificates that are valid within 6 years of issuance. Due to the difference between the issuance time of the green certificate and the initial filing time of the project, the carbon market-related rights will be lost after 2030.

[0129] Finally, for new renewable energy power projects, the internationally commonly used 21-month system is adopted. When calculating the indirect emissions of green certificates, the matching electricity volume must be deducted within 21 months. This method is implemented according to the interval between the issuance time of the green certificate and the initial filing and power generation time of the project.

[0130] To prevent data duplication and overlap, green certificates participating in indirect carbon emission accounting must set reasonable time boundaries according to the characteristics of renewable energy power projects, and then cancel them immediately at the green certificate management end.

[0131] 2. Proportional constraints on the link between new green certificates and the carbon market:

[0132] Increasing the proportion of wind power and photovoltaics in the energy mix can reduce carbon footprints, but this will directly reduce the dispatchable output of thermal power plants, leading to a decline in overall thermal power efficiency, which in turn will drive carbon emissions higher. Therefore, to achieve true emissions reductions, the carbon reductions from green power must be greater than the increases from reduced thermal power efficiency.

[0133] The model construction of proportional constraints includes the following:

[0134] Assume that the electricity emission factor is in represents the total emissions from the power industry, Q' t Let θ represent the total power supply of the power industry and θ represent the proportion of renewable energy generation. Then the necessary and sufficient conditions for achieving emission reduction are:

[0135]

[0136] Because Q' t =Q' f +Q' g ,Q' f Indicates the thermal power supply, Q' g Indicates the green electricity supply, then:

[0137]

[0138] Derivatives of both sides with respect to θ:

[0139]

[0140] Solve the differential equation:

[0141]

[0142] We can get:

[0143]

[0144] From the above calculations, it can be concluded that if and only if Only when θ is equal to the thermal power emission factor, can real emission reduction be achieved. In this formula, c is a constant. The preceding analysis shows that renewable energy development across provinces and cities exhibits significant imbalances. Using a single mean as the proportional constraint standard would exacerbate regional imbalances. Each province's proportional constraint must correspond to its renewable energy consumption responsibility weight.

[0145] Therefore, the proportion of green electricity purchased by enterprises must exceed the provincial mandatory proportion of renewable energy before the corresponding green electricity can be included in the scope of indirect emission deductions. If the conditions are not met, it will be regarded as thermal power and changed in accordance with the country's dynamic control requirements for the annual absorption ratio. This will promote the effective absorption of clean energy, enhance the connection with the current mechanism, optimize the matching relationship between green electricity production and consumption, and focus on the implementation of absorption requirements by power-consuming units to improve carbon emission control results.

[0146] When the proportion of green electricity in data centers exceeds the provincial mandatory proportion of renewable energy, the proportion of green electricity in indirect emissions may be reduced accordingly; if it does not meet the standard, it will be calculated according to thermal power emissions. This proportion standard will be dynamically adapted according to the country's annually revised management and control requirements. This measure can effectively promote the consumption of renewable energy and seamlessly connect with the current system. At the same time, it can optimize the matching relationship between the power supply end and the power consumption end, focus on implementing the consumption obligations to key power-consuming enterprises, and further enhance the quality of emission reduction.

[0147] Example 2:

[0148] In order to verify the effectiveness of the evaluation method of the present invention, this example conducted the following specific applications and analyses:

[0149] 1. Practical application examples of carbon emission accounting for diesel generator sets

[0150] Take a diesel generator set that consumes 1000 tons of diesel per year as an example. Its basic parameters are activity data, AD = 1000t, low calorific value (NCV i ) is 42.652GJ / t, and the carbon content per unit calorific value (CC i ) is 0.0201tC / GJ, carbon oxidation rate (OF i ) is 99%, the total amount of purchased electricity AD 电 =500MWH, the grid emission factor is EF 电 =0.6101tCO2 / MWh, the calculation process is as follows.

[0151] Specific calculation formula for emission factors:

[0152]

[0153] Specific calculation formula for direct emissions:

[0154] E 燃烧 =∑(AD i ×EF i )=1000t×42.652GJ / t×0.0727tCO2 / GJ=3100tCO2

[0155] Specific calculation formula for indirect emissions:

[0156] E 电 =AD 电 ×EF 电 =500MWh×0.6101tCO2 / MWh=305.05tCO2

[0157] In this example, the diesel generator set emits 3,100 tons of carbon dioxide in a year, accounting for 93.7% of the data center's scope 1 emissions. This demonstrates the importance of the diesel generator set and reflects it in the data center's carbon footprint. This accounting method provides a standardized tool for data center facility-level accounting.

