Primer: The now and future impacts of energy storage
With renewables growth comes the need to move variable generation to when customers use power most
Grid operators and forward-thinking utilities are already getting a glimpse of what energy storage can do – but they ain’t seen nothin’ yet.
Multi-megawatt energy storage installations are now providing services and finding value propositions and use case scenarios around the world and at all levels of the grid, according to Energy Storage Association (ESA) Executive Director Matt Roberts. And there are gigawatts of projects already on planners’ horizons.
“Wholesale markets are leading the rush for storage,” Roberts said. And the New York ISO, MISO, SPP, ISO New England, ERCOT, and the California ISO already have or are in the process of instituting “market structures that value the services energy storage provides."
Cost estimates vary, but all show storage rapidly becoming more affordable, he said, citing four sources. A Navigant study predicted a 4 hour battery storage system would be as low as $700 per kWh by 2020 while Oncor foresees a $350 per kWh cost in 2020. Morgan Stanley estimates battery-only costs will eventually approach $125 per kWh and Tesla has said its lithium-ion battery-only cost is already $110 per kWh.
The battery cells make up around 30% to 40% of a battery system and the rest is in the cost of the balance of system (BOS) hardware like inverters, switches, control systems and power electronics, Roberts said. “Those costs continue to decline, not just because of the energy storage industry but because a lot of those technologies are shared across the renewables and energy industries.”
An ESA-calculated 5.8% drop in BOS costs by 2030 takes the system installed cost to around $400 per kW, he added.
Though the incipient industry’s numbers are still relatively small, it is seeing record-breaking installation numbers quarter after quarter. Grid-scale storage has been the driver but behind-the-meter additions are growing, Roberts said. Between 220 MW and 240MW will be built in 2015 but “we will have 800 MW to 900 MW go operational in 2019.”
With this increase in storage, a new paradigm is emerging for balancing generation and load because “energy storage falls on both sides of the equation,” explained Black & Veatch Management Consulting Division Manager Benson Joe. “This leads to the question of how energy storage will be used.”
There are as many as 30 to 40 different applications for energy storage on the grid “but there are multiple applications within each segment that storage can provide value in,” Roberts said.
End users are looking for technologies that allow them to manage their energy consumption, he explained. That is driving behind-the-meter technologies.
In the general category of flexible resources, reserve and response services require faster technologies. Supercapacitors, flywheels, and many of the electro-chemistries, particularly the lithium ion chemistry, can presently serve the need for “really fast ancillary services like frequency regulation.”
Transmission and distribution grid support can come from multiple chemistries, some flywheel technologies, and eventually flow batteries that can provide “longer duration storage in the one hour to four hour range,” Roberts said.
For bulk power management, there are massive scale storage technologies already in the field, like pumped hydro and compressed air, that provide “durations beyond six or eight hours,” he said.
Because storage is application-specific, different markets are driving different technologies. “In the ISO markets, storage is supplying capacity, regulation, supplemental reserves, voltage regulation, and black start,” Roberts explained. “For IOU rate recovery, it will supply services like transmission and distribution system upgrade deferral, congestion relief, and voltage support.”
“Smaller scale, customer sited storage can provide power quality, reliability, or demand charge management to end users, he added.
Flexibility and capacity
Energy storage is already succeeding in the marketplace as a flexibility resource because it has a very fast response time.
“Some technologies respond in milliseconds but all are well under a minute,” Roberts said.
Storage has a much wider flexible range than the natural gas peaker plants now serving grids’ flexible services needs. “There is no minimum,” he said. “It can provide a kilowatt or a megawatt,” depending on battery size.
“It provides exactly how much energy is needed exactly when it is needed and a traditional gas peaker does not need to be idling, waiting to be called on. That saves emissions,” Roberts added.
Storage as a capacity resource is likely to be an increasingly important market factor because it addresses the burden on the electrical system of peaking demand, he went on.
“According to the U.S. Energy Information Administration, the entire transmission and distribution system is about 40% to 50% over-built,” Roberts said. There is generation, distribution, and transmission infrastructure to meet demand peaks that only occur a few hours each year.
Energy storage can limit further over-building because it “enables arbitrage,” Roberts explained. Energy can be generated whenever and wherever it is most cost effective and used during high demand periods or an outage, when supplying power would otherwise be most expensive and emissions-intensive.
Stored blocks of energy can be “ready to go on demand,” he said. That also turns energy storage into load to capture otherwise curtailed or spilled renewables, he added.
