The next big challenge for energy storage, after bringing down the cost so that storage is economic and finding a suitable business model, is financing.
There are two ways to look at project finance.
One is that borrowing a large amount of money to build a project requires locking down costs and locking in a revenue stream so that the bank can determine how much money the sponsor will have to pay debt service. Traditional project finance involves borrowing to build a project on a “nonrecourse” basis: the lender looks to the project company that owns the project rather than the ultimate owners for repayment. Therefore, it is keenly interested in the certainty of the revenue stream and the predictability of costs. It focuses on the net amount the project company will have to pay debt service after covering costs. This allows it to calculate a debt-service-coverage ratio. That, in turn, determines how much can be borrowed. For example, if the lender requires a debt-service-coverage ratio of 1.4x, then the net revenue stream each payment period must be at least 1.4 times the required interest and principal payments on the debt.
Another way to think about project finance is it is an exercise in risk allocation. Nothing gets financed until all the risks have been identified and allocated among the parties to the transaction. A rule of thumb in the project finance market is the party that best understands the risk is the one that takes it. For example, if an asset must not be in service before the tax equity investor funds, the sponsor takes the risk because it is in the best position to know when the asset was put in service.
The first challenge with storage projects is to find a fixed revenue stream.
There are four basic business models currently for utility-scale standalone storage facilities in the United States.
One is a regulation service model. For example, a 20-megawatt battery might be connected to the grid in PJM by an independent storage company to participate in the ancillary services market. The storage company bids into the market each hour indicating it is willing to accept or deliver up to 20 megawatts of electricity that hour. The market sets the price for the regulation services by auction. Say the price is $25 a megawatt that hour. On average over the hour, the storage facility will never be near the limit. In practice, what happens is the grid might shed power to the battery for three minutes, then draw back for two minutes, then shed for 30 seconds, and so on.
The grid pays the owner of the storage facility $25 times 20 megawatts for the right to control the battery that hour.
The actual energy charged to or discharged from the battery is netted, and the battery owner or the grid also makes a payment at the end of the hour for the net electricity used that hour at the wholesale market rate for that hour. In practice, the battery owner expects to make net payments over time to the grid as power is lost during conversion from AC to DC for storage and back to AC as the electricity is returned to the grid.
Standalone batteries using this model are most common in organized electricity markets: PJM, ISO New England, New York ISO, MISO, CAISO and ERCOT.
The Federal Energy Regulatory Commission issued two orders to encourage storage. FERC Order 755 in 2011 is an attempt to create a level playing field in organized markets, like PJM, by allowing storage to compete to provide ancillary services on the same terms as power plants. FERC Order 819 in 2015 addresses storage in other parts of the country where there is no independent system operator or regional transmission organization managing the statewide or regional grid. It allows individual utilities in those areas to negotiate terms with standalone storage units without having to get prior approval for the rates from FERC. The parties merely have to let FERC know the terms on which they agreed.
Another standalone model is a tolling agreement. In the typical tolling agreement used in the power sector, the owner of gas pays a power plant a fee to convert the gas into electricity, and the gas owner takes back the electricity. In the storage market, a utility owning electricity might pay a battery owner a fee to store the electricity, and then the utility takes back the electricity. The battery owner might be paid a fixed fee per hour, like a capacity payment or reservation charge. It might be paid a fee based on the quantity of electricity stored each hour, like an energy payment. Or it might be paid a combination of the two.
There are not a lot of tolling arrangements. They are more likely to be found in deregulated markets, like California, where utilities have had to divest all their generating assets and are merely wires companies. The Southland project in southern California is an example of this model. The project was financed in late June. It involves a 100-megawatt battery in California and a 10-megawatt battery in Arizona as adjuncts to two combined-cycle gas-fired power projects with a combined capacity of 1,284 megawatts. Southern California Edison has a tolling agreement with the project company that owns the 100-megawatt battery where the project company receives a large capacity payment, in addition to a smaller variable operations and maintenance payment. Southern California Edison is responsible for supplying and paying for the energy to be charged, and has the right to charge or discharge the battery at its discretion.
Another standalone model is a buy-sell model. The battery owner buys electricity during off-peak periods when the electricity is cheap and then sells it back to the grid during peak hours.
