Method for testing reliability of carbon emission data of power plant

Abstract

The invention discloses a method for checking the reliability of carbon emission data of a power plant. The method comprises the following steps: 1, collecting data of coal as fired and coal as fired; 2, calculating the relative mean square deviation of monthly data according to the relative difference between the data of the coal as fired and the data of the coal as fired through month-by-month comparison; 3, calculating the relative difference distribution of annual data according to the obtained relative mean square deviation of the monthly data; 4, obtaining a confidence interval of the relative difference of the annual data according to the relative difference distribution of the annual data; 5, judging whether the carbon emission data of the power plant is reliable or not; according to the method, the accumulated characteristic difference of the data deviation between the in-factory end and the in-furnace end caused by different reasons is utilized, the in-factory-in-furnace coal data difference part caused by the time difference between in-factory coal and in-furnace coal is effectively eliminated, the measurement error is separated out, and the accuracy and the credibility are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of reliability judgment of power plant data, and in particular to a method for verifying the reliability of power plant carbon emission data by comparing data of incoming coal and incoming coal. Background Art

[0002] At present, the emission factor method, mass balance method and actual measurement method are the three most widely recognized accounting methods. The emission factor method is a macro method and is not very applicable to the carbon emission accounting of a certain power plant. The actual measurement method requires accurate measurement of the carbon dioxide concentration (distribution) and flue gas volume in the flue gas, which is difficult and costly.

[0003] The mass balance method is highly applicable in the power production end, and the emissions are calculated by the carbon content and usage of the fuel. It can provide relatively accurate measurement at a low cost. According to the "Guidelines for the Accounting and Reporting of Greenhouse Gas Emissions for Enterprises - Power Generation Facilities" (hereinafter referred to as the "Guidelines"), according to the "Guidelines", the elemental carbon content should be the received elemental carbon content. For coal-fired power plants, the coal is measured in one of three ways: 1) Daily testing, using the daily coal test data to calculate the monthly average received elemental carbon content of the coal entering the furnace, and the weight is the daily coal consumption; 2) Monthly reduced sample testing, collect the reduced samples of the coal entering the furnace every day, and merge and mix the daily reduced samples every month to test their elemental carbon content. The received elemental carbon content of fuel oil and gas should be tested at least monthly; 3) Each batch of testing, using the monthly batch of coal test data to calculate the monthly average received elemental carbon content of the coal entering the factory, and the weight is the amount of coal received in each batch of the factory. When measuring fossil fuel consumption, the data used are prioritized as follows: a) data recorded in the production system; b) data in the purchase and sales ledger; c) settlement voucher data provided by the supplier.

[0004] Since the fossil fuel consumption and its composition must correspond one to one when calculating carbon emissions, when using the element balance method to calculate coal-fired power carbon emissions, if the consumption data recorded by a higher-priority production system is used, daily testing or monthly reduced sampling testing must be used to measure the elemental carbon content.

[0005] In order to control the cost of testing, power plants usually use the monthly reduced sampling method to calculate carbon emissions. However, this method requires the daily amount of coal entering the furnace as a weight for sample mixing, which increases the possibility of human intervention and may lead to inaccurate calculation of the carbon content of coal types. The quality measurement of coal entering the furnace may also have errors. If carbon emissions are calculated based on the coal entering the plant, the carbon content of the fuel uses the above method 3) (test each batch), and the fuel consumption uses the data from sources b (purchase and sales ledger) and c (supplier settlement voucher) above. This method may lead to inaccurate measurement of the carbon content of coal types due to poor sampling representativeness, and the annual production data and carbon emission data may not fully correspond.

[0006] Therefore, using the quality and quantity of coal entering the furnace as the basis for carbon emission calculations, and using the quantity and quality of coal entering the factory as verification of the reliability of carbon emission data, can detect and eliminate data errors as much as possible, and improve the accuracy and credibility of carbon emission detection. Summary of the invention

[0007] In order to solve the problems existing in the prior art, the present invention provides a statistical method for comparing and analyzing the carbon emissions calculated by the amount of coal entering the plant and the quality of coal with the carbon emissions calculated by the amount of coal entering the furnace and the quality of coal to verify the reliability of the data.

