A method for optimizing energy storage configuration in electricity price bidding scenario

Abstract

The present invention provides an energy storage configuration optimization method in an electricity price bidding scenario, constructs a mathematical model with the goal of optimizing energy storage revenue, and uses a branch and bound algorithm to solve the model. The present invention can quickly solve the optimal charging and discharging power, optimal total revenue, charging and discharging cost, number of cycles, etc. of the energy storage battery at each moment of each energy storage configuration by knowing the electricity price at each moment in history, and find out which energy storage configuration has the best economy. This method is applied to the actual energy storage strategy control operation system, and the future electricity price forecast curve is input to quickly obtain the optimal energy storage battery charging and discharging strategy curve, thereby realizing effective charging and discharging control of the energy storage battery. Compared with the manual estimation method, it has the advantages of high accuracy and high efficiency.

Description

Technical Field

[0001] The present invention belongs to the field of electric power technology, and in particular relates to an energy storage configuration optimization method in an electricity price bidding scenario. Background Art

[0002] In foreign power markets such as Australia, with strong economic growth and rising user enthusiasm, the number of distributed photovoltaic system installations has increased dramatically, and has correspondingly driven the development of user-side energy storage systems. The emergence of integrated business models such as virtual power plants can fully tap and utilize the wholesale market value of thousands of distributed power generation systems, making the power supply highly volatile, leading to an imbalance in power supply and demand, which is reflected in the power market as large fluctuations in real-time changes in electricity prices. Electricity prices change every 5 minutes, and the price difference before and after can reach hundreds or even thousands or tens of thousands. When the power supply side has sufficient power generation and the power generation is far greater than the load, the power price may be negative, and consuming electricity can also bring benefits. On the contrary, if the power supply side is insufficient and the power consumption demand is strong, the power price will be very high.

[0003] In order to balance the supply and demand of electricity, the government encourages users to configure energy storage systems to charge when the power generation is greater than the load and the electricity price is low, and to discharge when the power generation is less than the load and the electricity price is high. This can effectively manage demand on the user side, balance the electricity supply and demand market, smooth the load, and balance the energy storage costs and achieve profits by buying low and selling high, thereby promoting the utilization and development of new energy.

[0004] However, in actual applications, the benefits of using different energy storage capacities, rated powers and different charging and discharging strategies vary greatly. There are countless combinations of charging and discharging strategies, which cannot be calculated using an exhaustive method. Manual estimation using some simple strategies will result in inaccurate estimates and low efficiency. Summary of the invention

[0005] The present invention aims to solve the above technical problems and provide a method for optimizing energy storage configuration in an electricity price bidding scenario.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] A method for optimizing energy storage configuration in an electricity price bidding scenario, comprising:

[0008] Collect historical electricity price data and future electricity price forecast data for a period of time at a data collection interval of Δt, and select the lowest electricity price at each time i as the charging electricity price, denoted as PriceElecMin i , select the highest electricity price at each time i as the discharge electricity price, recorded as PriceElceMax i, calculate the buy low and sell high profits, demand costs and charge and discharge cycle costs;

[0009] Calculate the total revenue of the electricity price bidding, which is the revenue of buying low and selling high minus the demand cost and the cost of charging and discharging cycle;

[0010] Use the branch and bound algorithm to solve the charging and discharging power at each moment to optimize the total benefit;

[0011] Based on historical electricity price data and different energy storage configurations, the optimal energy storage configuration is obtained to achieve the best input-output ratio. Energy storage batteries are purchased according to this optimal energy storage capacity and rated power and put into operation.

[0012] Preferably, during operation, the electricity price data for the past 24 hours is first obtained and given to the electricity price prediction algorithm to predict the electricity price curve for the next 4 hours or a day. The data is then input into the optimization calculation model to obtain the charge and discharge power at each moment in the next 4 hours. The charge and discharge power instructions are then sent to the energy storage inverter, which then controls the battery according to the predicted charge and discharge power.

[0013] As a preferred method, the method for calculating the profit of buying low and selling high is as follows:

[0014] If charging is done at low prices and discharging is done at high prices, the profit from buying low and selling high is the sum of the discharge profits minus the sum of the charging costs:

[0015]

[0016] Among them, u i is the discharge power at the i-th moment, v i is the charging power at the i-th moment, and Δt is the data collection interval.

