Active power control in renewable power plants for grid stability

Abstract

Various aspects of the invention relate to a method for controlling a renewable power plant connected to an electrical grid to reduce deviation of a measured frequency of the electrical grid from a target frequency. The method comprises determining a predicted power gradient over a prediction interval defined between a first time point and a second time point; and at a third time point during the prediction interval, controlling the power plant to output active power according to a minimum active power level if the measured frequency at the third time point is lower than the target frequency, and according to a maximum active power level if the measured frequency at the third time point is higher than the target frequency. The maximum active power level and the minimum active power level are based on the predicted power gradient.

Description

Technical Field

[0001] This invention relates to a method for controlling a renewable energy power plant, a renewable energy power plant controller, a wind turbine power plant, and more generally to a renewable energy power plant. Background Technology

[0002] Recently commissioned renewable power plants, more specifically wind power plants, are expected to operate and adapt to a variety of different conditions when connected to the grid. A wind power plant typically comprises multiple wind turbine generators and is also known as a wind park or wind farm. The regulation and general operation of the power plant are controlled by the power plant control system or controller (PPC), which enforces the operating restrictions and requirements set forth by the grid operator (e.g., transmission system operator (TSO) or distribution system operator (DSO)) or national or regional grid interconnection requirements (known as “grid specifications”).

[0003] Some grid operators implement wind power plant operation requirements related to grid stability. Grid stability is typically measured in frequency relation. Generally, if the grid frequency changes to be above or below set limits (e.g., frequency dead zone), the grid operator requires specific controls. As will be readily understood by those skilled in the art, the dead zone includes upper and lower limits around the nominal frequency.

[0004] Historically, within frequency dead zones, grid operators have not expected wind farms and other renewable energy plants to take measures to improve grid stability. However, the increasing prevalence of renewable energy, and its relatively unstable supply compared to traditional non-renewable energy sources, has led grid operators to impose stricter requirements on its operation. Some grid operators now require wind farms to operate to offset any deviations from the nominal frequency, even within the frequency dead zone. These measures help improve overall grid stability—eliminating deviations early can prevent more harmful ones. However, because historically, wind farms did not need to take any action within the dead zone, current control methods are not suitable for applications within the dead zone.

[0005] The purpose of this invention is to address one or more drawbacks associated with the prior art. Summary of the Invention

[0006] According to one aspect of the present invention, a method is provided for controlling a renewable power plant connected to a power grid to reduce the deviation of a measured frequency of the power grid from a target frequency. The method includes: determining a predicted power gradient over a predicted interval defined between a first time point and a second time point; and, at a third time point during the predicted interval, if the measured frequency at the third time point is lower than the target frequency, controlling the power plant to output active power based on a minimum active power level, and if the measured frequency at the third time point is higher than the target frequency, controlling the power plant to output active power based on a maximum active power level. The maximum and minimum active power levels are based on the predicted power gradient.

[0007] The beneficial outcome of implementing this method in renewable power plants is an improved contribution of the power plant to grid stability. This method is particularly suitable for use when frequency deviations are within the frequency dead zone because it is a direct control approach that does not require complex calculations or the complex operation of generators (such as wind turbines, photovoltaic cells, or batteries) or auxiliary equipment. In the long term, reducing deviations in the dead zone by implementing this method to ensure grid frequency stability provides overall stability and reduces the magnitude of severe frequency deviations.

[0008] Determining the predicted power gradient may include: receiving active power levels for a first time point and active power levels for a second time point; and determining the predicted power gradient as the gradient between the received active power levels. Alternatively, determining the predicted power gradient may include: receiving the predicted power gradient from the transmission system operator.

[0009] The active power level at the first time point may include the measured active power level. Alternatively, the active power level at the first time point may include the active power level received at a second time point for a previously predicted interval.

[0010] The active power level at the second time point can be received from a prediction source outside the wind power station.

[0011] The maximum active power level can be the value of the predicted power gradient at the third time point. The minimum active power level can also be the value of the predicted power gradient at the third time point.

[0012] The maximum active power level can be less than the predicted power gradient at the third time point. The minimum active power level can be greater than the predicted power gradient at the third time point.

[0013] Controlling the output active power of a power plant may include: receiving a measured frequency at a third time point; comparing the measured frequency with a target frequency; and determining whether the measured frequency is higher than, lower than, or equal to the target frequency.

[0014] If it is determined that the measured frequency is equal to the target frequency, controlling the power plant's output active power may include controlling the power plant to maintain its current operation.

