A frequency adaptive protection method for weak inertia supported power grid

Abstract

The present invention discloses a method for adaptive frequency protection of a weak-inertia-supported power grid, belonging to the field of power system operation and maintenance. The method first estimates the system's power deficit using the power deficit borne by the generator and the power deficit borne by the load node. Based on the traditional low-frequency load shedding scheme, the adaptive low-frequency load shedding scheme proposed in this invention differs from the traditional "successive approximation" approach by providing two basic control stages and two correction control stages. While enabling the frequency to quickly restore to the rated value and accurately shedding the load, it can effectively prevent secondary power deficit accidents caused by low-frequency protection of wind turbines, thereby improving the stability of power system operation.

Description

Technical Field

[0001] The present invention belongs to the field of power system operation and maintenance, and in particular relates to a frequency adaptive protection method for a weak inertia supported power grid. Background Art

[0002] As the proportion of wind turbines in power systems continues to rise, the frequency stability issues brought about by wind turbines connected to the grid are attracting more and more attention, profoundly affecting the safety and stability of the power system. Among them, low-frequency load shedding is the main control method of the third line of defense in the safety and stability control of the power system. However, the traditional low-frequency load shedding strategy is no longer applicable to today's power grids with a high proportion of wind turbines. In traditional power systems, related research analyzes the sensitivity of loads to system frequency, determines the amount of load removal in each round of low-frequency load shedding after a power disturbance occurs, and formulates the load shedding tasks for each round. In order to achieve timely action of low-frequency load shedding, the impact of cluster temperature control loads is taken into account, thus realizing a frequency control strategy that takes into account the characteristics of the temperature control load's early response and cooperates with traditional low-frequency load shedding. However, in systems with a high proportion of wind turbines, due to the decrease in inertia, the system frequency will drop faster, and the accuracy of the traditional solution will be deviated. With the development of artificial intelligence (AI), neural network prediction models can be used to implement control strategies for under-frequency load shedding. However, these models are dependent on the system's operating mode, the type, location, and magnitude of disturbances. Furthermore, neural network training is complex and requires a large amount of data, making it unsuitable for new power systems with a high proportion of wind turbines. With the large-scale integration of wind turbines into power systems, numerous studies have conducted in-depth analyses of the impact of wind turbine connection on grid frequency. However, optimization of under-frequency load shedding methods after wind turbines are connected to the grid has not yet been conducted.

[0003] In summary, there is an urgent need in the prior art for a low-frequency load reduction method applicable to weak inertia systems containing a high proportion of wind turbines to solve such problems. Summary of the Invention

[0004] The technical problem to be solved by the present invention is to provide a frequency adaptive protection method for a weak inertia supported power grid, which can achieve rapid frequency recovery and avoid over-loading, thereby reducing the load shedding amount as much as possible.

[0005] A frequency adaptive protection method for a weak inertia support power grid, characterized by comprising the following steps:

[0006] Step 1: Calculate power shortage

[0007] When a power shortage occurs in the system, the total power shortage of the system includes the power shortage borne by the generator set and the active power shortage borne by the load node caused by the voltage deviation.

[0008]

[0009] Where ΔP is the total power shortage of the system; ΔP g,i is the power shortage borne by the i-th generator in the system; ΔP L,i is the power shortage borne by the i-th load node in the system; Ω g is the set of generators that are still running when there is a power shortage in the system; Ω L is the set of load nodes of the system;

[0010] Step 2: Adaptive low-frequency load shedding basic control level

[0011] The basic control level is set to two rounds. In the first round, the load shedding amount of low frequency load shedding is 0.9*ΔP according to the preset action frequency value and action delay time; in the second round, the load shedding amount of low frequency load shedding is 0.1*ΔP according to the preset action frequency value and action delay time;

[0012] Step 3: Adaptive low-frequency load reduction correction control level

[0013] The correction control level is set to two rounds. In the first round, the load shedding amount of low frequency load reduction is 0.05*ΔP according to the preset action frequency value and action delay time; in the second round, the load shedding amount of low frequency load reduction is 0.1*ΔP according to the preset action frequency value and action delay time.

