Batteries & Solar: The Cheapest Capacity and Peak Energy Resource Available

admin
Print Friendly
Share on LinkedIn0Tweet about this on Twitter0Share on Facebook0

______________________________________________________________________________

Eric Selmon Hugh Wynne

Office: +1-646-843-7200 Office: +1-917-999-8556

Email: eselmon@ssrllc.com Email: hwynne@ssrllc.com

SEE LAST PAGE OF THIS REPORT FOR IMPORTANT DISCLOSURES

______________________________________________________________________________

July 2, 2018

Batteries & Solar: The Cheapest Capacity and Peak Energy Resource Available

Six PPAs recently signed by NV Energy document a dramatic reduction in the costs of solar energy and energy storage. In this note, we assess the implications for the long-term competitiveness of both technologies.

Portfolio Manager’s Summary

  • NV Energy has signed six PPAs to procure a total of 1 GW of solar generation capacity and 100 MW of associated battery storage. These PPAs call for:
    • Solar energy prices as low as $23.76/MWh flat over 25 years, and
    • Battery capacity prices as low as $6.11/kW-month, subject to 2% annual escalation over 15 years.
  • Assuming (i) weighted average cost of capital of ~7% (in line with a target sponsor IRR of ~10% and a 50% debt to capital ratio), (ii) 100% bonus depreciation, (iii) a 30% ITC for solar and associated storage capacity and (iv) DC-to-AC conversion losses of 14%, we estimate the installed cost of solar capacity at these projects to be ~$0.80/kW and the cost of battery storage to be ~$210/MWh or $840/MW.
  • Another key data point is the 33% solar capacity factor (AC output to AC capacity) expected to be achieved at NextEra Energy’s Dodge Flat project, whose solar panels will be mounted on trackers.
  • Calibrating our forecasts to reflect these data points, we expect the cost of capacity from batteries, when installed at solar facilities and thus qualifying for the solar ITC, to be 20% lower than that of gas turbine peakers in 2021 and 34% lower in 2024.
  • Even if a battery facility qualifies only for the standard 10% ITC, we expect the capacity cost of batteries in $/kW-month to be 23% lower than that of gas turbine peakers by 2024 (Exhibit 3).
  • Further, we estimate the levelized cost of solar energy in the southwestern United States to be 32% below average peak hour power prices in the region by 2021 and 45% below by 2024.
  • Even in the absence of the solar ITC, we estimate the LCOE of solar energy in the southwestern United States to be 35% below peak hour prices in 2024 (see Exhibit 4).
  • Absent the solar ITC, we expect the levelized cost of solar energy to remain ~15%, or ~$4-6/MWh, above average peak hour prices in 2024 in the southern, midwestern and northeastern regions of the country, but low enough to remain attractive to utilities in light of its environmental attributes.
  • We see sales of gas turbine peakers to be at risk from lower cost battery storage.
    • We estimate that gas turbine peakers comprise at least 50% of total gas turbine orders, the remainder being combined cycle gas turbine units.
    • We expect that storage will capture some share of the market for peaking capacity served by gas turbines. In our report GE, SIE, MHVYF: The Impending Recovery in Gas Turbine Orders, we modeled a scenario where 50% of gas turbine peaker volumes are lost to batteries, but we believe this to be a highly conservative downside case, with actual losses to batteries likely to be below that. (Click here for the full report).
    • We note that the four-hour duration of discharge typical of lithium ion batteries will constrain their deployment as peaking capacity in certain contexts, such as in regions with winter peaks, which tend to be of longer duration, and particularly in RTOs such as PJM and the New England ISO, that require capacity resources to be capable of 16 hours of continuous discharge.

Exhibit 1: Heat Map: Preferences Among Utilities, IPP and Clean Technology

Source: SSR analysis

Details

Following a competitive solicitation initiated in January, on May 31st NV Energy signed contracts to procure a total of 1 GW of solar generation from six greenfield projects in Nevada. Three of these include associated energy storage facilities totaling 100 MW or 400 MWh of capacity. The six solar energy projects and three energy storage facilities are scheduled to enter operation by the end of 2021. Summary terms of these contracts, which remain subject to Nevada PUC approval, are presented in Exhibit 2 below.

