Processing method and processing system for processing measurements acquired in electrical network, and system comprising electrical network

Abstract

A processing method and a processing system operative for processing measurements acquired in an electrical grid (11) are provided. The processing system (20) is operative to: determine a perturbation parameter; introducing disturbance into the power grid according to the disturbance parameters; and determining, based on the measurement, at least one response parameter that affects a grid frequency response of at least a portion of the grid to further disturbances.

Description

TECHNICAL FIELD

[0001] Embodiments of the invention relate to systems and methods operating on processing measurements acquired in an electrical grid. Embodiments of the invention relate in particular to such processing systems and methods operating on determining electrical grid frequency change characteristics in response to a disturbance. BACKGROUND

[0002] Electrical grids are important infrastructure components. Most installed electrical grids are built as alternating current (AC) electrical grids with an electrical grid frequency. The electrical grid frequency can differ from one electrical grid to another, but is typically set to have a prescribed value for any region in which it is deployed.

[0003] Disturbances can cause changes in the electrical grid frequency. Such changes in the electrical grid frequency can be problematic and can affect the stability, efficiency and safety of the electrical grid. Such disturbances can be caused by, for example, disconnection of tie lines to other network regions or main generators or load rejection.

[0004] After a disturbance that shifts the balance between mechanical and electrical energy in the electrical grid, the electrical grid frequency shows an initial shift from its nominal value, which is then typically recovered to close to its nominal value through a combination of automatic control operations and active rescheduling of remaining power generation. The dynamics of this change in electrical grid frequency as a function of time is relevant and it is desirable to reduce the change in electrical grid frequency as much as possible. The severity of a given disturbance on the electrical grid frequency and thus the additional risk to power system operation can be quantified by parameters such as the rate of change of frequency (RoCoF) or the nadir of the electrical grid frequency. The electrical grid response to a disturbance depends on the balance between electrical power and mechanical power, the latter typically stored in rotating components such as generators or applicable flywheel storage systems.

[0005] US 10 218 181 B2, US 10 997 674 B2 and US 11 112 438 B2 disclose exemplary techniques that can be used to examine electrical grid frequency response. These techniques employ actively introducing power into the electrical grid in a known pattern and evaluating the response. There is still a need for more versatile techniques for determining electrical grid frequency response parameters, for example techniques that can be implemented even in the case of lack of phasor measurement unit (PMU) instrumentation in multiple relevant parts of the electrical grid and / or that are able to determine electrical grid frequency response parameters with higher accuracy. There is also a need for improved techniques that provide further automation in the field of determining electrical grid response parameters. SUMMARY

[0006] It is an object of the present invention to provide methods, devices and / or systems that provide enhanced techniques for determining grid frequency response parameters that influence grid frequency changes to disturbances. It is a particular object of the present invention to provide methods, devices and / or systems that address one or more of the above needs. It is an optional object of the present invention to provide methods, devices and / or systems that can be implemented even when phasor measurement unit (PMU) instrumentation is lacking in at least some parts of the grid and / or that allow for determining grid frequency response parameters with higher accuracy.

[0007] According to exemplary embodiments, the methods and systems recited in the claims are provided.

[0008] According to an aspect of the present invention, a method of processing measurements taken in a grid is provided. The method is performed by a processing system and comprises causing a disturbance to be introduced into the grid in dependence on a disturbance parameter determined by the processing system, receiving data comprising frequency data, wherein the frequency data is indicative of grid frequencies measured at one or more locations in the grid, and processing the received data to determine, for at least a portion of the grid, at least one response parameter that influences a grid frequency response of the at least a portion of the grid to a further disturbance.

[0009] A variety of effects and advantages are achievable by the method. The method enables disturbances to be introduced in a controllable manner, facilitating a high accuracy determination of the at least one response parameter.

[0010] The method can be a method of processing measurements for determining an inertia and / or a response parameter that influences a frequency change of the grid to a further disturbance.

[0011] Thereby, response parameters can be determined that are particularly useful for improving frequency stability and / or for assessing an influence of a further disturbance on the grid frequency before the further disturbance occurs.

[0012] The method can further comprise setting a disturbance frequency of the disturbance based on an electrical resonance and / or a mechanical resonance of the grid.

[0013] Thereby, undesirable resonance excitation can be avoided. The at least one response parameter can be determined with higher accuracy.

[0014] The disturbance frequency can be set to prevent an excitation of an electrical resonance and / or a mechanical resonance of the grid, and / or the disturbance frequency is set to space the disturbance frequency from any of the electrical resonance and / or the mechanical resonance of the grid by at least a threshold value.

[0015] Thereby, undesirable resonance excitation can be avoided. The at least one response parameter can be determined with higher accuracy.

[0016] The disturbance being introduced into the power grid can comprise controlling at least one first device comprised by or coupled to the power grid to introduce the disturbance.

[0017] The disturbance being introduced into the power grid can comprise controlling at least one device comprised by or coupled to the power grid to introduce the disturbance and to reduce the impact of the disturbance on the voltage magnitude of the power grid.

[0018] Thereby, undesired variations of the voltage magnitude can be reduced, thereby mitigating potential negative effects on loads connected to the power grid, while allowing for a more accurate determination of the at least one response parameter.

[0019] The at least one device controlled by the processing system can comprise a device operating to influence the reactive power in the power grid.

[0020] Thereby, undesired variations of the voltage magnitude can be reduced, thereby mitigating potential negative effects on loads connected to the power grid, while allowing for a more accurate determination of the at least one response parameter.

[0021] The disturbance being introduced into the power grid can comprise controlling at least one first device comprised by or coupled to the power grid to introduce the disturbance, and controlling at least one second device comprised by or coupled to the power grid to reduce the impact of the disturbance on the voltage magnitude of the power grid. The at least one second device can comprise a device operating to influence the reactive power in the power grid, such as a reactance.

[0022] The processing of the received data can comprise determining the at least one response parameter based on a total power deviation, wherein the total power deviation comprises the power of the introduced disturbance and a change in power in response to the introduced disturbance.

[0023] Thereby, changes in load power and / or loss power caused by the disturbance can be taken into account, thereby making the determination of the response parameter more accurate. By determining the at least one response parameter based on the change in power, the method can be used even in cases where changes in load power and / or loss power caused by the disturbance cannot be measured by a respective PMU.

[0024] The at least one response parameter can comprise a parameter of a generator equivalent model. The at least one response parameter can comprise a parameter of a single generator equivalent model of a generator in the power grid or a portion of the power grid. The determined at least one response parameter can comprise one, more or all of: an inertia, a damping, a governor time constant, a governor gain, a load sensitivity to voltage magnitude.

[0025] Thereby, one or more response parameters can be determined which contribute to assessing and potentially mitigating future changes in the power grid frequency in response to further disturbances.

[0026] Determining the at least one response parameter can comprise determining parameters of a time-discrete response function. The time-discrete response function can establish a relationship between a variation of the grid frequency (the variation of the grid frequency being time-dependent) in a time interval following an initial time of the disturbance and a variation of a total power (the variation of the total power also being time-dependent) in the time interval following the initial time of the disturbance, the total power being defined as a sum of a disturbance power, a load power and a loss power.

[0027] Determining the at least one response parameter can comprise determining parameters of a power swing equation relating a variation of the grid frequency (the variation of the grid frequency being time-dependent) in a time interval following an initial time of the disturbance and a variation of a total power (the variation of the total power also being time-dependent) in the time interval following the initial time of the disturbance, the total power being defined as a sum of a disturbance power, a load power and a loss power.

[0028] Thereby, the at least one response parameter can be determined efficiently using, for example, linear response theory.

[0029] The method can further comprise repeating the steps of introducing a disturbance, receiving data and processing the received data for determining the at least one response parameter for each of two or more regions in the grid.

[0030] Thereby, the inertia and / or other response parameters can be determined individually for each of the two or more regions.

[0031] Processing the received data can comprise determining a variation of the load power and / or a variation of the loss power caused by the disturbance, wherein, optionally, processing the received data comprises determining the variation of the load power based on voltage magnitude measurements, and / or wherein, optionally, processing the received data comprises determining the variation of the loss power based on a state estimation.

[0032] The response parameter can comprise at least an inertia of a generator equivalent model, such as an inertia of a single generator equivalent model in the grid or a part thereof.

[0033] Thereby, the response parameter can be determined based on, for example, a power swing equation, which is particularly useful for determining a variation of the grid frequency.

[0034] The response parameter can comprise a parameter depending on a balance between mechanical power and electrical power in the grid, such as an inertia of a generator equivalent model.

[0035] Thereby, the response parameter can be determined which is particularly useful for determining a variation of the grid frequency caused by a deviation from a balance between mechanical power and electrical power in the grid.

[0036] The change in power can comprise a total power deviation. The total power deviation can comprise the disturbance power and a change in power in response to the disturbance in at least a portion of the power grid.

[0037] By considering the total power deviation comprising both the disturbance power itself and the change in power as a reaction to the disturbance, the response parameters (such as the inertia of the individual generator equivalent model) can be determined more accurately compared to techniques that only consider the disturbance power itself.

[0038] The total power deviation can comprise the power of the disturbance and a change in load power and / or loss power caused by the disturbance.

[0039] By considering the change in load power and / or loss power caused by the disturbance, the response parameters (such as the inertia of the individual generator equivalent model) can be determined more accurately compared to techniques that only consider the disturbance power itself.

[0040] Processing the received data can comprise determining the change in load power and / or loss power caused by the disturbance. Determining the change in load power and / or loss power caused by the disturbance can comprise calculating the change in load power and / or loss power from PMU measurements comprised in the received data.

[0041] Thereby, the change in load power and / or loss power caused by the disturbance can be determined efficiently. The accuracy in determining the at least one response parameter is enhanced.

[0042] Alternatively or additionally, determining the change in load power and / or loss power caused by the disturbance can comprise calculating the change in load power and / or loss power from voltage measurements comprised in the received data.

[0043] Thereby, the change in load power and / or loss power caused by the disturbance can be determined even in case PMU measurements are not available. The accuracy in determining the at least one response parameter is enhanced.

[0044] The method can further comprise storing the at least one response parameter in a control, monitoring and / or protection device or system for use in assessing the power grid frequency response to further disturbances.

[0045] This facilitates taking appropriate actions to stabilize the power grid frequency in the future and / or to provide information (such as alarms and / or warnings) in case the planned generator unit commitment is likely to result in unacceptable magnitudes of the power grid frequency (e.g. unacceptable frequency nadirs) and / or unacceptable rates of change of the power grid frequency (e.g. unacceptable rates of change of frequency (RoCoF)).

[0046] The method can further comprise using, by the processing system performing the method, the at least one response parameter to assess the grid frequency response to a further disturbance. The assessed grid frequency response can comprise a rate of change of frequency, a RoCoF and / or a frequency nadir.

[0047] This facilitates taking appropriate action to stabilise the grid frequency in the future and / or to provide information such as an alarm and / or a warning in the event that the planned generator unit commitment is likely to result in an unacceptable magnitude of grid frequency (and / or an unacceptable rate of change of grid frequency).

[0048] The method can further comprise implementing, by the processing system, an action based on the at least one response parameter. The action can comprise a corrective action and / or a mitigation action to reduce the change in grid frequency brought about by the further disturbance.

[0049] The method thereby facilitates maintaining grid frequency stability.

[0050] A grid control, protection and / or monitoring method according to an embodiment comprises using, by a control, protection and / or monitoring device or system, at least one response parameter to assess the impact of a further disturbance on grid frequency, and performing a corrective action and / or a mitigation action based on using the at least one response parameter determined according to the method in the above disclosed aspects or embodiments.

[0051] According to another aspect of the application, there is provided a method for processing measurements taken in a grid. The processing system comprises an interface; and at least one processing circuitry operative to determine disturbance parameters of a disturbance to be generated; generate a command and output the command via the interface to cause the disturbance to be introduced into the grid in accordance with the determined disturbance parameters; receive data comprising frequency data, wherein the frequency data is representative of a grid frequency measured at one or more locations in the grid; and process the received data to determine, for at least a portion of the grid, at least one response parameter affecting a grid frequency response of the at least a portion of the grid to a further disturbance.

[0052] A number of effects and advantages are obtained by the processing system. The processing system allows for introducing a disturbance in a controlled manner, facilitating determining the at least one response parameter with high accuracy.

[0053] The processing system can be operative to perform the method in any one of the above disclosed aspects or embodiments.

[0054] According to another aspect of the application, there is provided a system comprising: a grid; a processing system of the aspects or embodiments disclosed herein, the processing system being operative to determine at least one response parameter.

[0055] The system can further comprise a control, monitoring and / or protection device or system that is operable to perform an operation based on the at least one response parameter to respond to a further disturbance.

[0056] The control, monitoring and / or protection device or system can be operable to perform an operation based on the at least one response parameter to keep a change in the grid frequency in response to a future disturbance below a frequency change threshold.

[0057] The operation performed by the control, monitoring and / or protection device or system can be a change in a generator unit commitment for one or more generator units.

[0058] The system can further comprise a measurement instrument operable to provide some or all of the data to the processing system. The measurement instrument can comprise a PMU. The measurement instrument can comprise a voltage measurement instrument.

[0059] Thus, by using a change in power (e.g. a sum of disturbance power, loss power and load power) to determine the at least one response parameter, the at least one response parameter can be determined based on measurements.

[0060] According to a further embodiment, there is provided machine readable instruction code which, when executed by at least one programmable circuit, causes the at least one programmable circuit to perform a method according to an embodiment.

[0061] According to a further embodiment, there is provided a non-transitory storage medium having stored thereon machine readable instruction code which, when executed by at least one programmable circuit, causes the at least one programmable circuit to perform a method according to an embodiment.

[0062] According to another aspect of the application (which can be used in combination with the preceding aspects and embodiments or independently thereof), there is provided a method of processing measurements taken in a power grid. The method comprises receiving data comprising frequency data. The frequency data is indicative of a grid frequency measured at one or more locations in the power grid in response to a disturbance. The method comprises processing the received data to determine, for at least a portion of the power grid, at least one response parameter that influences a grid frequency response of the at least a portion of the power grid to a further disturbance. The at least one response parameter is determined based on a change in power in response to the disturbance.

[0063] As used herein, a "change in power in response to a disturbance" includes a change caused by the disturbance power and distinct from the introduced disturbance power. Examples of such a change in power in response to a disturbance are a change in load power, a change in loss power and / or a change in tie voltage power caused by the disturbance but distinct from the disturbance itself.

[0064] Various effects and advantages can be obtained by the method. The method is versatile in the sense that it can be operated to determine grid frequency response parameters both in case the disturbance is actively introduced for determining the grid frequency response parameters (active method) and in case the disturbance is caused by a load, generator or other device connected to or comprised by the grid that consumes or produces power (passive method). By determining the at least one response parameter based on the power variation in response to the disturbance, the variation of the load power and / or the loss power caused by the disturbance can be taken into account, thereby making the determination of the response parameter more accurate. By determining the at least one response parameter based on the power variation in response to the disturbance, the method can even be used in case the variation of the load power and / or the loss power caused by the disturbance cannot be directly measured by the respective PMU.

