CONTROL METHOD FOR A SYSTEM COMPRISING A FREQUENCY CONVERTER CONNECTED TO A POWER GRID

Abstract

A method which is suitable for a system having a frequency converter and a generator, both of which are connected to a power grid, includes obtaining sub-synchronous components of the grid voltage and determining damping current set points according to the sub-synchronous components to compensate for sub-synchronous resonances of the grid. Damping current set points are determined by receiving the sub-synchronous components of the grid voltage and returning damping current set points as outputs. A variable damping gain is adjusted according to the sub-synchronous frequency of the grid, such that the required compensation level can be adapted to the frequency converter for damping sub-synchronous resonance of the grid.

Claims

1-10. (canceled)

11. A control method for controlling a system having a frequency converter connected to a power grid, the method comprising, obtaining sub-synchronous components of a grid voltage of the power grid; determining damping current set points based on a variable damping gain generated from the sub-synchronous components of the grid voltage; and outputting the damping current set points to compensate for sub-synchronous resonances of the power grid, wherein the variable damping gain is dynamically adjusted according to a sub-synchronous component of the power grid such that a required compensation level of the frequency converter is adapted for damping sub-synchronous resonance of the power grid.

12. The method according to claim 11, wherein the variable damping gain is adjusted based on at least one electric variable of the system reflecting a sub-synchronous resonance behavior of the power grid, said electric variable being selected from the sub-synchronous components of the grid voltage, a zero-sequence current at a point in the system where it is possible to determine the zero-sequence current, and a bus voltage of the frequency converter.

13. The method according to claim 12, further comprising determining at least a module of said electric variable to adjust the variable damping gain based on the electric variable, the variable damping gain being calculated based on said module by a regulator receiving said module as input and returning the damping gain.

14. The method according to claim 13, wherein the variable damping gain used for obtaining the damping current set points is limited by a maximum limit and a minimum limit, both of which are pre-established according to characteristics of the power grid and of a power generation farm where the frequency converter is located.

15. The method according to claim 13, wherein if it is determined that the module of an electric variable based on which the variable damping gain is calculated, or the derivative of said module, exceeds a predetermined safety threshold, determination of the damping current set points is stopped and the system is decoupled from the power grid.

16. The method according to claim 13, wherein if it is determined that the module of an electric variable based on which the variable damping gain is calculated, or the derivative of said module, exceeds a predetermined safety threshold for a predetermined safety time, determination of the damping current set points is stopped and the system is decoupled from the power grid.

17. An electrical power generation system comprising, a frequency converter connected to a power grid, a voltage detector for detecting a grid voltage of the power grid, a current detector for detecting a grid current of the power grid, and a central control unit for controlling the frequency converter, configured for implementing a method according to claim 13.

18. The method according to claim 12, wherein the variable damping gain comprises a first damping gain and a second damping gain, the method further comprising calculating the first damping gain based on a first electric variable reflecting the sub-synchronous component of the power grid and the second damping gain based on a second electric variable also reflecting the sub-synchronous frequency of the power grid, said first and second electric variables being selected from the sub-synchronous components of the grid voltage, the zero-sequence current, and a bus voltage of the frequency converter, and selecting a higher of the first damping gain and the second damping gain as the adjusted damping gain.

19. A method according to claim 18, wherein a first of the damping current set points is determined based on one of the sub-synchronous components, and a second damping current set point is determined based on another of the sub-synchronous components, the first damping gain being determined based on the sub-synchronous components of grid voltages, and the second damping gain being determined based on the zero-sequence current, the damping current set points being determined based on both the first damping gain and the second dampening gain.

20. The method according to claim 18, wherein the variable damping gain used for obtaining the damping current set points is limited by a maximum limit and a minimum limit, both of which are pre-established according to characteristics of the power grid and of a power generation farm where the frequency converter is located.

21. The method according to claim 12, wherein the variable damping gain used for obtaining the damping current set points is limited by a maximum limit and a minimum limit, both of which are pre-established according to characteristics of the power grid and of a power generation farm where the frequency converter is located.

