Energy storage system configuration method for assisting primary frequency modulation of thermal power plant

Abstract

The invention discloses an energy storage system configuration method for assisting primary frequency modulation of a thermal power plant. According to the method, long-time scale 15s output difference data and 30s output difference data are formed for rated power configuration, probability density estimation is carried out on the 15s output difference data and the 30s output difference data respectively to obtain a Gaussian probability density function and parameters thereof, and the Gaussian probability density function and the parameters thereof are converted into standard Gaussian distribution. According to the probability density function and the assessment amount, a relation curve between the energy storage rated power value and the 15s and 30s output response assessment amount of each year is obtained, and finally, energy storage power configuration is obtained through a direct search optimization method. For rated energy configuration, contribution electric quantity data of a unit in a relatively long time scale are obtained, and energy storage energy configuration is obtained by adopting a process similar to rated power configuration. The energy storage system is configured from the perspective of the whole life cycle, the later operation and maintenance cost is considered, and the economy of the whole life cycle of the project is improved.

Description

Technical Field

[0001] The present invention relates to the field of energy storage configuration for assisting frequency regulation of thermal power plants, and more specifically, to a configuration method for an energy storage system for assisting primary frequency regulation of thermal power plants. Background Art

[0004] With the gradual increase in the penetration rate of new energy power generation access in the power system, the contradiction between the power grid frequency regulation demand and the power grid frequency regulation resources has become increasingly prominent. In the "two detailed rules" of the North China region, the performance assessment of primary frequency regulation is divided into two cases: small disturbance and large disturbance. Among them, the maximum frequency deviation of the power grid does not exceed 0.06Hz for small disturbance, and the maximum frequency deviation of the power grid is greater than 0.06Hz for large disturbance. The assessment coefficients of the two cases are different. The large-scale access of non-synchronous power sources has made the frequency fluctuation of the power grid increasingly significant, and has put forward higher requirements for the primary frequency regulation process of grid-connected generators. Since the "two detailed rules" stipulate certain assessment indicators and assessment coefficients, such as 15s output response index, 30s output response index, power contribution index, etc., grid-connected power plants will face increasingly severe primary frequency regulation assessment pressure.

[0005] Due to the limitation of unit performance, configuring energy storage devices to assist the primary frequency regulation of grid-connected generator sets has become an effective means to improve the unit frequency regulation performance and reduce the power consumption of primary frequency regulation assessment. In the scenario of assisting the primary frequency regulation of units, electrochemical energy storage systems with a charge and discharge rate of 1C / 2C or power-type energy storage devices such as supercapacitors and flywheel energy storage are usually used. High-rate electrochemical energy storage and high-power physical energy storage often have large investment costs. It is necessary to reasonably determine the power capacity and energy capacity of the energy storage device on the premise of meeting the unit regulation performance requirements and reducing the unit assessment costs, so as to achieve the optimal overall economy throughout the life cycle.

[0006] In the auxiliary unit primary frequency regulation scenario, the existing technology has the following problems:

[0007] a) The existing technical solutions are not fully combined with the requirements of regional frequency regulation performance indicators, resulting in the problem that the planning of joint primary frequency regulation of thermal storage in specific areas is not targeted;

[0008] b) The existing technical solutions do not conduct statistical density analysis on long-term operation data, resulting in the inability to scientifically plan the configuration capacity of auxiliary primary frequency regulation energy storage;

[0009] c) The existing technical solutions do not configure the energy storage power capacity and energy capacity separately, resulting in the problem that the power capacity or energy capacity of the energy storage system cannot be fully utilized;

[0010] d) The existing technical solutions do not configure the energy storage system from the perspective of the entire life cycle, resulting in insufficient consideration of the subsequent operation and maintenance costs and poor economic efficiency throughout the entire life cycle. Summary of the invention

[0011] The technical problem to be solved by the present invention is to overcome the deficiencies in the prior art and provide a method for configuring an energy storage system for assisting primary frequency regulation of a thermal power plant.

