California’s investor-owned utilities are ready to test how effectively aggregated demand management resources can be used to serve bulk grid needs.
San Diego Gas and Electric (SDG&E), Southern California Edison (SCE), and Pacific Gas and Electric (PG&E) have picked the private sector players whose aggregated resources will go into the state’s Demand Response Auction Mechanism (DRAM) pilot program.
In compliance with a December 2014 regulatory order, the IOUs promised the California Public Utilities Commission (CPUC) an auction for sellers of aggregated demand response. Last October, providers filed applications for participation in the new market, and utilities announced the winners earlier this month.
The winning bidders will be used in the day-ahead electricity market of the California Independent System Operator (CAISO), the state’s grid operator.
The IOUs will use the selected 40 MW of demand response (DR) product to meet a small portion of their obligations to the grid operator for resource adequacy. SCE accepted bids for more than 20 MW, PG&E took 17.7 MW, and SDG&E, with a smaller customer base, took just under 3 MW.
The winning private sector participants fall into four basic categories. The biggest players are the traditional demand response (DR) providers, who will use their customers’ commitments to reduce load.
Johnson Controls/EnergyConnect, IPKeys Power Partners, and EnerNOC, which provide energy management services to commercial-industrial customers, will be able to offer the CAISO demand reduction by assisting their customers in strategically dialing down energy use.
A new provider on the scene was eMotorWorks, which makes electric vehicle (EV) chargers and charger components. Its smart charging technology will recruit EV batteries’ storage to help grid operators shift loads.
More distributed DR will come from residential and small business energy management service companies like OhmConnect, Chai Energy, and EnergyHub. These companies will reward customers with smart thermostats for reducing their distributed heating and cooling and other electricity loads.
The novel category in California's DR program is aggregated behind-the-meter battery energy storage at the commercial-industrial and residential levels. Winners in the IOUs’ DRAM auctions were Green Charge Networks and Stem.
EnerNoc has been offering DR to utilities and organized wholesale electricity markets like PJM, NY ISO, ISO New England, and ERCOT through bilateral, power purchase agreement-like contracts for over a decade. It also provides DR in international markets from Germany to Japan.
“They buy a firm capacity product which takes the form of customer reductions in consumption on demand,” said EnerNOC Sr. Director of Regulatory Affairs Mona Tierney-Lloyd.
Buyers contract for use reductions for a specified number hours and time of the year, typically during system peaks. Contracts also specify an advance notice of a day, hours, or minutes.
These terms make meeting DRAM contractual obligations —such as ensuring load reductions of four hours per day for up to three consecutive days — familiar to EnerNOC. It will be working in the program with PG&E and SCE.
Other specific contract terms and prices for all three utilities and all private sector participants remain confidential.
“Our obligation is to find customers willing to curtail their load when we call them and we pay them for that load reduction,” Tierney-Lloyd said.
EnerNOC has two sets of concerns going into the first round of DRAM operations, which begin this summer, Tierney-Lloyd said.
The first set of concerns is around the parameters of participation for this first-of-its-kind California wholesale electricity market program, she explained. “It is testing a lot of new things at one time.”
The commission’s cautiousness is evident in the program procedures. “We don’t go through as formal a process in other markets,” Tierney-Lloyd said. California is more demanding “to make sure customers, especially residential customers, are adequately informed and protected.”
EnerNOC’s other set of concerns is that the utilities are already very near having or already have acquired the number of accounts the CPUC asked them to be prepared to manage.
“Because there was a significant amount of small residential customer participation, those accounts were taken up in DRAM 1,” Tierney-Lloyd said. “We will strongly be considering participating in round 2 but my concern is that the utilities will not be able to handle a large number of DRAM 2 customers.”
This concern may be because it is a large customer business and is unfamiliar with the aggregation of smaller accounts. But its uncertainty is increased because “there is no clear direction to the utilities about their obligation for DRAM 2,” she said. “It is entirely possible the utilities can accommodate large numbers but we should get a better understanding of that capability."
EVs for DR
The eMotorWerk (eMW) success in the DRAM bidding and other competitive solicitations for energy services is due to the fact that its chargers are designed to be smart grid resources, explained CEO Val Miftakhov. “Across our entire network, we can modulate load 100% in three seconds.”
