by Stephen Lacey
January 20, 2017

Stephen Lacey: This is The Interchange, conversations about the changing business of energy and cleantech from Greentech Media. I'm Stephen Lacey joined in our Boston office by my co-host Shayle Kann, who moonlights as our Senior VP of the research outfit here at GTM, when he's not recording with me. Hey Shayle.

Shayle Kann: Hey Stephen.

Stephen Lacey: Brett Simon is with us this week. He's an analyst on the energy storage team. Brett, you were busy before the holidays tracking an up-and-coming sector of the storage market: behind-the-meter installations. And that's what we're going to talk about today. How's it going?

Brett Simon: I'm doing well, Stephen. Thanks for having me.

Stephen Lacey: In the next five years, storage systems installed on the customer side of the meter are going to outpace batteries on the utility side of the meter. So people have long talked about grid-scale, power-plant-sized batteries to control the grid as the future of the storage industry. But actually distributed storage, distributed batteries, are going to slowly overtake the market here in the U.S. And that of course shapes financing models, rate design, the taxonomy of the companies operating in this space, and how utilities integrate storage. So, we're going to try and get to all of those today. But largely I think we're going to try and focus on the players involved and how they're selling and integrating these systems. First, Brett, when we talk about behind-the-meter storage, what exactly are we talking about? What is getting deployed today?

Brett Simon: Great question, Stephen. I think that's an excellent way to frame the conversation. So really, when we talk about behind-the-meter storage, we are talking about storage systems sited on the customer side of the meter. So those are systems deployed either for residential customers, whether it is in a home or an apartment complex, or for what we term at GTM as non-residential customers, which includes everything from commercial and industrial customers, military installations, supermarkets, religious institutions, and so on and so forth. And those systems are used for a variety of use cases. But in general they are often used for either a resiliency play to offer some sort of backup power for the customer, or some sort of bill-management, which can either take the form of shaving demand peaks, which is of a lot of interest to these C&I customers, or for things like [time-of-use] shifting or to even offer grid services in some cases.

Shayle Kann: So let me offer a little bit of additional nuance on what "behind-the-meter" means. Because with energy storage, I think that as time goes on, it's going to be a little bit trickier to segment off these markets. As Brett mentioned, when we talk about behind-the-meter, and it's a term that's constantly used in the energy storage industry, most of what we mean is that it's a battery or an energy storage project, for which the primary application goes directly to the end customer in which the process is sited, so either that's an economic value, demand charge management, something like that, or that's a resiliency value. Either way, that's feeding the customer where the project is. The reason this is going to become more and more nuanced is, as time goes on, it's looking likely that a fair number of these projects, where the primary application goes to the customer, will essentially moonlight for grid services. So they will, in addition to providing a residential customer with backup, maybe they will be aggregated and provide some peak application for utility or ancillary services or something like that. So we'll have these behind-the-meter projects that are both serving some value to the customer and to the utility. We're still going to define those as behind-the-meter. The question is, what is the primary application? And if there's something for the customer, that's going to be called "behind-the-meter."

Stephen Lacey: It's an interesting way of describing it. And that actually gets to the question of business model. So in order to scale in behind-the-meter storage, do you necessarily have to look at grid services, in order to get a critical mass of customers?

Brett Simon: That's a great question, Stephen. I would answer that it seems to be the way that a lot of the big players in behind-the-meter storage are going. So you look at companies like Stem or Green Charge or Advanced Microgrid Solutions, and they are applying and winning bids on these large utility contracts, which are going to be aggregating behind-the-meter systems to provide grid services for one reason or another. We're seeing states like California, as well as New York, with utilities who are starting to very seriously look at how behind-the-meter storage can be leveraged to reduce peak or offer transmission or distribution upgrade deferral options. So, I think it's becoming increasingly interesting. And as we see these projects get implemented over the next few years, if they do indeed bear fruit, we'll see an increasing interest in leveraging these systems for grid services.