[0012] The present invention provides a method for configuring an energy storage system for assisting primary frequency regulation of a thermal power plant, which is performed according to the following steps:

[0013] a) Acquisition and processing of data;

[0014] For the energy storage rated power configuration of the auxiliary thermal power plant primary frequency regulation, obtain the 15s actual output extreme MW, 15s theoretical output extreme MW, and the actual output MW at the start of the primary frequency regulation data of the unit for a long period of 1 or 2 years, and calculate the 15s output response index. At the same time, obtain the 30s actual output extreme MW, 30s theoretical output extreme MW, and the actual output MW at the start of the primary frequency regulation data of the unit for 1 or 2 years, and calculate the 30s output response index;

[0015] b) Extract the 15s output difference, 30s output difference and contribution power difference when the assessment occurs;

[0016] Screen out the cases where the 15s output response index is lower than 75% and the 30s output response index is lower than 100%, which are the cases where the 15s output and 30s output are assessed; calculate the 15s output difference and 30s output difference each time the assessment occurs, and form the long-term 15s output difference data and 30s output difference data;

[0017]

[0018] In the formula, is the 15s output difference during the i-th assessment, is the theoretical output extreme value of 15s during the i-th assessment, is the actual output extreme value of 15s during the i-th assessment, is the 30-second output difference during the i-th assessment, is the theoretical output extreme value of 30s during the i-th assessment, is the actual output extreme value of 30s during the i-th assessment, and n is the total number of assessments occurring within a certain period of time;

[0019] c) Probability density estimation is performed on each group of data, and about 200-300 probability density data points are selected; that is, probability density estimation is performed on the 15s output difference data, the probability density function of the 15s output difference data is calculated, and about 200-300 probability density data points are uniformly selected within the maximum and minimum value range of the 15s output difference data;

[0020] d) Perform one-dimensional Gaussian distribution fitting on the selected probability density data points to obtain the Gaussian distribution probability density function, probability distribution function and probability density parameters of the 15s output difference: mean and variance; perform the same processing on the 30s output difference data to obtain the Gaussian distribution probability density function, probability distribution function and probability density parameters of the 30s output difference: mean and variance;

[0021] e) extracting fitting parameters to form probability density curve and probability distribution curve;

[0022]

[0023] In the formula, ΔP 15s is the 15s output difference random variable, μ 15s is the Gaussian distribution mean of the 15s output difference, σ 15s is the standard deviation of the Gaussian distribution of the 15s output difference, ΔP 30s is the 30s output difference random variable, μ 30s is the Gaussian distribution mean of the 30s output difference, σ 30s is the standard deviation of the Gaussian distribution of the 15s output difference, and N is the normal distribution;

[0024] f) Convert the probability density curve into a standard normal distribution;

[0025] The obtained 15s output difference and 30s output difference are converted into standard normal distribution and used to calculate the probability value within a certain interval;

[0026]

[0027] In the formula, is the standard normal distribution random variable of the 15s output difference, is the standard normal distribution random variable of the 30s output difference;

[0028] g) Give the corresponding relationship between power configuration and electricity cost assessment;

[0029] According to the probability density function of the 15s output difference, the relationship between the rated power value of the energy storage and the probability of compensating the 15s output difference is calculated. According to the average number of times the 15s output response index of the unit is assessed each year, the amount of electricity lost in each 15s output response assessment and the corresponding assessment amount, the relationship curve between the rated power value of the energy storage and the annual 15s output response assessment amount is obtained;

[0030] Based on the probability density function of the 30s output difference, the relationship curve between the energy storage rated power value and the annual 30s output response assessment amount is obtained;

[0031] EXA 15s =N 15s *Q15s *(1-P(P S =P0)) (7)

[0032] EXA 30s =N 30s *Q 30s *(1-P(P S =P0)) (8)

[0033] Where, EXA 15s is the total cost of annual 15s output assessment, N 15s Q is the total number of annual 15s output assessments before allocation of storage. 15s The cost of a single 15s output assessment, EXA 30s is the total cost of annual 30s output assessment, N 30s Q is the total number of annual 30s output assessments before storage allocation, 30s is the cost of a single 30s output assessment, P S is the power configuration of the energy storage system, P0 is a selected power value, P(P S =P0) is P S When P0 is taken, the probability of compensating the number of 15s output assessment times can be achieved;

[0034] Based on the relationship between the rated power value of energy storage and the output response assessment amount at the above two time scales, an optimization objective function is constructed, and the optimization objective function is constructed with the levelized unit power profit (i.e., levelized unit power benefit - levelized unit power cost). The levelized unit power benefit is the assessment fee reduced by the annual average unit power energy storage, and the levelized unit power cost is the annual average unit power energy storage configuration cost, as shown in formula (9);

