by Dr. Steve Isser, Good Company Associates
“”˜When I use a word,’ Humpty Dumpty said, in a rather scornful tone, “˜it means just what I choose it to mean, neither more nor less.’
“˜The question is,’ said Alice, “˜whether you can make words mean so many different things.’
“The question is,’ said Humpty Dumpty, “˜which is to be master– that’s all.'”
–Lewis Carroll, “Through the Looking Glass”
When I volunteered to work on the North American Energy Standard Board (NAESB) demand-side management (DSM) measurement and verification (M&V) effort, I thought the definition of demand response (DR) was pretty simple: measures that reduce demand, on command.
As I examined the issue, I realized that it was far more complex than I first thought, that different people used different names for different activities, and that names mattered.
When you peruse various documents from the Federal Energy Regulatory Commission (FERC), the North American Electric Reliability Corp. (NERC), independent system operators (ISOs), state agencies or private groups, it becomes obvious that there are no generally accepted definitions for many terms used in the industry. DR, load management, smart grid, etc., often have different meanings for different groups.
DR is generally used as an umbrella term. For example, the FERC definition of DR covers the complete range of load-shape objectives and customer objectives, including strategic conservation, time-based rates, peak-load reduction, as well as customer management of energy bills.
I use DR as a term covering the spectrum of dispatchable measures (the term “active demand response” is used in a similar fashion) that reduce load for specific periods when called. Using my terminology, load management (also referred to as “passive demand response”) refers to measures that encourage participants to voluntarily shift load from one period to another. The key is “control.”
If I send a price signal in a voluntary program, I may get a predictable response but one that may change as relative prices change.
DR implies a contractual obligation to provide a set quantity of load reduction when requested.
System operators want to know when they call on a resource that it will be available, not that it will probably be available, and NERC rules reflect that distinction.
NERC includes only direct load control and “interruptible demand” (customer demand, in accordance to contractual agreement, can be interrupted on peak) as resources in its reliability assessments.
The distinction between active and passive, however, is merely one dimension of DR diversity. There are a number of other key dimensions: the type of market, type of load, technology and load-reduction service.
NAESB split its business-standards development effort between wholesale and resale markets. This was an attempt to distinguish the requirements of the ISOs that run formal wholesale markets from vertically integrated utilities and distribution companies, public power companies, electric cooperatives and municipal electric companies that sell electricity directly to final customers. Many of these utilities are embedded inside wholesale markets, so they may be involved with DR and load management at wholesale and retail levels.
DR resources can bid into ancillary service markets at the wholesale level. Higher-quality DR resources (in terms of the speed and reliability of response to a curtailment request) are capable of meeting more demanding service requirements, such as power regulation (requiring very rapid response) and spinning reserves.
There are also emergency DR markets providing payments for responding during emergency conditions.
Load-management programs tend to have much longer lead times, often with notification a few hours or even the day before an event. Some programs rely on prices that vary by time of day and season (time of use).
Other programs provide advanced warning to consumers when they will be charged superhigh rates for a few hours (critical peak pricing).
Ontario makes payments to loads that shift consumption from targeted time periods. The New England Forward Capacity Market purchases both DR and load-management services.
Loads differ between customer class and electricity-consumption equipment. Residential electricity use is largely for cooling and heating (depending on climate), lighting and appliances.
As income increases, appliances and specialized applications (for example, pool pumps) become more important.
Commercial buildings also use electricity for lighting and HVAC purposes, but some customers such as grocers may have refrigerator and freezer applications.
Industrial consumption can incorporate a wide range of machinery and processes from motors and compressors to crushers, rollers and furnaces. Different loads have different response times, quantities of load that can be shed and acceptable duration of curtailments.
Technology has provided a means to deal with small customers and large, complex loads in industrial facilities. Direct-load control, using switches or smart thermostats, allows thousands of residential loads to be monitored, aggregated and controlled.
