The newest numbers on U.S. energy storage growth show the emerging business is still growing at breakneck speed.
Compared to conventional generation, storage's capacity numbers are still small, but the upward trend is undeniable, and utility interest is palpable. Installed capacity remains on track to triple in 2015. At that rate, the 2019 annual market will be 13 times bigger than the 2014 market and four times bigger than the 2015 market, according to the recently released report, "Q1 2015 U.S. Energy Storage Monitor" from GTM Research and the Energy Storage Association (ESA).
“This year will be the biggest yet for storage,” said GTM Research Senior Analyst and lead author Ravi Manghani. “Through 2014, there was perhaps one or two big projects per year but this year there will a handful of projects getting interconnected,” he explained. “The trend is clear. This is a market on a growth trajectory.”
First quarter numbers are often unimpressive because companies tend to put more effort into planning than building in the first three months of the year. But the U.S. deployed 5.8 MW of storage in Q1 2015, 16% more than Q1 2014. Utility-scale, in-front-of-the-meter storage, the sector’s biggest component, added 4.2 MW in six projects. Behind-the-meter (BTM) storage had its biggest Q1 ever, with 1.6 MW of capacity installed, a jump of 132% from Q1 2014.
All the storage sector’s market segments had strong quarters relative to preceding Q1s. It was the third-biggest quarter ever for residential storage and the non-residential market had its second-biggest quarter.
“There won’t be a lot of diversity in the market this year,” Manghani said. It is “largely driven by a couple of geographies and a few applications in those geographies.”
“California is still in the early stages of what will be the biggest U.S. storage market in the next two years because of the mandate requiring its three IOUs to deploy 1,325 MW of energy storage by 2020,” Manghani said.
The mandate allows the utilities to choose the applications “but we are only just beginning to see what they will decide about the best use cases for storage in their territories.”
Storage by the numbers
Only about 11% of the U.S. market will be distributed, BTM storage in 2015. The other 89%, about 220 MW, will be utility-scale storage, Manghani said. That is more than three times 2014’s 64 MW.
“Those 220 MW will be a few systems in the PJM Interconnection ancillary services market,” he added. An example is the 31.5 MW system brought online by Invenergy in May, which represents almost 15% of the expected 2015 market.
California and PJM are the market leaders. Together, they have 99 MW of the 115 MW deployed since Q1 2013. PJM dominates that period with 68.7 MW of utility-scale storage. California has installed 21.2 MW of utility-scale capacity in that time. California dominates both the residential (1.0 MW) and non-residential (6.9 MW) markets. PJM has no residential storage and only 0.8 MW of non-residential storage at commercial and industrial facilities.
After PJM’s cumulative 69.4 MW and California’s cumulative 29.2 MW, all other markets have only 7.8 MW of energy storage.
The price of energy storage is generally coming down from the weighted average 2014 system price of $2,064 per kilowatt, Manghani found.
For lithium ion battery systems with 1 hour to 2 hour discharge durations and no special interconnection requirements, and based on limited (because it is proprietary) data, the median utility-scale price is presently about $900 per kWh, the median non-residential price is about $1100 per kWh, and the median residential price is about $1,500 per kWh.
But those median values don't represent the cutting edge of storage technolgies, industry advocates say. At the Edison Electric Institute's annual conference in June, Steve Hellman, Chairman of Eos, a utility scale energy storage provider, told attendees that his company's zinc hybrid batteries cost $160 per kWh, lasts about 15 years and is 75%-85% efficient. That battery is available in 1 MW or 4 MW sizes.
Higher deployment volumes are also driving down battery-pack costs, GTM found, including the price for batteries, wiring, racking, and battery management systems. Better power conversion systems are lowering the cost of grid integration and helping cut balance-of-system costs.
Growth from services to grid operators
“Energy storage is such a big deal that in the electric utility business we called it the holy grail,” explained former Southern California Edison Senior Vice President Jim Kelly recently. “We knew that if we tried to get higher amounts of wind and solar on the grid, eventually it would break unless we had energy storage.”
Federal Energy Regulatory Commission (FERC) orders 755 and 784 changed that, Manghani said. In 755, grid operators were mandated to consider energy storage as a grid services option. In 784, they were ordered to find ways to monetize fast-response frequency regulation services provided by energy storage systems.
“Leadership matters. PJM, the biggest system in the world, was the first to enact Order 784 effectively,” Manghani said. “The other independent system operators (ISOs) and regional transmission operators (RTOs) tweaked their market rules. But PJM made sure both the signal, the operational element, and the capacity multiplier, the compensation element, are in line with what a storage system needs and can deliver.”
The key is the type of signal the grid operator dispatches, Manghani explained. Because of efficiency losses a storage system discharges slightly less energy than the energy used to charge it, while other generating units aren't subject to the same costraints. Because operating a grid is about balancing, the operator has to get precisely the amount of frequency regulation expected and absolutely no less than what is needed.