[0035]

[0036] In the formula, d is the discount rate, m is the operating life of the energy storage equipment, C P is the unit power cost (yuan / kW), C E is the unit energy cost (yuan / kWh), T S is the energy storage discharge time;

[0037] h) Taking the maximization of levelized unit power revenue as the goal, a direct search method is used to determine the energy storage power configuration;

[0038] The constraints include: primary frequency modulation action frequency threshold constraint, speed inequality constraint, primary frequency modulation adjustment depth constraint, energy storage power constraint, energy storage self-discharge constraint, and energy storage SOC constraint; calculate the optimal energy storage rated power value in the 15s output and 30s output scenarios; calculate the 15s output and 30s output data respectively, and take the larger one as the energy storage rated calculated power, and consider that in a certain primary frequency modulation action, the energy storage device can meet both the 15s output index requirement and the 30s output index requirement, and consider that the instantaneous output of the energy storage device meets the requirement, which can meet the 15s output and 30s output requirements;

[0039] i) Taking the contribution of electricity as the research object:

[0040] Step 7: For the energy storage rated energy configuration for assisting the primary frequency regulation of the thermal power plant, obtain the actual power contribution and theoretical power contribution of the unit for one or two years and calculate the power contribution index;

[0041] Screen out the cases where the power contribution index is lower than 75%, which is the case where the contribution power is assessed; calculate the difference in the contribution power each time the assessment occurs, and form the long-term contribution power difference data;

[0042] Select a kernel function, perform kernel density estimation on the contribution power difference data, calculate the probability density function of the contribution power difference data, and uniformly select about 200-300 probability density data points within the maximum and minimum value range of the contribution power difference data; perform one-dimensional Gaussian distribution fitting on the selected probability density data points to obtain the contribution power difference Gaussian distribution probability density function, probability distribution function and probability density parameters: mean and variance;

[0043] The purpose of the normal distribution verification process is to verify whether the normal distribution probability density function matches the original output difference data. If the matching degree is not good, it can be adjusted to a distribution function such as Weibull distribution or t distribution, and the matching degree with the original data can be verified. The probability distribution with a better matching degree is selected, and the probability calculation within a certain interval is performed;

[0044] Step 8: According to the probability density function of the difference in contribution electricity, the relationship between the rated energy value of the energy storage and the probability of compensating the difference in contribution electricity is calculated. According to the average number of times the unit's electricity contribution index is assessed each year, as well as the electricity lost in each contribution electricity assessment and the corresponding assessment amount, the relationship curve between the rated energy value of the energy storage and the annual electricity contribution assessment amount is obtained. Based on the above relationship, the energy storage rated energy variable is added to the optimization objective function to form a levelized unit energy benefit objective function, and the rated energy configuration is optimized considering relevant constraints to obtain the optimal energy capacity configuration.

[0045]

[0046] Where N 60s is the annual average number of electricity assessments, Q 60s E is the single electricity assessment fee, S is the energy configuration of the energy storage system, E0 is a selected energy value;

[0047] j) With the goal of maximizing the levelized unit energy benefit, a direct search method is used to determine the energy storage energy configuration.

[0048] Compared with the prior art, the present invention has the following beneficial effects:

[0049] a) Solve the problem that the existing technical solutions are not fully integrated with the requirements of regional frequency regulation performance indicators, and improve the pertinence of the joint primary frequency regulation planning of fire storage in specific areas.

[0050] b) Solve the problem that the existing technical solutions do not conduct statistical density analysis on long-term operation data, and scientifically plan the configuration of auxiliary primary frequency regulation energy storage capacity.

[0051] c) Solve the problem that the existing technical solutions do not configure the energy storage power capacity and energy capacity separately, and give full play to the role of the power capacity or energy capacity of the energy storage system.

