The solar revolution has enabled households to take part in the transition away from fossil fuels. From the early adopters that wanted to contribute to the green movement more than 10 years ago, electricity from rooftop photovoltaics has become a scalable business.
In the US, 3.1GW of residential solar systems were deployed in 2020, an 11% rise on the previous year, despite the pandemic causing slowdowns and significant periods when installers and sales teams weren’t able to get out to customer sites.
Enjoy 12 months of exclusive analysis
- Regular insight and analysis of the industry’s biggest developments
- In-depth interviews with the industry’s leading figures
- Annual digital subscription to the PV Tech Power journal
- Discounts on Solar Media’s portfolio of events, in-person and virtual
Or continue reading this article for free
For a long time, we’ve been writing here at Energy-Storage.news about virtual power plants (VPPs) being a logical next big step forward for distributed solar. By adding batteries, customers can get a greater degree of energy independence — including some backup if the grid goes down — and their utility can use the combined solar-plus-storage asset as a resource too, thanks to the bi-directional power flow between battery system and grid.
In order to leverage those assets though, once again, creating scale is the key. In the first part of this interview with Swell Energy CEO Suleman Khan, published last week, we heard about how Swell Energy has been working to ‘productise’ the virtual power plant proposition: making it attractive to utilities and to their end-customers and then wrapping that into a long-term agreement.
This time out, we speak to Suleman about some of the finer details of the VPP proposition and where he thinks the market is heading.
The beauty of energy storage is in its versatility and flexibility: being able to perform multiple applications and benefit multiple stakeholders. At grid-scale, this is often discussed as putting services together to build a ‘revenue stack’, combining various different grid services. What goes into building the revenue stack for a Swell Energy VPP portfolio?
In terms of the revenue stack, in Southern California Edison the value being delivered to the utility, in one of our programmes, is non-wires solutions (NWS). It's CapEx deferral, they're also making money off of Resource Adequacy (the mechanism by which utilities are recognised by the California Independent System Operator (CAISO) to be meeting obligations to supply energy reliably to customers).
These are four-hour dispatches. In CAISO if you have a four-hour dispatch, you can apply that towards Resource Adequacy value. So Southern California Edison is getting wholesale value in the form of Resource Adequacy, they're getting value in the form of a non-wires solution, which is CapEx deferral, and then they're getting value in the form of evening wholesale energy procurement.
If you look at time-of-use rates, wholesale rates are actually much higher, and the variance between the price variance and wholesale rates at peak time, it's much higher than what is translated to the retail customer.
So the utility is getting at least three types of fantastic value: Resource Adequacy, CapEx deferral due to non-wires solutions and wholesale energy arbitrage.
So we then take that and say, “Ok, this is super-monetisable,” enter into a 10 to 15-year contract with them to deliver this resource and turn that into monthly payments back to the customer. We essentially take that risk, the delivery risk, etc. The art, of course, comes down to how do you take 8,000 systems and orchestrate them into this resource, to get that capacity payment. So that's what we do.
How open to these ideas have utilities been? What sort of reception did you get from them as you first brought these ideas to them?
The utility reception: a lot of it has been responses to requests for offers (RFOs). So they were seeking a solution and we stepped up with what we believe is the smartest, most cost-effective solution. Around 2017 we started seeing and responding to those RFOs and requests for proposals (RFPs) where VPPs fit in as a solution to address various grid challenges.
Last year we started seeing bilateral, unsolicited interest from utilities, which has led to more discussions around the merits of behind-the-meter storage. That bilateral, unsolicited interest through 2020 was amazing.
The time has actually come for VPPs as an asset class. The utilities can now benefit from behind-the-meter rooftop solar, consider it a resource like any other resource, a large power purchase agreement (PPA) that they might have with a large solar farm. We have become a large solar farm, it just happens to be on people's roofs!
Then there's a resilience aspect, when you're behind-the-meter, you can keep people's lights on. And what they all struggled with initially was: how do you get customer acquisition, project fulfilment, financing, and grid services all going together?
So what we're trying to be in a sense is like a VPP in a box. I think the model is similar to community solar, like 10 years ago, you saw community solar essentially come up with a solution that it could templatise and take to many different utilities and it made community solar, something that the utility could have as part of their portfolio all of a sudden.
Well, we're trying to do that with distributed energy resources (DERs). The way to do that is to make them dispatchable and then to figure out how to get a programme going right.
Heading out on the electron highway
How do you then show the utilities and the customer that there’s a long-term value in this? One difficulty that has been seen in energy storage at larger scale is that with a lot of different merchant opportunities it’s not always easy to model the long-term upside and it’s now always easy to see the downside risk. How visible is that to you and your customers and partners?
You're right, in a lot of markets it's hard to get a contract with a utility, it's hard to monetise it on a long-term basis. Our utility contracts are super-valuable, because they're long-term fixed payments that investors can look at and say, “These are nice with a single A credit counterparty. I can give you a cheaper cost of capital here,” and that's been a transformative piece of our business also.
But in some markets, where it's merchant risk, what we're doing is saying to the customer: “Look, we will take on some exposure and give you like a US$500 discount upfront to participate. Then over the course of the next 20 years, we are absolutely going to try to monetise these assets. And we'll split it with you ‘X, Y’” — we still kind of give them a path to monetisation as and when that becomes available.
And for us, this is option value. In places where we don't have a utility contract, it's option value over the next 20 years, that we're securing with the customer.
From the technical side, is there a little bit of a risk in playing some of these systems designed for residential use into the grid services market? How does that work out?
Absolutely, the answer is, yes, they're perfect. They're perfect for it for a couple of reasons.