Shayle Kann: I'd make a stronger statement there. And this is just my personal opinion. I do think that in order for behind-the-meter storage to scale, if you are a developer, an integrator, something like that, and you're trying to think long-term about a big market for behind-the-meter energy storage, you absolutely need to be able to figure out how to be able to aggregate and how to play in wholesale markets or direct bilateral contracts with utilities. I think you can make an economic case for behind-the-meter storage, with falling battery costs and so on, that doesn't require grid services. But I don't think that business model scales nearly as fast. And I don't think the total addressable market and the customer demand is high enough that you'd end up with a market the size that you really want. So, in part, you mention Stephen at the beginning, is if this is a foregone conclusion that behind-the-meter will ultimately overtake front of the meter. In other words, the prediction that we've made is that by early next decade, there will be a year in which we will see, in capacity terms, more energy storage installed at the customer premises than we will for just big centralized utility type applications. That's not a foregone conclusion. That is absolutely a counterintuitive prediction, due to how big the front-of-the market meter is today. And part of the reason we've made the prediction is it's partially predicated on the idea that a lot of the behind-the-meter assets that get deployed will be value stacking, in other words, will also be providing some service to the grid.

Stephen Lacey: A big piece of this report is figuring out the go-to-market strategies and helping people understand how companies are developing projects. You've got a really great taxonomy in here. And I want to walk through how companies are approaching this space. So you kind of break down four different go-to-market strategies in behind-the-meter storage.

Brett Simon: There are some companies that are selling directly to the end customer. And we're seeing a pretty substantial number, particularly in the C&I space, who have an effective internal sales force and are working to go directly to the end customer. And this actually works well for some of the grid service pieces we talked about earlier. Because in some of these cases, the actually solution provider is working with a utility to actually find these customers, actually deploy these systems, and then work with the utility to say, "We can guarantee that you can use these systems for Utility A, B or C at these times."

Shayle Kann: Wait, so let me ask a question about that, because I didn't realize that was actually happening. Just imagine that you're Stem and you're working for Southern California Edison or something, and you submit a bid and win to provide some peak reduction in a particular area in Southern California next year. Is Southern California Edison then helping you find customers that you can go after to deploy these systems? Like are they actually helping to identify customers for their partners?

Brett Simon: My understanding is that, no, the utilities are not helping them. It's the companies themselves. And if I was unclear, I apologize. But the companies themselves are going out there and acquiring those customers, whether through their own internal direct sales force, or in some cases, they're working for either a solar developer or installers to actually find these customers and sign them up for these programs.

Shayle Kann: But that seems like missed opportunity ultimately. Maybe there's some privacy reason why you can't do this, but the utility has everybody's load information. And so the utilities should have the unique capacity to find customers in a geographic region who have particularly peaky demand, which is exactly what you're looking for if you're deploying these systems. At some point, you would think the utility, either for their own benefit or for their partner's benefit, should help in screening customers.

Stephen Lacey: Isn't that what Advanced Microgrid Solutions is attempting to do? They've basically said that they're going to be the face of the utility, work with the utility to find where they have capacity requirements. And then develop projects based on local needs, which is also a product of California's local capacity requirement, but also very different in terms of how they market themselves compared to some of their competitors doing behind-the-meter storage.

Brett Simon: AMS is a different case. And that's a good point, Stephen, because that company's entire value proposition is trying to essentially work with the utilities. And rather than going out and saying, "We're going to find the customer and later sign them up for grid services," instead they're doing the reverse. They want to supply the grid services first and foremost. So there is that work with the utility to identify key customers to work with and where to deploy systems.

Shayle Kann: Wait, we're talking about two different things here. If you're a storage developer and you're going to work with utility to deploy systems on the customer side of the meter to provide some grid service, the first thing you can do is you can work with the utility to identify areas within the grid where you can provide this service. And I think that's mostly what Advanced Microgrid Solutions has been talking about. They will work with utility to proactively identify somewhere geographically where the conditions of the grid are such that energy storage could provide that value. So that would be step one where they could work together. But then step two, it's still not clear to me whether this is happening or not, could theoretically be, "Okay, now we've agreed that Pomona, California is the place to do this. Now utility, dig through all of your customers in that location, filter for customers who have a load of this size that has this large of a peak or something like that and give those names to me, so I can go off and sell them." Imagine how much that reduces your customer acquisition costs if you could do that. But it's not clear to me that that could happen.

Brett Simon: To my knowledge, that isn't happening, Shayle. That first step, which you and Stephen both correctly identified, is occurring. But that second step, I don't know that that's happening.