[0008] The technical solution of the present invention is as follows:

[0009] A method for verifying the reliability of carbon emission data of a power plant comprises the following steps:

[0010] Step 1: Collect data on coal entering the plant and the furnace, including one or more of the coal quantity, carbon content of the coal, and carbon emissions;

[0011] Step 2: Compare the relative differences between the coal entering the plant and the coal entering the furnace on a monthly basis and calculate the relative mean square deviation of the monthly data;

[0012] Step 3: Based on the relative mean square deviation of the monthly data, calculate the relative difference distribution of the annual data;

[0013] Step 4: Obtain the confidence interval of the root mean square of the relative difference of annual data based on the distribution of the relative difference of annual data;

[0014] Step 5: Determine whether the power plant carbon emission data is reliable based on whether the root mean square of the relative difference of the annual data falls within the confidence interval.

[0015] Furthermore, taking coal quantity data as an example, the carbon content and carbon emissions of coal can be applied to the following formula. When applying, the coal quantity data can be replaced with the carbon content of coal (average carbon content data) or carbon emission data. The calculated process data and result data are the carbon content or carbon emission data of coal, and the reliability judgment becomes the reliability of the corresponding data.

[0016] The step 2 comprises the following steps:

[0017] Step 2.1: The difference between the coal entering the plant and the coal entering the furnace is calculated as follows:

[0018] a=yx (1)

[0019] a is the absolute difference of monthly data, x is the monthly coal quantity entering the furnace, and y is the monthly coal quantity entering the plant;

[0020] Step 2.2: The relative mean square deviation of the monthly data is:

[0021]

[0022] MSEr is the relative mean square deviation of coal quantity, a i is the difference between the coal entering the plant and the coal entering the furnace in the i-th month, m is the total number of months for statistics, is the average value of the coal data x arriving at the plant within m months;

[0023] Step 2.3: The relative mean square deviation is the sum of the part caused by the error and the part caused by the time difference; that is:

[0024] MSEr=MSEr1+MSEr2 (3)

[0025] MSEr is the relative mean square error, MSEr1 is the relative mean square error caused by error, and MSEr2 is the relative mean square error caused by time difference.

[0026] Furthermore, the step 3 comprises the following steps:

[0027] Step 3.1: The relative difference distribution of annual data is:

[0028]

[0029] B is the relative difference of annual data, expressed as follows:

[0030]

[0031] Among them, Y is the coal data arriving at the factory each year, and X is the coal data used each year;

[0032] M(0,σ 2 ) is the mathematical expectation of 0 and the variance of σ 2 The probability distribution function of

[0033] The standard deviation σ0 of the probability distribution of B is:

[0034]

[0035] is the probability distribution variance of the annual data deviation caused by the error, is the probability distribution variance of the annual data deviation caused by the time difference;

[0036] The confidence interval of the relative difference of annual data is [-k×σ0,k×σ0], where k is the coverage factor.

[0037] Furthermore, step 4 includes the following steps:

[0038] Step 4.1: Set the annual allowable error E;

[0039] Step 4.2: Set MSER1 = 12 × E 2 Substitute into formula (3) and combine with formula (2) to calculate MSEr2;

[0040] Step 4.3: Substitute MSEr1 and MSEr2 into formula (7) to calculate the standard deviation σ0 of the relative difference of annual data;

[0041] Step 4.4: The root mean square of the relative difference B of the annual data for n consecutive years is defined as:

[0042]

[0043] n is the number of years. When |RMSEB|≥k×σ, the data are considered unreliable, otherwise the data are considered reliable.

[0044] When calculating, it is assumed that the MSER1 and MSER2 data of the power plant for several consecutive years have not changed, so the MSER1 and MSER2 of n years are averaged as the MSER1 and MSER2 data of the plant every year.

[0045] Furthermore, the k value is obtained by computer as follows:

[0046] Generate n normally distributed random numbers with an expectation of 0 and a variance of 1, and find the root mean square of these numbers;

[0047] Repeat the above steps 10,000 times and record all generated RMS values;

[0048] Sort the 10,000 RMS values ​​from largest to smallest and number them from 1 to 10,000 in order;

[0049] The root mean square numbered 10000×sig is k, where sig is the significance level, which ranges from 0.001 to 0.05.