[0158] 2. Practical application example of annual carbon emissions accounting for a data center

[0159] The study selected a large data center located in Jiangsu Province. This data center has 20MW of IT equipment and operates 8,760 hours annually. Its energy mix includes diesel generators (backup power) and purchased electricity (including green power). The main parameters are shown in Table 2.

[0160] Table 2 Basic parameters of a data center

[0161]

[0162] Specific calculation formula for direct emissions:

[0163] Diesel combustion emission calculation:

[0164]

[0165] Refrigerant leakage and emission:

[0166] E 制冷 = Leakage volume (kg) × GWP HFC-134a (GWP=1430CO2-eq / kg)=50×1430=71.5tCO2

[0167] Calculation of total Scope 1 emissions:

[0168] E 范围1 =E 柴油 +E 制冷剂 =1,062.8+71.5=1,134.3? tCO2

[0169] Specific calculation formula for indirect emissions:

[0170] Calculation of total energy consumption:

[0171] Q 总 =∑(IT 耗电 ×PUE)=20? MW×8760? h×1.35=236,520? MWh

[0172] Calculation of total Scope 2 emissions:

[0173] E 范围2 =(AD 总电 -AD 绿电 )×EF 电网 ×PUE=(120000-18000)×0.5812=58292.4t? CO2

[0174] Specific calculation of value chain emissions:

[0175] The implicit carbon emissions of servers, referring to the industry average of 1.5tCO2 / unit, are calculated as follows:

[0176]

[0177] User-side allocated emissions, based on the proportion of IT equipment energy consumption, are calculated as follows:

[0178]

[0179] Calculation of total Scope 3 emissions:

[0180] E 范围3 =E 服务器 +E 用户 =2,400+9,857.1=12,257.1t? CO2

[0181] Total carbon emissions summary:

[0182] Table 3 Carbon emission accounting results

[0183]

[0184] The carbon emissions accounting results in Table 3 verify the feasibility of this model. It accurately quantifies Scope 1-3 emissions, with Scope 2 emissions accounting for 85.5% of the total. This confirms the importance of purchased electricity and the crucial role of clean electricity in reducing carbon emissions in data centers. Scope 3 emissions account for 12.4% of the total, demonstrating the critical importance of server management and the importance of guiding energy consumption on the user side. These are both directions for future emissions reduction. This accounting system employs two approaches: producer responsibility, which targets direct emissions, and a proportional allocation approach, which addresses value chain emissions. This clearly defines responsibilities and provides a comprehensive picture of the carbon footprint distribution.

[0185] 3. Case Study Analysis of Green Electricity Consumption Indicators for Data Centers

[0186] The electricity consumption structure of any regulated enterprise is comprised of four components: contracted green power purchases, self-generated power generation, on-site power generation, and general power contracts or power purchasing agents. The only difference is the proportions. Of course, in addition to purchasing green power, they also consider purchasing green certificates.

[0187] Therefore, the accounting of indirect emissions can be done according to Figure 1 Steps to complete, including:

[0188] Step 1: Determine the parameters of the four parts of the data center. For example, the electricity source composition of the data centers from the four pilot carbon markets in 2024 is 58.17% from general electricity contracts or grid-purchased electricity, 40.96% from contracted green electricity, 0.65% from on-site third-party power generation facilities, and 0.22% from self-owned photovoltaic power generation.

[0189] The second step is to determine whether the green power consumption ratio exceeds the province's renewable energy consumption ratio. If the data center's green power consumption ratio is lower than the province's green power consumption ratio weight, it will be treated as thermal power. If it is higher, the next step is to determine the timeliness of the data center.

[0190] Step 3: To determine the timeliness of green electricity, contracted green electricity purchases, self-generated power generation, and on-site power generation all involve green electricity deductions. You can directly obtain the corresponding "new green certificate" based on the green electricity project type. The green certificate must include the project's card creation date, project type, and certificate issuance date. Linking is done using the linking method described above.

[0191] Step 4: The green certificates purchased by enterprises through the green certificate market also need to be linked according to the linking method.

[0192] Step 5: Calculate indirect emissions based on the data center's purchased electricity accounting formula.

[0193] There are currently four main calculation methods, as shown in Table 4.

