A joint bidding method for multi-type power market applicable to pumped storage power stations

Abstract

The present invention discloses a multi-type electricity market joint bidding method applicable to pumped storage power stations. First, electricity price forecasting is performed to establish a CVaR risk model; the risk model is introduced into the optimization decision-making target, the objective function and constraints are set, and a risk-return model for the pumped storage power station to participate in the electricity market and the auxiliary service market is established; the constraints such as power generation / pumping rate constraints and reservoir capacity constraints are considered, and the model is solved with the goal of maximizing the benefits. There is a solution, that is, the pumping capacity, power generation capacity, frequency regulation capacity and the benefits obtained by participating in the market of the pumped storage power station in each period within an optimization cycle are obtained, and the bidding strategy with the best economic efficiency is obtained. The bidding method of the present invention introduces confidence and risk preference coefficients to consider the fluctuation and uncertainty risks of market electricity prices. By optimizing capacity allocation, the production is arranged scientifically to ensure the best economic efficiency within the feasible range, and the economic benefits of the pumped storage power station are further improved.

Description

Technical Field

[0001] The present invention relates to the field of electrical engineering technology, and in particular to a multi-type electricity market joint bidding method applicable to a pumped storage power station. Background Art

[0002] Normally, when a generator sets formulates a power sales plan for the power market, it only needs to determine the optimal power generation capacity at each moment based on the power market price and the fuel / natural resource situation. However, pumped storage is different from such conventional generator sets. It cannot generate electricity independently, but needs to generate electricity by "storing water". Therefore, pumped storage must have two periods of "pumping" and "generating electricity" during the operation cycle. This determines that the operation plan of pumped storage is more complicated, but it must be determined in advance. Under the traditional planning and dispatching mode, the power dispatching agency is responsible for formulating the operation plan of the pumped storage power station. Based on the principle of minimizing the system dispatching cost, the pumping period and the water release period of the pumped storage are arranged. However, despite the obvious peak-valley price difference, the probability of a large price difference is not high, and the peak-valley electricity price difference arbitrage cannot become the main source of income for pumped storage power stations. Due to the lack of market mechanism, the income of ancillary services is also difficult to guarantee.

[0003] With the gradual construction of the electricity market, pumped storage can participate in market competition as an independent entity, hoping to maximize its own economic profits while providing regulation services for the system, and formulate capacity allocation plans for electricity and frequency regulation auxiliary services based on the market prices of electricity and frequency regulation. However, at present, the market awareness of pumped storage power stations is weak, and the operation strategy is not clear. The existing operation strategy model generally conducts post-analysis based on the existing electricity price, ignoring the volatility of electricity prices in the actual market, and is insufficient for pumped storage power stations to formulate bidding strategies and report market information in advance. Summary of the invention

[0004] In view of the problems existing in the prior art, the present invention proposes a joint optimization bidding method for pumped storage power stations to participate in the electric energy market and the ancillary service market. The bidding method introduces confidence and risk preference coefficients to consider the volatility and uncertainty risks of market electricity prices, and scientifically arranges production within a feasible range to ensure economic efficiency by optimizing capacity allocation, thereby further improving the economic benefits of pumped storage power stations.

[0005] To achieve the above object, the present invention adopts the following technical solution: a multi-type electricity market joint bidding method applicable to pumped storage power stations, comprising the following steps:

[0006] (1) forecasting the on-grid electricity price based on historical data such as load, surplus power generation capacity, and electricity price information during each trading period; the on-grid electricity price is described by a probability density function;

[0007] (2) Using the Monte Carlo sampling method, M electricity price forecast scenarios are generated based on the estimated electricity price probability density function for each period of the day, and the K-means clustering scenario reduction method is used to reduce them to N typical scenarios; based on the N typical scenarios, a CVaR risk model is established to measure the risk caused by the uncertainty of market electricity prices and complete the electricity price forecast;

[0008] (3) Construct an objective function based on electricity energy benefits, frequency regulation benefits, and CVaR risk;

[0009] (4) Setting generation / pumping rate constraints, reservoir capacity constraints, quantitative relationship constraints between generation / pumping / frequency regulation power and reservoir capacity, and CVaR-related constraints;

[0010] (5) Based on the objective function set in step (3) and the constraints set in step (4), a risk-return model for the pumped storage power station to participate in the electricity market and ancillary service market is established;

[0011] (6) Using the CPLEX solver, the risk-return model established in step (5) is solved to obtain the production simulation results. If there is a solution, the pumping capacity, power generation capacity, frequency regulation capacity and market benefits of the pumped-storage power station in each period of an optimization cycle are obtained, and the most economically efficient bidding strategy is obtained.