[0014] Furthermore, in step 1, the power shortage borne by the generator set is calculated by the frequency change rate and moment of inertia at the time of the generator disturbance. Ignoring the electrical damping and mechanical damping of the generator, the rotor motion equation of the generator is:

[0015]

[0016] Where M, ω, and P m 、P e and ΔP g They are the generator's moment of inertia, rotor angular velocity, mechanical power, electromagnetic power and unbalanced power on the shaft, and t represents the differential time;

[0017] Under the per-unit value, there is ω i =f i ,ω i is the speed of the i-th generator, f is the bus frequency of the i-th generator, then the unbalanced power of the system can be expressed as:

[0018]

[0019] Where M eq and f coi are the equivalent inertia and inertia center frequency of the system, Mi is the moment of inertia of the i-th generator;

[0020] When a power disturbance occurs in the system, due to voltage deviation, the power shortage borne by the load can be calculated from the voltage change at the load node. The static characteristic of the load is:

[0021]

[0022] Where, P L,i is the real-time active power value of the load, P L0,i is the active power value of the load at rated frequency and initial voltage, a p,i 、b p,i and c p,i are the constant impedance, constant current and constant power ratios in the load respectively; V i0 and V i are the voltages at the load node i before and after the disturbance, respectively; k p,i is the active frequency influencing factor; f i is the bus frequency of the i-th generator, f i0 is the rated frequency of the i-th generator.

[0023] Furthermore, in step 2, the first round of action frequency value of the basic control level is set to 49.5 Hz, and the action delay is 0.2 s; the second round of action frequency value is set to 49.3 Hz, and the action delay is 0.2 s.

[0024] Furthermore, in step three, the first round of action frequency value of the correction control stage is set to 49.8 Hz, and the action delay is 15 s; the second round of action frequency value is set to 49.5 Hz, and the action delay time is 20 s.

[0025] The above-mentioned design scheme can bring the following beneficial effects: The proposed method for adaptive frequency protection of a weak-inertia-supported power grid takes into account the power shortfall borne by the generator and the power shortfall borne by the load due to voltage fluctuations during power shortfall estimation, resulting in a calculated power shortfall close to the actual power shortfall in the system. Furthermore, the proposed new adaptive low-frequency load reduction method can achieve rapid frequency recovery and stability in weak-inertia systems with a high proportion of wind turbines, effectively preventing the occurrence of secondary power shortfall events in wind turbines and improving the frequency stability of power system operation.

[0026] Compared with the prior art, the present invention has the following characteristics:

[0027] The present invention adds the active power shortage of the load caused by the voltage deviation to the estimation of the power shortage, so that the estimated power shortage is very close to the actual power shortage, reducing the error caused by the calculation. At the same time, in the low-frequency load reduction method, the basic control level is set to two rounds, rather than the multi-round and successive load shedding method in the traditional low-frequency load reduction, which ensures the rapid recovery of the frequency. At the same time, considering the frequency hovering caused by the under-cut load and the secondary power shortage accident caused by the triggering of the low-frequency protection of the wind turbine when the frequency is too low, a correction control level is set. Therefore, the setting of the correction control level is suitable for low-inertia power systems with a high proportion of wind turbines, which greatly improves the applicability of the present invention while improving the frequency stability of the power system. BRIEF DESCRIPTION OF THE DRAWINGS

[0028] Figure 1 This is the IEEE-39 node system topology diagram;

[0029] Figure 2 This is a comparison chart of load shedding frequencies between the solution of the present invention and the traditional solution according to scenario A in the embodiment;

[0030] Figure 3 This is a comparison chart of load shedding frequencies between the solution of the present invention and the traditional solution according to scenario B in the embodiment;

[0031] Figure 4 1. A comparison chart of load shedding frequencies between the solution of the present invention and the traditional solution according to scenario C in the embodiment;

[0032] Figure 5 2. A comparison chart of load shedding frequencies between the solution of the present invention and the traditional solution according to scenario D in the embodiment;

[0033] Figure 6 2. A comparison chart of load shedding frequencies between the solution of the present invention and the traditional solution according to scenario E in the embodiment;

[0034] Figure 7 This is a diagram showing the load shedding effect of the traditional scheme under different wind power penetration rates;

[0035] Figure 8 Flowchart of the frequency adaptive protection method for weak inertia supported power grid. DETAILED DESCRIPTION

[0036] To make the objects, features, and advantages of the present invention more apparent and understandable, the technical solutions of the present invention are described clearly and completely below in conjunction with the accompanying drawings in accordance with the embodiments of the present invention. Obviously, the present invention is not limited to the following embodiments, and specific implementation methods can be determined based on the technical solutions of the present invention and actual conditions. To avoid obscuring the essence of the present invention, well-known methods, processes, procedures, components, and circuits are not described in detail.

