Research on Wind Power Energy Storage Joint Optimization Operation under the Double Detailed Rules Assessment Taking into Account the Benefits of Green Certificate

Abstract

Due to the uncertainty of wind power outputs, there is a large deviation between the actual output and the planned output during large-scale grid connections. In this paper, the green power value of wind power is considered and the green certificate income is taken into account. Based on China’s double-rule assessment system, the maximum net income of the wind farm is used as the target and the optimization model is established with the maximum income of the wind-energy storage consortium as the objective function at 96 points on a typical day. The particle swarm optimization algorithm with improved weight and learning factor is adopted to solve the model and the energy storage output characteristics are analyzed. At the same time, the effect of changing the deviation coefficient and the penalty factor on the combined income is compared. The example results show that the wind storage consortium improves the stability of output, effectively reduces the double-rule assessment cost, and increases the green certificate income of the new energy station, which verifies the validity and correctness of the proposed method.

1. Introduction

1.1. Background

According to statistics, by the end of 2020, China’s total installed capacity of wind power is 281.72 million kilowatts, accounting for 12.8% of the national installed capacity of power generation, and the new installed capacity of wind power is 7167 kw. Accounting for 37.5% of the newly installed capacity in China. However, in 2020, the annual air abandonment volume reached 16.6 billion kWh, with an average of 3.5% [1]. At the same time, due to the uncertainty of its output power, the large-scale grid connection of wind power will bring negative effects on the stable operation of the system. The configuration of energy storage units in wind farms can not only promote the consumption of wind power but also stabilize the fluctuation of new energy output and improve the economy and safety of the system [2].

1.2. Literature Review

At present, there has been some examples in the literature on the combined operation of wind power and energy storage. In [3,4], from the perspective of the power grid, constructed the objective function by using the lowest total operation cost of thermal power units in the whole network or the lowest average daily output of conventional power plants. The conclusions all showed the positive effect of pumped storage on the power grid. The authors of [4] also analyzed how energy storage can effectively reduce coal consumption and realize energy savings and emission reduction in the power grid through the results of calculation examples. The authors of [5] established the optimal allocation model of energy storage capacity in the combined system of wind power and energy storage and analyzed the influence of sensitive factors such as feed-in price, energy storage cost, power shortage penalty factor, and regulation mode on capacity allocation by combining examples. The authors of [6] put forward an energy storage capacity optimization strategy considering the impact of the power shortage penalty cost and wind abandoning penalty cost on the combined operation economy of wind power and energy storage and proved through examples that when the reduction in penalty cost was greater than the increase in energy storage investment, the energy storage configuration would be economical. Both papers were from the perspective of the optimal allocation of energy storage capacity and both took into account the cost of wind power abandonment and the power deficiency of combined output of wind power and energy storage in the objective function. The results both analyzed the relationship between different configuration capacities of energy storage and the cost of wind abandonment and power deficiency. In [7], from the perspective of wind farm operation, based on the peak-valley electricity price, a multi-objective optimal dispatching model was established, which has the maximum operation benefits of wind power and pumped storage and the minimum fluctuation variance of the combined output of wind power and pumped storage. The results show that the fluctuation range of the combined grid power of wind power and pumped storage has been significantly improved. Because of the uncertainty of wind power output, the authors of [8] introduced the output penalty of wind power and an energy storage system and compared the economy of independent operation of wind power and energy storage systems and wind power under different output deviation coefficients and different penalty factors, which proved that the proposed model of wind power and energy storage system had a good economy. Both kinds of studies stood at the perspective of the power generation side. Considering the fluctuation of wind power and energy storage combined output and wind power planned output, the results show that the addition of energy storage can effectively improve the economic benefits of wind farms and effectively smooth the wind power fluctuations. Based on the operation characteristics of the combined wind power and energy storage system, the authors of [9] built a revenue maximization evaluation model for the wind power and energy storage system and studied the impact of the energy storage capacity level and the prediction accuracy of wind power on the economy of the combined system through sensitivity analysis. The authors of [10] built a wind power deviation electricity assessment model for the impact of deviation electricity assessment on the operating economy of wind power generation enterprises and conducted a sensitivity analysis on the assessment cost per unit of deviation electricity and the deviation exemption coefficient in combination with an example, proving that controlling the number of times of energy storage charging and discharging can bring higher benefits to wind power and energy storage systems. Additionally, from the perspective of the power generation side, these two papers built the joint optimization operation model of wind power and energy storage under the spot market environment of electricity. In the above examples in the literature, there are different ways to study the joint optimization operation of wind power and energy storage. Peak-valley electricity prices or spot market electricity prices are used to analyze the market income of wind power and energy storage consortium. Energy storage control strategies obtained by different electricity price curves have obvious differences. However, few articles are based on the market background of double regulation assessment and the assessment cost of a wind power grid connected to the model objective function. In addition, few examples in the literature consider the economic and environmental value or green electricity value of wind power and energy storage combined operation at the same time. Some examples in the literature analyze the low carbon emission reduction effect of wind power and energy storage combined operation in the result analysis. However, the environmental value or green electricity value is not taken into account in the objective function.