[0012] Furthermore, the on-grid electricity price in step (1) is preferably described by a normal distribution function, and the formula is as follows:

[0013]

[0014] In the formula, i is the time period identifier, P i e is the market clearing price in the ith period, i.e., the on-grid electricity price; μ i The expected value of the on-grid electricity price for power generation; is the variance.

[0015] Furthermore, the objective function in step (3) includes two parts, as shown below:

[0016]

[0017] Where, the first part f is the net income of the pumped storage power station in the electricity market and frequency regulation market, π w is the probability of scenario ω describing the uncertainty of electricity price, N ωis the number of typical electricity price scenarios; the second part is the product of the conditional value at risk CVaR and the risk preference coefficient β. The risk preference coefficient β indicates the risk preference of the pumped storage power station owner.

[0018] Furthermore, the net income f of the pumped storage power station participating in the electricity market and ancillary service market is as follows:

[0019]

[0020] Where i represents the unit time period, P i e represents the power generated in time period i, D i represents the pumping power in period i, P i re Indicates the frequency regulation capacity under power generation conditions, P i rd represents the frequency regulation capacity under pumping conditions, represents the forecast value of the on-grid electricity price in time period i, represents the frequency regulation capacity price, represents the predicted electricity price based on frequency regulation mileage, and R represents the ratio of frequency regulation mileage to frequency regulation capacity of pumped storage.

[0021] Furthermore, in the risk value model, VaR refers to the income obtained by the pumped storage power station in market changes under a certain confidence level; CVaR represents the expected value of the profit distribution tail 1-α of the pumped storage power station holder under a certain time and confidence level α; the risk value CVaR related constraints are as follows:

[0022]

[0023]

[0024] η ω ≥0

[0025] In the formula, ξ represents the value at risk (VaR), η ω is the difference between the return and risk value under scenario ω.

[0026] Furthermore, the quantitative relationship constraint between the power generation / pumping / frequency regulation power and the reservoir capacity is as follows:

[0027] V i+1 =V i -P i e ·α / η1-P i re 0.5α / η2+D i ·α·η2+P i rd·0.5α·η2

[0028] Where α is the coefficient of pumped storage power generation or conversion of pumped water into storage capacity, η1 is the loss rate of pumped storage during the pumping process, and η2 is the loss rate of pumped storage during the power generation process.

[0029] Furthermore, the power generation / pumping rate constraint expression is as follows:

[0030] P i e ·D i =0,P i e +D i ≥0

[0031] P min ≤P i e +P i re ≤P max

[0032] 0≤D i +P i rd ≤D max

[0033] P i re ·P i rd =0,P i re +P i rd ≥0

[0034] D i ·P i re =0

[0035] P i e ·P i rd =0

[0036] P i e ,D i ,P i rd ,P i re ≥0

[0037] P i re ≤εP max

[0038] P i rd ≤εDmax

[0039] Where P min is the minimum power output of the pumped storage unit, P max is the maximum power output of the pumped storage unit, D max is the maximum pumping capacity of the pumped storage unit, and ε is the upper limit of the capacity of the pumped storage unit participating in the frequency regulation market.

[0040] Furthermore, the storage capacity constraint expression is as follows:

[0041] V1=V 24

[0042] V min ≤V i ≤V max

[0043] Where V i represents the storage capacity in time period i, V min Represents the minimum storage capacity, V max Indicates the maximum storage capacity.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] 1. The bidding method takes into account the risk of uncertainty in electricity prices and predicts electricity prices in the market, providing technical support and auxiliary decision-making reference solutions for pumped storage power stations to track market trends and formulate operation strategies in advance.

[0046] 2. The bidding method establishes a risk-return model for pumped-storage power stations to participate in the electricity market and ancillary service market, fully utilizes the resource flexibility of pumped storage, realizes the optimal capacity allocation of the electricity market and ancillary service market, and maximizes benefits while buffering risks.

