The value proposition for energy storage
SCE and Duke Energy discuss what they’re learning
Published In: Intelligent Utility Magazine September / October 2012
RIGHT NOW, MYRIAD UTILITIES ARE TESTING THAT VALUE proposition regarding what role energy storage could play for them and whether it's technically and financially feasible.
Two of the nation's utilities-Southern California Edison and Duke Energy-are among those utilities implementing specific energy storage technologies for a widely sought application: the integration of renewable resources.
In a July Intelligent Utility Realities webcast on the subject, Mark Irwin, director of technology advancement at Southern California Edison, and Chris Rees, strategic planning manager at Duke Energy, discussed their utilities' projects, as well as the early lessons learned.
Here, edited for length, are excerpts from their comments.
REES There's a very wide range of costs out there. I think one of the challenges with looking at storage devices and evaluating them is that there are very different applications, and certain technologies are better suited to different applications. If you look at a generation resource like a combustion turbine facility, you think about that at a dollars-per-kilowatt basis. It's very easy to compare apples to apples and say that this facility costs $X per kilowatt and another facility costs $Y per kilowatt, and make that comparison.
With energy storage, where you're looking at different applications, making that comparison of apples to apples, it's a little bit more tricky. And so, with our most recent project, for example, the project cost is $45 million for a 36 MW output project, and for us, that output was key to being able to generate capacity revenues in the ERCOT market. So we look at that on a dollars-per-kilowatt basis, 45 divided by 36 generates a cost of roughly $1,250 or $1,300 per kilowatt for that project.
Now, if you're looking at a longer-duration project, you compare that on a dollars-per-kilowatt basis, that might look like several thousand dollars per kilowatt, and so you would say, "Well, clearly, that asset is cheaper." Well, it's cheaper for that application that you're looking at, being a short duration application, but for a longer-duration application, you might want to look at it on a dollar-per-kilowatt-hour basis instead. Again, you'd see very wide ranges on that, as well.
Another thing is the life cycle each battery might have. Some batteries may have fewer cycles that you can use per the life of that system. And so, even though it may be cheaper up front, you may not be able to use that system for as long, and so the life cycle cost might be a little bit higher, as well.
IRWIN One of the things that continues to be a challenge is that there's lots of work going on out there in the market so that people can now tell you how their batteries have performed under different lab testing and such, but I think at the end of the day, what we implement into our system, whether we're a utility or a competitive business (like this Duke project is), we integrate systems. And so seeing people with proven systems that operate is not a small step. It's not a small step from the technology to the actual integrated system. I think that's one important thing that we've seen, and I'd say we've seen that not just necessarily at Tehachapi, but across all our battery projects. That's been critical. And I think the other thing for us, which is more Tehachapi specific, is seeing how we can have a device that integrates both into systems operations-a lot of our purposes are system-and also has some activities in the market at the same time is not a small challenge.
And I think we'll go through that in the testing process and we'll show that. But I think doing that on a long-term basis is difficult. Because, in our company at least, and probably in a lot of larger utilities that participate in a market, we have one group in our company that participates in the market, and we have another group in our company that participates in transmission, and we have very clear FERC regulations about communication. And so getting a device that crisscrosses those groups is something that will be interesting moving forward, and be a good challenge.
REES Some of our lessons learned have been more about what do we need to do to get a storage resource like this registered in ERCOT and be able to participate in the market, what the stakeholder process in ERCOT is like in terms of being able to make some of those changes that would accommodate a storage resource being able to participate more fully. So we're still trying to work through what some of those market rule changes are going to be that might accommodate the storage device, understanding more what the potential pilot opportunity might look like with ERCOT.
So as that becomes more clearly defined, I think we're continuing to learn and we'll continue to establish how we want to move forward with testing the devices and how we're going to optimize it in terms of participating in the ERCOT market.
As well, a big piece of it is understanding what the application is you're going to be using (energy storage) for. Certainly if you only need 15 minutes of storage for an application, putting in a lithium ion solution may not be the most cost-effective solution for that.
Applications question essential
IRWIN In one of our slides, we listed 13 applications or so. The comparison depends on what you're doing. The reason I see the complexity is, it's which application, so what device can do that, and then with the storage device, when you say we're going to do four or five applications, you have actually multiple devices you have to compare it to. So one thing for variable resources might be a peaking plant. But it really depends.
REES In addition to thinking through what the specific applications are and what the requirements of those applications are, what is the true life cycle cost of that going to be?
What are the variable costs of maintaining that system, and what are the maintenance costs for that system going forward? How long will you need that system to last? How long do you need it to last for the application?
So as we considered the project and as we looked at other potential investment opportunities for storage in other markets, we really looked at that combination of factors.
At this point, each project almost has to be evaluated on the usual basis, and there's not really a great resource out there for making a true apples-to-apples comparison across multiple different projects or multiple technologies.