When it comes to energy storage systems, the cost of the battery itself receives a great deal of attention, but a new report shows that the plummeting price of other storage components could prove just as significant for the proliferation of the resource.
The 15 to 20 different items in the balance of the system (BOS) represent one-half to three-fourths of the installation’s cost. Those costs are likely to plummet over the next five years – if the growth of storage deployment continues.
“If storage can keep up its momentum there is reason to believe BOS costs will fall from their 2015 level of $670/kW to under $400/kW by 2020, a 41% cost drop,” said Luis Ortiz, lead author of a recent report from GTM Research: "Grid-Scale Energy Storage Balance of Systems 2015-2020: Architectures, Costs and Players."
The researchers assumed $350/kWh for the battery cost, an aggressive number based on the performance of industry leaders Samsung and LG and the Panasonic-Tesla product. For an installation intended to supply energy with a two-hour duration battery, BOS would be about half the total project cost, Ortiz explained.
But for a project intended to supply power with a half-hour duration battery, BOS is almost three-fourths of the installation’s cost, he said.
Continued growth of battery storage to drive this drop in the BOS cost is likely, according to the Q3 2015 U.S. Energy Storage Monitor from the Energy Storage Association and GTM Research. The 60.3 MW deployed in Q3 2015 was a twofold increase from Q3 2014 and a 46% increase from Q2 2015.
2015’s total deployment of 192 MW will more than triple 2014's 62 MW. By 2020, annual deployment is expected to be up to 1,349 MW, which is “21 times the size of the 2014 market, and seven times the size of 2015 market,” the monitor reports.
The BOS study is intended to add dimension to the battery energy storage cost discussion “to make it possible to figure out how financeable a project has become by including everything that figures into consideration of a project when capital is involved,” Ortiz said.
The researchers did two analyses. One was an item by item cost projection through 2020 of the 15 to 20 discreet factors beyond the battery in an energy storage system. The other was a learning curve analysis, based on the solar PV industry’s BOS cost decline from 2009 to 2015, that was applied to BOS for battery energy storage through 2020.
The two analyses produced “very similar numbers” which gives Ortiz "good reason to believe storage BOS will hit the cost declines we expect,” he said.
Is solar PV growth a good template for storage growth?
Even MacDonald's uses the learning curve analysis, according to Ortiz. “The cost of each hamburger is reduced by each sales volume increment, and that happened in solar BOS costs,” he said.
The solar PV BOS cost was halved with each increase of installed MW by a factor of ten, and “the learning curve dynamics of percentage declines should hold true for storage,” according to the BOS analysis. That cost will also benefit from the “established ecosystem of best practices,” it reports. “The learning curve dynamics of percentage declines should hold true."
Storage will piggyback on the solar experience because storage BOS is similar and they are often co-located, Ortiz added.
Though the bidirectional inverters necessary for storage are more expensive than PV inverters, the technology is familiar to solar developers and the price difference is expected to compress somewhat over the next five years.
“The value chain and vendor ecosystem for storage projects is ripe for new entrants,” the BOS analysis reports.
The giant push in research and development from the Department of Energy’s SunShot program helped drive down the solar PV BOS costs, but the biggest driver was “the sheer volume of gigawatts deployed,” Ortiz said. “It gave solar companies the opportunity to buy for scale and design for scale and one-off installations were standardized and that cut costs more."
The boom in production that shook the solar industry between 2011 and 2013 — leading to rampant consolidation and driving down solar prices — will not necessarily be part of what drives the decline in battery and storage system BOS costs, Ortiz said.
That industry consolidation drove PV module prices down, but it is not clear that will happen in grid-scale storage supply, Ortiz said. Grid scale battery storage would likely not be the driver of consolidation in lithium ion (Li-ion) manufacturing.
“Automotive battery storage is most likely to predict if there will be consolidation," he said.
Ortiz expects demand for Li-ion grid-scale storage systems to remain strong because solar developers “will start looking at storage projects as an opportunity to improve their margins.”
To sustain deployment beyond 2020, a new battery technology will be needed because Li-ion will “bottom out on cost reductions,” Ortiz said. “A different battery chemistry for grid scale storage will be needed to hit the cost levels for continuous deployment.”
