EP4490826A1

EP4490826A1 - A smart load system and method for regulating power supplied from an ac voltage distribution system

Info

Publication number: EP4490826A1
Application number: EP22726728.3A
Authority: EP
Inventors: Alvaro RUBINOS SICRE; Alvaro MORALES MUNOZ; Francisco Daniel FREIJEDO FERNÁNDEZ
Original assignee: Huawei Digital Power Technologies Co Ltd
Current assignee: Huawei Digital Power Technologies Co Ltd
Priority date: 2022-05-02
Filing date: 2022-05-02
Publication date: 2025-01-15
Also published as: CN117561660A; WO2023213377A1

Abstract

This disclosure relates to regulating power supplied from an alternating current (AC) voltage distribution system to a main load and a smart load. The disclosure proposes a smart load system comprising a controller and the smart load and a corresponding method for operating the smart load system, wherein the controller is configured to: estimate an AC frequency of the AC voltage distribution system; and determine a reference power based on the estimated AC frequency; wherein the smart load is configured to either consume power from the AC voltage distribution system or provide power to the main load, depending on the reference power, and wherein the smart load is electrically connectable in parallel to the main load.

Description

A SMART LOAD SYSTEM AND METHOD FOR REGULATING POWER SUPPLIED FROM AN AC VOLTAGE DISTRIBUTION SYSTEM

TECHNICAL FIELD

This disclosure relates to regulating power supplied from an alternating current (AC) voltage distribution system to a main load and a smart load. The disclosure proposes a smart load system comprising a controller and the smart load, and a corresponding method for operating the smart load system.

BACKGROUND

Nowadays, electric grids and/or power systems are facing a high penetration of renewable energy sources and, simultaneously, a reduction of energy generated by traditional power plants. Thus, the number of synchronous machines, which are usually used for traditional power plants, is reducing.

From an overall power system perspective, the removal of synchronous machines leads to the following: the aggregated inertia provided by multiple synchronous machines is being removed and, in principle, not replaced; and the lack of inertia in the power system perturbs the power system frequency control, as the power system balancing among generation and consumption becomes unfeasible.

Inertia is the ability of a system to resist against changes in power consumption or provision. For example, the inertia of the frequency of an alternating current (AC) voltage distribution system. Conventional synchronous machines store energy as kinetic energy in a rotating mass. A difference between power production and power consumption leads to an acceleration, or deceleration of a power system frequency. These fluctuations can be absorbed by the rotating masses present in the synchronous machines that interface generation units with the power system. Comparatively, power electronic devices lack such a mechanism. Conventionally, this is mitigated by the power generation side by requesting power plants to comprise a certain level of inertia, thus making energy storage systems required in many electronic based power plants.

However, it is not convenient to use energy storage devices in the production side, while it is common to use energy storage devices in the consumption side. The frequency support can be shared between both the production and the consumption side.

New methods for frequency control and grid support are needed, in order to complete the transition from traditional power plants to renewable energy sources, while achieving a similar performance as synchronous machines.

SUMMARY

In view of the above, this disclosure aims to provide a smart load system and a corresponding method that is able to reliably and/or precisely regulate power supplied from an AC voltage distribution system to a main load and a smart load.

These and other objectives are achieved by the solutions of this disclosure as described in the independent claims. Advantageous implementations are further defined in the dependent claims.

A first aspect of this disclosure provides a smart load system for regulating power supplied from an alternating current, AC, voltage distribution system to a main load and a smart load, wherein the smart load system comprises a controller configured to: estimate an AC frequency of the AC voltage distribution system; and determine a reference power based on the estimated AC frequency; and wherein the smart load system further comprises the smart load configured to either consume power from the AC voltage distribution system or provide power to the main load, depending on the reference power, and wherein the smart load is electrically connectable in parallel to the main load.

The AC voltage distribution system may be supported, stabilized, and/or regulated, while simultaneously ensuring that the main load receives sufficient power. The smart load may be configured to additionally or alternatively provide power to the AC voltage distribution system. For example, if the main load already receives sufficient power and the smart load comprises further excess energy that can be used. Thus, the voltage distribution system may be further supported and/or stabilized.