In terms of applying a particle swarm optimization algorithm to study wind power grid connection problems, the authors of [11] proposed an improved particle swarm optimization algorithm by adding an adaptive particle velocity and adding a displacement update adaptive to the uncertainty problem after wind power grid connection. In [12], the basic particle swarm optimization algorithm was used to solve the problem of maximizing the absorption of wind power and, based on the optimization problem of minimizing the cost of thermal power, hydropower, and pumped storage to participate in peak regulating power generation, the strategy of multi-agent joint peak regulating was developed; the authors of [13] used the improved IP-SO-LSTM prediction model to produce a short-term prediction of wind power in typical scenarios under different seasons. The results showed that the improved model has high accuracy. In [14], the ideas of chaotic initialization and chaotic disturbance were applied to improve the particle swarm optimization algorithm to solve the joint optimization problem of wind power and pumped storage; based on meeting the requirements of full grid connection of wind power, the authors of [15] constructed multi-objective optimization functions from the perspective of environmental protection and low-carbon, converted multi-objective optimization problems into single objective optimization problems using the maximum satisfaction method, and solved the optimization problems using improved particle swarm optimization algorithm.

Renewable Portfolio Standard (RPS) and Green Certificate Trading (GCT) have provided a new way to improve the consumption rate of renewable energy generation and reduce carbon emissions. In the research on green certificate trading, the authors of [16] introduced system dynamics, built an interactive trading model, and simulated it to clarify the multi-scale market coupling interaction relationship of electricity green certificate excess consumption under the new quota system (RPS); the authors of [17] established quota system models for different subjects, designed an enhanced quota system that takes into account the differences and penalties of power producers, and built a comprehensive evaluation index to reflect the number of green certificates obtained by different power producers. The results showed that the designed enhanced quota system could effectively promote the consumption of new energy; the authors of [18] established a “source load” bilateral complementary coordination optimization scheduling model based on green certificate trading and carbon trading system. The results showed that the model taking green certificate trading and carbon trading into account could effectively reduce carbon emissions; the authors of [19] combined the carbon emission mechanism and the green certificate trading mechanism to build an IES joint trading optimization operation model considering the value at risk. The results showed that considering these two trading mechanisms at the same time had high economic value and environmental value. Based on the research so far, most of these documents combine green certificate trading with a renewable energy quota system or carbon trading mechanism, taking into account the green certificate trading cost and the constraints of the renewable energy quota system, to build a market entity trading model, thus effectively playing the role of green certificate trading and promoting the consumption of new energy. However, few articles, from the perspective of new energy, combine its income in the electric energy market and the green certificate market at the same time to build an optimization model and analyze the interaction between the two markets.

Therefore, based on the above research, this paper considers both the electricity energy market returns of wind farms and the green certificate market returns. To encourage new energy stations to improve prediction accuracy, a penalty model for green certificate acquisition quantity is designed to calculate the green certificate returns. It aims to maximize the operating returns of wind power-energy storage systems and considers the joint output penalties and constraints of wind power and energy storage system based on the double-rule assessment system. The optimization model of wind-energy storage combined daily operation was established, the particle swarm optimization algorithm with improved weight and learning factor was used to solve the model, the energy storage output characteristics were analyzed, and the economy of independent operation and operation of wind-storage combined system under different scenarios was compared. Meanwhile, the deviation coefficient and penalty factor were changed. By comparing the wind farm electric energy market income with the penalty cost and the change of green certificate income, the feasibility of the model is verified by the simulation example.

The main contributions of this paper include the following:

(1)

Based on the double-rule assessment system for power plants adopted by many provinces in China, this paper first considers the assessment costs when new energy is connected to the grid, which leads to the important role of energy storage in improving the accuracy of new energy prediction, which can effectively reduce the assessment costs of new energy, improve the revenue from electricity sales, reduce wind abandonment, guide new energy stations to configure energy storage, and solve the security problems of large-scale new energy grid connection in the future. Promoting the sustainable development of green energy plays an important role.

(2)

Secondly, this paper considers the benefits of green electricity in the green certificate market, constructs a green certificate revenue model, and designs a green certificate quantity assessment model based on the accuracy of forecast output to promote energy storage to better track the new energy output, further reduce the grid connection assessment fees of new energy stations, which is of great significance for paying attention to the value of green certificates, more reasonable arrangement of energy storage charging and discharging, and promoting the consumption of new energy.

(3)

In addition, considering the disadvantage of local convergence of traditional particle swarm optimization algorithms, this paper improves the learning factor and inertia coefficient of the algorithms. The results of numerical examples show that the improved algorithm has better stability and superiority.

(4)

This paper couples the electric energy market and the green certificate market, build the optimal operation model of the wind storage consortium based on the maximum revenue of the combined market, and obtains the optimal charging and discharging strategy of energy storage. The feasibility of the model is verified by an example, which improves the electricity sales revenue and green certificate revenue of new energy manufacturers and is of great significance for new energy stations to guide energy storage operations.

(5)

This paper compares and analyzes the benefits under different deviation coefficient and penalty factor scenarios and draws reasonable conclusions. Furthermore, the reliability of the model used in this paper shows the flexible regulation of energy storage.

Comparison of the advantages of this work with recent works is shown in

Table 1. Comparison of the advantages of this work with recent works.

Refs.	On the Power Generation Side	Energy Storage	Environmental Value/Green Power Value	Include Green Power Value or Environmental Value in the Objective Function	Wind Power Prediction	Energy Storage Cost
[3]	✕	✓	✓	✕	✕	✕
[4]	✕	✓	✕	✕	✕	✕
[5]	✓	✓	✕	✕	✓	✕
[6]	✓	✓	✕	✕	✓	✓
[7]	✓	✓	✕	✕	✓	✕
[8]	✓	✓	✕	✕	✓	✕
[9]	✓	✓	✕	✕	✓	✓
[10]	✓	✓	✕	✕	✓	✓
This work	✓	✓	✓	✓	✓	✕

.

