The problem demand response solves
An electricity grid has to balance supply and demand at every moment. Demand is not steady: it tends to peak at certain times, such as early evening, or during very hot or very cold weather. Meeting those peaks may require extra generation that is expensive to run and only needed for a few hours a year.
Demand response takes the opposite approach to building more supply. Instead, it asks some users to reduce their consumption, or move it to a quieter time, when the grid is under strain. A utility or grid operator offers a program, customers sign up, and devices or people respond when a request arrives.
Two basic kinds of response
Load shedding means using less during the event. A building might raise its cooling setpoint a few degrees, or dim non-essential lighting, for an hour or two.
Load shifting means moving a task to another time. Water heating, electric vehicle charging, dishwashers and some industrial processes can often run later without anyone noticing. The energy is still used, just at a time when the grid has room.
Some programs also reward turning usage up when there is a surplus, for example when a lot of wind or solar power is available. Details differ by region and by program, so what applies to you depends on your utility.
How connected devices fit in
Demand response can be done manually, such as a phone call asking a factory to cut load. IoT makes it quicker and more precise. A typical chain looks like this:
- The utility or an aggregator sends a signal: a price change, a request for reduced load, or a notice that an event is coming.
- A gateway, controller or cloud service receives the signal and checks it against the customer’s settings.
- Connected devices act on it. Smart thermostats can pre-cool or pre-heat a building and then ease off. Controllable water heaters, vehicle chargers and smart plugs can delay or pause their loads.
- Meters record what actually happened, so the customer’s response can be verified and credited.
That last step relies on accurate metering. Interval data from smart meters lets a program compare usage during an event with what would otherwise have been expected. That expectation, called a baseline, is an estimate, and how it is calculated is one of the more technical and sometimes contested parts of any program.
Types of programs
Programs vary, but common patterns include:
- Time-based pricing. Electricity costs more during peak hours. Customers who shift usage save money. Nothing is forced; the price is the signal.
- Event-based programs. The utility calls a limited number of events and pays credits or bill discounts for participation. Customers are usually told in advance, and often can opt out of individual events.
- Direct load control. The customer allows the utility or an aggregator to cycle certain equipment, such as an air conditioner, for short periods in return for a payment or discount.
- Commercial and industrial programs. Larger sites commit to reducing a certain amount of load when called. Penalties may apply if they fail to deliver, so these participants usually plan carefully.
Why the communication layer matters
Demand response is only as reliable as its links. A request that never reaches a device does nothing. Because homes use a mix of wireless protocols and brands, getting devices to respond consistently is harder than it sounds, and device interoperability is a recurring practical obstacle. Devices also need a sensible plan for losing contact. A thermostat that cannot reach the internet should keep following its last known schedule rather than freezing in an extreme setting, which is a design issue related to offline operation.
Benefits and trade-offs
For the grid, flexible demand can reduce peak strain and may lower the need for the least efficient generation. For participants, benefits can include bill savings or credits. For the wider system, shifting demand toward times of abundant clean power can help, although the effect depends on the local mix of generation.
The trade-offs are real:
- Comfort. Pre-cooling a home or delaying a laundry cycle is usually unnoticeable, but a hot afternoon with the air conditioning cycled off is not. Good programs let you set limits.
- Savings are uncertain. Rewards depend on the program, your usage and the weather. They should not be assumed.
- Privacy. Detailed interval data shows when a household is active. Read what a program collects and how long it keeps it, a point covered under data retention.
- Rebound. After an event, delayed loads often restart together, which can create a smaller spike of their own. Well-run programs stagger the restart.
Checking a program before joining
Look at who controls the devices, whether you can override an event, what happens if you do, how rewards are calculated, and what data is shared. A reasonable program makes these answers easy to find. If it does not, that is worth noting before signing up.