Exhibit 2: Terms of Solar Energy and Energy Storage Contracts Signed by NV Energy (1)

_________________________________

1. Capacity figures represent maximum AC (alternating current) output.

Source: Application of NV Energy before the Public Utilities Commission of Nevada seeking approval for 1,001 MW of renewable power purchase agreements, June 1, 2018, available at http://pucweb1.state.nv.us/PDF/AxImages/DOCKETS_2015_THRU_PRESENT/2018-6/30452.pdf

The NV Energy PPAs offer valuable and often surprising insights into the cost of new, greenfield solar energy projects and associated energy storage facilities. Key data points include:

  • Ground-breaking prices for battery storage: NextEra Energy has committed to supply 50 MW of four-hour energy storage at its Dodge Flat project and 25 MW of four-hour energy storage at its Fish Springs project at prices of $6.11 and $6.20 per kW-month, respectively. Both 15-year contracts are subject to 2.0% annual price escalation. Importantly, these prices reflect the eligibility of batteries for the solar ITC when built in conjunction with solar energy projects.[1]
  • Highly competitive solar energy prices: The prices at which NV Energy has contracted for solar energy range from a low of $23.76/MWh to $29.89/MWh, with no escalation over the 25-year term of these contracts. (NV Energy’s PPA with the Copper Mountain project calls for an even lower initial price of $21.55/MWh, but is subject to 2.5% annual escalation.)
  • Surprisingly high solar capacity factors: NEE’s Dodge Flat project is expected to generate 375 GWh annually from a 200 MW AC solar PV facility, for a capacity factor of 33% (AC output to AC capacity). [2]

We have used the information disclosed on these projects by NV Energy in its filing with the Nevada Public Utilities Commission[3] to estimate the installed cost per MW of the solar PV and battery storage capacity deployed by the most competitive projects. Assuming (i) weighted average cost of capital of ~7% (in line with a target sponsor IRR of ~10% and a 50% debt to capital ratio), (ii) 100% bonus depreciation, (iv) a 30% investment tax credit (ITC) for solar energy,[4] and (iv) 14% DC-to-AC conversion losses, we estimate the installed cost of solar capacity at the Dodge Flat and Eagle Shadow Mountain projects to be ~$0.80/Watt of DC capacity. Based on the contract price for storage capacity at NextEra’s Dodge Flat project ($6.11/kW-month) and utilizing the same financing assumptions (including the ability of batteries built in conjunction with solar energy facilities to qualify for the solar ITC), we estimate NEE’s installed cost of battery storage to be $210/kWh or $840/MW AC.

We have updated our estimate of the future cost trajectory of utility scale solar PV and battery storage capacity to reflect these new estimates of 2021 costs (see Exhibit 3). From 2021 through 2024, we assume that the installed cost of utility scale solar PV declines at a moderate annual rate of ~5% to $0.70/Watt DC, well below ~10% annual cost reductions over 2018-2021 implied by the NV Energy PPAs. Similarly, we assume that from 2021 through 2024 the installed cost of grid scale battery storage declines at an annual rate of ~8% to $165/MWh, below the ~10% annual cost reductions for battery storage implied by the NV Energy PPAs over 2018-2021.

Exhibit 3: Estimated Cost Trajectory of Utility Scale Solar and Battery Storage

________________________________________________________

Source: SSR analysis and estimates; Lazard’s Levelized Cost of Storage Analysis, Version 3.0 and Lazard’s Levelized Cost of Energy Analysis, Version 11.0; NEE solar and storage costs estimated from Application of NV Energy before the Public Utilities Commission of Nevada seeking approval for 1,001 MW of renewable power purchase agreements, June 1, 2018, available at http://pucweb1.state.nv.us/PDF/AxImages/DOCKETS_2015_THRU_PRESENT/2018-6/30452.pdf

Based on the revised cost trajectory for grid scale battery storage set out in Exhibit 3, we have prepared a comparison of the all-in cost of short-term (four-hour) peaking capacity from:

(i) a grid scale lithium ion battery storage facility;

(ii) a gas-fired combustion turbine, the current standard to serve large scale capacity needs (>100 MW), and

(iii) a gas-fired internal combustion engine, used for smaller capacity requirements (<100 MW).