Power electronics are not restricted to only providing a source of power, but frequency control features may be added to “smart loads”. A combination of power electronics control and energy storage may permit implementing functionalities that contribute to the inertia “as seen” by the power system, for example, the grid. Thus, a smart load system may enrich the basic functionalities of conventional power supply concepts with frequency control or emulated inertia capabilities.

In an implementation form of the first aspect, the main load is a critical load and/or a direct current (DC) load, and/or the smart load is a non-critical load and/or a DC load.

Conventionally, when the demand for electricity nears available supply levels, it is sometimes necessary to temporarily interrupt the delivery of electricity to maintain the integrity of an electric grid and to prevent catastrophic grid failures and extended outages for customers, which is referred to as “load shedding”. Such load shedding may be prevented by using the smart load system of this disclosure.

The main load and/or the smart load may be a low voltage DC load.

For example, a smart load system may be used as a high-tech solution for interfaces between AC medium voltage distribution grids and low voltage DC systems where critical loads are placed.

A highest priority may be to prevent disconnecting critical loads and/or to guarantee sufficient power is supplied to the critical loads at all times.

Disconnecting and/or shutting off critical loads may lead to, for example, economic losses, security losses, and/or loss of critical information. In a further implementation form of the first aspect, the smart load is connectable to the AC voltage distribution system by at least a power electronics converter, and/or the smart load is electro-magnetically connectable to the AC voltage distribution system, and/or the smart load is electro-magnetically connectable to the AC voltage distribution system by at least an AC to DC power electronics converter.

The power electronics converter may be an AC to DC power electronics converter and/or a solid-state transformer.

Alternatively or additionally, the smart load may be connectable to the AC voltage distribution system by at least a rectifier.

In a further implementation form of the first aspect, the controller is configured to determine the reference power based on the estimated AC frequency and a predetermined reference frequency, for example, based on a difference between the estimated AC frequency and the predetermined reference frequency.

In a further implementation form of the first aspect, the predetermined reference frequency is a nominal AC frequency of the AC voltage distribution system.

The nominal AC frequency may be 50 Hz.

In a further implementation form of the first aspect, the controller is configured to determine the reference power based on a linear relation between power and frequency, and/or based on a linear droop control.

For example, the difference between the estimated AC frequency and the predetermined reference frequency may be converted to the reference power based on a linear relation between power and frequency.

For example, the difference between the estimated AC frequency and the predetermined reference frequency may be converted to the reference power by multiplication with a constant predetermined factor. Thus, the reference power may be determined quickly and easily. Further, the determined reference power may be close to the optimal reference power.

A reference power may be considered optimal if it ensures sufficient power is supplied to the main load, while stabilizing the AC frequency of the AC voltage distribution grid as much as possible.

In a further implementation form of the first aspect, the smart load is configured to consume power from the AC voltage distribution system if the reference power is positive, and is configured to provide power to the main load if the reference power is negative, and/or the smart load is configured to consume power from the AC voltage distribution system or provide power to the main load equal to the absolute value of the reference power.

The power consumption and provision may be adjusted by the smart load or by the controller.

The smart load may absorb and/or store energy and/or power from the AC voltage distribution system.

In a further implementation form of the first aspect, the controller is configured to estimate the AC frequency by estimating and/or predicting a real-time frequency of the AC voltage distribution system.

The AC frequency may be estimated and/or predicted continuously and/or in real-time. Thus, the reaction time of the smart load system to frequency deviations of the AC voltage distribution system may be reduced or minimized.

By predicting the AC frequency, the reaction time of the smart load system to frequency deviations of the AC voltage distribution system may be reduced or minimized.

Predicting or estimating the real-time frequency of the AC voltage distribution system may be based on, for example, at least one measurement of the real-time frequency of the AC voltage distribution system, and/or a real-time frequency change of the AC voltage distribution system. In a further implementation form of the first aspect, the controller is configured to estimate the AC frequency based on a phase-locked loop algorithm.

In a further implementation form of the first aspect, the controller is configured to continuously and/or in real-time adjust the reference power based on the estimated AC frequency.

The AC frequency of the AC voltage distribution system may fluctuate. Thus, the estimated AC frequency and the reference power may be updated continuously and/or in real-time to react to the AC frequency fluctuations.