The rest of this paper is designed as follows: Section 1 introduces the research status of the wind power and energy storage combined operation, particle swarm optimization algorithm, and green certificate trading. Section 2 introduces the basic particle swarm optimization algorithm and the improved particle swarm optimization algorithm applied in this paper. Section 3 constructs the green certificate trading model and the daily optimal operation model of the wind power and energy storage combined day. Section 4 carries out numerical simulations to prove the feasibility of the model and method in this paper.

2. Strategy and Modeling

This section introduces the basic particle swarm optimization algorithm and particle swarm optimization algorithm with improved learning factor and inertia coefficient.

2.1. Basic Particle Swarm Optimization

The origin of the particle swarm optimization algorithm is the research of birds foraging. This method finds the optimal solution through information sharing and cooperation among individuals in the group. In this algorithm, n particles fly at a certain speed in the D-dimension space to find the optimal solution [20]. Each particle has a fitness value, which is determined by the objective function. At the same time, each particle knows the best location p_best and the current location Xi, which can be seen as the flight experience generated by each particle itself. At the same time, each particle can also be clear at this time in the population of all the particles found in the best position for g_best. Each particle in the whole population determines the direction and trajectory of the next action through its own experience and the best experience generated by other particles. The velocity and position update formulas of the particles are, respectively [21,22]:

$$\left\{\begin{array}{c}{v}_i^{t+1}=\omega v_i^t+c_1{r}_1\left(pbes{t}_i^t-x_i^t\right)+c_2{r}_2\left(gbes{t}_i^t-x_i^t\right)\\ x_i^{t+1}=x_i^t+v_i^{t+1}\end{array}\right.$$

(1)

where $v_i^t$, $v_i^{t+1}$, respectively, represents the velocity of the ith $t$, $t+1$ iteration; $x_i^t$, $x_i^{t+1}$ represents the position of the ith particle in the $t$, $t+1$ iteration, respectively; $pbes{t}_i^t$, $gbes{t}_i^t$, respectively, represents the optimal solution of the ith particle in the population and the optimal solution of the population at the iteration; $c_1$, $c_2$ represent learning factor; $r_1$, $r_2$ are random numbers generated between 0 and 1.

In this method, different particles represent different possible solutions and the current velocity of particles determines the velocity of particles at the next time. During the flight, particles constantly modify their flight direction and speed according to their own and group information to obtain the optimal solution. The basic principle and flow of the particle swarm optimization algorithm are shown in Figure 1.

2.2. Improved Particle Swarm Optimization Algorithm

The basic particle swarm optimization algorithm has strong computing power in solving the optimization function and it can quickly find the approximate solution through the iteration cycle. However, this algorithm is prone to fall into local optimization in some application scenarios. To overcome the problem that a traditional particle swarm optimization algorithm is prone to have a low solving degree, this paper adopts the method of improving the learning factor and inertia weight to improve the particle swarm.

The convergence and speed of the algorithm depend on the speed of the particles. If the speed of the particles is too large, the algorithm converges quickly, but it is easy to fly over the optimal solution. If the particle speed is too slow, the algorithm convergence is slow and it is easy to enter the local optimum. Shi and Eberhart’s introduced the inertia weight coefficient into the algorithm ω (generally, it adopts a fixed coefficient between 0.8 and 1.2), which can control the speed of particles and effectively improve the global search ability of the algorithm [11]. However, the fixed inertia weight coefficient will affect the convergence speed of the algorithm. If the value is unreasonable, the algorithm will converge too slowly or even fail to converge. The larger the particle is ω, the faster the particle flies; the larger the step size, the larger the global search will be, which is conducive to jumping out of the local minimum. The smaller the ω, the smaller the flight speed of the particles, the smaller the particle step size, and the smaller the range of movement of the particles in this iteration, thus allowing the particles to search in a smaller range, causing this iteration to be more inclined to a fine local search. In this paper, the weight coefficient adopts the linear decreasing weight. With the increase in the number of iterations, the inertial weight changes from the maximum to the minimum. The formula for the change is:

$$\omega ={\omega }_{max}-\frac{t\times \left({\omega }_{max}-{\omega }_{min}\right)}{t_{max}}$$

(2)

where ${\omega }_{max}$ and ${\omega }_{\min }$ represent the maximum and minimum values of $\omega$, respectively; $t$ represents the current iteration steps, and $t_{\max }$ represents the maximum number of iteration steps.

The size of learning factors $c_1$ and $c_2$ determine the particle’s self-learning ability and social learning ability, respectively. They have the characteristics of improving summary and learning from the outside world so that the particle is close to the best in the group or field. At the same time, they can adjust the maximum step size when the particle flies. When the learning factor is too small, the particles may wander in the area far away from the target and fall into the local optimum. When the learning factor is too large, the particles will quickly approach the target area and, if the speed is too fast, they may exceed the area [23]. Therefore, the values of $c_1$ and $c_2$ will directly affect the results of the PSO algorithm. In this paper, asynchronous learning factors are used to change the values of $c_1$ and $c_2$, as shown in the following formula [24]:

$$\begin{array}{l}{c}_1=c_{1,start}+\frac{\left(c_{1,end}-c_{1,start}\right)t}{t_{max}}\\ c_2=c_{2,start}+\frac{\left(c_{2,end}-c_{2,start}\right)t}{t_{max}}\end{array}$$

(3)

where $c_{1,start}$ and $c_{1,end}$ are the initial and final values of $c_1$, $c_{2,start}$, and $c_{2,end}$ are the initial and final values of $c_2$, $t$, and $t_{max}$ are the current iteration number and the maximum iteration number, respectively, the value of $c_{1,start}$, $c_{1,end}$, $c_{2,start}$, and $c_{2,end}$ in this paper are $c_{1,start}=2.5$, $c_{1,end}=0.5$, $c_{2,start}=0.5$, and $c_{2,end}=2.5$.

