The following relates, generally, to power inverters; and more particularly, to a system and method for determining active and reactive currents during asymmetrical low-voltage ride through (LVRT) conditions.
Modern grid codes (GCs) usually standardize the operation of inverter-based resources (IBRs) during low-voltage ride-through (LVRT) conditions. Conventionally, GCs require that the IBRs support the voltage during LVRT by injecting positive-sequence reactive current. More recently, however, some GCs mandate the generation of negative-sequence reactive current as well. As shown in the example of
Regardless of the GC, an inverter's phase currents must be limited, necessitating a prioritization scheme for the different components of current. Some GCs give higher priority to the reactive component of an IBR's LVRT current over its active component. Once the reactive current requirement in
In an aspect, there is provided a method for determining active and reactive currents during asymmetrical low-voltage ride through (LVRT) conditions at an inverter, the method executable on a controller or executed as a model on a computer, the method comprising: receiving an indication of an LVRT condition; and where there is an active current such that the largest phase current magnitude does not exceed a phase current limit, determining a maximum active current for associated positive-sequence and negative-sequence reactive currents by determining a largest active current magnitude and outputting the largest active current and associated positive-sequence and negative-sequence reactive currents to the inverter, otherwise: scaling down each of the positive-sequence and negative-sequence reactive currents, or superimposed positive-sequence and negative-sequence reactive currents, to determine revised positive-sequence and negative-sequence reactive currents; where the magnitudes of all of the phase currents are below the phase current limit and a β condition for each phase after the scaling down is within a predetermined range, determining non-zero positive-sequence revised active current, the β condition based on a negative voltage angle (θV
In a particular case of the method, determining whether there is an active current such that the largest phase current magnitude does not exceed a phase current limit comprises determining a maximum active current by determining ranges for the active current such that each phase current does not exceed the phase current limit and selecting the upper bound of the range.
In another case of the method, the β condition equals π+θV
In yet another case of the method, scaling down each of the positive-sequence and negative-sequence reactive currents comprises scaling uniformly based on a change in the positive-sequence and negative-sequence reactive currents.
In yet another case of the method, scaling down each of the positive-sequence and negative-sequence reactive currents comprises uniformly scaling by determining a scaling factor that causes at least one of the phase currents equal to the phase current limit.
In yet another case of the method, determining the scaling factor comprises: determining all of the possible solutions for the scaling factor using a quadratic relationship to the phase current limit; discarding any solutions to the scaling factor that are outside of the [0,1] range; and selecting a largest solution to the scaling factor that makes at least one of the phases equal to or below the phase current limit and the other phases below the phase current limit.
In yet another case of the method, the quadratic relationship comprises the scaling factor (ρ) related to the phase current limit (Imax) by Imax2=λ2φρ2+Δ1φρ+λ0φ, where coefficients λ2φ, λ1φ, and λ0φ, are related to parameters of superimposed reactive currents, pre-fault positive-sequence reactive currents, and capacitor reactive currents.
In yet another case of the method, determining the non-zero positive-sequence revised active current comprises determining a magnitude of an active current that satisfies the relationship of |Iφ|=Imax=√{square root over (|Ix|2+|Iy|2)}, where Ix comprises the positive-sequence reactive current (IQ+) and the angle between the positive-sequence reactive current and either the negative-sequence reactive current (β), the negative-sequence reactive current (β) minus 2π/3, or the negative-sequence reactive current (β) plus 2π/3, and Iy comprises the negative-sequence reactive current (IQ−) and the angle between the positive-sequence reactive current and either the negative-sequence reactive current (β), the negative-sequence reactive current (β) minus 2π/3, or the negative-sequence reactive current (β) plus a 2π/3
In yet another case of the method, where the A-phase is the largest active phase current, the Ix equal
and the Iy equals (|IQ−|sin β)∠θV
and the Iy equals
and where the C-phase is the largest active phase current, the Ix equals
and the Iy equals
θV
In yet another case of the method, determining the non-zero positive-sequence revised active current such that the current of the A-phase is below the phase current limit comprises the magnitude of the active phase current limited to −2|IQ−|sin β, such that the current of the B-phase is below the phase current limit comprises the magnitude of the active phase current limited to
and such that the current of the C-phase is below the phase current limit comprises the magnitude of the active phase current limited to
In yet another case of the method, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling where the negative-sequence reactive current is decreased at a higher rate than the positive-sequence reactive current.
In yet another case of the method, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling where the positive-sequence reactive current is decreased at a higher rate than the negative-sequence reactive current.
In yet another case of the method, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling where the positive-sequence reactive current is decreased at a higher rate than the negative-sequence reactive current and superimposed positive-sequence reactive currents are greater than superimposed negative-sequence reactive currents.
In yet another case of the method, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling using optimization, where the negative-sequence reactive current is decreased at a higher rate than the positive-sequence reactive current.
In yet another case of the method, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling using optimization, where the positive-sequence reactive current is decreased at a higher rate than the negative-sequence reactive current.
In yet another case of the method, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling using optimization, where the positive-sequence reactive current is decreased at a higher rate than the negative-sequence reactive current and superimposed positive-sequence reactive currents are greater than superimposed negative-sequence reactive currents.
In yet another case of the method, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling where the negative-sequence current is prioritized to be decreased before the positive-sequence current.
In yet another case of the method, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling where the positive-sequence current is prioritized to be decreased before the negative-sequence current.
In yet another case of the method, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling using optimization, where the negative-sequence current is prioritized to be decreased before the positive-sequence current.
In yet another case of the method, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling using optimization, where the positive-sequence current is prioritized to be decreased before the negative-sequence current.
In another aspect, there is provided a system for determining active and reactive currents during asymmetrical low-voltage ride through (LVRT) conditions at an inverter, the system comprising a processing unit in communication with a non-transitory computer-readable medium comprising instructions to cause the processing unit to execute: receiving of an indication of an LVRT condition; where there is an active current such that the largest phase current magnitude does not exceed a phase current limit, determining a maximum active current for associated positive-sequence and negative-sequence reactive currents by determining a largest active current magnitude; where there is an active current such that the largest phase current magnitude does not exceed the phase current limit, outputting the largest active current and associated positive-sequence and negative-sequence reactive currents to the inverter; where there is no active current such that the largest phase current magnitude does not exceed the phase current limit, scaling down each of the positive-sequence and negative-sequence reactive currents, or superimposed positive-sequence and negative-sequence reactive currents, to determine revised positive-sequence and negative-sequence reactive currents; where there is no active current such that the largest phase current magnitude does not exceed the phase current limit and a β condition for each phase after the scaling down is within a predetermined range, determining a non-zero positive-sequence revised active current where the magnitudes of all of the phase currents are below the phase current limit, the β condition based on a negative voltage angle (θV
In a particular case of the system, determining whether there is an active current such that the largest phase current magnitude does not exceed a phase current limit comprises determining a maximum active current by determining ranges for the active current such that each phase current does not exceed the phase current limit and selecting the upper bound of the range.