[0047] The bidding method of the present invention takes into account the open competition environment and electricity price uncertainty of the market after the electricity market construction gradually matures, and proposes an optimal bidding strategy suitable for pumped storage power stations. The established decision model and method can provide methodological support for the bidding decision of flexibility resources in the electricity market. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 It is a framework diagram of the joint optimization method of the present invention. DETAILED DESCRIPTION

[0049] The present invention is further described in detail below with reference to the accompanying drawings and specific implementation examples.

[0050] The multi-type electricity market joint bidding method applicable to pumped storage power stations proposed in the present invention first considers the impact of uncertainty in electricity prices on the bidding strategy of pumped storage power stations, and establishes a CVaR risk model; introduces the risk model into the optimization decision-making target, and establishes a risk-return model for pumped storage power stations to participate in the electricity market and the auxiliary service market, wherein the electricity market revenue comes from the peak-valley price difference revenue, and the auxiliary service market revenue only considers the capacity revenue and mileage revenue of participating in the frequency regulation auxiliary service; considers the constraints of power generation / pumping rate, storage capacity, and the quantitative relationship between power generation / pumping power and storage capacity, and solves the model with the goal of maximizing revenue. There is a solution, that is, the pumping capacity, power generation capacity, frequency regulation capacity and the revenue obtained from participating in the market of the pumped storage power station in each period within an optimization cycle are obtained, and the bidding strategy with the best economic efficiency is obtained. The bidding method of the present invention introduces confidence and risk preference coefficients to consider the fluctuation and uncertainty risks of market electricity prices, and scientifically ensures the economically optimal production arrangement within a feasible range by optimizing capacity allocation, further improving the economic benefits of pumped storage power stations.

[0051] like Figure 1 The framework diagram of the joint optimization bidding method for pumped storage power stations to participate in the electric energy market and the ancillary service market of the present invention is shown in FIG. In combination with actual cases, the bidding method specifically includes the following steps:

[0052] (1) The on-grid electricity price is predicted based on historical data such as load, surplus capacity, and electricity price information in previous trading periods. The on-grid electricity price can be described by a probability density function. Then the optimal operation mode of the pumped storage power station in the market environment is transformed into a stochastic optimization problem. Assuming that the on-grid electricity price is described by a normal distribution function, the expression is as follows:

[0053]

[0054] In the formula, i is the time period identifier, P i e is the market clearing price in the ith period, i.e., the on-grid electricity price; μ i The expected value of the on-grid electricity price for power generation; Assume that the power grid price P of different periods of time i e (yuan / kWh) is subject to the mean value μ i ,variance At the same time, assuming that the bid of the pumped storage unit will not affect the hourly power generation grid-connected electricity price, the expected value of the power generation grid-connected electricity price at each moment is shown in Table 1.

[0055] Table 1: Expected values ​​of on-grid electricity prices

[0056]

[0057] (2) Monte Carlo sampling method is used to generate M (M is a custom constant, 1000 in the embodiment of the present invention) electricity price forecast scenarios according to the estimated electricity price probability density function in each period of the day, and the K-means clustering scenario reduction method is used to reduce them to N (N is a custom constant, 10 in the embodiment of the present invention, M>>N) typical scenarios. Based on 10 typical scenarios, a CVaR risk model is established to measure the risks caused by the uncertainty of market electricity prices, and the electricity price forecast is completed. Among them, CVaR represents the expected value of the pumped storage power station holder for the tail 1-α of the profit distribution under a certain time and confidence level α.

[0058] The probability of each scenario in the embodiment of the present invention is shown in Table 2. The confidence interval level of CVaR is α=0.95, β=1.

[0059] Table 2: Probabilities of ten typical scenarios

[0060]

[0061]

[0062] (3) Construct an objective function based on the electricity energy income, frequency regulation income and CVaR risk; the objective function formula is as follows:

[0063]

[0064] The objective function consists of two parts. The first part f is the net income of the pumped storage power station in the electricity market and frequency regulation market, π w is the probability of scenario ω describing the uncertainty of electricity price, N ω is the number of typical electricity price scenarios; the second part is the product of the conditional value at risk CVaR and the risk preference coefficient β. The risk preference coefficient β indicates the risk preference of the pumped storage power station owner. When β>0, the power station is risk-averse, corresponding to the attitude of prioritizing the stability of income. At this time, the power station adopts a more conservative strategy; when β=0, the power station is risk-neutral. At this time, the power station adopts a more aggressive strategy, with the goal of maximizing expected income. Considering that considering CVaR in the objective function itself is a risk-averse bidding strategy, the situation of β<0 is not considered.