3. Construction of Daily Profit Model of Wind Storage Joint Operation Taking into Account Green Certificate Income

In this section, we first built a green certificate income model and then built a joint optimization operation model of wind storage based on this model.

3.1. The Construction of the Green Certificate Income Model

The Green Certificate Transaction (GCT) is an electronic certificate with unique code identification issued by the state to renewable energy power generation enterprises for each megawatt hour of renewable energy power generation. The value of the green certificate represents the external environmental contribution of hydropower, wind power, photovoltaic power generation, and other renewable energy compared with traditional fossil energy power, generation, which is reflected through the transaction price of the green certificate in the market [25].

The source of the green certificate is to prove the green power attribute of renewable energy. Although the green power attribute of renewable energy can be confirmed when it is generated from the power plant, after it is integrated into the grid together with other power sources such as thermal power generation, due to the physical properties of the current, it is difficult to distinguish the electricity generated by different power generation methods. When purchasing electricity, it is also difficult for users to know or prove the energy properties of the purchased electricity. In order to solve the problem of “identity” proof of electricity generated from renewable energy generation projects, the green certificate is designed as the “identity card” of green electricity.

In addition to the function of proving the attribute of green electricity, the green certificate has been endowed with more abundant functions in the process of its institutional development. First, to stimulate the development of the renewable energy industry, countries represented by the United States have gradually established a renewable energy quota system. The countries that have established quota systems in the world usually use the green certificate trading system as a supporting system of the quota system, that is, the government issues green certificates corresponding to the amount of electricity generated by renewable energy power generation enterprises. The quota-responsible subject can prove that the quota target has been achieved by purchasing green certificates from power generation enterprises. China has also gradually explored the establishment of a quota system in recent years.

In addition, a green certificate is one of the main incentives for the construction of renewable energy power generation projects. Power generation enterprises can obtain additional income by selling green certificates and gradually reducing their dependence on subsidies or alternative subsidies. At present, China’s green certificate market is still in the initial stage of development, but the policy system of “quota system + green certificate” has been clarified and is gradually improving. Renewable energy obtains green certificates while participating in power market transactions and settling at market prices to obtain electric energy income and obtains additional income by selling green certificates on the green certificate trading platform (the economic rights and interests of green certificates are fully vested in renewable energy enterprises), to realize the separate pricing and transaction of the electric energy value and green environmental value of renewable energy, thus promoting the sustainable development of the entire renewable energy industry [26].

The specific green certificate trading mechanism in China is as follows:

(1)

Issuing subject and object: The National Renewable Energy Information Management Center will issue the green certificate and issue the corresponding certificate to the enterprise according to the standard of 1 MWH of settlement electricity for each certificate.

(2)

Transaction subject and transaction method: Onshore wind power and ground centralized photovoltaic power stations and also wind power generation projects with low prices and low prices that are included in the national renewable energy price subsidy catalog can apply for green certificates. Distributed photovoltaic projects and offshore wind power projects are not qualified to apply for the green certificate. The buyer of green certificates refers to government agencies, enterprises, institutions, social institutions, and individuals at all levels. The trading methods are mainly listed as sale and agreement transfers.

(3)

Pricing of green certificates and separation of securities and electricity: China adopts a voluntary subscription system for green certificates. The subscription price is determined by the buyer and the seller through self-negotiation or bidding by the amount of additional fund subsidies of renewable energy electricity price not higher than the corresponding electricity quantity of the certificate. In addition, each green certificate transaction does not correspond to the physical transportation or use of green electricity one by one. That is, the purchase of green certificates declaring the use of green electricity does not mean that enterprises consume the corresponding green electricity. As an asset certificate, green certificates show the support of government agencies, enterprises, or individuals on the attitude of new energy power generation.

It can be seen that the characteristics of green syndrome in China are as follows:

(1)

Separation of syndrome and electricity

The biggest advantage of the separation of securities and electricity is that it can solve the restriction of power transmission capacity on green certificate trading, promote the consumption of renewable energy, reduce the cost of building green power transmission channels, and promote the process of carbon neutralization. However, the separation of securities and electricity is also likely to lead to the disconnection between green securities trading and green electricity trading. At present, except for those electricity sellers in New Jersey and Pennsylvania who can purchase bundled green certificates and renewable energy electricity through the wholesale electricity market, most countries still adopt the trading system of separating securities from electricity.