In another case of the system, the β condition equals π+θV
In yet another case of the system, scaling down each of the positive-sequence and negative-sequence reactive currents comprises uniformly scaling based on a change in the positive-sequence and negative-sequence reactive currents.
In yet another case of the system, scaling down each of the positive-sequence and negative-sequence reactive currents comprises uniformly scaling by determining a scaling factor that causes at least one of the phase currents equal to the phase current limit.
In yet another case of the system, determining the scaling factor comprises performing: determining all of the possible solutions for the scaling factor using a quadratic relationship to the phase current limit; discarding any solutions to the scaling factor that are outside of the [0,1] range; and selecting a largest solution to the scaling factor that makes at least one of the phases equal to or below the phase current limit and the other phases below the phase current limit.
In yet another case of the system, the quadratic relationship comprises the scaling factor (ρ) related to the phase current limit (Imax) by Imax2=λ2φρ2+λ1φρ+λ0φ, where coefficients λ2φ, λ1φ, and λ0φ, are related to parameters of superimposed reactive currents, pre-fault positive-sequence reactive currents, and capacitor reactive currents.
In yet another case of the system, determining the non-zero positive-sequence revised active current comprises determining a magnitude of an active current that satisfies the relationship of |Iφ|=Imax=√{square root over (|Ix|2+|Iy|2)}, where Ix comprises the positive-sequence reactive current (IQ+) and the angle between the positive-sequence reactive current and either the negative-sequence reactive current (β), the negative-sequence reactive current (β) minus 2π/3, or the negative-sequence reactive current (β) plus 2π/3, and Iy comprises the negative-sequence reactive current (IQ−) and the angle between the positive-sequence reactive current and either the negative-sequence reactive current (β), the negative-sequence reactive current (β) minus 2π/3, or the negative-sequence reactive current (β) plus 2π/3
In yet another case of the system, where the A-phase is the largest active phase current, the Ix equals (|IQ+|+|IQ−|cos β)∠(θV
and the Iy equals
and where the C-phase is the largest active phase current, the Ix equals
and the Iy equals
θV
In yet another case of the system, determining the non-zero positive-sequence revised active current such that the current of the A-phase is below the phase current limit comprises the magnitude of the active phase current limited to −2|IQ−|sin β, such that the current of the B-phase is below the phase current limit comprises the magnitude of the active phase current limited to
and such that the current of the C-phase is below the phase current limit comprises the magnitude of the active phase current limited to
In yet another case of the system, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling where the negative-sequence reactive current is decreased at a higher rate than the positive-sequence reactive current.
In yet another case of the system, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling where the positive-sequence reactive current is decreased at a higher rate than the negative-sequence reactive current.
In yet another case of the system, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling where the positive-sequence reactive current is decreased at a higher rate than the negative-sequence reactive current and superimposed positive-sequence reactive currents are greater than superimposed negative-sequence reactive currents.
In yet another case of the system, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling using optimization, where the negative-sequence reactive current is decreased at a higher rate than the positive-sequence reactive current.
In yet another case of the system, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling using optimization, where the positive-sequence reactive current is decreased at a higher rate than the negative-sequence reactive current.
In yet another case of the system, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling using optimization, where the positive-sequence reactive current is decreased at a higher rate than the negative-sequence reactive current and superimposed positive-sequence reactive currents are greater than superimposed negative-sequence reactive currents.
In yet another case of the system, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling where the negative-sequence current is prioritized to be decreased before the positive-sequence current.
In yet another case of the system, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling where the positive-sequence current is prioritized to be decreased before the negative-sequence current.
In yet another case of the system, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling using optimization, where the negative-sequence current is prioritized to be decreased before the positive-sequence current.
In yet another case of the system, scaling down each of the positive-sequence and negative-sequence reactive currents comprises non-uniform scaling using optimization, where the positive-sequence current is prioritized to be decreased before the negative-sequence current.
In another aspect, there is provided a method for determining active and reactive currents during asymmetrical low-voltage ride through (LVRT) conditions at an inverter, the method executable on a controller or executed as a model on a computer, the method comprising: receiving an indication of an LVRT condition; and capping the positive-sequence and negative-sequence reactive currents at prespecified limits and outputting the capped positive-sequence and negative-sequence reactive currents to the inverter, and determining a maximum active current for associated positive-sequence and negative-sequence reactive currents by determining a largest active current magnitude and outputting the largest active current and associated positive-sequence and negative-sequence reactive currents to the inverter.
These and other aspects are contemplated and described herein. It will be appreciated that the foregoing summary sets out representative aspects of the system and method to assist skilled readers in understanding the following detailed description.
A greater understanding of the embodiments will be had with reference to the Figures, in which:
For simplicity and clarity of illustration, where considered appropriate, reference numerals may be repeated among the Figures to indicate corresponding or analogous elements. In addition, numerous specific details are set forth in order to provide a thorough understanding of the embodiments described herein. However, it will be understood by those of ordinary skill in the art that the embodiments described herein may be practised without these specific details. In other instances, well-known methods, procedures and components have not been described in detail so as not to obscure the embodiments described herein. Also, the description is not to be considered as limiting the scope of the embodiments described herein.
Various terms used throughout the present description may be read and understood as follows, unless the context indicates otherwise: “or” as used throughout is inclusive, as though written “and/or”; singular articles and pronouns as used throughout include their plural forms, and vice versa; similarly, gendered pronouns include their counterpart pronouns so that pronouns should not be understood as limiting anything described herein to use, implementation, performance, etc. by a single gender. Further definitions for terms may be set out herein; these may apply to prior and subsequent instances of those terms, as will be understood from a reading of the present description.