[0065] Among them, the net income f of the pumped storage power station participating in the electricity market and ancillary service market is calculated as follows:

[0066]

[0067] In the formula, i represents the unit time period, P i erepresents the power generated in time period i, D i represents the pumping power in period i, P i re Indicates the frequency regulation capacity under power generation conditions, P i rd represents the frequency regulation capacity under pumping conditions, represents the forecast value of the on-grid electricity price in time period i, represents the frequency regulation capacity electricity price (calculated according to 12 / MW in the embodiment of the present invention), represents the predicted price of frequency regulation mileage, and R represents the ratio of frequency regulation mileage to frequency regulation capacity of pumped storage (calculated as 2.94 in the embodiment of the present invention). The income of the pumped storage unit is the income from power generation sales, frequency regulation capacity and frequency regulation mileage, and the variable cost is the cost of purchasing electricity.

[0068] (4) Setting generation / pumping rate constraints, reservoir capacity constraints, quantitative relationship constraints between generation / pumping / frequency regulation power and reservoir capacity, and CVaR-related constraints;

[0069] In the risk value model, VaR refers to the profit obtained by the pumped storage power station in the market changes under a certain confidence level. Then CVaR represents the expected value of the profit distribution tail 1-α of the pumped storage power station holder under a certain time and confidence level α.

[0070] The constraints related to the CVaR are as follows:

[0071]

[0072]

[0073] η ω ≥0

[0074] Where ξ represents the value at risk (VaR), η ω is the difference between the return and risk value under scenario ω.

[0075] The quantitative relationship constraint between power generation / pumping / frequency regulation and reservoir capacity Considering that pumped storage provides frequency regulation, including upward and downward regulation, a certain amount of electricity will be consumed during the frequency regulation process, and the amount of electricity lost in frequency regulation cannot be accurately calculated. By estimation, during the power generation period, the amount of electricity consumed by frequency regulation is calculated as half of the frequency regulation capacity. During the pumping period, the frequency regulation of pumped storage does not consume electricity, but in the form of occupying part of the pumping capacity, so that the pumped storage cannot be pumped at full load, and the amount of electricity pumped in the vacant area is also calculated as half of the frequency regulation capacity. The quantitative relationship constraint between the capacity of pumped storage and power generation / pumping / frequency regulation is as follows:

[0076] V i+1 =V i -Pi e ·α / η1-P i re 0.5α / η2+D i ·α·η2+P i rd ·0.5α·η2

[0077] Where θ is the coefficient of pumped storage power generation or conversion of pumped water into storage capacity, η1 is the loss rate of pumped storage during the pumping process, and η2 is the loss rate of pumped storage during the power generation process.

[0078] The power generation / pumping rate constraint means that pumped storage can only generate power or pump water at any time, but frequency regulation can be provided at any time; During the power generation period, the sum of the power generation capacity and the frequency regulation capacity shall not exceed the maximum power generation capacity; During the pumping period, the sum of the pumping capacity and the frequency regulation capacity shall not exceed the maximum pumping capacity; the power generation power, pumping power, and frequency regulation capacity must meet the non-negative constraint; The frequency regulation service capacity provided by pumped storage cannot exceed a certain capacity. The power generation / pumping rate constraint expression is as follows:

[0079] P i e ·D i =0,P i e +D i ≥0

[0080] P min ≤P i e +P i re ≤P max

[0081] 0≤D i +P i rd ≤D max

[0082] P i re ·P i rd =0,P i re +P i rd ≥0

[0083] D i ·P i re =0

[0084] P i e ·P i rd =0

[0085] P i e ,D i ,P i rd ,P i re ≥0

[0086] P i re ≤εP max

[0087] P i rd ≤εD max

[0088] Where P min is the minimum power output of the pumped storage unit, P max is the maximum power output of the pumped storage unit, D max is the maximum pumping capacity of the pumped storage unit, and ε is the upper limit of the capacity of the pumped storage unit participating in the frequency regulation market.