(2)

Voluntary transactions

The second characteristic of green certificate trading in China is voluntary. China’s green certificates are currently subject to a voluntary subscription system. The subscription work has been officially carried out since 1 July 2017. According to the market subscription situation, the quota assessment of renewable energy power and the mandatory binding transaction of green power certificates have been launched in 2018, however, the mandatory constraint transaction has not yet started. The restrictions and incentives of the voluntary subscription trading system on the market participants in the transaction are weak. Due to the lack of additional preferential policies that can be enjoyed when purchasing green certificates, the main motivation for government agencies, enterprises, or individuals to purchase green certificates is to fulfill their social responsibilities and establish their corporate image. The level of green certificate trading is low. If only relying on the voluntary green certificate trading market, the development of green certificate trading in China will be very limited. At present, the European Green Certificate GO (Guarantees of Origins) mechanism is a voluntary market for green certificates. At the same time, Norway, Sweden, and other countries have established a mandatory market for green certificates with quota obligations, but it is independent of the GO system; the green certificate mechanism in the United States is also a combination of voluntary transactions and forced markets. At the same time, the purchase channels and methods in the voluntary market are relatively flexible and the purchase methods of voluntary non-bundled renewable energy certificates still account for the largest proportion of the total green certificate market.

The main participants in China’s GCT are new energy power generation companies, power-selling enterprises, and power users. In the case of the combination of green certificate trading and renewable energy quota system, when the number of green certificates obtained from the National Renewable Energy Information Management Center is more than the number required by the assessment of consumption responsibility weight, the enterprise can sell the excess green certificates to obtain profits. On the contrary, the corresponding quantity of green certificates should be purchased to meet the assessment requirements [27]. The transaction process is shown in Figure 2:

According to the definition of green certificate trading, the number of green certificates required by the system to meet the renewable energy quota system is [27]:

$$N_{GCT}=P_q\cdot \Delta t$$

(4)

where $P_q$ is the new energy power generation capacity of the system based on the renewable energy quota system, with the unit of MWh, $\Delta t$ is the unit dispatching period, and $N_{GCT}$ is the number of green certificates required in the system dispatching period.

After the establishment of the green certificate trading mechanism, renewable energy power generation enterprises can not only obtain physical electricity price income through the electric energy market but also obtain market subsidies by selling green certificates. To encourage new energy power generators to improve prediction accuracy and encourage new energy stations to configure energy storage with a certain capacity, this paper refers to [27] to punish the number of green certificates according to the accuracy of power generation prediction of new energy stations. When allocating green certificates, consider the current new energy forecast situation. For enterprises whose accuracy fails to meet the industry standard, deduct the number of green certificates obtained per unit of green electricity. The green certificate income of renewable energy power generation enterprises is as follows:

$$Q_{GCT}={\alpha }_{GCT}{\displaystyle \sum _{i=1}^T{p}_w(t)\Delta t}-\theta {\alpha }_{GTC}\left(p_w(t)-\gamma p_k^f(t)\right)\Delta t$$

(5)

where $T$ is the dispatching period, $\Delta t$ is the unit dispatching period, ${\alpha }_{GCT}$ is the green certificate trading price, $p_w(t)$ is the actual wind power output, unit: MWh, $p_k^f(t)$ is the planned wind power output, unit: MWh, $\gamma$ is the minimum prediction accuracy of new energy stations that will not reduce the green certificate, which is recorded as 80%, $\theta$ is the penalty coefficient, which is recorded as 0.6 [27].

3.2. Construction of Daily Revenue Model of Wind Storage Joint Operation

In the context of the national energy transformation strategy, to promote the construction of a new power system with new energy as the main body, ensure the safe, high-quality, and economic operation of the power system and effectively standardize the grid-connected operation management and auxiliary service management of the power system, the National Energy Administration of China, The State Power Regulatory Commission, and regional power regulatory authorities (Northeast and North China and Northwest, East, Central, and South Energy Regulatory Bureaus) have successively formulated and issued the Implementation Rules for Grid-connected Operation Management of Power Plants and the Implementation Rules for Grid-connected Auxiliary Service Management of Power Plants (hereinafter referred to as the “two rules”). The “two detailed rules” include new energy as the main grid-connected entity in the assessment to balance the output and profit distribution of all power sources in the power system and assume the security responsibility of the grid. The Detailed Rules for the Implementation of Grid-connected Operation Management of Power Plants put forward requirements on the prediction accuracy of new energy. Using the Northwest double-rule newly issued by the Northwest Energy Regulatory Bureau in 2019 as an example, Article 33 requires that wind farms and photovoltaic power plants should submit short-term power prediction curves to the power regulatory authority on time. The maximum error of daily prediction curves provided by wind farms should not exceed 25% and the maximum error of daily prediction curves provided by photovoltaic power plants should not exceed 20%. If it fails to meet the standard, it will be assessed according to the deviation integral electricity 0.2 point/10,000 kWh. The assessment score of grid-connected operation management of the power plant will be converted into the cost. Therefore, if the new energy forecast rate fails to meet the standard, it may bear high grid-connected assessment costs. Therefore, in combination with China’s double-rule assessment system rules, this paper comprehensively considers the deviation between the joint output of wind power storage and the planned output, builds an optimization model for maximizing the full-day revenue of the joint operation of wind power storage energy storage system, and simultaneously considers the value of green power. The above green certificate benefits are taken into account in the objective function and a series of constraints such as energy storage and joint output penalties are considered [8]. The synergy between wind power, energy storage, and the grid is shown in Figure 3.