[0015] If it is determined that the measured frequency at the third time point is higher than the target frequency, controlling the power plant to output active power may include: comparing the active power output level of the power plant at the third time point with the value of the predicted power gradient at the third time point; and if it is determined that the active power output level is higher than the predicted power gradient, transmitting the maximum active power level as the active power setpoint to the power plant, or if it is determined that the active power output level is equal to or lower than the predicted power gradient, controlling the power plant to maintain its current operation.

[0016] Controlling the active power output of a power plant may include limiting the active power output level of the power plant to the active power setpoint.

[0017] If it is determined that the measured frequency at the third time point is lower than the target frequency, controlling the power plant to output active power may include: comparing the active power output level of the power plant at the third time point with the predicted power gradient; and if it is determined that the current active power output level is lower than the predicted power gradient, transmitting the minimum active power level as the active power setpoint to the power plant; or, if it is determined that the active power output level is equal to or higher than the predicted power gradient, controlling the power plant to maintain its current operation.

[0018] Controlling the active power output of a power plant may include: operating compensation equipment to provide additional active power. Controlling the active power output of a power plant may also include: overloading one or more generators at the power plant to provide additional active power, thereby raising the active power output level of the power plant to the active power setpoint. Controlling the active power output of a power plant may further include: using any suitable technology that allows for increased active power output from one or more generators.

[0019] The target frequency can be the nominal frequency of the power grid.

[0020] According to another aspect of the invention, a power plant controller is provided, which is configured to implement the method described above.

[0021] Within the scope of this application, it is expressly intended that the various aspects, embodiments, examples, and alternatives listed in the foregoing paragraphs, claims, and / or the following description and drawings, especially their various features, can be implemented independently or in any combination. That is, all embodiments and / or features of any embodiment can be combined in any manner and / or combination unless these features are incompatible. The applicant reserves the right to amend any originally filed claim or accordingly file any new claim, including modifying any originally filed claim to be subordinate to any other claim and / or incorporating any feature of any other claim (even though these claims were not initially claimed). Attached Figure Description

[0022] One or more embodiments of the invention will now be described by way of example only with reference to the accompanying drawings, in which:

[0023] Figure 1 This is a schematic diagram of a wind power station, its connection to the power grid, and its control system.

[0024] Figure 2 This is an operation according to an embodiment of the present invention. Figure 1 The flowchart of the control system operation is used to control the wind power station to reduce the deviation of the grid frequency from the nominal frequency;

[0025] Figure 3 This illustrates one embodiment of the invention, in Figure 2 A flowchart illustrating how the steps for controlling a power plant are performed; and

[0026] Figure 4a and 4b They respectively showed the situation without implementation Figure 2 and 3 In the case of the method and in the implementation Figure 2 and 3 Two schematic diagrams illustrating the method, showing exemplary active power output of a wind power station and the corresponding frequency level of the power grid at a prediction interval. Detailed Implementation

[0027] Figure 1The illustration depicts a typical architecture in which a wind power plant (WPP) is connected to the main transmission grid as part of a wider power grid. As those skilled in the art will understand, a WPP includes at least one wind turbine generator (WTG) and is also referred to as a wind farm or wind farm. A WTG is commonly referred to as a wind turbine. The example shown is merely representative, and those skilled in the art will understand that other specific architectures are possible, relating to wind power plants, other renewable energy power plants, wind turbine generators, and other renewable energy power sources, such as solar power plants incorporating photovoltaic cell units and power plants incorporating batteries. Therefore, the present invention also generally relates to renewable energy power plants and renewable energy generators, and not specifically to the wind power plant and generator shown in the figure. Furthermore, those skilled in the art will understand that the methods, systems, and techniques described below are applicable to many different grid configurations. Moreover, the components of wind power plants and grids are conventional and will therefore be familiar to those skilled in the art. It is anticipated that, in addition to Figure 1 The components shown and described, or as Figure 1 The components shown and described may be substituted, and other known components may be added. Such modifications will be within the capabilities of those skilled in the art.

[0028] Figure 1 The diagram shows a power grid 10 containing WPP 12. WPP 12 includes multiple WTGs 14. Each of the multiple WTGs 14 converts wind energy into electrical energy, which is then transmitted as active power from the WTG 14 to the main transmission network (also known as the power grid or main grid 16) for distribution.

[0029] WPP 12 also includes compensation devices 17, such as synchronous condensers and / or static compensation devices (STATCOM) and / or other generator types, such as battery or photovoltaic cell units (in hybrid power plants), configured to provide active and / or reactive power support as needed. In some WPP 12s operating according to embodiments of the invention, compensation devices 17 may be absent.