[0194] Table 4 Comparison of different methods

[0195]

[0196] Imagine that a data center in Shanghai consumes 100 million kWh of electricity a year, 40% of which is certified by the new green certificate, and 60% of the electricity that does not meet the standard is classified as thermal power, that is, 60 million kWh. In view of data integrity, 2024 was selected for accounting work. Referring to the 2025 edition of the China Energy Statistical Yearbook, the average carbon emission factor of thermal power calculated in 2024 was 0.794 kg / ; in 2024, the average carbon emission factor of the national power grid confirmed by the Ministry of Ecology and Environment was 0.58 kg / ; the emission factor of the power grid in East China was 0.77865 kg per kilowatt-hour; the unit emission of the Shanghai power grid was 0.42 kg / . The Carbon Disclosure Project (CDP) 2024 green certificate emission reduction accounting value shows that the difference in regional power grid emission factors will cause the CCER amount to fluctuate between 0.657 tons and 0.872 tons.

[0197] Table 5 Comparison of results of different linking methods

[0198]

[0199] It can be seen from the data in Table 5 that the calculation results of each method are significantly different. The result obtained by Method 4 is at the lowest level, Method 1 reaches the highest value, and Method 3 adopts a dual-factor accounting model. From the perspective of data duplication, Method 1 is different from other methods and does not have the phenomenon of repeated calculation. The other three models all have the problem of repeated statistics. They all deduct the green electricity related to green certificates at the power grid level, and simultaneously reduce the green electricity portion of the company's purchased electricity. In terms of fairness, Method 1 and Method 2 have no negative effect on fairness. The fairness of Method 3 and Method 4 needs further evaluation. The actual application effects of the same green certificates in different regions are different, which easily causes green certificates to converge to power grids with high carbon intensity.

[0200] Method 1 ensures the uniqueness of environmental rights by directly deducting the amount of electricity corresponding to green certificates from the total electricity purchased. The remaining electricity is then calculated uniformly using the average emission factor for thermal power generation, ensuring the uniqueness of environmental rights. This mechanism design does not rely on regional grid emission factors, fundamentally avoiding the risk of multiple accounting for the same green electricity's environmental attributes (e.g., the reuse of green certificates, regional grid factors, and CCERs). Furthermore, by using a nationally unified thermal power emission factor, it is unaffected by regional grid structure variations. For example, in this example, regardless of whether the data center is located in Shanghai (grid factor 0.42) or North China (grid factor 0.778), the emission reduction effect of green certificates is reflected through the thermal power factor (0.794), avoiding incentive distortions caused by geographic location. Method 1 does not affect the emission reduction effects of green electricity consumption due to regional power structure variations. For example, if a region has a low grid factor, the emission reduction benefits obtained by enterprises through green certificates may be underestimated; conversely, in high-emission regions, the benefits may be overestimated. By standardizing the thermal power factor, Method 1 ensures that all enterprises are accounted for using the same standard, thus maintaining market fairness. Furthermore, Method 1 only requires calculating the amount of electricity corresponding to the Green Certificate, eliminating the need to track changes in regional grid factors or deal with complex regional differences, reducing data center accounting costs and regulatory complexity. Therefore, Method 1 is superior and offers universal applicability, fairness, and operability.

[0201] According to the above design, the electricity consumption of any enterprise that consumes electricity can be clearly divided into a green portion and a non-green portion. The green portion can be directly accounted for using a zero-emissions factor, while the non-green portion is accounted for using the average emission factor of thermal power plants. This approach not only avoids the double counting associated with indirect electricity emissions, but also provides greater clarity and incentives for electricity users to actively consume electricity.

Claims

1. A method for evaluating green electricity consumption indicators of data centers based on carbon emission calculation, characterized in that: The steps include: S1: Calculate the carbon emissions of diesel-generator units based on the total fuel consumption, low calorific value, and carbon oxidation rate, and establish a carbon emissions calculation model for diesel-generator units; S2: Considering the sources of carbon emissions from data centers, a "producer + apportionment" hybrid accounting model is used to calculate the carbon emissions of data centers. S3: Based on the linkage mechanism between green certificates and carbon trading markets, and according to the carbon emission accounting results of diesel generator sets and data centers, the evaluation results of the green electricity consumption indicators of data centers are obtained.

2. The method for evaluating green electricity consumption indicators of a data center based on carbon emission calculation according to claim 1 is characterized in that: In step S1, the emission sources of the diesel generator set are divided into direct emissions and indirect emissions, and the carbon emissions of the diesel generator set are calculated by quantifying the carbon dioxide emissions generated by diesel combustion.