In a further implementation form of the first aspect, the smart load is configured to continuously and/or in real-time adjust its power consumption from the AC voltage distribution system or its power provision to the main load, based on the reference power.

The power provision or consumption of the smart load may be adjusted faster and/or more accurately compared to conventional systems.

The smart load may continuously and/or in real time adjust its power consumption or provision based on the reference power to precisely provide or consume power as indicated by the reference power.

Thus, a difference between the reference power and power provision or consumption may be continuously and/or in real-time reduced or minimized.

The reaction time of the smart load system to frequency deviations of the AC voltage distribution system may be reduced or minimized.

In this disclosure “minimized” may refer to “minimized according to the operational limitations of the main load, the smart load, the AC voltage distribution system, and/or a corresponding power electronics system (PES)”. For example, a “minimized reaction time of the smart load system” may refer to a reaction time of the smart load system to frequency deviations of the AC voltage distribution system that is faster or on a same order of magnitude as a reaction time and/or a processing speed of one or more power electronics actuators, for example one or more or all power electronics actuators that are comprised in the main load, the smart load, the AC voltage distribution system, and/or a corresponding PES.

In this disclosure “reduced” may refer to “reduced compared to conventional solutions”.

The AC frequency of the AC voltage distribution grid may be continuously and/or in real time adjusted by the smart load system. Thus, a deviation of the AC frequency of the AC voltage distribution system from a nominal frequency may be continuously and/or in real time reduced or minimized by the smart load.

In this disclosure “continuously” may refer to continuous or non-discrete actions in time. Alternatively, “continuously” may refer to discrete actions in time with a fine granularity, wherein the number of actions in time may be maximized or increased compared to conventional systems. A large number of discrete actions may approximate continuous actions. “Actions” may refer to, for example, “measurements”, “adjustments”, “estimations”, and/or “calculations”.

In a further implementation form of the first aspect, the controller is configured to provide the reference power to the smart load, or the controller is configured to provide a reference DC current, which is based on the reference power and a DC voltage of the smart load, to the smart load.

The reference power may be converted by the controller to the reference DC current.

In a further implementation form of the first aspect, the smart load is a flexible storage device, for example, a battery.

In a further implementation form of the first aspect, the AC voltage distribution system is a grid and/or an AC medium voltage distribution system. In a further implementation form of the first aspect, the smart load system is configured to not further communicate with the AC voltage distribution system and/or with an operator of the AC voltage distribution system, for example a distribution system operator.

Thus, the smart load system can be used efficiently and/or can be easily integrated with a conventional power system or any AC voltage distribution system.

One or more smart load systems may conduct a distributed frequency control action that supports and/or regulates a power-system consumption/generation equilibrium in realtime.

A second aspect of this disclosure provides a method of operating a smart load system for regulating power supplied from an alternating current, AC, voltage distribution system to a main load and a smart load, wherein the smart load system comprises a controller and the smart load, wherein the smart load is electrically connectable in parallel to the main load, and wherein the method comprises: estimating, with the controller, an AC frequency of the AC voltage distribution system, determining, with the controller, a reference power based on the estimated AC frequency, and adjusting the smart load to either consume power from the AC voltage distribution system or to provide power to the main load, depending on the reference power.

The method of the second aspect may have implementation forms that correspond to the implementation forms of the device of the first aspect. The method of the second aspect and its implementation forms achieve the advantages and effects described above for the device of the first aspect and its respective implementation forms.

A third aspect of this disclosure provides a computer program product comprising a program code for controlling a smart load system according to the first aspect or implementation forms of the first aspect, or for performing, when the program code is executed on a computer, a method according to the second aspect or implementation forms of the second aspect.

In this disclosure the phrase “power command” and “reference power” may be used interchangeably. Further, the phrase “DC current command” and “reference current” may be used interchangeably. In this disclosure the “main load” and “the smart load” combined may be referred to as a “load system”.

In this disclosure the “AC distribution system”, for example a grid, may be referred to as “AC side”, “AC grid side”, or “AC grid”. The ’’main load” and “smart load” may be referred to as “DC side”, “DC grid side”, “DC grid”. The AC side and DC side may be connected.