[0030] Each WTG 14 is associated with a corresponding WTG controller 15. In some examples, a group of WTGs may share a single semi-centralized WTG controller, thus the number of WTG controllers is less than the number of WTGs. As will be understood by those skilled in the art, the WTG controller 15 can be considered a computer system capable of operating the WTG 14 in the manner described herein, and may include multiple modules controlling individual components of the WTG or just a single controller. The computer system of the WTG controller 15 may be operated according to software downloaded via a communication network or programmed thereon from a computer-readable storage medium.

[0031] During normal operation of WPP 12, WTG controller 15 operates to fulfill active and reactive current requests or setpoints received from power plant controller (PPC) 22. In abnormal conditions, WTG controller 15 operates to meet predetermined network requirements and also takes action to protect WTG 14 from any potentially harmful conditions. Normal operation can be defined as the situation where the voltage and frequency levels of the main grid 16 are within acceptable levels (typically a voltage or frequency dead zone around the nominal voltage or frequency value). Abnormal conditions occur when the voltage and / or frequency are not within acceptable levels. The methods described herein are generally relevant to the normal operation of a power plant.

[0032] WPP 12 is connected to the main grid 16 via connection network 18. WPP 12 and main grid 16 are connected at interconnection point (PoI) 20, which is the interface between WPP 12 and main grid 16.

[0033] PPC 22 is connected to the power grid at the point of measurement (PoM) 24 and directly to WPP 12. The role of PPC 22 is to act as a command and control interface between WPP 12 and the power grid 16, and more specifically, between WPP 12 and the power grid operator 26 (as described above, which may be a TSO or DSO). PPC 22 is a suitable computer system for executing the aforementioned controls and commands, and therefore includes a processing module 28, a connection module 30, a memory module 32, and a sensing module 34. PPC 22 can also receive information about the power grid 16 and / or local bus, substations, and networks from the energy management system (not shown). WPP 12 is able to change its power output or current output according to commands received from PPC 22.

[0034] PPC 22 is configured to measure various parameters, including the representative power output that WPP 12 will provide to the main grid 16 at PoI 20. Since PoM 24 is not located at PoI 20, the measured parameters are only representative, as losses in the lines between PoM 24 and PoI 20, and between PoM 24 and PPC 22, can affect the measurements. These losses can be appropriately compensated for to ensure accurate measurements.

[0035] PPC 22 transmits control commands to WPP 12 in an appropriate manner. It is important to note that... Figure 1This is a schematic diagram, therefore the method of transmitting control commands to the WPP 12 is not explicitly described. However, it should be understood that suitable cabling can be provided to interconnect the PPC 22 and WPP 12. The interconnection can be a direct connection or a "point-to-point" connection, or it can be part of a local area network (LAN) operating according to a suitable protocol (e.g., CAN bus or Ethernet). Furthermore, it should be understood that control commands can be transmitted wirelessly via a suitable wireless network, such as WiFi, instead of using cabling. TM Or ZigBee TM (They run according to the IEEE 802.11 and 802.15.4 standards, respectively).

[0036] Grid operator 26 operates to coordinate the supply and demand of electricity / power from grid 10, typically to avoid unnecessarily high levels of deviation in frequency and voltage from the main grid 16. To coordinate supply and demand, grid operator 26 balances available supply and demand by generating forecasts of available supply at a future point in time. These forecasts are generated for each power station connected to the main grid 16; therefore, in Figure 1 In the power grid 10, the grid operator 26 generates at least a forecast of the expected output of WPP 12 at a future point in time. In generating the supply forecast, the grid operator 26 can identify potential under- or over-generation and take appropriate action. The forecast is generated for a predetermined future point in time at the end of a predetermined time period (hereinafter referred to as the forecast interval). Therefore, the forecast interval can be considered as defined between a first point in time marking the start of the forecast interval and a second point in time marking the end of the forecast interval. For the end of the forecast interval (i.e., at the second point in time), the forecast output is generated.

[0037] Figure 1 The schematic diagram should only be used as a representative example of wind power station 12. Alternative configurations for wind power stations are known and are expected to include, in addition to, [other configurations]. Figure 1 Other than or as components shown and described Figure 1 The components shown and described can be substituted, and other known components can be added. Such modifications will be within the capabilities of those skilled in the art.