[0050] Beneficial effects of the present invention:

[0051] The present invention compares and calculates the mean square deviation of short-term (monthly) data at the plant entrance and furnace entrance and the deviation of long-term (yearly) data to determine whether the calculated errors of the data at the furnace entrance and the plant entrance are within the allowable range. The innovation lies in that the accumulated characteristic differences of the data deviations at the plant entrance and furnace entrance caused by different reasons are utilized to effectively eliminate the difference in the plant-furnace coal data caused by the time difference between coal entering the plant and the furnace, and the calculated errors are separated, filling the deficiency of the reliability verification means of carbon emission data calculated by the mass balance method, solving the problem that the time difference between coal entering the plant and the furnace in different power plants is different, making it difficult to directly compare the furnace-plant coal data difference and judge the data reliability with a unified standard, thereby improving the accuracy and credibility.

[0052] Since the invention uses data commonly recorded by power plants and their coal suppliers, it hardly increases costs and workload, thereby strengthening the accuracy advantage of the mass balance method in calculating carbon emission data while retaining its low-cost advantage. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0054] The present invention is further described below in conjunction with the accompanying drawings and embodiments.

[0055] Example:

[0056] This invention is further explained through the following actual test on the reliability of the carbon emission data of Power Plants A and B for two years. This test used the quality and carbon content data of all batches of coal entering Power Plants A and B during 2021-2022 (the amount of coal quality data for each plant exceeded 1,300 per year), as well as the monthly incoming coal quality and carbon content of the reduced coal samples obtained by the monthly reduced sampling detection method, and the total monthly carbon emissions data calculated based on this, and the fuel oxidation rate was 99%.

[0057] Using the coal data entering the plant, the monthly carbon emissions of the two plants (at the plant entrance) were calculated using formula (1).

[0058] The absolute differences in coal quantity, carbon content and carbon emissions between the factory and furnace end each month are calculated using formula (1).

[0059] The mean square deviation (MSEr) of the coal quantity, carbon content and carbon emission data at the plant and furnace ends each month is calculated by formula (2).

[0060] The annual data allowable error of coal quantity is E' = 0.2%, and the annual data allowable error of coal carbon content and carbon emission is E = E" = 0.5%. Taking the coal quantity data as an example, MSER1 = 12 × E 2 , the carbon content and carbon emission data of coal are calculated using the same method.

[0061] Substitute the MSER value and MSER1 value into formula (3) to calculate MSER2. The data are shown in Tables 1 and 2.

[0062] Table 1 Carbon emission data of Power Plant A in 2021-2022

[0063]

[0064]

[0065] Table 2 Carbon emission data of Power Plant B in 2021-2022

[0066]

[0067] The MSEr2 of the two years is averaged to obtain a new MSEr2 value, which is substituted into equation (7) to calculate the standard deviation σ0 of the relative difference between the annual data of the two power plants using the data of the two years.

[0068] With the help of a computer, we can find k in the following way:

[0069] Generate two normally distributed random numbers with expectation 0 and variance 1, and find the RMS of the two numbers.

[0070] Repeat the above steps 10,000 times and record all the generated RMS.

[0071] Sort the 10,000 RMS values ​​from large to small and number them from 1 to 10,000 in order. The RMS value numbered 10,000×0.05 is k. The Matlab code to implement this function is as follows:

[0072] for i=1:10000; % loop statement

[0073] r(i,:)=randn(1,2); % Generate two standard normal distribution random numbers in the i-th row of matrix r

[0074] s(i)=rms(r(i,:)); % Calculate the root mean square of all elements in the i-th row of matrix r and store it as s

[0075] end

[0076] s1=sort(s,'descend'); % Arrange all numbers in matrix s in descending order and store them as matrix s1

[0077] k=s1(0.05*10000)%Find the 500th number in s1 and assign its value to k

[0078] The above method obtains k=1.71, and the allowable root mean square deviation of the annual data is k×σ.

[0079] Then, according to the deviation of the actual annual data, the actual root mean square deviation is calculated: The obtained data are shown in Table 3 and Table 4.

