There are roughly 28 GW of demand response capacity in North America, and the bulk of it comes from large commercial and industrial customers who make up the backbone of the resource.
According to new data from Navigant, 21 GW of North American demand response capacity comes from C&I programs, and 7 GW from residential. The industrial side of demand response has been active for decades, maturing for a variety of reasons: higher payments make the programs valuable to businesses, economies of scale favor larger resources, and utilities have more experience attracting those customers.
But C&I demand response growth is slowing, while the residential side accelerates, and within a decade, the resources will be much closer in size. C&I demand response programs are growing at about 2.5% annually, compared with 14% on the residential side.
By 2025, Navigant is forecasting 26 GW of C&I demand response, and 23 GW for residential.
"The C&I was the low hanging fruit," said Brett Feldman, principal research analyst with Navigant and an author of the report. "All the largest customers are already participating or they’re well-educated on the programs and have decided not to participate for their own business reasons. I don’t think there are a lot of large customers who aren’t aware."
But with demand response overall still growing rapidly—about 75% over the next decade—does it matter where the resource comes from? The issue is cost. With C&I accounts already enrolled, the price of a demand response MW could be set to rise, as utilities are forced to chase smaller resources. Residential aggregation may be a part of the solution, as utilities look to third parties to help control marketing and management costs. But demand response providers say there is still plenty of room to maneuver in the industrial space, especially with new rules coming into play, international opportunities growing and utilities seeking to optimize their demand management portfolios.
"There’s still plenty of growth in C&I sector in North America but we certainly also see a lot of potential internationally," said EnerNOC General Manager of Demand Response Christian Weeks. "The growth rate for C&I internationally will be higher, because the U.S. is a market leader. …. There’s still opportunity, but most international markets have yet to embrace demand response."
“We’re seeing increases in growth for all segments,” said Nir Matalon, director of product management at AutoGrid, which builds demand response management systems. “Technology as software really enables us to go after a much larger base of customers, including C&I.”
Commercial and industrial customers make up a majority of AutoGrid’s contracts, and across the industry software is lowering the cost of sales, while open standards are making it easier to connect installed devices, Matalon said. “And finally, software is also allowing the provider to manage it all.”
Demand response growth will also depend significantly on the market in question, Feldman said. PJM and ISO New England are "more mature markets" and "probably peaked a few years ago." But other areas have ample room for growth, including California and Texas.
Adam Miller, director of Product Development at Gexa Energy, said commercial demand response is growing in the Electric Reliability Council of Texas market, where the operator will pay $50 million to participants in its Emergency Response Service this year.
"Given the growth, ERCOT is considering changes to advance the program," Miller said. "With the US Supreme Court upholding FERC Order 745, we see continued opportunity to help customers get compensation for reducing use and responding to grid events."
Growth in the residential sector
Miller said there may be more growth on the residential side, but C&I's entrenched position means it will continue to grow as well. He sees connected heating and air conditioning, water heaters and battery storage as the industry's greatest opportunity, but said C&I customers "have existing infrastructure and scale to participate."
"The markets are efficiently compensating residential customers for their effort and inconvenience," said Miller. "While there have been significant advances with smart homes, we still think there is opportunity to make it easier for residential customers to participate in demand response."
Matalon said AutoGrid is seeing growth in a couple of areas, including Bring Your Own Thermostat programs and distributed energy resources.
BYOT programs are helping demand response companies “reduce the cost of the megawatt,” Matalon said, because the customer already owns the device. On the distributed resource side, technology is pushing response time and dispatch speed, meaning the resources can be used in a wider range of applications.
“It allows us to optimize the lowest-cost dispatch or highest-reliability and control them,” Matalon said. “And then use software that bids those into markets.”
"Utilities are always going to have larger residential programs," Feldman said, in terms of number of customers. But "1 MW from a large industrial customer costs a lot less than trying to get a megawatt from 1,000 residential customers."
Investment in enablement hardware and software spending is higher on the residential side, Feldman added, with about $74 million annually in spending done annually on the C&I side versus $103 million.
Revenues from the residential side will grow much faster as well, between now and 2025.
On the C&I side, market and utility payments to aggregators and customers will rise from $1.2 billion to $1.7 billion. On the residential side, those revenues will rise from $300 million to $970 million, according to Navigant's research.
Making older C&I more reliable
C&I opportunity extends beyond signing new customers. Because the programs are often older, they may be built on less-effective technology, utilize outdated measuring and verification processes, or be based around less-reliable pricing signals.
"The rules are getting more strict," said Feldman. "In the early days the rules were pretty loose, RTOs and ISOs didn’t fully understand how to measure demand response performance compared to generators …. Things like baselines have gotten tighter and they’ve added more requirements to make it more comparable to generation, posting financial insurance, more risk and process."
"There are a lot of interruptible rate programs out there," said EnerNOC's Weeks. "We’ve seen a few utilities wondering if we could help convert some of those programs into more reliable DR resources. There are thousands of megawatts in these interruptible programs."
Weeks said the programs were often launched as economic development incentives, and are infrequently dispatched. Some utilities believe "they may not be as reliable, if they were needed in earnest," he said, as they are often not using real-time metering or telemetry.
"The utility is going to need a more reliable resource," Weeks said. "The expectation might be they’re going to dispatch it more regularly going forward."
Matalon said utilities are using platforms like AutoGrid's as a way to manage a range of existing programs.
"Utilities, even though they already have all of these programs, they use AutoGrid as a way to manage them all together," he said. "They come to us to make sure that these assets are dispatachable. If it's a generation resource, you flip a switch and it's on." Older demand response programs can be trickier, he said. "Without a smart meter its difficult to assess how those end points performed."