[0052] d) Solve the problem that existing technical solutions do not configure the energy storage system from the perspective of the entire life cycle, consider the subsequent operation and maintenance costs, and improve the economic efficiency of the project throughout its entire life cycle. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a flow chart of the power configuration process of auxiliary primary frequency regulation energy storage;

[0054] Figure 2 It is a flow chart of the energy configuration process of auxiliary primary frequency regulation energy storage;

[0055] Figure 3 is the probability density curve of the difference in contribution power;

[0056] Figure 4 It is the probability density curve of 15s output difference;

[0057] Figure 5 It is the probability density curve of 30s output difference. DETAILED DESCRIPTION

[0058] In order to be able to more clearly understand the above-mentioned objects, features and advantages of the present invention, the present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that, in the absence of conflict, the embodiments of the present application and the features in the embodiments can be combined with each other. In the following description, many specific details are set forth in order to fully understand the present invention, and the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in the art without making creative work are within the scope of protection of the present invention. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as those generally understood by technicians in the technical field of the present invention. The terms used in the specification of the present invention herein are only for the purpose of describing specific embodiments and are not intended to limit the present invention.

[0059] The present invention Figure 1-2 As shown, follow the steps below:

[0060] a) Acquisition and processing of data;

[0061] Step 1: For the energy storage rated power configuration of auxiliary thermal power plant primary frequency regulation, the data of 15s actual output extreme value (MW), 15s theoretical output extreme value (MW), actual output at the beginning of primary frequency regulation (MW) and other data of the unit over a longer time scale (such as 1 year or 2 years) should be obtained, and the 15s output response index (%) should be calculated. At the same time, the data of 30s actual output extreme value (MW), 30s theoretical output extreme value (MW), actual output at the beginning of primary frequency regulation (MW) and other data of the unit over a longer time scale (such as 1 year or 2 years) should be obtained, and the 30s output response index (%) should be calculated.

[0062] Step 2: Screen out the cases where the 15s output response index (%) is lower than 75%, and the 30s output response index (%) is lower than 100%. The above cases are the cases where the 15s output and 30s output are assessed. Calculate the 15s output difference and 30s output difference each time the assessment occurs to form the long-term 15s output difference data and 30s output difference data.

[0063]

[0064] In the formula, is the 15s output difference during the i-th assessment, is the theoretical output extreme value of 15s during the i-th assessment, is the actual output extreme value of 15s during the i-th assessment, is the 30-second output difference during the i-th assessment, is the theoretical output extreme value of 30s during the i-th assessment, is the actual output extreme value of 30 seconds during the i-th assessment, and n is the total number of assessments occurring within a certain period of time.

[0065] Step 3: Perform probability density estimation on the 15s output difference data, calculate the probability density function of the 15s output difference data, and uniformly select about 200-300 probability density data points within the maximum and minimum range of the 15s output difference data. Perform one-dimensional Gaussian distribution fitting on the selected probability density data points to obtain the 15s output difference Gaussian distribution probability density function, probability distribution function, and probability density parameters (mean, variance). Perform the same processing on the 30s output difference data to obtain the 30s output difference Gaussian distribution probability density function, probability distribution function, and probability density parameters (mean, variance).

[0066]

[0067] In the formula, ΔP 15s is the 15s output difference random variable, μ 15s is the Gaussian distribution mean of the 15s output difference, σ 15s is the standard deviation of the Gaussian distribution of the 15s output difference, ΔP 30s is the 30s output difference random variable, μ 30s is the Gaussian distribution mean of the 30s output difference, σ 30s is the standard deviation of the Gaussian distribution of the 15s output difference, and N is the normal distribution.

[0068] Step 4: In fact, the mean of the Gaussian distribution of the obtained 15s output difference and 30s output difference is usually not 0 (generally greater than 0). The main reason is that in the power grid frequency disturbance, most of the disturbance directions are in the direction of frequency reduction, that is, the theoretical output of primary frequency regulation is in the direction of power increase, and the disturbance direction is less likely to be in the direction of frequency increase. Therefore, the obtained 15s output difference and 30s output difference can be converted into a standard normal distribution and used to calculate the probability value within a certain interval.

[0069]

[0070] In the formula, is the standard normal distribution random variable of the 15s output difference, is a standard normally distributed random variable of the 30s output difference.

