IEEE Power & Energy Magazine - Grid Edge 2023 - 95

Real-World Applications
We have now started to move out of the laboratory to demonstrate
the deployment of AEGs in the real world. The team
has been working with Holy Cross Energy (HCE), a utility
cooperative near Aspen, Colorado, to deploy the AEG technology
in a group of smart homes in Basalt, Colorado. The smart
homes in Basalt Vista (Figure 10) are a pilot for an altogether
new approach to the grid. These homes optimize energy for
residents and their neighbors, but the principles behind Basalt
Vista go much further. Within homes, each new connected
device or energy resource, such as a residential battery, water
heater, or solar PV system, can be controlled for unprecedented
energy efficiency. At a larger scale, entire communities could
rapidly share power, creating reliable energy for everyone.
HCE had been searching for a solution to managing new
devices on its system. This has included a mix of customer
energy technologies and bulk generation resources since the
decreasing costs of connected customer-owned devices have
made these systems much more affordable. HCE's grid has
seen 10-15 rooftop solar installations per week, and it has
been increasing its solar base for years, planning for a 150-MW
summer peaking system through 2030.
In another real-world experiment, we used our real-time
optimization algorithms to coordinate assets in the Stone
Edge Farm Estate Vineyards & Winery in Sonoma Valley,
California (Figure 11). The winery is a microgrid with tens
of DERs, including PV systems, batteries, a hydrogen electrolyzer,
a gas turbine, and controllable loads. The experiment,
conducted in collaboration with the Massachusetts
Institute of Technology-born startup company Heila Technologies,
showed how our approach can help self-optimize a
cell within the future AEG vision, achieving voltage regulation
and allowing the microgrid to become a virtual power
plant that can provide services to the distribution system.
Conclusion
AEGs of the future will need to control and optimize millions
of controllable devices in real time. A traditional central
optimization approach to this problem is infeasible because
of the computational cost. Therefore, robust, scalable, and
predictive hierarchical and distributed control algorithms
with provable convergence are needed to optimize the grid
in real time. NREL has developed these scalable algorithms
to enable the proliferation of DERs on a massive scale.
A fundamental underpinning of AEGs is the ability to accurately
model the cellular building blocks and their interactions
with the rest of the systems so that control, optimization, and
forecasting methods might be applied in operation. NREL is
also building computational tools that can cosimulate multiple
technologies on the grid to design and evaluate these scalable,
distributed control and optimization algorithms using high-performance
computing. In addition, NREL has taken the preliminary
steps of demonstrating these algorithms in real time for
real-world devices in the laboratory and now in smart homes.
Additional work will be needed in controls, optimization, data
november/december 2020
figure 9. The inverters under test in the ARPA-E NODES
experiment. (Source: NREL; used with permission.)
analytics, complex systems, and cybersecurity to implement the
AEG across the entire U.S. grid.
Building on the distributed optimization techniques that
have been developed, additional thought needs to be given to the
design of future market mechanisms to systematically account
for payment/rewards to exchange energy and the provision of
ancillary services among autonomous cells and devices. The
ideas of transactive energy will need to be considerably
figure 10. The smart homes in Basalt, Colorado. (Source:
NREL; used with permission.)
ieee power & energy magazine
95
IEEE Power & Energy Magazine - Grid Edge 2023

Table of Contents for the Digital Edition of IEEE Power & Energy Magazine - Grid Edge 2023