[0037] The present invention proposes a method for adaptive frequency protection of a weak inertia-supported power grid. This method calculates the power shortfall borne by the generator and the power shortfall shared by the load separately, thereby determining the magnitude of the power disturbance suffered by the system. Furthermore, a new low-frequency load shedding scheme is proposed based on the traditional "successive approximation" method used for low-frequency load shedding. The low-frequency load shedding method proposed in this invention comprises a basic control stage and a correction control stage. The basic control stage is primarily used to ensure rapid frequency recovery, while the correction control stage is primarily used to prevent secondary frequency shortfall accidents from occurring during the low-frequency load shedding process, effectively improving the stability of the system frequency. Figure 8 , the calculation process includes the following steps:

[0038] Step 1: Calculation of power shortage

[0039] When a power shortage occurs in the system, the power shortage can be mainly divided into two parts: (1) the generator sets in the system bear part of the power shortage; (2) the active power shortage borne by the load nodes caused by voltage deviation;

[0040]

[0041] Where ΔP is the total power shortage of the system; ΔP g,i is the power shortage borne by the i-th generator in the system; ΔP L,i is the power shortage borne by the i-th load node in the system; Ω g is the set of generators that are still running when there is a power shortage in the system; Ω L is the node set of the system load;

[0042] The power shortage borne by the generator set can be calculated by the frequency change rate and moment of inertia of the generator at the time of disturbance. Ignoring the electrical damping and mechanical damping of the generator, the rotor motion equation of the generator is:

[0043]

[0044] Where M, ω, and P m 、P e and ΔP g They are the generator's moment of inertia, rotor angular velocity, mechanical power, electromagnetic power and unbalanced power on the shaft, and t represents the differential time;

[0045] Under the per-unit value, there is ω i =f i ,ω i is the speed of the i-th generator, f is the bus frequency of the i-th generator, then the unbalanced power of the system can be expressed as:

[0046]

[0047] Where M eq and f coi are the equivalent inertia and inertia center frequency of the system, M i is the moment of inertia of the i-th generator;

[0048] When a power disturbance occurs in the system, due to voltage deviation, the power shortage borne by the load can be calculated from the voltage change at the load node. The static characteristic of the load is:

[0049]

[0050] Where, P L,i is the real-time active power value of the load, P L0,i is the active power value of the load at rated frequency and initial voltage, a p,i 、b p,i and c p,i are the constant impedance, constant current and constant power ratios in the load respectively; V i0 and V i are the voltages at the load node i before and after the disturbance, respectively; k p,i is the active frequency influencing factor; f i is the bus frequency of the i-th generator, f i0 is the rated frequency of the i-th generator;

[0051] Step 2: Adaptive low-frequency load shedding basic control level

[0052] The adaptive under-frequency load shedding method proposed in this invention is based on the traditional under-frequency load shedding scheme. It includes two parts: a basic control level and a correction control level. The basic control level has two rounds. The first round has an action frequency value set at 49.5Hz, an action delay of 0.2s, and removes a load of 90% of the power shortage to ensure rapid frequency recovery. The second round has an action frequency value set at 49.3Hz, an action delay of 0.2s, and removes a load of 10% of the estimated power shortage to continue to restore the system frequency.

[0053] Step 3: Adaptive low-frequency load reduction correction control level

[0054] The correction control stage also has two rounds. To prevent the occurrence of system frequency hovering due to system under-shedding, the first round of action frequency is set to 49.8Hz, the action delay is 15s, and the load shedding size is 5% of the estimated power shortage. To prevent the occurrence of secondary power shortage accidents caused by the wind turbine triggering underfrequency protection and being shedding, the second round of correction action frequency is set to 49.5Hz, the action delay time is 20s, and the load shedding size is 10% of the estimated power shortage. The specific parameters are shown in Table 1.