Unlike most of its competition in the charger station market, he said, its dispatch speed allows eMW to access revenue streams from the real time five minute energy market and the frequency markets which require a four second response.
“We are able to provide capacity at a very low cost for a couple of reasons,” Miftakhov said.
First, eMW is already in the market as EV charging. “Our 10 kW, 40 amp, 250 volt charger retails for about $600 and is among the leaders in a market where 60% to 70% of EV buyers also buy a charging station,” Miftakhov said.
Second, eMW offers a discount for participation.
“Every resource I put into the market, through multiple revenue streams, can generate several hundred dollars per year,” Miftakhov said. “I can share that with my customer if I want to increase my market penetration. I can also share it with EV manufacturers to encourage them to include my technology in their vehicles or their charging stations.”
Distributed battery storage
In terms of sheer capacity, contributions in the DRAM program’s first round from distributed battery energy storage providers are relatively small. But the 530 kW from Green Charge Networks and the 350 kW from Stem, split between SDG&E and SCE, could be huge contributions to the future of DR.
“We install behind the meter distributed energy storage systems at commercial and industrial facilities,” Green Charge CEO Vic Shao said.
Its core customers include school districts, municipalities, hospitals, retailers, offices, and industrial spaces. Its core strengths are its software, aggregated network, and access to financing.
Green Charge's commitment to the DRAM program is a small fraction of the 30 MWh of battery storage it now has in operation or in construction.
“The bulk of the winning first round bids went to traditional DR but when the characteristics of battery storage are fully realized, it will beat the current demand response tools hands down,” Shao predicted.
Two advantages distributed battery storage advocates claim over traditional DR are its automated instantaneous response time and its ability to fulfill its DR commitment with no disruption to the utility customer.
Traditional DR requires giving the customer notification, waiting for them to react, and then curtailing their electricity use if they react,” Shao said. “When we reduce the load at our host customers’ meters, it has no impact on operations or behavior at all. The customer simply draws energy from storage instead of from the grid.”
A combination of demand response and energy storage might be ideal, he acknowledged. In the DRAM, the utilities are buying both and aggregating it.
The DRAM was re-designed as a new way to do DR that is integrated with the CAISO markets, said Stem Director of Policy Ted Ko. Originally conceived for traditional DR, the program evolved as policymakers realized battery storage could provide the same or better performance.
Stem’s focus is software that allows consumers to use storage more effectively to reduce their demand charges “but, in the future, we could be doing the same thing with other kinds of DR-like resources,” Ko said.
For DR as a grid resource, storage has another advantage beyond speed and reliability of response, he said: It can be used more.
“It can also be called on a lot more often," Ko said. "Traditional DR end customers will only accept usage so often but storage-based DR can be called on as often as necessary because the end customer doesn’t see any difference at their site.”
Traditional DR’s larger scale customers may allow it to provide utilities and grid operators with more load volume, Ko acknowledged, "but there is that reluctance of customers to be called on repeatedly."
Storage is also not limited to large scale commercial-industrial customers. “With aggregated distributed storage,” Ko said, “the potential universe of load is much bigger.”
But no DR program, including the DRAM, has set a value for the performance advantages of battery storage, Ko said. It should be credited for those capabilities in competitive bidding for DR and other energy services.
The first round of DRAM is aimed at proving resource adequacy (RA) and reflects the RA policies’ failure to fully value battery storage, he said. Going forward, “the DRAM could consider procuring flexible RA and valuing the different performance characteristics and response times of resources like battery storage that can provide it.”
Stem’s bids are cost effective because its intelligent energy management capabilities will allow it to capture multiple value streams with the same storage asset. Ko said. Stem will be able to use the storage in its contracted fleet to both meet its DRAM obligations and reduce its customers’ demand charges.
“We bid at prices that will be cost effective when we stack those two value streams,” Ko added. “That is what we are demonstrating in this pilot.”
But while battery storage garners attention for its new role in demand response, research shows traditional DR is more cost effective than storage, Tierney-Lloyd said. And the limits of its customers’ responsiveness is not necessarily an issue because that is designed into the EnerNOC commitments.