22. The method according to claim 21, wherein if it is determined that the value of the variable damping gain reaches the maximum limit and maintains said maximum limit throughout a predetermined safety time, determination of the damping current set points is stopped and the system is decoupled from the power grid.

23. The method according to claim 12, wherein the variable damping gain used for obtaining the damping current set points is limited by a maximum limit and a minimum limit, both of which are pre-established according to a current capacity of the frequency converter.

24. The method according to claim 23, wherein if it is determined that the value of the variable damping gain reaches the maximum limit and maintains said maximum limit throughout a predetermined safety time, determination of the damping current set points is stopped and the system is decoupled from the power grid.

25. The method according to claim 12, wherein the variable damping gain used for obtaining the damping current set points is limited by a maximum limit and a minimum limit, both of which are pre-established according to characteristics of the power grid and of a power generation farm where the frequency converter is located and according to a current capacity of the frequency converter.

26. The method according to claim 25, wherein if it is determined that the value of the variable damping gain reaches the maximum limit and maintains said maximum limit throughout a predetermined safety time, determination of the damping current set points is stopped and the system is decoupled from the power grid.

27. An electrical power generation system comprising, a frequency converter connected to a power grid, a voltage detector for detecting a grid voltage of the power grid, a current detector for detecting a grid current of the power grid, and a central control unit for controlling the frequency converter, configured for implementing a method according to claim 11.

28. An electrical power generation system comprising, a frequency converter connected to a power grid, a voltage detector for detecting a grid voltage of the power grid, a current detector for detecting a grid current of the power grid, and a central control unit for controlling the frequency converter, configured for implementing a method according to claim 12.

Description

DESCRIPTION OF THE DRAWINGS

[0033] FIG. 1 shows a one-line circuit diagram of a power grid compensated with series capacitors.

[0034] FIG. 2 shows a one-line circuit diagram of a system comprising a frequency converter connected to a power grid, particularly for a wind power generation application based on a doubly-fed topology, where the method of the invention can be implemented.

[0035] FIG. 3 shows a block diagram depicting a mode of generating the sub-synchronous components of the grid voltage.

[0036] FIG. 4 shows a block diagram describing the sub-synchronous resonance damping loop of a frequency converter of the state of the art.

[0037] FIG. 5 shows a block diagram where the generation of the total current set points in the state of the art is depicted based on generating active and passive current set points.

[0038] FIG. 6 shows a block diagram where the generation of active and reactive current set points in the state of the art is depicted by means of an active power regulation loop and a reactive power regulation loop.

[0039] FIG. 7 shows an embodiment of the method of the invention based on a block diagram representation, in which the damping current set points are generated based on a damping gain from sub-synchronous components of the grid voltage.

[0040] FIG. 8 shows variation in damping gain by applying the method depicted in FIG. 7, depending on the number of wind turbines connected to the grid.

[0041] FIG. 9 shows an embodiment of the method of the invention based on a block diagram representation, in which the damping current set points are generated based on a damping gain from the zero-sequence current of the grid, the frequency converter or the stator of the turbine generator.

[0042] FIG. 10 shows the total active power generated by a wind farm in a sub-synchronous resonance event for different numbers of turbines on the farm coupled to the grid, seen from a system in which the method of the state of the art is applied to compensate for said sub-synchronous resonance.

[0043] FIG. 11 shows the total active power generated by a wind farm in a sub-synchronous resonance event for different numbers of turbines on the farm coupled to the grid, seen from a system in which an embodiment of the method of the invention is applied, in which the damping gain for determining damping current set points is calculated based on the sub-synchronous components of the grid voltage.

[0044] FIG. 12 shows the total active power generated by a wind farm in a sub-synchronous resonance event for different numbers of turbines on the farm coupled to the grid, seen from a system in which an embodiment of the method of the invention is applied, in which the damping gain for determining damping current set points is calculated based on the zero-sequence current of the grid, the frequency converter or the stator of the turbine generator.