[0055] Table 1 Low frequency load reduction method of the present invention

[0056]

[0057] In summary, the design of the adaptive low-frequency load reduction strategy for the weak inertia system of high-ratio wind turbines is completed.

[0058] Below in conjunction with embodiment, the present invention is described in further detail:

[0059] by Figure 1 The simulation is conducted using a 10-generator, 39-bus system as an example. The system's generators have a rated active power of 6140 MW, and the total active power of the load is 6097 MW. The load model used is a static load consisting of a 40% constant power model and a 60% constant impedance model. G01-G10 represent the 10 generators numbered 1-10.

[0060] In order to verify the effectiveness of the load shedding method proposed in the present invention, the load shedding method is implemented in the following five scenarios.

[0061] Scenario A: This scenario has no wind turbines, and the proportion of wind turbines in the system is 0. It is set that generator sets G01 and G03 in the system are disconnected at 0.2s. The rated active powers of generator sets G01 and G03 are 1000MW and 650MW respectively. The disconnection of generator sets G01 and G03 is equivalent to a power disturbance of 26.87% in the system at this time.

[0062] Scenario B: In this scenario, generator units G07 and G08 are replaced by wind turbines with rated active power of 560 MW and 540 MW, respectively. The replaced wind turbines have the same active power as the original generator units. At this time, the wind turbine proportion is 17.92%. At 0.2s, generator units G01 and G03 are set to fall off, which is equivalent to a power shortage disturbance of 26.87% in the system.

[0063] Scenario C: In this scenario, generator sets G04, G07, and G08 are replaced by wind turbines. The active power provided by generator sets G04, G07, and G08 are 508MW, 560MW, and 540MW, respectively. The replaced wind turbines have the same active power as the original generator sets. The proportion of wind turbines in the system is 28.21%. The loss of generator sets G01 and G03 at 0.2s is equivalent to a power shortage disturbance of 26.87% in the system.

[0064] Scenario D: In this scenario, generator sets G04, G07, and G08 are replaced by wind turbines. The replaced wind turbines have the same active power as the original wind turbines. In this case, the wind turbines account for 28.21%. Generator set G09 is set to fall off at 0.2s. Its rated active power is 830MW, which is equivalent to a power shortage disturbance of 13.52% in the system.

[0065] Scenario E: In this scenario, generator sets G04, G07, and G08 are replaced by wind turbines. The replaced wind turbines have the same active power as the original generator sets. At this time, the wind turbines account for 28.21%. It is set that generator sets G03, G06, and G09 fall off at 0.2s. Their rated active powers are 650MW, 650MW, and 830MW respectively, which is equivalent to a power shortage disturbance of 34.69% in the system.

[0066] Power disturbances of varying magnitudes occurred in five scenarios with varying proportions of wind turbines. The system power deficit was calculated by collecting measurement information from the generators and load nodes at the moment of the disturbance. The power disturbances borne by the generators and load nodes were calculated and compared with the actual power deficit suffered by the system. The comparison results are shown in Table 2.

[0067] Table 2 Active power shortage calculation results

[0068]

[0069] Table 2 shows that there is still a certain error between the calculated power shortage and the actual value. The main sources of the error are the error in the generator frequency change rate measured at the time of the fault and the error in the load node voltage information. In addition, after the system is disturbed, the system power flow distribution changes, resulting in changes in line losses, which leads to errors between the measured and calculated data and the actual value. At the same time, comparing scenario A with the other four scenarios, it can be seen that due to the connection of wind turbines to the system, the error between the estimated system power shortage and the actual value after the system is disturbed becomes larger.

[0070] The low-frequency load shedding method of the present invention is applied to the five scenarios set above and compared with the traditional low-frequency load shedding scheme. The comparison diagram between the low-frequency load shedding method of the present invention and the traditional low-frequency load shedding scheme is shown in FIG. Figure 2-Figure 6 The comparison of load shedding effects under the five scenarios is shown in Table 3, and the load shedding amounts under the five scenarios are shown in Table 4.