[0080] Table 3 Reliability analysis of carbon emission data of Power Plant A in 2021-2022

[0081]

[0082]

[0083] Table 4 Reliability analysis of carbon emission data of Power Plant B in 2021-2022

[0084]

[0085] According to the data in the table, the actual RMS deviation is smaller than the allowable RMS deviation, and the carbon emission data of the two power plants should be considered reliable.

Claims

1. A method for verifying the reliability of carbon emission data of a power plant, characterized in that: The following steps are involved: Step 1: Collect data on coal entering the plant and the furnace, including one or more of the coal quantity, carbon content of the coal, and carbon emissions; Step 2: Compare the relative differences between the coal entering the plant and the coal entering the furnace on a monthly basis and calculate the relative mean square deviation of the monthly data; Step 3: Based on the relative mean square deviation of the monthly data, calculate the relative difference distribution of the annual data; Step 4: Obtain the confidence interval of the root mean square of the relative difference of annual data based on the distribution of the relative difference of annual data; Step 5: Determine whether the power plant carbon emission data is reliable based on whether the root mean square of the relative difference of the annual data falls within the confidence interval.

2. A method for verifying the reliability of carbon emission data of a power plant according to claim 1, characterized in that: The step 2 comprises the following steps: Step 2.1: The difference between the coal entering the plant and the coal entering the furnace is calculated as follows: a=yx (1) a is the absolute difference of monthly data, x is the monthly coal quantity entering the furnace, and y is the monthly coal quantity entering the plant; Step 2.2: The relative mean square deviation of the monthly data is: MSEr is the relative mean square deviation of coal quantity, a i is the difference between the coal entering the plant and the coal entering the furnace in the i-th month, m is the total number of months for statistics, is the average value of the coal data x arriving at the plant within m months; Step 2.3: The relative mean square deviation is the sum of the part caused by the error and the part caused by the time difference; that is: MSEr=MSEr1+MSEr2 (3) MSEr is the relative mean square error, MSEr1 is the relative mean square error caused by error, and MSEr2 is the relative mean square error caused by time difference.

3. A method for verifying the reliability of carbon emission data of a power plant according to claim 2, characterized in that: The step 3 comprises the following steps: Step 3.1: The relative difference distribution of annual data is: The relative difference of annual data is expressed as follows: Among them, Y is the coal data arriving at the factory each year, and X is the coal data used each year; M(0,σ 2 ) is the mathematical expectation of 0 and the variance of σ 2 The probability distribution function of The standard deviation σ0 of the probability distribution of B is: is the probability distribution variance of the annual data deviation caused by the error, is the probability distribution variance of the annual data deviation caused by the time difference; The confidence interval of the relative difference of annual data is [-k×σ, k×σ], where k is the coverage factor.

4. A method for verifying the reliability of carbon emission data of a power plant according to claim 3, characterized in that: The step 4 comprises the following steps: Step 4.1: Set the annual allowable error E; Step 4.2: Set MSER1 = 12 × E 2 Substitute into formula (3) and combine with formula (2) to calculate MSEr2; Step 4.3: Substitute MSEr1 and MSEr2 into formula (7) to calculate the standard deviation σ0 of the relative difference of annual data; Step 4.4: The root mean square of the relative difference B of the annual data for n consecutive years is: n is the number of years. When |RMSEB|≥k×σ, the data are considered unreliable, otherwise the data are considered reliable; During the calculation, it is assumed that the MSER1 and MSER2 data of the power plant for several consecutive years have not changed, so the MSER1 and MSER2 of n years are averaged as the MSER1 and MSER2 data of the plant every year.

5. A method for verifying the reliability of carbon emission data of a power plant according to claim 3, characterized in that: The k value is obtained by computer as follows: Generate n normally distributed random numbers with an expectation of 0 and a variance of 1, and find the root mean square of these numbers; Repeat the above steps 10,000 times and record all generated RMS values; Sort the 10,000 RMS values ​​from largest to smallest and number them from 1 to 10,000 in order; The root mean square numbered 10000×sig is k, where sig is the significance level, which ranges from 0.001 to 0.05.

Images