This model is not widely used. Its main use is on a demonstration basis. The model is not considered economic currently, but it could become economic in the future as batteries and other storage technologies become more efficient. The model involves time-based arbitrage.
The last basic standalone model for utility-scale storage is where a battery is added to a wind, solar or other power plant. The battery controls the ramp rate at which the electricity is fed into the grid and puts the project in a position to earn additional revenue for ancillary services to the grid.
A battery might be added to an older fossil fuel power plant in order to give the plant the ability to respond more quickly to instructions from the grid to ramp up or down instead of having to do an expensive rebuild of the plant to comply with new, tighter response times the grid imposes on power plants that are interconnected with it.
There is also a distributed behind-the-meter model in the rooftop solar market where a large number of batteries are combined to offer storage capacity to the local utility. The solar company receives capacity and energy payments. The batteries can also provide demand services to host customers, by discharging to the customer when the customer’s onsite load will peak. Host customers pay a monthly fee for this service. The battery owner must ensure that the battery is available when called by the utility, or the battery owner will be subject to penalties. (For further discussion about business models, see “Emerging Storage Business Models” in the April 2017 NewsWire.)
The challenge for storage is the only revenue that banks will credit in deciding how much to lend is a fixed capacity payment that is locked in for a specific contract term. Merchant power plants that sell electricity into the spot market can be financed, but only with a hedge that sets a floor under the electricity price. Storage needs the equivalent of such a hedge.
CIT financed 50 megawatts of distributed behind-the-meter batteries earlier this year. It lent for the term of a contract under which Southern California Edison made capacity payments for use of the capacity.
Most of the risks in energy storage projects are not dissimilar from any other project financing. Lenders focus first on anything that might interrupt the revenue stream. They confirm that the ability to use the site is secure and that the project has all the permits required to operate. They analyze the counterparty credit on the contract that is the source of revenue to pay debt service.
The market is not settled as to whether lenders will require a fully-wrapped engineering, procurement and construction contract for most energy storage systems, or whether deals will typically be financed with separate battery supply and construction contracts. The Southland project did not have a fully-wrapped engineering, procurement and construction contract.
However, there are also regulatory, technology and operating risks that are unique to storage. The Federal Energy Regulatory Commission and regional transmission organizations are struggling with whether to classify storage as generation, transmission or a hybrid. Projects are more likely to be financed the clearer the regulatory framework.
Most lenders consider lithium-ion technology bankable and require an extended warranty from a supplier with a strong credit rating. A 10-year warranty appears to be standard for lithium-ion technologies. Lenders are less comfortable with other emerging technologies and may not be ready to lend against them without an additional performance guarantee.
The role of the asset manager is extremely important. The asset manager optimizes dispatch. Lenders will insist on an asset manager with a good track record, although this is difficult in the short term given the nascent nature of the industry. (For more analysis of risks, see “Financing Energy Storage Projects: Assessing Risks” in the June 2017 NewsWire.)
Turning to forms of financing, there are various sources of capital. The chief financial officer at a storage company would normally stack capital from cheapest to most expensive until he or she covers the full cost of the storage facility.
Government grants or subsidized debt are likely to be the cheapest. Export credit agencies may be willing to offer subsidized debt for imported storage units. If the project qualifies for federal tax credits, then it might be best to focus in the first instance on how to get value for them and then build the rest of the capital stack around tax equity.
There are three main tax equity structures. A sale-leaseback is the simplest. The storage facility is sold to a bank leasing company and leased back. This raises the full cost of the storage facility in theory, but the developer must usually prepay 15 percent to 20 percent of the rent. A sale-leaseback and be arranged up to three months after the storage unit is put in service. If the storage company wants to keep the storage facility after the lease ends, it must buy it back form the lessor.
Partnership flip transactions are more complicated structures. A tax equity investor owns the storage project in a partnership with the developer and is allocated 99 percent of the tax benefits and a share of the cash until a flip date anywhere from five to nine years out. The investor is allocated 5 percent of the economics after the flip. The developer has a “call” option to buy the remaining interest of the tax equity investor after the flip. Partnership flips raise 40 percent to 50 percent of the capital cost of a typical solar project, and 50 percent to 60 percent of the capital cost of a typical wind farm. The tax equity investor must be in the partnership before the project is placed in service.