[0065] As used herein, the "power variation in response to the disturbance" comprises a variation caused by the disturbance power and different from the introduced disturbance power. Examples of such power variation in response to the disturbance are a variation of the load power, a variation of the loss power and / or a variation of the tie voltage power caused by the disturbance but different from the disturbance itself.

[0066] The method can be performed by or using a processing system. Thus, the response parameter or the plurality of response parameters can be determined automatically or semi-automatically.

[0067] The method can be a method of processing measurements to determine an inertia and / or a response parameter that influences the frequency variation of the grid to a further disturbance.

[0068] Thereby, response parameters can be determined that are particularly useful for improving the frequency stability and / or for assessing the impact of a further disturbance on the grid frequency before the further disturbance occurs.

[0069] The response parameter can comprise an inertia of at least a generator equivalent model, such as an inertia of a single generator equivalent model of the grid or a part thereof.

[0070] Thereby, response parameters can be determined that are particularly useful for determining a grid frequency variation based on, for example, a power swing equation.

[0071] The response parameter can comprise a parameter that depends on a balance between mechanical power and electrical power in the grid, such as an inertia of a generator equivalent model.

[0072] Thereby, response parameters can be determined that are particularly useful for determining a grid frequency variation caused by a shift of the balance between mechanical power and electrical power in the grid.

[0073] The power variation can comprise a total power deviation. The total power deviation can comprise the disturbance power and a power variation in response to the disturbance in at least a portion of the power grid.

[0074] By considering the total power deviation comprising both the disturbance power itself and the power variation as a reaction to the disturbance, the response parameters, such as the inertia of the individual generator equivalent model, can be determined more accurately compared to techniques considering only the disturbance power itself.

[0075] The total power deviation can comprise the power of the disturbance and a change in load power and / or loss power caused by the disturbance.

[0076] By considering the change in load power and / or loss power caused by the disturbance, the response parameters, such as the inertia of the individual generator equivalent model, can be determined more accurately compared to techniques considering only the disturbance power itself.

[0077] Processing the received data can comprise determining the change in load power and / or loss power caused by the disturbance. Determining the change in load power and / or loss power caused by the disturbance can comprise calculating the change in load power and / or loss power from PMU measurements comprised in the received data.

[0078] Thereby, the change in load power and / or loss power caused by the disturbance can be determined efficiently. The accuracy in determining the at least one response parameter is enhanced.

[0079] Alternatively or additionally, determining the change in load power and / or loss power caused by the disturbance can comprise calculating the change in load power and / or loss power from voltage measurements comprised in the received data.

[0080] Thereby, the change in load power and / or loss power caused by the disturbance can be determined even in case PMU measurements are not available. The accuracy in determining the at least one response parameter is enhanced.

[0081] Calculating the change in load power and / or loss power from the voltage measurements can comprise determining a proportionality constant between a voltage variation obtained from the voltage measurements and a load power variation.

[0082] Thereby, the change in load power caused by the disturbance can be determined efficiently even in case no PMU measurements are available. By considering the change in load power caused by the disturbance determined thereby, the accuracy in determining the at least one response parameter is enhanced.

[0083] Determining a proportionality constant between the voltage variation and the load power variation can comprise identifying from a set of proportionality constants the proportionality constant that minimizes a metric quantifying a temporal correlation deviation between the measured grid frequency variation as a function of time and the calculated determined grid frequency variation as a function of time, wherein the calculated determined grid frequency variation is obtained from a linear dependence of the voltage measurements on the variation of the load power.

[0084] Thereby, the variation of the load power caused by the disturbance can be effectively determined even without PMU measurements. By taking into account the thus determined variation of the load power caused by the disturbance, the accuracy in determining the at least one response parameter is enhanced.

[0085] Alternatively or additionally, processing the received data can further comprise determining a variation of the load power and / or the loss power based on the state estimation.

[0086] Thereby, the variation of the load power and / or the loss power caused by the disturbance can be determined and taken into account when determining the at least one response parameter. By taking into account the thus determined variation of the load power and / or the loss power caused by the disturbance, the accuracy in determining the at least one response parameter is enhanced.

[0087] Processing the received data can comprise determining the power variation in a time-dependent manner.

[0088] Thereby, the time-dependent variation of the grid frequency can be combined with the time-dependent variation of the power (e.g. the variation of the sum of the disturbance power, the loss power and the load power). A more accurate determination of the at least one response parameter (e.g. a parameter of a single generator equivalent, such as the inertia) is possible.

[0089] Processing the received data can comprise processing the time-dependent variation of the power as indicated by the measurements and the time-dependent variation of the grid frequency.

[0090] Thereby, the time-dependent variation of the grid frequency can be combined with the time-dependent variation of the power (e.g. the variation of the sum of the disturbance power, the loss power and the load power). A more accurate determination of the at least one response parameter (e.g. a parameter of a single generator equivalent, such as the inertia) is possible.

[0091] The variation of the power can be determined as time series data or non-periodically repeating data. The variation of the frequency can be time series data or non-periodically repeating data.

[0092] Thereby, by taking into account the dynamics of the variation of the power (e.g. the variation of the sum of the disturbance power, the load power and the loss power) and the synchronous dynamics of the grid frequency as a function of time, the at least one response parameter can be effectively determined.

[0093] The method can comprise obtaining disturbance magnitude and disturbance time information. The at least one response parameter can be further determined based on the disturbance magnitude (e.g. disturbance power) and the disturbance time information (e.g. initial time or an estimate of the initial time).

[0094] This enables the at least one response parameter to be determined based on the known characteristics of the grid's response to the disturbance.

[0095] Different processing can be employed depending on whether the disturbance is a disturbance that is actively introduced into the grid for the purpose (e.g. the only purpose) of determining the at least one response parameter (active way of determining the at least one response parameter), or whether the disturbance is caused by one of e.g. a load (such as an electric arc furnace) or a generator (passive way of determining the at least one response parameter).

[0096] For the active way, the disturbance magnitude and the disturbance time information (e.g. initial time and / or time dependency of the disturbance power) can be determined by the processing system before the disturbance is introduced into the grid. It can not be necessary (but still possible) to also rely on measurements to determine the disturbance power. Measurements are used to determine the change in load power and / or loss power caused by the disturbance.

[0097] For the passive way, the disturbance magnitude and the disturbance time information can be determined (e.g. estimated) by processing measurements included in the received data.

[0098] The processing system can be capable of operating according to both the active way and according to the passive way. The processing system can switch its mode of operation depending on whether the at least one response parameter is determined according to the active way or according to the passive way.

[0099] The time information can comprise an initial time of the disturbance or a time series of at least one electrical characteristic caused by the disturbance.

[0100] Thereby, the change in load power and / or loss power can be associated with the time at which the disturbance occurred to help identify the change in load power and / or loss power caused by the disturbance.

[0101] The received data can comprise an output of a measurement instrument, wherein the measurement instrument is coupled to the grid. The measurement instrument can comprise a PMU. The measurement instrument can comprise a voltage measurement instrument. The measurement instrument can comprise a measurement instrument capable of measuring the grid frequency at one or more locations of the grid.

[0102] Thereby, the measurements required for determining the inertia or other response parameter that determines the change in grid frequency caused by a further disturbance are obtained without necessarily requiring (but still allowing) full PMU observability.

[0103] The method can further comprise causing the disturbance to be introduced into the power grid.

[0104] Thereby, the disturbance can be introduced actively for the purpose of determining the at least one response parameter, e.g. for this sole purpose.

[0105] The method can further comprise setting a disturbance frequency of the disturbance based on an electrical resonance and / or a mechanical resonance of the power grid.

[0106] Thereby, undesired resonance excitation can be avoided. The at least one response parameter can be determined more accurately.

[0107] The disturbance frequency can be set to prevent excitation of an electrical resonance and / or a mechanical resonance of the power grid.

[0108] Thereby, undesired resonance excitation can be avoided. The at least one response parameter can be determined more accurately.

[0109] Causing the disturbance to be introduced into the power grid can comprise controlling at least one device comprised by or coupled to the power grid to introduce the disturbance and to reduce the impact of the disturbance on a voltage magnitude of the power grid.

[0110] Thereby, undesired variations of the voltage magnitude can be reduced, thereby mitigating potential negative effects on loads connected to the power grid, while allowing the at least one response parameter to be determined more accurately.

[0111] The at least one device controlled by the processing system can comprise a device operating to affect a reactive power in the power grid.

[0112] Thereby, undesired variations of the voltage magnitude can be reduced, thereby mitigating potential negative effects on loads connected to the power grid, while allowing the at least one response parameter to be determined more accurately.

[0113] Causing the disturbance to be introduced into the power grid can comprise controlling at least a first device comprised by or coupled to the power grid to introduce the disturbance and controlling at least a second device comprised by or coupled to the power grid to reduce the impact of the disturbance on a voltage magnitude of the power grid. The at least one second device can comprise a device operating to affect a reactive power in the power grid, such as a reactance.

[0114] The at least one response parameter can comprise a parameter of a generator equivalent model. The at least one response parameter can comprise a parameter of an individual generator equivalent model of a generator in the power grid or a portion of the power grid. The determined at least one response parameter can comprise one, more or all of the following parameters: an inertia, a damping, a governor time constant, a governor gain, a load sensitivity to a voltage magnitude.

[0115] Thereby, one or more response parameters can be determined which are helpful to assess and potentially mitigate future changes of a power grid frequency in response to further disturbances.

[0116] Determining the at least one response parameter can comprise determining parameters of a time-discrete response function. The time-discrete response function can establish a relationship between a variation of the grid frequency (the variation of the grid frequency being time-dependent) over a time interval following an initial time of the disturbance and a variation of a total power (the variation of the total power also being time-dependent) over the time interval following the initial time of the disturbance, the total power being defined as a sum of a disturbance power, a load power and a loss power.

[0117] Determining the at least one response parameter can comprise determining parameters of a power swing equation, the equation relating a variation of the grid frequency (the variation of the grid frequency being time-dependent) over a time interval following an initial time of the disturbance and a variation of a total power (the variation of the total power also being time-dependent) over the time interval following the initial time of the disturbance, the total power being defined as a sum of a disturbance power, a load power and a loss power.

[0118] Thereby, the at least one response parameter can be determined efficiently using, for example, linear response theory.

[0119] The method can further comprise repeating the steps of receiving data and processing the received data for two or more regions of the grid to determine the at least one response parameter for each of the two or more regions.

[0120] Thereby, the inertia and / or other response parameters can be determined individually for each of the two or more regions.

[0121] The method can further comprise storing the at least one response parameter in a control, monitoring and / or protection device or system for use in assessing a grid frequency response to a further disturbance.

[0122] This facilitates taking appropriate action to stabilize the grid frequency in the future and / or to provide information (such as alarms and / or warnings) in case a planned generator unit commitment is likely to result in an unacceptable magnitude of the grid frequency (e.g. an unacceptable frequency nadir) and / or an unacceptable rate of change of the grid frequency (e.g. an unacceptable rate of change of frequency (RoCoF)).

[0123] The method can further comprise assessing a grid frequency response to a further disturbance using the at least one response parameter by a processing system executing the method. The assessed grid frequency response can comprise a rate of change of frequency, a RoCoF and / or a frequency nadir.

[0124] This facilitates taking appropriate action to stabilize the grid frequency in the future and / or to provide information such as alarms and / or warnings in case the planned generator unit commitment can lead to unacceptable magnitudes of the grid frequency (and / or unacceptable rates of change of the grid frequency).

[0125] The method can further comprise implementing, by the processing system, an operation based on the at least one response parameter. The operation can comprise a corrective operation and / or a mitigation operation to reduce the change in grid frequency brought about by the further disturbance.

[0126] Thereby, the method facilitates maintaining the stability of the grid frequency.

[0127] According to another aspect of the present invention, there is provided a grid control, monitoring and / or protection method. The grid control, monitoring and / or protection method comprises determining, by a control, monitoring and / or protection device or system, at least one response parameter using a method as disclosed in the above-mentioned aspects or embodiments; and performing an operation based on the at least one response parameter to reduce the change in grid frequency brought about by a further disturbance.

[0128] Thereby, the method facilitates maintaining the stability of the grid frequency.

[0129] According to another aspect of the present invention, there is provided a processing system for processing measurements taken in a grid, the processing system comprising at least one processing circuitry operative to: receive data comprising frequency data, wherein the frequency data is indicative of a grid frequency measured at one or more locations in the grid in response to a disturbance; and process the received data to determine, for at least a portion of the grid, at least one response parameter affecting the grid frequency response of the at least a portion of the grid to a further disturbance, wherein the at least one processing circuitry is operative to determine the at least one response parameter based on a change in power in response to the disturbance.

[0130] Various effects and advantages are obtained by this processing system. The processing system is versatile in the sense that it can be used both in cases where the disturbance is actively introduced to determine the grid frequency response parameter (active approach) and in cases where the disturbance is caused by a load connected to or comprised by the grid (passive approach). By determining the at least one response parameter based on the change in power, the processing system can take into account the change in load power and / or loss power caused by the disturbance, thereby making the determination of the response parameter more accurate. By determining the at least one response parameter based on the change in power, the processing system can be used even in cases where the change in load power and / or loss power caused by the disturbance cannot be measured by a corresponding PMU.

[0131] The processing system can be operable to process the measurements to determine inertia and / or response parameters that affect the change in frequency of the power grid to further disturbances.

[0132] Thereby, the processing system can determine response parameters that are particularly useful for improving the frequency stability and / or for assessing the impact of further disturbances on the frequency of the power grid before the further disturbances occur.

[0133] The processing system can be operable such that the determined response parameters comprise at least one inertia of a generator equivalent model, such as an inertia of a single generator equivalent model of the power grid or a portion thereof.

[0134] Thereby, the processing system can determine response parameters that are particularly useful for determining the change in frequency of the power grid based on, for example, the power swing equation.

[0135] The processing system can be operable such that the determined response parameters comprise parameters that depend on a balance between mechanical power and electrical power in the power grid, such as an inertia of a generator equivalent model.

[0136] Thereby, the processing system can determine response parameters that are particularly useful for determining the change in frequency of the power grid caused by a shift in the balance between mechanical power and electrical power in the power grid.

[0137] The change in power can comprise a total power deviation. The total power deviation can comprise the disturbance power and a change in power that is a response to the disturbance in at least a portion of the power grid.

[0138] By taking into account the total power deviation, which comprises the disturbance power itself and the change in power as a reaction to the disturbance, the processing system can more accurately determine the response parameters, such as the inertia of a single generator equivalent model, compared to techniques that only take into account the disturbance power itself.

[0139] The processing system can be operable such that the total power deviation can comprise the power of the disturbance and a change in load power and / or loss power caused by the disturbance.

