Robust Dynamic Resilient Power Grids Enabled By Modern Control Framework

Abstract

The inexorable effects of climate change are disrupting weather patterns with disastrous consequences such as hurricanes, wildfires, ice storms, flooding, etc. The growing frequency, duration, and intensity of such catastrophes put stress on the traditional aging electrical power infrastructure and cause prolonged power outages worldwide. In many well-publicized reports of remediation plans for power outages, microgrids have been envisioned as a critical resource for supplying facilities with resilient and reliable power during such emergencies and outages. The main reason for fostering microgrids is the benefits it provides that include (1) the opportunity to deploy more zero-carbon emission electricity sources, thereby reducing greenhouse gas emissions, (2) enhanced grid resiliency to extreme weather attacks by flexibly engaging/disengaging to/from the main grid, (3) utilization of on-site distributed energy resources (DERs) for reliable and uninterrupted power supply to the local consumers. Worldwide roll-out projects and implementations indicate that from the stage of a niche novelty, today microgrid is emerging as a mainstream solution.New challenges have emerged from this transformation in the way of making and consuming power using microgrids. Increasing installation of renewable DERs via inverter-based-resources (IBRs) in microgrids has resulted in a significant amount of decline in the inertia of the system by the displacement of high inertia-based synchronous machines. Penetration of intermittent and unpredictable renewable to such low inertia-based microgrid systems lead to undesirable fluctuations in the frequency and the voltage. Besides, there is a renewed emphasis on today’s microgrids that provide operational flexibility leveraged by the control schemes for IBRs which introduces the plug-n-play capability of IBRs. As a result, today’s microgrids experience a significant amount of uncertainties from the generation side due to intermittent renewable integration, from the electrical network side due to plug-n-play of IBRs, and from the load side due to constantly varying demand. It deteriorates the robustness in performance of the microgrid by influencing the control of IBRs, leading to instability in the worst cases. Another topical area of development for microgrid is the enhancement of its resurrective capability to restore power rapidly and seamlessly after outages. The improvement of the flexibility in the way of connecting/disconnecting to/from the main grid or other local microgrids, when needed, is the crucial remedial action that makes a microgrid dynamic and resilient. The major challenge here is to develop a capability for the IBRs that enables a smooth (controlled transients and distortion in voltage waveform) and seamless (without any electrical interruption) transition of a microgrid from on-grid (grid-tied) to off-grid (islanded) conditions. Moreover, when the main grid returns to normalcy, the microgrid needs to be reconnected/resynchronized smoothly and seamlessly. Large uncontrolled transients in the output voltage and/or current of the IBRs are usually observed during these jumps in the mode of operations of microgrids, leading to power collapse in the worst cases. These transients are the major impediments that restrict today’s microgrid from a smooth and seamless transition. In this thesis, Chapter 2 contributes towards the development of a robust microgrid against various inevitable uncertainties. A µ-synthesis-based robust controllers for the IBRs in a microgrid are developed by quantifying some major sources of uncertainties using a generalized control framework. It includes the uncertainties caused by the equivalent impedance parameters at the point-of-connection of the IBRs and the equivalent loading parameters of the IBRs. The efficacy and viability of the robust controllers for IBRs are evaluated using an experimental case study involving an emulated 1-ϕ, 240V, 10kVA, 14-bus residential microgrid with two physical 2kVA IBRs. The developed robust controller for IBRs shows robust performance and dynamic response against large uncertainties of a microgrid. Chapter 3 and Chapter 4 of this thesis contribute towards developing a dynamic microgrid that can engage/disengage to/from the main grid or the other microgrids very flexibly. In Chapter 3, a mode-dependent integrator-based droop controller is developed for the IBRs that facilitates a smooth and seamless on-grid to off-grid transition capability in a microgrid. In Chapter 4, a dynamic compensation-based active synchronization method is developed using droop controllers of the IBRs that enables a smooth and seamless off-grid to on-grid transition capability in a microgrid. Methods are evaluated using an emulated experimental case study involving a 3-ϕ, 480V, 500kVA, 55-bus microgrid with six IBRs and using a laboratory-scale experimental case study involving a 3-ϕ, 480-V, 270-kVA microgrid with two 3-ϕ, 480V, 125kVA IBRs. The developed controllers in the IBRs enable relatively fast and smooth connection/disconnection capability of a microgrid with less transients. Chapter 5 introduces a large-scale demonstration only to showcase that the seamless transition capabilities of Chapter 3 is the key capability for a resilient microgrid to support disaster response efforts and facilitate recovery. The demonstration is conducted on a laboratory-scale 3-ϕ, 480V, 750kVA microgrid with critical loads of 550kW, 215kVAr, four 3-ϕ, and 125 kVA, 4-quadrant IBRs and, one 3-ϕ, 135kW diesel genset. This thesis further investigates another emerging domain that deals with advanced and modern validation techniques for any engineering innovations. Any technological innovation is required to be validated either via offline simulation environments or via experimental demonstration before being employed in the real world. For instance, a newly developed controller for IBRs requires to be validated by conducting diverse tests before prototyping and real-world implementation. In this context, power hardware-in-the-loop (PHIL)-based digital real-time simulation (DRTS) is considered to be an attractive platform for conducting diverse tests on a laboratory scale, providing results that closely resemble real-world scenarios, all while avoiding the high costs and risks associated with experimentation on a full-scale power system. The hardware system, also called hardware-under-test (HUT), is usually interfaced with a software system that is simulated in real-time inside a real time simulator. The interfacing compensator/controller is the backbone of the PHIL-based DRTS that manages the closed-loop interaction between the hardware system (e.g., single/group of IBRs) and the software system (e.g., realistic microgrid model). Considerable challenges emerge in the process of synthesizing the PHIL interfacing controller due to the inherent uncertainties present in both the software and the hardware system. Interfacing controllers, designed without addressing these uncertainties a priori, often suffer from a lack of robustness in the performance. This leads to inaccurate results from the testing of the HUT using the PHIL-based DRTS that don’t resemble with the expected results of the same HUT in real-world operations. Therefore, the interface between the software and the hardware system must be carefully synthesized to ensure robust and accurate PHIL-based DRTS. Chapter 6 of this thesis makes an important contribution in addressing these challenges by synthesizing a µ-synthesized robust PHIL interface controller between the hardware and software components of a PHIL-based DRTS while employing a modern control perspective for managing inherent uncertainties. It provides a methodology that quantifies model uncertainty especially caused by the point-of-connection in the emulated power network of the software system and the model uncertainty stemming from the varying nature of the HUT. The designed interfacing controller is implemented to interface a software system based on an 110V, 1MW, 225-bus residential sub-network of the University of Minnesota and suburb Minneapolis and a hardware system consisting of two 1-ϕ 1.67kVA IBRs, one linear and nonlinear load of 1.8 kVA each. It is observed that the robust PHIL interface shows improved accuracy, and tracking performance under varying conditions of software and hardware system.