by Julian Spector
June 15, 2018

I'm about to show you something you're not supposed to see. 

It's daring and thrilling and you can't get a peep at it in most states, but Nevada takes a different approach to what must be concealed.

I'm talking, of course, about a solar-plus-storage power-purchase agreement.

NV Energy's new solar procurement made headlines because it was cheaper than any other public solar deals in the U.S. Hidden in the appendix, dismissed as a vestigial organ by many, the full-length contracts sat, waiting to be discovered.

Evidently, Nevada requires a greater degree of openness in its utilities' contracts than do most other states. In this week's Storage Plus, we're going to dive into that document like a drunk tourist into the Bellagio's dancing fountain. It says a lot about what developers think about pricing, how utilities gauge lithium-ion degradation and how hybrid storage deals are getting structured.

But first, the numbers

The target to beat for solar-plus-storage was the Xcel solicitation's median pricing of $36 per megawatt-hour. That figure, released six months ago, became a touchstone for measuring the advancement of aggressive storage pricing.

Just last week, Xcel released a follow-up plan with more specific project and pricing information, Julia Pyper reported. Now the utility is looking at levelized pricing of $30-$32 per megawatt-hour for solar-plus-storage, setting an even more aggressive benchmark.

The tricky thing with Nevada’s results is that the utility broke out the solar and storage into two separate contracts, rather than reporting a blended number. You can't compare a solar energy contract and a storage capacity contract with a single solar and storage energy contract.

However, should you take the time to dig through the hundreds of pages of appendices, the utility clarifies the levelized cost of energy for the projects.

That number includes network upgrade costs and maintenance, so as a general rule it should be a little higher than the pure blended PPA price. With that in mind, here are the numbers:

  • Dodge Flat (NextEra): LCOE is $29.23/megawatt-hour for solar only, $34.87/megawatt-hour including storage. Due online December 2021. PPA here.
  • Battle Mountain (Cypress Creek): LCOE is $26.50/megawatt-hour for solar alone, $30.94/megawatt-hour for solar and energy storage. Due online June 2021. PPA here.
  • Fish Springs Ranch (NextEra): LCOE is $30.67/megawatt-hour for solar, $36.94/megawatt-hour for solar and energy storage. Due online December 2021. PPA here.

As you can see, Dodge Flat and Battle Mountain beat Xcel's initial median price point handily, and Fish Springs Ranch slightly exceeds it. 

Those projects don't look as good compared to the newest pricing from Xcel, but since that number comes as a range, and doesn't refer to a finalized project, there isn't a firm comparison to make.

I can say that Battle Mountain has the cheapest solar-plus-storage pricing I've seen so far, but it's a modest and believable improvement on numbers we'd seen elsewhere. As GTM Research Storage Director Ravi Manghani put it, "The numbers didn’t shock me, but they’re still pleasantly surprising."

That said, Xcel is looking at a 2023 in-service date for its solar-plus-storage, so that makes Nevada look more ambitious by comparison.

How did they get here?

These deals have time on their side — almost three years of industry learnings and cost declines, depending on how long the developers can wait before buying their components.

So far, historically, time has been a battery developer's best friend. That won't necessarily hold true forever. A surge in demand for electric vehicles or a supply chain bottleneck for cobalt could make high quality batteries harder to come by. The president could impose unexpected tariffs on key materials. Who knows.

These sorts of risks, though, are nothing new for energy contracting, which routinely happens years ahead of the due date.

"There is still a lot of uncertainty out there, but factoring in these uncertainties, the costs are within the band of what we can expect in the market," Manghani said.

The structure of the deal is important to understand. Instead of blending it, as Xcel and Tucson Electric Power seem to have done, NV Energy splits the solar and storage contracts. That means it isn't paying for the total amount of energy that passes through the batteries; it's paying for the capacity delivered.

The PPA for Battle Mountain, for instance, stipulates that the battery must be at least 98 percent available during on-peak hours in the summer. It also caps the number of cycles for the system at 300 for the year and no more than two per day.

The utility appears to have thought hard about what it really needs the batteries for, and then framed the contract around that. It's paying for capacity during peak hours. The 300-cycle cap allows for some additional use beyond that, but a limited amount.

"The smartest way of procuring storage, in my opinion, is looking at it as a capacity resource and not as an energy resource," Manghani said.

"The per-megawatt-hour framework became relevant with the inception of solar and wind contracts," he added. "Those assets are non-dispatchable. As soon as we talk about dispatchable assets, you want to look at it as a capacity."

Otherwise, it's worth noting that both developers, NextEra and Cypress Creek, have experience with large-scale solar projects. NextEra has already won several large storage deals as well. And the solar side, which is the bigger part of the cost stack, benefits from wide open space and a lot of sunshine.

The more targeted storage capacity contracting will be a trend to watch. There simply haven't been many solar-plus-storage contracts publicly described in great detail, so we haven't seen that approach taken yet. 

Not shy about degradation

The Battle Mountain PPA directly addresses the degradation that will occur on the lithium-ion batteries.

It features a chart (page 186) in which the maximum stored energy level drops by 2 megawatt-hours each year, from 100 in Year 1 to 82 in Year 10.

That signals the developer and utility agree that batteries will fade, and won't expect otherwise. Battery scientists may take issue with the implication of linear degradation; very little in the real world works that way, and it's implausible to assume the system will perform identical charge and discharge in identical weather conditions year after year.

I didn't see that kind of chart in the NextEra PPA agreements. They have a different approach: a 2 percent annual escalator on the storage capacity price.

It's not clear what purpose that serves. It could incentivize NextEra to keep the battery capacity refreshed in order to earn more later on in the project's life. But if the batteries did degrade, the company would still earn more later on than if there wasn't an escalator.

Different lifespans

There's another good reason, illustrated in this procurement, for splitting the storage and solar contracts. They have different life expectancies.

Battle Mountain includes a 25-year solar PPA and a 10-year storage capacity contract. The NextEra projects have 25-year solar PPAs with storage contracts for 15 years.

We don't have good real-world examples of how storage paired with solar ages in the field. Clearly, the developers aren't expecting that resource to last as long as the solar, and NV Energy is comfortable with that. Then the question is, what happens when the storage contract ends? 

It could be that the storage system still works pretty well and the parties can just sign a new contract and pop some new batteries in. Or maybe the next-generation storage will be so much cheaper and better than today's offerings that it will be cost-effective to buy a whole new system to ride through the rest of the solar lifespan.

There isn't a playbook for this yet, but it's something the industry will need to clear up eventually.