[0140] By taking into account the change in load power and / or loss power caused by the disturbance, the processing system can more accurately determine the response parameters, such as the inertia of a single generator equivalent model, compared to techniques that only take into account the disturbance power itself.

[0141] The processing system can be operable to determine the change in load power and / or loss power caused by the disturbance based on the received data. The processing system can be operable such that determining the change in load power and / or loss power caused by the disturbance comprises calculating the change in load power and / or loss power from PMU measurements comprised in the received data.

[0142] Thereby, the processing system can efficiently determine the change in load power and / or loss power caused by the disturbance. The accuracy in determining the at least one response parameter is enhanced.

[0143] Alternatively or additionally, the processing system can be operative such that determining the change in load power and / or loss power caused by the disturbance can comprise calculating the change in load power and / or loss power from the voltage measurements comprised in the received data.

[0144] Thereby, the processing system can efficiently determine the change in load power and / or loss power caused by the disturbance. The accuracy in determining the at least one response parameter is enhanced.

[0145] The processing system can be operative such that calculating the change in load power and / or loss power from the voltage measurements comprises determining a proportionality constant between a change in voltage as obtained from the voltage measurements and a change in load power.

[0146] Thereby, the processing system can efficiently determine the change in load power caused by the disturbance, even in the absence of PMU measurements. By taking into account the change in load power caused by the disturbance thus determined, the accuracy in determining the at least one response parameter is enhanced.

[0147] The processing system can be operative such that determining a proportionality constant between a change in voltage and a change in load power can comprise determining a proportionality constant from a set of proportionality constants that minimizes a metric quantifying a temporal correlation deviation between a measured change in grid frequency as a function of time and a calculated determined change in grid frequency as a function of time, wherein the calculated determined change in grid frequency is obtained from a linear dependence of the change in load power on the voltage measurements.

[0148] Thereby, the processing system can efficiently determine the change in load power caused by the disturbance, even in the absence of PMU measurements. By taking into account the change in load power caused by the disturbance thus determined, the accuracy in determining the at least one response parameter is enhanced.

[0149] Alternatively or additionally, based on the received data, the processing system can be operative to determine the change in load power and / or loss power based on a state estimation.

[0150] Thereby, the processing system can determine the change in load power and / or loss power caused by the disturbance and take it into account in determining the at least one response parameter. By taking into account the change in load power and / or loss power caused by the disturbance thus determined, the accuracy in determining the at least one response parameter is enhanced.

[0151] The processing system can be operative to determine the change in power in a time-dependent manner.

[0152] Thereby, the processing system can combine (e.g., correlate) the time-dependent change of the grid frequency with the time-dependent change of the power (e.g., the change of the sum of the disturbance power, the loss power and the load power). The at least one response parameter (e.g., a parameter of a single generator equivalent, such as an inertia) can be determined more accurately.

[0153] The processing system can be operable to process both the time-dependent change of the power as indicated by the measurements and the time-dependent change of the grid frequency.

[0154] Thereby, the processing system can combine (e.g., correlate) the time-dependent change of the grid frequency with the time-dependent change of the power (e.g., the change of the sum of the disturbance power, the loss power and the load power). The at least one response parameter (e.g., a parameter of a single generator equivalent, such as an inertia) can be determined more accurately.

[0155] The processing system can be operable to determine the change of the power as time series data or as non-periodically repeating data. The processing system can be operable to determine the change of the frequency as time series data or as non-periodically repeating data.

[0156] Thereby, by taking into account the dynamic nature of the change of the power (e.g., the change of the sum of the disturbance power, the load power and the loss power) as a function of time and the synchronous dynamic nature of the change of the grid frequency, the processing system can effectively determine the at least one response parameter.

[0157] The processing system can be operable to obtain disturbance magnitude and disturbance time information. The processing system can be operable to determine the at least one response parameter further based on the disturbance magnitude (e.g., the disturbance power) and the disturbance time information (e.g., the initial time or an estimate of the initial time).

[0158] This enables the processing system to determine the at least one response parameter based on the known behavior of the grid in response to the disturbance.

[0159] Depending on whether the disturbance is a disturbance that is actively introduced into the grid for the purpose (e.g., the only purpose) of determining the at least one response parameter (active way of determining the at least one response parameter) or whether the disturbance is caused by one of the loads (such as an electric arc furnace) (passive way of determining the at least one response parameter), the processing system can be operable to perform different processing.

[0160] For the active approach, the processing system can be operable to determine the disturbance magnitude and disturbance time information (e.g. initial time and / or time dependency of the disturbance power) before the disturbance is introduced into the power grid. It can also be possible (but still possible) to also rely on measurements to determine the disturbance power. The processing system can be operable to use the measurements to determine the change in load power and / or loss power caused by the disturbance.

[0161] For the passive approach, the processing system can be operable to determine (e.g. estimate) the disturbance magnitude and disturbance time information by processing the measurements comprised in the received data.

[0162] The processing system can be operable to operate according to both the active approach and according to the passive approach. The processing system can switch its mode of operation depending on whether the at least one response parameter is determined according to the active approach or according to the passive approach.

[0163] The processing system can be operable such that the time information can comprise an initial time of the disturbance or a time sequence of at least one electrical property caused by the disturbance.

[0164] Thereby, the processing system can correlate the change in load power and / or loss power with the time at which the disturbance occurred to help identify the change in load power and / or loss power caused by the disturbance.

[0165] The processing system can be operable such that the received data comprises an output of a measurement instrument coupled to the power grid. The measurement instrument can comprise a PMU. The measurement instrument can comprise a voltage measurement instrument. The measurement instrument can comprise a measurement instrument capable of measuring the power grid frequency at one or more locations of the power grid.

[0166] Thereby, the processing system can obtain the measurements required for determining the inertia or other response parameter determining the change in power grid frequency caused by a further disturbance without necessarily requiring (but still allowing) full PMU observability.

[0167] The processing system can be operable to cause the disturbance to be introduced into the power grid.

[0168] Thereby, the processing system can actively introduce the disturbance for the purpose of determining the at least one response parameter (e.g. for this sole purpose).

[0169] The processing system can be operable to set the disturbance frequency of the disturbance based on an electrical resonance and / or a mechanical resonance of the power grid.

[0170] Thereby, undesirable resonance excitation can be avoided. The at least one response parameter can be determined more accurately.

[0171] The processing system can be operable to set the disturbance frequency to prevent excitation of an electrical resonance and / or a mechanical resonance of the power grid.

[0172] Thereby, the processing system can avoid undesired excitation of a resonance. The at least one response parameter can be determined more accurately.

[0173] The processing system can be operable such that, in order to introduce the disturbance into the power grid, the processing system operates to control at least one device comprised by or coupled to the power grid to introduce the disturbance and to reduce the effect of the disturbance on the voltage magnitude of the power grid.

[0174] Thereby, undesired variations of the voltage magnitude can be reduced, thereby mitigating potential negative effects on loads connected to the power grid, while allowing the at least one response parameter to be determined more accurately.

[0175] The at least one device controlled by the processing system can comprise a device operating to influence the reactive power in the power grid.

[0176] Thereby, undesired variations of the voltage magnitude can be reduced, thereby mitigating potential negative effects on loads connected to the power grid, while allowing the at least one response parameter to be determined more accurately.

[0177] The processing system can be operable such that, in order to cause the disturbance to be introduced into the power grid, the processing system operates to control at least a first device comprised by or coupled to the power grid to introduce the disturbance, and to control at least a second device comprised by or coupled to the power grid to reduce the effect of the disturbance on the voltage magnitude of the power grid. The at least one second device can comprise a device operating to influence the reactive power in the power grid, such as a reactance.

[0178] The processing system can be operable such that the at least one response parameter determined by the processing system comprises a parameter of a generator equivalent model. The processing system can be operable such that the at least one response parameter determined by the processing system comprises a parameter of a single generator equivalent model of a generator of the power grid or of a part of the power grid. The processing system can be operable such that the at least one response parameter determined by the processing system comprises one, more or all of the following parameters: an inertia, a damping, a governor time constant, a governor gain, a load sensitivity to the voltage magnitude.

[0179] Thereby, one or more response parameters can be determined which facilitate assessing and potentially mitigating future changes of the power grid frequency in response to further disturbances.

[0180] The processing system can be operable to determine parameters of a time-discrete response function for determining the at least one response parameter. The time-discrete response function can establish a relationship between a change of the power grid frequency (the change of the power grid frequency being time-dependent) over a time interval following an initial time of the disturbance and a change of a total power (the change of the total power also being time-dependent) defined as a sum of a disturbance power, a load power and a loss power over the time interval following the initial time of the disturbance.

[0181] The processing system is operable such that, in determining the at least one response parameter, it determines a parameter of a power swing equation that relates a variation in grid frequency over a time interval following an initial time of a disturbance (the variation in grid frequency being time dependent) to a variation in total power (the variation in total power also being time dependent) over the time interval following the initial time, the total power being defined as a sum of a disturbance power, a load power and a loss power.

[0182] Thus, the at least one response parameter can be determined by the processing system using, for example, linear response theory.

[0183] The processing system is operable to repeat the receiving of data and processing of received data for two or more regions of the grid to determine the at least one response parameter for each of the two or more regions.

[0184] Thus, the processing system can determine the inertia and / or other response parameters separately for each of the two or more regions.

[0185] The processing system is operable to store the at least one response parameter in a control, monitoring and / or protection device or system for use in assessing the grid frequency response to a further disturbance.

[0186] This facilitates taking appropriate action to stabilise the grid frequency in future and / or to provide information such as alarms and / or warnings in the event that the planned generator unit commitment is likely to result in an unacceptable magnitude of grid frequency (e.g. an unacceptable frequency nadir) and / or an unacceptable rate of change of grid frequency (e.g. an unacceptable rate of change of frequency (RoCoF)).

[0187] The processing system is operable to use the at least one response parameter to assess the grid frequency response to a further disturbance. The assessed grid frequency response can comprise a rate of change of frequency, a RoCoF and / or a frequency nadir.

[0188] This facilitates taking appropriate action to stabilise the grid frequency in future and / or to provide information such as alarms and / or warnings in the event that the planned generator unit commitment is likely to result in an unacceptable magnitude of grid frequency (and / or an unacceptable rate of change of grid frequency).

[0189] The processing system is operable to implement an action based on the at least one response parameter. The action can comprise a corrective action and / or a mitigation action to reduce the variation in grid frequency brought about by a further disturbance.

[0190] Thus, the processing system facilitates maintaining grid frequency stability.

[0191] The processing system is operable to implement an operation based on the at least one response parameter, wherein the operation comprises changing a scheduled future generator unit commitment.

[0192] Thereby, the processing system facilitates maintaining grid frequency stability.

[0193] The processing system can be operable to automatically or semi-automatically perform a method of any of the aspects or embodiments disclosed herein.

[0194] According to another aspect of the application, there is provided a system comprising a power grid and a processing system of the aspects or embodiments disclosed herein, operable to determine at least one response parameter.

[0195] The system can further comprise a control, monitoring and / or protection device or system to perform an operation in response to a further disturbance based on the at least one response parameter.

[0196] The control, monitoring and / or protection device or system can be operable to perform an operation based on the at least one response parameter to maintain a change in grid frequency in response to a future disturbance below a frequency change threshold.

[0197] The operation performed by the control, monitoring and / or protection device or system can be to change a generator unit commitment for one or more generator units.

[0198] The system can further comprise a measurement instrument operable to provide some or all of the data to the processing system. The measurement instrument can comprise a PMU. The measurement instrument can comprise a voltage measurement instrument.

[0199] Thereby, by using a change in power (e.g. a total of disturbance power, loss power and load power) to determine the at least one response parameter, the at least one response parameter can be determined based on measurements.

[0200] A number of embodiments are provided by the following numbered list of embodiments: Embodiment 1 : A method for processing measurements taken in a power grid, the method performed by a processing system and comprising: causing a disturbance to be introduced into the power grid in dependence on a disturbance parameter determined by the processing system; receiving data comprising frequency data, wherein the frequency data is representative of a power grid frequency measured at one or more locations in the power grid; and processing the received data to determine at least one response parameter for at least a portion of the power grid, the at least one parameter influencing a power grid frequency response of the at least a portion of the power grid to a further disturbance.

[0201] Embodiment 2: The method of embodiment 1, further comprising setting a disturbance frequency of the disturbance based on an electrical and / or mechanical resonance of the power grid.

[0202] Embodiment 3: The method of embodiment 2, wherein the disturbance frequency is set to prevent exciting electrical and / or mechanical resonances of the power grid, and / or wherein the disturbance frequency is set such that the disturbance frequency is at least a threshold away from any of the electrical and / or mechanical resonances of the power grid.

[0203] Embodiment 4: The method of any of the preceding embodiments, wherein causing the disturbance to be introduced into the power grid comprises controlling at least one device comprised by or coupled to the power grid to introduce the disturbance.

[0204] Embodiment 5: The method of embodiment 4, wherein the at least one device is controlled to cause the disturbance to be introduced into the power grid in a repetitive manner.

[0205] Embodiment 6: The method of embodiment 4 or embodiment 5, wherein causing the disturbance to be introduced into the power grid comprises controlling a plurality of devices arranged in different regions of the power grid to determine the at least one response parameter in a region-specific manner.

[0206] Embodiment 7: The method of any of embodiments 4 to 6, further comprising controlling the at least one device to reduce an impact of the disturbance on a voltage magnitude of the power grid.

[0207] Embodiment 8: The method of embodiment 7, wherein the at least one device comprises a device operating to affect a reactive power in the power grid.

[0208] Embodiment 9: The method of any of the preceding embodiments, wherein processing the received data comprises determining the at least one response parameter based on a total power deviation comprising a power of the introduced disturbance and a change in power in response to the introduced disturbance.

[0209] Embodiment 10: The method of embodiment 9, wherein processing the received data comprises determining a change in load power and / or a change in loss power caused by the disturbance, optionally wherein processing the received data comprises determining the change in load power based on voltage magnitude measurements, and / or optionally wherein processing the received data comprises determining the change in loss power based on a state estimation.

[0210] Embodiment 11: The method of any of the preceding embodiments, further comprising storing the at least one response parameter in a protection and / or monitoring device for evaluating an impact of a further disturbance on a frequency of the power grid, and / or using the at least one response parameter by the processing system to evaluate an impact of a further disturbance on a frequency of the power grid, optionally wherein the evaluated impact comprises a rate of change of frequency, a RoCoF, and / or a frequency nadir.

[0211] Example 12: The method of any of the preceding examples, further comprising implementing, by the processing system or a control, monitoring and / or protection system or device coupled to the processing system, an operation based on the at least one response parameter, optionally wherein the operation comprises a corrective and / or mitigating operation for reducing a change in grid frequency resulting from a further disturbance.

[0212] Example 13: A grid protection and / or monitoring method, comprising: using, by a grid control, monitoring and / or protection system or device, the at least one response parameter to assess an impact of a further disturbance on a grid frequency; and performing an operation based on the at least one response parameter.