Any module, unit, component, server, computer, terminal or device exemplified herein that executes instructions may include or otherwise have access to computer readable media such as storage media, computer storage media, or data storage devices (removable and/or non-removable) such as, for example, magnetic disks, optical disks, or tape. Computer storage media may include volatile and non-volatile, removable and non-removable media implemented in any method or technology for storage of information, such as computer readable instructions, data structures, program modules, or other data. Examples of computer storage media include RAM, ROM, EEPROM, flash memory or other memory technology, CD-ROM, digital versatile disks (DVD) or other optical storage, magnetic cassettes, magnetic tape, magnetic disk storage or other magnetic storage devices, or any other medium which can be used to store the desired information and which can be accessed by an application, module, or both. Any such computer storage media may be part of the device or accessible or connectable thereto. Further, unless the context clearly indicates otherwise, any processor or controller set out herein may be implemented as a singular processor or as a plurality of processors. The plurality of processors may be arrayed or distributed, and any processing function referred to herein may be carried out by one or by a plurality of processors, even though a single processor may be exemplified. Any method, application or module herein described may be implemented using computer readable/executable instructions that may be stored or otherwise held by such computer readable media and executed by the one or more processors.
Generally, implementations of the requirements for generation of the negative-sequence current for inverters during LVRT have not considered the requirements with the same hierarchy. For example, currents may not ensure full utilization of the IBR's phase current capacity under all LVRT conditions. In addition, ΔIQ+ and ΔIQ− are not necessarily prioritized over IP+. As another example, some schemes violate the GCs by generating active negative-sequence current, IP−.
Generally, conventional inverter design follows the following approach: the inverter attempts to meet a requirement for current, for example as shown in
The present embodiments provide an approach to advantageously maximize the active current of an IBR while the reactive current requirements are met. Moreover, if the phase currents hit the limit due to large superimposed reactive currents ΔIQ± shown by
In an example, the LVRT mode can be engaged when one or more phase to phase voltages are outside of a static voltage range and/or there is a sudden change in voltage. In this example, the LVRT mode can be deactivated when all the phase to phase voltages are in the static voltage range or after five seconds if the sudden voltage change did not result in any voltage exceeding the static voltage range. The sudden voltage jump can be defined by an absolute difference between an actual value of the positive and negative sequence voltage and a 50 period average of the positive and negative sequence voltage relative to a declared voltage.
As described herein, the present inventors conducted example experiments to verify the effectiveness and advantages of the present embodiments. In the example experiments, PSCAD/EMTDC simulations were conducted of a modified version of the IEEE 39-bus system, depicted in the test system of
Common approaches used to determine the reference for the positive-sequence active current of an IBR that generates negative-sequence current have substantial limitations. When the reactive current is prioritized, a straightforward way to derive the reference for |IP+| is based on Equation (1), which maintains the scalar sum of the positive- and negative-sequence current magnitudes below the phase current limit of the IBR:
|IP−| is zero or very small in comparison to |IQ−|, so the only unknown in Equation (1) is |IP+|. The following case studies evaluate the performance of this approach:
Example Case 1 elaborates on the effective utilization of an IBR's current capacity when the superimposed reactive currents ΔIQ+ and ΔIQ− given by
The phase currents resulting from the above sequence currents are displayed in
Since the maximum phase current, |Ic|, is 0.13 pu less than Imax, one might expect that |IP+| can be increased by at most 0.13 pu, and then the inverter's capacity is fully utilized. However, increasing |IP+| by 0.13 pu while |IQ+| and |IQ−| are kept the same as in
Although GCs generally prioritize the reactive power, they also generally require maximizing active power. This is critical in maintaining the load-generation balance; i.e., the ultimate objective of the LVRT requirement. Effective utilization of the seemingly small 0.2-pu excess current capacity of the inverter in the phase domain offers sizable active power in the sequence domain. This can be made clear only when the excess current capacity is maximally used, as described herein.
Example Case 2 focuses on when the superimposed reactive currents ΔIQ+ and ΔIQ− given by
One obvious problem of the above process is that the 0.857 scaling factor obtained using Equation (1) prevents maximizing the reactive currents, and so 15% of the inverter's excess current capacity remains unused. If |IQ+| and |IQ−| were scaled down by a factor of 0.879 to 0.63 pu and 0.60 pu, respectively, then the current of phase C would reach to 1.2 pu. This would satisfy the IBR's phase current limit, even though it violates Equation (1).
The second (and more substantial) problem is not as obvious. It is taken for granted that once 0.879 is used to scale down |IQ+| and |IQ−| to 0.63 pu and 0.60 pu, and the phase C current hits the limit, no room is left to generate |IP+|. However, as shown in
Turning to
The system 100 executes a method that complies with the requirement to maximize an IBR's active current during LVRT while the requirements of applicable grid codes (for example, as outlined in the diagram of
At block 202, the processing unit 102 receives pre-LVRT quantities values for the currents from respective current sensors associated with the inverter.
At block 204, the processing unit 102 receives an indication of an LVRT condition. Determination of LVRT conditions can vary depending on a given GC. For example, in the Verband der Elektrotechnik, Elektronik und Informationstechnik e.V. (VDE) code, one of the following two events are specified as the criterion for the start of the LVRT condition:
In another example, under the Institute of Electrical and Electronics Engineers (IEEE) P2800 standard, except for 500 kV nominal voltage, the continuous operating region is when the applicable voltage is ≥0.9 per unit and ≤1.05 per unit. For 500 kV nominal voltage, the continuous operating range is when the applicable voltage is ≥0.9 per unit and ≤1.10 per unit. During temporary low voltage disturbances, the applicable voltage is defined as the lowest magnitude fundamental frequency phasor component phase-phase or phase-ground voltage at the RPA relative to the corresponding nominal system voltage. According to the standard, LVRT can be defined as when the applicable voltage drops below 0.9 pu.
At block 205, the processing unit 102 determines whether the largest phase current magnitude exceeds a phase current limit.
At block 206, where the largest phase current magnitude Imax does not exceed a phase current limit, the processing unit 102 determines a maximum active current for the given reactive currents. Generally, inverters have an applicable limit for their phase current to prevent damage to power electronic switches and other componentry. In many cases, this phase current limit is between 100% to 200% of the inverter's rated current. In some cases, this limit can be time-variant; for example, the phase current limit can be 160% of the rated current during the first 20 milliseconds of LVRT and decrease to 120% of the rated current for the remainder of the LVRT period. While the present disclosure generally refers to a time-invariant phase current limit, it is understood that this limit can be time-dependent.
