Microgrid synchronization using power offset through a central controller

A high penetration of renewable energy in power systems gives rise to the inverter based distributed generators (DG) in the current microgrid system. With the ability of microgrid to operate in both islanded and grid connected mode, a robust and reliable coordination and synchronization between the DGs and the utility grid is crucial to provide a seamless transition between operation modes. Droop based technique is the most popular method in microgrid system, however, voltage or current transient are present during mode transition. A power offset synchronization technique is proposed to provide a seamless transition between modes of operation. The results show that the mode transition is smooth and seamless without any voltage or current transients.


Introduction
Typical microgrid systems consist of multiple distributed generators (DG), power storage and multiple loads connected in parallel. With renewable energy such as solar are becoming more widespread in power systems, inverter based DGs have grown in popularity and plays an important role in the current generation of microgid system [1]. Since these inverters are electronically commutated, coordination and synchronization are a crucial in the microgrid system to maintain system reliability and operation [2].
Microgrid DGs can operate with or without the utility grid. In grid connected mode, the DGs operate in grid supporting mode where the DGs support the grid by injecting current into the grid to supply the loads [3]- [5]. Both the voltage and frequency are regulated by the utility grid. When the microgrid operates independently without the need of the utility grid, it is known as islanded mode. In islanded mode, the DGs have to established their own voltage and frequency [6]- [10]. This mode is also known as a standalone mode and coordinating between the DGs are important to avoid instabilities in the microgrid system. It is preferred for a microgrid system that can operate in both grid connected and islanded mode for a more robust distribution network. Coordination and synchronization between the DGs to provide seamless transition between these modes is important to reduce the current and voltage fluctuations in the system [11].
Phase locked loop (PLL) are mainly used in microgrid to provide frequency and voltage tracking for the DGs in both grid connected and islanded mode [12]- [15]. As for DGs controls, the most common way for a DG to work in both modes is to use two distinct controllers to cater for each mode [16]- [18]. One controller regulates the voltage and frequency output for it to work in islanded mode, while the other controller regulates the output current injected to the grid. Only one controller is active at a given time. Switching in between these controllers introduces voltage or current spikes during transition due to the different control parameters between the controllers. In [19], a large leader DG,  [20], low bandwidth communication is implemented between the DGs and a secondary controller is used to provide synchronization signals. In [21][22], a P-w and Q-V droop control are implemented with the use of communication through a central controller for synchronization.
Most researchers focus on using the droop technique with a secondary or central controller to provide the synchronization signals required for the DGs. However, the transition between islanded and the grid connected mode causes current and voltage transient during transition. Furthermore, the droop technique encompasses drawbacks such as poor power sharing performance, frequency variation and sensitive to changes in line parameters [23]. In this paper, a synchronization technique based on power offset is introduced to provide seamless transition and synchronization between modes of operation. A microgrid central controller (MGCC) and a secondary controller are used as a synchronization and power sharing platform.

Proposed microgrid system
A microgrid system is proposed as in figure 1, where the system consists of 3 DGs, common load, MGCC and a secondary controller. Each DG is connected to the PCC through line impedance and an interfacing inductance to make the line behave more inductive in nature. Each DG is controlled by the MGCC where the DG reference active power P and reactive power Q are calculated independently by MGCC.  Based from the AC power flow through an impedance as in figure 2 where the line is mostly inductive, it is well known that the DG active power can be adjusted by varying the DG power angle α while the DG reactive power can be adjusted by varying the voltage output of the DG. The overall control for each DG is shown in figure 3 where P and Q generated by the DG are controlled through the P and Q controller. The controller adjusts the voltage amplitude and phase angle of the DG reference generator. The inverter primary control consists of an inner current control [22] with an outer voltage control.

Power offset synchronization
Transitioning from islanded mode to grid connected mode requires the voltage, phase and frequency on the PCC to match the main utility grid before connection to avoid a high power surge and damaging the DGs. For a PQ controlled DG, a synchronization signal to correct for voltage, phase, and frequency is required. This will introduce multiple signals to be transmitted through the low bandwidth communication lines and slows down system response. Since each DG in the microgrid are controlled using P and Q references, changing these power references allows the DG to change its phase angle and voltage output to meet the new set references. Thus, by introducing a power offset in each DG power references, the phase and voltage on the PCC can be adjusted to match the utility grid without affecting the DGs P and Q generation. The addition of power offset are managed by the MGCC where the P and Q power references for each inverter is calculated as = + 1 + 2 +..

(2)
Where n is the number of available DGs in the microgrid system. The P offset and Q offset are regulated through a PI controller which the controller inputs are obtained from the secondary controller. The P and Q power offset are only applied during synchronization with the grid in islanded mode. During normal operation, MGCC sets both the power offsets as zero so that the power sharing scheme between DGs is implemented. The secondary controller calculates the PCC phase and voltage difference with respect to the utility grid by using PLL and Clarke transformation respectively. The voltage and phase difference is calculated as The synchronization with the utility grid is completed when ∆V and ∆θ are approximately zero and thus P offset and Q offset are also converging to zero. At this moment, MGCC will close the utility grid switch (G SW ) and connect the utility grid with the PCC of the microgrid. Equation (1) and (2) are still applied by the MGCC even after successfully connected to the grid to avoid large voltage and current transient during grid connection. MGCC then switches to grid connected mode by utilizing equation (5) and (6) for P and Q reference for each DG. The P offset and Q offset term are removed since it is only used during synchronization and replaced by P grid and Q grid so power can flow from the utility grid to PCC.

Results
Simulation of the proposed microgrid system is constructed in Matlab Simulink as in figure 4. The parameters used in the simulation are specified in table 1. Note that the different values of line impedances and interfacing inductances are used in the simulation. This helps to show the performance of the proposed system for DG with different line impedance. All other parameters including the LC filter, PQ controller and primary controller are the same for all DGs.   figure 5, at t<0.3s all the DGs are initially operating in islanded mode and G SW is open. Using equation (1) and (2) with the P and Q offsets are zero, the DGs operate in equal power sharing mode supplying the common load of 10kW and 2kVar. At t=0.3s, the grid synchronization is started and P and Q power offset are applied. Figure 6 shows the voltage and phase difference between PCC and utility grid. With a negative phase difference, a negative P offset is injected while a positive Q offset is injected to correct for the positive voltage difference. A slight increase in power is observed due to the increase in voltage amplitude on the PCC to match with the grid voltage.   At t=0.5s, G SW is closed hence connecting the utility grid to the PCC. Note that equation (1) and (2) still applies which allow for power sharing only between DGs while P grid and Q grid is zero. From figure  7, it is observed that during when the Gsw is closed, no current transients are present on both the grid and load currents during connection. At t=0.55s, MGCC switches to using equation (5) and (6) where P and Q for the utility grid is included in the power sharing with the DGs. Equal power sharing is At t=0.9s, the transition from grid connected mode to islanded mode occur. During transitioning, the DGs promptly have to increase their P and Q power to cater for the disconnection of the utility grid and MGCC will revert to using equation (1) and (2). All the DGs can quickly reach their respective P and Q reference power and there is no current fluctuation or transients for the load current during transition.

Conclusion
A microgrid grid synchronization method is proposed by using P and Q power offset to adjust the DGs phase angle and voltage amplitude to provide smooth transitions between modes of operation. The proposed method is able to quickly synchronize the voltage and phase with the grid without compromising on power sharing and drooping of the frequency. Furthermore, the method allows for smooth and seamless transitions in between operation modes without any transients. Individual P and Q control for each DG can be maintained for different line impedance in both islanded and grid connected modes and during synchronization.