Below are answers to the questions submitted during our webinar on April 23. The questions have been edited for readability.
I thank my co-presenters Ken Elser of DNV GL and Gintaras Sadauskas and Vadim Ovchinnikov of Alfa Energy Advisors and my colleague Deanne Barrow for their assistance in preparing these answers.
The questions are organized by the following topics: storage, the IRS’s start of construction rules, the capital stack, fuel cells, engineering and PG&E’s bankruptcy.
The presentation from the webinar is available here.
Question: What’s the latest in financing for standalone storage and storage combined with solar?
Answer: We have started to see some project finance transactions executed with respect to standalone storage projects and combined solar and storage projects in select markets (e.g., Hawaii, Massachusetts, California) where the state’s regulatory policy makes storage services valuable. Many utilities are now routinely including storage in their requests for proposal.
Projects that can lock in a fixed revenue stream under a long-term offtake agreement with a creditworthy utility are attractive to financiers. However, capital costs for batteries are still relatively high and regulatory, technical and operational risks unique to these assets have hindered the pace of financing and investment.
Question: Have many utility scale storage or storage combined with solar project finance transactions have been executed?
Answer: We have seen just a handful of large-scale standalone and combined solar and storage projects financed to date. The largest utility-scale storage financing remains the AES Southland project that closed in 2017 (for more on that financing, see our blog post here).
Nonetheless, we anticipate the number of financings for solar combined with storage will grow rapidly due decreases in the cost of equipment, clearer policies and design improvements gleaned from the operating experience of combined solar and storage projects, particularly those that are DC-coupled. Further, the rise of the combined solar and storage power purchase agreement (PPA), which specify the manner in which storage is used and priced, should lead to more project financings. Such PPAs (i) set limits typically based on total cycles (or throughput) available on an annual basis and (ii) accommodate the investment tax credit (ITC) charging requirements, which were discussed in our presentation, during the five-year ITC recapture period.
Question: Is there a general or preferred structure with respect to financing storage?
Answer: We see three basic structures currently for utility-scale storage. One is a regulation service model used in organized markets, such as PJM, in which each battery bids into the market each hour indicating it is willing to accept or deliver up to its capacity in electricity that hour.
Another model is a tolling agreement. Under this structure, the project receives a large capacity payment and a small, variable operations and maintenance payment. As would be the case with a conventional power tolling agreement, the utility-offtaker is responsible for supplying and paying for the input, in this case, charging electricity. This model is more likely to be found in deregulated markets, like California.
The third basic model is where a battery is added to a wind, solar or other power generation plant. The battery controls the ramp rate at which electricity is fed into the grid and enables the project to earn additional revenue for ancillary services to the grid.
Question: Is treatment of storage as a generator a consistent treatment across independent system operators (ISOs) and regional transmission organizations (RTOs)?
Answer: The treatment of storage is currently not consistent across ISOs and RTOs. However, FERC Order 841 which was issued in February 2018 is expected to change that. The order’s basic premise is to require RTO and ISOs to implement technology-agnostic participation models that do not discriminate against storage. The order provides the ISOs and RTOs a healthy amount of latitude to decide how they want to structure their participation models to meet this basic premise.
However, one area in which the RTOS and ISOs are not granted latitude is regarding the flexible treatment of storage as both generation and load. Under Order 841, the RTOs and ISOs must tailor their tariffs and market rules to allow storage to be dispatched as both supply (i.e., generation) and demand (i.e., load); further, storage projects must able to participate in setting wholesale market clearing prices as both a wholesale seller and wholesale buyer. Some ISOs, like CAISO, PJM and ISO-NE, are already in compliance with this aspect of Order 841; however, others, like NYISO, are not. The RTOs and ISOs made their initial compliance filings in December 2018, but full implementation of the order is not expected until next year in some cases.
Start of construction for tax credit qualification purposes
Question: Many developers used the construction of road for wind projects to “start construction” prior to the deadline for the decline in the tax credit that wind projects are eligible for (e.g., 2016 to qualify for a full tax credit). Are utility scale solar developers using the construction of roads for this purpose with respect to the phase-down of the ITC?
Answer: As they were in IRS Notice 2013-29 for wind, roads are referenced in Notice 2018-59 as a means for a project owner to start construction. In order to qualify, the roads must be needed “to operate and maintain” the propjet. We suspect some solar developers will use them as a start of construction strategy; however, similarly to what was observed in wind, the construction of roads will likely be combined with another strategy such as entering into a binding written contract for the purchase of a custom step-up transformer.
A frustrating aspect of the IRS notices is that they do not specify a minimum threshold of road that must be constructed to qualify. Obviously, more road construction is better than less. Also, it is a best practice to ensure the segment of road constructed for this purpose is actually included in the final project and the work is substantial enough to survive any snow or rain storms that the project site is subjected to prior to its completion.
Question: What’s the debt-equity ratios for projects are being observed with the new revenue models (e.g., corporate PPAs)?
Answer: As of the commercial operation date, sponsor equity typically accounts for to 60 to 70 percent of project’s capital stack for solar and 40 to 50 percent for wind with the balance of the capital stack funded by tax equity. Project level debt is used in a minority of solar transactions and almost never in wind transactions. Sponsors in transactions without project level debt will often borrower back-leverage (holdco-level debt) that is secured only by the sponsor’s interest in the tax equity partnership(s) that owns the project(s).