In this disclosure a “power electronics system” (PES), “PES based power supply” or “PES technology” may refer to a power electronics based technology that aims to extract electric power from an AC voltage distribution system by using a power converter, for example a AC to DC power converter and/or a solid state transformer. The load system may be galvanically isolated from the AC voltage distribution system side. In the load system one or more main loads may be connected, for example servers in a data center or electric power drives in a traction application.

In this disclosure one or more main loads combined may be referred to as the “main load”.

In this disclosure one or more smart loads combined may be referred to as the “smart load”.

It has to be noted that all devices, elements, units and means described in the disclosure could be implemented in the software or hardware elements or any kind of combination thereof. All steps which are performed by the various entities described in the disclosure as well as the functionalities described to be performed by the various entities are intended to mean that the respective entity is adapted to or configured to perform the respective steps and functionalities. Even if, in the following description of specific embodiments, a specific functionality or step to be performed by external entities is not reflected in the description of a specific detailed element of that entity which performs that specific step or functionality, it should be clear for a skilled person that these methods and functionalities can be implemented in respective software or hardware elements, or any kind of combination thereof. BRIEF DESCRIPTION OF DRAWINGS

The above described aspects and implementation forms will be explained in the following description of embodiments in relation to the enclosed drawings, in which

FIG. 1 shows a smart load system according to an embodiment of this disclosure.

FIG. 2 shows a smart load system according to an embodiment of this disclosure.

FIG. 3 shows a smart load system according to an embodiment of this disclosure.

FIG. 4a shows a frequency response for a smart load system of this disclosure and a conventional system.

FIG. 4b shows a power response for a smart load system of this disclosure and a conventional system.

FIG. 4c shows a battery power variation for a smart load system of this disclosure and a conventional system.

FIG. 5a shows a frequency response for a smart load system of this disclosure and a conventional system.

FIG. 5b shows a power response for a smart load system of this disclosure and a conventional system.

FIG. 5c shows a battery power variation for a smart load system of this disclosure and a conventional system.

FIG. 6 shows a method according to an embodiment of this disclosure. DETAILED DESCRIPTION OF EMBODIMENTS

FIG. 1 shows a smart load system 100 according to an embodiment of this disclosure. The smart load system 100 comprises a smart load 101 and a controller 102. Further, FIG. 1 shows an AC voltage distribution system 200 and a main load 201, wherein the smart load system 100 regulates power supplied from the AC voltage distribution system 200 to the main load 201 and the smart load 101, and the smart load 101 is electrically connectable in parallel to the main load 201. The controller 102 is configured to estimate an AC frequency 202 of the AC voltage distribution system 200, and to determine a reference power 104 based on the estimated AC frequency 103. Further, depending on the reference power 104, the smart load 101 is adapted to either consume power from the AC voltage distribution system 200 or provide power to the main load 201.

The load system may be a DC load system, for example a low voltage DC system. A low voltage DC system may run on, for example, 600 V or less of direct current.

The AC voltage distribution system 200 may have a medium voltage, for example a line to line root mean square (RMS) voltage between 600 V and 35 kV. A “line to line RMS voltage” may be referred to as a “reference value” of an AC voltage distribution system 200.

A flexible energy storage system 101 or smart load 101 may be an electric system able to absorb or deliver power as a function of a reference power/current 104 or a power/ current command 104.

A power supply based on PES technology may include at least one smart load, for example, one flexible bidirectional energy storage system 101, which is connected in parallel to the main load 201. The command that drives the smart load 101 is obtained from an AC frequency 202 estimation of an AC voltage distribution system 200, for example from a phase-locked loop or equivalent signal-processing based method.

FIG. 2 shows an exemplary application of the smart load system according to this disclosure. In particular, FIG. 2 shows a conventional PES based system, a flexible energy storage system 101 and a smart control method that drives this flexible energy storage 101, and is performed by a controller 102. The conventional PES based system includes a PES, a main load 201, and a main control method. The main control method controls the two conversion stages from AC to DC, and from DC to DC, that transfer energy from an AC (three-phase) side to a low voltage (LV) DC grid side. Thanks to the isolation in the DC to DC conversion stage, all of the modules converge to a common DC grid where the main loads 201 are connected. Further, FIG. 2 shows multiple branches of a LV DC voltage.