[0213] Example 14: A processing system for processing measurements taken in a grid, the processing system comprising: an interface; and at least one processing circuitry operative to: determine a disturbance parameter; generate and output, via the interface, a command to cause a disturbance to be introduced into the grid in accordance with the disturbance parameter; receive data comprising frequency data, wherein the frequency data is representative of a grid frequency measured at one or more locations in the grid; and process the received data to determine, for at least a portion of the grid, at least one response parameter that affects a grid frequency response of the at least a portion of the grid to a further disturbance.

[0214] Example 15: A system comprising: a grid; a processing system as recited in Example 14, the processing system operative to determine at least one response parameter; and a control, monitoring and / or protection device operative to perform a corrective operation or a mitigating operation in response to a further disturbance based on the at least one response parameter.

[0215] Various effects and advantages can be obtained by embodiments of the invention. For example, the systems and methods according to embodiments provide an enhanced technique for determining inertia and / or one or more other response parameters in a highly accurate and efficient manner. The systems and methods are particularly suitable for use with a grid having renewable energy sources access. The systems and methods allow the at least one response parameter that determines a frequency change in response to a further disturbance to be determined even in the absence of PMU instrumentation at all places where needed. The systems and methods disclosed herein are operative to use voltage measurements associated with processing to replace at least partially missing PMU measurements, the processing allowing power changes to be estimated from measured voltage changes.

[0216] The systems and methods can be associated with use with a grid having renewable energy sources and / or battery-based or mechanical energy storage systems, but are not limited thereto. BRIEF DESCRIPTION OF DRAWINGS

[0217] Embodiments of the present application will be described with reference to the accompanying drawings, wherein like or similar elements are designated by like or similar reference numerals throughout the various figures.

[0218] Figure 1 is a block diagram of a processing system.

[0219] Figure 2 is a block diagram of a system including a processing system.

[0220] Figure 3 is a block diagram of a processing system.

[0221] Figure 4 is a schematic diagram of an electrical grid.

[0222] Figure 5 is a schematic diagram of an electrical grid.

[0223] Figure 6 is a schematic diagram of an electrical grid.

[0224] Figure 7 is a graph showing changes in electrical grid frequency.

[0225] Figure 8 is a flowchart of a method.

[0226] Figure 9 is a flowchart of a method.

[0227] Figure 10 is a flowchart of a method.

[0228] Figure 11 is a flowchart of a method.

[0229] Figure 12 is a flowchart of a method.

[0230] Figure 13 is a graph illustrating derivation of a load power change process from a voltage measurement.

[0231] Figure 14 is a block diagram of a system including a processing system.

[0232] Figure 15 is a flowchart of a method.

[0233] Figure 16 is a block diagram of a system including a processing system.

[0234] Figure 17 is a flowchart of a method.

[0235] Figure 18 is a block diagram of a system including a processing system and a control system.

[0236] Figure 19 is a flowchart of a method. DETAILED DESCRIPTION

[0237] Embodiments of the present invention will be described with reference to the accompanying drawings. In the accompanying drawings, similar or identical reference numerals indicate elements having similar or identical configurations and / or functions.

[0238] While embodiments will be described in conjunction with processing systems and methods for determining inertia of a generator equivalent model, the embodiments are not limited thereto. The techniques disclosed herein can be used to determine other response parameters that quantitatively determine the amount of change in grid frequency in response to a disturbance. Examples of such response parameters include any one or any combination of the following: inertia, damping, governor time constant, governor gain, and load sensitivity to voltage amplitude. The determined response parameters can be parameters of a power swing equation that establishes a relationship between grid frequency as a function of time and the change in power as a function of time, but are not limited thereto.

[0239] While the embodiments will be described in conjunction with a power grid having power generation units including renewable energy sources, the embodiments are not limited thereto.

[0240] The processing methods and systems disclosed herein are operable to determine at least one response parameter.As used herein, the terms "determine" and "determine" encompass an approximate determination of a corresponding quantity, such as an estimate of the corresponding quantity.

[0241] As used herein, the term "response parameter" refers to a quantity on which a change in grid frequency in at least a portion of a grid in response to a disturbance depends. The response parameter may depend on the balance and / or imbalance between mechanical and electrical energy in at least a portion of the grid. The response parameter may be associated with a specific region of the grid or with the entire grid. In the former case, separate response parameters may be determined for several or all of the multiple regions of the grid, thereby allowing for spatially resolved determination of inertia or other response parameters.

[0242] As used herein, the term "response parameter" encompasses inertia, but is not limited thereto. Response parameters may include parameters of the power swing equation and / or the governor equation, or parameters derived therefrom (such as any parameter affecting the rate of change of frequency (RoCoF) or the frequency minimum).

[0243] As used herein, the term "at least one response parameter" can include or can be a parameter or a combination of parameters that allows for a predictive determination of the impact of a further disturbance on the grid frequency based on at least one response parameter and the magnitude (e.g., power), frequency information (e.g., frequency or spectrum of further disturbance), and location of a disturbance in the grid. The "at least one response parameter" can include or can be a parameter or a combination of parameters that allows for a predictive determination of the impact of a further disturbance on the grid frequency using linear response theory using at least one response parameter and the magnitude (e.g., power), frequency information (e.g., frequency or spectrum of further disturbance), and location of a disturbance in the grid as input.

[0244] As used herein, the term "grid" encompasses a transmission grid and / or a distribution grid. The grid can include both a transmission grid and a distribution grid.

[0245] As used herein, the term "inertia" refers to the inertia of a generator equivalent model (such as a single generator equivalent model) in at least a portion of the grid.

[0246] As used herein, the term "power" refers to electrical power, unless explicitly stated otherwise.

[0247] According to the present invention, processing methods and systems are provided for determining (in particular, estimating) one or more response parameters of the grid response to a disturbance in the entire grid or a certain area of the grid using measurement data. Inertia is one example of such a response parameter. The systems and methods are operable to determine one or other parameters of a single generator equivalent of a power system. The systems and methods can perform reliable inertia estimation in a grid without full phasor measurement unit (PMU) observability.

[0248] Embodiments of the present invention thus address the need for a more general and more accurate technique for determining one or more response parameters that determine how much a grid frequency is shifted by a further future disturbance. The one or more response parameters allow for a prediction of the change in the grid frequency in response to the further disturbance and / or allow for an appropriate countermeasure (such as changing a generator unit commitment) to be taken automatically or semi-automatically to keep the change in the grid frequency at a low level (e.g., below a threshold and / or to a minimum change achievable).

[0249] The systems and methods are operable to perform an inertia estimation of the grid or a certain area of the grid based on frequency measurements and knowledge about a disturbance that is occurring in the grid.

[0250] The systems and methods are operable such that the impact of the original disturbance is taken into account in the estimation model. The impact of the original disturbance can include a change in load power and / or a change in loss power.

[0251] The system and method can operate such that the knowledge about the disturbance comes entirely from the measured data.

[0252] The system and method can operate such that at least some of the knowledge about the disturbance comes from measured values that are used as a basis for approximating quantities not measured in the grid (e.g. voltage magnitude as a measured data used to derive changes in load power).

[0253] The system and method can operate such that the disturbance is exogenous to the grid.

[0254] The system and method can operate such that the disturbance can be controlled by controlling one or more devices capable of changing their power output.

[0255] The system and method can operate such that one or more filtering techniques are applied to the received measured data to improve the quality of the estimation. For example, narrow band pass filtering can be used to extract the grid response to a periodic controllable disturbance.

[0256] The system and method can operate to set the frequency of the disturbance in a controlled manner so as to not excite any undesirable dynamics in the grid.

[0257] The system and method can operate such that at least one response parameter is determined by system identification techniques, while taking into account equations of a simplified equivalent of the managed grid / region.

[0258] The system and method can operate such that at least one response parameter (such as the estimated inertia of an equivalent generator) is dependent on the amount of kinetic energy in the grid. Alternatively or additionally, the system and method can also operate such that one or more response parameters (such as inertia) are used in conjunction with other parameters of the equivalent generator model to model the frequency response of the grid to further disturbances.

[0259] The system and method can operate to allow the disturbance to be generated in a controlled manner such that the impact of the disturbance on the voltage magnitude of the grid is kept below the voltage change magnitude and / or minimized.

[0260] The systems and methods can operate to determine at least one response parameter using an estimate based on a swing equation and / or an estimate based on an actively introduced probing signal. In the swing equation-based technique, an evaluation can be made for one disturbance or multiple disturbances that are not specifically created for the purpose of inertia estimation, but are introduced by, for example, a load connected to the power grid that intermittently (e.g., periodically or aperiodically, but on an intermittent basis) draws increased power, causing a shift in the grid frequency. In the probing signal-based technique, the systems and methods cause a probing signal to be actively injected into the power grid to cause an excitation of the grid frequency. The probing signal can be a one-time injection or a repeatedly injected (e.g., in an intermittent manner) sinusoidal signal or other periodically varying signal.

[0261] The systems and methods can operate to process power measurements and, where applicable, approximate power values from, for example, voltage measurements. In the latter case, the time-dependent changes in the grid frequency in response to the disturbance can be combined with the measured time-dependent voltage to approximately determine the time-dependent changes in power.

[0262] The systems and methods can operate to determine at least one grid response parameter (such as inertia) based on a total power deviation that is composed of the power of the disturbance itself and the changes in load power and / or loss power as a result of the disturbance. Thereby, at least one grid response parameter is determined with enhanced accuracy as compared to techniques that rely only on the power of the disturbance and ignore the effects of the disturbance on the load and / or losses, and in particular, the effects of the disturbance on the load power and / or loss power.

[0263] The systems and methods can estimate at least one response parameter even in the absence of full PMU observability (that is, not all of the changes in load power and / or loss power and / or disturbance power are available via respective PMUs). The systems and methods can operate to replace missing power data in determining at least one response parameter. The systems and methods can operate to approximately infer changes in load power and / or loss power from available measurements, such as changes in voltage magnitude and / or individual changes in losses caused by the disturbance.

[0264] The systems and methods can operate to use an identification technique that can ensure convexity of an optimization problem solved to obtain at least one response parameter. Thereby, a globally optimal solution to identify at least one response parameter is guaranteed even in the absence of PMU measurements that constitute one or more summation terms of a total power deviation in response to the disturbance.

[0265] The system and method can be operable to actively introduce a disturbance, for example by one or more controllable devices. The disturbance can be generated to have a periodicity, in particular a sinusoidal signal shape. The disturbance can be generated such that the power amplitude of the disturbance is smaller than a threshold value. All measurement time series used as input for the estimation of at least one response parameter can be filtered using a bandpass filter. The passband of the bandpass filter can be at (e.g. centered around) the frequency of the actively introduced disturbance (e.g. the frequency of the sinusoidal power variation introduced by the disturbance). The bandpass filter can have a bandwidth that is smaller (e.g. at most 0.1 times or at most 0.01 times) the frequency of the sinusoidal disturbance. The bandpass filter allows to distinguish the reaction of the power grid to the disturbance from e.g. background load variations.

[0266] The system and method can be operable such that when the disturbance is generated under control of a processing system using controllable devices, the disturbance and / or the processing of the measurements is generated in a way that the influence of the disturbance on the voltage amplitude of the power grid is kept below a voltage threshold value, which helps to reduce load variations due to the disturbance and thereby improves the estimation accuracy.

[0267] The system and method can be operable such that parameters of individual generators or the equivalent of individual generators in a certain region of the power grid can be determined. This can be done in a way that at least one inertia is determined, which depends on (e.g. reflects) the total kinetic energy stored in the power grid (such as in the rotating part of the power grid). Alternatively or additionally, at least one response parameter can also be determined, which is not directly equivalent to the kinetic energy, but can be used to predict the frequency evolution of the power grid after a further disturbance of known size and location.

[0268] The system and method can be operable such that when the disturbance is generated under control of a processing system using controllable devices, the frequency of the disturbance (e.g. the frequency of a disturbance introducing a sinusoidal variation of power at a certain location in the power grid) is chosen such that it does not excite any resonance modes in the power grid, for example. The disturbance is preferentially generated to have a very narrow frequency band in order to facilitate the distinction of the response to the disturbance from noise and background load variations. To this end, the time period during which the disturbance is applied can be much longer (e.g. at least 20 times, at least 30 times, at least 40 times or at least 50 times) than the duration of a period of the sinusoidal wave. The disturbance can be generated to have a smooth envelope, for example a Gaussian or a Lorentzian curve envelope, to reduce the introduction of unwanted frequencies into the power grid.

[0269] Various effects can be obtained by the system and method according to embodiments. The system and method according to embodiments allow determining at least one response parameter indicative of the amount of kinetic energy in the rotating masses of the electrical machines stored in the power grid. The inertia of the equivalent model of the generator is an example of such a parameter. These rotating masses can include rotating masses of synchronous generators and / or flywheel energy storage systems. The kinetic energy as quantified by the inertia can represent the energy buffer for changes in frequency after a generation / demand change.

[0270] The system and method according to the present embodiments provide enhanced techniques for determining a response parameter indicative of the kinetic energy of the rotating components of the power grid. This helps in planning, such as generator unit commitment planning, and prevents low inertia situations. Low inertia can cause several problems. For example, the rate of change of frequency (RoCoF) and the frequency nadir depend on the inertia. By implementing power grid operation based on the determined response parameter, such as the inertia, the risk of a frequency limit (e.g., ±0.2 Hz) being violated can be reduced or even eliminated. Violating the frequency limit is undesirable as it can lead to load shedding, equipment disconnection, and the activation of potential cascading outages. Higher inertia levels avoid such problems and provide a higher safety margin for the transmission system operator (TSO) when operating the system. The present techniques allow efficient and accurate determination of the inertia, thus helping to enable enhanced control and operation of the transmission system.

[0271] The system and method disclosed herein provide techniques to estimate at least one response parameter affecting the change in the grid frequency under further disturbances based on a single disturbance or a series of disturbances. The disturbances can include disturbances of which the size and start time are known, as they are determined by the processing system. The disturbance parameters, such as the disturbance frequency and / or amplitude, can be determined based on known mechanical and / or electrical resonances to avoid exciting any of these resonances. For example, the knowledge about the resonances can be obtained from historical data.

[0272] The system and method disclosed herein can be applied to the entire power grid or any area in the power grid, provided that the current through the tie lines connecting the area to the rest of the power grid is monitored.