DETAILED DISCLOSURE OF THE INVENTION

[0045] The description of the invention uses as a reference an energy generation application based on doubly-fed topology. A person skilled in the art will understand that the described invention can be applied to any application which includes at least one frequency converter 4 connected to the grid, even if it is not based on doubly-fed topology. Examples that can be cited would be energy generation or consumption applications in which all the energy flows through the frequency converter 4, i.e., full converter, HVDC applications for electrical power distribution or HVAC applications for electrical power distribution.

[0046] Doubly-fed topology is made up of a doubly-fed asynchronous generator in which the terminals of the stator are connected directly to the power grid, and in which the terminals of the rotor are connected to a frequency converter 4 which will in turn be connected to the power grid.

[0047] Patent document US20130176751A1 is incorporated by reference. The method of the invention is suitable for being implemented in electrical power generation systems of different applications as discussed above, such as the one shown in FIG. 2, for example. Said FIG. 2 shows an electrical power generation system 1000 with a turbine 900 including a doubly-fed generator 1, the stator of which is connected to the grid by means of a contactor 2 and a transformer 3. The transformer 3 adapts the output voltage of the stator to the grid voltage value. The rotor of the doubly-fed generator 1 is connected to a frequency converter 4 comprising a grid-side converter or rectifier 5 and a machine-side converter or inverter 6. The system 1000 further comprises a central control unit 10 for generating switching commands 11 for the switches of the rectifier 5 and for generating switching commands 12 for the switches of the inverter 6. In one embodiment, the inverter 6 and the rectifier 5 can include static switches of the IGBT type, with their opening and closing controlled by switching commands 11 and 12 generated by the central control unit 10 (by means of corresponding regulation algorithms).

[0048] Operation of the system 1000 is controlled from the central control unit 10, which processes the measurements taken through sensors installed in said system 1000 (of voltage and/or current) and executes programmed control algorithms according to said measurements for controlling the flow of power between the generator 1 and the grid. The final result of executing these algorithms is in the form of switching commands 11 and 12 for the switches comprised both in the rectifier 5 and in the inverter 6. Said switching commands 11 and 12 are calculated by means of modulation steps using pulse width modulation techniques for synthesizing reference voltages that must be applied at the output of the inverter 6 and rectifier 5 for controlling the currents of each based on the voltage of the AC stage. Pulse width modulation techniques are widely used in the art today, being able to choose between scale-based techniques or vector-based techniques. Scale-based modulation techniques are those using the comparison of carrier signals with modulating signals as a basis, for example, PWM (Pulse Width Modulation). Vector-based techniques are those which apply specific switching patterns or vectors during specific previously calculated times in the mentioned modulation steps, for example, SVPWM (Space Vector Pulse Width Modulation).

[0049] The method of the invention comprises at least the steps of identifying sub-synchronous components V.sub.xs and V.sub.ys of the grid voltage 38 based on detections of said grid voltage 38, and of determining damping current set points 40′ and 41′ according to said sub-synchronous components V.sub.xs and V.sub.ys, to compensate for resonance frequencies of the grid. Said steps are implemented in the central control unit 10. How sub-synchronous components V.sub.xs and V.sub.ys are obtained is something known in the field, and for that purpose Clarke and Parke transformations can be used, as mentioned in the state of the art, for example, depicted in FIG. 3 and explained in patent document US20130176751A1, or any other known mathematical method can be used.

[0050] In the method of the invention, the damping current set points 40′ and 41′ are determined by regulation means 45′ receiving the sub-synchronous components V.sub.xs and V.sub.ys of the grid voltage 38 and returning the damping current set points 40 and 41. Said regulation means 45′ comprise at least one regulator with at least one variable damping gain 46a′, and said variable damping gain 46a′ is adjusted dynamically according to the sub-synchronous frequency of the power grid at all times. Therefore, regulation means 45′ receive on one hand the sub-synchronous components V.sub.xs and V.sub.ys of the grid voltage 38 and the adjusted damping gain 46a′, and return damping current set points 40′ and 41′ as output. The required compensation level can thereby be adapted to the frequency converter 4 for damping sub-synchronous resonance of the grid, according to actual conditions of the grid to which the system is connected 1000 and of the farm to which the system 1000 belongs. Damping current set points 40′ and 41′ are generated in a sub-synchronous resonance damping loop 39′ such as the one shown by way of example in FIGS. 7 and 9, which replaces the sub-synchronous resonance damping loop 39 of patent document US20130176751A1. The damping set point regulation block 45 of the state of the art is replaced with regulation means 45′, which are suitable for receiving the variable adjusted damping gain 46a′, and a compensation regulation block 46′ is furthermore included for generating the variable damping gain 46a′, thereby a modification to the generation of damping current set points 40′ and 41′ to compensate for sub-synchronous frequencies of the grid being proposed.