[0017] Preferably, the method for calculating demand cost is as follows:

[0018] In addition to the charging cost, the demand cost also needs to be borne every month. In the electricity price bidding scenario, charging is basically based on the rated power, and the demand is priced based on the apparent power. Therefore, the maximum demand is the apparent power corresponding to the rated power:

[0019]

[0020] Among them, vmax is the maximum charging power, PFactor is the conversion factor of apparent power, and PriceDemand is the demand electricity price per kVA.

[0021] Preferably, the method for calculating the charge-discharge cycle cost is as follows:

[0022] Eliminate arbitrage below the cost of charge and discharge cycles:

[0023] The number of cycles in the entire life cycle of the battery is limited. Each charge and discharge has a certain cost. In order to avoid low returns from one charge and discharge and waste of cycles, the cost of each charge and discharge needs to be taken into account. The total cost of charge and discharge is:

[0024]

[0025] Among them, u i is the discharge power at the i-th moment, v i is the charging power at the i-th moment, and PriceCycle is the charging and discharging cost per kilowatt-hour.

[0026] Preferably, the energy storage configuration optimization method in the electricity price bidding scenario further includes constructing constraint conditions of the evaluation model, wherein the constraint conditions include: charging and discharging cannot be performed at the same time:

[0027] In order to limit the simultaneous charging and discharging at the same time, two control variables need to be added, m i Indicates whether the battery can be discharged at the i-th moment, m i ∈0,1,n i Indicates whether the battery can be charged at the i-th moment, n i ∈0,1, then the following conditions must be met if charging and discharging cannot be performed at the same time:

[0028] m i +n i <=1.

[0029] Preferably, the energy storage configuration optimization method in the electricity price bidding scenario further includes constructing constraint conditions of the evaluation model, and the constraint conditions include:

[0030] The battery's charge and discharge power at each moment is within the limit:

[0031] The battery charge and discharge power at each moment is limited between 0 and the rated power, and if charging is performed at this moment, the discharge power should be 0, otherwise, if discharging is performed at this moment, the charging power should be 0, and PRate is the rated power:

[0032] 0<=u i <=PRate*m i

[0033] 0<=v i <=PRate*n i .

[0034] Preferably, the energy storage configuration optimization method in the electricity price bidding scenario further includes constructing constraint conditions of the evaluation model, and the constraint conditions include:

[0035] The battery capacity at each moment is within the limit:

[0036] Q0 is the initial capacity, and the energy storage capacity at each moment is between the rated capacity and the minimum remaining capacity:

[0037]

[0038] Among them, QRate is the rated capacity, SOCRemain is the minimum SOC, and SOC is the minimum remaining battery power.

[0039] As a preferred method, split a month's data by day, first find the optimal daily charging and discharging strategy and optimal profit, then concatenate the daily charging and discharging strategies to get the optimal monthly charging and discharging strategy, and add up the daily optimal profits to get the optimal monthly profit.

[0040] Preferably, the iteration termination condition of the optimization model solution is set to terminate the iteration when the error between two iterations is less than 5%.

[0041] After adopting the above technical solution, the present invention has the following advantages:

[0042] The present invention constructs a mathematical model with the goal of optimizing energy storage benefits, and uses a branch and bound algorithm to solve the model.

[0043] By knowing the electricity price at each moment in history and a variety of energy storage configurations, such as capacity and rated power, the present invention can quickly solve the optimal charging and discharging power, optimal total revenue, charging and discharging cost, number of cycles, etc. of the energy storage battery for each energy storage configuration at each moment, and find out which energy storage configuration has the best economy.

[0044] By applying this method to the actual energy storage strategy control operation system and inputting the future electricity price forecast curve, the optimal energy storage battery charge and discharge strategy curve can be quickly obtained, thereby realizing effective charge and discharge control of the energy storage battery. Compared with the manual estimation method, it has the advantages of high accuracy and high efficiency.

[0045] At the same time, the present invention optimizes the model solving speed of the computing device, and improves the calculation speed by solving the model on a daily basis and iterative parameter optimization control, reducing the calculation time from several hours to tens of seconds.

