October 24, 2022

Federal incentives are supercharging U.S. energy storage. These 3 challenges will determine how much and how fast.

Jason Burwen, our VP of Policy and Strategy, shares his thoughts on what’s standing in the path of the over 10 GW of annual energy storage deployments we need in the U.S.–which are now possible, thanks to the storage ITC and other Inflation Reduction Act incentives.

The short version: we need to solve barriers in the supply chain, interconnection & permitting, and power system modeling. 

With the #InflationReductionAct now law in the United States, #energystorage projects have their own federal investment tax credit for the first time ever. Battery storage deployments have grown to 7 GW cumulatively prior to this moment, with more than half of that capacity installed in just the past two years. Now that the storage industry will have access to tax credits (with transferability, bonuses, and other features that will make them widely utilized), consultancies like WoodMackenzie and BloombergNEF are projecting battery storage installations in excess of 10 GW per year for the foreseeable future. It’s looking like the “100 GW by 2030” vision that my colleagues and I outlined at the U.S. Energy Storage Association back in 2018 is now within reach…

…if we can remove the barriers standing in the way. The IRA solved the question of “Should we build energy storage?” by substantially advantaging the project economics of storage versus its competitors, primarily gas-fired generation (which was already becoming unexpectedly costly due to global gas supply shortages stemming from Europeans seeking alternatives to Russian pipelines). The questions are now “How much storage should we build?” and “How fast can we build it?” The volume and pace of U.S. energy storage deployment will depend on resolving the challenges of supply chains, interconnection & permitting, and modeling.

Many companies in the battery storage industry today tell me their top three concerns are supply chain, supply chain, and supply chain.

To some extent, this is a typical business cycle story, where demand (suddenly) increases and outstrips (time-lagged growing) supply. Not only long-anticipated tax credits for energy storage, but also supportive state policies (100% clean energy mandates, storage deployment targets and incentives, etc) and continually diving costs have led to much higher expectations of storage deployments—occurring at the same time as private investors are highly capitalized and need places to park capital, particularly to meet new ESG investment goals. But there’s an atypical aspect to this demand surge—because in lithium-ion batteries, the hashtag#electricvehicle sector is already purchasing well over 10x the volume as the stationary energy storage sector, and growing. That cross-sectoral competition means that even if a storage developer can afford the price of batteries, automotive OEMs can sign contracts for much higher volumes and crowd them out. IRA incentives for battery manufacturing could augur some additional U.S. supply—particularly given bonus incentives for storage deployments that use domestic content—but nearly all of the battery manufacturing capacity announced in the past year is intended for the automotive sector. The result is that nearly all of the grid battery supply for perhaps the next 24 months should be expected to already be under contract—and the many new entrants into this space may extend that contracting farther into the future. Stories are circulating of developers making very aggressive bids to divert battery orders originally intended for other developers. Similar pain is also affecting other clean energy industries, and shared critical components are also in short supply; orders for new power transformers are on a minimum one-year delivery lag. And that’s nothing to say of workforce shortages among skilled tradespeople, particularly as IRA incentives create unprecedented demand for apprentices.

Beyond that of course is the global weirding of logistics and commodities. The COVID-induced stoppages in Chinese factories and ports show up overseas with a time lag, which itself may be compounded by unprecedented congestion at U.S. ports. Delays in commissioning projects are increasingly common, although at present seem to mostly account for several months. Meanwhile, lithium ore prices have skyrocketed and are expected to remain elevated over the medium term. It may only take 2 years to commission a battery or electrode materials manufacturing plant, but it can take 5-7 years to commission a new lithium mine (and that’s outside the U.S., where permitting timelines are much longer). Most folks only learned that Russia is a key supplier of nickel after ore prices spiked following Russia’s initiation of war against Ukraine. Sustained high prices will presumably induce innovations that increase longer-term supply—whether in new battery metals mining, refining, and recycling capabilities or in improvement of cheaper electrochemistries—but over the medium-term this pain is not expected to subside. Storage industry members were of course thrilled to celebrate IRA passage—if only because the new investment tax credits offset the near-term battery price increases they are facing, allowing them to avoid having to deal with the unpleasant reality of renegotiating contracts and reducing IRRs in their project finance models.

Behind all of this are geostrategic tensions leading to trade re-nationalization, with the U.S. showing more and more willingness to accept pain in reducing its reliance on Chinese imports. Nearly every single battery installed on the U.S. grid today came from foreign sources, with China-based suppliers accounting for an increasing share relative to their Korean and Japanese counterparts. Lithium-ion batteries already face 7.5% tariffs imposed as part of the trade war opened up under the Trump Administration. Observers anticipate stiffening tariffs on Chinese batteries may be coming from the Biden Administration, which proclaimed in an early Executive Order that batteries are one of four critical supply chains needing special attention and continues to make public announcement of interest in avoiding Chinese suppliers. And that’s all before recent higher-level attention to overseas labor concerns associated with mining battery metals; recent solar industry pain over Withhold Release Orders could similarly and suddenly upend battery imports. Suffice it to say that industry members worry that this Administration may see an electric vehicle sector largely dependent on domestically produced batteries and forget that additional tariffs will create significant pain for medium-run U.S. stationary storage development. Certainly the addition of new tariffs, on top of commodity price increases, could overwhelm the value of federal incentives intended to make the U.S. a leader in energy storage deployments.