[0051] The damping gain 46a′ is adjusted based on at least one electric variable of the system 1000 reflecting the sub-synchronous frequency of the power grid to which the system 1000 is connected, said electric variable being selected from sub-synchronous components V.sub.xs and V.sub.ys of the grid voltage 38, the zero-sequence current at a point of the system 1000 where it is possible to determine zero-sequence current (cases of the grid itself, the frequency converter 4 or the stator of the generator 1, if any), and the bus voltage V.sub.BUS of the frequency converter 4. Determination of the zero-sequence current depends on detections of current available in the system 1000, and also on the neutral operation of the system 1000 itself (of both the turbine 900 and frequency converter 4) because a neutral point connection is necessary for there to be a zero-sequence current. The requirements for there to be a zero-sequence current and the determination thereof is something that is already known in the state of the art, therefore it will not be explained in further detail.

[0052] To adjust a damping gain 46a′ based on one of said electric variables, the module of said electric variable is determined, the damping gain 46a′ being calculated based on said module preferably by means of a regulator which receives said module as input and returns the damping gain 46a′ and which is comprised in the compensation regulation block 46′. Said regulator can be a proportional regulator, a PI regulator or a PID regulator, the gain (or gains) of which is determined previously according to the grid to which the system 1000 is connected and to the farm it belongs (to the number of systems 1000 forming said farm). Instead of a regulator, the compensation regulation block 46 can comprise a look-up table, for example, or another known alternative, for establishing an output value (damping gain 46a′) according to the inputs.

[0053] In one embodiment shown by way of example in FIG. 7, the compensation regulation block 46′ receives sub-synchronous components V.sub.xs and V.sub.ys of the grid voltage 38 coming from the sub-synchronous resonance identification block 44 as input for calculating the variable parameter 46a′, and determines the module of said sub-synchronous components V.sub.xs and V.sub.ys, said module reaching the regulator of the compensation regulation block 46′. Regulator output is the damping gain 46a′. FIG. 8 shows the results of a simulation of a variation in damping gain 46a′ in a system 1000 during a time interval t by applying the regulation algorithm defined in FIG. 7 for one and the same value of the module of the sub-synchronous components V.sub.xs and V.sub.ys and for different numbers of turbines 900 connected to the grid at the time of the event (one turbine, evolution E1; five turbines, evolution E5; ten turbines, evolution E10; fifteen turbines, evolution E15; twenty turbines, evolution E20; twenty-five turbines, evolution E25; and thirty turbines, evolution E30). The number of systems 1000 connected to the grid is not known, and said FIG. 8 shows evolution of variation in damping gain 46a′ for different cases, demonstrating that the compensation regulation block 46′ provides a different damping gain 46a′ according to the systems 1000 connected to the grid, it being automatically adjusted to the actual need without having to know the total capacity of the farm to which the system 1000 belongs, to compensate for resonance at a specific time.

[0054] In another embodiment shown by way of example in FIG. 9, the compensation regulation block 46′ receives the zero-sequence current as input for calculating the damping gain 46a′. A zero-sequence component identification block 47′ receives the current measurements of a point of the system 1000 (of the grid current, stator or converter, for example), and calculates the corresponding zero-sequence current based on said measurements. Various techniques for calculating zero-sequence current are known in the art, and any of them can be used in this case. The zero-sequence current enters the compensation regulation block 46′, and said compensation regulation block 46′ calculates the value of the damping gain 46a′ for the damping set point regulation block 45′ according to said zero-sequence current. The compensation regulation block 46′ determines the module of said zero-sequence current, and said module reaches the regulator comprised in the compensation regulation block 46′ itself.