Stephen Lacey: And the go-to-customer strategy is kind of a function of the nascent nature of the market, because you're doing a lot of one-off projects. There's a long sales cycle. You've got to approach each individual customer and figure out what their needs are. And it's only now that you have broader rate designs that allow you to maybe aggregate these projects. Or you have pricing that's favorable so that you can reach a broader range of customers. But ultimately, these early projects are one-off and it takes a while to get that customer to want to sign across the dotted line. So you have to work very closely with them.

Brett Simon: Exactly. And I think that's why we are seeing an evolution in some of these go-to-market strategies, that increasingly companies are looking at who they can work with, whether that be a solar developer or an installer, to bring their systems to market in another manner. Like you said, if you're doing these one-off projects, that can be very challenging. And as the market evolves, there will be increasing standardization. We're seeing that with an increase of modularity of systems, especially at the residential level, where in that case, going directly to customers is even more challenging and even more time-consuming. So there, really, you're not seeing so much direct outreach rather there is that move to work with installers.

Shayle Kann: Can we spend a minute to talk about what these companies are actually offering to customers? What that model looks like? In particular, what the financial value proposition is for a customer? We could maybe start with commercial, because this is a little bit more of a mature market. Can you just sort of walk through the primary mechanisms these companies are offering of value to customers?

Brett Simon: For sure. And I think the one that everyone comes back to immediately is demand charge management. When it comes to C&I customers, that's really the major driver of value for an end customer in C&I right now. And for those of our listeners that may not be aware, this is essentially where you would use stored energy in your system to discharge during your period of peak usage to reduce your dollars-per-kilowatt month charge, which is usually assessed on either a 30 minute or a one hour basis. Actually last year, GTM put out a report on actually the economics of C&I demand charge management and found, interestingly enough, that in the next five years it will become increasingly viable in states where traditionally we haven't seen deployments today, so a state like Michigan, where we haven't necessarily seen massive deployments today.

Other value streams of interest to C&I customers is time-of-use shifting, and that involves storing energy during off-peak periods and discharging it during peak periods, where in that case your volumetric charge or dollars-per-kilowatt-hour is assessed at a different rate for your peak and your off peak times. And then, the third piece, which is kind of hard to assess the economic value of, is resiliency. There is this interest and need for resiliency. But there has to be kind of balanced against economics a lot of times, because, even today, diesel gensets are much deeper compared to energy storage. And then the last piece would be, of course, grid services, which we talked about earlier. And, in some cases, in the contract there will be some way that the end customer can actually share grid service revenue with the actual system operator.

Stephen Lacey: So those are all the value streams that customers can get from these assets. Let's talk a little bit about how they're financing those assets. And how the actual cash flow works for the customer.

Brett Simon: That's a great question. So, one of the models that kind of ties pretty directly into demand charge management, is the shared savings model. So we see players like, Green Charge gets a lot of press about this, actually having this model, where effectively you make a contract with your customer. And you effectively say, "We will cut your peak demand charges, and we will share the savings." So, some percentage will go to the company like Green Charge. Some percentage will effectively just be saved by the customer. And they'll keep that.

Shayle Kann: So the customer pays $0 upfront and gets, let's just say, half of the savings that come from discharging the battery at peak times. So it's $100 in savings. The customer gets maybe $50 dollars in savings. They pay nothing upfront. So it seems like a total no-brainer. Why wouldn't you do that? Where there is no risk there. I think it's worth noting that there's some risk. And the customer generally does bear at least some of that. For example, rate risk. You might not get those savings. If utility changes the demand rate, and your demand charge cost goes down, the savings may be lower than expected. Still, it seems like a pretty good deal for the customer. Is there any way, Brett, for the customer to end up in the red in a cost-saving transaction?

Brett Simon: When the contracts are signed with the company, I think there's generally a certain guarantee of saying, "We will share a certain percentage of savings." I believe if the customer breaks the contract by either vastly changing something like their electricity consumption or switching something like their electricity rate, there can be penalties assessed. Though, to be honest, a lot of the contract specifics are kept pretty close to the chest by these companies and aren't generally given out even to handsome and smooth-talking analysts like myself.

Shayle Kann: But still, I think shared savings, what it's got going for it is it looks to the customer like the lowest risk cost-sharing option available. It's interesting that shared savings exists in energy storage before it exists in solar. It's something we've talked about for solar for a long time. It has not yet been deployed. I think there's a couple companies that are trying to pioneer this. In solar, you can sign up for a power purchase agreement or a lease or a loan, any of which puts the risk on you. If you don't actually have savings because of your solar, you lose that money. So why not offer something like shared savings in solar? And I'm curious, Brett. A lot of these companies are looking at deploying solar-plus-storage. Have you seen any of them talking about doing a shared savings model that incorporates solar?

