Control and Power Hardware-in-the-Loop Testing of Grid-Forming Voltage Source Converters (CPHILT-GFVSC)

The integration of renewable energy resources (RES) in power systems is displacing conven tional generation assets based on synchronous generators. RES use voltage source convert ers (VSCs) as the main interface between the primary energy source and the grid. Therefore, these power electronic devices should guarantee in future converter-dominated power systems, robust and secure operation. For this purpose, the provision of ancillary services is of partic ular relevance, especially concerning frequency regulation. Recently, the concept of a virtual synchronous generator (VSG) with grid-forming functionality has emerged to operate in three phase systems. However, the extension of this concept to single-phase devices, especially those operating within low-voltage networks, has not been widely explored. Taking into account that a relevant part of the new RES generation is connected to distribution systems, it might be necessary to endow a certain number of single-phase RES with grid-forming capability. This work proposes the use of two reference frames, rotating and stationary, to implement the con trol levels of a single-phase network forming VSC. These levels are made up of inner and outer control loops that allow the grid-forming VSC to adapt its voltage at the POI according to the load of the system.

The rotating reference frame for single-phase systems is generated through virtual axes dq. These are obtained by means of a filter to the voltages and currents of the system. This makes it possible to implement classic control strategies based on PI controls. While the stationary ref erence frame is based on the currents and voltages of the single-phase system, implementing PR controllers in the inner control loops to track sinusoidal references.

The evaluation and transfer of code between the User Group and the Host Group has been carried out using the following methodology:

• Simulations carried out by the User Group with the aim of validating the control of single phase GF VSC in the proposed reference frameworks.
• Transfer and adaptation of the User Group simulations to the Host Group simulation en vironment. The host group has a detailed simulation of the Triphase converter, device used in the experimental test-bed, which has been provided by Triphase. This simulation allows to integrate all the control blocks identically to the files used for the experimental validation.
• Comparison of the dynamic and steady state response of the single phase GF VSC in both reference frames between the User Group simulations and the simulation environ ment used by the Host Group. To carry out this comparison process, an analysis of the performance of each level is carried out individually. This allows evaluating each level of control separately as well as its behavior.
• Experimental evaluation of the single-phase GF VSC for both frames of reference. The control is progressively validated by increasing voltage levels and currents in the proto type. In this way, control performance for both reference frames is safely and progres sively evaluated. Similar to the previous stage, each level of control is validated individu ally.

In addition, it was discussed how to implement the evaluation of the single-phase GF VSC in isolation mode for a P-HIL environment. This discussion revolved mainly around how to emulate a passive load through an amplifier. Several strategies were established, which are being evaluated in a simulated environment.

The simulation results of the User Group show an adequate dynamic response and good power quality for each of the control levels in both reference frames. However, the dynamic response of the stationary frame is faster than the rotating axes. This is due to the delay introduced into the control by the filter that generates the rotating virtual reference frame in dq axes. Identical responses are obtained with the simulation results of the Host Group once the code transfer was performed. The dynamic and steady-state response of each control level in the experi mental results of the single-phase GF VSC presents a very similar behavior to those obtained in both simulations. The only appreciable difference is relative to the minimum value of the frequency reached with an increase in load. This value is reached in simulation for the rotat ing frame, while in the experimental results it is reached for the stationary frame. Both groups believe that the mismatch is due to a difference in the gains of some control stage. However, further research on this question remains pending.

It can be concluded that a work methodology has been established for the transfer of research from the User Group to the experimental validation in the laboratory infrastructure of the Host Group. This methodology has allowed to successfully validate the single-phase GF VSC control in two reference frames. In addition, there have been discussions between both groups about the best way to implement a load in P-HIL for isolated systems. These discussions have also led to how to validate through P-HIL the control of a GF VSC in a grid-connected mode. As a consequence, both groups have agreed to participate in a new ERIGrid call for the year 2023 with the aim of proposing solutions to these open questions during the execution of this project.