[0055] In another embodiment not shown in the drawings, the compensation regulation block 46 can calculate the damping gain 46a′ based on the bus voltage V.sub.BUS of the frequency converter, because, in the event of resonance, the bus voltage V.sub.BUS comprises oscillations. Said compensation regulation block 46′ determines the module of the bus voltage V.sub.BUS, and said module reaches the regulator comprised in the compensation regulation block 46′.

[0056] In other embodiments not shown in the drawings, the compensation regulation block 46′ calculates at least two compensation gains 46a′, based on one of the electric variables selected from the sub-synchronous components of the grid voltage, the zero-sequence current or bus voltage V.sub.BUS. The compensation regulation block 46′ determines the module of the corresponding electric variable and the derivative thereof, and said module and said derivative reach their respective regulator comprised in the compensation regulation block 46′ itself. Each regulator calculates a respective damping gain, and the highest of them is received by the damping set point regulation block 45′ for generating damping current set points 40′ and 41′. Evolution of sub-synchronous frequency can be estimated when considering the derivative, and a faster response can be provided (damping current set points 40′ and 41′ are anticipated).

[0057] In other embodiments not shown in the drawings, a plurality of damping gains 46a′ is calculated based on the sub-synchronous components V.sub.xs and V.sub.ys, zero-sequence components and/or bus voltage, each of them by means of the compensation regulation block 46′ thereof. The highest damping gain 46a′ calculated for generating damping current set points 40′ and 41′ is preferably applied in the damping set point regulation block 45′. Each compensation regulation block 46′ can be implemented with only the module of the corresponding electric variable, or with the module and derivative of said electric variable.

[0058] In other embodiments not shown in the drawings, the damping set point regulation block 45′ receives at least two damping gains 46a′: one for the sub-synchronous component V.sub.xs and the other one for the sub-synchronous component V.sub.ys. Therefore, one of the damping current set points 40′ and 41′ is determined by means of a first regulator of the regulation means 45′ based on one of the sub-synchronous components V.sub.xs and V.sub.ys with the corresponding damping gain 46a′, and the other damping current set point 40′ and 41′ is determined by means of a second regulator of the regulation means 45′ based on the other sub-synchronous component V.sub.xs and V.sub.ys with the other corresponding damping gain 46a′. One of the damping gains 46a′ is calculated based on the sub-synchronous components V.sub.xs and V.sub.ys, (preferably the one that is later associated with the sub-synchronous component V.sub.xs), and the other damping gain 46a′ is calculated based on the zero-sequence current (preferably the one that is later associated with the sub-synchronous component V.sub.xy). How to calculate a damping gain 46a′ based on sub-synchronous components V.sub.xs and V.sub.ys and zero-sequence current has previously been discussed.

[0059] For the method, in any of its embodiments, a maximum limit and minimum limit are preferably pre-established for each one of the variable parameters 46a′ has a maximum limit and a minimum limit. The limits are established according to the characteristics of the grid and the farm where the corresponding turbine 900 is installed. The maximum limit, for example, is established according to the gain required in the event that compensation must be done by means of a single wind turbine 900. The minimum limit, for example, is established according to the gain required in the event that compensation is done by means of all the wind turbines 900 on the farm.

[0060] The current capacity of the frequency converter 4 must also be considered to establish the limits because the frequency converter 4 must work in conditions in which a sub-synchronous compensation component must be added to the current set point of the regulation loops. This current affects the losses of the frequency converter 4, and therefore thermal performance, and it must be assured that safe working conditions are applied at all times.