The relation between the different components of an IBR's sequence currents and the phase currents that flow through the inverter switches is shown in Equation (2); where α=ej2π/3; By denotes the voltage angle at the POC; pre in the subscript denotes the pre-LVRT quantities; and cap indicates the quantities associated with the shunt capacitor of the inverter's filter.
From Equation (2), the phase currents can be written as:
where Φ∈{a, b, c}, and φ is 0, −2π3, and 2π3 for phases A, B, and C, respectively. The reactive components of IΦ are derived using Equation (4), which includes |ΔIQ±| given by
The processing unit 102 determines the active current |IP+| in Equation (3) such that the largest phase current given by Equation (3) equals Imax. For each phase current, the first two of the three vectors on the right side of Equation (3), i.e., |IQ+|φ(θV
The square magnitudes of the phase currents in Equation (6) are:
The IBR limit for the three phase currents can be expressed as the three inequalities embedded in Equation (9) for different values of φ.
In Equation (9), |IQ+| and |IQ−| are given by
To show the difference made using the above approach, consider the fault of example Case 1 while the processing unit 102 with respect to IBR-4 uses Equation (10) to determine |IP+|. Using the angles of the sequence voltages shown in
In certain cases, there are conditions under which Equation (10) returns an empty set, and as such, there are approaches used by the processing unit 102 for these conditions to satisfy the GC. Generally, an inverter's maximum current is fairly small. In addition, when a fault is not very far from the IBR, and so the change in the voltage is significant, the reactive currents determined by
At block 207, the processing unit 102 communicates the reactive currents and the active currents to the respective IBR 150.
At block 208, where the largest phase current magnitude Imax exceeds a phase current limit, the processing unit 102 scales down the positive-sequence and negative-sequence reactive currents, or superimposed positive-sequence and negative-sequence reactive currents. In some cases, this scaling can be performed uniformly on both the positive-sequence and negative-sequence. In other cases, as described herein, the positive-sequence and negative-sequence can be scaled using other current limitation strategies, as described herein. In contrast, other approaches scale the total positive-sequence and negative-sequence reactive currents, IQ±. Thus, the formulations in these schemes include scaling the pre-fault current, IQ-pre, and the current through the capacitor of the inverter's LC filter, IQ-cap, neither of which is generally controllable during LVRT. Advantageously, in the present embodiments, the processing unit 102 provides for such current to be scaled.
When positive-sequence and negative-sequence reactive currents are not scaled down, the following relation holds as long as the same K-factor is used for the positive and negative sequence in
Therefore, the IBR's equivalent impedances in the two sequence circuits are similar, replicating a synchronous generator. An IBR should ideally maintain the same relation after the currents are scaled down, so that the similarity with the synchronous generators is preserved. Scaling the total reactive currents violates Equation (10) since the total reactive currents are not limited to only the superimposed currents given by
Equation (12) is solved for ρ such that max{|Ia|, |Ib|, |Ic|}=Imax to ensure maximum utilization of the inverter's capacity. Equating the magnitude of phase currents given by Equation (12) with Imax yields the three equations embedded in Equation (13) for φ=0, −2π3, and +2π3, corresponding to phases A, B, and C, respectively. This relation can be written with respect to p, as in:
where the coefficients λ2φ, λ1φ, and λ0φ, are expressed by Equation (15) in terms of the known parameters of Equation (13).
In Equation (15), if IQ-pre+ is capacitive, μ=−1, and the upper sign must be used in ± and ∓. For an inductive IQ-pre+, however, the lower sign must be used in ± and ∓, and μ=−1 when |IQ-pre+|+ρΔIQ+<0, and μ=1 when |IQ-pre+|+ρΔIQ+>0. As ρ is the unknown of Equation (14), the sign of |IQ-pre+|+ρΔIQ+ cannot be determined before Equation (14) is solved. Thus, Equation (14) must be solved for both conditions, i.e., |IQ-pre+|+ρΔIQ+<0 and |IQ-pre+|+ρΔIQ+>0. For each of these inequalities, the calculated ρ is acceptable if the respective inequality is held for that ρ.
To find an optimal ρ that satisfies all of the constraints, the equation for each phase embedded in Equation (14) is solved independently of the other two equations in Equation (14) (for the other two phases) but the constraint on the current magnitude must be satisfied for all three phases. Therefore, the scaling factor can be found by, first, finding all of the possible solutions for ρ in each equation embedded in Equation (14); then, discarding any ρ that is outside the [0,1] range because such ρ's do not scale down the current magnitude; then, choosing the largest ρ that is the solution of Equation (14) for one of the phases but also simultaneously keeps the current magnitude in the other two phases below Imax.
Since Equation (12) was solved to satisfy the max{|Ia|, |Ib|, |Ic|}=Imax condition, the scaled-down positive- and negative-sequence reactive currents obtained above make the current of at least one phase equal to the inverter's limit. When one of the phase currents reaches the limit, generally, it has previously been assumed that the inverter has no room to inject active current. At block 210, the processing unit 102 can calculate a non-zero positive-sequence active current, IP+ to be generated by the IBR 150, while the reactive currents obtained above remain intact and the inverter's phase current limit is satisfied.
Assume, without loss of generality, that |Ia| is the largest phase current when IQ± is scaled down and no IP+ is generated, i.e.:
The angle between IQ+ and IQ− in Equation (16) is Q defined in Equation (8). Either 0≤β≤π, as in
for
For π≤β≤2π in
can be determined by the processing unit 102. This would keep |Iy| unchanged because for the |IP+| given by Equation (20), Iy=(|IQ−|sin β−2|IQ−|sin β)∠θV
Although the above formulation does not ensure that Ib and Ic are less than Imax, it proves that Equation (10) can be used to determine the |IP+| that can be generated after scaling down IQ±. Substituting the Imax given by Equation (17) and Equation (8) into Equation (10) shows that the |IP+| given by Equation (20) is the upper bound of the second range on the right side of Equation (10), which corresponded to phase A. It can be similarly shown that if the derivations in Equations (16) to (20) are carried out for phases B and C (which would require only shifting β by +2π3 and −2π3, respectively), substituting the Imax given by Equation (17) and Equation (8) into Equation (10) makes the upper bound of the third and the fourth range on the right side of Equation (10) equal to the |IP+| in Equation (20), respectively. The third and the fourth range in Equation (10) corresponded to phases B and C, respectively. Consequently, if the scaled-down reactive current determined herein are plugged into Equation (10) as IQ±, this equation will provide the maximum |IP+| that keeps all of the phase currents below Imax.