Our comparison is based upon the cost, expressed in $/kW-month, of contracting for capacity alone from facilities deploying these three alternative technologies. Reflecting the structure of standard capacity contracts, the variable cost of energy, expressed in $/kWh, is excluded from this comparison.[5]

To compare the all-in cost of capacity from each of the three alternative technologies, we have calculated the revenue required by each to recover the costs of capital recovery, an after-tax, weighted average cost of capital of 7% and income taxes. In Exhibit 4, we have expressed the revenue requirement of each technology in $/kW-month, and trace its expected trajectory over time.

Exhibit 4: Comparison of the Cost of Capacity from a Battery Storage Facility, Gas-fired Combustion Turbine and Gas-fired Internal Combustion Engine, in $kW-month (1)

_____________________________________________________________________________

1. ITC refers to the investment tax credit available to battery storage facilities that are integrated into a solar energy project. We assume that both the battery storage and gas-fired peakers are owned by an independent power producer, rather than a regulated utility, and are thus able to capitalize on the bonus depreciation provisions of the Tax Cuts and Jobs Act of 2017. Please see the first paragraph on page 6 for an explanation of our tax assumptions.

Source: SSR analysis and estimates; Lazard’s Levelized Cost of Storage Analysis, Version 3.0 and Lazard’s Levelized Cost of Energy Analysis, Version 11.0; SSR analysis and estimates.

For purposes of this comparison, we have assumed that the gas-fired internal combustion engine costs ~$950/kW, while the gas-fired combustion turbine has an installed cost of ~$875/kW. We have further assumed that the installed cost of the combustion turbine and internal combustion engine remain constant in real terms (i.e., expressed in constant 2018 dollars). As noted above, based on the contract price for storage capacity at NextEra’s Dodge Flat project, we estimate NEE’s installed cost of battery storage to be $210/kWh or $840/MW AC. We assume that from 2021 through 2024 the installed cost of grid scale battery storage declines at an annual rate of ~8% to $165/MWh.

The provisions of the tax code have a significant impact on our comparison of capacity costs. The increase in the cost of capacity from gas-fired peakers visible in Exhibit 4 is attributable to the step-down in bonus depreciation, which drops from 100% over 2018-2022 to 80% in 2023 and 60% in 2024. Similarly, the cost of capacity from battery storage facilities integrated into solar energy projects rises over time as the solar ITC is phased out. The solar ITC is 30% if construction of the project commences by the end of 2019. It then steps down to 26% for projects that begin construction in 2020, to 22% for projects that begin construction in 2021 and to 10% for commercial solar energy projects that comment construction after 2021. However, recently released IRS guidance allows solar energy developers up to four years to complete the construction of their projects (see footnote 4 on page 3 for more detail). We therefore assume that a battery storage facility integrated into a solar energy project that begins construction by 2019 and comes on line in 2020 will qualify for the 30% solar ITC, and that one begins construction by 2020 but comes on line as late as 2024 will qualify for the 26% ITC. A battery storage facility integrated into a solar energy project that commences construction after 2021 will qualify for the 10% ITC. Finally, battery storage facilities that are not integrated into solar energy projects do not qualify for an investment tax credit.

Our operating assumptions for each of the three alternatives, which are the same for each year of the forecast, include:

    • A uniform capacity factor of 10% across the three technologies;
    • Fixed operation and maintenance expense of $15/kW-year for the internal combustion engine, $10/kW-year for the combustion turbine and $10/kW-year for the battery;
    • A useful life of 15 years for the storage facility, corresponding to ~3300 cycles; and
    • A useful life of 25 years for the gas-fired peakers.
    • Fuel and purchased power costs are excluded, as we modeled these resources as selling power under tolling contracts, where all costs of generation are borne by and all power output is owned by the purchasing utility.

As is evident from Exhibit 4, our analysis suggests that grid scale battery storage, built at a solar PV facility to be eligible for the solar ITC, is a materially lower cost source of four-hour peaking capacity than either of the conventional gas-fired peakers. We expect the cost of energy storage, expressed in $/kW-month, to be 20% lower than that of gas turbine peakers in 2021 and 34% lower in 2024, provided the project began construction by 2020. If we assume the solar energy and battery storage project started construction after 2021, and qualifies only for the standard 10% ITC, we expect the cost of battery storage in $/kW-month to be 23% cheaper than gas turbine peakers in 2024. Finally, a stand-alone battery storage facility that is not integrated into a solar energy project, and which therefore does not qualify for the solar ITC, would nonetheless be 13% cheaper that a gas turbine peaker in 2024.