The pace of U.S. energy storage deployment will depend on the speed of interconnection and permitting processes, which are likely to slow down before they speed up.

Prior to the IRA, developers were already lining up to connect projects to the transmission system—and now more and more companies will line up behind them. Unfortunately, transmission wires infrastructure is built far more slowly in the U.S. than generators, and the prospects looking forward are grim. Much has been written or podcasted about why it’s so damn hard to build transmission infrastructure in the U.S., but the way energy storage developers will see it is through the lens of interconnection.

At the end of 2021, about 421 GW of energy storage was in some stage of requesting transmission interconnection. While some amount is always speculative—companies getting a spot in line with the bare minimum to hold it, possibly to sell it on to someone else later—the larger share of those interconnection requests reflect genuine development efforts. The timeline for processing those requests has lengthened significantly, to the point where developers entering the queue now must wait 4 years on average, often more, before they will have any chance of connecting their projects to the bulk electric system. And that’s not counting delays as regional electricity market operators reform their interconnection rules and processes—well-intentioned efforts that will likely mean slowing down through the middle of the decade before speeding up again. The mid-Atlantic grid operator PJM made headlines when it announced that its (good!) effort to reform interconnection to a parallel rather than serial process would require postponing any new interconnection requests from being processed before 2026, let alone finishing the already queued requests. It doesn’t help that the RTOs/ISOs are losing talent to many of the clean energy companies that are lining up, just at the time when their expertise is most needed. Indeed, the war for talent in the clean energy industries is real, and an unintended byproduct of storage and other industries’ ambitions from the IRA is a steady brain drain from the many less flashy or lower-resourced institutions that serve as gatekeepers for clean energy deployment. Storage industry members are increasingly wary of efforts at “jumping the queue,” such as might be interpreted in storage-as-transmission or utility-owned storage sited at distribution. It’s also leading numerous developers to themselves seek interconnection at distribution voltages, seeking smaller project sizes in exchange for shorter interconnection processes with distribution utilities—although this has resulted in running battles across numerous states over the right fee structures for storage providing wholesale service while connecting to retail infrastructure. And of course customer-sited storage developers will find that, while distribution interconnection may not take as long as transmission interconnection, there is no open access principle animating distribution interconnection the way it animates transmission infrastructure in the RTO/ISO territories that cover two-thirds of the U.S.—every state and utility territory within that state is its own world.

Separately from interconnection, the geographic siting of energy storage is subject to permitting processes of a totally different flavor. Typically, local authorities have jurisdiction, and across the country these “AHJs” are being asked to issue permits for energy storage facilities for the first time. Most AHJs have yet to adopt the safety codes & standards associated with stationary energy storage installations; indeed, while codes & standards have been updated recently and are ready for adoption, state and local authorities often lag behind codes development bodies by many years. Newness is rarely beneficial in permitting, and storage is no exception; there has been a noticeable uptick in localities across the US pushing back against siting of storage over concerns regarding safety, construction, and other issues—even including local moratoria. While opposition to energy infrastructure siting is not new, nevertheless the stampede of storage developers looking for optimal sites (especially for easier transmission access) means city councils and fire departments are being asked to give more attention to storage than ever before. Lack of familiarity and training could create a parallel dynamic of local permitting slowing down before speeding up later this decade.

Power system models will decide if energy storage or other resources win the race to back up massive amounts of variable wind & solar power generation.

Models are not reality, but power system models strongly influence what electric supply resources get realized. More to the point, the unprecedented levels of new wind and solar power projects that will come online are driving debates over what is needed to keep the electric system reliable and affordable—often times under explicit requirements in states or regions to fully decarbonize by certain dates and meet certain milestones. Whether in organized electricity markets or regulated utility territories, grid operators and state regulators will turn to modelers to answer the question “What can we rely on and how quickly can we reduce reliance on hydrocarbons?”