3.2.1. Objective Function

$$\begin{array}{l}\max {\displaystyle {\prod }_{(w-e)}=}{\displaystyle {\sum }_{t=1}^T\lambda (t)\left(p_w(t)+u_k^d(t)p_k^d(t)-u_k^c(t)p_k^c(t)\right)}-\\ m_k\left(\left|p_w(t)+u_k^d(t)p_k^d(t)-u_k^c(t)p_k^c(t)\right|-\epsilon p_k^f(t)\right)+Q_{GCT}\end{array}$$

(6)

where Item 1 represents the electricity sales income of net on-grid electricity. The second item represents the punishment for the maximum allowable deviation from the provisions and the third item represents the estimated green certificate income of the wind farm.

Where $T$ is the Total period divided, this paper takes a total of 96 periods in 24 h, $\lambda (t)$ is the energy market price of the electric energy market, and $p_w(t)$ is the actual output of the wind turbine in the t period; $p_k^d(t)$ is the discharge power of the energy storage system in period t, $p_k^c(t)$ is the charging power of the energy storage system in period t, and $p_k^f(t)$ is the planned output of wind power in period t; $u_k^d(t)$ is the 0–1 integer variable of the period t when the energy storage system is in the discharge state; $u_k^c(t)$ is the 0–1 integer variable of the period t when the energy storage system is in the charging state; $m_k$ is the penalty coefficient of the Kth period; and $\epsilon$ is the maximum deviation coefficient allowed to deviate from the planned output under specified conditions. Among them, considering the green certificate benefits after the operation of energy storage:

$$Q_{GCT}={\alpha }_{GCT}{\displaystyle \sum _{t=1}^T\left[\left(p_w\left(t\right)+p_c\right)\Delta t/4-\theta {\alpha }_{GTC}\left(p_w(t)+p_c-\gamma p_k^f(t)\right)\Delta t/4\right]}$$

(7)

where $p_c$ is the energy storage output.

3.2.2. Constraints

(1)

Energy storage system constraints

In this paper, energy storage constraints mainly consider the state equation of energy storage system operation, upper and lower limit constraints of energy storage system output, etc.:

$$\left\{\begin{array}{c}0\le p_k^d(t)\le p_{discharge}^{max}\\ \begin{array}{l}0\le p_k^c(t)\le p_{charge}^{max}\\ u_k^d(t)+u_k^c(t)=1\\ u_k^d(t),u_k^c(t)\in \left\{0,1\right\}\end{array}\end{array}\right.$$

(8)

where $p_{discharge}^{max}$, $p_{charge}^{max}$ is the maximum discharge power and the maximum charging power of the energy storage device, respectively, assuming that $p_{discharge}^{max}=p_{charge}^{max}=p_{rated}$; $p_{rated}$ is the rated power of the energy storage system.

At the same time, storage energy constraints, energy conversion equations, and energy balance constraints are also considered during the operation of the energy storage system [28]:

$$\left\{\begin{array}{l}{E}_{min}\le E(t)\le E_{max}\\ E(t+1)=E(t)-p_k^d(t)/{\eta }_{Ei}^{dis}-{\eta }_{Ei}^{cha}{p}_k^c(t)\\ E_T=E_0\end{array}\right.$$

(9)

where $E(t)$ is the energy stored under period t; $E_{max}$ is the upper limit of stored energy and $E_{min}$ is the lower limit of stored energy; ${\eta }_{Ei}^{dis}$ is the energy storage discharge efficiency; ${\eta }_{Ei}^{cha}$ is the energy storage charging efficiency. $E_0$, $E_T$ represents the capacity at the beginning and end of the energy storage cycle, respectively.

(2)

Penalty constraint for a combined output of wind power and energy storage

$$m_k=\left\{\begin{array}{l}0 p_W(t)+u_k^d(t)p_k^d(t)-u_k^c(t)p_k^c(t)-p_k^f(t)|\le \epsilon p_k^f(t)\\ \omega \lambda \left(t\right) |p_W(t)+u_k^d(t)p_k^d(t)-u_k^c(t)p_k^c(t)-p_k^f(t)|>\epsilon p_k^f(t)\end{array}\right.$$

(10)

where $\omega$ is the penalty factor and $m_k$ is set as the penalty factor to reduce the difference between the joint output of the wind storage and the planned output and the maximum allowable deviation of the specified wind farm.

4. Case Study

This section first compares the benefits of independent operation of wind farms and joint operation of wind farms and then analyzes the changes in penalty fees, electricity sales revenue, and green certificate revenue of grid-connected assessment by changing the penalty factor and deviation coefficient of grid-connected assessment.

4.1. Model Basic Parameter Setting

This paper selects a wind farm with an installed capacity of 300 MW in a province of China and selects its typical 24 h wind power plan and actual output data of 96 points, as well as energy storage configuration parameters to verify the proposed model. The planned and actual output power of wind power generation are shown in Appendix A.

The configuration parameters of energy storage are shown in

Table 2. Energy storage configuration parameters.

E0/MWh	Emin/MWh	Emax/MWh	P0/MW	ndis	ncha
45	15	135	80	0.9	0.8

. E₀ is the initial capacity of energy storage, E_min and E_max are the minimum and maximum capacity of energy storage, P₀ is the rated power of energy storage, n_dis and n_cha are the discharge and charging efficiency of energy storage:

The on-grid price of wind power in this paper is the peak and valley time of use price. The setting is shown in

Table 3. The on-grid price of wind power.