We therefore see sales of gas turbine peakers to be at risk from lower cost battery storage, particularly after 2024, when we expect storage to be competitive with gas turbines on an unsubsidized basis. We estimate that gas turbine peakers comprise at least 50% of total gas turbine orders (the remainder being combined cycle gas turbine units), and we expect that storage will capture some share of this market. In our report GE, SIE, MHVYF: The Impending Recovery in Gas Turbine Orders, we have modeled a scenario where, from 2024 on, 50% of gas turbine peaker volumes are lost to batteries. (Click here for the full report).

However, we believe this to be a highly conservative downside case, with the displacement of gas turbine peakers by batteries likely to fall well below this level. This reflects the fact that the four-hour duration of discharge typical of lithium ion batteries will constrain their deployment as peaking capacity in certain contexts. Four-hour peaking capacity may be insufficient, for example, for utilities operating in regions where load peaks in the winter months, as winter peaks tend to be of longer duration than summer peaks. Similarly, competitive generators will be unable to obtain capacity credit for four-hour discharge batteries in RTOs such as PJM and the New England ISO, which require capacity resources to be capable of 16 hours of continuous discharge.

Nonetheless, for utilities requiring short-term peaking capacity, we see grid scale battery storage as a highly competitive alternative to conventional gas-fired peakers. Moreover, as the cost of solar and batteries declines, we expect increased pressure on RTOs to change their rules to allow the participation of storage combined with solar.

Finally, we have used our updated estimate of the cost trajectory of solar PV capacity to estimate the levelized cost of energy from ground mounted, utility scale solar facilities in various regions of the country. As can be seen in Exhibit 5, our analysis suggests that in the southwest – the region of the United States with the highest solar insolation – the LCOE of utility scale solar is today highly competitive with average peak hour power prices, and will become more so over time. We calculate the LCOE of solar in 2018 to be 18% below average peak hour power prices in the region. By 2021, we estimate that the LCOE of solar will be 32% below average peak hour prices (based on the currently prevailing forward price curve for the 2021 energy year); by 2024, we expect the LCOE of solar to fall to 45% below average peak hour prices. Even in the absence of the solar ITC, we estimate that the LCOE of solar in the southwestern United States will be 35% below peak hour prices in 2024 (see Exhibit 5).

In other regions of the United States, where the solar resource is inferior to the southwest, solar is less competitive with peak hour prices. Across the southern states stretching from Texas to Georgia, for example, we see the LCOE of solar falling slightly below peak hour prices as long as the solar ITC remains in place (see the left-hand chart of Exhibit 6). Absent the solar ITC, however, we estimate that in 2024 the levelized cost of solar energy will be ~14% above average peak hour prices in the region. In the Midwest and northeastern states, we estimate the LCOE of solar to be at or slightly above the average level of peak hour prices through 2024 (see the right-hand chart of Exhibit 6). With the expiry of the solar ITC, we estimate that in 2024 the levelized cost of solar energy will be ~16% above average peak hour prices in the region.

Nonetheless, given the highly competitive cost of grid scale battery storage and the limited premium for solar (~$4-6/MWh), we expect solar supplemented by storage to become an increasingly compelling option for regulated utilities seeking to expand their use of renewable energy while addressing short-term non-winter peaks in power demand.

Exhibit 5: Levelized Cost of Solar Energy in California/Nevada vs. On-Peak Power Prices at CAISO SP-15 (1)

__________________________________________________________________________________

1. LCOE of solar estimated based upon a ground-mounted, utility scale facility equipped with horizontal single axis tracking. CAISO on-peak power prices reflect the average of monthly forward prices for on-peak power at SP-15 over the June 1 to May 30 energy year.