[0071] Step 5: According to the probability density function of the 15s output difference, calculate the relationship between the rated power value of the energy storage and the probability of compensating the 15s output difference. According to the average number of times the 15s output response index of the unit is assessed each year, the amount of electricity lost in each 15s output response assessment and the corresponding assessment amount, obtain the relationship curve between the rated power value of the energy storage and the annual 15s output response assessment amount. In fact, frequency deviation is divided into large disturbances and small disturbances. The maximum frequency deviation of the power grid does not exceed 0.06Hz, which is a small disturbance, and the maximum frequency deviation of the power grid is greater than 0.06Hz, which is a large disturbance. The coefficients used in calculating the assessment electricity are different, but generally speaking, the maximum frequency deviation of the power grid is usually not greater than 0.06Hz, that is, they are basically small disturbances, so this step is calculated according to the small frequency disturbance. Similarly, based on the probability density function of the 30s output difference, the relationship curve between the rated power value of the energy storage and the annual 30s output response assessment amount is obtained.

[0072] EXA 15s =N 15s *Q 15s *(1-P(P S =P0)) (7)

[0073] EXA 30s =N 30s *Q 30s *(1-P(P S =P0)) (8)

[0074] Where, EXA 15s is the total cost of annual 15s output assessment, N 15s Q is the total number of annual 15s output assessments before allocation of storage. 15s The cost of a single 15s output assessment, EXA 30s is the total cost of annual 30s output assessment, N 30s Q is the total number of annual 30s output assessments before storage allocation, 30s is the cost of a single 30s output assessment, P S is the energy storage system capacity configuration, P0 is a selected power value, P(P S =P0) is P S When P0 is taken, the probability of compensating for the number of 15s output assessment times can be compensated.

[0075] Formula (7) describes the relationship between the total annual 15s output assessment cost and the total 30s output assessment cost and the energy storage configuration capacity, where the probability P(P S=P0) can be calculated by the standard normal distribution shown in equations (5) and (6). In general, the total assessment cost decreases with the increase of energy storage configuration capacity, and the minimum value is 0, that is, the energy storage fully compensates for the 15s output assessment times, but from the properties of Gaussian distribution, it can be seen that in order to meet the larger but less likely frequency deviation compensation demand, the energy storage configuration capacity will be too large, affecting the economy.

[0076] Step 6: Based on the relationship between the energy storage rated power value and the output response assessment amount at the above two time scales, an optimization objective function is constructed. The optimization objective function is constructed with the levelized unit power profit (levelized unit power income-levelized unit power cost). The levelized unit power income is the assessment cost reduced by the annual average unit power energy storage, and the levelized unit power cost is the annual average unit power energy storage configuration cost, as shown in formula (9). The constraints mainly include: primary frequency regulation action frequency threshold value constraint, speed inequality constraint, primary frequency regulation depth constraint, energy storage power constraint, energy storage self-discharge constraint, energy storage SOC constraint and other constraints. Calculate the optimal energy storage rated power value in the 15s output and 30s output scenarios. During a certain frequency modulation process, both 15s output and 30s output may be assessed, or only 15s output or 30s output may be assessed. Therefore, the 15s output and 30s output data should be calculated separately, and the larger one should be taken as the rated calculated power of the energy storage. It is considered that during a certain frequency modulation action, the energy storage device can meet both the 15s output index requirement and the 30s output index requirement. It can be considered that the instantaneous output of the energy storage device meets the requirement, which can meet the 15s output and 30s output requirements.

[0077]

[0078] In the formula, d is the discount rate, m is the operating life of the energy storage equipment, C P is the unit power cost (yuan / kW), C E is the unit energy cost (yuan / kWh), T S is the energy storage discharge time.

[0079] In the case of a large disturbance in the grid frequency (more than 0.06 Hz), since large disturbances are currently a low-probability event, the unit assessment results under frequency disturbances over a longer time scale are statistically analyzed separately, and the relationship between the energy storage rated power value and the output response assessment amount during large disturbances is calculated and superimposed on the above-mentioned optimization objective function to calculate the optimal energy storage rated power value in the 15s output and 30s output scenarios.

[0080] Step 7: For the rated energy configuration of energy storage that assists the primary frequency regulation of thermal power plants, the actual power contribution and theoretical power contribution of the unit over a longer time scale (such as 1 or 2 years) should be obtained and the power contribution index (%) should be calculated.

[0081] Screen out the cases where the power contribution index (%) is lower than 75%, which is the case where the contribution power is assessed. Calculate the contribution power difference each time the assessment occurs to form the long-term contribution power difference data.