Number one: shorter, more frequent dispatches have a lesser impact on battery degradation in terms of cycling, than full 80% depth-of-discharge dispatches.
So what we’re doing for Hawaiian Electric Company (HECO, for which Swell Energy is deploying 80MW of aggregated home storage into the US’ largest VPP programme to date), it's a completely different delivery to the three values we’re providing for SoCal Ed.
Basically in HECO, it's ramp up, ramp down and frequency response, we are providing the grid with energy when it needs it at those critical times, just like in SoCal Ed. But then other times, we're keeping parts of the battery empty so we can take down excess wind energy off the grid.
So we're basically helping HECO not clip excess wind energy. We're storing that excess wind energy, which is a big deal, given how much wind HECO has and consumption at night. So when the wind’s blowing, our VPP is going to be called on to actually charge.
The frequency response piece, those are short, direct hits into the grid and from a cycling standpoint, they're actually more benign than full long-term, long-duration dispatches.
Where that example comes in, actually, is when you think about vehicle-to-grid (V2G). When you think about V2G, you might say, “Hey, do you want to dispatch your EV battery the way that the grid wants it to? From an energy density and from a power and energy standpoint, EVs are high power kind of applications. Residential is like low power, a lot of energy type applications. But so long as you're using any of the technologies that we offer, there's no incremental damage to the battery itself.
When Tesla, who you were working for at the time, brought out the Tesla Powerwall in 2015, it seemed like it had been designed with virtual power plants in mind. But it wasn't until 2017 until the company's first VPP with a utility was launched (with Green Mountain Power in Vermont). Do you think they kind of left it to providers like yourself to figure out that sort of piece of it?
Yes, I would say all of the above. So, we certainly see them in the field as well. We have collaborated with them on several VPP applications, and we continue to.
We find ourselves collaborating with various OEMs on VPP projects: Tesla and a number of others. What the business is becoming is that there is an ‘electron highway’ in the form of virtual power plants, it is happening, it is real.
We have half a billion dollars worth of obligation in this space and we are kind of positioning ourselves as the developers and therefore the toll booths on that electron highway.
And a key part of that are all these OEMs. And we're trying to drop a little bit of VPP goodness into all of them. Right. So saying, hey, we've got the VPP thing covered. You're a fantastic battery manufacturer. Let's go to the utilities together.
Finally, the three major markets that we've heard about Swell Energy projects in have been California, Hawaii, and New York. These are obviously all areas that have been spoken about, as already very fertile ground for energy storage, with a lot of deployments in Hawaii and California, and there's a strong expectation that that will be the case in New York. I’m wondering about your views on the dynamics for other parts of the US. Outside the US, we always hear about how it's more than 50 different states and more than 50 different markets. So where do you see opportunities coming up besides those early leaders? Could you guys just afford to stick with those early leader kind of states or does it need geographic scale as well as numbers?
It's a great question. I love giving this example: a utility in Oklahoma came to us and talked about their wholesale demand charges, which is kind of the watermark in the river. Energy cost is the flow of the energy but then however high your demand was at any given point, in a 15 minute interval, is what you have to pay. There were utilities from literally the middle of America coming to us and saying, “Look, we have identified that having batteries at half a percent of our commercial and residential load that we've identified would save us millions”.
This is not one of your 15 solar states. This is not a state in which there's there is really high retail electricity costs. But this is a state in which the wholesale value of energy storage is very material and for a small utility, this could be one way, versus, if you're thinking about, for example, non-wires solutions, this could be one way to make it to next year or make it for the next five years without having to do massive CapEx and rate hikes.
So I would say what's interesting is a California, Hawaii, Massachusetts, New York have been great, because they've adopted policies that really encourage energy storage.
The other states, it might not happen from a policy standpoint, it might, but utility business development solves it. When we get the opportunity to go to a utility… and you can take Texas, for example, there are a deceptively high number of residential solar systems in Texas; I don't think folks fully realise it.
It's a state where I would say it's interesting, because there hasn't been really net energy metering (NEM) in any meaningful form. So there hasn't been the level of solar adoption because of that value proposition. You might think no NEM means limited adoption, but batteries solve that.
So batteries are a way to go into a deregulated state like Texas, and say, “You guys have had solar, but there's this massive market out there, and batteries unleash it because you don't need NEM with batteries”. On the other hand, batteries can actually assist NEM and preserve it in a viable form. So I think we're seeing that, and we're even seeing international interest. There are utilities in Europe and Australia and East Asia, coming to us and saying, “Look, your VPP in a box model is interesting, is it applicable?”
And what I found fascinating is that some of them find different parts of our technology stack, of our software, and our operations to be of different value. Things that in the US, I assumed to be just like, this is just what you need to run the business, like sales and fulfilment, in Australia that was deemed to be one of our most interesting software functionalities, because it's such a fragmented market. So there's applicability beyond the states that we're in, there's applicability beyond just the solar states, in fact.
Again, it's on the utility side you find it, and there's certainly applicability in my mind in Southeast Asia, Australia and Europe.
Cover image: With the highest per capita rate of solar adoption among US states, Hawaiian Electric has contracted Swell Energy to aggregate 80MW of customer-sited solar batteries to provide an orchestrated resource for the utility's portfolio. Image: Hawaiian Electric via Twitter.
Read the first part of this interview, 'Time to take virtual power plants seriously: Swell Energy creates asset class from customer assets,' here.
This article has been amended from its original form to reflect that Swell Energy has a “number of” OEM suppliers besides Tesla, not “two other” suppliers as was mistakenly reported.