POWER MANAGEMENT AND CONTROL IN MICROGRIDS

A variety of electric power generators are available on various microgrids such as diesel engines, gas micro turbines, photovoltaic cells, fuel cells, and wind turbines. The power converters through which most of these generators are tied to the network also provide control functionalities required in the microgrids. These converters work as inverters when the power flow required is from the DC side to AC side and as rectifiers for a reverse power flow. The structure of these converters and their control capabilities are very much generic irrespective of the modes of operation. In grid- connected mode, the inverters are expected to control their individual power outputs as commanded by their RE maximum power point trackers (MPPT) and maintain balance within the microgrid through P/Q control.

The salient control responsibilities of such inverters are: (a) AC voltage generation, (b) independent control of active and reactive powers, (c) synchronization to grid, (d) meeting or exceeding the harmonics standards, (e) grid forming, (f) low/ high voltage ride through, (g) minor voltage/frequency disturbance ride through, (h) control under grid fault and distorted grid conditions, and (i) islanding detection and isolation. The control capabilities of the grid-connected power converters are achieved through feedback controllers with a wide range of controlled and controlling parameters.

The performance of such feedback controllers of power converters is assessed based on stability, reference tracking accuracy and dynamic response. The inverter feeding the grid in the grid-connected mode is equivalent to a current source with the control objective as to deliver a predefined amount of current for a given power reference as detailed in Section 2.2.3.1. Therefore, the controller has to be designed as a current regulator and be provided with inputs as reference current and actual current. There is a conflict in this control problem that the controlled parameters are the line currents delivered by the power converters which are AC quantities whereas the legacy control techniques like P, PI, and PID regulators handle only DC quantities. So, the control challenges are: (i) Is it possible to have DC control loops rather than AC? (ii) If AC control loops are to be used then what will be the structure of the AC regulator? (iii) Will this regulator be as effective as the time-tested DC regulators? (iv) If DC regulators only are to be used, then how can be the AC quantities accommodated in DC control loops?

This section describes the possible controller structures and their requirements to answer the aforesaid questions to work satisfactorily in microgrid environment.

Control under Grid-Connected Mode

General Structure of Grid-Connected Converter Control

Voltage source converters are traditionally used in electric motor drive systems applying diverse control techniques ranging from the natural reference frame control to the synchronous reference frame control and employing conventional controllers or even artificial intelligence-based controllers. With minor tweaking, many of the control techniques developed for the motor drive systems can be made suitable for utility/ grid-connected applications as well. However, when operated in the grid-connected applications, the controllers have to overlook transient conditions, different types of disturbances and parameter variations due to grid conditions varying from time to time. At the same time, these have to fulfill the grid connection standards.

The detailed control scheme of the grid-connected converter is shown in Figure 2.24, wherein a DC link is shown feeding a voltage source inverter (VSI) controlled by a multi loop controller. Irrespective of the number of preceding power

FIGURE 2.24 Detailed control scheme of grid-connected converter.

stages ahead of the DC link, the grid-side converter and its control structure will be similar to that of Figure 2.24. Most control structures share some common components and features which are discussed below.

Grid Synchronization

When an inverter has to establish and maintain an asynchronous link with the grid for a controlled power transfer, then the information about the phase, the form and the magnitude of the grid voltage need to be continuously monitored and made available to the controller. The grid voltage vector (V_grid) will be rotating at the rated frequency or within a specific band around the rated frequency and the inverter output voltage vector (V_inv) will be rotating at a different frequency prior to synchronization. That means, the relative phase angle between the grid voltage and the inverter voltage is continuously varying from 0° to 180°. At the instant of closure of the switch connecting these two active sources, the two voltage vectors should be made to rotate at the same frequency so as to avoid large circulating currents triggered due to the instantaneous voltage mismatches. Further, the phase angle (<5) between these two

FIGURE 2.25 Grid-connected converter control and phasor diagram: (a) Equivalent circuit of inverter connected to grid, (b) Power delivery to grid, and (c) Power delivery from grid.

voltage vectors (known as power angle) decides the magnitude and direction of the current (/, or /,) between the two sources, which in turn decides the power transfer as depicted in Figure 2.25. The role of grid synchronizer is imperative in generating an accurate sinusoidal current reference for the grid-connected inverters. Accurate grid information will ensure reliable synchronization and harmonic-free AC current injection. Such information is provided by a grid synchronizer through grid measurements which generate the reference template for the control schemes, whatsoever be the type, to achieve synchronization. Thus, every grid-connected converter system will have a synchronization unit as seen in Figure 2.24 which houses one of the diverse ranges of synchronization techniques discussed earlier.