[0046] In addition, the present invention combines the real-time electricity price prediction for the next 4 hours and applies it to the actual charging and discharging strategy of the energy storage battery, which can effectively manage the demand-side electricity consumption. When the power supply is greater than the load demand, the electricity price drops, and charging is performed to increase the load demand and adjust the electricity price to rise; when the power supply is less than the load demand, the electricity price rises, and discharging is performed to increase the power supply and adjust the electricity price to fall. Therefore, the load and power supply can be smoothed, the power supply and demand relationship can be balanced, and the volatility of the electricity price can be smoothed, thereby improving the investment income of energy storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0047] Figure 1 A flow chart of an energy storage configuration optimization method in an electricity price bidding scenario;

[0048] Figure 2 The real-time electricity price for a certain period of time of a project using the energy storage configuration optimization method in the electricity price bidding scenario;

[0049] Figure 3 The energy storage configuration of a project using the energy storage configuration optimization method in the electricity price bidding scenario has the charging and discharging power and specific benefits at each moment. DETAILED DESCRIPTION

[0050] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments.

[0051] The energy storage system can choose different services for charging at the same time, and different services have different corresponding prices. There are 4 modes of battery charging electricity prices to choose from, namely NEM (National Energy Market National Energy Trading Market Electricity Price), including one price for charging and discharging, FCAS (Frequency Control Ancillary Service Frequency Control Ancillary Service Market Electricity Price), the frequency reduction response time is 6 seconds, 60 seconds, and 5 minutes, corresponding to three charging prices, and the frequency increase response time is 6 seconds, 60 seconds, and 5 minutes, corresponding to three discharge prices.

[0052] In order to optimize the energy storage configuration, the present invention provides an energy storage configuration optimization method in an electricity price bidding scenario, comprising:

[0053] Step 1: Collect historical electricity price data and future electricity price forecast data for a period of time at a data collection interval of Δt. Pre-process the electricity price at each moment. To obtain the highest profit, the lowest electricity price at that moment should be selected when charging, and the highest electricity price at that moment should be selected when discharging. Therefore, the electricity prices at all moments are processed as follows:

[0054] Select the lowest electricity price at each time i as the charging electricity price, denoted as PriceElecMin i ;

[0055] Select the highest electricity price at each time i as the discharge price, denoted as PriceElceMax i ;

[0056] Step 2: Construct a bidding scenario energy storage revenue evaluation model:

[0057] Step 2.1, calculate the profit of buying low and selling high:

[0058] If charging is done at low prices and discharging is done at high prices, the profit from buying low and selling high is the sum of the discharge profits minus the sum of the charging costs:

[0059]

[0060] Among them, u i is the discharge power at the i-th moment, v i is the charging power at the i-th moment, Δt is the data collection interval;

[0061] Step 2.2, calculate demand cost:

[0062] In addition to the charging cost, you also need to bear the demand cost every month. In the electricity price bidding scenario, charging is basically based on the rated power. In addition, in the Australian electricity market, demand is priced based on apparent power, so the maximum demand is the apparent power corresponding to the rated power:

[0063]

[0064] Among them, vmax is the maximum charging power, PFactor is the conversion factor of apparent power, and PriceDemand is the demand price per kVA;

[0065] Step 2.3, eliminate arbitrage below the cost of charge and discharge cycle:

[0066] The number of cycles in the entire life cycle of the battery is limited. Each charge and discharge has a certain cost. In order to avoid a low return on one charge and discharge and waste of cycles, the cost of each charge and discharge needs to be taken into account. The total charge and discharge cycle cost is:

[0067]

[0068] Among them, PriceCycle is the charging and discharging cost per kilowatt-hour;

[0069] Step 2.4, calculate the total revenue of electricity price bidding:

[0070] The total revenue of electricity price bidding is the revenue of buying low and selling high minus the demand cost, minus the circulation cost, and RTotal is the total revenue:

[0071]

[0072] Step 3: Construct the constraints of the evaluation model:

[0073] Step 3.1, charging and discharging cannot be done at the same time:

[0074] In order to limit the simultaneous charging and discharging at the same time, two control variables need to be added, m iIndicates whether the battery can be discharged at the i-th moment, m i ∈0,1,n i Indicates whether the battery can be charged at the i-th moment, n i ∈0,1, then the following conditions must be met if charging and discharging cannot be performed at the same time:

[0075] m i +n i <=1;

[0076] Step 3.2: The battery's charge and discharge power at each moment is within the limit:

[0077] The battery charge and discharge power at each moment is limited between 0 and the rated power, and if charging is performed at this moment, the discharge power should be 0, otherwise, if discharging is performed at this moment, the charging power should be 0, and PRate is the rated power:

[0078] 0<=u i <=PRate*m i

[0079] 0<=v i <=PRate*n i ;

[0080] Step 3.3, the battery capacity at each moment is within the limit:

[0081] Q0 is the initial capacity, and the energy storage capacity at each moment is between the rated capacity and the minimum remaining capacity:

[0082]

[0083] Among them, QRate is the rated capacity, SOCRemain is the minimum SOC, and SOC is the minimum remaining battery capacity;

[0084] Step 4, model solution:

[0085] This is a mixed integer linear programming optimization problem. It is necessary to solve the charging and discharging power at each moment to optimize the total benefit. The branch and bound algorithm is used to solve it. During the solution process, it is found that the solution speed is very slow when inputting one year of data or even a few days of data. Therefore, the following two optimizations are made:

[0086] (a) Solving by day

[0087] Since the optimal solution for each month is approximately equal to the set of the optimal solutions for each day, the data for one month can be split by day, and the optimal charging and discharging strategy and optimal benefit for each day can be calculated first. Then, the optimal charging and discharging strategy for each day can be concatenated to obtain the optimal charging and discharging strategy for each month. The optimal benefit for each day can be accumulated to obtain the optimal benefit for each month.

[0088] After adopting the data splitting strategy, the operation situation changed from system crash to being able to calculate the results in a few hours;

[0089] (b) Iterative Control

[0090] To further reduce the running time, the iterative control parameters were optimized. First, the iterative termination condition was optimized. The termination condition was set to terminate the iteration when the error between two iterations was less than 5%. The running speed was rapidly improved. The optimal solution for some days' data was completed within a few minutes, but the data for some days required dozens of minutes to calculate.

[0091] It was found that the target optimal return was relatively large for the data of the days with a long calculation time, and the termination condition did not work when the error was less than 5%. Therefore, the target optimal return was reduced, and the overall optimal solution speed could be completed within a few minutes;

[0092] Step 5, run control:

[0093] Based on historical electricity price data and different energy storage configurations, the optimal energy storage configuration can be obtained with the best input-output ratio. Energy storage batteries can be purchased and put into operation according to this optimal energy storage capacity and rated power.

[0094] During operation, the electricity price data of the past 24 hours is first obtained and fed into the electricity price prediction algorithm to predict the electricity price curve for the next 4 hours or a day. The data is then input into the optimization calculation device to obtain the charge and discharge power at each moment in the next 4 hours. The charge and discharge power instructions are then sent to the PCS (Power Conversion System), which then controls the battery according to the predicted charge and discharge power.

[0095] The flowchart of the above method is as follows Figure 1 As shown;

[0096] The following is an example of a project using the energy storage configuration optimization method in the above electricity price bidding scenario:

[0097] The real-time electricity price of the project for a period of time is as follows: Figure 2 As shown, there are 7 electricity price data at the same time every 5 minutes, including 4 charging prices and 4 discharging prices.

[0098] The input parameters to the optimization device are as follows:

[0099] The battery rated power is between 1000kw and 1500kw, with a power interval step of 100kw;

[0100] Battery full charge hours: between 1h and 2h, with a step length of 1h;

[0101] Demand electricity price: 0.076 yuan / kva / month;

[0102] Active power converted into apparent power factor: 0.85;

[0103] Energy storage initial SOC: 0.1;

[0104] Discharge remaining SOC: 0;

[0105] The output is as follows:

[0106] The optimal energy storage configuration is found to be 1500kw rated power and 2 hours full charge time; Figure 3 This is the charging and discharging power and specific revenue situation of this energy storage configuration at each moment; the revenue data for each month is shown in the following table;

[0107]

[0108]

[0109] In addition to the above preferred embodiments, the present invention has other implementation modes. Those skilled in the art may make various changes and modifications based on the present invention. As long as they do not depart from the spirit of the present invention, they should all fall within the scope defined by the claims attached to the present invention.