[0089] The storage capacity constraint means that at the beginning and end of the optimization cycle, the storage capacity of the pumped storage reservoir must be equal. During the pumping / discharging process of pumped storage, the reservoir capacity satisfies the capacity constraint. The storage capacity constraint expression is as follows:

[0090] V1=V 24

[0091] V min ≤V i ≤V max

[0092] Where V i represents the storage capacity in time period i, V min Represents the minimum storage capacity, V max Indicates the maximum storage capacity.

[0093] (5) Based on the objective function set in step (3) and the constraints set in step (4), a risk-return model for the pumped-storage power station to participate in the electricity market and ancillary service market is established.

[0094] The ancillary services mentioned above mainly refer to frequency regulation services, and the capacity allocation of pumped storage in the electric energy and frequency regulation markets is studied. The electric energy market price and the frequency regulation market price are known exogenous variables, and pumped storage is a price receiver, and its bidding strategy will not affect the market clearing price.

[0095] (6) Using the CPLEX solver, the risk-return model established in step (5) is solved to obtain the production simulation results. If there is a solution, the pumping capacity, power generation capacity, frequency regulation capacity and market benefits of the pumped-storage power station in each period of an optimization cycle are obtained, and the most economically efficient bidding strategy is obtained.

[0096] Xianyou Pumped Storage Power Station is taken as an example for calculation analysis. Xianyou Pumped Storage Power Station is the first large-scale pumped storage power station in Fujian Province. It is equipped with 4 mixed flow reversible pump turbine generator sets with a single unit capacity of 300MW, a total installed capacity of 1200MW, and an initial storage capacity of 5 million m 3 .

[0097] The final optimization result is that the total pumping time of pumped storage is 12 hours, including periods 1 to 7, 10, 11, 16, 21, and 22, and the total pumping volume in one day is 4811.52MWh. The total power generation time is 8 hours, including periods 8, 9, 12, 17 to 20, and 23, and the total power generation in one day is 5956.41MWh. The pumped storage unit provides frequency regulation services throughout the day, of which the total capacity of frequency regulation under power generation conditions is 3370.31MW, and the total capacity of frequency regulation under pumping conditions is 9456.42MW. During the entire operation process, the pumped storage reached the maximum storage capacity at period 8 and the minimum storage capacity at period 23h. The storage capacity at the end of the optimization cycle was 500MWh, which was equal to the initial storage capacity.

[0098] The economic benefits of pumped storage participating in the market can be calculated from the distribution results of electric energy and frequency regulation capacity. After calculation, the daily income of electric energy market is 708,700 yuan, the daily income of frequency regulation capacity is 153,910 yuan, the daily income of frequency regulation mileage is 900,800 yuan, the daily electricity purchase cost is 1,019,700 yuan, and the total daily profit after the daily optimized pumped storage unit participates in the market is 1,743,600 yuan.

[0099] The optimization bidding method of the present invention is measured by taking Xianyou Pumped Storage Power Station as an example. The optimization model is used to solve the different market capacity allocation and corresponding benefits in each period of 24 hours a day. The results show that the average daily economic benefits of the same pumped storage power station under the traditional planning and dispatching mode are higher. The optimization bidding method of the present invention enables pumped storage power stations to have more participation autonomy and profit space.

[0100] In summary, the bidding method of the present invention takes into account the risk of uncertainty in electricity prices, predicts electricity prices in the market, and provides technical support and auxiliary decision-making reference solutions for pumped-storage power stations to track market dynamics and formulate operation strategies in advance. The method of the present invention establishes a risk-return model for pumped-storage power stations to participate in the electricity market and ancillary service market, gives full play to the resource flexibility of pumped storage, realizes the optimal capacity allocation of the electricity market and ancillary service market, and maximizes benefits while buffering risks.

[0101] The above specific implementation modes are used to explain the present invention rather than to limit the present invention. Any modification and change made to the present invention within the spirit of the present invention and the protection scope of the claims shall fall within the protection scope of the present invention.