At block 212, the processing unit 102 communicates the revised reactive currents and the revised active currents to the respective IBR 150. The IBR 150 generally has a positive-sequence control loop which receives the references for the revised positive-sequence currents and generates such references. The IBR 150 generally also has a similar loop for generating the revised negative-sequence currents. In the block diagram of
The method 200 can be applicable to a variety of LVRT conditions. In an example, for example Case 2, |IQ-pre+|=0.038 pu (capacitive), ΔIQ−=0.68, and ΔIQ−=0.68. The angles of the sequence voltages in
Substituting the above values into Equation (5), the coefficients (λ2,φ, λ1,φ, λ0,φ) of the quadratic equation, Equation (14), for phases A, B, and C (corresponding to φ=0, −2π3, and +2π3) will be (0.8748, 0.0103, 0.0000), (1.7495, 0.0206, 0.0000), and (0.1501, 0.0018, 0.0000), respectively. Solving the three quadratic equations in (13), the solutions for ρ will be (−1.2889, 1.2771), (−0.9131,0.9014), and (−3.1035, 3.0917), for φ=0, −273, and +2π3 associated with phases A, B, and C, respectively. Among these values, ρ=0.9014 is a solution of Equation (14) for phase B inside the [0,1] range. For this p, the scaled-down current references on the right side of Equation (10) are |IQ+|=|−|IQ-pre+|+ρΔIQ+|−|IQ-cap+|=0.62 pu, and |IQ−|=ρΔIQ−+|IQ-cap−|=0.62 pu. When these reference currents are plugged into Equation (12), the current of phase B is limited to Imax=1.2 pu while the current of the other two phases are smaller than the limit (|Ia|=0.85 pu and |Ie|=0.35 pu). Thus, ρ=0.9014 is the acceptable scaling factor.
Since the current of phase B is maximum, the BR will be able to generate IP+ if 180°<β−120°<360°. β=93.1°, and so this condition holds. Using the above IQ±, the largest |IP+| that satisfies Equation (10) is 0.43 pu, displayed along with |IQ±| in
The advantages of the present embodiments both increase active current generation and maximizes reactive current of the inverter beyond levels provided by other approaches. As described herein, when the ΔIQ+ and ΔIQ− given by
In example Case 3, consider if IBR-4 of
If IBR-4 uses method 200, i.e., by substituting |IQ+|=1.23 pu, |IQ−|=0.93 pu, |IQ-cap+|=0.033 pu, and |IQ-cap−|=0.004 pu, along with sequence voltage angles of
Since the GCs generally prioritize reactive current over active current during LVRT, to satisfy the GCs' LVRT requirement other approaches generally, first, check if the reactive currents given by
Conversely to other approaches, the system 100 first attempts to find the maximum active current for the reactive currents of
Generally, GCs allow scaling down IQ± only if it is necessary to do so, which is not the case for conditions like the one displayed in
In example Case 4, consider an AG fault with Rf=5Ω at 40% of the line connecting B23 to B24 in
Meanwhile, if the reactive currents and sequence voltage angles shown in
As described herein, IBR can generate non-zero |IP+| after IQ± are scaled down only if β+φ is inside the [180°, 360°] range, where φ is 0, +120°, or −120° when the maximum phase current after scaling down IQ± occurs in phase A, phase B, or phase C, respectively. The present inventors investigated the likelihood of 180°<(β+φ)<360° in real power systems; which was determined to depend on the fault type.
For a bolted AG fault, the phase lead of the negative-sequence voltage over the positive-sequence voltage (θV
Additionally, the phase lead of the negative-sequence reactive current over the positive-sequence reactive current, ∠IQ−−∠IQ+, obtained through Equation (12), is β, β−120°, and β+120°, for phases A, B, and C, respectively. Given the above range for β, the following angles are obtained at the POC:
The smaller angle between the two sequence components of phase B current makes Ib larger than the phase A and C currents and equal to Imax. As described herein, when Ib=Imax, the IBR is able to generate |IP+| if 180°<β−120°<360°. This condition is normally satisfied as the above described how β is inside [60°, 120° ] for AG faults.
For a YNd11 transformer, θV
For BC and BCG faults, θV
The angle between the negative- and positive-sequence reactive currents is smaller for phase C, indicating that the current of phase C is maximum and equal to Imax. As described herein, when the phase C current is maximum, an IBR can generate IP+ if 180°<β+120°<360°. Meanwhile, it was shown in the above that 180°<β<240°, and so 300°<β<360°. Thus, the condition on β to generate IP+ holds.
If the transformer's vector group is YNd11, β at the POC is normally inside [60°, 120° ] for BC and BCG faults. This makes Ib equal to Imax. Thus, the IBR is able to generate |IP+| if 180°<β−120°<360°, which is usually satisfied since β is inside [60°, 120° ] as mentioned above. A similar result is obtained for other double-phase faults as well
Advantageously, the method 200 can factor in the current of the LC filter's shunt capacitor (IQ-cap) in the formulation that derives the reference current. This would marginally impact the reference currents derived herein. IQ-cap is an uncontrolled current that is added to the currents of the inverter switches before the POC. Thus, if IQ-cap is neglected when the reference currents are determined, the superimposed reactive currents measured at the POC will deviate slightly from what the GC may require. This error can be compensated for the positive-sequence circuit in Equation (4) by subtracting |IQ-cap+| from the required reactive current at the POC, i.e., |±|IQ-pre++ΔIQ+|. Similarly, ΔIQ-cap− can be added to the required negative-sequence reactive current at the POC (i.e., ΔIQ−) in Equation (5) to derive |IQ−|. Thus, when IQ-cap± is taken into account, the BR generates smaller positive- and larger negative-sequence reactive currents, compared to when IQ-cap± is neglected.
Since the positive-sequence voltage at the POC never drops below the negative-sequence voltage, |IQ-cap+| always exceeds |IQ-cap−|. Therefore, when IQ-cap± is taken into account, the decrease in the positive-sequence reactive current is larger than the increase in the negative-sequence reactive current. Therefore, besides the absolute precision in complying with the GC, considering IQ-cap± determined in the method 200 creates more room to generate active current.