[0273] Figure 1A processing system 20 is shown operating on data including measurements taken or generated by one or more measurement instruments of a data acquisition system. The data includes frequency measurements indicative of changes in grid frequency as a function of time (e.g., in the form of time series data) caused by a disturbance. The data includes PMU measurements and / or other measurements that allow determination of changes in load power and / or loss power that occur in response to the disturbance. The processing system 20 can receive data indicative of the disturbance, such as one or more of: a disturbance power as a function of time (e.g., in the form of time series data), a magnitude of the disturbance, frequency resolution information for the disturbance (such as a center frequency of the disturbance power in Fourier space), a duration of the disturbance. The data indicative of the disturbance can be from a measurement instrument (such as at least one PMU) or from a SCADA system or other control and / or monitoring system. The processing system 20 need not receive data indicative of the disturbance when the processing system 20 determines parameters of the disturbance and controls at least one device to introduce the disturbance into the grid specifically for the purpose of determining at least one response parameter. In this case, the processing system 20 can store and use the determined parameters of the disturbance (such as a disturbance power as a function of time (e.g., in the form of time series data), a magnitude of the disturbance, frequency resolution information for the disturbance (such as a center frequency of the disturbance power in Fourier space), a duration of the disturbance) when determining at least one response parameter.

[0274] The processing system 20 can be implemented as a device including a device housing in which the components shown in FIG. 1 are housed. Figure 1 The processing system 20 can be implemented as a combination of multiple devices communicatively coupled to one another. The processing system 20 can operate in conjunction with and in cooperation with a control system for at least a portion of a power grid, such as a transmission control system, a regional control center, or a national control center.

[0275] As further explained herein, the processing system 20 operates on data including frequency measurements in response to a disturbance to determine at least one grid response parameter that affects a magnitude of changes in grid frequency in at least a portion of the grid in response to a further disturbance. The at least one response parameter determined by the processing system 20 can be one or more response parameters that: can be combined with other parameters; allow prediction of future changes in grid frequency given a magnitude (such as power), frequency, and / or time information (such as a frequency spectrum), and location of the further disturbance. Inertia of a single generator equivalent is one example of such a response parameter.

[0276] The processing system 20 comprises at least one processing circuitry 30. The at least one processing circuitry 30 is operative to be coupled to the at least one interface 21. The at least one processing circuitry 30 is configured to receive data acquired or generated by the measuring instrument, such as frequency measurements and / or time-dependent data indicative of an electrical property, such as PMU measurements or voltage measurements. The at least one processing circuitry 30 is operative to perform a processing 31 of the frequency measurements and the data indicative of a disturbance. The at least one processing circuitry 30 is operative to perform a determination 32 of at least one response parameter based on the processing 31 of the frequency measurements and the data indicative of a disturbance. The at least one processing circuitry 30 is operative to perform a determination 33 of at least one operation based on the determined at least one response parameter.

[0277] The at least one processing circuitry 30 is operative to perform a disturbance source control 39. The disturbance source control 39 is operable to determine a disturbance parameter of a disturbance to be introduced into the power grid, and can introduce the disturbance in accordance with the determined disturbance parameter. The disturbance source control 39 is operable to determine the disturbance parameter, such as a disturbance frequency and / or amplitude, based on known mechanical and / or electrical resonances to avoid exciting any of these resonances. For example, the knowledge about the resonances can be obtained from historical data, or from system modeling.

[0278] The at least one processing circuitry 30 is operable to store the determined disturbance parameter in a storage system accessible to the at least one processing circuitry 30, and to retrieve the disturbance parameter for determining the at least one response parameter.

[0279] The at least one processing circuitry 30 can comprise any one or any combination of an integrated circuit, an integrated semiconductor circuit, a processor, a controller, a special-purpose integrated circuit (ASIC), a field-programmable gate array (FPGA), a circuit comprising qubits and / or quantum gates, but is not limited thereto.

[0280] The processing system 20 can comprise an output interface 22. The processing system 20 is operable to output a command causing performance of the operation determined by the operation determination 33.

[0281] The processing system 20 is operable to output a command causing one or more of the following: providing the at least one response parameter to a control and / or monitoring system, such as a SCADA system, a national control center, a regional control center, to ensure power grid stability; controlling a human-machine interface (HMI) to output information about the determined at least one response parameter; controlling one or more controllable devices of the power grid to mitigate future changes in the power grid frequency and / or to ensure that the power grid frequency remains within a predetermined frequency range (which can have a width of e.g. ±0.2 Hz or at most 0.2 Hz or below).

[0282] Figure 2is a schematic diagram of system 10.

[0283] System 10 includes an electrical grid 11. System 20 includes a processing system 20 that operates to determine at least one response parameter (such as an inertia) that influences or determines a change in the grid frequency in response to a further disturbance.

[0284] A disturbance source 13 introduces a disturbance into the grid. Processing system 20 can receive information about the disturbance from disturbance source 13 or from a system that controls and / or monitors the disturbance source (e.g., a SCADA system), such as one or more of the following: a disturbance power as a function of time (e.g., in the form of time series data), an amplitude of the disturbance, frequency resolution information for the disturbance (such as a center frequency of the disturbance power in Fourier space), a duration of the disturbance. Disturbance source 13 can include a dedicated source provided for the specific purpose of introducing a disturbance for the purpose of determining at least one response parameter. Disturbance source 13 can include a load connected to the grid that intermittently draws increased amounts of power, such as an electric arc furnace.

[0285] Disturbance source 13 can include a controllable disturbance source. The frequency and / or amplitude of the disturbance introduced by disturbance source 13 can be determined by processing system 20, as described in greater detail elsewhere herein.

[0286] System 10 includes a measurement instrument 12. Measurement instrument 12 can include a PMU and / or a voltage measurement device, but is not limited thereto. Measurement instrument 12 can operate to obtain grid frequency measurements and provide the grid frequency measurements, as well as power and / or voltage measurements, to processing system 20.

[0287] Processing system 20 can output at least one grid response parameter or instructions dependent thereon. Processing system 20 can output an inertia 15 (e.g., an inertia of a single generator equivalent or an inertia of a power swing equation) for use by control system 14.

[0288] Control system 14 can operate to use at least one response parameter (e.g., inertia 15) to take a mitigation action that reduces a change in the grid frequency caused by a further disturbance before the further disturbance occurs. Control system 14 can operate such that the mitigation action can include adjusting a generator unit commitment and / or increasing a kinetic energy of a rotating component (such as a rotating component of a synchronous generator) in order to maintain the grid frequency within a target frequency range even when a further disturbance is introduced.

[0289] The processing system 20 operates to take into account not only the disturbance power, but also the effect of the disturbance on the load power and / or the loss power. The processing system 20 can operate to determine a total power deviation as the sum of the disturbance power, the change in loss power, and the change in load power, each as a function of time. The total power deviation as a function of time can be processed in combination with a measured change in grid frequency as a function of time, measured over a time interval following the introduction of the disturbance. Thus, a power swing equation or other processing can be used to determine at least one response parameter, as described in more detail herein.

[0290] Figure 3 is a schematic block diagram of at least one processing circuit 30 of the processing system 20. The at least one processing circuit 30 operates to process measurements to perform a determination 35 of a change in loss power caused by the disturbance and a determination 36 of a change in load power caused by the disturbance. Filtering or other pre-processing techniques can be applied to the power and / or voltage measurements to distinguish the change in load power and / or loss power caused by the disturbance from other root causes of changes in load power and / or loss power. The determinations of changes in load power and loss power can determine the changes in load power and loss power as a function of time, e.g., in the form of time series data.

[0291] The at least one processing circuit 30 operates to perform a determination 34 of the disturbance power to determine the disturbance power as a function of time. The determination 34 can but need not be based on measurements. For example, the determination 34 of the disturbance power can be based on data from a control system (e.g., a SCADA system) or from the disturbance source 13 or from disturbance parameters previously determined by the processing system 20 and stored in a storage system 23 of the processing system 20.

[0292] The at least one processing circuit 30 operates to determine a total power deviation by summing the disturbance power and the changes in load power and loss power. The total power deviation can be determined as a function of time, e.g., in the form of time series data.

[0293] The processing system 20 can operate to process the time-dependent total power deviation with time-dependent changes in grid frequency to determine a response parameter, such as inertia.

[0294] Figure 4 is a schematic diagram of a system 40, in which power lines are shown as single lines. The system 40 includes a grid, a transmission grid 50, and one or more distribution grids 60. One or more power generation units 51, 52, 53 can be configured to be coupled to a conductor 54 (e.g., a bus or power line) in a high- or ultra-high-voltage domain of the transmission grid 50. Rotating components of the power generation units 51, 52, 53 can store kinetic energy, contributing to frequency stability.

[0295] The distribution grid 60 is connected to the conductor 54 of the transmission grid via power stations 61, 63 operating to step down the voltage to a low voltage, such as for example 50 kilovolts. Loads 68 are supplied via distribution buses 62, 64, for example. The loads 68 can include industrial loads, such as arc furnaces 69 that intermittently draw increased levels of power or other power consumers that intermittently draw increased levels of power.

[0296] The grid can also include renewable energy sources 65, 66. The techniques disclosed herein can provide robust and reliable determination of at least one grid response parameter, such as inertia, even in the case of renewable energy sources interfacing the grid.

[0297] The system 40 includes a measurement instrument that includes at least one PMU 71, 72, 73. The measurement instrument is operable to measure power flow and / or grid frequency at the voltage bus 54 and / or the distribution buses 62, 64.

[0298] The system 40 includes a processing system 20. The processing system 20 is communicatively coupled to the measurement instrument. The processing system 20 and the measurement instrument are operable such that the processing system 20 receives measurements in messages that are compliant with IEC 61850 (e.g., compliant with IEC 61850-90, such as compliant with IEC 61850-90-2:2016 messages).

[0299] The processing system 20 is operable to determine at least one response parameter based on the power measurements and taking into account changes in load power and / or loss power caused by the disturbance. The disturbance can be introduced by a dedicated device, or by a consumer 69 that intermittently draws increased levels of power, for example.

[0300] The processing results can then be provided to and used by a control system, for example, to change generator unit commitments and / or to otherwise reduce future grid frequency variations (e.g., by ensuring that the grid frequency remains within a target frequency range).

[0301] Figure 5 is a schematic diagram of the system 40. The plurality of power generation units 51, 52, 53 are replaced by a single generator equivalent 59. The processing system 20 is operable to determine one or more parameters of the single generator equivalent model, such as inertia, based on measurements of changes in grid frequency as a function of time, and total power deviation including both disturbance power and changes in load power and loss power.

[0302] Figure 6 is a schematic diagram of another embodiment of the system 40. Figure 6The system 40 in the figure does not have full PMU observability. Even in the absence of full PMU observability, the processing system 20 is operable to determine at least one response parameter based on changes in grid frequency, disturbance power, changes in loss power and changes in load power (as a function of time and in response to the disturbance, respectively).

[0303] In this case, the processing system 20 is operable to approximate the change in load power caused by the disturbance as a function of time based on voltage measurements obtained using the voltage measurement instruments 74, 75. This can be done by using a processing technique that ensures convexity of the optimization problem solved for determining the at least one response parameter.

[0304] Alternatively or additionally, the processing system 20 is also operable to approximate the change in loss power caused by the disturbance by performing a state estimation. The required state information can be retrieved by the processing system 20 from a control and / or monitoring system, such as a SCADA system.

[0305] Various processing techniques that allow the processing system 20 to determine the at least one response parameter will be described in detail below. For example, the processing system 20 can apply any one or any combination of processing techniques that solve an optimization problem with the objective of minimizing a metric that quantifies the extent to which the measured grid frequency change observed in response to the disturbance deviates from the theoretically expected grid frequency change to the disturbance when applying a candidate parameter in the at least one response parameter. The candidate response parameter that results in the smallest value of the metric computed over a time interval following the disturbance is then the determined at least one response parameter. The determined at least one response parameter can subsequently be used to automatically predict further deviations of the grid frequency in response to further disturbances and / or to take mitigation actions (such as adjusting generator unit commitments) that ensure that the grid frequency remains within a target frequency range.

[0306] Other metrics or other processing techniques can also be used, which can depend on which response parameter needs to be determined.

[0307] Figure 7 is a plot representing measurements of the grid frequency 81 in response to the disturbance. Characteristics of the plot 81 that are widely used in the art are the rate of change of frequency (RoCoF) and / or the frequency nadir represented by the nadir of the plot 81.

[0308] Frequency curves 82, 83 are computed using the information about the disturbance and one or more response parameters, such as inertia. The processing system 20 operates to identify at least one response parameter for which the frequency change computed for the at least one response parameter is most similar to the measurements 81 when evaluated according to a certain metric and over a time interval after the disturbance. The metric can be an L2 metric or another metric, depending on the type of response parameter.

[0309] The processing system 20 can operate to determine inertia as a response parameter based on the RoCoF after the disturbance. When the power system is subjected to a large disturbance, the power imbalance in the inertia response phase is mainly caused by the electrical power deviation:

[0310] wherein, is the equivalent system inertia, is the change in the grid frequency and its derivative is defined as the RoCoF,

[0311] represents the power change of the frequency dependent load and the voltage dependent load, is the size of the disturbance.

[0312] The processing system 20 can operate to determine the RoCoF based on frequency measurements taken at different times after the initial time of the disturbance. To reduce noise and sharp transient type changes in the frequency measurements, the processing system 20 can apply a processing that can include techniques such as filters and / or pre-processing of the frequency measurements, thereby generating a more accurate estimate of the RoCoF. The accuracy of the power imbalance estimate also has a significant impact on the inertia estimate. Thus, by taking into account the effect of the disturbance on the load and / or losses, a more accurate determination of the inertia can be obtained. By taking into account the effect of the disturbance on the load and losses of the grid, the drawbacks of conventional methods, such as those that only consider the effect of the power of the disturbance, can be mitigated, which can lead to inaccurate results.

[0313] Even in the case where the disturbance is a single event with a certain magnitude, such as a disconnection or a circuit breaker (CB) trip, it is possible to determine the inertia based on equations (1) and (2), which is an example of a response parameter for determining the change in the grid frequency in response to further disturbances.

[0314] When the disturbance is introduced in a controlled manner (by using a disturbance that is purposefully introduced to change the balance between electrical and mechanical power, causing a change in grid frequency), the processing system 20 can operate to additionally or alternatively process a total power deviation that is composed of a disturbance power, a change in loss power, and a change in load power. The processing system 20 can operate to cause a sustained change in disturbance power, such as by controlling a controllable device to produce a continuous, modulated disturbance of a given frequency or frequency band. The measurements can be filtered to distinguish the grid response to the disturbance from noise. This technique allows the use of disturbances with smaller magnitude to determine the at least one response parameter, as compared to the above-described method of deriving the at least one response parameter based on a response to a disconnection or CB trip. The at least one response parameter (such as inertia) can be determined with enhanced accuracy even if the disturbance (which is purposefully introduced to estimate the grid response parameter) has a smaller magnitude.

[0315] The processing system 20 can operate to identify inertia or another grid response parameter using a sinusoidal or multi-sinusoidal disturbance of the system state based on the total power deviation. The processing system 20 can perform a selection of a frequency or frequencies taking into account the potential influence of the frequency or frequencies on a resonance mode of the grid. The processing system 20 can perform bandpass filtering to distinguish the grid response to the disturbance from noise.