“If the grid operator sends the signal and gets less than expected, there are issues with up and down time and charge-discharge cycle time and cycles and battery life and a variety of issues,” Manghani said.
FERC and the system operators have acknowledged storage is not a traditional generation asset, he explained. “The efficiency loss is well understood. But the storage system has to respond to the signal almost instantaneously. An improperly tweaked signal can lead to poor performance. That is one area where PJM has done better than others.”
But in Q1 2015, Manghani said, the New England system operator (ISO NE) and the Southwest Power Pool (SPP) have implemented compensation and market mechanisms similar to PJM’s.
“We expect growth for storage in ancillary services to these two systems soon," he said. "It may not be as big as PJM because that system offers an order of magnitude more opportunity than any other. But both those systems changing their rules indicates we are moving toward more storage for ancillary services.”
And, he added, “We are beginning to see rules being written in a way so that other storage applications can participate.”
With over 1.6 MW of BTM capacity added in Q1, it is clear that market is also growing as expected. Though much of it is used on-site, BTM storage can also be sold into wholesale ancillary services markets and capacity markets.
The Solar Grid Storage installations just acquired by SunEdison serve renewables projects as backup storage but also sell frequency regulation services to PJM.
Pilot deployments by PG&E, SCE, and SDG&E in California, by HECO in Hawaii, by ERCOT in Texas, by S&C and Viridity in Illinois, and by Direct Energy in Maryland are testing the newest technology to sell aggregated BTM storage system capacity into wholesale ancillary services markets.
It is still very early in the use of aggregated BTM storage, Manghani said. The technology needs to get better, the right signals are needed, and they need to be optimized.
“This new idea is working but there have been and will be learning opportunities,” he explained. “There will be more pilots, they will become actual market opportunities, and individual projects will become parts of fleets. But that will take work on the technology, the market rules, and the business models.”
Limits to Growth
Each application and market has its own limits and there are obvious and well known limits to battery storage technologies such as cycle life and lifetime and efficiency losses and cost and recycling and safety, among others, Manghani said.
“They are constantly changing and there will be linear and non-linear shifts and who knows where the market will be when any one of those limitations is mitigated? There are a million possibilities.”
But a specific limitation to frequency regulation, the market’s biggest driver, is becoming clear. At some volume of available capacity, the multiplier paid for fast-responding services that makes it a cost-effective choice for developers would become unnecessary, Manghani explained.
“As you get more fast-responding systems, the multiplier would need to shrink to send a signal to decrease the supply. There is a point where there is no more need for fast-responding services.” If the multiplier is accurately handled, he acknowledged, there would be no market driver.
Tesla Motors sent the electric sector into a tizzy on May 1 with the annoucement of two new batteries — one for residential use and another scalable option for commercial, industrial, and utility storage facilities. But while it attracted a lot of attention, some utilites are still cool on the company's technology.
“The Tesla 7 kWh battery costs $7,000 installed and is meant to be cycled daily,” Tucson Electric Power’s Carmine Tilghman recently told Utility Dive. "At the average all-in electricity rate of $0.12 per kWh, the buyer saves $0.84 per day. And $7,000 divided by $0.84 per day means it will take 8,333 days or about 22.9 years to get the initial investment back for a battery that comes with a ten year warranty.”
The Tesla battery’s 2 kWh discharge cycle provides 3.5 hours of discharge, which is not enough electricity to run many appliances, Tilghman added. “And Tesla’s 100 kWh commercial system costs $25,000 plus about $25,000 to install. Will a residential customer spend $50,000 to get off the grid and save a $150 per month electric bill?”
“Those numbers seem accurate,” Manghani said. But the Tesla announcement did not change the GTM Research-ESA forecast through 2019 “in any measureable way,” he added.
But Tesla officials would likely take issue with those assessments.
JD Straubel, CTO of Tesla said at the EEI convention that his business is targeting $100 per kWh for lithium ion energy storage batteries by the end of the decade. Much of the price reduction will be driven by the increasing scale of manufacturing facilities, Straubel and Tesla CEO Elon Musk told the attendees. At a later shareholder meeting, Musk indicated that the company's gigafactory will begin production in mid 2016. The company expects to produce 50 GWh worth of battery packs annually by 2020.
Additionally, Musk told the conference that much of the media has interpreted the value proposition for its residential batteries incorrectly. In part due to some statements from SolarCity that he called "not correct," many outlets have interpreted the battery's intitial use to be to back up residential solar arrays or to power households during peak hours to take stress off the grid. That may be the case one day, Musk said, but currently Tesla is marketing the home batteries as an option to keep the lights on for a limited amount of time during power interruptions.
But even if the Tesla batteries have yet to make a big splash with utilities across the country, Maghani says the attention its launch attracted was significant.
“It made storage a more mainstream topic. But the things that need to happen for growth are going to happen. Tesla’s announcement did not accelerate growth. And it did not change the fact that battery storage will provide services to customers, utilities, and system operators for a more efficient and cleaner grid operation.”