IECON 2017
Paper Title: A Reactive Power Compensation Method for a Smart Grid Connected Inverter Using a Residential PV System.
Authors: Claudio Valenzuela, Pablo Vela, José Espinoza
Full paper HERE.
Abstract
This work explores reactive power compensation on a small photovoltaic generation at residential installation to enhance reliability on a Smart Grid. In order to accomplish this goal, the modulator values of the inverter are determined to obtain the allowable range of reactive power compensation in a photovoltaic system connected to the grid trough out a single-phase voltage source inverter. Mathematical validation in addition to simulations and experimental results are included to validate the proposed approach. Although the method is used in small generation systems, it can be extended to large power systems.
Research
Topology
The topology to be analyzed in this work can be summarized to a single-phase circuit as shown in the figure below which corresponds to a monophasic PV system synchronized to the grid through an inverter that injects active power or compensate reactive power. This topology is replicated three times in the grid, one in each phase. It is this feature that allows to create a Smart grid of several compensators (residential PV systems) interconnected. Each residence will have this system and finally each neighborhood will be connected to the distribution grid to increase the reliability and availability of the system against disturbances.
Theoretical Results
In this section the colors of the graphics are just for visualization and don’t represent a variable of interest.
When solving the system of equations, the condition imposed on the system to make the power injecting range valid in the grid is that there is no over-modulation (17).
The active current injected varies according to the DC-Link voltage and the reactive current being injected is as shown in the figure
Fig. 4 and Fig 5 show the values that the modulators must have as a function of the DC-Link voltage and the reactive current that is required. Consequently, with the values of DC-Link voltage and reactive current Isq known, the active current Isd that is being injected into the grid can be determinated.
On the other hand, when calculating the module of the decomposition of the modulators through (4) the valid operating range is finally obtained for the proposed system.
Fig 6 and Fig 7 show the operating range where what is under the flat grid is the valid range and the points that are over the surface produce overmodulation.
As shown in Fig 8, by taking the values that do not produce overmodulation, the valid range of the power in p.u. is obtained for the proposed system. Finally, by taking the values that do not produce overmodulation, the valid range of the power in p.u. is obtained for the proposed system that can be injected as a function of the DC-Link voltage and the reactive current.
In Fig 9 it is possible to observe the minimum power required according to the desired reactive current variation Isq, both inductive and capacitive. On the other hand, the system does not operate at 1.0 p.u., this is due to the losses by switching and those associated with the DC-Link, these losses are represented by the DC resistance Rdc.
It is important to emphasize that the greater the voltage range in the DC-Link is, the greater the range in which the reactive power and the active power injected can be varied. For example, if the maximum DC voltage range of the system drops to 500V the new range is shown in Fig 10, where the reduction of the valid operation range is clearly seen in contrast to Figure 6.
Simulations
The validation of the previous model was verified on PSIM 9.0. The results obtained in Fig 13 and Fig 14 are congruent with the expected ones and also it is evident the lag between the voltage and the current injected to the grid, Fig 13.
Experimental Results
In this setup, a DSP TMS320F28335 is used for generating the PWM pulses of the inverter (blue box), the grid is implemented using a variac connected to a transformer for a galvanic isolation (green box) and an inductor with a rheostat to emulate the distribution line (red box). The complete setup is shown in Fig. 15.
The results are shown in Fig. 16, Fig. 17 and Fig. 18. In Fig 16 the current and voltage at the output of the inverter are in phase i.e. only active power is injected to the grid but in Fig. 17
and Fig. 18 the current and voltage are out of phase i.e. reactive power is being injected to the grid. This confirms the operation as a capacitive reactance or an inductive reactance.
Conclusions
The modeling of the system from the proposed mathematical arrangements allows obtaining the range of active and reactive power to be injected for systems based on solar panels.
The amount of reactive power that can be injected is finite and is limited by the level of voltage of the DC-Link in the PV array. In fact, the greater the voltage in the DC-Link is, the greater is the range of reactive power that can be injected into the network without causing overmodulation.
For the parameters used in the mathematical model in this system when requesting the maximum of reactive current the active power injected to the grid decreases 0.013 p.u. with respect to injecting active power with unity power factor.
This new feature implemented in the photovoltaic generation, allows to promote the use of solar panels not only in large generators but also in houses due to the rapid development of Smart Grids. Finally, it improves the robustness of the interconnected system without altering the electrical generation system current.
Experimental verifications demonstrated the feasibility of the proposed approach.