[0316] The processing system 20 can operate to determine the at least one response parameter for both a sudden disturbance (which can be represented as a step change) and an intermittent (e.g., periodic) disturbance signal. Thus, the processing system 20 is adapted to make a determination of the at least one response parameter from the total power deviation for any of a variety of types of electrical disturbances. The disturbances can include passive (uncontrollable) electrical disturbances such as a generator trip or a disturbance caused by a load step change. The processing system and processing method can operate to determine the at least one response parameter based on measurements of at least an active (controllable, injected) disturbance that is created by the processing system controlling a controllable device such as an HVDC device or a battery.

[0317] The processing system and processing method can additionally also operate to process measurements that reflect the grid frequency dynamics in response to passive electrical disturbances. In the case of passive disturbances such as a generator trip, the processing system 20 can operate to perform a single estimation per disturbance, such that the inertia estimation is sporadic and infrequent. For disturbances that recur intermittently (e.g., periodically), a repetitive (e.g., periodic) estimation can be performed by the processing system. The determination of the at least one response parameter can be made on a repetitive basis, e.g., with a repetition time of hours or minutes. An example of a periodic, uncontrolled disturbance is the operation of an electric arc furnace. The disturbance size can be determined based on an output of a measurement device at the source of the disturbance or by an estimation technique.

[0318] As the disturbance is actively introduced under the control of the processing system 20 for the purpose of determining at least one response parameter, the rate at which the at least one response parameter can be set or updated in response to user input received by the processing system 20. The active introduction of the disturbance will result in a frequency deviation, changes in the load power and / or loss power and changes in the frequency which can be used in conjunction with known parameters of the disturbance to determine the at least one response parameter. The processing system 20 can operate to select the amplitude and frequency of the disturbance (e.g. the sine wave frequency of a sinusoidal disturbance power) based on characteristics of the power grid, in particular based on known electrical and / or mechanical resonances of the power grid.

[0319] Example techniques of the processing which can be performed by the processing system 20 will be described in more detail next. Although some techniques are described in relation to a response parameter comprising inertia, embodiments are not limited to this. Other response parameters can be determined instead of or in addition to inertia.

[0320] Model The processing system 20 can use a model in which the power grid or a region of the power grid is modelled using a single equivalent generator (as schematically illustrated in Figure 5 Fig. 1) having an inertia equal to the inertia of the power grid / region. This approach simplifies the problem to be solved, making it easier to handle measurements during live operation. It is not required to determine the at least one response parameter (such as inertia) at every bus in the power grid / region. An aggregate value such as the inertia of the single generator equivalent provides the most relevant information for predicting changes in the power grid frequency and represents the quantity of interest for the TSO. Given the typical time scale of frequency events, it is not necessary to model fast acting controls (such as automatic voltage regulation) in the equivalent generator.

[0321] The main equation for the estimation is the so-called swing equation which relates the generator inertia and damping to the imbalance between the electrical energy demand in the power grid and the mechanical energy from the prime movers:

[0322] where is the inertia, is the damping, denotes a deviation from a pre-disturbance value of a quantity, is the mechanical power applied to the rotating parts (e.g. the turbine), is the total electrical demand, is a disturbance occurring in the power grid (such as a load step change, a generator or interconnection line trip, and are the frequency and its derivative, respectively.

[0323] For periodic disturbances, assume that the mechanical power deviation is zero because the turbine governor will not react to it due to high frequency disturbance, short measurement window, or governor dead band. For large disturbances, the :

[0324] where and are the time constant and gain of the governor, respectively.

[0325] The electrical power deviation is composed of several addends

[0326] where: • is the total change in active power consumed by the load as a result of the disturbance. The active power can depend on both the frequency and the amplitude of the voltage at its bus, both of which can change as a result of the disturbance.

[0327] • is the total change in active power loss (e.g., in transmission lines and transformers) as a result of the disturbance. Large disturbances can cause a redistribution of power flow, changing the losses, due to the primary frequency control of the generators.

[0328] • is the total change in active power flow in the tie lines as a result of the disturbance. This term can be omitted. In particular, when determining response parameters (such as inertia) for a certain region of the power grid, rather than the entire power grid, this term can be optionally considered. When determining at least one response parameter (such as inertia) for a certain region, including this term helps to isolate the region from other regions of the power grid and provides a more accurate estimate of the at least one response parameter for the respective region.

[0329] In the above equations (3), (4), and (5), the quantities , , and are scalars representing parameters of the equivalent generator, while all other quantities are time series. The processing system 20 can be operable to determine one, some, or all of the parameters , , and using measurements included in the data. The processing system 20 can be operable to determine at least the parameter based on measurements included in the data. The measurements included in the data can include measurements that define or allow the processing system 20 to at least approximately determine and time-dependent changes. The measurement data used by the processing system can include at least time-dependent grid frequency in response to the disturbance (e.g. in the form of time series data).

[0330] Equations (3)-(5) above are used as examples, and the present application is not limited to using these exact equations. As an alternative or in addition to the parameters of the above-described models, parameters of other parametric equations modeling the behavior of the grid frequency in response to the disturbance can be determined by the processing system.

[0331] For illustration, the processing system 20 can operate to determine one or more parameters of a high-order governor equation. These parameters can be used to predict the frequency changes in response to a further disturbance, the parameters of which (such as amplitude, frequency, location) are known in advance.

[0332] For further illustration, the processing system 20 can operate to determine one or more parameters of a high-order swing equation. These parameters can be used to predict the frequency changes in response to a further disturbance, the parameters of which (such as amplitude, frequency, location) are known in advance.

[0333] For further illustration, the processing system 20 can operate to determine one or more parameters of a high-order swing equation or governor equation with additional terms (which can be non-linear terms).

[0334] A basic assumption in simplifying the grid / region to a single equivalent generator is that the parameters of this generator are constant. The processing system 20 can operate to take into account that at the beginning of the disturbance, only the inertia in its immediate vicinity is available through the processing disclosed herein. In other words, in processing the measurements, the processing system 20 can operate to take into account that the changes in grid frequency as well as changes in loss and / or load power are initially more pronounced in the vicinity of the disturbance, and only at a certain delay further away from the disturbance. The processing system 20 can use spatially dependent processing of the measurements to accommodate these effects. For example, the measurements can be weighted in a spatially dependent manner depending on whether they were captured closer or further away from the disturbance.

[0335] Data used by the processing system and optional pre-processing The processing system 20 operates to determine at least one response parameter using the following data: • data related to the electrical disturbance .

[0336] This data can but need not come from a measurement. For example, when the processing system 20 determines the at least one response parameter using an active disturbance of the power grid introduced for determining the at least one response parameter, the processing system 20 can use a priori knowledge of this disturbance, such as location, amplitude, frequency, duration. The processing system 20 can be operable to optionally use knowledge of the electrical and / or mechanical resonance of the power grid to determine these characteristics of the disturbance, and can cause the disturbance to be introduced. The determined characteristics can be stored by the processing system 20 in an internal memory 23 or an external memory for later use in determining the at least one response parameter.

[0337] The specific implementation of the disturbance-related data can be different, depending on the type of disturbance and its amplitude. For disturbances such as load or generation step changes, the amplitude of the change and its initial time or an approximation of the amplitude and initial time can be sufficient.

[0338] For small, recurring disturbances, the processing system 20 uses time series measurement data of the disturbance. The frequency of the controlled disturbance is preferentially chosen by the processing system 20 or a control system communicatively coupled to the processing system 20 to avoid exciting any known power grid resonances. The frequency range to be avoided can be known a priori from system studies or obtained from a real-time oscillation monitoring system based on a wide-area monitoring system (WAMS). To reduce the impact of the controlled disturbance on the voltage amplitude of the power grid, reactive power control can be used to mitigate the impact of the disturbance on the load and thereby improve the estimation accuracy.

[0339] • Frequency measurement data The frequency measurement can be a measurement obtained at a single measurement location or at multiple measurement locations. In the latter case, the processing system 20 can be operable to average the measured frequencies to obtain an average power grid frequency of the power grid or a certain region thereof. In determining the power grid frequency to be processed later, the processing system can take into account the distance of the measurement location from the disturbance. Spatially dependent weighting or averaging techniques can be used.

[0340] The processing system 20 can be selectively operable to determine the at least one response parameter using one or more of the following data: • Line flow measurements / values and active power demand. These quantities can be measured directly, or the processing system 20 can be operable to estimate these quantities using state estimation techniques.

[0341] • Voltage amplitude measurements at different buses in the power grid. As described in more detail elsewhere herein, the processing system 20 can be operable to estimate load changes as a result of the disturbance based on the voltage amplitude measurements.

[0342] • snapshots of total losses. These data can be measured or estimated directly by the processing system using state estimation techniques. The processing system 20 is operable to use the snapshots (i.e. time-discrete loss values) to approximately determine the time series of load changes as a result of the disturbance.

[0343] • measurements of tie-line power flows P tie lines These measurements are optional and useful in case at least one response parameter is determined for a certain area of the power grid. The processing system 20 is operable to optionally take into account tie-line power flows to estimate the inertia or other response parameters in a given power grid area. Thereby accuracy can be further improved.

[0344] Preferably, the time series measurement data used by the processing system are time-synchronized to facilitate processing by the processing system. Time-synchronization is achieved by using corresponding measurement devices (such as PMUs) or applying synchronization methods after the data is collected. If the initial time of the disturbance is unknown, the processing system 20 is operable to estimate the initial time using signal processing or similar techniques.

[0345] Once the data for estimation is collected and synchronized, the processing system 20 can optionally perform pre-processing. Pre-processing is useful for both active estimation and passive estimation (i.e. using or not using a dedicated disturbance source controlled for the specific purpose of determining at least one response parameter) to reduce estimation errors. The processing system 20 is operable to apply filtering techniques to distinguish power flows caused by at least one disturbance from other background signals. However, the principles disclosed herein can in principle also work without such pre-processing. If the measurement noise is low, the power grid is small or larger estimation errors can be accepted, filtering can be omitted.

[0346] The processing system 20 is operable such that the type of pre-processing (such as normalizing data, filtering and other conditioning or pre-processing steps) can depend on the specific type of power grid and measurement devices.

[0347] The processing system 20 is operable to narrowband filter active disturbances (i.e. when the disturbance is introduced by a controllable device controlled in a dedicated way for the specific purpose of determining at least one response parameter). This enables the processing system 20 to distinguish information (such as power flows and / or power grid frequency changes) at the frequency at which the power grid is excited by the disturbance from noise and other phenomena in the power grid that are not related to the disturbance.

[0348] The processing system 20 can also be operable to perform low-pass filtering, for example, for passive estimation (i.e., when the disturbance is caused by equipment that is not controlled in a dedicated manner for the specific purpose of introducing the disturbance for the purpose of determining at least one response parameter). This enables the system identification technique to preserve the slow time constant shape of the frequency response.

[0349] The processing system 20 can be operable to perform grid-specific pre-processing techniques. Such grid-specific pre-processing techniques can include using weighted averages of frequency and voltage measurements based on knowledge of the grid topology and operating conditions, delaying the active estimation procedure for a certain time until the transient response of the grid to the disturbance has attenuated, but are not limited thereto.

[0350] Metrics used by the processing system The processing system 20 can be operable to use a metric that is maximized or minimized to obtain an optimal value for the at least one response parameter. Inertia or other parameters of the equivalent generator model are examples of such response parameters.

[0351] For inertia estimation, which is particularly relevant to predicting how the grid will react to a given further disturbance. For this, any one or any combination of two types of inertia estimates can be determined: • a value that reflects the total kinetic energy of the rotating components in the grid or a certain region of the grid (e.g., of synchronous generators and optionally rotating energy storage systems, if present).

[0352] • a value that does not have a physical meaning but, in combination with other parameters of the equivalent generator, enables a prediction of the expected frequency evolution in response to a further disturbance, given the magnitude and location of the further disturbance.

[0353] To assess the quality of these two types of inertia estimates, the processing system can use the mean absolute error (MAE) between the true value and the estimated inertia. For example, the true value of the inertia can be approximated by the processing system 20, for example, by summing the generator inertias and assuming a percentage of the load inertias.

[0354] Even in the absence of information about the true value of the inertia, the processing system 20 can be operable to determine the inertia as a response parameter. For this, various types of metrics can be used.

[0355] The processing system 20 can be operable to optimise a measure based on the previous step prediction error (e.g. the response parameters are determined to have those values which minimise the prediction error). Thus, given the frequency at the previous time step, for the next time step the processing system 20 can be operable to use the difference between the actual frequency measurement and the predicted frequency of the equivalent generator. Implementation of this type of measure and determination of the values which optimise (i.e. minimise) the measure are available to the skilled person.

[0356] The processing system 20 can be operable to use a measure related to the overall frequency response prediction. The processing system 20 can determine values for the at least one response parameter such that a window of frequency measurements is used to determine the at least one response parameter, rather than just focussing on the previous step prediction.

[0357] An example of such a measure for a large disturbance can be the RoCoF error, the error in the time of the nadir, the error in the nadir value and a weighted sum of the steady state value of the frequency:

[0358] where the variables the tilde indicates the estimated value and the value without the tilde is the true value. The weights are adjustable parameters which can be tuned. At least some of these weights can be zero.

[0359] Thus, the processing system 20 can be operable to identify values for one or more response parameters (such as the inertia and / or other parameters of the generator equivalent) such that a weighted average of at least two of the RoCoF error, the error in the time of the nadir, the error in the nadir value and the steady state value of the frequency is optimised (typically minimised).

[0360] By using the measure in equation (6), the processing system 20 can typically determine values for the at least one response parameter which provide a minimum value for equation (6). For a multi-dimensional parameter space, optimisation techniques such as gradient descent can be used. The processing system 20 can perform a parameter sweep to determine values for the at least one response parameter which maximise the overall frequency fit of the identified system. That is, the identified response parameters are those which cause the changes in grid frequency observed for the disturbance to be best reproduced by the values of the identified response parameters which have the measure of equation (6) at an absolute minimum.

[0361] The processing system 20 can employ model reduction techniques. The determination of the response parameters can be performed at different orders (at least including third order) and then reduced to a first or second order model. Thus, conventional approximation techniques can be used to determine the required response parameters which are then mapped onto the model of the grid generator being used. In this process, there are multiple parameters and the best combination of parameters is that which maximizes the overall frequency response prediction. While the parameters of the reduced system do not necessarily correspond directly to physical quantities, they act as good predictors of the frequency evolution of the grid following further disturbances.

[0362] Response parameter determination • For the equivalent generator model (or just the swing equation, or with the governor equation), the processing system 20 can operate to determine response parameters which determine the evolution of the grid frequency in response to a disturbance. Various implementations can be used by those skilled in the art, such as grey-box identification, Kalman filtering (un-scented or extended), auto-regressive exogenous (ARX) models or other methods.