Power Regulation

Power controllers regulate the active as well as the reactive power delivered by the modulation in all the three modes of operation of microgrids; namely, grid-feeding, grid-forming, and grid-support. The various control functions of grid-connected converters of single and multistage type are depicted in Figure 2.26. Multistage conversion provides a decoupling between RE generator and the grid, which relieves the grid-side converter of the burden of voltage regulation; improving the voltage regulation at the DC link will help reduction of harmonics by appropriate choice of inverter modulation index. At the same time, multistage converters need large energy storage device at its DC link; however, such large DC link can provide virtual inertia as well as fault/low voltage ride through support for microgrids.

For power regulation, the grid converter operation is to be identified first based on the mode of operation of the microgrid, and then the method to obtain the reference power can be identified. The commonly used strategies are instantaneous power balancing, maximum power point tracking (MPPT) of the RE generator, direct power control with droop characteristics, etc. Once the power reference is obtained, then either a voltage or a current control loop can be used to force the actual quantities to follow the respective references. These control methods are presented in Figure 2.27. In current control, the magnitude and the phase of the current injected will be modified as demanded by the reference power to be delivered by the converter; w'hereas in voltage control, the inverter voltage magnitude and its phase with respect to the grid voltage are modified as demanded by the reference power.

The next level of implementation will be the selection of the control structures like natural reference frame, stationary reference frame or synchronous reference frame.

FIGURE 2.26 Control capabilities of converters in grid-connected systems: (a) Multistage converters and (b) single stage converter.

FIGURE 2.27 Control methods of grid-tied converters.

Active Power Control

The active and reactive power (P and Q) transfer between the inverter and the grid, which are decoupled through an inductive filter (L) as seen in Figure 2.25, can be expressed as,

where 0 is the phase angle between Fj„_v and /,.

For small values of S, sin <5 ^ <5 and cos 5=1; then equations 2.5 and 2.6 can be reduced to,

It can be confirmed from equations 2.7 and 2.8, that the active power delivered by the converter depends on the angle between the voltage vectors of inverter and grid, while the reactive power delivered/absorbed depends on the algebraic difference between the voltage vectors’ magnitudes.

Both active and reactive power controls are usually implemented as outer loops and provide appropriate current references to the inner control loops. The inner current loops should account the series voltage drop that occurs across the output filter when the inverter delivers the reference power. This series drop is a variable quantity and a function of the inverter current, which in turn is a function of the power. As the inverter output voltage is the vector sum of the grid voltage and the variable series drop, the control scheme should ensure that the inverter voltage vector is always higher in magnitude and phase with respect to the grid voltage vector to facilitate the power flow from DC link to the AC side.

On the other hand, in multistage RE systems, the converter on the source side is entrusted with the responsibility of extracting maximum power from the source. Typically, MPPT is provided to keep the operating point of the RE generator around MPP. However, only if the power reference of the inverter control is identified rightly at MPP, the tracking operation will be successful. One such method is instantaneous DC link power balance that is depicted in Figure 2.28; it can identify the right inverter current reference for the MPP power.

Basically, the capacitor voltage depends on the energy balance between the power received by the inverter at the DC link and the power delivered by it. The DC link voltage will be constant if and only if these two powers are made equal. If power

FIGURE 2.28 Concept of instantaneous DC link power balance.

received by the DC link is greater than the power delivered by it, then the extra energy will be charging the capacitor, which in turn will elevate its voltage. On the other hand, if the power delivered by the VSI is greater than the power received, then the deficit power is supplied by the DC link capacitor momentarily which results in reduction of its voltage. Thus, the DC link voltage serves as an index of power balance for the grid tied VSI, and by regulating it the maximum power extracted from the RE source can be pushed into the grid.

