From Pilots to Portfolios

A gap exists today between the load flexibility utilities need to meet AI-driven load growth and what they can realistically achieve through existing systems and solutions. After nearly two decades of flat demand, utility load forecasts have turned sharply upward: a recent report by Grid Strategies finds the aggregate five-year forecast of peak demand growth across U.S. utilities has increased sixfold over the last three years, driven largely by data centers. Utilities will need to add over 200 gigawatts of capacity to meet the increase in peak demand expected by 2030, according to the U.S. Department of Energy (DOE).

By leveraging load flexibility, utilities can unlock additional capacity rapidly and at lower cost relative to generation and T&D buildout — but this resource has been underutilized to date. Per EIA data, utilities dispatched just 12.3 gigawatts of demand response at peak in 2024 — less than two percent of the record 745-gigawatt national peak, and below the level of a decade ago — while studies by Brattle and the DOE estimate cost-effective flexibility potential at 10 to 20 percent of peak.

Load flexibility for utilities today largely consists of technology-specific virtual power plants (VPPs), managed EV charging pilots, and time-varying rates (TVR) as a patchwork of siloed efforts rather than an integrated whole. While many utilities have implemented distributed energy resource management systems (DERMS) to integrate these resources dynamically within the grid, their interactions are rarely modeled, and inclusion in planning is limited. Without jointly optimizing these tools and including interactive effects within an integrated modeling and planning framework, utilities cannot realize the full value of load flexibility.

Ideal Future State — Load Flexibility Portfolios

To deliver speed to power while maintaining energy affordability, utilities and regulators will need to work together to enable, establish, and evolve portfolio-based approaches to load flexibility. By adopting a portfolio of load flexibility tools and modeling how those assets interact, utilities can build the stack of load flexibility resources best suited to unlock capacity when and where it is needed at lowest cost.

Designed, deployed, and orchestrated in unison, load flexibility tools shift and shave peak load more effectively and efficiently than any individual solution could in isolation. For decades, utilities have worked to create energy efficiency and demand-side management portfolios by layering in additional programs thoughtfully, modeling potential interactions, and assessing how the EE/DSM portfolio as a whole affects not only absolute energy consumption, but just as importantly system peak demand. That same collaborative discipline must now be applied to load flexibility — building a portfolio capable of meeting regional and system peak capacity requirements immediately, and over the next decade as data center demand increases.

Utility Load Flexibility — Current State

Molly Podolefsky: To deliver speed to power while maintaining energy affordability, utilities and regulators will need to work together to enable, establish, and evolve portfolio-based approaches to load flexibility.

Utility Context: Utility load flexibility tools, operations software, and planning processes exist largely in silos. Utilities often implement load flexibility tools and programs such as managed EV charging, smart thermostat VPPs, EV time-of-use (TOU) rates, and residential demand response (DR) in isolation, without modeling the joint impacts and synchronizing to shift load optimally.

Rather than complementing one another, these programs may contribute to new suboptimal system peaks, even working at cross-purposes at times due to lack of coordination. In the absence of a DERMS, utility control room operators may not be able to integrate flexible resources dynamically within the grid to maximize load-shifting potential.

While DERMS and real-time integration of flexible loads within grid operations are becoming more common, most utilities do not have the software required to model interactions between all the load flexibility programs and tools in their portfolio and accurately embed these resources within system planning.

Deepening the challenge for utilities, functions such as generation, transmission and distribution, operations, planning, rates, and customer experience are often siloed in departments with separate leadership, goals and objectives, metrics, budgets, and procurement and planning processes, hampering change management and the development of portfolio-based approaches to load flexibility.

Regulatory Landscape: The siloed and piecemeal state of load flexibility within utilities is in part driven by the regulatory environment. State regulators have not traditionally provided the incentives necessary for utilities to develop a portfolio approach to load flexibility — most utilities are not subject to portfolio-level load flexibility targets and are rarely required to anchor load flexibility tools to a single unifying dynamic price signal.

Complicating the picture, different tools and resources, from VPPs to C&I direct load control (DLC) programs, are often subject to different cost-effectiveness tests by regulators. Many utilities are also subject to multiple, stand-alone state regulatory filings on resource planning, capital investments, grid modernization and other utility imperatives — each of which treats flexibility resources in isolation.

Bridging the Gap

Steps for Utilities: Utilities can begin taking steps to drive formation of integrated, co-optimized utility-level load flexibility resource portfolios. Through AI-based rate engines, utilities can design and implement AMI-informed TVR to shave and reshape peak load. Using dynamic modeling tools, utilities can jointly model and co-optimize TVR and other load-shifting tools such as VPPs to maximize load-shifting potential.

In tandem, utilities can begin implementing software systems that together will form the utility’s operational load flexibility fabric — a foundational DERMS layer for system-level orchestration and operational integration, and multiple edge-DERMS solutions, such as VPPs or managed EV charging programs, communicating through open protocols with the DERMS layer to synchronize vehicle-to-grid (V2G) charging, battery energy storage systems (BESS), smart thermostats, and other grid-edge devices.

At the same time, utilities must undertake organizational changes that support a portfolio approach to load flexibility — breaking down silos to allow for and incentivize cross-department planning, budgeting and implementation, modeling and incorporating interactions between different load flexibility tools as standard for planning, adopting processes and metrics to value all load flexibility assets the same way, and standardizing requirements such as open protocols for technology procurement.

Steps for Regulators: Public utilities commissions (PUCs) and other state regulators can require consolidated and integrated portfolio filings rather than stand-alone annual reports on performance, resource adequacy, investment planning, and other reporting obligations. They can adopt performance-based incentive mechanisms to ensure ratepayers are paying for outcomes rather than programs and quantify portfolio-level load flexibility targets so that utilities will select the portfolio of resources to meet those targets at the lowest cost.

Finally, regulators can require a single unifying dynamic price signal across flexibility tools, and reform cost-effectiveness testing to place diverse load flexibility resources and tools on equal footing. With California’s Avoided Cost Calculator as an example, regulators can create a single commission-maintained dataset of temporally and locationally differentiated avoided costs — inclusive of energy, generation capacity, transmission and distribution deferral value, and ancillary services — that all demand-side valuation must reference.

When every program, rate, and cost-effectiveness test is anchored to that same cost curve, the flexibility portfolio is unified analytically; published in machine-readable form and, where applicable, embedded in dynamic rate design, the same backbone becomes the operational price signal that flexibility tools dispatch against.

Looking Forward — Building a New Paradigm

Organizational change, especially where regulatory requirements must evolve in addition to utilities’ internal systems and processes, takes time — but time is not on our side in meeting the mounting demands of data centers and AI-driven load growth. Utilities and regulators can take steps now to build a portfolio approach to load flexibility that harnesses the full power of DER deployment across the grid, unlocking capacity to meet the demands of load growth while maintaining affordable rates for consumers.

Data centers themselves may be the next horizon for load flexibility, as hyperscalers architect the ability to orchestrate and shift load dynamically so that data centers can serve as VPPs. However, if utilities take the same approach, adding data-center VPPs as just another isolated tool, they run the risk of creating massive load flexibility silos that fail to achieve their full potential.

Utilities must instead create holistic strategies for load flexibility portfolio development and begin taking steps today to build a new paradigm that is better positioned to meet the evolving needs of a system undergoing rapid transformation and growth.