Brett Simon: That's a good question. I have not yet seen a solar-plus-storage shared savings model. Though in some of the cases where storage is deployed for shared savings, the customers do have solar. But the thing with shared savings is, in these cases, a decent amount of it is actually in deployed standalone, because there isn't necessarily that need to have a solar system and store that solar energy. If the rates are set such that the demand rate is high enough, the customer can actually get away with simply storing energy from the grid and then discharging it at a peak time.

Shayle Kann: And then final question on peak savings, and then we can talk about the other models quickly. And I know this is one that Stephen probably could have an opinion on, because you've looked at the issues with measurement valuation and energy efficiency. And I think with something like shared savings in energy, you probably have a similar challenge, which is that you're measuring savings against a hypothetical. You're saying, "How much do we estimate that we've saved you, because we've deployed this energy storage. Because we've discharged at this time. What would your peak demand have been and what would that have been charged at? And let's give that to you as your savings." Do you see that from a customer perspective as something that gives them cold feet, that they'd just have to trust that the savings are what their developer told them?

Brett Simon: Certainly that's something that has to be understood. And that's why the measurement and verification piece is so important, and why the modeling that comes before the contract can also be pretty important, because you're right, Shayle. You can never know for sure what the customer's peak would have been. There's all kinds of factors that can change. What if in year one before the system is deployed, there's a heavy storm event and the customer has to use more of their stored energy? What if the utility has a rate case that doesn't seem like it'll go through, but then it ends up getting approved in the fifth year after the system is installed? So you can never really know for sure. And I think that is something that might give some reticence to the financiers as well. Because on that front too, it's hard to know exactly what the cash flow will be from the system in the future, even as good as the modeling could be.

Stephen Lacey: One of the models that is not very clear to myself and to Shayle, is the PPA-like financing structure for behind-the-meter commercial storage. What are people offering or what are people proposing for a PPA-like structure?

Brett Simon: That's a great question. So PPAs have taken off pretty substantially in solar. But when it comes to a storage, what I term the PPA-esque model, it's almost more accurate to call it a solar-plus-storage PPA-esque model, because I don't know anyone who's doing this in absence of having some sort of solar as well as part of the system. But with that aside, essentially solar PPA is you make an agreement, you agree to a certain price, dollars per kilowatt hour, for that energy. For the PPA-esque model, for storage, it's a little more complicated. Because, for storage, you almost have to think about what you can get from that system in terms of dollars per kilowatt savings. So this model, which I admit is still in its early stages and still isn't as widely deployed, essentially the modeling works that the company will come to the customer and say, "Okay, we will take your full electric bill. We take your final demand cost, which is the dollars-per-kilowatt cost, and the energy cost, the dollars-per-kilowatt-hour cost. And we'll bundle that into a single dollar value. And we'll divide that by your kilowatt-hour consumption per month to effectively get your bundled demand and energy cost." Following me so far?

Stephen Lacey: Yeah, I think so.

Brett Simon: So taking that value, you now have the dollars-per-kilowatt hour value that's bundled. And essentially the company says, "Say that combined value is 20 cents per kilowatt-hour. We will offer you an agreement with solar storage that equates to you say 18 cents per kilowatt-hour, or something similar." Usually with some kind of escalator over the coming years. So effectively it's trying to account for both the demand savings value of the storage, as well as the kilowatt-hour energy saving value of the solar, and bundling those together into a single dollar-per-kilowatt-hour price.

Stephen Lacey: So what's attractive about that to a customer, presumably, is the visibility into their own cost. In other words, it's similar to a solar PPA on its own. What you do know is you'll get these things put on your roof or your wall or whatever. And you'll pay 18 cents per kilowatt-hour for all the power that's produced or discharged. That's the benefit. But, of course, you don't get the certainty of savings that you would get from a shared savings model.

Brett Simon: Exactly. And because it's just that whatever rate you agreed to, with the shared savings model you're discharging against the peak demand charge. So it's a little more straightforward in that way on the storage piece. But one advantage of the shared savings model is there is also that value that people who are familiar with solar PPAs may be more amenable to it, because they say, "I've done solar PPAs for however many years. This is effectively the same thing, just adding on storage."

