The declining cost of solar Photovoltaics (PV) generation is driving its worldwide deployment. As conventional generation with large rotating masses is being replaced by renewable energy such as PV, the power system’s inertia will be affected. As a result, the system’s frequency may vary more dramatically in the case of a disturbance, and the frequency nadir may be low enough to trigger protection relays such as under-frequency load shedding. The existing frequency-watt function mandated in power inverters cannot provide grid frequency support in a loss-of-generation event, as PV plants usually do not have power reserves. In this article, a novel adaptive PV frequency control strategy is proposed to reserve the minimum power required for grid frequency support. A machine learning model is trained to predict system frequency response under varying system conditions, and an adaptive allocation of PV headroom reserves is made based on the machine learning model as well as real-time system conditions including inertia. Case studies show the proposed control method meets the frequency nadir requirements using minimal power reserves compared to a fixed headroom control approach.

Load unbalance in electric distribution systems is unavoidable. Unbalanced load currents will lead to negative sequence (NS) voltages that may damage electric equipment. Unbalanced power flow analysis is a common tool to detect and mitigate NS voltage issues and requires accurate models of grid components. While traditional source models are available, grid-forming (GFM) inverter models are not well developed. GFM inverters implement various control strategies, which affect their power flow models. In this article, a novel GFM inverter model considering control effects is proposed. We show that for some control methods, unbalanced system loading will lead to unbalanced terminal voltages of the GFM inverters, which are modeled through an equivalent negative sequence impedance. The proposed models are initially validated using a simple test circuit. Then, they are applied in the power flow analysis on the IEEE 13-bus and 34-bus systems to demonstrate the accuracy improvement over the state-of-the-art. Using time-domain simulations as benchmarks, we show that the proposed models reduce the calculation error of negative sequence voltages by at least 25% in unbalanced distribution systems.

In contrast with conventional microgrids (MGs) with fixed boundaries, a smart and flexible MG with dynamic boundary is introduced in this paper. Such a MG can dynamically change its boundary by picking up or shedding load sections of a distribution feeder depending on its available power, leading to more flexible operation, better utilization of renewables, smaller size of energy storage system, higher reliability, and lower cost. To achieve a flexible MG, the main challenges in MG design are addressed, including recloser placement, MG asset sizing considering resilience, system grounding design, and protection system design. Meanwhile, a hierarchical structure is employed to design and implement the MG controller. On top of the functions defined in IEEE 2030.7-2018, a few new functions, e.g., online topology identification and PQ balance, are added, while the planned/unplanned islanding and reconnection functions are enhanced. The controller is implemented on a CompactRIO, a general-purpose hardware platform provided by National Instruments (NI), and tested on a controller hardware-in-the-loop setup based on an OPAL-RT real-time simulator and a reconfigurable power electronic converter-based hardware testbed. The test results have validated the performance of the developed controllers. Such a flexible MG and its controller have been deployed at a municipal utility, and part of the controller’s functions have been tested on-site.

This paper discusses the design, development, and testing of a flexible combined heat and power (F-CHP) system. It includes two parts: the power conditioning system (PCS) converter, which connects the system to the grid and is the key component to provide grid support services, and the F-CHP central controller, which accommodates the control of local sources, local loads, and the PCS converter. The functions and structures of the central controller are discussed. The central controller’s system-level performances are verified with hardware-in-the-loop (HIL) tests, and the mode transition and grid transient performances are validated with power electronic converter-based hardware testbed (HTB) tests. The PCS converter is designed with 10 kV SiC MOSFETs. The grid requirements are considered in the 13.8 kV/1 MW PCS converter design, and the impact of different requirements on the converter design is identified. Also, converter scalability is studied with the paralleling operation. Two 13.8 kV/100 kW converter prototypes are developed to verify the converter design and scalability. The second prototype is an improved design of the first one; the volume is reduced by 49% and the efficiency at the rated output power is increased from 96.4% to 98.4%. The test results of the converter performances, grid support functions, and parallel operation are also discussed.

The Remedial Action Scheme (RAS) is designed to take corrective actions after detecting predetermined conditions to maintain system transient stability in large interconnected power grids. However, since RAS is usually designed based on a few selected typical operating conditions, it is not optimal in operating conditions that are not considered in the offline design, especially under frequently and dramatically varying operating conditions due to the increasing integration of intermittent renewables. The deep learning-based RAS is proposed to enhance the adaptivity of RAS to varying operating conditions. During the training, a customized loss function is developed to penalize the negative loss and suggest corrective actions with a security margin to avoid triggering under-frequency and over-frequency relays. Simulation results of the reduced United States Western Interconnection system model demonstrate that the proposed deep learning–based RAS can provide optimal corrective actions for unseen operating conditions while maintaining a sufficient security margin.