The amount of back-leverage is project-specific: it depends on factors such as the pricing and tenor of the PPA and the cash distribution sharing percentages between the sponsor and the tax equity investor. The ability of sponsors to raise back-leverage has declined in recent years as PPA rates moved lower and PPA terms became shorter. We have observed wind projects for which project sponsors were not able to raise any back-levered debt.
Project market finance market for fuel cells
Question: How is the project finance market responding to fuel cell projects?
Answer: Several banks have executed debt and tax equity financings for fuel cells, and several more banks are actively pursuing such transactions. These investments reflect the fact that fuel cell technology has improved, while the cost of fuel cell projects has declined. Further, tax equity investors appreciate the ITC eligibility of fuel cell projects.
Given the small footprint required for a fuel cell project, they can be deployed at utility scale in urban areas with high population density.
As fuel cells provide clean around-the-clock power with a small footprint and tax advantages, we expect that the market acceptance of the technology will continue to grow.
Question: What is the average actual additional generation captured by solar trackers?
Answer: The typical added production for single-axis tracker vs. fixed tilt is in the range of 10 to 25 percent. The percentage of added production depends on the site location, the operating characteristics of the tracker and other conditions.
Question: Currently, what do engineers believe is expected lifespan of an inverter?
Answer: For “central inverters” (i.e., larger capacity inverters handling all or a large portion of a project’s capacity, as opposed to “string inverters” which have a smaller capacity and are distributed within the project), the typical lifetime assumption is 30 years. This assumption appears reasonable for better quality inverters installed in mild weather sites with typical loads. For a central inverter to achieve that lifespan, normally one major overhaul is expected, which typically the project’s financial model provides for by assuming some of the project’s cash flow will be reserved for such an expense, rather than distributed to the project’s owners. Harsher environments, heavier generation loads and poor design can result in significantly more maintenance and repairs being required for a central inverter to achieve a 30-year life.
Question: What has the failure record been of wind turbine foundations over the past 20 years?
Answer: We have not observed widespread failures of foundations in the United States, but there have been some isolated cases of such failures. The earliest multi-megawatt wind turbines in the U.S. market are now approaching their twenty-year operating period, and such turbines had a 20-year “design life.”
We have observed instances of the need for retrofits to maintain foundation viability for continued or extended operation. There are range of failure modes that a foundation can suffer, and not all of them are catastrophic for the turbine.
The design life of newer wind turbines is typically 25 to 30 years.
Question: What do you expect the impact on tax equity investors to be if PG&E’s bankruptcy results in PPAs being rejected or renegotiated at lower rates?
Answer: As an initial matter, it is imprudent to lump all tax equity together. The impact could vary significantly transaction-by-transaction. Some of the variables are ITC versus production tax credit (PTC) projects, the age of the tax equity transaction when the PPA is rejected and how much of the tax equity investor’s return of and on its capital remains to be funded by cash flow.
As a general matter, we believe tax equity investors will suffer less due to PPA changes than the sponsor will. A renegotiation of the PPA may have minimal adverse effect on the tax equity investor, so long as the project never cease selling power or ceases for only a short period of time. Similarly, a rejection would not be that painful for most tax equity investors, so long as the project can sell power on a merchant basis or contract with another offtaker.
However, the degree of the tax equity investor’s pain would increase dramatically if the PPA rejection (i) occurs during the five-year ITC recapture period or the ten-year period that PTCs are available and (ii) the project is unable to sell power for an extended period of time due to the rejection.
The PTC for wind is $24 per megawatt hour generated in the project’s first ten year of operations measured from the placed in service date; there is no “tolling” of the ten-year period due to an offtaker bankruptcy or any other reason. The PTC only accrues for third-party sales, but any electron that eventually reaches a consumer or third-party business customer is deemed sold to a third-party. See IRS Notice 2008-60, 2008-2 C.B. 178 (interpreting, inter alia, §45(a)(2)(B)).
Accordingly, if the rejection of the PPA caused the project not to be able to sell its power to third-parties then while the project is out of operation those foregone PTCs would be lost forever. If subsequently some cash is generated from power sales, the tax equity investor may have first right to some or all of that cash to reach its agreed after-tax return given the delay in achieving such return due to the foregone PTCs. Such a cash sweep could be quite expensive for the sponsor as the after-tax value of $1 of PTCs is equivalent to $1.59 of project cash, using a 37 percent corporate tax rate.
Solar or other projects claiming the ITC would face a related but different issue. If the rejection of the PPA caused the project to cease selling power, at any price, the IRS could assert the project was “removed from service” and recapture of the unvested ITC was triggered. See e.g., I.R.M., Ch. 9 (Credit Recapture on Disposition) (“recapture would be triggered … when the property ceases to be income producing”); see I.R.C. § 50(a)(1). As the IRS’s manual uses the phrase “ceases to be income producing,” we hope the IRS would not raise such an issue based on merely an indefinite pause while the project owner actively is looking for an alternative means to dispatch power to the market.