Time Interval		Electricity Price/(CNY/KWh)
Peak time	9:00–13:00	0.80
	18:00–22:00
Flat Valley Period	6:00–9:00	0.61
	13:00–18:00
Low period	0:00–06:00	0.32
	22:00–24:00

. The specific electricity prices for 96 h of the day are in Appendix A. The green certificate transaction price is 50 CNY/book [27]:

The grid price curve of wind power is shown in Figure 4:

For the penalty coefficient $m_k$, according to the statistical results of frequency modulation, synchronous reserve, and average electricity price in the American PJM market, it will be recorded as 0.44. At the same time, it is assumed that the maximum deviation of allowable output ε = 5% [8].

4.2. Result Analysis

4.2.1. Independent Operation of Wind Power Plant

First, ignore the impact of energy storage output on the wind farm and analyze the income of the wind farm when it operates alone. The model of the wind farm when it operates alone is as follows:

$${\displaystyle \sum {}_{t=1}^T}\lambda (t)\times p_W(t)-m_k(|p_W(t)-p_k^f(t)|-\epsilon p_k^f(t))+Q_{GCT}$$

(11)

including:

$$m_k=\left\{\begin{array}{l}0 \left|p_W(t)-p_k^f(t)\right|\le \epsilon p_k^f(t)\\ \omega \lambda \left(t\right) |p_W(t)-p_k^f(t)|>\epsilon p_k^f(t)\end{array}\right.$$

(12)

The planned and actual output results are shown in Figure 5:

It can be seen that the deviation between the planned output and the actual output of the wind power plant is large from the 28th to 48th periods and from the 80th to 96th periods and the corresponding periods are 7:00–12:00 and 20:00–0:00. In the first period, the electricity price is higher, the actual output of wind power is greater than the planned output, and the revenue from electricity sales is more than the penalty cost. However, due to the large deviation of prediction, the revenue from the green certificate is reduced; in the second period, the actual output is smaller than the planned output, the electricity price is lower, the revenue from the electricity sales is lower, and the penalty cost is increased; and the revenue from the green certificate is reduced.

When the wind power plant operates independently, the actual electricity sales revenue is CNY 2,790,400, the penalty cost is CNY 100,540, and the green certificate revenue is CNY 222,890.

4.2.2. Combined Wind and Energy Storage Output

When energy storage is combined with the wind power, the combined wind power-energy storage output curve is shown in Figure 6 and the charge–discharge power of the energy storage system is shown in Figure 7.

It can be seen from Figure 6 that the deviation between the combined output of wind storage and the planned output of the wind power is smaller than that when the wind farm operates independently. After the energy storage is added, the planned output can be tracked more efficiently, which can reduce the penalty cost of grid connection assessment. The income from the combined output of wind storage is CNY 2,878,100 and the green certificate income is CNY 235,790. The penalty is CNY 50,380. Compared with not adding energy storage, the income of combined wind storage increased by CNY 87,700, or 3.14%; at the same time, the penalty cost decreased by CNY 50,160, or 50.27%.

According to the analysis of Figure 7, it is in the low electricity price period from 0:00 to 06:00 and the planned output is higher than the actual output. Therefore, the energy storage stores a large amount of wind power during this period. During the peak electricity price period from 18:00 to 22:00, the energy storage releases power. In the electricity price flat section, the charging and discharging state and power are determined by comparing the penalty cost and income and the low storage and high discharge are carried out, thus increasing the electricity sales income of the consortium.

The influence of the maximum deviation coefficient ε and penalty factor ω on wind farm income and penalty cost is analyzed.

Firstly, by changing the ε value from 5% to 10%, the income of the wind-energy storage combined output is CNY 2,907,600 and the penalty cost is CNY 31,800. The energy storage output and the wind-energy storage combined output are shown in Figure 8 and Figure 9.

According to the analysis in Figure 8 and Figure 9, with the relaxation of the maximum deviation coefficient, the deviation between the output of wind power and energy storage consortium and the planned output increases, the penalty cost is greatly reduced, which decreases by 36.7% compared with the ε = 5%, and the income from electricity sales increases by CNY 29,500.

Second, after appropriately increasing the value of the penalty factor ω (ω = 0.88, ω = 1.32) and while keeping the other variables as set values of the initial example (ε = 5%), the combined output of the wind power and energy storage are, respectively, as shown in Figure 10 and Figure 11.

It can be seen from Figure 10 and Figure 11 that the deviation between the two curves of the combined output of wind power and energy storage and the planned output of wind power is decreasing. As the penalty factor increases, the penalty cost increases and the total income of the wind farm decreases.

To sum up, the results of the combined operation revenue and penalty cost of wind power and energy storage in each scenario are shown in

Table 4. Operation results under each scenario.

Scene		Green Certificate Income/CNY 10,000	Electricity Sales Income/CNY 10,000	Punishments for Grid Connection Assessment/CNY 10,000	Total Income of Electric Energy Green Certificate Market/CNY 10,000
ω = 0.44	ε = 5%	23.579	269.269	5.038	287.81
ω = 0.44	ε = 10%	22.76	271.18	3.18	290.76
ω = 0.88	ε = 5%	22.425	268.885	7.87	283.44
ω = 1.32	ε = 5%	21.509	268.028	9.057	280.48

:

It can be seen that under various scenarios of wind power-energy storage joint operation, when the penalty factor ω increases, the output of wind power and the energy storage consortium becomes closer and closer to the planned output, the grid connection assessment penalty cost of wind power still increases, and the revenue from electricity sales decreases. Although the penalty amount for green certificates decreases, the revenue from green certificates still decreases as the combined output of wind power and energy storage decreases. The total revenue of the electric energy-green certificate combined market decreased; when the maximum deviation coefficient ε increases, the deviation between wind power and energy storage combined output and wind power planned output increases, the grid connection assessment penalty cost of wind power decreases, and the revenue from electricity sales increases. However, the large deviation from planned output results in an increase in the green license penalty, a decrease in the green license revenue, and an increase in the total revenue from the electric energy-green license combined market.