Claims

1. A multi-type electricity market joint bidding method applicable to pumped storage power stations, characterized in that: The steps include: (1) forecasting the on-grid electricity price based on historical data including load, surplus generation capacity, and electricity price information during each trading period; the on-grid electricity price is described by a probability density function; (2) The Monte Carlo sampling method is used to generate M electricity price forecast scenarios based on the estimated electricity price probability density function for each period of the day, and the K-means clustering scenario reduction method is used to reduce them to N typical scenarios. Based on the N typical scenarios, a CVaR risk model is established to measure the risk caused by the uncertainty of market electricity prices and complete the electricity price forecast; (3) Constructing an objective function based on the electric energy income, frequency regulation income and CVaR risk; the objective function in step (3) includes two parts, as shown below: Where, the first part f is the net income of the pumped storage power station in the electricity market and frequency regulation market, π w is the probability of scenario ω describing the uncertainty of electricity price, N ω is the number of typical electricity price scenarios; the second part is the product of the conditional value at risk CVaR and the risk preference coefficient β, and the risk preference coefficient β indicates the risk preference of the pumped storage power station owner; The net income f formula of the pumped storage power station participating in the electricity market and ancillary service market is as follows: Where i represents the unit time period, P i e represents the power generated in time period i, D i represents the pumping power in period i, P i re Indicates the frequency regulation capacity under power generation conditions, P i rd represents the frequency regulation capacity under pumping conditions, λ i e represents the forecast value of the on-grid electricity price in time period i, represents the frequency regulation capacity price, represents the predicted electricity price of frequency regulation mileage, R represents the ratio of frequency regulation mileage to frequency regulation capacity of pumped storage; (4) Setting generation / pumping rate constraints, reservoir capacity constraints, quantitative relationship constraints between generation / pumping / frequency regulation power and reservoir capacity, and CVaR-related constraints; In the risk value model, VaR refers to the income obtained by the pumped storage power station in market changes under a certain confidence level; CVaR represents the expected value of the profit distribution tail 1-α of the pumped storage power station holder under a certain time and confidence level α; the risk value CVaR related constraints are as follows: or ω ≥0 In the formula, ξ represents the value at risk (VaR), η ω is the difference between the return and risk value under scenario ω; The quantitative relationship constraint between the power generation / pumping / frequency regulation power and the reservoir capacity is as follows: V i+1 =V i -P i e ·a / n1-P i re ·0.5a / n2+D i ·a·η2+P i rd ·0.5a·n2 Where α is the coefficient of pumped storage power generation or conversion of pumped water into storage capacity, η1 is the loss rate of pumped storage during the pumping process, and η2 is the loss rate of pumped storage during the power generation process; The power generation / pumping rate constraint expression is as follows: P i e ·D i =0,P i e +D i ≥0 P min ≤P i e +P i re ≤P max 0≤D i +P i rd ≤D max P i re ·P i rd =0,P i re +P i rd ≥0 D i ·P i re =0 P i e ·P i rd =0 P i e ,D i ,P i rd ,P i re ≥0 P i re ≤εP max P i rd ≤εD max Where P min is the minimum power output of the pumped storage unit, P max is the maximum power output of the pumped storage unit, D max is the maximum pumping capacity of the pumped storage unit, ε is the upper limit of the capacity of the pumped storage unit participating in the frequency regulation market; The storage capacity constraint expression is as follows: V1=V 24 In min ≤V i ≤V max Where V i represents the storage capacity in time period i, V min Represents the minimum storage capacity, V max Indicates the maximum storage capacity; (5) Based on the objective function set in step (3) and the constraints set in step (4), a risk-return model for the pumped storage power station to participate in the electricity market and ancillary service market is established; (6) Using the CPLEX solver, the risk-return model established in step (5) is solved to obtain the production simulation results. If there is a solution, the pumping capacity, power generation capacity, frequency regulation capacity and market benefits of the pumped-storage power station in each period of an optimization cycle are obtained, and the most economically efficient bidding strategy is obtained.

2. The multi-type electricity market joint bidding method applicable to pumped storage power stations according to claim 1 is characterized in that: In step (1), the on-grid electricity price is described by a normal distribution function, and the formula is as follows: In the formula, i is the time period identifier, P i e is the market clearing price in the ith period, i.e., the on-grid electricity price; μ i The expected value of the on-grid electricity price for power generation; is the variance.

Images