[0038] During WPP 12 operation, the frequency level of the main grid 16 may deviate from its nominal frequency level. The nominal frequency level of a national grid is typically 50 Hz or 60 Hz, but the specific value may vary by country. Within the frequency dead zone, frequency deviations relative to the nominal frequency are generally acceptable. The frequency dead zone can be an area on either side of the nominal frequency. For example, in a 50 Hz network, the frequency dead zone could be from 49.85 Hz to 50.15 Hz, or 15 Hz on either side of the nominal frequency. If the grid frequency deviates beyond the dead zone, power plants connected to both the grid and the network have specific protection devices and / or control mechanisms to ensure that the frequency deviation does not damage the grid and thus allows it to quickly return to the dead zone.

[0039] As stated above, efforts to reduce any deviation from the nominal frequency are useful, even within the dead zone of the power grid. WPPs that take action to reduce deviation contribute to the overall stability of the power grid they are connected to, reducing the likelihood of future grid instability and more severe deviations.

[0040] Figure 2 Method 100 is a method suitable for controlling WPP 12 to reduce the deviation relative to the nominal frequency within the frequency dead zone of the main grid 16, with the aim of improving or at least contributing to the stability of the main grid 16.

[0041] Overall, Figure 2 Method 100 is used to ensure that when the frequency of the main grid 16 (measured or received by PPC 22) differs from the nominal frequency, PPC 22 takes action to control WPP 12, thereby contributing to the stability of the main grid 16 rather than causing it to become unstable. WPP 12 is able to contribute to the stability of the main grid 16 by providing active power to the main grid 16 at a specific level according to the end-user's demand on the main grid 16.

[0042] Excessive or insufficient active power generated by WPP 12 compared to the end-user demand on the main grid 16 can cause changes in the frequency level of the main grid 16. If the frequency level of the main grid 16 is higher than the nominal frequency level, and WPP 12 is generating more active power than required (i.e., over-generating), then WPP 12 is further increasing the frequency level, thereby disrupting the stability of the main grid 16 by causing an increase in the deviation from the nominal frequency level. Similarly, if the frequency level of the main grid 16 is lower than the nominal frequency level, and WPP 12 is generating less active power than required (i.e., under-generating), then WPP 12 is further decreasing the frequency level, thereby disrupting the stability of the main grid 16 by causing an increase in the deviation from the nominal frequency level.

[0043] On the other hand, it is desirable for the main grid 16 to generate less power when the frequency level is above the nominal level and to generate more power when the frequency level is below the nominal level, because WPP 12 is thus considered to contribute to the stability of the main grid 16.

[0044] Figure 2 Method 100 is used to offset the instability effects caused by over- or under-generation when the frequency level is respectively above or below the nominal level. By employing actions by the grid operator 26 to balance available supply and end-user demand, over- or under-generation that leads to instability can be identified and / or prevented.

[0045] As mentioned above Figure 1 The grid operator 26 balances available supply and demand by generating forecasts of available supply to identify potential under- or over-generation and take appropriate action. Forecasts are made at regular forecast intervals for each power station connected to the main grid 16. The length of each forecast interval can be any time period. For example, the time period of each forecast interval can be selected from the following: 5 minutes; 10 minutes; 15 minutes; 30 minutes; or 1 hour. In other examples, the forecast interval can be of any length. Forecasts can be received from sources outside the grid operator 26 (e.g., a national forecasting service).

[0046] therefore, Figure 2 Method 100 begins with step 102, in which the predicted power gradient over the prediction interval is determined. As described above, the prediction interval is defined over a period of time between a first time point and a second time point, which respectively mark the start and end of the prediction. Therefore, the predicted power gradient is provided as the gradient or slope between the active power level at the first time point and the active power level at the second time point. It should be understood that the term "gradient" here is used to refer to the actual straight line or slope between two points, rather than a numerical value of the steepness of the straight line. Therefore, a point on the predicted power gradient at a specific time can be compared with a measured value at that time.

[0047] The predicted power gradient defines a threshold to indicate whether WPP 12 is generating too much or too little active power. In this respect, the predicted gradient between the start and end points of the prediction is used as an approximation of end-user demand and can therefore be used to indicate whether generation is excessive or insufficient. The gradient can be flat if the active power levels at the first and second time points are the same.

[0048] In step 102, determining the predicted power gradient may include: PPC 22 receiving the predicted power gradient directly from grid operator 26 or from elsewhere (e.g., an external forecasting service associated with PPC 22, WPP 12, or grid operator 26). Alternatively, determining the predicted power gradient may include: PPC 22 receiving active power levels at a first time point and a second time point, the active power level at the second time point being a predicted level received from a forecasting service, and calculating the predicted power gradient based on the received active power level. In another alternative, determining the predicted power gradient may include: PPC 22 receiving a predicted wind profile from which the power gradient can then be calculated.