The DC link capacitor can be designed to operate at any desired voltage ripple and that can provide a hold-up time to facilitate the multistage grid-connected inverter system to search for the MPR This holdup time can be the permissible MPPT delay in addition to the control and process time lag of the preceding DC-DC converter in a typical multistage grid-connected RE system. During the holdup time, the inverter keeps providing the regulated output current even during a momentary power shortage owing to a fall in irradiance in solar PV system, a fall in wind velocity in WTG system, etc.

The DC link capacitor (C_dc._link) can be designed with the assumption that the inverter has no losses; its value can be expressed as a function of the desired hold up time as,

I

where V'_dcmin is the minimum voltage across C_dc__Hnk as chosen by the allowed ripple in the DC link voltage, typically in the range of 75%-90% of its nominal voltage, Vdc,nominal and, Prated >^s the rated power delivered to grid. The product of P_rated and holdup time in Equation 2.9 decides the size of ^Cdc-li„k- The transient energy support (AE) that the capacitor should receive or deliver to restore V_dc as its nominal value is then expressed as,

Generally, the volume of capacitor is proportional to its voltage rating while its energy storage capacity is proportional to the square of the voltage rating. Large holdup time results in large size of C_dc._link, at the same time, gives adequate time for MPP search; also, the converters can be designed to operate with lower switching frequency. Conversely, small holdup time though reduces the size of C_dc._link demands higher switching frequency and superior transient response of MPPT.

Reactive Power Control

Inverters can be made to deliver currents at any phase angle with respect to the grid voltage by generating an inverter voltage at appropriate amplitude and phase. This feature that realizes VSI to feed power at any desired power factor helps the reactive power control. Reactive power control of RE inverters in microgrids can dynamically compensate the reactive power where it is needed the most, i.e., close to the loads and thus perform voltage control. Such distributed voltage controls significantly improve the voltage profile within the microgrid; it can further be extended as an ancillary service to the main grid which outperforms the centralized voltage control generally carried out in conventional power system.

Reactive power control is performed by the inverter according to one or more of the following requirements: (i) The reactive power support required from the converter, (ii) the power factor at which the current need to be delivered, and (iii) voltage control on the AC grid. Inverters can be made to consume or deliver reactive power based on the system needs. In microgrids often both consumption as well as delivery of reactive power will be required when operated in grid-support and grid-forming modes of operations. The control of grid tied inverter in any of the methods specified in Figure 2.27 will include a reactive power loop with a reference which can be varied in accordance with the microgrid system demand. This will ensure the required reactive power be provided by the inverter.

Reactive power control is considered as an advanced inverter function as of today. However, these advanced functions will be adopted in the near future as they can provide ride-through capability against minor voltage fluctuations, low/high voltage and grid faults which eliminates avoidable disconnects. An advanced inverter is a standard inverter that has been enabled witli advanced features with no significant increase in cost. However, the extent to which an inverter can provide reactive power support/delivery depends on the percentage real power loading of the inverter.

Reactive power capacity curve of the inverter is presented in Figure 2.29. The inverter limit circle indicates the kVA rating of the RE inverter; any vector within this circle indicates the possible active and reactive power combinations without violating the capacity constraint of the inverter. In Figure 2.29a, 5, is the kVA delivered by the inverter for an available RE active power of P_t and the corresponding reactive power delivered is Q_v The possible maximum reactive power delivery at this inverter loading is £2_lmax, and a Q demand beyond 2,_Max will ask for a reduction in active power to R_derate. Figure 2.29b corresponds to another condition of RE power transfer, where the available active power is very close to the rated capacity of the inverter. It can be seen that the maximum possible reactive power support is reduced to Q_2Max from Q_IMax of Figure 2.29a. It therefore infers that when the inverter is required to provide large reactive power support, an oversized inverter is necessary to avoid real power curtailment as well as associated revenue loss.

Conversely, when the RE generator is on part load, then the active power delivered will be below the inverter kVA rating (assuming that both generator and inverter have equal power ratings). Under such operating conditions, the remaining inverter capacity can be utilized for supplying reactive power and this will contribute to improving the RE system economics. Therefore, this control feature of inverter can provide reactive power at any time of the day, regardless if the sun shines or the wind blows.