EnergyHub is first-to-market with state-of-the art dynamic load shaping VPP technology.

Energy demand is rising as data centers proliferate, domestic manufacturing grows, and the U.S. electrifies its vehicles, homes, and buildings. Peak demand, in particular, is expected to rise by 38 GW by 2028 according to a report from the Clean Grid Initiative. The U.S. power sector is not prepared for this level of load growth, which could require billions of dollars of investment per GW of additional load.

While many utilities rely on VPPs during times of peak demand, they also have the potential to offer around-the-clock flexibility comparable to natural gas generators or grid-scale batteries. For this to happen, however, VPPs need to deliver precise, sustained load shape adjustments that are similar to a conventional power plant and designed around customer retention.

EnergyHub’s new dynamic load shaping functionality is making this future a reality, allowing utilities to automatically coordinate and dispatch distributed energy resources (DERs) — like batteries and smart thermostats — to achieve desired load shapes using machine learning and advanced optimization methods.

How does dynamic load shaping build on demand response?

Unlike traditional demand response — which requires utilities to manually set device-specific event parameters for peak periods — dynamic load shaping uses machine learning and optimization to dispatch granular groups of devices to achieve a desired load shape. This all happens while accounting for customer preferences and device characteristics.

With dynamic load shaping, demand side management teams and grid operators can precisely control VPP load shift profiles across DER types, allowing them to schedule VPP power production and mitigate snapback effects.

Delivering real-world results with dynamic load shaping

EnergyHub’s Edge DERMS platform is the first to demonstrate dynamic load shaping technology at the MW scale and across multiple device types. Our new dynamic load shaping functionality proved successful across three proof-of-concept tests during the summer 2024 season, delivering multiple grid benefits to three utilities.

1. National Grid uses its cross-DER VPP to extend peak reduction period

On August 28, 2024, we helped National Grid dynamically shape load using a combination of 20,000 thermostats and 2,400 residential batteries. During this four-hour event, we delivered a consistent load shape and minimal snapback to ensure load shed during coincident peak hours.

Throughout the event, we were able to match the target load shape while ensuring each device type participated for no more than three hours in alignment with National Grid’s program participant agreement. Our proprietary algorithm accomplished this by strategically segmenting resources into smaller sub-groups and optimally staggering the dispatch strategy for each group. By combining machine learning with optimization techniques, it identified the optimal start times, durations, and event parameters for each sub-group to achieve the desired outcome.

The transition from thermostats to batteries was orchestrated automatically to create a smooth and uninterrupted handover between these device classes. This approach allowed certain thermostats to continue providing additional load relief during the snapback period and as the battery dispatch began, ensuring a balanced and efficient shift in resource management. These results were only possible because National Grid manages both thermostats and batteries within the EnergyHub platform, unlocking greater value for the grid.

Graph of National Grid's dynamic load shaping results compared to the targeted load shape.

Fig. 1 Average VPP output compared to target (National Grid)

Graph of National Grid's dynamic load shaping results broken down by device group.

Fig. 2 Average VPP output by DER group (National Grid)

2. Large East Coast IOU test delivers sustained load shed target with minimal snapback

Also on August 28, 2024, EnergyHub deployed a test with a large East Coast IOU using 10,000 smart thermostats designed to achieve three hours of constant VPP output with no snapback for one hour after the event.

To address the need for longer periods of peak reduction, we dispatched a series of staggered thermostat groups. The test delivered three hours of consistent load shed, with each device having no more than a two-hour control period. There was no snapback for the first hour after the event, highlighting VPPs’ ability to deliver consistent and sustained levels of VPP output akin to a traditional generation resource.

Graph of East Coast IOU's dynamic load shaping results compared to the target load shape.

Fig. 3 Average VPP output compared to target (East Coast IOU)

Graph of East Coast IOU's dynamic load shaping results broken down by DER group.

Fig. 4 Average VPP output by DER group (East Coast IOU)

3. APS captures solar to boost load shed and beat the duck curve with minimal snapback

On September 27, 2024, we demonstrated that EnergyHub’s Edge DERMS platform can turn VPPs into resources that can be scheduled to dynamically solve grid problems. The objective of this proof-of-concept test with Arizona Public Service (APS) was to optimize solar availability during peak periods while aligning dispatch with customer time-of-use (TOU) rates.

To boost load shed with minimal snapback, the EnergyHub platform computed a strategy that involved pre-cooling homes during midday when solar production was high. Afterward, there was a gap before the customer time-of-use period began, during which the VPP output was leveled to a controlled, constant value. Once the TOU rate period started, the platform dispatched additional groups of thermostats to gradually increase load shed and help customers avoid peak rates.

This strategy allowed us to achieve an optimization period of more than seven hours, with increasing load shed throughout and no snapback. Our test with APS is a significant improvement from traditional demand response programs, which generally see load shed drop over time and a large snapback.

Graph of APS' dynamic load shaping results compared to the target load shape.

Fig. 5 VPP output compared to target (APS)

Graph of APS' dynamic load shaping results broken down by DER group.

Fig. 6 Average VPP output by DER group (APS)

Delivering the next generation grid flexibility

EnergyHub’s optimization tests prove that virtual power plants can dynamically shape load across different DER types and deliver sustained MW reductions comparable to conventional generation resources. Download our recent white paper to learn more about the future of flexibility and how to unlock the full value of VPPs.