[0049] The active power level at the first time point can be determined by measurements from WPP 12, or it can be determined as the reusable active power level at the end of the immediately preceding prediction interval. The active power level at the second time point is the active power level that WPP 12 expects to generate at the second time point of the prediction interval.

[0050] In some embodiments where the PPC 22 receives active power levels for first and second time points, the active power levels will be received before the first time point of the prediction interval so that a gradient can be generated before the start of the prediction interval. Alternatively, in other embodiments, one or both of the active power levels may be received at the start of the prediction interval so that a gradient can be generated as close as possible to the start of the prediction interval. Similarly, in embodiments where the grid operator 26 provides the predicted power gradient directly to the PPC 22, the gradient is provided before or at the first time point of the prediction interval.

[0051] Figure 2 The next step 104 of method 100 is performed during the prediction interval. It is conceivable that step 104 will be repeated at multiple regular intervals during the prediction interval. At least, this step will be performed at one point in the prediction interval. This point in time can be considered a third point in the prediction interval. In some embodiments, the third point in time can be the same as the first or second point in time.

[0052] In step 104, during the prediction interval, the active power output of WPP 12 is controlled based on either the minimum active power level or the maximum active power level. Which active power level to use for controlling WPP 12 is based on the measured frequency of the main grid 16 at that time. Specifically, if the measured frequency at the time point of step 104 is lower than the target frequency or a frequency threshold (which is typically the nominal frequency), then step 104 of method 100 includes controlling the active power output of WPP 12 based on the minimum active power level, i.e., at or above the minimum active power level. If the measured frequency is higher than the target frequency, then step 104 of method 100 includes controlling the active power output of WPP 12 based on the maximum active power level, i.e., at or below the maximum active power level.

[0053] The maximum and minimum active power levels are based on the predicted power gradient determined in the previous step 102, which is used as an active power threshold. In some embodiments, the maximum and minimum active power levels are the values ​​of the gradient at that point in time. In other embodiments, the maximum and minimum active power levels are based on the values ​​of the gradient at that point in time, but are proportionally smaller or larger than the gradient to provide a buffer.

[0054] Figure 3 It shows how to execute Figure 2 Example of control step 104 of method 100. Figure 3 The steps in the flowchart are performed during the prediction interval after step 102, which determines the predicted power gradient.

[0055] After determining the predicted power gradient, Figure 3 Two comparison steps, 108 and 110, are performed. In comparison step 108, the current active power output from WPP 12 is compared with the current value of the predicted power gradient. The current active power output and the current value of the predicted power gradient represent the values ​​of the active power output and the predicted power gradient at a certain time point (i.e., the third time point) during the prediction interval. In other words, at time point T during the prediction interval, the active power output P of WPP 12 at that time point T is compared. T The value of the predicted power gradient m at time point T T Comparisons can be made. Active power output can be measured directly by the PPC 22 or received in other ways.

[0056] In another comparison step 110, the current grid frequency level is compared with the target frequency level. In other words, at time point T, the frequency level f of the main grid 16 at time point T is compared. TThe frequency is compared with the target frequency (in this case, the target frequency is the nominal frequency f0). The frequency level of the main grid 16 can be measured directly by the PPC 22 or received by other means.

[0057] Two comparison steps, 108 and 110, are performed to determine whether the relevant parameter is above or below its relevant threshold or target. During the prediction interval, the comparison steps can be repeated at regular interval time points within each prediction interval. For example, the regular interval time points can be spaced out at intervals of 0.5s, 1s, 3s, 4s, 5s, 10s, 1 minute, or 5 minutes.

[0058] The next step 112 of the method is to determine whether the result of the comparison requires a change in the control of WPP 12. In step 112, three results cause the flowchart to return to comparison steps 108 and 110 without requiring any changes to the control of WPP 12, i.e., in response to the "yes" of step 112. Here, control over WPP 12 is provided when controlling WPP 12 to continue its operation without modification.