[0082] Select a kernel function, perform kernel density estimation on the contribution power difference data, calculate the probability density function of the contribution power difference data, and uniformly select about 200-300 probability density data points within the maximum and minimum range of the contribution power difference data. Perform one-dimensional Gaussian distribution fitting on the selected probability density data points to obtain the Gaussian distribution probability density function, probability distribution function and probability density parameters (mean, variance) of the contribution power difference.

[0083] In fact, the mean of the Gaussian distribution of the contribution power difference is usually not 0 (generally less than 0), mainly because in the power grid frequency disturbance, most of the disturbance direction is the frequency reduction direction, that is, the theoretical output of primary frequency regulation is the power increase direction, and the disturbance direction is the frequency increase direction. Therefore, the contribution power difference can be converted into a standard normal distribution and used to calculate the probability value within a certain interval.

[0084] Here, a normal distribution verification process should be added. The main purpose is to verify whether the normal distribution probability density function matches the original output difference data. If the matching degree is not good, it can be adjusted to Weibull distribution, t distribution and other distribution functions, and the matching degree with the original data can be verified. Select the probability distribution with a better matching degree and perform probability calculation within a certain interval. In addition, unlike the energy storage rated power selection process, the horizontal coordinate of the Gaussian distribution in the rated energy configuration process is the contribution power, that is, the energy value, rather than the power value in the rated power configuration process. At the same time, in the process of responding to a frequency modulation, the energy storage cannot know the direction of frequency change in advance. Therefore, the energy storage device needs to maintain an appropriate SOC level during operation. Therefore, after obtaining the energy storage energy that meets a certain probability level, its rated calculated energy should be twice the energy value.

[0085] Step 8: According to the probability density function of the difference in contribution electricity, calculate the relationship between the energy storage rated energy value and the probability of compensating the difference in contribution electricity. According to the average number of times the unit's annual electricity contribution index is assessed, as well as the electricity lost in each contribution electricity assessment and the corresponding assessment amount, obtain the relationship curve between the energy storage rated energy value and the annual electricity contribution assessment amount. Based on the above relationship, add the energy storage rated energy variable to the optimization objective function to form a levelized unit energy revenue objective function, and consider the relevant constraints to optimize the rated energy configuration to obtain the optimal energy capacity configuration.

[0086]

[0087] Where N 60s is the annual average number of electricity assessments, Q 60s E is the single electricity assessment fee, S is the energy configuration of the energy storage system, and E0 is a selected energy value.

[0088] In the case of a large disturbance in the grid frequency (more than 0.06Hz), since large disturbances are currently low-probability events, the unit assessment results under frequency disturbances on a longer time scale are separately counted, and the relationship between the rated energy value of the energy storage and the contribution electricity response assessment amount during large disturbances is calculated and superimposed on the optimization objective function to calculate the optimal energy storage rated power value and rated energy value in the long-time unit primary frequency regulation scenario.

[0089] Example: Calculation Example

[0090] The annual output data of a gas-fired unit in a thermal power plant is adopted, with an installed capacity of 300MW, a primary frequency regulation action threshold of 0.033Hz, and a speed inequality rate of 5%. The basic data is used to generate the contribution power difference, 15s output difference, and 30s output difference data when the assessment occurs.

[0091] The probability density of the contribution power difference, 15s output difference, and 30s output difference data is estimated respectively, and about 200 probability density data points are taken to perform one-dimensional Gaussian distribution fitting on the probability density data points to form Gaussian distribution probability density curves and parameters, as shown in the figure. Figure 3-Figure 5 And as shown in Table 1.

[0092] Table 1 Probability distribution parameters of each assessment item data

[0093]

[0094] from Figure 1-Figure 3 It can be seen that the extracted probability density data points have a high degree of fit with the probability density curve, and the obtained probability distribution parameters of each assessment item data can represent the statistical characteristics of each assessment item on a longer time scale.

[0095] The average annual unit energy cost of energy storage and the discount rate d, the operating life of energy storage equipment n, and the annual operation and maintenance cost coefficient C O , discharge time T, unit power cost C P , unit energy cost C E Assuming the unit power cost C P and unit energy cost C E The composition and value of are shown in Table 2.