In FIG. 2 the smart load 101 is a flexible energy storage system 101, which may be for example a battery. The flexible energy storage system 101, which is placed in the DC side, is driven based on an estimation of a grid-frequency of the AC grid side.

In order to estimate the grid-frequency 202, a frequency estimation method may be used. The frequency estimation method may be a signal processing algorithm running in a controller 102. The controller 102 may be, for example a microcontroller 102, and may be the same controller 102 that drives the conventional PES based system. Alternatively, another controller may control the conventional PES based system. A conventional choice for frequency estimation algorithms are, for example, phase-locked loops (PLL), which may be used for the smart load system 100. Alternatively, any frequency estimation method may be used. After determining an accurate estimation 103 of the grid-frequency 202, for example in real-time, a power command 104 or reference power 104 for the flexible energy storage system 101 is calculated.

The frequency control block of FIG. 2 receives as input the estimated grid-frequency 103 f_g(t) and provides a power command or reference power 104 P_s*_t(t). Alternatively, a DC current command may be provided. The DC current command may be calculated based on power being equal to current multiplied with voltage, and in a DC-grid, voltage being approximately constant. Thus, a reference current may be approximately proportional to a reference power 104. The power command may be determined by the controller 102.

If the estimated real-time grid-frequency 103 is above a nominal value, for example, 50 Hz in a European grid, which is a well-known predetermined value, the power command 104 may be positive, which means that the energy storage system 101 may absorb and/or receive an excess of power from the grid 200. Thus, the overall power generation and power consumption of an overall power system may be balanced by increasing the load of the energy storage system 101 for the AC voltage distribution side 200. The overall power system includes the AC voltage distribution system 200 and all connected loads and power generator systems.

If the estimated real-time grid-frequency 103 is below the nominal value, the flexible energy storage system 101 may provide power to the local DC grid, for example to the main load 201. Subsequently the power received by the PES from the AC grid 200 decreases, as the load for the AC grid 200 decreases. Thus, on an overall power system level, the power generation vs. power consumption equilibrium may be balanced by reducing the overall power system load.

FIG. 3 shows a more specific example of the smart load system 100 of this disclosure for a Data Center. A PES interfaces a medium voltage (MV) AC grid for which 10 kV is determined as a reference value. The AC/DC conversion is made by the PES by a series of AC/DC modules of the unity power factor rectifier (UPFR) family. For the DC/DC conversion involving galvanic isolation at least one resonant converter can be considered. The symbol “//” represents galvanic isolation. The LV DC grid may include a nominal voltage of 400V, which is atypical value for such an application. Multiple racks of servers are connected as main loads 201, wherein the main application of the PES is to feed these main loads 201, for example, servers. The flexible energy storage system 101 is a battery system connected to the LV DC grid. The frequency control algorithm may be based on a linear relation between power and frequency, for example a linear droop function. This linear relation may be restricted around a nominal frequency. The grid-frequency 202 is estimated by a phase-locked loop algorithm connected to the medium voltage AC grid and may be already available for the main/conventional functionality. The controller 102 determines an estimation 103 of the grid-frequency 202 and a power command or reference power 104 for the flexible energy storage system 101.

An advantage of UPFR based AC-DC conversion includes that by constraining the power supply to work with a power factor equal to or almost equal to one, a more compact module with higher power density can be designed.

The advantages of series resonant converters include its robustness, power density, simplicity of operation, and high efficiency. The advantages of using a PLL includes that this algorithm is widely employed in many grid-connected applications.

The advantage of using a linear droop control to drive the battery power command 104 includes simplicity and accuracy. A linear relation between both variables around the nominal equilibrium point, for example 50 Hz, is accurately mimicking the behaviour of a synchronous machine.

A battery storage system is a cost-effective solution already available for many applications such as Photovoltaics and data centers.

Frequency control of an AC voltage distribution system 200 may comprise using a conventional PES based power supply, which is intended to feed a main load 201, for example, one or more servers connected to a low voltage DC grid of a data center. Further, said frequency control may comprise using a smart load 101, for example a flexible storage device 101, which may be physically placed in a low voltage DC side, combined with a control algorithm determining its charge/discharge based on the frequency estimation of an AC voltage distribution system 200. For example, the flexible load 101 absorbs energy if the frequency is higher than a threshold, and/or it contributes to a “braking” of the AC voltage distribution system 200. In another example, if the grid-frequency 202 is below a threshold, then flexible loads deliver an excess of energy that flows to the main load 201, such the energy absorbed from the grid is reduced, and/or the flexible load 101 contributes to an “acceleration” of the AC voltage distribution system 200.