An inverted lease is the third tax equity structure. There are two tax benefits for which a storage project qualifies potentially: a tax credit worth 30¢ per dollar of capital cost and depreciation worth 26¢. The developer keeps the depreciation and transfers the tax credit to an investor. The attraction of an inverted lease is that the tax credit can be calculated on the fair market value of the assets rather than their cost to construct, and the developer gets the assets back at the end of the lease without having to pay anything for them. An inverted leased raises the least amount of capital, in part because the tax benefits are bifurcated: the developer keeps the depreciation. The inverted lease must be in place before the assets are put in service.
The tax equity investor usually insists on being ahead of any debt in the capital structure. The rest of the capital stack is usually back-levered debt and true equity. (For more detail about tax equity structures and issues, see “Solar Tax Equity Structures” in the September 2015 NewsWire.)
Batteries qualify for a 30 percent investment tax credit at the federal level if they are considered part of the generating equipment at a solar project.
The battery should be on the project side of the step-up transformer or customer side of the inverter. It should be owned by the same legal entity that owns the solar project. It should be physically adjacent. It should work like a knob on a motor in the sense that its primary use is to regulate the ramp rate at which the solar electricity is fed into the grid. A battery at a wind farm also qualifies, but only if an investment tax credit, rather than production tax credits, will be claimed on the wind farm.
The Internal Revenue Service issued two private letter rulings confirming that batteries added to large wind farms qualify. In both cases, the projects received Treasury cash grants under section 1603 of the Obama economic stimulus program rather than claimed production tax credits.
The IRS confirmed in a separate private ruling issued to a solar company that an investment tax credit can be claimed on batteries installed as part of rooftop solar systems, but because the solar company was unable to represent that the batteries would be used primarily to store solar electricity from the rooftop systems, the IRS said a “75 percent cliff” would apply. At least 75 percent of the electricity used to charge the battery must come the first year from the solar rooftop system and whatever percentage solar charge there is the first year is the percentage of tax credit that can be claimed. For example, if the solar charge is 80 percent, then the tax credit is 80 percent x 30 percent = 24 percent. If the percentage of solar charge in any of the next four years drops below the benchmark set the first year, then all or part of the unvested investment tax credit will have to be repaid to the US Treasury. The tax credit vests ratably over five years. (For more details about the rules in this area, see “Batteries and Tax Credits” in the October 2016 NewsWire.)
The IRS is rewriting its regulations on when investment tax credits can be claimed. The issues are complicated and could take well into 2018 to resolve.
Solar projects must be under construction by December 2019 and in service by December 2023 to qualify for tax credits at the full 30 percent rate. A lower percentage tax credit may be claimed on projects that start construction in 2020 and 2021. A storage coalition has been pressing Congress to allow tax credits on standalone storage. The proposal faces an uphill climb.
A tax reform framework released by Republican leaders in Congress and the Trump administration in late September suggests Congress will allow companies to write off—or depreciate—the full cost of investments in new equipment immediately for investments made after September 27, 2017. It said this policy will remain in place for at least five years.
Another tax issue in play in Washington is the cost of interconnecting large batteries to the grid. The IRS said in June 2016 that utilities should not have to pay taxes on interconnection payments from storage projects, but there is an unresolved technical issue. (For earlier coverage, see “IRS Updates Tax Treatment of Interconnection Payments” in the August 2016 NewsWire.)
Independent generators and storage owners connecting their projects to the grid must reimburse the utility for the cost of any upgrades to a utility substation or the grid to accommodate the project. The utility will charge a tax “gross up” if it must report the reimbursement as income. It does not have to report the reimbursement as income as long as, among other things, no more than 5 percent of the expected total power flows in both directions over the intertie will be power flowing back to the generator or storage owner. This was intended to identify situations where an independent generator is a customer of the utility. Payments that utilities receive from customers must be reported by utilities as income. The test obviously does not work for standalone storage or an independent generator that has added storage. The IRS is working on updating the 2016 notice.