[0096] Table 2 Energy storage system cost composition and value

[0097]

[0098]

[0099] If the discount rate d is 3%, this discount rate is the actual discount rate, which can be 2%-3% to consider the time value of money; the operating life of the energy storage equipment is 10 years, and the battery cell is not replaced during the operation period; the annual operation and maintenance cost coefficient is 1.5%, and the annual operation and maintenance cost coefficient is usually 1%-3%; the discharge time is calculated as 1h, which is used as the conversion coefficient between power cost and energy cost. In the energy storage project, it can be calculated according to the actual discharge time. Substituting the above parameters into formula (10), the annual average unit energy cost ECOS is obtained. i

[0100]

[0101] Therefore, under the above-mentioned assumed cost structure, the annual average unit energy cost of the energy storage system is not less than 201 yuan / kWh. When different discharge durations are taken, the annual average unit energy cost of the energy storage system will be affected.

[0102] Different power and duration configurations correspond to different assessment power and assessment costs, and form different configuration combinations. Taking the maximization of levelized unit power and energy benefits as the optimization goal, the benefits of different combinations are calculated, and the optimal energy storage configuration in this example scenario is that the rated power capacity is 1.25% of the installed power of the gas turbine, and the charging and discharging time corresponding to the rated energy capacity is 30 minutes.

[0103] The above is only a preferred embodiment of the present invention. It should be pointed out that, for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principle of the present invention. These improvements and modifications should also be regarded as the scope of protection of the present invention.