To sum up, the introduction of energy storage can effectively reduce the cost of grid connection assessment and improve the revenue of the green certificate market for the operation of individual wind farms. When the penalty factor and deviation coefficient of the grid connection assessment change, energy storage can take into account the revenue changes of the green certificate market and electric energy market at the same time, play a better role in regulation, maximize the comprehensive economic benefits of new energy, and maintain the stability of the revenue of new energy stations, thus promoting the consumption of clean energy.

5. Conclusions

Based on the current policy trend of requiring new energy enterprises to allocate a certain capacity of energy storage in China, this paper mainly focuses on the role of energy storage in reducing assessment costs in the grid connection of new energy. Based on the double-rule assessment rules, a joint optimization model of wind power and energy storage is constructed. A penalty mechanism for the number of green certificates is designed considering the value of green electricity in the green certificates market and the particle swarm optimization algorithm is applied to solve the problem. The results verify the effectiveness of the proposed model. This paper shows that energy storage can track the output of new energy flexibly, which can effectively improve the benefit of green certificates and reduce the grid connection assessment cost of new energy stations. With the change of deviation coefficient and penalty factor, the energy storage output changes accordingly to maximize the comprehensive return of new energy stations, which verifies the stability of the model proposed in this paper and further demonstrates the flexibility regulation effect of energy storage. In addition, for new energy stations, the formulation of a reasonable energy storage charging and discharging strategy can effectively improve prediction accuracy, smooth wind power fluctuations, and effectively improve market returns. In addition, energy storage also has positive benefits for the power grid, which solves the disadvantages of the instability of new energy grid connections, promotes the stability of the power grid operation, and ensures the stable transmission of electric energy.

The main conclusions of this paper are as follows:

(1)

The energy storage can track the output of new energy well and charge and discharge flexibly according to the market price, which has a good regulatory role. Compared with the independent operation of wind farms, it can greatly reduce the cost of new energy double-rule assessment systems.

(2)

After adding the punishment mechanism for the number of green certificates, due to the double detailed rules assessment and the double assessment of green certificate transactions, the prediction accuracy of the new energy and energy storage consortium can be further improved, so that the actual output of the wind storage consortium is closer to the planned output of wind power, reducing the assessment costs and promoting the consumption of new energy.

(3)

Changing the maximum deviation coefficient and penalty factor in the grid connection assessment will subsequently affect the electricity sales revenue, penalty cost, and green certificate revenue of the wind storage consortium. The corresponding change trend is: with the increase in the maximum deviation coefficient, the grid connection assessment cost of the wind farm will decrease, the electricity sales revenue will increase, the green certificate revenue will decrease, and the total revenue of the electricity green certificate joint market will increase; with the increase in the penalty factor, the assessment cost of wind farm grid connection increases, the income from power sales decreases, the income from green certificate decreases, and the income from the combined market of electric energy and green certificate decreases, which reflects the good regulation ability of energy storage. Reasonable charging and discharging can better offset the penalty cost, which has good positive suggestions for guiding new energy stations to configure energy storage with a certain capacity.

(4)

Through the simulation, the validity of the model proposed in this paper is verified, which can provide reference suggestions for setting up a green certificate trading assessment mechanism in the future.

This paper is of great significance to the study of the optimal operation of wind storage and the coupling of the electric energy market and the green certificate market. The market background of this paper is based on China’s planned power market and has not yet involved the spot power energy market that is being vigorously promoted. Therefore, the future learning work will discuss the optimal operation of wind storage in the spot market, involving the day ahead, day in, and even real-time operation strategies. The electricity price in the spot market is more flexible than that in the planned market. With the increase in the price difference between peak and valley, wind power will be configured to store energy for operation and better market returns may be obtained through the low reserve and high release. In addition, the spot market may also face more sources of assessment costs, such as deviation assessment costs, unbalanced funds, etc. On the other hand, this paper discusses the 24 h optimal operation of wind storage, without taking into account the investment in energy storage, charging and discharging costs, etc., and the investment cost of energy storage is often high. Therefore, when discussing the long-term optimal operation of energy storage in the future, it is necessary to comprehensively consider various costs of energy storage to help new energy manufacturers make better investment decisions. Third, in many documents, green certificate trading is usually considered together with carbon trading. The green certificate records the non-electricity attribute including environmental rights and interests, while the environmental rights and interests attribute is generally expressed by showing the carbon emission reduction corresponding to the one-megawatt hour of green power on the green certificate. Therefore, the green certificate itself has the role of proving carbon emission reduction, such as establishing a linkage mechanism between the green certificate and the carbon market. The green certificate can also be used as a carbon emission reduction proof to participate in the carbon market. This paper does not consider carbon emissions and carbon trading for the time being. In the future, we can further study the coupling of these two markets and even more markets. Finally, with the vigorous implementation of independent energy storage in the power market, independent energy storage can obtain multiple benefits. In the future, it is also of great research value to consider the access point of energy storage and compare the benefits under these two business models, whether the new energy stations will continue to implement the integrated operation of wind storage or build independent energy storage.