In 36 states, utilities are responsible for developing an integrated resource plan that identifies long-term system needs and what resources it will build or buy service from to meet those needs. While in many cases these IRPs are something of a box-checking exercise, increasingly state regulators and stakeholders see them as in fact the most central place where decisions about clean energy transition are made. I’m fond of saying “IRP is the new RPS” because the speed and volume at which utilities plan to build wind, solar, and storage is increasingly being determined in these planning efforts, particularly in states without strong statutory mandates to procure renewables. Wind and solar are now the cheapest sources of electrons, which has led to numerous utilities planning multiple GW of procurement in just the next decade. As a result, IRPs are increasingly the domain where modeling works out what the optimal options will be for complementing the variable output of wind and solar and maintain reliability, primarily to replace old power plants slated for retirement. Heretofore utilities have typically relied on their IRP models to demonstrate that gas-fired power resources will be the primary resource for complementing renewables. Those IRPs do not, however, capture either the recent and prospective long-term declines in energy storage costs driven by IRA nor the elevated price of natural gas over the medium term. Moreover, many IRPs still fail to abide by the best practices for how to accurately model both the need for flexibility in power systems and the operations of energy storage, all the more critical with high levels of wind and solar to come. It boggles my mind that the planning of billions of dollars of investment decisions is still done on the basis of models that extrapolate conclusions from just a few representative hours in each year. And distributed resources, including behind-the-meter storage, still remain an afterthought in most IRPs, even as virtual power plants are demonstrating that customer-sited resources can be effectively offered to utilities as supply resources. If these 10-20 year views into the future indicate only modest needs for energy storage, sales pitches to utilities about the attractive price tag of storage with IRA incentives may fail to get traction.

Meanwhile, every RTO/ISO regulated by the Federal Energy Regulatory Commission is responsible for accounting for the system reliability of their entire territories, ranging in size from the single-state system of California ISO to the Midcontinent ISO’s territory over some or all of 15 states. While states vary in their role in approving the resources procured for system reliability—with eastern states largely having delegated that authority to their RTO/ISO and central states largely having retained that role for their state regulators—each RTO/ISO sets the level of reserves needed to keep their whole region reliable. Since wind, solar, gas, coal, nuclear, hydropower, and energy storage all present different operating capabilities and constraints, it is up to RTOs/ISOs to determine how much each type of resource contributes to system reliability—which then informs how much to discount the rated power of each resource when adding them up to meet the target level of reserves. All RTOs/ISOs are either adopting or have adopted new modeling approaches—notably Effective Load Carrying Capability analysis—to calculate what energy storage resources (and renewables and hybrid storage-plus-renewables facilities) contribute to system reliability. A benefit of these new models is that they can capture interactions among the fleet—discounting the reliability value resources to the extent their availability and output is correlated, and enhancing the reliability value of resources to the extent that their output is complementary; indeed, on the whole these models find that, under level of higher wind and solar production, current-day energy storage remains valuable for reliability. However, at present these models are used only to calculate the reliability value of storage and renewables; as such, no such discounts are applied to fuel-based power generation. Winter Storm Uri in Texas in 2021 was a catastrophic demonstration of the correlation of outages among gas-fired generators under extreme weather conditions, and it is likely that the absence of that correlation from these planning models overcounts how much reliability RTOs/ISOs should expect from those resource when they account for significant parts of the supply mix. Other modeling choices can also shift the calculations significantly for storage; for example, if the reliability contribution of every storage resource is valued according to the marginal unit, ELCC models can discount the value of storage close to zero after even modest deployments. Energy storage and fuel-based generation are mostly competitors when it comes to providing system reliability. The modeling choices of RTOs/ISOs will effectively determine what amount of storage gets procured—even in states that have clean energy goals and might otherwise want storage as a source of reliability to also lower greenhouse gas emissions. IRA incentives for storage won’t matter very much if RTOs/ISOs tell states and markets that storage isn’t valuable.

So how much storage can the U.S. build, and how fast?

These challenges—supply chains, interconnection & permitting, and modeling—are not the only determinants of the pace and volume of U.S. energy storage deployment post-IRA enactment. The macroeconomic environment is…uncertain, to say the least (although the volume of private equity piling into storage may insulate the sector). Certainly how the IRA itself is implemented could have meaningful impacts on how easy or hard it is to avail the incentives for storage. Moreover, it remains to be seen how the incentives for green hydrogen and other resources in the IRA affect the relative price competition of storage with other resources. There are a panoply of electricity market design changes underway that will affect the prices that merchant storage project economics are based on. And of course there may be upsides we have yet to see that come with the determined shift in the U.S. government to undertake robust industrial policy. States will likely revisit their plans and tools for attaining clean energy and related policy goals, including with storage. It is always hard to peer around the corner very far—and I very much welcome folks in the comments to point out major considerations you think I’ve missed in this discussion.

Any way you cut it, the U.S. energy storage industry is in an incredible moment. As someone who has spent the last 7 years strategizing how to accelerate market development for the U.S. energy storage industry by scraping and scrapping in state legislatures, regulatory agencies, and market stakeholder processes, enactment of IRA feels like we’ve finally pushed the boulder to the top of the mountain, and now it is beginning to roll downhill. The momentum of deployment and the supercharged incentives of IRA have created their own challenges—but they are very much “good problems” to have, compared to those of years past. Decarbonizing the U.S. power system will require 150-350 GW of energy storage by 2035. We have no time to lose in tackling these challenges and clearing the path ahead for energy storage project development on a scale we’ve only imagined. Onward.