[0071] Table 3 Comparison of load reduction effects

[0072]

[0073] Table 4 Comparison of load shedding

[0074]

[0075] Simulation results analysis

[0076] from Figures 2 to 6 As can be seen from Tables 3 and 4, the load shedding method proposed in the present invention can restore frequency stability faster than the traditional solution. Now, the above five scenarios are further analyzed in detail:

[0077] Scenario A: In this scenario, there is no wind turbine in the system. Figure 2 As can be seen from Table 3, the steady-state frequency offset value of the traditional load shedding method is greater than 0.2Hz, which does not meet the requirements for stable operation of the power system. This is because the inaccurate load shedding amount of the traditional load shedding method causes the steady-state frequency to ultimately remain at a higher frequency value, and the excessive load shedding rounds cause the traditional load shedding method to slowly decrease after reaching the highest frequency value. However, the highest and lowest frequencies in the load shedding process of the present invention are significantly better than those of the traditional load shedding method, and the steady-state frequency is reached about 20s after the fault occurs, meeting the system operation standard. As can be seen from Table 4, compared with the load shedding method of the present invention, the load shedding amount of the traditional load shedding method is much greater than the actual power shortage, while the load shedding method proposed by the present invention is closer to the actual power shortage.

[0078] Scenario B: In this scenario, 28.21% of the wind turbines are connected to the system. Figure 3 As can be seen from Table 3, the steady-state frequencies of both the present invention and the traditional load shedding method meet system operation requirements. However, the maximum and minimum frequencies during the load shedding process of the present invention are superior to those of the traditional method. Analysis of the frequency response curves shows that the steady-state frequencies achieved by the two methods are similar, but the frequency recovery curve of the present invention is smoother. As can be seen from Table 4, the load shedding capacity of the present invention is much smaller than that of the traditional method, and is closer to the actual value.

[0079] Scenario C: In this scenario, 28.21% of the wind turbines are connected to the system. Figure 4 As shown in Table 3, the steady-state frequencies of both the present invention and the traditional load shedding method meet the system operation requirements. The maximum and minimum frequencies during the load shedding process of the present invention are both superior to those of the traditional method. Furthermore, compared to the traditional load shedding method's steady-state frequency of 50.10 Hz, the present invention's steady-state frequency of 49.99 Hz is closer to the system's rated frequency of 50 Hz. Table 4 also shows that the present invention achieves a lower load shedding capacity than the traditional method, closer to the actual value. Furthermore, the frequency recovery curve is more stable, while the traditional method's frequency recovery curve exhibits more pronounced oscillation.

[0080] Scenario D: In this scenario, 28.21% of the wind turbines are connected to the system. Figure 5As shown in Table 3, the steady-state frequencies of both the present invention and the traditional load shedding method meet the system's operational requirements. Compared to the traditional load shedding method's steady-state frequency of 49.80 Hz, the present invention's steady-state frequency of 49.95 Hz is closer to the system's rated frequency. Both the maximum and minimum frequencies during the present load shedding process are superior to those of the traditional load shedding method, which also shows a trend of continued decline. Table 4 shows that in this scenario, the system's actual power shortfall is 830 MW. The present invention removes 814 MW of load, while the traditional load shedding method removes 701 MW. The traditional load shedding method removes far less load than the actual load, which is the primary reason for its lower steady-state frequency.

[0081] Scenario E: In this scenario, 28.21% of the wind turbines are connected to the system. Figure 6 As can be seen from Table 3, the steady-state frequencies of the present invention and the traditional scheme are very close, but the load reduction method proposed in the present invention has a faster frequency recovery speed and its performance is significantly better than that of the traditional scheme. From Table 4, it can be seen that the actual power shortage of the system at this time is 2210MW. The method proposed in the present invention removes a load of 2030MW, while the traditional scheme removes a load of 1859MW. The load shedding amount of the traditional scheme is inaccurate and far from sufficient, which leads to its lower steady-state frequency.