In addition to maintaining the balance between the load and generation, maximum utilization of the IBRs' current generation capacity by the method 200 enhances the grid stability from the following perspectives:
Provided herein is an example comparative analysis to show the system-wide impact made by the method 200 in increasing the power of the IBRs across the grid shown in
Columns 5 and 8 of Table 1 indicate the for all fault conditions P+ of the method 200 is larger. Consider, for example, case 6 with K=6. The total active power generated by the IBRs using the method 200 is 443 MW (124%) larger than the total active power when the other techniques are used. A noticeable pattern for P+ in Table 1 is that as the K-factor increases, the active power generated by the method 200 becomes more superior. For instance, for a BC fault at bus B16, and for K=2, K=4, and K=6, the method 200 generates 163 MW (11%), 306 MW (52%), and 318 MW (110%), respectively, more than the other techniques. The reason is that for larger K-factors, the other techniques do not attempt to generate any active current after scaling down the reactive currents. For the method 200, however, large amounts of active current can be generated.
As discussed herein, one of the advantages of the method 200 lies in the generation of not only larger active power, but also higher levels of reactive power. As shown in Table 1, the improvement in the reactive power is more significant for cases involving single-line-to-ground faults and large K-factors. These are the cases for which the reactive currents need to be scaled down due to the large K-factor even though the voltage drop at the POC is not very severe. In Case 12, for example, the IBRs generate 259 MVAr (15%) of extra reactive if they use the method 200. For the same case, the method 200 generates 64 MW (13%) of additional active power as well. Only in three cases with lower K-factor, the Q+ generated by the method 200 is 1 to 3 MVAr smaller. For instance, the Q+ given by the method 200 is 882 MVAr, while the other techniques yield 885 MVAr; i.e., a mere 0.34% difference. This stems from the fact that for the other techniques, the dip in the positive-sequence voltage during the transients of the first cycle of the fault is larger. The first cycle of the fault is when the K-factor diagram in
Comparison of columns 7 and 10 in Table 1 reveals that the method 200 generates larger Q− for all of the cases. Meanwhile, the pattern of variations for Q− is different from what is observed for Q+. As the K-factor increases, Q+ generated by IBRs naturally increases for all of the three approaches in Table 6. Similarly, Q− increases in columns 7, 10, and 13 as K is increased from 2 to 4. However, in most cases, from K=4 to K=6, Q− decreases. The reason is that from K=4 to K=6, the generation of larger amount of Q+ causes the power system to become more balanced, with larger positive-sequence voltage and smaller negative-sequence voltage throughout the grid.
As described herein, compensating for the current of the filter's capacitor enables the IBR to generate larger active currents. Furthermore, a comparison between the reactive currents of the two approaches confirms that when an BR compensates for the current of the shunt capacitor, Q+ decreases and Q− increases. The numbers for Q− seem to stay unchanged in some cases. This is likely due to the rounding error. For instance, Q− in Case 1 for the method 200 without and with compensation of the capacitor's current is 103.00 MVAr and 103.13 MVAr, respectively.
Instead of using closed form mathematical relations, in some cases, scaling down the reactive currents and re-calculating the active current can be combined and carried out using an optimization technique. In these cases, an optimization problem can be optimized with the objective function of maximizing ρ and the following constraints:
Constraints of Equation (21) and Equation (22) are features of ρ and |IP+|, and Equation (23) implies that the phase currents have to satisfy their limit. Convex optimization problems have a globally optimum answer and can be solved through any suitable standard convex solver.
In addition to the objective function of finding the largest scaling factor ρ, this optimization problem should guarantee that the maximum phase current capacity of the BR is used; i.e., at least one of the phase currents becomes exactly equal to the Imax limit. However, one concern might be that ρ and |IP+| are found such that the phase currents become less than the limit, while none of them is equal to the limit. The following proves that this scenario is impossible. Based on the geometric interpretation of the optimality condition, it can be concluded that the optimal point of a convex problem exists on the boundary of the feasibility set of the optimization problem. As the optimization problem is convex, the optima exist on the boundary of the feasibility set. This implies that the optimal (ρ, |IP+|) is in either of the following sets:
If at the optima, |IP+|=0, then the largest value for ρ is what was determined from Equation (14), which would make at least one of the phase currents saturate at the limit. Therefore, if the optima is in 1, then at least one of the phase currents saturates at the limit. Furthermore, the optima being in
2 trivially results in at least one of the phase currents being equal to the limit.
Advantageously, the computation time of this optimization approach is not substantial. The optimization problem is quadratic with only two variables, i.e., ρ and |IP+|. Such small optimization problems can be solved in around one millisecond using available solvers.
For example, example Case 2, in which the superimposed reactive currents obtained from
In some cases, it may be possible to neglect the shunt capacitor from the LC filter 1902 of the inverter; and the inverter will still be able to operate, but with some degree of distortion in the currents. Neglecting the LC filter's capacitor's currents may be a safe practice because some GCs allow small deviations from the required reactive currents as illustrated in
For
For
In some cases, Blocks 2108 and 2112 can be performed either in parallel with Block 2106, or sequentially after Block 2106. This is because the output can be obtained either at Block 2108 (if |IP+|∉Ø) or Block 2112 (if |IP+|∈Ø). In the block diagram of
In
The current limitation approach, for example represented by Blocks 2108 and 2110, and that uses Equations (14) and (12), assumes that the reactive positive- and negative-sequence currents are uniformly scaled down. However, it is to be understood that the present embodiments can be used with other current limitation strategies. The following current limitation strategy examples describe changes required in the above formulations and equations in order to implement these current limitation strategies; any of which may be used as appropriate.
In a first example current limitation strategy, non-uniform scaling of the sequence reactive currents can be performed when the negative-sequence reactive current is to be decreased at a higher rate. In this case, a new parameter γ− s.t. 0<γ−<1 is selected. The smaller is γ−, the higher is the reduction rate of ΔIQ− (compared with the reduction rate of ΔIQ+). In order to make sure that the selected γ− does not cause the maximum phase current to fall below Imax, a lower bound for γ−, i.e., γL−, is determined. γL− should satisfy 0<γL−<1 and Equation (26) such that max{|Ia|, |Ib|, |Ic|}=Imax.
In order to calculate γL−, for any phase current which was above the limit for the original superimposed reactive currents, i.e., ΔIQ±, the respective Equation (27) has to be solved.
