Experimental Multi-site Testbed for Advanced Control and Optimization of Hybrid Energy Systems

TL;DR

The paper tackles the challenge of validating multi-resource Hybrid Energy Systems (HES) under renewable variability by introducing a dual-site experimental testbed at the University of Vermont, combining the Accelerated Testing Laboratory (ATL) on campus with the Hybrid Solar Test Center (HSTC) off-campus. It leverages hardware-in-the-loop and CHIL approaches via the OPAL-RT real-time simulator to enable plug-and-play integration of PV, batteries, an electrolyzer, and grid-tied inverters, driven by real weather and PV data. A solar-smoothing CHIL case demonstrates coordinated PV-battery operation, achieving a ramp-rate reduction from the raw PV ramp of 56%/min to 3.75%/min while maintaining safe state of charge and meeting grid limits. Overall, the platform provides a versatile, real-time, multi-timescale validation environment that supports rapid prototyping, model validation, and co-optimization of HES components for diverse grid-services with real hardware and validated models.

Abstract

This paper presents a hybrid energy system (HES) experimental testbed developed at the University of Vermont to support prototyping and validation of advanced control and optimization strategies for grid services. The platform integrates hardware-in-the-loop (HIL) simulation with a reconfigurable set of kilowatt-scale assets, including solar photovoltaic (PV), battery storage, an electrolyzer as a controllable load, and grid-tied inverters. A unified monitoring and communication architecture supports real-time data acquisition, model validation, and control implementation. The testbed's capabilities are demonstrated through a controller hardware-in-the-loop (CHIL) experiment in which a battery system participates in PV power smoothing.

Figures (5)

• Figure 1: Operational HES across the U.S. Berkeley2025.
• Figure 2: HES testbed showing ATL components, including inverters interfaced with the RGS on the AC side, battery modules powered through the DC supply, an electrolyzer as a controllable load, and OPAL-RT integration for HIL testing; and HSTC components including the PV array and planned battery system.
• Figure 3: Data and communication architecture between HSTC and ATL.
• Figure 4: Experimental setup for solar smoothing CHIL validation showing two series-connected Enphase batteries with a BMS, bidirectional DC power supply, TI control card, and OPAL-RT real-time simulator.
• Figure 5: Two-hour snapshot of HIL experimental PV power smoothing using historical 5-second HSTC data and ATL battery setup. (a) raw and smoothed PV power, here, $P_{\text{PV}}$ denotes the raw PV power, $\hat{P}_{\text{PV}}^{e}$ and $\hat{P}_{\text{PV}}^{s}$ denote the smoothed PV power from experiment and simulation, respectively, (b) dispatched battery power, (c) battery SOC, and (d) distribution of ramp rates.