Glossary term
Demand Response
A grid flexibility method that changes electricity consumption in response to grid needs, prices, control signals, or emergencies.
Definition
methodDemand response is the controlled change of electricity consumption in response to grid needs, prices, control signals, or emergencies.
Demand response provides grid flexibility by reducing, shifting, increasing, or shaping load while preserving acceptable service. It can support peak reduction, congestion management, renewable integration, reserve, emergency operation, and local distribution constraints. Its value depends on response time, power change, duration, recovery behavior, measurement, customer limits, and reliability.
Demand response changes electricity use to support a grid or site objective. It may reduce load during peak demand, shift energy to another time, increase load when renewable generation is abundant, or shed noncritical demand during emergencies.
The response should be described by power, duration, response time, notification time, recovery energy, and service impact. A load that can reduce 1 MW for ten minutes is not equivalent to a load that can shift 1 MWh over six hours.
A delivered response is often estimated against a baseline:
The equation is simple, but the baseline is not. It must account for weather, occupancy, production rate, calendar effects, data center workload, equipment state, and normal operating variability. Poor baselines can overstate flexibility or penalize a participant for changes unrelated to the event.
Engineering use
Common demand-response resources include HVAC systems, refrigeration, electric vehicle charging, industrial batches, water pumping, data center workloads, thermal storage, and batteries. Critical loads require careful limits because safety, comfort, production, asset life, or data integrity may be affected.
A useful demand-response design states the controllable load, unavailable load, notification time, ramp rate, minimum and maximum event duration, recovery behavior, rebound energy, telemetry, fallback mode, and operator override. For industrial processes, the constraint may be product quality or batch timing. For buildings, it may be thermal comfort. For data centers, it may be workload placement, UPS reserve, cooling margin, and network service level.
Measurement and verification should confirm that the response was real, repeatable, and available when called. Reliability matters: a resource that frequently opts out or rebounds into a new peak may have less grid value than its nominal connected load suggests.
Common mistakes
A common mistake is counting connected load as available response. Only verified, controllable, and recoverable load change should be treated as demand response. Another mistake is ignoring rebound after the event, which can move the stress rather than remove it. A strong demand-response review states the baseline method, controllable equipment, operating constraints, response test evidence, telemetry quality, rebound treatment, customer limits, and failure mode when the response is not delivered.