Claims

1. A method for configuring an energy storage system for assisting primary frequency regulation of a thermal power plant, characterized in that: Follow the steps below: a) Acquisition and processing of data; For the energy storage rated power configuration of the auxiliary thermal power plant primary frequency regulation, obtain the 15s actual output extreme MW, 15s theoretical output extreme MW, and the actual output MW at the start of the primary frequency regulation data of the unit for a long period of 1 or 2 years, and calculate the 15s output response index. At the same time, obtain the 30s actual output extreme MW, 30s theoretical output extreme MW, and the actual output MW at the start of the primary frequency regulation data of the unit for 1 or 2 years, and calculate the 30s output response index; b) Extract the 15s output difference, 30s output difference and contribution power difference when the assessment occurs; Screen out the cases where the 15s output response index is lower than 75% and the 30s output response index is lower than 100%, which are the cases where the 15s output and 30s output are assessed; calculate the 15s output difference and 30s output difference each time the assessment occurs, and form the long-term 15s output difference data and 30s output difference data; In the formula, is the 15s output difference during the i-th assessment, is the theoretical output extreme value of 15s during the i-th assessment, is the actual output extreme value of 15s during the i-th assessment, is the 30-second output difference during the i-th assessment, is the theoretical output extreme value of 30s during the i-th assessment, is the actual output extreme value of 30s during the i-th assessment, and n is the total number of assessments occurring within a certain period of time; c) Probability density estimation is performed on each group of data, and about 200-300 probability density data points are selected; that is, probability density estimation is performed on the 15s output difference data, the probability density function of the 15s output difference data is calculated, and about 200-300 probability density data points are uniformly selected within the maximum and minimum value range of the 15s output difference data; d) Perform one-dimensional Gaussian distribution fitting on the selected probability density data points to obtain the Gaussian distribution probability density function, probability distribution function and probability density parameters of the 15s output difference: mean and variance; perform the same processing on the 30s output difference data to obtain the Gaussian distribution probability density function, probability distribution function and probability density parameters of the 30s output difference: mean and variance; e) extracting fitting parameters to form probability density curve and probability distribution curve; In the formula, ΔP 15s is the 15s output difference random variable, μ 15s is the Gaussian distribution mean of the 15s output difference, σ 15s is the standard deviation of the Gaussian distribution of the 15s output difference, ΔP 30s is the 30s output difference random variable, μ 30s is the Gaussian distribution mean of the 30s output difference, σ 30s is the standard deviation of the Gaussian distribution of the 15s output difference, and N is the normal distribution; f) Convert the probability density curve into a standard normal distribution; The obtained 15s output difference and 30s output difference are converted into standard normal distribution and used to calculate the probability value within a certain interval; In the formula, is the standard normal distribution random variable of the 15s output difference, is the standard normal distribution random variable of the 30s output difference; g) Give the corresponding relationship between power configuration and electricity cost assessment; According to the probability density function of the 15s output difference, the relationship between the rated power value of the energy storage and the probability of compensating the 15s output difference is calculated. According to the average number of times the 15s output response index of the unit is assessed each year, the amount of electricity lost in each 15s output response assessment and the corresponding assessment amount, the relationship curve between the rated power value of the energy storage and the annual 15s output response assessment amount is obtained; Based on the probability density function of the 30s output difference, the relationship curve between the energy storage rated power value and the annual 30s output response assessment amount is obtained; EXA 15s =N 15s *Q 15s *(1-P(P S =P0)) (7) EXA 30s =N 30s *Q 30s *(1-P(P S =P0)) (8) Where, EXA 15s is the total cost of annual 15s output assessment, N 15s Q is the total number of annual 15s output assessments before allocation of storage. 15s The cost of a single 15s output assessment, EXA 30s is the total cost of annual 30s output assessment, N 30s Q is the total number of annual 30s output assessments before storage allocation, 30s P is the cost of a single 30s output assessment. S is the power configuration of the energy storage system, P0 is a selected power value, P(P S =P0) is P S When P0 is taken, the probability of compensating the number of 15s output assessment times can be achieved; Based on the relationship between the rated power value of energy storage and the output response assessment amount at the above two time scales, an optimization objective function is constructed, and the optimization objective function is constructed with the levelized unit power profit (i.e., levelized unit power benefit - levelized unit power cost). The levelized unit power benefit is the assessment fee reduced by the annual average unit power energy storage, and the levelized unit power cost is the annual average unit power energy storage configuration cost, as shown in formula (9); In the formula, d is the discount rate, m is the operating life of the energy storage equipment, C P is the unit power cost (yuan / kW), C E is the unit energy cost (yuan / kWh), T S is the energy storage discharge time; h) Taking the maximization of levelized unit power revenue as the goal, a direct search method is used to determine the energy storage power configuration; The constraints include: primary frequency modulation action frequency threshold constraint, speed inequality constraint, primary frequency modulation adjustment depth constraint, energy storage power constraint, energy storage self-discharge constraint, and energy storage SOC constraint; Calculate the optimal energy storage rated power value in the 15s output and 30s output scenarios; calculate the 15s output and 30s output data respectively, and take the larger one as the energy storage rated calculated power, and consider that in a certain frequency modulation process, the energy storage device can meet both the 15s output index requirement and the 30s output index requirement, and consider that the instantaneous output of the energy storage device meets the requirement, which can meet the 15s output and 30s output requirements; i) Taking the contribution of electricity as the research object: Step 7: For the energy storage rated energy configuration for assisting the primary frequency regulation of the thermal power plant, obtain the actual power contribution and theoretical power contribution of the unit for one or two years and calculate the power contribution index; Screen out the cases where the power contribution index is lower than 75%, which is the case where the power contribution is assessed; Calculate the difference in the amount of electricity contributed each time an assessment occurs, and form data on the difference in amount of electricity contributed over a long time scale; Select a kernel function, perform kernel density estimation on the contribution power difference data, calculate the probability density function of the contribution power difference data, and uniformly select about 200-300 probability density data points within the maximum and minimum value range of the contribution power difference data; perform one-dimensional Gaussian distribution fitting on the selected probability density data points to obtain the contribution power difference Gaussian distribution probability density function, probability distribution function and probability density parameters: mean and variance; The purpose of the normal distribution verification process is to verify whether the normal distribution probability density function matches the original output difference data. If the matching degree is not good, it can be adjusted to a distribution function such as Weibull distribution or t distribution, and the matching degree with the original data can be verified. The probability distribution with a better matching degree is selected, and the probability calculation within a certain interval is performed; Step 8: According to the probability density function of the difference in contribution electricity, the relationship between the rated energy value of the energy storage and the probability of compensating the difference in contribution electricity is calculated. According to the average number of times the unit's electricity contribution index is assessed each year, as well as the electricity lost in each contribution electricity assessment and the corresponding assessment amount, the relationship curve between the rated energy value of the energy storage and the annual electricity contribution assessment amount is obtained. Based on the above relationship, the energy storage rated energy variable is added to the optimization objective function to form a levelized unit energy benefit objective function, and the rated energy configuration is optimized considering relevant constraints to obtain the optimal energy capacity configuration. Where N 60s is the annual average number of electricity assessments, Q 60s E is the single electricity assessment fee, S is the energy configuration of the energy storage system, E0 is a selected energy value; j) With the goal of maximizing the levelized unit energy benefit, a direct search method is used to determine the energy storage energy configuration.

Images