[0363] Further improvements are obtained by the processing system 20 operating to ensure the robustness of the solution. Thus, the risk of physically meaningless solutions being elicited by non-convexity when using conventional processing techniques is reduced The equations (3)-(5) relating to the equivalent of the generator can be expressed as follows

[0364] When all the power time series in these equations are known based on measurements or derived based on measurements, techniques such as ARX model techniques can be used in conjunction with the second order model to find the transfer function coefficients of the corresponding discrete dynamic system:

[0365] where, is the time step, is a noise term, is the total power of the disturbance, , , , are the coefficients to be identified. These coefficients can be found using, for example, a QR decomposition method. The parameters of the equivalent generator can be determined by this unique mapping:

[0366] where, is the data sampling time. Thus, in this setting of full observability, the above techniques or similar techniques performed automatically by the processing system 20 can be used to find the parameters of the equivalent generator.

[0367] While four parameters are determined in the above technique, for small perturbations, a first order model is sufficient, which is characterized by and Thus, the processing system 20 is operable to determine at least the parameters and as response parameters. These parameters can be used to adequately predict the frequency evolution in response to further perturbations, provided that the magnitude of the further perturbation is smaller than the magnitude threshold.

[0368] When the power grid has full PMU observability, time series data for and are available.

[0369] When the power grid does not have full PMU observability, the processing system 20 is operable to approximately determine and and use the determined values to compensate (at least partially compensate) for the lack of full measurement observability of and

[0370] The processing system 20 is operable to approximately determine the time series of as a step-like response time series, which is established using two snapshot values of the total loss: one shortly before the perturbation and one shortly after the perturbation. The processing system 20 is operable to obtain these two snapshot values from a steady state estimator. Such a steady state estimator is typically run at the control center of any TSO. Alternatively or additionally, the processing system 20 is also operable to sum the time series of losses on the lines monitored by the PMUs and then scale the sum accordingly based on the snapshot of the total loss from the steady state estimator.

[0371] The processing system 20 is operable to approximately determine the deviation of the load power from the perturbation using time series measurements of the voltage magnitude. In case of multiple measurement locations, the processing system 20 is operable to determine a single voltage time series using an average or a weighted average.

[0372] In the latter case, the processing system 20 can use a multiple-input single-output (MISO) ARX formulation, where one input is the load power deviation and the other input is the voltage deviation. Alternatively, the processing system is also operable to perform grey-box identification based on physical equations.

[0373] Preferably, and as will be described in detail elsewhere herein, the processing system is operable to represent the load change as a result of the perturbation as a linear function of the voltage change:

[0374] wherein ​for the voltage time series, is an unknown proportionality factor. To preserve the convexity of the optimization problem, the processing system can be operable to perform a parameter sweep over different values of the proportionality factor and then, for example, solve the ARX formula, which approximates the change in load power caused by the deviation by equation (14), and implicitly identify the proportionality factor that maximizes or minimizes a metric, which is considered to be a function of not only the generator equivalent parameters, but also . In this way, the convexity of the optimization problem is preserved, thereby ensuring the reliability of the solution. The scan range of for the parameter sweep can be set by the processing system 20 based on, for example, the total load in the system.

[0375] The processing system 20 can be operable such that the optimization problem that minimizes or maximizes the metric is a constrained optimization problem. One or more constraints can be added to, for example, enhance stability, etc. Alternatively or additionally, the processing system need not identify the generator equivalent and / or the parameters of the power swing equation, but can determine other response parameters. Alternatively or additionally, the voltage magnitude measurements need not be used in equation (14). These techniques will remain operable with other measurement quantities.

[0376] Figure 8 is a flowchart of a method 90. The method 90 can be performed automatically by the processing system 20.

[0377] At processing block 91, the processing system 20 receives data. The data includes measurements. The measurements include at least frequency measurements indicative of the evolution of the grid frequency in response to the disturbance. The measurements can include power measurements or measurements of other electrical quantities that allow the determination of the time-dependent evolution of the change in load power in response to the disturbance. The measurements can include power measurements and / or data from a steady-state estimator that allows the determination of the time-dependent evolution of the change in loss power in response to the disturbance.

[0378] At processing block 92, the processing system 20 can acquire information about the disturbance. The information about the disturbance can be acquired from a memory 23 internal to the processing system 20 or a data repository external to the processing system 20. This is particularly applicable in the case where the disturbance is introduced intentionally for the purpose of estimating at least one response parameter. Alternatively or additionally, the processing system 20 can also be operable to derive the information about the disturbance from the measurements received at processing block 91. The information about the disturbance can include at least the amplitude of the disturbance and timing information, such as the initial time or an estimate of the initial time.

[0379] At processing block 93, the processing system 20 determines one or more response parameters. The one or more response parameters can include parameters of the generator equivalent model, parameters of the power swing equation, or other parameters of linear or non-linear response theory that allow for a prediction of the evolution of the grid frequency in response to a further disturbance, provided that the magnitude, location, and optionally frequency of the further disturbance are known.

[0380] At processing block 94, the processing system 20 can perform an operation based on the at least one response parameter. Alternatively or additionally, the processing system 20 can also provide the at least one response parameter to another system, such as a control center of a TSO. The at least one response parameter can be used by the processing system 20 or a control system communicatively coupled to the processing system 20 to perform any one or any combination of the following: output a warning alert if the evolution of the grid frequency predicted based on the at least one response parameter indicates that the grid frequency is projected to deviate from an acceptable frequency range; automatically change a generator unit commitment; automatically trigger a mitigation operation that ensures that the grid frequency remains within an acceptable frequency range; use the at least one response parameter for decision logic to determine whether it is necessary to take a mitigation operation to ensure that the grid frequency remains within an acceptable frequency range.

[0381] Figure 9 is a flowchart of a method 100. The method 100 can be performed automatically by the processing system 20.

[0382] At processing block 101, the processing system 20 determines a time-dependent evolution of power in response to a disturbance introduced to the grid. The time-dependent evolution of power can be a time-dependent evolution of total power deviation, which is composed of the disturbance power, a change in load power, and a change in loss power. The time-dependent evolution of power can be a time-dependent evolution of total power deviation, which is composed of the disturbance power, a change in load power, a change in loss power, and a change in tie-bar power.

[0383] At processing block 102, the processing system 20 determines a time-dependent evolution of the grid frequency in at least a portion of the grid. The processing system 20 can determine the time-dependent evolution of the grid frequency based on the measurements. Processing such as averaging or weighted averaging can be performed based on the measurement locations relative to the disturbance location to determine the evolution of the grid frequency in response to the disturbance.

[0384] At processing block 103, the processing system 20 determines at least one response parameter based on the time-dependent evolution of the total power determined at processing block 101 and the time-dependent evolution of the grid frequency determined at processing block 102.

[0385] Figure 10is a flowchart of a method 110. The method 110 can be performed automatically by the processing system 20. The method 110 can be performed by the processing system 20 to implement the determination of the change in load power at processing block 111 in the method 100. Figure 9

[0386] At processing block 111, the processing system 20 can determine the time-dependent evolution of the disturbance power as a function of time. The time-dependent evolution of the disturbance power can be determined based on, for example, parameters of the means for actively controlling the introduction of a time-varying (e.g., sinusoidal or multi-sinusoidal) disturbance into the power grid.

[0387] At processing block 112, the processing system 20 can determine the time-dependent evolution of the loss power in response to the disturbance. The processing system 20 can determine the time-dependent evolution of the loss power based on the steady-state estimator output and / or the PMU measurements and / or other electrical measurements. The processing system 20 can also determine the time-dependent evolution of the load power in response to the disturbance. The processing system 20 can determine the time-dependent evolution of the load power based on a single measurement and / or other electrical characteristics, assuming a linear relationship between the change in load power and the change in voltage magnitude or other electrical characteristics (such as, for example, in equation (14)), where the proportionality constant is determined by the processing system 20.

[0388] Figure 11 is a flowchart of a method 120. The method 120 can be performed automatically by the processing system 20. The method 120 can be performed by the processing system 20 to implement the determination of the change in load power at processing block 111 in the method 110. Figure 10

[0389] At processing block 121, the processing system 20 obtains voltage measurements. The voltage measurements can be or can include the voltage magnitude of at least one bus to which the load is connected.

[0390] At processing block 122, the processing system 20 estimates the change in load power caused by the disturbance from the change in voltage magnitude caused by the disturbance. The proportionality constant that couples the change in load power to the change in voltage magnitude can be determined by the processing system 20 in such a way that the metric is optimized not only for the at least one response parameter to be determined, but also for the proportionality constant.

[0391] Figure 12 is a flowchart of a method 130. The method 130 can be performed automatically by the processing system 20. The method 130 can be performed by the processing system 20 to implement the determination of the change in load power based on the voltage magnitude at processing block 122 in the method 120. The method is an optimization routine for determining the proportionality factor between the change in load power caused by the disturbance and the measured change in voltage magnitude caused by the disturbance. Figure 11 ​​​

[0392] At processing block 131, processing system 20 selects a candidate scale factor. The candidate scale factor may be selected from a scale factor range. The scale factor range may be set based on the total load.

[0393] At processing block 132, processing system 20 executes an optimization routine to find at least one parameter of a parametric equation (e.g., a generator model and / or a power swing equation) that couples the time-dependent evolution of the grid frequency and the time-dependent evolution of the total power deviation caused by the disturbance. At processing block 133, the processing system 20 determines a value of a metric that quantifies the difference between the measured evolution of the grid frequency and the evolution calculated for the identified optimal parameters. This value depends on the candidate scaling factor.

[0394] At processing block 134, it is determined whether any other candidate scale factors remain. For example, processing system 20 may determine whether the entire scan range has been scanned. If any other candidate scale factors remain, the method returns to processing block 131.

[0395] At processing block 135, a scaling factor is selected for which the value of metric 133 is at a global optimum (which may be minimum or maximum, depending on the metric under consideration). At least one response parameter defining the global optimum of the metric is defined by the parameter value determined at processing block 132 for the corresponding scaling factor.

[0396] Figure 13 Shown Figures 8-12 Any operation in any method.

[0397] Figure 13 The time-dependent evolution of a measurement 141 of the grid frequency in response to a disturbance is shown. Figure 13 Also shown are the time-dependent evolutions of the grid frequency 142, 143, which are determined for different values ​​of at least one response parameter and, if applicable, different values ​​of a proportionality constant coupling the variation in load power to the measured variation in an electrical characteristic, such as in equation (14). The method disclosed herein provides at least one response parameter for computationally determining (if necessary in a predictive manner) a time-dependent evolution 143 of the grid frequency that mimics the actual measured evolution 141.

[0398] In any of the processing systems, methods, and systems disclosed herein, the processing system 20 is operable to interact with the controllable disturbance source 13. The processing system 20 is operable to determine parameters of the disturbance generated by the disturbance source 13.

[0399] Figure 14A block diagram of system 10 is shown. Disturbance source 13 can be controlled by a disturbance source controller 39 of processing system 20. Disturbance source 13 can also feed data or measurements to processing system 20. Processing system 20 can be operable to receive data or measurements from disturbance source 13 and can use this information to determine at least one response parameter.

[0400] Although Figure 14 Although a single disturbance source 13 is shown schematically in Fig. 1, processing system 20 can be operable to control multiple disturbance sources, which can be arranged at different locations of power grid 11.

[0401] Figure 15 A flowchart of a method 160 is shown. Method 160 can be performed automatically by processing system 20. Method 160 can be performed automatically by processing system 20 to determine parameters for controlling disturbance source 13, such as the amplitude, time dependence and / or location of a disturbance introduced into power grid under control of processing system 20.

[0402] At processing block 161, processing system 20 determines one or more disturbance parameters of the disturbance. The one or more disturbance parameters can be determined based on one or more target purposes, such as avoiding exciting any known electrical and / or mechanical resonances in the power grid. Processing system 20 can determine the frequency of a sinusoidal disturbance or all frequencies of a multi-sinusoidal disturbance and / or an envelope of the disturbance, such that all spectral components of the disturbance are spaced apart from known electrical and / or mechanical resonances in the power grid by at least a given minimum spacing. The given minimum spacing can be fixed or can be configurable.

[0403] At processing block 162, processing system 20 can control one or more devices to introduce the disturbance into the power grid in accordance with the disturbance parameters determined at processing block 161.

[0404] In any of the processing systems, methods and systems disclosed herein, processing system 20 can be operable to perform a mitigation operation that reduces or can even eliminate a change in voltage amplitude at the load caused by the at least one disturbance. In other words, processing system 20 can be operable to take countermeasures to reduce a change in voltage amplitude of the load that would otherwise be caused by the disturbance (a reduction in the change in voltage amplitude can be obtained compared to not taking countermeasures). This is particularly applicable in the case that the disturbance is actively introduced into the power grid for the dedicated purpose of determining at least one response parameter.

[0405] Figure 16 A block diagram of system 10 is shown. Disturbance source 13 can be controlled by a disturbance source controller 39 of processing system 20.

[0406] The system 10 includes at least one device 153 that is selectively controlled by, for example, the processing system 20 or a control system communicatively coupled to the processing system 20. The at least one device 153 operates so that, when activated, it reduces the change in voltage magnitude of the voltage load that would otherwise be caused by the disturbance introduced by the disturbance source 13. The at least one device 153 can include a device that operates to adjust reactive power. The at least one device 153 can include a controllable reactive component, such as a controllable reactance.

[0407] The processing system 20 can include a countermeasure controller 152. The countermeasure controller 152 can operate to determine whether to control the at least one device 153 to reduce the change in voltage magnitude. The countermeasure controller 152 can operate to determine timing and / or other operating parameters of the at least one device 153. Although Figure 16 The device 153 is only shown schematically in the figure, but the system 10 can include multiple reactive devices 153 that can be operated under the control of the processing system 20 and deployed at different locations of the power grid 11.

[0408] Figure 17 is a flowchart of a method 170. The method 170 can be performed automatically by the processing system 20 alone or in combination with a control system, such as a TSO control center, that is communicatively coupled to the processing system 20.

[0409] At processing block 171, the processing system 20 receives data specifying a disturbance and frequency data that is a response of the power grid to the disturbance. The processing system 20 can use the data specifying the disturbance to generate a time series of disturbance power values. The processing system 20 can use the data specifying the response of the power grid to determine a time series of total power deviation in response to the disturbance and a time series of power grid frequency values. The processing system 20 can determine at least one power grid response parameter based on these time series.

[0410] At optional processing block 172, the processing system 20 can collect additional data, such as tie-line power flow, bus voltage, and loss snapshots. The collection of bus voltage magnitude has utility: when there is no appropriate metering instrument to directly measure the load power, the processing system can use the bus voltage magnitude to approximately determine the load power change in response to the disturbance. However, if the power grid is fully observable, then the load change can be directly measured or calculated instead of using the bus voltage-based approximate determination. Additionally or alternatively, loss snapshots determined based on, for example, state estimation of the TSO control center can be used by the processing system 20 to approximately determine the loss power change in response to the disturbance. If the loss change can be directly measured, then the processing system 20 can use the measurement of the loss change for better accuracy.