[0061] In some embodiments, the method is furthermore suitable for stopping the determination of damping current set points 40′ and 41′ and for generating an alarm whereby disconnection of the corresponding turbine 900 from the grid is preferably caused, if it is determined that resonance cannot be compensated. Different techniques can be used to determine whether or not resonance can be compensated, such as: [0062] If it is determined that the module of the sub-synchronous components V.sub.xs and V.sub.ys, or the derivative of said module, exceeds a predetermined safety threshold, it is determined that resonance cannot be compensated. System shut-down would be instantaneous under these conditions. [0063] If it is determined that the module of the sub-synchronous components V.sub.xs and V.sub.ys, or the derivative of said module, exceeds a predetermined safety threshold throughout a predetermined safety time, it is determined that resonance cannot be compensated. Predetermined thresholds and safety time are set at a value which assures operation of both the mechanical and electrical components of the turbine 900 in safe conditions, and they also depend on the existence and adjustment of additional protections of the wind farm. For example, it could be adjusted for a case in which the sub-synchronous component V.sub.xs, V.sub.ys of the grid voltage 38 exceeds 8% of the rated voltage value for 10 seconds or the derivative is positive for 250 ms. [0064] If it is determined that the value of the damping gain 46a′ reaches its maximum limit and maintains said maximum limit throughout a predetermined safety time, it is determined that resonance cannot be compensated. The maximum limit could be set, for example, at 50, which allows compensating for resonance with 10% of the turbines 900 on the farm coupled to the grid.

[0065] In summary, any of the embodiments of the proposed method allows changing compensation of sub-synchronous components V.sub.xs and V.sub.ys through the compensation regulation block 46′, such that the greater the module (and/or derivative) of the sub-synchronous components V.sub.xs and V.sub.ys of the grid voltages 38, the greater the compensation. In the event that all the wind turbines 900 on a farm are coupled, small compensation of each of them will be enough to compensate for resonance. In the opposite case in which few turbines 900 are coupled, a greater compensation component will be required of them. The regulation system reaches a balance in which compensation is distributed among the available wind turbines 900 without having to know the power generated by each of them and the total wind farm power.

[0066] FIG. 10 shows the performance of a wind generation application based on doubly-fed topology controlled by a frequency converter 4, the operation of which is controlled by the regulation algorithm with resonance compensation considering the total available farm capacity (without implementing the invention). FIG. 10 shows the total power P generated by a wind farm in a sub-synchronous resonance event during a specific time interval t1 and for a specific number of turbines 900 coupled to the grid in each case, in which resonance is only compensated and stabilized for a minimum number of turbines 900 coupled to the grid. Specifically, the results are shown for the following numbers of turbines 900 coupled to the grid: 20 (total power P20), 21 (total power P21), 22 (total power P22), 23 (total power P23), 24 (total power P24) and 25 (total power P25).

[0067] FIG. 11 shows the performance of the wind farm with compensation with variable parameters 46a′ according to the sub-synchronous components V.sub.xs and V.sub.ys of the voltage of the power grid during a specific time interval t2 and for a specific number of turbines 900 coupled to the grid in each case. It is found that performance improves with respect to the previous case of the state of the art (FIG. 10), stability of the system 1000 improving regardless of the power being generated and the number of wind turbines 900 connected to the grid. Specifically, the results are shown for the following numbers of turbines 900 coupled to the grid: 1 (total power P1), 5 (total power P5), 10 (total power P10), 15 (total power P15), 20 (total power P20), 25 (total power P25), 30 (total power P30), 35 (total power P35), 40 (total power P40), 45 (total power P45) and 50 (total power P50).

[0068] FIG. 12 likewise shows performance of the wind farm with compensation with variable parameters 46a′ according to the zero-sequence component of the current during a specific time interval t3 and for a specific number of turbines 900 coupled to the grid in each case. It is also found that performance improves with respect to the previous case of the state of the art (FIG. 10), stability of the system 1000 improving regardless of the power being generated and the number of wind turbines 900 connected to the grid. Specifically, the results are shown for the following numbers of turbines 900 coupled to the grid: 1 (total power P1), 5 (total power P5), 10 (total power P10), 20 (total power P20), 30 (total power P30), 56 (total power P56) and 126 (total power P126).