Integrated Demand Flexibility

For nearly three decades, electric power planning operated under a relatively stable planning paradigm: modest load growth, predictable asset replacement cycles, and infrastructure expansion paced over years rather than months. Today, that playbook is being tested by a very different set of system conditions.

Data center expansion, industrial reshoring, and rapid electrification are forcing power system planners to confront a sharp surge in load growth. At the same time, many substations, transformers, and feeders are aging or approaching thermal and operational limits.

The default industry reaction has been to accelerate traditional infrastructure investment: proposing new generation, expanding high-voltage transmission, and upgrading substations. Supply-side expansion remains essential, but relying primarily on new capital infrastructure creates structural bottlenecks across development timelines, supply chains, permitting processes, and customer bills.

The problem is not that traditional infrastructure is unnecessary; it is that it cannot always arrive on the timeline required by new load. Complex permitting, supply-chain backlogs, and protracted interconnection queues have stretched development timelines for major grid infrastructure. New generation projects can spend five years or more moving from an interconnection request to commercial operation, while major transmission projects can take a decade to plan, permit, and build.

At the same time, the cost of grid expansion is becoming increasingly visible in customer rates. As capital expenditures expand utility rate bases, new infrastructure projects intensify customer bill pressure and increase scrutiny from regulators and consumer advocates.

Grid planning must therefore evolve to weigh supply-side and demand-side solutions on a comparable basis. A growing pool of flexible load and distributed energy resources already exists behind customer meters. The planning challenge is to determine how much of that flexibility can be made dependable, locationally useful, and economically comparable with conventional infrastructure.

Moving Beyond Siloed Demand-Side Management

State regulatory frameworks and utility organizational structures have long managed demand-side resources in distinct operational silos. Energy efficiency programs are evaluated through their own cost-effectiveness frameworks; demand response is often treated primarily as a peak-management resource; time-varying rates are developed through separate rate-design and regulatory processes; and virtual power plants are often managed through standalone pilot or emerging-technology projects.

These independent structures can make it difficult for demand-side resources to receive consistent capacity treatment in Integrated Resource Plans and to be modeled as operational resources alongside generation and storage. As a result, demand flexibility is often evaluated as a program rather than as a resource capable of meeting a defined system need.

A unified framework — one that plans, dispatches, and evaluates distributed energy resources, energy efficiency, time-varying rates, and automated end-use controls as an integrated portfolio — would allow planners to test when, where, and under what conditions demand-side resources can substitute for or defer conventional investments.

Operationalizing Demand Flexibility in Power System Planning

Building a demand flexibility portfolio that satisfies both regulatory scrutiny and system operations means moving beyond static program design toward resource planning based on measurable system needs. Rather than evaluating customer resources in isolation, planners must align across five interconnected variables: when the system is constrained, where the constraint occurs, which end uses can respond, how customers are likely to participate, and how that response should be valued.

The work begins by reconciling time and location. Legacy demand response was often designed around broad, system-wide peak estimates and fixed dispatch windows, but modern grids face localized bottlenecks that vary sharply by circuit, season, and hour. On some feeders, shortfalls may appear during only a few dozen critical hours each year. On others, rapid electrification can create recurring evening constraints or push historically summer-peaking systems toward dual-peak or winter-constrained profiles.

Pairing hour-by-hour load modeling with AMI telemetry and GIS asset mapping allows utilities to identify the circuits, substations, and hours where incremental load reduction has the highest system value. That changes the question from “How many customers can we enroll?” to “How many kilowatts of dependable flexibility do we need at this location and during these hours to defer an identified investment?”

With physical constraints mapped, planners can build a bottom-up inventory of controllable end uses across customer classes. Evaluating the coincident-peak contribution of residential smart thermostats, heat-pump water heaters, and EV chargers alongside commercial HVAC automation, industrial pumping loads, batteries, and contracted flexible loads allows operators to combine resources with different durations and response characteristics into a portfolio matched to the duration and shape of the system constraint.

Installed capability is not the same as dependable capacity. Planners must account for enrollment, opt-out rates, device availability, rebound effects, event fatigue, weather sensitivity, and persistence over time. Automated controls, direct load control, and time-varying rates can all reshape load, but their contribution to planning should reflect measured or conservatively estimated performance under the hours that drive the system need.

Program design also matters. Automation, default enrollment where permitted, OEM-embedded controls, and well-designed customer incentives can materially increase participation and reduce the gap between theoretical and deliverable flexibility. The relevant planning metric, however, is not enrollment alone; it is the load reduction or shifting that can be relied upon during the specific hours for which the resource is being credited.

Ultimately, the regulatory question is straightforward: Is procuring demand flexibility less expensive than the infrastructure, generation, or energy it can avoid? Answering that question requires utilities to identify when and where flexibility is available, how reliably it will perform, and the specific system costs it can defer or avoid — from a distribution upgrade in a constrained area to new capacity needed during a handful of critical hours.

That comparison should be explicit. If a substation upgrade would cost $100 million and is driven by a 30-megawatt shortfall during 60 hours per year, planners should be able to test whether a portfolio of efficiency, load control, storage, and time-varying rates can reliably eliminate enough of that shortfall, for long enough, and at a lower net present cost.

The same logic should apply in resource adequacy and integrated resource planning: demand-side resources should receive capacity value based on the performance they can reliably deliver during system-critical hours, rather than being treated only as reductions to the load forecast.

With that locational and temporal specificity, regulators can compare demand-side resources directly with conventional investments and determine which provides the lowest-cost path to meeting system needs. The objective is not to replace traditional infrastructure. It is to make demand flexibility a resource that planners can select when it is faster, less expensive, or more targeted than the next conventional investment.