[0059] These three results are the result that WPP 12 is stabilizing the main grid 16 and / or that the frequency level therein is at the nominal frequency. From step 112, it can be seen that these three results are: (1) the frequency level of the main grid 16 is equal to the nominal frequency, i.e., f T =f0; (2) The active power output of WPP 12 is equal to or less than the gradient value at that time point, and the frequency level of the main grid 16 is greater than the nominal frequency, i.e., P T ≤m T And f T >f0; or (3) the active power output of WPP 12 is equal to or greater than the gradient level at that time point, and the frequency level of the main grid 16 is less than the nominal frequency, i.e., P T ≥m T And f T <f0。

[0060] If the comparison results do not meet these requirements, for example, if both the active power output of WPP 12 and the frequency level of the main grid 16 are higher than or lower than their respective thresholds / targets, i.e., P T >m T And f T >f0 or P T <m T And f T< f0, the answer in step 112 is "No". Among these results, WPP 12 is considered to cause instability in the main power grid 16, and it is necessary to change the active power output of WPP 12. Therefore, in step 114, an active power setpoint for WPP 12 is generated according to the gradient and output to WPP 12. The active power set value is set to m T , that is, the value of the gradient at time T. When considered in the context of step 104 of Figure 2 , the setting of the active power setpoint is essentially: setting the maximum active power output of WPP 12 for the case where both the active power output and the frequency level are higher than their respective thresholds / targets, and setting the minimum active power output of WPP 12 for the case where both the active power output and the frequency level are lower than their respective thresholds / targets.

[0061] It should be understood that Figure 3 the flowchart of T > m T and f T > f0 or P T < m T and f T < f0), to determine whether WPP 12 is disrupting the stability of the main power grid 16, so that answering "Yes" leads to setting the active power setpoint, and answering "No" leads to repeating the comparison step.

[0062] In some embodiments, the frequency comparison step 110 can precede the active power comparison step 108, and different comparisons are performed according to the result of the frequency comparison. In other embodiments, the active power comparison step 108 can precede the frequency comparison step 110, and different comparisons are performed according to the result of the active power comparison.

[0063] As described above, for the three results illustrated in step 112, namely f T = f0, P T ≤ m T and f T > f0 or P T ≥ m T and f T<f0, WPP 12 is actually operating at or above the minimum active power output level defined by the predicted power gradient and / or at or below the maximum active power output level defined by the predicted power gradient. In an alternative embodiment, the maximum active power level or the minimum active power level may be provided to the wind turbine as needed, regardless of the level of active power output, whereby WPP 12 operates within the required boundaries. In other words, Figure 2 the control step 104 of method 100 will be only a frequency-based comparison, and based on whether the frequency is above, below, or equal to the nominal frequency, respectively, the setpoint output to WPP 12 will be the maximum setpoint or the minimum setpoint or no setpoint.

[0064] To provide Figure 2 the context of method 100 and Figure 3 step 104, an implementation example of method 100 will now be described in conjunction with Figure 4a and 4b An implementation example of method 100 will now be described in conjunction with

[0065] Figure 4a The illustration shows two graphs 116, 118, respectively illustrating the typical active power output of WPP 12 during the prediction interval and the frequency level of the main power grid 16 within the frequency deadband. The upper graph 116 illustrates the active power output P normal between the start T1 of the prediction interval and the end T2 of the prediction interval, the active power level P1 at the start of the prediction interval, the active power level P2 at the end of the prediction interval, and the predicted power gradient m between the active power levels P1, P2. The lower graph 118 illustrates the frequency level f of the main power grid 16, the nominal frequency f0, and the upper and lower deadband values f upper 、f lower . Here, the nominal frequency f0 is used as the target frequency.

[0066] It should be noted that in Figure 4a , the active power level display of the normal output of WPP 12 is shown without implementing Figure 2 and Figure 3 the method, but for comparison purposes, the predicted power gradient is included. In other words, the information required to execute Figure 2 the method is shown, but the results are not shown. Instead, the results of the method are described.

[0067] Provided Figure 4b to illustrate the same prediction interval to which the Figure 2 and Figure 3 methods are applied. In Figure 4bThe document provides two charts, 120 and 118. The first chart, 120, illustrates the application of... Figure 2 and Figure 3 The active power output of WPP12 is shown in the following figure, and the second figure 118 illustrates the frequency level of the main grid 16 within the frequency dead zone during the prediction interval, as... Figure 4a As shown in Figure 120 above, the active power output P between the start of the prediction interval T1 and the end of the prediction interval T2 is illustrated. adjusted And the active power level P1 at the beginning of the prediction interval, the active power level P2 at the end of the prediction interval, and the prediction power gradient m between the active power levels P1 and P2, as... Figure 4a As shown. Figure 4a As shown in Table 118 below, the frequency level f, nominal frequency f0, and upper and lower dead zone values ​​f of the main power grid 16 are illustrated. upper f lower .