The performance of a smart load system 100 according to an example of this disclosure is confirmed by simulations and studying the response of a grid generated via synchronous machine generators to a sudden demand variation. In the following examples, a plurality of server racks or main loads demand 0.8 MW, and a battery system or smart load is able to inject or store up to 0.2 MW.

FIG. 4a shows a frequency response to a sudden demand increase from 1 MW to 1.5 MW with a smart load 101 and without a smart load 101. FIG. 4a shows a frequency variation from a nominal frequency of 50 Hz. The frequency drop caused by the load change is reduced and/or improved compared to the case without a smart load 101. FIG. 4b shows the variation in the power demanded by the PES as a consequence of the power injected by the battery system 101. FIG. 4c shows the power consumed (+) or injected (-) by the battery in the low voltage DC (LVDC) side.

FIGs. 5 a to 5 c show performance graphs for a sudden reduction in power demand from 1.5 MW to 1 MW, and shows a similar improved frequency response due to the smart load system 100.

The smart load system 100 may suitable for, for example, data centers. Future applications may include microgrid systems fed by an external grid, intelligent buildings, bulky energy storage systems, green hydrogen generation systems, or other systems that are able to integrate a smart load 101.

Participation of consumers in grid-services, such as frequency control may be required or incentivized in the future. Currently, Transmission System Operators of several countries, for example UK, Germany, Australia, Spain, financially compensate for participating in grid-services, as it requires to invest in the infrastructure. Taking into account the already existing infrastructure of data centers, the applications of a smart load system 100 in, for example, data centers may provide economic benefits without involving much additional capital investment.

The controller 102 may be a processor 102.

Generally, the processor 102 may be configured to perform, conduct or initiate the various operations of the smart load system 100 described herein. The processor 102 may comprise hardware and/or may be controlled by software. The hardware may comprise analog circuitry or digital circuitry, or both analog and digital circuitry. The digital circuitry may comprise components such as application-specific integrated circuits (ASICs), field- programmable arrays (FPGAs), digital signal processors (DSPs), or multi-purpose processors. The smart load system 100 may further comprise memory circuitry, which stores one or more instruct! on(s) that can be executed by the processor 102, in particular under control of the software. For instance, the memory circuitry may comprise a non- transitory storage medium storing executable software code which, when executed by the processor 102, causes the various operations of the smart load system 100 to be performed. In one embodiment, the smart load system 100 may comprises one or more processors 102 and a non-transitory memory connected to the one or more processors 102. The non- transitory memory may carry executable program code which, when executed by the one or more processors 102, causes the smart load system 100 to perform, conduct or initiate the operations or methods described herein

FIG. 6 shows a method 300 according to an embodiment of this disclosure. The method

300 may be performed by the smart load system 100. The method 300 comprises a step

301 of estimating an AC frequency 202 of the AC voltage distribution system 200. Further, the method 300 comprises a step 302 of determining a reference power 104 based on the estimated AC frequency 103. Further, the method 300 comprises a step 303 of adjusting the smart load 101 to either consume power from the AC voltage distribution system 200 or to provide power to the main load 201, depending on the reference power 104.

The disclosure has been described in conjunction with various embodiments as examples as well as implementations. However, other variations can be understood and effected by those persons skilled in the art and practicing the claimed matter, from the studies of the drawings, this disclosure and the independent claims. In the claims as well as in the description the word “comprising” does not exclude other elements or steps and the indefinite article “a” or “an” does not exclude a plurality. A single element or other unit may fulfill the functions of several entities or items recited in the claims. The mere fact that certain measures are recited in the mutual different dependent claims does not indicate that a combination of these measures cannot be used in an advantageous implementation.