Shayle Kann: And I wonder whether, if you're a solar developer, many of whom are now looking at storage, and you want to go back and retrofit and add storage to an existing customer, are you seeing those companies talk about just wrapping the cost of storage to the existing PPA, bumping up the price of the existing one?

Brett Simon: I have not seen that yet. I have not heard anyone talking about doing this as a retrofit. Any of the instances I've heard talking about the PPA-esque model is all for new installations. I don't know that that would be impossible. But it seems like it would be difficult value proposition to sell the customer and also might be kind of challenging to structure that kind of a contract.

Shayle Kann: You've got an existing financing structure for those projects. Probably hard to stop in the middle and do something different. All right, so then let's talk about the last major model that's being offered to customers, which is the lease.

Brett Simon: The lease or some companies call it a subscription model is effectively a customer has the system in their home and they pay a fee to the developer every month for that system. And any benefits they get from that system are theirs. So unlike a shared savings model, if you have a really good month when you shave a lot of demand, that savings would all accrue to you. But then the converse is also true. If there is a month where the savings wouldn't necessarily be as high, you might be paying a subscription price that's a larger percentage of your overall savings.

Shayle Kann: Since I'm in the mode of talking about what I think may be attractive about each of these options, one of the benefits I think of the lease model is that it's probably the easiest one to finance, because you have fixed guaranteed cash flow coming from the customer. So on the back end, if you're the developer, and you need to finance the asset, that is going to be the easiest thing to finance. It's probably easier for you to raise capital. You might get a lower cost of capital that way. That could flow through to the customer. So despite the fact that shared savings sounds theoretically more attractive to a customer, because it provides some certainty and potentially lower risk, economically it might end up being a slightly worse deal if it ends up being so much easier to finance a lease.

Stephen Lacey: So for the most part this entire conversation has revolved around C&I storage because that's what dominates the market. But the big question is what happens in residential, wen we stop getting behind pilot projects. What is residential going to look like through 2021? And what kind of financing models will we see? Will these systems be providing grid services? Or will we just see simple leased systems? Help us understand where residential fits into this picture.

Brett Simon: Sure. And I agree, Stephen, residential is definitely the murkier piece behind-the-meter right now compared to C&I. And I think one of the big challenges with residential is there's a lack of clearly monetizable value streams. So we've talked a lot about the opportunity for demand charge management, which is a monetizable value stream in a lot of cases for C&I systems. But for residential, there's nothing that would really clearly say to a financier, "This is a system or a project that I should get behind right now."

I think something that is increasingly worth watching is residential developers and system integrators who are trying to work with utilities to actually deploy systems that can be networked and provide grid services. We've seen Green Mountain Power, the utility in Vermont, actually deploy Tesla firewalls and have the opportunity to actually aggregate them to lower peak. We've seen Sunverge do a number of projects working with utilities. We've seen one with Con Edison in New York. And we've seen one in Glasgow EPB in Kentucky. And we're increasingly, interestingly enough too, seeing this interest in utilities and trying to figure out how they can offer storage to their end customer. We're already seeing this in countries like Germany and Australia, where utilities are actually getting fairly seriously into the storage game and actually selling or leasing systems to their end customers. And similarly, we are seeing residential storage winning bids, not in a large number of utility procurements, but at least it's starting. For instance, the PRP in California, there was a 20 megawatt hour award to small energy. And all of those systems are going to be behind-the-meter residential systems.

So to come back to your question of how will these systems actually be financed? That's a great question. I think shared savings is pretty much out for these residential systems. I mentioned earlier demand. There are residential demand charges in a few locations. But it's very unlikely to become widespread. And even so, the economics probably won't pencil out for doing demand charge management alone. But I think the interesting thing to look out for residential will be either leasing or some kind of loan financing, where at the end the customer will end up owning the assets, because that speaks a little bit more to how the customers at the end of the day will want to have control of their systems and use it either for self-consumption or in some cases back-up.

Stephen Lacey: The report is called the behind-the-meter energy storage landscape 2016-2021. And you can get all sorts of really fantastic charts and data at gtmresearch.com. With Shayle Kann and Brett Simon, I'm Stephen Lacey, and this is The Interchange, a weekly conversation on the changing business of energy and cleantech. From GTM Squared. We'll catch you next time.