At that point, Ortiz posited that there could be a consolidation of Li-ion battery manufacturers as players who have learned how to supply grid scale storage buy up the new technologies. By then, new liquid metal and flow battery technologies that have achieved the necessary cost-competitive, 20-year life performance will be needed.
From 2020 to 2025, according to a recent Lux Research study that supports this Ortiz hypothesis, the technologies to watch are Aquion Energy’s aqueous sodium and lithium ion electrolyte chemistry, Ambri’s liquid-metal design, and Eos Energy Storage’s carbon-based cathode, zinc anode, and aqueous electrolyte chemistry.
In the interim, the numbers are pointing to sustained adoption and, if that deployment growth continues, BOS costs will come down, Ortiz said. “The driver for that deployment will be the usefulness adopters find in storage when they begin using it.”
Soft costs set to drop
The biggest single target for BOS cost reduction is in hardware like inverters but the next biggest potential is in soft cost reduction, the analysis found.
Because data is limited, it is only just beginning to be clear how big the storage system soft costs are and how much they can be cut, Ortiz said.
The storage industry has not tracked its soft cost numbers adequately due to “limited incentives to share openly and due to imprecise tracking,” the analysis reports.
The industry’s limited deployment means the allocation of things like a developer’s overhead costs is likely to be made to a single installation or across a relatively small project volume. With increased deployments, almost the same overhead costs would be different because they would be divided among more installations.
There is also much uncertainty about the costs of interconnection, Ortiz said. It is still an incompletely defined process for many grid operators and utilities. As deployment volume increases, system operators and regulators “will streamline the process just as they did with solar PV and that should bring that soft cost down,” he added.
If standardization around the integration of storage into the grid does not follow a pattern similar to the learning curve for solar interconnection, it could result in the stunting of storage adoption, Ortiz said.
Without clarity on interconnection procedures and the rules for getting a project online, financiers would lose confidence and the vital early capital necessary to keep a project going until it generates returns would become less available, Ortiz said. But that is unlikely, he stressed, because there are efforts already ongoing to standardize aspects of storage and overcome that barrier.
What utilities need to know
Utilities are likely to be key to driving deployment as purchasers, Ortiz believes. California utilities, pushed by a 1,325 MW state mandate, are already having positive impacts on costs.
The great diversity of utilities across the country will display a diversity of reactions but they will slowly discover “the storage going into the ground now is a more robust and affordable technology than the storage they have seen before,” Ortiz said.
“Though not homogenously, they will one-by-one embrace storage and drive megawatt deployment when they find it can do things they want to do but haven’t been able to do.”
What will most attract utilities is getting the service of several assets from the purchase of a single asset, Ortiz said.
Normally, he explained, a utility would likely have to buy a separate dedicated asset to provide reactive power at a substation, another for frequency and local control management, another for load balancing, and another for managing transformer cycling. “Storage can do all those things so when it buys a battery to help with reactive power, the utility gets the other services sort of ‘for free.’”
This is corroborated in a recent Rocky Mountain Institute study that found 37 different services, including 13 fundamental services, storage can provide to customers, utilities, and system operators.
The maximum value of storage comes when it can provide more of its full “stack” of potential services, co-author Garrett Fitzgerald told Utility Dive.
“Batteries deployed for only a single primary service generally do not provide a net economic benefit (i.e., the present value of lifetime revenue does not exceed the present value of lifetime costs),” the RMI research concluded. But stacking on other services “shifts the economics in favor of storage.”
This is more than just separate value streams, Ortiz said. As storage assets get deployed for a primary purpose, their multiple uses will quickly be seized upon, “especially as more DERs are deployed,” he explained.
Storage at a substation that can dynamically and automatically control co-located DERs becomes “almost a magic elixir,” he said.
Utilities may only inch their way toward this realization but eventually some will begin planning for storage’s multiple uses and others will see the success and follow, Ortiz believes.
“Storage will find markets and applications all over the country,” he said. “Batteries will get bigger and run for longer durations and BOS costs will continue to come down as utilities and other customers discover it can do things besides what it was purchased and deployed for.”