[0068] like Figure 4a and 4b As shown, the active power level P1 at the beginning of the prediction interval is known, and the active power level P2 is also provided. The predicted power gradient m has also been calculated between the two levels P1 and P2, as follows. Figure 4a and 4b As shown in both cases, step 102 has already been performed because the predicted power gradient has been determined.

[0069] During this interval, WPP 12 typically outputs active power P at a variable level. normal ,like Figure 4a As shown in Table 116 above. Meanwhile, during the forecast interval, the frequency level f of the main power grid 16 is within the dead zone limit f. upper f lower Internal variations. For clarity and ease of discussion, the prediction intervals are divided into five distinct periods, 141 to 145, each illustrating a different scenario. These periods will now be discussed in sequence.

[0070] In the first period 141, Figure 4a and Figure 4b At the start of the prediction interval, the active power level begins with the same value as the active power level P1. During the first period 141, the frequency level f of the main power grid 16 is equal to the nominal frequency level f0.

[0071] Therefore, when applications such as Figure 3 In step 104, as shown, at all times T during the period 141, the frequency level f is never higher or lower than the nominal frequency, but is equal to the nominal frequency, and therefore...Figure 3 The answer to step 112 is "yes". According to Figure 3 the process, steps 108 and 110 are repeatedly compared without taking any action, as Figure 4b shown in chart 120 of

[0072] In the second period 142, the frequency level f drops below the nominal frequency f0, and thus is between the nominal frequency f0 and the lower limit f lower of the dead zone. According to Figure 2 , the WPP 12 should thus be controlled according to the minimum active power level. As Figure 4a shown, without any control, the active power level P normal drops below the predicted power gradient m, and thus the WPP 12 increases the instability of the main power grid 16.

[0073] When applying the Figure 2 and Figure 3 method, the answer to step 112 is "no" because in the second period P T <m T and f T <f0. Therefore, control step 106 sets the active power setpoint to the value m T of the gradient during the entire period 142. Therefore, as Figure 4b shown, as long as the frequency value remains below the nominal frequency, the minimum active power can be identified and the WPP 12 is controlled according to the minimum setpoint.

[0074] To increase the active power level to the active power setpoint, the WPP 12 can use the compensation device 17 to increase its active power output to the setpoint and / or overload one or more of the WTGs 14 to provide an increased active power output in a short period of time.

[0075] Alternatively or additionally, other means can be used to increase the active power level output from the WPP 12. Any device that provides an additional active power generation or control method can be used here, and these devices are capable of releasing additional active power from the WTGs 14 of the WPP 12. For example, in some embodiments, overload or power boost of the output from the WPP 12 can be achieved by pitch-adjusting the wind turbine blades and controlling the generator and converter systems in the WTGs 14 to increase the active power level to the active power setpoint, thereby allowing more power to be generated regardless of the additional wear that will occur.

[0076] Alternatively, additional active power can be utilized by performing short-term forecasts within the forecast interval to determine the expected output during these shorter periods, and determining whether a power boost is needed based on the short-term forecasts and the predicted power gradient. In these embodiments, a separate forecasting system forecasts the power output for short periods within the forecast interval and compares the predicted short-term output with the actual output over the entire period to estimate whether the predicted power output will be achieved at the end of the period. If the predicted power output will be achieved and the output will be higher than the predicted power gradient, no power boost is needed. However, if the predicted power output is expected to be lower than the predicted power gradient, and / or if the predicted power output is expected to be higher than the predicted power gradient, but subsequent estimates based on measurements and meteorological data indicate that the predicted power gradient will not be met, a power boost for the output of WTG 14 for WPP 12 is performed. The above methods can be combined to assess turbine wear based on forecasts and, where permissible, to generate additional power based on control methods.

[0077] Figure 4a The third time period 143, as illustrated, begins with the active power level rising above the gradient again. Since the frequency level is still below the nominal frequency and the active power level is now above the value indicated by the gradient, the operating level of WPP12 is above the minimum level specified by the frequency level below the nominal frequency. Therefore, the answer to step 112 is "yes," and no further action is taken, and the comparison is performed again. Throughout the entire third time period 143, the answer is "yes," and therefore no action is taken during the entire time period 143. Figure 4a and 4b As shown, the active power level of WPP 12 is the same during this period 143.

[0078] In the fourth time period 144, the active power output level of WPP 12 remained above the gradient value, but the frequency level rose to the nominal frequency f0 and the dead zone upper limit f. upper The levels between. This means that both the active power level and the frequency level are above their respective thresholds, i.e., P in step 112. T >m T And f T >f0, therefore the answer to step 112 is "no", and thus the active setpoint is set to m. T This means setting the maximum active power setpoint for WPP 12 to limit its power generation.