[0411] At optional processing block 173, the processing system 20 performs signal pre-processing. Depending on the type of disturbance (active disturbance or passive large disturbance), the pre-processing can include low pass filtering, band pass filtering and / or other processing such as normalization and removal of outliers.

[0412] At processing block 174, the processing system 20 performs inertia determination through system identification. Any of the techniques described in detail herein can be used to determine the inertia and / or other grid response parameters based on the time series obtained at processing blocks 171 and 172.

[0413] The processing system 20 can be operable to extract parameters from the identified system and map them to parameters of an equivalent generator.

[0414] At processing block 175, the determined at least one response parameter can be used to take action. The action can include comparing the estimated inertia to a pre-determined threshold, simulating grid response to various disturbances, changing the balance of kinetic and electrical energy to improve the stability of the grid frequency and / or ensuring that the grid frequency remains within a target frequency range even when any of a set of further disturbances are introduced, changing generator unit commitment to ensure that the grid frequency remains within a target frequency range even when further disturbances are introduced, without limitation.

[0415] If there is tie-line power flow, the processing system 20 can separate the grid into various different regions (separated from each other by tie-lines) and can perform the determination of the at least one grid response parameter in a region specific manner. The processing can consider the tie-line power flow, preferably in determining the at least one response parameter, based on any of the techniques disclosed herein. Separation into multiple regions further improves the accuracy of the estimation.

[0416] Figure 18 is a block diagram of a system 10 according to an embodiment. The system 10 includes a processing system 20 that can be operable according to any of the embodiments disclosed herein.

[0417] The system 10 includes at least one server 191 that is operable to provide forecast information that can impact the grid frequency. The forecast information can include weather forecast information, particularly in the case of a grid having renewable energy access. The forecast information can include load forecast information that can be based on historical data.

[0418] The processing system 20 and / or the control system 180 of the system 10 can be communicatively coupled with at least one server 191 through a private network, a wide area network 190, or the Internet. The processing system 20 and / or the control system 180 can be operable to retrieve prediction information from the at least one server 191 and use the prediction information in conjunction with the at least one response parameter determined by the processing system 20 to predict an expected future evolution of the grid frequency in response to a further disturbance, respectively assuming that the operating conditions of the grid will be consistent with the prediction information.

[0419] The system 10 includes a human-machine interface (HMI) 192 that can be communicatively coupled with the processing system 20 and / or the control system 180. The processing system 20 and / or the control system 180 can be operable to cause an alert, a warning, or other information to be output via the HMI 192. For example, the processing system 20 can cause the at least one response parameter to be output via the HMI 192 for further use. Alternatively or additionally, the processing system 20 and / or the control system 180 can cause the frequency nadir or RoCoF predicted based on the at least one response parameter to be output via the HMI 192.

[0420] The system 10 includes a control system 180. The processing system 20 can be operable to provide the at least one response parameter or information derived therefrom, such as a frequency nadir or RoCoF predicted based on the at least one response parameter for an expected further disturbance, to the control system 180.

[0421] The control system 180 can be operable to receive the at least one response parameter or information derived therefrom from the processing system 20 at a control system interface 181. The control system 180 can be operable to store the at least one response parameter or information derived therefrom in a control system memory 182.

[0422] The control system 180 can include one or more control circuits 183. The at least one control circuit 183 can include an integrated circuit, an integrated semiconductor circuit, a processor, a controller, an application specific integrated circuit (ASIC), a field programmable gate array (FPGA), any one or any combination of circuits including qubits and / or quantum gates, without limitation.

[0423] The at least one control circuit 183 can be operable to perform a frequency change prediction 184 to determine one, more, or all of the following based on the at least one response parameter and information about a possible further disturbance: a prediction of an evolution of the grid frequency in response to the further disturbance; a prediction of a RoCoF in response to the further disturbance; a prediction of a frequency nadir in response to the further disturbance.

[0424] The at least one control circuit 183 is operable to execute a generator unit and / or load schedule 185. The at least one control circuit 183 is operable to adjust, for example, a generator unit commitment based on results of the frequency change prediction 184. The at least one control circuit 183 is operable to execute control operations, such as control operations on at least one of the at least one power generation unit, the at least one rotating energy storage system, and / or at least one disconnect switch or circuit breaker in the power grid. The control operations are operable to reduce changes in the power grid frequency and / or to keep the power grid frequency within a target frequency band. The control operations can be selectively taken and / or selected based on results of the frequency change prediction 184.

[0425] When the power grid has renewable energy sources access, the at least one control circuit 183 is operable to determine adjusted settings for an energy management system (EMS) 193 and / or a power management system (PMS) 194. The system 10 can include the EMS 193 and / or the PMS 194. The control system 180 is operable to cause the EMS 193 and / or the PMS 194 to use the adjusted settings to reduce the risk of the power grid frequency deviating from the target band.

[0426] The at least one control circuit 183 is operable to execute HMI control 186. The at least one control circuit 183 can cause an alert, warning, or other information based on the at least one response parameter to be output via the HMI 192.

[0427] Figure 19 is a flowchart of a method 200. The method 200 can be executed automatically by the control system 180.

[0428] At process block 201, the control system 180 receives one or more response parameters determined by the processing system 20. The control system 180 can receive the at least one response parameter by actively querying the processing system 20. The control system 180 can receive the at least one response parameter by a push process from the processing system 20.

[0429] At process block 202, the control system 180 predicts an expected upcoming change in the power grid frequency that will be caused by a further disturbance. The prediction is based on the at least one response parameter and parameters of the further disturbance, such as magnitude, location, frequency, and / or duration. The prediction can optionally be based on prediction information, such as weather data, that can affect power generation by renewable energy sources.

[0430] At processing block 203, the control system 180 performs a control operation in response to the prediction determined at processing block 202. The control operation can be selectively performed, if desired, to ensure that the grid frequency remains within the target frequency band. The control operation can be selected in accordance with the prediction result determined at processing block 202. The control operation can include one or more of the following: changing a generator unit commitment; adjusting a setting of the EMS; adjusting a setting of the PMS.

[0431] Various effects and advantages can be obtained by the processing system, method and system according to the embodiments. The processing system and processing method allow to determine at least one response parameter that contributes to predicting a future grid frequency change with a higher accuracy. Thus, the processing system and processing method allow to take control operations that contribute to keeping the grid frequency within a desired frequency band.

[0432] Although the embodiments have been described in detail referring to the accompanying drawings, various modifications can be implemented in other embodiments. For illustration and not limitation: • While embodiments have been described that operate to determine at least the inertia of a generator equivalent model, the processing system and processing method can operate to determine any one or any combination of the following parameters of the grid response: inertia, damping, governor time constant, governor gain, load sensitivity to voltage magnitude. The processing system and processing method can operate to estimate any other parameter of the grid frequency response and / or to perform the determination in a region-specific manner, e.g. by determining the inertia and optionally other parameters in a region-specific manner. The processing system and processing method can operate to determine the parameters, such as the inertia, for a plurality of equivalent generators per grid or per region.

[0433] • While embodiments have been described that determine at least one response parameter based on a generator equivalent model and a power swing equation, other equations that couple the time-dependent evolution of the grid frequency with the time-dependent evolution of the power can also be used.

[0434] • While embodiments have been described that the processing system operates to approximately determine the change in load power based on measurements of the bus voltage magnitude, these techniques can also be extended to other electrical measurements that are additionally or alternatively measured. While techniques have been described that the processing system assumes a linear relationship between the change in bus voltage magnitude and the change in load power caused by a disturbance, more complex modeling techniques can also be used, such as higher order dependencies in which not only a scaling factor can be determined, but also a plurality of parameters of a model that describe the dependency of the change in load power caused by a disturbance on the change in bus voltage magnitude and / or other electrical measurements.

[0435] • While embodiments have been described in which the processing system operates to approximately determine the change in loss power based on the output of a steady state estimator, other techniques for approximately determining the change in loss power can also be used. For example, imputation techniques can be used.

[0436] • While embodiments have been described in which the at least one response parameter is used to adjust generator unit commitments and / or adjust settings of the EMS and / or PMS, the at least one response parameter can also be used for other purposes. For example, the at least one response parameter can be stored in an IED and used by decision logic of the IED during field operation of the power grid.

[0437] • While embodiments have been described in which the disturbance source 13, the measuring instrument 12, the control system 14, 180 and the reactive power device 153 are shown separately from the power grid 11, some or all of these components can also be integrated with the power grid.

[0438] Embodiments can be used in association with power grids having renewable energy access, such as power grids including renewable energy systems, such as DERs, without being limited thereto.

[0439] This specification and the accompanying drawings should not be construed as limiting the claims of the protected invention. In other words, while the present invention has been illustrated and described in detail in the drawings and foregoing description, such illustration and description are to be considered illustrative or exemplary only and not restrictive. Various mechanical, compositional, structural, electrical, and operational changes can be made without departing from the spirit and scope of the present specification and claims. In some instances, well-known circuits, structures, and techniques have not been shown in detail in order not to obscure the present invention. Thus, it will be understood that changes and modifications can be made by those of ordinary skill within the scope and spirit of the claims. In particular, further embodiments of the present invention covering any combination or sub-combination of the features from the different embodiments described above and below are contemplated. In particular, further embodiments of the present invention covering any combination or sub-combination of the features from the different embodiments described above and below are contemplated.

[0440] The present disclosure also covers all additional features individually shown in the figures, although they can not be described in the foregoing or following description. Furthermore, individual alternatives to the embodiments described in the figures and description can be abandoned from the subject matter of the invention or from the disclosed subject matter. The present disclosure comprises the subject matter consisting of the features defined in the claims or embodiments and the subject matter consisting of the features described.

[0441] The term "comprising" does not exclude other elements or steps, and the indefinite article "a" or "an" does not exclude a plurality. A single unit or processing block can fulfil the functions of several features recited in the claims. The fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage. Components described as coupled or connected can be electrically or mechanically directly coupled or they can be indirectly coupled via one or more intermediate components. Any reference signs in the claims should not be construed as limiting the scope.

[0442] Machine readable instruction code can be stored / distributed on a suitable medium, such as an optical storage medium or a solid-state storage medium supplied together with or as part of other hardware, but can also be distributed in other forms, such as via wired or wireless communication networks. Further, machine readable instruction code can also be a data structure product or signal per se, for example embodying a specific method, such as a method according to an embodiment.

Claims

1. A method for processing measurements acquired in an electrical network (11), the method being performed by a processing system (20) and comprising: causing a disturbance to be introduced into the power grid (11) according to a disturbance parameter determined by the processing system (20), receiving data comprising frequency data, wherein the frequency data represents a grid frequency measured at one or more locations in the grid (11), and processing the received data to determine at least one response parameter for at least a portion of the grid (11), the at least one response parameter influencing a grid frequency response of the at least a portion of the grid (11) to a further disturbance; wherein causing the disturbance to be introduced into the power grid (11) comprises controlling at least one device (13, 153) comprised by or coupled to the power grid (11) to introduce the disturbance, The method further comprises controlling the at least one device (13, 153) to reduce the effect of the disturbance on the voltage amplitude in the power grid (11).

2. The method of claim 1, further comprising setting a disturbance frequency of the disturbance based on an electrical resonance and / or a mechanical resonance of the power grid (11).

3. The method according to claim 2, wherein: The disturbance frequency is set to prevent excitation of the electrical resonance and / or mechanical resonance of the power grid (11).

4. The method according to claim 2 or claim 3, wherein: The disturbance frequency is set such that the disturbance frequency is spaced from any one of the electrical resonance and / or mechanical resonance of the power grid (11) by at least a threshold value.

5. A method as claimed in any one of the preceding claims, wherein The at least one device (13, 153) is controlled to cause the disturbance to be introduced into the power grid (11) in a repetitive manner.

6. A method as claimed in any one of the preceding claims, wherein Causing the disturbance to be introduced into the power grid (11) comprises controlling a plurality of devices (13, 153) arranged in different regions of the power grid (11) to determine the at least one response parameter as a region-specific response parameter.

7. A method as claimed in any one of the preceding claims, wherein The at least one device (13, 153) comprises means (13, 153) operative to influence reactive power in the grid (11).

8. A method as claimed in any one of the preceding claims, wherein Processing the received data includes determining the at least one response parameter based on a total power deviation, wherein the total power deviation includes a power of the introduced disturbance and a change in power in response to the introduced disturbance.

9. The method of claim 8, wherein: Processing the received data includes determining a change in load power and / or a change in power loss caused by the disturbance.

10. The method of claim 9, wherein: Processing the received data comprises determining a change in the load power based on a voltage amplitude measurement, and / or wherein processing the received data comprises determining a change in the lost power based on a state estimation.

11. The method according to any one of the preceding claims, further comprising: storing said at least one response parameter in a protection and / or monitoring device for use in evaluating the impact of said further disturbance on the grid frequency; and / or The at least one response parameter is used by the processing system (20) to assess an impact of the further disturbance on the grid frequency, optionally wherein the assessed impact includes a rate of change of frequency, RoCoF and / or a frequency minimum.

12. The method of any of the preceding claims, further comprising performing an operation by the processing system (20) or a control, monitoring and / or protection system (180) or device coupled to the processing system (20) based on the at least one response parameter, optionally wherein, The operations include corrective operations and / or mitigation operations to reduce grid frequency variations caused by the further disturbance.

13. A method for protecting and / or monitoring an electric grid (11), comprising: using the at least one response parameter by a grid (11) control, monitoring and / or protection system (180) or device to assess the impact of the further disturbance on the grid frequency; and An operation is performed based on the at least one response parameter.

14. Machine-readable instruction code, which, when executed by at least one programmable circuit, causes the at least one programmable circuit to perform the method according to any one of the preceding claims.

15. A processing system (20) for processing measurements acquired in an electrical network (11), the processing system (20) comprising: Interfaces (21, 22); and At least one processing circuit (30), the at least one processing circuit operating to: Determine the perturbation parameters, generating a command and outputting the command via the interface (21, 22) to cause a disturbance to be introduced into the power grid (11) according to the disturbance parameter, receiving data comprising frequency data, wherein the frequency data represents a grid frequency measured at one or more locations in the grid (11), and processing the received data to determine at least one response parameter for at least a portion of the grid (11), the at least one response parameter influencing a grid frequency response of the at least a portion of the grid (11) to a further disturbance, The processing system (20) is operated such that, in order to introduce the disturbance into the power grid, the processing system is operated to control at least one device (13, 153) included by the power grid (11) or coupled to the power grid (11) to introduce the disturbance and reduce the effect of the disturbance on the voltage amplitude in the power grid (11).

16. The processing system (20) of claim 15, wherein: The processing system (20) is operative to perform the method according to any one of claims 1 to 13.

17. A system comprising Power grid (11); The processing system (20) of claim 15 or claim 16, the processing system being operative to determine the at least one response parameter; and A control, monitoring and / or protection device (180) is operable to perform a corrective action or a mitigation action in response to the further disturbance based on the at least one response parameter.

Images