“Being able to respond to grid signals flexibly and rapidly is critical,” Roberts said. “Storage can absorb energy when that is what is needed and can be a valuable place to put energy where there are high penetrations of wind or solar.”
Flexibility and capacity, a case study
With its newly-imposed 50% renewables by 2030 mandate, California will need a lot of flexible resources. Energy storage as a capacity resource will help fill that need, said Black & Veach's Joe, co-author of the “Impact Of High Solar And Energy Storage Levels On Wholesale Power Markets” from B&V and the Solar Electric Power Association (SEPA).
B&V expects solar PV costs will drop about 50% in the next 5 years to 7 years across all segments of the market, driving unprecedented growth, Joe said. By 2030, solar will significantly displace natural gas generation during the middle of the day, exacerbating “depressed electricity prices and the Duck Curve effect.”
The California Independent System Operator (the ISO) will have to work to integrate solar with minimum impact to its wholesale electricity market while ensuring grid stability, Joe said. The solutions it reaches will help grid operators across the country who face increasing penetrations of variable renewables after 2020, as solar and wind reach grid parity more widely.
B&V and SEPA ran four scenarios for California in 2030 based on a scenario with natural gas for all new generation, a 33% by 2030 renewables mandate, a 50% renewables mandate, or a 50% mandate combined with 10 GW of energy storage.
Their most consistent insight, whether looking at a 24-hour cycle or a seasonal cycle, is that a boosted renewables mandate in California means increased solar over-generation near mid-day, as well as the need for ramping between 4 p.m. and 7 p.m., at the load peak.
“That two hour to four hour period is where energy storage can be of value to shift some of the energy at the solar peak to the system peak,” Joe said. “There will be opportunities to export solar but neighboring markets will likely be long on solar, too, so a lot of solar will be curtailed unless there is energy storage on the system.”
With a 33% RPS, “there is a duck belly,” Joe said. The duck’s belly is the result of an excessive supply of solar at mid-day that causes the electricity price to plummet. It sets up an excessive need to ramp generation upward in the late afternoon-early evening peak demand period. The will drive electricity prices excessively high.
“With high solar meeting a 50% RPS, the duck’s belly gets bigger,” Joe added. “But with a lot of storage, the load curve is flattened by using storage to shift solar to when it is needed.”
With the solar penetration needed to meet the 50% requirement, 25,000 MW of storage or flexible capacity would be needed for peak demand ramping, the paper’s calculations concluded. “But 10,000 MW of storage flattens the net load shape and helps bring prices up during the times when solar is generating and helps reduce the super peak prices.”
The analysts expect some impacts on California to be seasonal.
“Solar is more problematic during the winter, spring, and sometimes the fall. Prices get down to $10 per MWh in January,” Joe said. “In the summer it is not as much of an issue, because load is bigger.”
Because of its ability to use more solar, energy storage also reduces curtailment, making renewables more economically viable, the B&V modeling shows.
“With storage in the mix, curtailment is reduced quite a bit throughout the year,” Joe said. “With lots of solar, there is curtailment during the middle of the day but storage significantly reduces it.”
Needed: regulatory reform
In further modeling, B&V found the effect of solar and storage on load was very much dependent on policy questions like the availability of incentives and time-of-use rates, Joe said.
Roberts agreed. “Figuring out regulatory structures that allow energy storage to be remunerated for the value it provides is the marketplace challenge it now faces,” he said. Because it can be both load and generation, it doesn’t fit neatly into asset classifications and or tariff systems.
At the federal level, FERC, Congress, and EPA regulations have had and will continue to have impacts, Roberts said. FERC rules, especially its Pay for Performance initiative, have been important.
EPA environmental regulations are expected to drive markets because emissions from natural gas peaker turbines will be more expensive. “In its Clean Power Plan, the EPA concluded storage itself does not reduce emissions but it enables emissions reductions,” Roberts explained.
There is little likelihood of Congressional action on storage in the foreseeable future but state policies are growing regional markets. “There are different drivers in different states,” he said.
California’s concerns with over-generation of renewables led to its landmark storage mandate. The ravages of Superstorm Sandy left states in the Northeast committed to greater system resilience. Hawaii’s high penetration of distributed generation is driving the pursuit of a more reliable system. All of them will have lessons for the rest of the country and its power companies.
Utilities, especially with the growth of solar and other distributed energy resources, have many ways to impact the load profile, so utility planning will also affect the value of storage, Joe said.
“Understanding where there might be overloading, both at the distribution and transmission system levels, and how energy storage can relieve it, will be a big part of planning in the future.”