3. The method for evaluating green electricity consumption indicators of a data center based on carbon emission calculation according to claim 2 is characterized in that: The calculation of the direct emissions portion of the carbon emissions accounting of the diesel generator set in step S1 includes: Direct emissions calculation formula: AND 燃烧 =∑(AD i ×EF i ) Among them, AD i Indicates that activity data refers to diesel consumption; EF i Expressed as an emission factor, the calculation formula is: Among them, CC i Indicates the carbon content per unit calorific value, which refers to the carbon content of diesel; i Indicates the carbon oxidation rate, which refers to the carbon oxidation rate of diesel combustion; NCV i Indicates low calorific value, which refers to the low calorific value of diesel.

4. The method for evaluating green electricity consumption indicators of a data center based on carbon emission calculation according to claim 2 is characterized in that: The calculation of the indirect emissions in the carbon emissions accounting of the diesel generator set in step S1 includes: Indirect emissions calculation formula: AND 电 =∑(AD 电 ×EF 电 ) Among them, AD 电 Indicates purchased electricity; EF 电 represents the emission factor.

5. The method for evaluating green electricity consumption indicators of a data center based on carbon emission calculation according to claim 1 is characterized in that: In step S2, the carbon emission sources of the data center are divided into direct emissions, indirect emissions and value chain emissions. Direct emissions include fuel combustion and refrigerant leakage, indirect emissions include purchased electricity and heat, and renewable energy offsets. Value chain emissions are divided into upstream and downstream. Upstream emissions include implicit carbon emissions from server manufacturing and building material production, while downstream emissions are energy consumption of terminal devices generated by users using cloud services.

6. The method for evaluating green electricity consumption indicators of a data center based on carbon emission measurement according to claim 5 is characterized in that: The calculation of the direct emissions portion of the data center carbon emissions accounting in step S2 includes: Direct emissions calculation formula: Diesel generator emissions: Refrigerant leakage and emission: E 制冷 = Leakage volume (kg) × GWP HFC-134a (GWP = 1430 CO2-eq / kg) The direct carbon emissions are as follows: AND 范围1 =And 柴油 +E 制冷剂 。 7. The method for evaluating green electricity consumption indicators of a data center based on carbon emission calculation according to claim 5, characterized in that: The calculation of the indirect emissions in the data center carbon emissions accounting in step S2 includes: Indirect emissions calculation formula: Electricity emissions: Q 总 =∑(IT 耗电 ×PUE) Among them, PUE stands for power usage effectiveness, which reflects the energy efficiency of the data center and is used to amplify the auxiliary power consumption outside of IT equipment; The indirect carbon emissions are as follows: AND 范围2 =(AD 总电 -TO 绿电 )×EF 电网 ×PUE。 8. The method for evaluating green electricity consumption indicators of a data center based on carbon emission calculation according to claim 5 is characterized in that: The calculation of the value chain emissions portion of the data center carbon emissions accounting in step S2 includes: Carbon emissions throughout the server life cycle: User-side emission allocation: Among them, Q 总 represents the total annual electricity consumption of the data center; P IT Indicates the power of IT equipment; P 辅助 Indicates the power of auxiliary equipment such as cooling / power supply, and satisfies P 辅助 =P IT ×(PUE-1); η represents the user's computing power usage ratio; EF 用户电网 Indicates the emission factor of the power grid in the user's area; The carbon emissions from the value chain are as follows:

9. The method for evaluating green electricity consumption indicators of a data center based on carbon emission measurement according to claim 1, characterized in that: The linking mechanism between green certificates and the carbon trading market in step S3 includes: Linking method of new green certificates and indirect emissions: For every additional green certificates purchased by the data center, the amount of electricity purchased by the enterprise is deducted according to the amount of green certificates, with the upper limit being the total electricity consumption of the enterprise. The indirect emission accounting formula for the purchased electricity of the data center is E=(Q t -Q g )F f Where E represents the indirect emissions of the emitting enterprises; Q t represents the total amount of electricity purchased by the emission-collecting enterprises; Q g Indicates the amount of electricity corresponding to the new green certificate; F f Represents the average carbon emission factor of thermal power generation. The relevant factors can be updated annually; Proportional constraints on the link between new green certificates and the carbon market: Assume that the electricity emission factor is in represents the total emissions from the power industry, Q' t Let θ represent the total power supply of the power industry and θ represent the proportion of renewable energy generation. Then the necessary and sufficient conditions for achieving emission reduction are: Because Q' t =Q' f +Q' g ,Q' f Indicates the thermal power supply, Q' g Indicates the green electricity supply, then: Derivatives of both sides with respect to θ: Solve the differential equation: We can get: From the above calculations, it can be concluded that if and only if Only when , can real emission reduction be achieved, where c is a constant.

Images