Grid Simulation and Power Hardware-in-the-Loop

This project evaluated battery energy storage system performance and validated holistic photovoltaics (PV)-battery energy storage inverter control applied across an electric distribution system. Individual inverter efficiency, battery utilization, and control response performance while operating in four key operational modes were evaluated. Then the system-wide impacts of controlling hundreds of distributed PV/storage units in four key operational modes were examined using PHIL. Detailed, real-time models of a complete electric distribution network under different loading scenarios and with varying penetrations of PV-only and PV-energy storage units were combined with a physical PV/battery inverter in a single closed-loop simulation.

The project demonstrated that coordinated control of many distributed PV-battery inverter units can provide valuable grid services, including voltage smoothing, reduced tap change operations of utility voltage regulators (and thus reduced operations and maintenance costs), and reduced peak distribution system power requirements.

NREL researchers developed real-time models of two Oahu distribution feeders and the Oahu bulk power system and used PHIL to investigate the effects of advanced solar PV inverters on Hawaii's grid. A variety of PV string inverters, microinverters, and three-phase inverters were tested using the PHIL system to investigate the effects of advanced inverter functions in an environment that emulates the actual voltage and frequency dynamics of the Oahu power system. To capture the effect of very large numbers of distributed inverters on the grid, additional simulated PV systems are included in the PHIL model based on the actual PV systems present on Oahu. Hawaiian Electric is in the process of updating its PV interconnection requirements to incorporate the inverter-based grid-support functions tested in this project.

NREL researchers also developed a PHIL system for testing inverter anti-islanding controls (which prevent inverters from unintentionally energizing a grid during an outage) in multiple-inverter scenarios. A suite of real-time PHIL models was developed to represent different interconnecting circuits between the inverters and different configurations of load within the island. The loads were tuned to closely match the real and reactive power of the inverters and to resonate at 60 Hz, intentionally creating a difficult situation for the inverters to detect. The inverters were tested for anti-islanding using a variety of interconnecting circuits designed to represent the range of possible configurations seen on the grid. In addition, anti-islanding tests were run with the inverters programmed for a variety of grid support functions.

Although the study did not investigate every possible island scenario, the scenarios that were tested experienced no islands longer than 700 milliseconds—well within the 2-second limit imposed by IEEE Standard 1547.

Publications
Experimental Evaluation of PV Inverter Anti-Islanding with Grid Support Functions in Multi-Inverter Island Scenarios, NREL Technical Report (2016)

In this study, PHIL-based testing was used to evaluate autonomous volt-VAR operations of multiple smart PV inverters connected to a real-time model of the IEEE 13-node test feeder. The results show that, depending on volt-VAR control parameters and grid parameters, interaction between inverters and between the inverter and the grid is possible—in some extreme cases with very high volt-VAR control slopes, fast response times, and large volt-VAR control response delays.

Publications
Power Hardware-in-the-Loop Testing of Multiple Photovoltaic Inverters' Volt-VAR Control with Real-Time Grid Model, Innovative Smart Grid Technologies Conference (2016)

NREL; University of Minnesota; and University of California, Santa Barba, are investigating the stability of low-inertia grids with inverters and traditional synchronous machine generators. Researchers developed scalable synchronous machine and inverter models to investigate their interactions as a function of relative power ratings. Using PHIL, machine models with frequency and voltage dynamics were connected to real-world inverters. PHIL also allows researchers to test a variety of cyber-physical scenarios in which interconnected machines and inverters can be modeled and combined with physical hardware in the laboratory. This work will facilitate the integration of large amounts of renewables into the grid by predicting integration issues and providing novel solutions.

NREL is a partner on two ARPA-E NODES projects that are developing unique approaches to distributed control of distributed energy resources and flexible loads to enable renewables penetration of at least 50%. In the NODES program, project technologies must be demonstrated and validated on a large scale using hardware-in-the-loop techniques. NREL is leveraging its expertise in power system modeling and PHIL simulation to evaluate the novel control solutions developed under these two projects.