Where σ2φ−, σ1φ−, σ0φ−, can be obtained by Equation (28):
Then, the minimum of the acceptable answers is taken. If, at least, one of the phases which was originally above the limit has no acceptable answer, then the lower bound will be zero, i.e., γL−=0. Then, the selected γ− is compared with the calculated γL−. If γ−<γL−, then γ−=γL−. Using Equation (29), ρ is found such that max{|Ia|, |Ib|, |Ic|}=Imax:
Which requires solving Equation (30):
While the coefficients ξ2φ−, ξ1φ−, ξ0φ− are obtained by Equation (31):
The proper selection for μ, ± and ∓ signs, and the optimal ρ is the same as described with respect to Equations (14) and (15). The scaled superimposed reactive currents will thus be ρΔIQ+ and ργ−ΔIQ−.
In a second example current limitation strategy, non-uniform scaling of the sequence reactive currents can be performed when the positive-sequence reactive current is to be decreased at a higher rate. Similar to the first example current limitation strategy, a new parameter γ+ s.t. 0<γ+<1 is selected. The smaller the γ+, the higher the reduction rate of ΔIQ+ (compared with the reduction rate of ΔIQ−). In order to make sure that the selected γ+ does not cause the maximum phase current to fall below Imax, a lower bound for γ+, i.e., γL+, is determined. γL+ should satisfy 0<γL+<1 and Equation (32) such that max{|Ia|, |Ib|, |Ic|}=Imax.
In order to calculate γL+, for any phase current which was above the limit for the original superimposed reactive currents, i.e., ΔIQ±, the respective Equation (33) has to be solved:
σ2φ+, σ1φ+, σ0φ+, can be obtained by Equation (34) when ±|IQ-pre+|+ΔIQ+≥0 (for capacitive and inductive |IQ-pre+|, the lower and upper signs in ± is applied, respectively):
Then, the minimum of the acceptable answers is used. If, at least, one of the phases which was originally above the limit has no acceptable answer, then the lower bound will be zero, i.e., γL+=0. Then, the selected γ+ is compared with the calculated γL+. If γ+<γL+, then γ+=γL+. Using Equation (35), ρ is found such that max{Ia|, |Ib|, |Ic|}=Imax:
Which entails solving Equation (36):
While the coefficients ξ2φ+, ξ1φ+, ξ0φ+ are obtained by Equation (37):
The proper selection for μ, ± and ∓ signs, and the optimal ρ is the same as described with respect to Equations (14) and (15). The scaled superimposed reactive currents will thus be ργ+ΔIQ+ and ρΔIQ−.
In a third example current limitation strategy, non-uniform scaling of the sequence reactive currents can be performed when the positive-sequence reactive current is to be decreased at a higher rate, while avoiding the superimposed positive-sequence reactive current becoming smaller than the superimposed negative-sequence reactive current. All the steps of the second example current limitation strategy are followed. However, at the end, if −ργ+ΔIQ+≥ρΔIQ−, no further action is required; otherwise ΔIQ+ is selected to be equal to −ΔIQ− and the approach of uniform scale down of the sequence reactive currents, as described herein with respect to Equations (12) to (15), is repeated.
In a fourth example current limitation strategy, optimized non-uniform scaling of the reactive currents are performed when the negative-sequence reactive current is to be decreased at a higher rate. A new parameter γ− s.t. 0<γ−<1 is selected. The following optimization problem is solved to find a lower bound for γ−, i.e., γL−:
If the optimization problem is infeasible, then γL−=0. The selected γ− is compared with the calculated γL−. If γ−<γL−, then γ−=γL−. Then, the following optimization problem is solved:
The scaled superimposed reactive currents will then be ρΔIQ+ and ργ−ΔIQ−.
In a fifth example current limitation strategy, optimized non-uniform scaling of the reactive currents is performed when the positive-sequence reactive current is to be decreased at a higher rate. A new parameter γ+ s.t. 0<γ+<1 is selected. The following optimization problem is solved to find a lower bound for γ+, i.e., γL+:
If the optimization problem is infeasible, then γL+=0. The selected γ+ is compared with the calculated γL+. If γ+<γL+, then γ+=γL+. Then, the following optimization problem is solved:
The scaled superimposed reactive currents will then be ργ+ΔIQ+ and ρΔIQ−.
In a sixth example current limitation strategy, optimized non-uniform scaling of the sequence reactive currents is performed when the positive-sequence reactive current is to be decreased at a higher rate, while avoiding the superimposed positive-sequence reactive current to become smaller than the superimposed negative-sequence reactive current. All the steps of the fifth example current limitation strategy are followed. At the end, if −ργ+ΔIQ+≥ρΔIQ−, no further action is required; otherwise ΔIQ+ is selected equal to −ΔIQ− and the optimization of uniform scale down of the sequence reactive currents, described with respect to Equations (21) to (23), is repeated.
In a seventh example current limitation strategy, prioritized scaling of the reactive currents is performed when the negative-sequence current is to be decreased first. γL− is determined as described with respect to the first example current limitation strategy. If it exists, then the scaled reactive currents are ΔIQ+ and γL−ΔIQ−, and no further action is required. Otherwise, γL+ is calculated as described with respect to the second example current limitation strategy; while ΔIQ− in Equations (34) and (35) is set to zero. The scaled superimposed positive- and negative-sequence reactive currents will then be γL+ΔIQ+ and zero, respectively. Note that this strategy may cause the IBR to generate zero superimposed negative-sequence current. If this is undesired, a limit may be applied to the reduction of ΔIQ−, e.g., ΔIQ,lim−. If γL− exists and
then the scaled superimposed reactive currents are ΔIQ+ and γL−ΔIQ−, and no further action is required. Otherwise, γL+ is determined as described in the second example current limitation strategy; while ΔIQ− in Equations (34) and (35) is substituted by ΔIQ,lim−. The scaled superimposed positive- and negative-sequence reactive currents will then be γL+ΔIQ+ and ΔIQ,lim−, respectively.
In an eighth example current limitation strategy, prioritized scaling of the reactive currents is performed when the positive-sequence current is to be decreased first. γL+ is calculated as described with respect to the second example current limitation strategy. If it exists, then the scaled reactive currents are γL+ΔIQ+ and ΔIQ−, and no further action is required. Otherwise, γL− is calculated as described with respect to the first example current limitation strategy; while ΔIQ+ in Equation (28) is set to zero. The scaled superimposed positive- and negative-sequence reactive currents will then be zero and γL−ΔIQ−, respectively. Note that this strategy may cause the IBR to generate zero superimposed positive-sequence reactive current (which may cause the IBR's phase rotation). If this is undesired, a limit may be applied to the reduction of ΔIQ+, e.g., ΔIQ,lim+. If γL+ exists and
then the scaled superimposed reactive currents are γL+ΔIQ+ and ΔIQ−, and no further action is required. Otherwise, γL− is calculated as described with respect to the first example current limitation strategy; while ΔIQ+ in Equation (28) is substituted by ΔIQ,lim+. The scaled superimposed positive- and negative-sequence reactive currents will then be ΔIQ,lim+ and γL−ΔIQ−, respectively.
