Smart Grid -- What is Demand Response?

Joao Batista Gomes | Apr 20, 2010


Consumers were previously treated in a homogeneous, today with the introduction of new technologies each consumer will have different features. For the success of Utilities in developing this new scenario, they must have the ability to understand the different distribution topologies and further understand the consumption profile of each class of consumer.

Demand Response involves encouraging customers to cut back or shift their electrical use or demand in response to grid emergencies or high market prices for electricity.

Today there are three different types of Demand Response services:

  • The first is where customers receive compensation for electing to standby to reduce a portion of their electric demand in a grid emergency. These are called "capacity resources" and are typically activated a few times a year for up to 6 hours at a time.

  • The second type involves sending customers price signals to encourage them to reduce demand during peak hours. The higher the hourly prices, the greater financial incentive a customer has for reducing their electric load. Customers can participate at their discretion for as little as one hour at a time. These are called "energy markets."

  • The last involves very short grid-initiated curtailment events with very short notification. These are typically 10 to 30 minute reductions with 10 minutes notice. These are called "ancillary services" and help the grid operator smooth out short-term imbalances of electrical supply.

Different Utilities or Regions, Substations, Feeders and Consumers should be treated differently, the faster the Utilities understand this, will have fewer problems with their consumers.

Related Topics


I was trying to think of why, if a utility had implemented your second method to all of its customers, either of the other two might be needed. The only eventuality which comes up is eg. an office building with full generator standby capability who requires to switch over before start of working hours or not at all because the switchover will interrupt the work on desktop computers during work hours. This sort of capacity reduction would not be available to any price broadcast system which didn't make its price curve known to the customers a day ahead. Of course IMEUC handles it well.

Demand Response should really be price response (alternative 2) as Len suggests. Even ancillary services can and should be handled this way, but for the fact that customers and utility engineers have been conditioned to expect moderate energy price volatility and capacity payments that provide the "missing money".

Until and unless energy suppliers are required to recover all of their costs from market revenues, customers that cannot or do not want to be flexible will either have to provide subsidies in the form of capacity payments for those customers that are flexible, or they will have to pay for new generation.

With demand curtailment tied to price, the demand side of the market will work like the supply side. As generators are dispatched in order of increasing price, customers will be shut off in order of increasing ability to pay. Demand response may be thought of as shutting off the power to the poorest first.

I was very intrigued by your opening paragraph and hope that you will continue to develop and contibute your ideas about how to categorize smart grid markets. Defining the abilities of the customer might be a good place to start. It is hard to imagine a set of demand response options configured for residential customers could match those needed for commercial entities. Likewise, the ability of the customer to respond that I would expect from retail store manager are far different and less substantial than what I anticipate could be provided by facility engineers running a univeristy campus, for example.

There is an additional customer response type I have never seen considered. About 20 years ago I asked engineers at the utility I worked for if customer meters could be economically designed to respond to frequency. The answer back then was that turning off in response would be easy. Turning back on would be difficult. The objective, of course, was to allow customers to provide load reductions voluntarily in response to a grid upset.

Over the last 20 years the value of service has increased for the average residential customer as we replaced gas pilot lights with electric ignitions and have come to depend on computers, other electronics, and the internet. But it would be relatively easy to build frequency response in to refrigeration equipment, dish & clothes washers, etc. If we had the protocols, customers could help avoid blackouts without even noticing the short interruptions of loads that need not run continuously and allow the grid to get back into balance. It could be built into appliances as part of the intelligent grid, without needing the smart meter.

Good ideas but let's not lose sight of the fact that all this does is flatten the demand curve. Expensive peaking generation (gas, oil) is not required (as much, perhaps not at all).
Creeping up on us all is our growing economies. Growth means that more, not less electricity is required. Certainly there are efficiencies to be had but it follows the law of diminishing returns as savings become harder and harder to find and justify economically.
Sooner or later new power plants must be built to replace existing units that are nearing the end of their design life and to provide the supply for growth.
Of course if the public is content with a zero growth economy and no further improvements to their standards of living then the problem goes away.

My observation is the opposite of that. People seem to want nicer homes more appliances and therefore more electricity to operate it all. While I am sure these ideas will work to some degree they should never be considered as a replacement for continuing to maintain and buid the existing electrical infrastructure.

Malcolm: "People seem to want nicer homes more appliances" -- While your observation is clearly correct, still, there are limits outside of people's wants. Present known reserves of most mineral fertilizer materials are only good, at the present rate of consumption, for "perhaps" 80 years. Even if we double that for a willingness to mine low-concentrat ores which require higher energy inputs, at some point there will be no more fuel to power the mines or the ships. Many exotic metals on which many present industries depend, eg. electronics, have much shorter time frames.

We should be starting now to carefully and accurately plan for all the known limits to growth, including planning for what happens when another 3 billion ++ presently non-participating persons decide to join the party.

Malcolm, I am not the pessimist that Len appears to be. We have consistently found new sources for raw materials, or substituted better ones for the declining ones. The 787 largely replaces aluminum with carbon fiber. Nuclear power can replace fossil fuels, etc.

But do not trivialize a large reduction in the need for peaking power plants. They are extraordinarily expensive per kWh produced and require a massive, expensive, transmission system that is rarely used. Average costs will come down significantly when/if the electric industry becomes efficient by greatly reducing peak generation. Such greater efficiency is the rare opportunity to be good for our wallets and the environment at the same time. Add thermal storage to use night time generation to serve air conditioning loads to the other efficiencies and we will need fewer new plants than you expect. And many of those can be cogeneration, using wasted process heat to additional budget and environmental benefits.

Eventually the U.S. will need additional generation, some of it central generation, and Asia needs a lot now. But we can largely avoid new peakers and that is an accomplishment.