Source: SSR analysis and estimates; solar LCOE estimated from data in Application of NV Energy before the Public Utilities Commission of Nevada seeking approval for 1,001 MW of renewable power purchase agreements, June 1, 2018, available at http://pucweb1.state.nv.us/PDF/AxImages/DOCKETS_2015_THRU_PRESENT/2018-6/30452.pdf

Exhibit 6: Levelized Cost of Solar Energy vs. On-Peak Power Prices (1)

__________________________________________________________________________________

1. Average of monthly forward prices for on-peak power over the June 1 to May 30 energy year.

Source: SSR analysis and estimates; solar LCOEs estimated from data in Application of NV Energy before the Public Utilities Commission of Nevada seeking approval for 1,001 MW of renewable power purchase agreements, June 1, 2018, available at http://pucweb1.state.nv.us/PDF/AxImages/DOCKETS_2015_THRU_PRESENT/2018-6/30452.pdf

©2018, SSR LLC, 225 High Ridge Road, Stamford, CT 06905. All rights reserved. The information contained in this report has been obtained from sources believed to be reliable, and its accuracy and completeness is not guaranteed. No representation or warranty, express or implied, is made as to the fairness, accuracy, completeness or correctness of the information and opinions contained herein.  The views and other information provided are subject to change without notice.  This report is issued without regard to the specific investment objectives, financial situation or particular needs of any specific recipient and is not construed as a solicitation or an offer to buy or sell any securities or related financial instruments. Past performance is not necessarily a guide to future results.

  1. Batteries qualify for the ITC when combined with solar if >75% of the power used by the battery is from the solar system. If the battery qualifies, the ITC is prorated by the actual percentage of power from the solar project (e.g. 90% of the power from the solar project results in 90% of the ITC for the battery). We have assumed 100% of the power used by the battery is from the associated system panels. 
  2. The discussion of the projects in the RFP results refers to “tracking mounted” projects, except for the Dodge Flat and Fish Spring projects, which are referred to as “horizontal single-axis tracking.” However, the output all of the projects suggest capacity factors before equipment losses that are only achievable with dual-axis tracking systems based on the calculations of NREL’s PVWatts calculator (pvwatts.nrel.gov). Therefore, either the projects are all using dual-axis tracking, there was a mistake in the Dodge Flat and Fish Springs descriptions and the costs of dual-axis tracking have declined significantly or there is a significant improvement in the output of solar systems beyond what NREL is capturing in the current version of PVWatts. Regardless, we assumed the improvement in single-axis solar output or dual-axis tracking costs were generally achievable for utility-scale solar projects and applied the same improvements to our estimates for other regions. 
  3. Application of NV Energy before the Public Utilities Commission of Nevada seeking approval for 1,001 MW of renewable power purchase agreements, June 1, 2018, available at http://pucweb1.state.nv.us/PDF/AxImages/DOCKETS_2015_THRU_PRESENT/2018-6/30452.pdf 
  4. Pursuant to the Omnibus Appropriations Act of 2015 (P.L. 114-113), investment in commercial and residential solar generation equipment is eligible for a 30% investment tax credit (ITC) if construction commences by the end of 2019. The ITC then steps down to 26% for projects that begin construction in 2020 and 22% for projects that begin in 2021. After 2021, the residential credit will drop to zero while the commercial credit will drop to a permanent 10%. Under IRS guidance issued in June of this year (IRS Notice 2018-59), “start of construction” is defined to mean either that (i) “physical work of a significant nature” has begun at the project site or on equipment for the project at a factory, or (ii) costs equivalent to at least 5% of the total project cost have been incurred. Once “start of construction” has occurred, moreover, the project must be completed within four years. As a result, projects that commence construction in 2019 and are completed by the end of 2023 may qualify for the 30% ITC; those begun in 2020 and completed by the end of 2024 may qualify for the 26% ITC; and those begun in 2021 and completed by the end of 2025 may quality for the 22% ITC. 
  5. Including the variable cost of energy in the comparison of the three technologies would make little difference, at least for battery storage facilities integrated into solar energy projects. The range of the PPA prices for solar power resulting from the NV Energy solicitation (~$24 to $30/MWh) is broadly in line with the estimated $27/MWh fuel cost of a gas turbine peaker (assuming a gas price of $3.00/MMBtu and a heat rate of 8.9 MMBtu/MWh). A battery operating independently of a solar project, however, might gain a modest advantage from the inclusion of energy costs in our analysis. While it would not be eligible for the solar ITC, it would be able to charge with energy from the grid during the four lowest cost hours of the day. Historically, in the California ISO, the price of electricity during the four lowest cost hours of the day has averaged ~$20/MWh. 
Print Friendly