In a ninth example current limitation strategy, optimized prioritized scaling of the reactive currents is performed when the negative-sequence current is to be decreased first. γL− is determined as described with respect to the fourth example current limitation strategy. If it exists, then the scaled reactive currents are ΔIQ+ and γL−ΔIQ−, and no further action is required. Otherwise, γL+ is determined as described with respect to the fifth example current limitation strategy; while ΔIQ− in the optimization problem of Equation (40) is set to zero. The scaled superimposed positive- and negative-sequence reactive currents will then be γL+ΔIQ+ and zero, respectively. Note that this strategy may cause the IBR to generate zero superimposed negative-sequence current. If this is undesired, a limit may be applied to the reduction of ΔIQ−, e.g., ΔIQ,lim−. If γL− exists and
then the scaled superimposed reactive currents are ΔIQ+ and γL−ΔIQ−, and no further action is required. Otherwise, γL+ is calculated as described with respect to the fifth example current limitation strategy; while ΔIQ− in the optimization problem of Equation (40) has been substituted by ΔIQ,lim−. The scaled superimposed positive- and negative-sequence reactive currents will then be γL+ΔIQ+ and ΔIQ,lim−, respectively.
In a tenth example current limitation strategy, optimized prioritized scaling of the reactive currents is performed when the positive-sequence current is to be decreased first. γL+ is determined as described with respect to the fifth example current limitation strategy. If it exists, then the scaled reactive currents are γL+ΔIQ+ and ΔIQ−, and no further action is required. Otherwise, γL− is calculated as described with respect to the fourth example current limitation strategy; while ΔIQ+ in the optimization problem of Equation (38) is set to zero. The scaled superimposed positive- and negative-sequence reactive currents will then be zero and γL−ΔIQ−, respectively. Note that this strategy may cause the IBR to generate zero superimposed positive-sequence reactive current (which may cause the IBR's phase rotation). If this is undesired, a limit may be applied to the reduction of ΔIQ+, e.g., ΔIQ,lim+. If γL+ exists and
then the scaled superimposed reactive currents are γL+ΔIQ+ and ΔIQ−, and no further action is required. Otherwise, γL− is determined as described with respect to the fourth example current limitation strategy; while ΔIQ+ in the optimization problem of Equation (38) has been substituted by ΔIQ,lim+. The scaled superimposed positive- and negative-sequence reactive currents will then be ΔIQ,lim+ and γL−ΔIQ−, respectively.
In an eleventh example current limitation strategy, as exemplified in the flowchart of
In a twelfth example current limitation strategy, when Equation (10) returns an empty set (if it does not return an empty set, then no current limitation needs to be performed), the positive- and negative-sequence reactive currents are capped at prespecified limits; e.g., |IQ+|≤Ilim+, |IQ−|≤Ilim− and Ilim++Ilim−≤Imax. In this example current limitation strategy, after capping 11 and |IQ−| at their respective limits, the capped reactive currents are then plugged into Equation (10) to calculate the revised maximum |IP+| using the approach described with respect to Equations (3) to (10).
In further cases, at least some of the present equations can be augmented, as appropriate, for simplicity. Equation (2), provided herein, can be augmented to:
Equation (4), provided herein, can be augmented to:
Equation (12), provided herein, can be augmented to:
Equation (13), provided herein, can be augmented to:
Equation (15), provided herein, can be augmented to:
As shown, in the above augmented versions of Equations (12), (13), and (15), the scaling of the reactive currents have been modified. Additionally, in the above augmented version of Equation (15), for capacitive and inductive pre-fault reactive currents, the upper and lower signs, respectively, must be used in ± and ∓.
Several reference frames can be adopted for the implementation of inverters' control, such as a synchronous reference frame (dq), stationary reference frame (αβ), or natural reference frame (abc).
As described herein, existing approaches to inverter control do not satisfy the requirements of most GCs for maximum active and reactive current generation during LVRT. These techniques use the scalar sum of the sequence currents to derive the inverter reference currents, leading to miscalculation of an IBR's capacity for generating active current. This may also cause unnecessary scale down of the reactive currents without any active current generation. The present embodiments address at least these problems and maximizes both active and reactive currents of an IBR. In particular, the example experiments show that when the reactive currents make the phase currents exceed the IBR's limit, the system 100 may be able to generate non-negligible amounts of active current and bring the phase currents below their limit without scaling down the reactive currents. Additionally, the system 100 can derive a larger scaling factor and so maximize the reactive current generated by the IBR. Additionally, after scaling down the reactive currents, although at least one of the phase currents reaches the IBR's limit, the BR is still usually capable of generating active current.
While the above disclosure generally describes the present embodiments applied to the application of a power inverter, it is appreciated that the presently described approaches can be applied to any suitable application; for example, applied to applications of high-voltage direct current (HVDC) stations. In other cases, the presently described approaches can be incorporated into fault determination software or model that uses the present embodiments to determine active and/or reactive currents. Fault determination software generally requires a software model of an inverter that complies with applicable grid codes and standards. This model can be integrated with models of other power system components (for example, transmission lines, conventional power plants, protective relays, and the like) to calculate the currents and voltages in different parts of the system during fault (i.e., LVRT) conditions. Utility engineers can use the results of these calculations to set up protective devices of the electrical grid. The software model generally receives one or more inputs described herein (e.g., LVRT voltage, pre-LVRT currents, current limitation strategy, K-factor, and the like) and outputs three-phase currents that would be generated by a code-compliant inverter.
Although the foregoing has been described with reference to certain specific embodiments, various modifications thereto will be apparent to those skilled in the art without departing from the spirit and scope of the invention as outlined in the appended claims. The entire disclosures of all references recited above are incorporated herein by reference.
Filing Document | Filing Date | Country | Kind |
---|---|---|---|
PCT/CA2022/051288 | 8/25/2022 | WO |
Number | Date | Country | |
---|---|---|---|
63264375 | Nov 2021 | US | |
63241793 | Sep 2021 | US |