“We know how customers perform and that is baked into the capacity of our portfolio," she said. "If we have a 10 MW commitment, we might have 12 MW of customers available.” Company tracking shows “over 100% commitment fulfillment across all dispatch.”
In the future, competition between traditional DR and batteries could give way to collaboration, such as EnerNOC's recent announcement that it will work with Tesla. They will combine battery storage with load curtailment to provide DR services to the grid.
“Although in the DRAM bidding they are competing technologies, there are also partnership opportunities,” Tierney-Lloyd acknowledged. While batteries are more useful for fast-response load following, automated traditional DR "can be an automated aggregated resource with a comparable response time.”
EnerNOC pilot programs in Canada, Germany, and with the Bonneville Power Authority in the Pacific Northwest are demonstrating grid stabilizing capabilities with automated responses within 200 milliseconds, she said. “It is expensive because the metering is costly but it is being done.”
eMW’s long term objective is to manage as much of the grid’s load as possible through electric vehicles, Miftakhov said. “We are ready for the next DRAM round and we are also entering markets in at least three other territories outside California.”
The eMW charging station software can ultimately be valuable as a way to prevent curtailment of renewables, he explained. EVs parked at owners’ homes can be charged with abundant winds that blow at night. EVs parked at work after a morning commute can be recharged with excess midday solar.
One of eMW’s already targeted markets is Hawaii because renewables curtailments and relatively shorter driving distances present eMW with a potentially big opportunity, Miftakhov said.
The Texas ERCOT market is also attractive.
“Wind is an increasingly large percentage of the Texas power mix and studies show the best way to balance wind variability without increasing fossil fuel consumption “is flexible loads like what we provide,” Miftakhov said. “In California, what catalyzed our market is programs that target flexible load.”
Because of its three second dispatch capability, eMW is also working with partners to get into PJM’s frequency regulation market. It already has enough capacity in the PJM region to provide the service, Miftakhov said.
“Distributed resources will be able to participate in the CAISO frequency regulation market within the next twelve months,” he added. But “the attractive thing about the PJM market is that distributed resources can easily be integrated into it.”
Future EV buyers will get free charging stations for agreeing to participate in energy services markets, Miftakhov believes. “These resources are so valuable to the grid when they are properly valued and integrated that whatever amount of money you can make selling hardware is not significant.”
Green Charge Networks will “absolutely” do more selling to utilities for the electricity markets, Shao said. “We offer this in New York and this year we will be going into a lot more markets.”
Aggregated storage works best as a grid service in markets with high demand rates, high solar penetrations, and a high cost of energy, according to Shao. Hawaii is almost the paradigm and there are comparable niche situations on the East Coast. Green Charge is also looking at international opportunities.
“We look at market fundamentals instead of policy,” Shao said. Favorable tariffs and rebates and incentive structures for storage like the California SGIP program and the New York DMP program are obviously appealing.
But the company's focus is “the business case and the market and the underlying economics,” Shao said. “Policies and politicians can change on a whim. For the long run, we rely on the underlying economics and market forces.”
Stem is readying bigger bids for the 2017 DRAM, Ko said. “This business model, stacking values and participating in wholesale markets, DRAM or other markets, is the core for Stem. Wherever we go to deploy, we will be looking for these kinds of opportunities.”
Stem’s potential participation is “orders of magnitude” greater than the 350 kW it bid into DRAM 1, Ko said. It is already committed to 85 MW for SCE from a 2014 bidding process. “Our systems will be built to manage fleets of that size and our technology will be scaled up to that level.”
Stem already has deployments in Hawaii and was a participant in a Hawaiian Electric Companies preliminary offering that has not been completed. It is also in New York and watching other Northeastern markets.
“We are waiting for policy to develop,” Ko said. “Most ISOs and RTOs do not make it easy for aggregated storage to participate in DR yet.”
A market will work for Stem if, besides its value stream from commercial-industrial customers for peak shaving, “there are rules from ISOs and RTOs to add another value stream, in the DRAM way.” Stem is waiting for new rules on distributed storage in PJM and ERCOT.
“We expect DRAM to be a model for where DR is going,” Ko said. “We want other organized wholesale markets to evolve away from this notion that only large resources can provide grid services. It can now be done by aggregated distributed resources. The technology is there. We just need to make sure the rules are in place to properly value them.”