Claims

1. A smart load system (100) for regulating power supplied from an alternating current, AC, voltage distribution system (200) to a main load (201) and a smart load (101), wherein the smart load system (100) comprises a controller (102) configured to: estimate an AC frequency (202) of the AC voltage distribution system (200); and determine a reference power (104) based on the estimated AC frequency (103); and wherein the smart load system (100) further comprises the smart load (101) configured to either consume power from the AC voltage distribution system (200) or provide power to the main load (201), depending on the reference power (104), and wherein the smart load (101) is electrically connectable in parallel to the main load (201).

2. The smart load system (100) according to claim 1, wherein the main load (201) is a critical load and/or a direct current, DC, load, and/or wherein the smart load (101) is a non-critical load and/or a DC load.

3. The smart load system (100) according to claim 1 or 2, wherein the smart load (101) is connectable to the AC voltage distribution system (200) by at least a power electronics converter, and/or wherein the smart load (101) is electro-magnetically connectable to the AC voltage distribution system (200), and/or wherein the smart load (101) is electro-magnetically connectable to the AC voltage distribution system (200) by at least an AC to DC power electronics converter.

4. The smart load system (100) according to any one of the preceding claims, wherein the controller (102) is configured to determine the reference power (104) based on the estimated AC frequency (103) and a predetermined reference frequency, for example, based on a difference between the estimated AC frequency (103) and the predetermined reference frequency.

5. The smart load system (100) according to claim 4, wherein the predetermined reference frequency is a nominal AC frequency of the AC voltage distribution system (200).

6. The smart load system (100) according to any one of the preceding claims, wherein the controller (102) is configured to determine the reference power (104) based on a linear relation between power and frequency, and/or based on a linear droop control.

7. The smart load system (100) according to any one of the preceding claims, wherein the smart load (101) is configured to consume power from the AC voltage distribution system (200) if the reference power (104) is positive, and is configured to provide power to the main load (201) if the reference power (104) is negative, and/or wherein the smart load (101) is configured to consume power from the AC voltage distribution system (200) or provide power to the main load (201) equal to the absolute value of the reference power (104).

8. The smart load system (100) according to any one of the preceding claims, wherein the controller (102) is configured to estimate the AC frequency (202) by estimating and/or predicting a real-time frequency of the AC voltage distribution system (200).

9. The smart load system (100) according to any one of the preceding claims, wherein the controller (102) is configured to estimate the AC frequency (202) based on a phase-locked loop algorithm.

10. The smart load system (100) according to any one of the preceding claims, wherein the controller (102) is configured to continuously and/or in real-time adjust the reference power (104) based on the estimated AC frequency (103).

11. The smart load system (100) according to any one of the preceding claims, wherein the smart load (101) is configured to continuously and/or in real-time adjust its power consumption from the AC voltage distribution system (200) or its power provision to the main load (201), based on the reference power (104).

12. The smart load system (100) according to any one of the preceding claims, wherein the controller (102) is configured to provide the reference power (104) to the smart load (101), or wherein the controller (102) is configured to provide a reference DC current, which is based on the reference power (104) and a DC voltage of the smart load (101), to the smart load (101).

13. The smart load system (100) according to any one of the preceding claims, wherein the smart load (101) is a flexible storage device, for example, a battery.

14. The smart load system (100) according to any one of the preceding claims, wherein the AC voltage distribution system (200) is a grid and/or an AC medium voltage distribution system (200).

15. The smart load system (100) according to any one of the preceding claims, wherein the smart load system (100) is configured to not further communicate with the AC voltage distribution system (200) and/or with an operator of the AC voltage distribution system (200), for example a distribution system (200) operator.

16. A method (300) of operating a smart load system (100) for regulating power supplied from an alternating current, AC, voltage distribution system (200) to a main load (201) and a smart load (101), wherein the smart load system (100) comprises a controller (102) and the smart load (101), wherein the smart load (101) is electrically connectable in parallel to the main load (201), and wherein the method comprises: estimating (301), with the controller (102), an AC frequency (202) of the AC voltage distribution system (200), determining (302), with the controller (102), a reference power (104) based on the estimated AC frequency (103), and adjusting (303) the smart load (101) to either consume power from the AC voltage distribution system (200) or to provide power to the main load (201), depending on the reference power (104).

17. A computer program product comprising a program code for controlling a smart load system (100) according to one of the claims 1 to 15, or for performing, when the program code is executed on a computer, a method (300) according to claim 16.