[0079] In this situation, a lower active power level is achieved by applying limiting measures (e.g., pitch adjustment of the blades of WTG 14), or, in extreme cases, by utilizing standby loads or applying braking force to the rotor, limiting the active power output of WPP 12. Similar to the case of an increase in active power, any method used to appropriately limit the active power generation of WPP 12 can be used here. Figure 4b As shown, during the entire period 144, this limit reduces the active power output of WPP12 to the gradient value.

[0080] Finally, in the fifth period 145, the active power level again drops below the gradient, but the frequency level remains above the nominal frequency and therefore does not exceed the maximum active power. The answer to step 112 is "yes," and any comparisons performed during this period 145 do not require any action. WPP 12 is providing an active power level below the maximum active power level.

[0081] In some embodiments, the setting of either the maximum or minimum active power level can be implemented based solely on how WPP 12 can be controlled. Typically, when compensation equipment is unavailable, WPP 12 may be controlled only at the maximum active power level as needed, because WPP 12 can actively limit active power output, but it is difficult to increase active power output as needed without using compensation equipment.

[0082] It should be understood that various changes and modifications can be made to this invention without departing from the scope of this application.

Claims

1. A method for controlling a renewable power plant connected to an electrical grid to reduce deviation of a measured frequency of the electrical grid from a target frequency within a frequency deadband of the electrical grid, the method comprising: determining a predicted power gradient over a prediction interval defined between a first time point and a second time point; and controlling the power plant to output active power according to a minimum active power level at a third time point during the prediction interval if the measured frequency at the third time point is below the target frequency, according to a maximum active power level if the measured frequency at the third time point is above the target frequency, wherein: the maximum active power level and the minimum active power level are based on the predicted power gradient.

2. The method of claim 1, wherein, Determining the predicted power gradient comprises: receiving an active power level for the first time point and an active power level for the second time point; and determining the predicted power gradient as a gradient between the received active power levels.

3. The method of claim 2, wherein, The active power level for the first time point comprises a measured active power level.

4. The method of claim 2, wherein, The active power level for the first time point comprises an active power level received for a second time point of a previous prediction interval.

5. The method of claim 2, wherein, The active power level for the second time point is received from a prediction source external to the power plant.

6. The method of claim 1, wherein, Determining the predicted power gradient comprises: receiving the predicted power gradient from a power system operator.

7. The method of claim 1, wherein, A value of the maximum active power level or a value of the minimum active power level is a value of the predicted power gradient at the third time point.

8. The method of claim 1, wherein, The value of the maximum active power level is less than the value of the predicted power gradient at the third time point, or the value of the minimum active power level is greater than the value of the predicted power gradient at the third time point.

9. The method of claim 1, wherein, Controlling the power plant to output active power comprises: receiving the measured frequency at the third time point; comparing the measured frequency to the target frequency; determining whether the measured frequency is above, below, or equal to the target frequency; and controlling the power plant to maintain its current operation if it is determined that the measured frequency is equal to the target frequency.

10. The method of claim 9, wherein, If it is determined that the measured frequency at the third time point is above the target frequency, controlling the power plant to output active power comprises: comparing an active power output level of the power plant at the third time point to the value of the predicted power gradient at the third time point; transmitting the maximum active power level as an active power setpoint to the power plant if it is determined that the active power output level is above the predicted power gradient; and controlling the power plant to maintain its current operation if it is determined that the active power output level is equal to or below the predicted power gradient.

11. The method of claim 10, wherein, Controlling the power plant to output active power further comprises limiting the active power output level of the power plant to the active power setpoint.

12. The method of claim 9, wherein, If it is determined that the measured frequency at the third time point is below the target frequency, controlling the power plant to output active power comprises: comparing the active power output level of the power station at the third point in time with the predicted power gradient; and if it is determined that the current active power output level is below the predicted power gradient, transmitting the minimum active power level as an active power set point to the power station; and if it is determined that the active power output level is equal to or above the predicted power gradient, controlling the power station to maintain its current operation.

13. The method of claim 12, wherein, controlling the power station to output active power further comprises operating compensation equipment to provide additional active power, and / or overloading one or more generators of the power station to provide additional active power, to increase the active power output level of the power station to the active power set point.

14. The method of claim 1, wherein, the target frequency is a nominal frequency of the power grid.

15. A power station controller configured to implement the method of any one of claims 1 to 14.

Images