[0082] The analysis of the five scenarios above demonstrates that the load shedding method proposed in this invention outperforms conventional solutions in terms of frequency recovery speed, maximum frequency, and minimum frequency during frequency response. The low-frequency load shedding method proposed in this invention achieves excellent control results in systems with varying wind turbine proportions and power disturbances of varying magnitudes.

[0083] At the same time from Figure 7 It can also be seen that as the proportion of wind turbines in the system increases, the lowest frequency in the frequency response process of the system's low-frequency load reduction becomes lower, the highest frequency becomes higher, and the frequency oscillation becomes more obvious and more unstable. This shows that as the proportion of wind turbines increases, it is more difficult to keep the system frequency stable. Therefore, in a weak inertia system with a high proportion of wind turbines, it is necessary to be able to act faster after a disturbance occurs in the system, to more accurately cut off the load of the same value as the actual power shortage of the system, and to restore the frequency to the rated value faster. Simulation experiments have proved that the adaptive low-frequency load reduction control strategy proposed in the present invention can achieve more accurate load shedding and faster frequency stability recovery. Therefore, the present invention solves the frequency stability problem in weak inertia systems by proposing a new low-frequency load reduction method.

Claims

1. A frequency adaptive protection method for a weak inertia supported power grid, characterized in that: The steps include: Step 1: Calculate power shortage When a power shortage occurs in the system, the total power shortage of the system includes the power shortage borne by the generator set and the active power shortage borne by the load node caused by the voltage deviation. Where ΔP is the total power shortage of the system; ΔP g,i is the power shortage borne by the i-th generator in the system; ΔP L,i is the power shortage borne by the i-th load node in the system; Ω g is the set of generators that are still running when there is a power shortage in the system; Ω L is the set of load nodes of the system; Step 2: Adaptive low-frequency load shedding basic control level The basic control level is set to two rounds. In the first round, the load shedding amount of low frequency load shedding is 0.9*ΔP according to the preset action frequency value and action delay time; in the second round, the load shedding amount of low frequency load shedding is 0.1*ΔP according to the preset action frequency value and action delay time; Step 3: Adaptive low-frequency load reduction correction control level The correction control level is set to two rounds. In the first round, the load shedding amount of low frequency load shedding is 0.05*ΔP according to the preset action frequency value and action delay time; in the second round, the load shedding amount of low frequency load shedding is 0.1*ΔP according to the preset action frequency value and action delay time; In step 1, the power deficit borne by the generator set is calculated by the frequency change rate and moment of inertia at the time of the generator disturbance. Ignoring the electrical damping and mechanical damping of the generator, the rotor motion equation of the generator is: Where M, ω, and P m 、P e and ΔP g They are the generator's moment of inertia, rotor angular velocity, mechanical power, electromagnetic power and unbalanced power on the shaft, and t represents the differential time; Under the per-unit value, there is ω i =f i ,ω i is the speed of the i-th generator, f i is the bus frequency of the i-th generator, then the unbalanced power of the system is expressed as: Where M eq and f coi are the equivalent inertia and inertia center frequency of the system, M i is the moment of inertia of the i-th generator; When a power disturbance occurs in the system, due to voltage deviation, the power shortage borne by the load can be calculated from the voltage change at the load node. The static characteristic of the load is: Where, P L,i is the real-time active power value of the load, P L0,i is the active power value of the load at rated frequency and initial voltage, a p,i 、b p,i and c p,i are the constant impedance, constant current and constant power ratios in the load respectively; V i0 and V i are the voltages at the load node i before and after the disturbance, respectively; k p,i is the active frequency influencing factor; f i is the bus frequency of the i-th generator, f i0 is the rated frequency of the i-th generator.

2. The method for frequency adaptive protection of a weak inertia support power grid according to claim 1, characterized in that: In step 2, the first-round action frequency value of the basic control level is set to 49.5 Hz, and the action delay is 0.2 s; the second-round action frequency value is set to 49.3 Hz, and the action delay is 0.2 s.

3. The method for frequency adaptive protection of a weak inertia support power grid according to claim 1, characterized in that: In step 3, the first round of action frequency value of the correction control stage is set to 49.8 Hz, and the action delay is 15 s; the second round of action frequency value is set to 49.5 Hz, and the action delay time is 20 s.

Images