ETD

• current limiters
• inverters
• power system dynamic

The increasing integration of inverter-based resources into modern power grids is a direct response to the necessity of decarbonization and the reduction of greenhouse gas emissions traditionally produced by fossil-fuel-based power plants. As a consequence, future grids are expected to have lower synchronous generation and higher penetration of sources like solar and wind, which are connected to the grid through power electronic devices. These electronic devices exhibit low current injection capability, as they can be easily damaged by the heat dissipated via the Joule effect. This issue can impact the dynamics of power systems, especially when the inverter is required to inject high currents in order to sustain the voltage or the frequency. This thesis focuses on implementing a current limitation logic that can provide both device protection and system support. Moreover, this current limiter is compared with limiters already implemented in the literature. The inverters studied are the grid-following type. The limitation logic is directly applied to the current reference entering the current controller of the inverter, as it can be considered a balanced solution among the limitations that act on the voltage at the output impedance of the inverter.

Various scenarios are simulated using a specific Python-based power system tool called Dome. To build the proposed limiter and the limiters from the literature, specific Python modules were implemented and interfaced with an existing grid-following inverter module. All of these limitation logics have since been added to the official version of the software.

The first step involved building a 9-bus power system and replacing one of the three synchronous generators with an inverter-based resource, such as a solar or wind plant.

To trigger the current limitation, two fault cases were simulated: one less severe and one more severe. This allowed for the study of the limiters' behavior under two different voltage dips. The relevant quantities for all limiters are shown both during and after the perturbation. Furthermore, all limiters are evaluated based on voltage and frequency metrics. The same faults were then applied to a lower-inertia system, confirming the general results and showing how the limitation impacts the stability of a low-inertia system.

Following the general comparison, the thesis focuses on the proposed limiter; specifically, a sensitivity study is addressed where the key parameter of the limiter is varied to determine the optimal value. The results indicate that values greater than one are most effective.

In conclusion, since all the faults simulated up to this point caused the system frequency to increase, it was fundamental to study the proposed limiter's behavior during an under-frequency event by simulating the loss of one generator in the system. The results show that in this case, the optimal parameter values are the opposite of the fault scenarios, as key parameter values near zero are preferred.

However, the differences between different parameter values for frequency and voltage metrics in the under-frequency event are small. In conclusion, since a parameter greater than one provides better voltage support under both under- and over-frequency conditions, and is also superior for frequency metrics in fault scenarios—while the differences in frequency metrics during a loss-of-generator event are negligible—a parameter value greater than one appears to be the optimal choice.