Claims

1. A method for optimizing energy storage configuration in an electricity price bidding scenario, characterized in that: include: Collect historical electricity price data and future electricity price forecast data for a period of time with a data collection interval of Δt, select the lowest electricity price at each moment i as the charging electricity price, recorded as PriceElecMini, select the highest electricity price at each moment i as the discharging electricity price, recorded as PriceElceMaxi, and calculate the low-buy-high-sell profit, demand cost, and charge-discharge cycle cost; Calculate the total revenue of the electricity price bidding, which is the revenue of buying low and selling high minus the demand cost and the cost of charging and discharging cycle; Use the branch and bound algorithm to solve the charging and discharging power at each moment to optimize the total benefit; Based on historical electricity price data and different energy storage configurations, the optimal energy storage configuration is obtained to optimize the input-output ratio. Energy storage batteries are purchased according to this optimal energy storage capacity and rated power and put into operation; The method for calculating the profit of buying low and selling high is as follows: If charging is done at low prices and discharging is done at high prices, the profit from buying low and selling high is the sum of the discharge profits minus the sum of the charging costs: Among them, u i is the discharge power at the i-th moment, v i is the charging power at the i-th moment, Δt is the data collection interval; The method for calculating demand cost is as follows: In addition to the charging cost, you also need to bear the demand cost every month. In the electricity price bidding scenario, charging is based on the rated power, and the demand is priced based on the apparent power. Therefore, the maximum demand is the apparent power corresponding to the rated power: Among them, vmax is the maximum charging power, PFactor is the conversion factor of apparent power, and PriceDemand is the demand price per kVA; The method for calculating the charge and discharge cycle cost is as follows: Eliminate arbitrage below the cost of charge and discharge cycles: The number of cycles in the entire life cycle of the battery is limited. Each charge and discharge has a certain cost. In order to avoid low returns from one charge and discharge and waste of cycles, the cost of each charge and discharge needs to be taken into account. The total cost of charge and discharge is: Among them, u i is the discharge power at the i-th moment, v i is the charging power at the i-th moment, PriceCycle is the charging and discharging cost per kilowatt-hour; Split a month's data by day, first find the optimal charging and discharging strategy and optimal income for each day, then combine the daily charging and discharging strategies to get the optimal charging and discharging strategy for each month, and add up the daily optimal income to get the optimal income for each month. During operation, the electricity price data for the past 24 hours is first obtained and given to the electricity price prediction algorithm to predict the electricity price curve for the next 4 hours or a day. The data is then input into the optimization calculation model to obtain the charging and discharging power at each moment in the next 4 hours. The charging and discharging power instructions are then sent to the energy storage inverter, which then controls the battery according to the predicted charging and discharging power.

2. The energy storage configuration optimization method in the electricity price bidding scenario according to claim 1, characterized in that: The energy storage configuration optimization method in the electricity price bidding scenario also includes constructing constraint conditions of the evaluation model, and the constraint conditions include: charging and discharging cannot be performed at the same time: In order to limit the simultaneous charging and discharging at the same time, two control variables need to be added, m i Indicates whether the battery can be discharged at the i-th moment, m i ∈0,1,n i Indicates whether the battery can be charged at the i-th moment, n i ∈0,1, then the following conditions must be met if charging and discharging cannot be performed at the same time: mi+ni<=1.

3. The energy storage configuration optimization method in the electricity price bidding scenario according to claim 2, characterized in that: The energy storage configuration optimization method in the electricity price bidding scenario also includes constructing constraint conditions of the evaluation model, and the constraint conditions include: The battery's charge and discharge power at each moment is within the limit: The battery charge and discharge power at each moment is limited between 0 and the rated power, and if charging is performed at this moment, the discharge power should be 0, otherwise, if discharging is performed at this moment, the charging power should be 0, and PRate is the rated power: 0<=u i <=PRate*m i 0<=v i <=PRate*n i 。 4. The energy storage configuration optimization method in the electricity price bidding scenario according to claim 1, characterized in that: The energy storage configuration optimization method in the electricity price bidding scenario also includes constructing constraint conditions of the evaluation model, and the constraint conditions include: The battery capacity at each moment is within the limit: Q0 is the initial capacity, and the energy storage capacity at each moment is between the rated capacity and the minimum remaining capacity: Among them, QRate is the rated capacity, SOCRemain is the minimum SOC, and SOC is the minimum remaining battery power.

5. The energy storage configuration optimization method in the electricity price bidding scenario according to claim 1, characterized in that: The iterative termination condition for solving the optimization model is set to terminate the iteration when the error between two iterations is less than 5%.

Images