Energy management systems in commercial buildings allow pinpoint control of loads. Even more sophisticated control systems allow industrial facilities to coordinate production activities to reduce the cost and increase the flexibility of load curtailments.
Advances in communication over wide areas and across homes or factories and software and equipment controls are reducing the cost and increasing the availability and quality of DR resources.
An unsophisticated DR resource might consist of a guy at the plant who has a mobile phone or pager, and when he gets called, he hits the off button.
As long as the communications are secure and dependable and the worker is reliable, this simple solution could be sufficient to allow participation in 10-minute nonspin markets and provide emergency reliability products.
A highly sophisticated DR resource might have multiple loads, some with variable controls, some with switches, connected to a central control system.
By combining portfolios of different loads with different characteristics, a trader can optimize sales of curtailment services into different markets depending on contractual requirements and prices. Custom software might be employed to determine the optimal strategy for bidding DR resources into various markets.
The nexus between DR and load management in practice stems from that a DR resource usually has the capability to respond to the price signals and incentives incorporated in most load-management programs. This is why it’s important to distinguish between the resource and the service or product it can provide.
If you can respond in 10 or 30 minutes, you can respond given an hour’s notice or if you have a reasonable expectation of peak energy prices in a few hours.
Being able to respond quickly, however, doesn’t necessarily mean it would be cost-effective to reduce load for an extended period.
The decision to participate in various markets involves more than the technical capability of curtailing load. Loads don’t exist in a vacuum; they reflect the purchase and use of electricity, and businesses and individual consumers don’t purchase electricity because they like electrons.
Curtailing load means reducing electricity-powered services that have some value, or there wouldn’t be a load.
Reducing consumption creates net losses for consumers of electricity because they buy electricity when its value is greater than its cost. Those costs of load curtailments will depend on the magnitude and timing of the curtailment and the electric service that is temporarily interrupted.
A homeowner at work might be willing to allow her air conditioner to be cycled for a few hours with temperatures rising to 80 degrees as long as the temperature is reduced to at least 76 degrees by the time she gets home.
Some industrial facilities have loads that can be shut down quickly or reduced at little cost, but after time, costs may escalate rapidly as their continued shutdown affects other business activities.
A frozen food warehouse might be able to shut off the freezers for a few hours because thermal mass allows temperatures to remain at safe levels, but a grocery store refrigerator might require constant cooling to meet public health standards.
Opportunity costs also explain why we see less price response in electricity markets than predicted.
Responding to price has two costs: monitoring prices and responding to them, and reduced electricity consumption.
If a price spike occurs with little warning, few electricity consumers will know about it until the event is over.
If a price spike occurs for a short period, it may cost consumers more to respond than one hour’s worth of expensive electricity.
Automation is one means to overcome these costs of responding to prices.
As load reduction becomes more automated, whether through smart thermostats at home or sophisticated controls at work, responding to price signals becomes easier and less costly.
The potential to automate price response is one of the reasons for investments in smart grids. Nevertheless, it will be a decade before the majority of consumers see or react to real-time prices.
There is always a lag time with prices, and prices are often too variable for many loads to bother to track and react to them. This is why most load-management programs use time-of-use and critical-peak pricing.
Advanced warning and predictable prices make it easier for consumers to decide to reduce or shift electricity consumption.
The more immediate impact of investments in smart grid technology will be to increase the quantity and quality of DR resources. Price-responsive load will not replace DR.
While load management has real value to a utility or electricity market by reducing peak demand and the investment in infrastructure required to service that demand, it lacks the “firmness” that dispatchable, or contractual, load curtailment provides to system operators.
Instead, the expansion of DR capacity and price response capability will allow utilities and ISOs to employ both tools to reduce costs and increase the reliability of the grid.
Dr. Steve Isser is vice president and general counsel of Good Company Associates and vice-chairman of the Peak Load Management Alliance (PLMA). Reach him at email@example.com