Douglas Dillon Fellow and Acting Director of the Program on Energy Security and Climate Change, Council on Foreign Relations
Director, Solar Energy Technologies Office and SunShot Initiative, Office of Energy Efficiency and Renewable Energy, U.S. Department of Energy
Senior Vice President and Head of GTM Research, Greentech Media
Chief Executive Officer, MAC Energy Advisors LLC
The cost of solar energy has plummeted in recent years, and solar is the fastest growing energy source around the world. Experts discuss innovations in solar technology, investment opportunities in the renewable energy industry, and the effects of solar power innovation on the U.S. energy portfolio.
The Emerging Technology series explores the science behind innovative new technologies and the effects they will have on U.S. foreign policy, international relations, and the global economy.
MCKIBBEN: Good afternoon. Welcome, everyone. This is the Solar Power and Clean Energy Innovation panel discussion that’s part of the Council on Foreign Relations Emerging Technology Series. We’ve got a wonderful group of panelists, very knowledgeable about the subject that we’re going to be covering.
You have all of their bios, so I won’t go through it. We’ve got Charlie Gay to my immediate left. He’s the director of Solar Energy Technologies Office and SunShot Initiative in the Office of Energy Efficiency and Renewable Energy at the U.S. Department of Energy. Next to him we’ve got Shayle Kann, who’s the senior vice president and head of GTM research at Greentech Media. And then last but not least we have Varun Sivaram, who’s the Douglas Dillon fellow and acting director here at the Program on Energy Security and Climate Change at the Council on Foreign Relations.
I thought we’d start the afternoon talking specifically about where the sort of solar technology is and where it’s been. And so we’ll have—our panelists will give first a couple—five minutes about their thoughts on where the technology has been, where it currently is, and the developments around the technology currently today. So if you’ll start it off, Charlie.
GAY: Sure, thank you. And thank you all for coming here, and with your interest in solar. I’m always happy to talk to people who care about solar. I’ve been working in solar for forty-three years. I started right after the time of the first oil embargo and thought this might be an area to get involved with. Most of my time has been in the manufacturing sector. They’re working to scale up technology for providing the equipment that people use to manufacture solar. I spent some time during the first Clinton administration as director at the National Renewable Energy Lab, and joined the Department of Energy about five months ago, just to give you a calibration on things.
MCKIBBEN: He picks his timing very well. (Laughter.)
GAY: Yes. Part of the backdrop here I think that I want to convey today is the speed of change. It’s—we’re at a tipping point here where the rate of change is very rapid and a lot of the system, the utility grid systems infrastructure that we’ve had in place in America since Samuel Insull, Thomas Edison’s secretary, laid out how ConEd could work and provide monopoly power to be able to aggregate customers. Now customers have choices. Over the course of the past year, the combination of wind and solar account for almost 60-plus percent of the new electricity generation being installed here in the states. And the rate of growth happening around the world is comparably large.
Much of what I’d like to reference here in sort of gauging change—people talk about incremental change, broadly speaking, and things continue to change incrementally until they don’t. And we’re at a stage where the incremental change is shifting over to a whole new wave of innovation, all the way across the supply chain from upstream and new materials that can be lower cost and higher efficiency in solar, to downstream ways of integrating distributed energy resources—solar, wind—and energy efficiency improvements that can be a part of our grid.
I’ve had a chance in my career to work—most of my career has been in rural areas. I’ve worked in over 80 countries installing rural systems for power and villages and oftentimes bringing the internet to the village as a way of helping that convergence of energy and information, where—that nexus is where we are today. And in thinking about how information can be utilized in the grid, essentially we’ve invested much of the time in wires. Over 40 percent of the capital asset base in the grid is wire, whether it’s in the form of the transmission or distribution network, or the rights of way associated with that. We can use those wires a whole lot better with distributed sources.
So at DOE among the things that we do are longer-term R&D. Since the time I got involved in solar, our basic science work is no longer done by corporate laboratories—Bell Labs or RCA Labs. It tends to be done in government laboratories. There are seventeen national labs within the Department of Energy’s span. And principal among them include the National Renewable Energy Lab, Sandia Lab, Pacific Northwest Lab, Lawrence Berkeley Lab, Lawrence Livermore, who are all engaged with grid modernization. And the reason that I joined DOE was to help participate in grid modernization, where the Office of Renewable Power and Energy Efficiency work together with the Office of Electricity. I look forward to having a chance of chatting further about a lot of the business opportunities that are emerging here as we bring these pieces together. Thank you.
KANN: Great. Thank you all for coming. Good afternoon.
So a brief way of introduction into my perspective on this, Greentech Media has been around for ten years, not quite as long as Charlie’s forty years in the industry, but a long time for solar nonetheless. We do two things, basically. We track the day-to-day movements of the broad sort of clean energy/green energy sector via a media business. Then we also have a market analysis and consulting operation, which is the part that I lead, where we focus on trying to understand trends within both the solar industry—sort of end-to-end, from manufacturing to demand and project finance—and also how that impacts and reverberates across the rest of the energy sector.
And the sort of underlying thesis under which we operate and try to direct ourselves is that the energy sector in general, and particularly electricity, is in some part of three simultaneous, interrelated though not identical, transformations, all of which are ultimately going to totally reshape the way that we deliver and use energy, both in the U.S. and globally. First of those transitions is decarbonization, the transition from an electricity sector that is primarily based on fossil fuels to one that has at least an increasing share, and presumably ultimately a majority share, of zero emissions resources, renewable energy.
Second is decentralization, moving from an electricity system that hasn’t actually changed all that much in the past century—the basic architecture still being the same as it was designed originally, which is you generate almost all of your power from a small number of large power stations. That electricity then goes through transmission lines, then distribution lines, to a largely passive customer base. Turn on the lights, they get the power, they pay for the power.
As Charlie alluded to, that’s changing both from a perspective of where electricity is generated—increasingly we’re seeing distributed energy resources—be they generating resources, load modifying resources, or energy-storage resources—but also the way that consumers interact with the grid. They have a lot more options now than they used to. They have different ways that they can manage their electricity. And so that’s causing all sorts of changes, both in the actual infrastructure of the grid, the way the utilities interact with customers, and the regulatory compact. So that’s a second transformation.
And the third is electrification, the transition toward things like electric vehicles. Solar, of course, lies really central to at least the first two of those three. Talking about decarbonization, you know, the vast majority of that is solar and wind, and you can have a conversation about the role that nuclear is or should play in that as well. Decentralization, solar is the base of that, because solar more than any other resource is completely scalable. You can scale it down to a single, tiny, you know, consumer electronics purpose, or something for an off-grid village in a developing country, all the way up to the size of centralized power stations. We’ve seen now one gigawatt-plus solar projects developed in India and China. We have, you know, 5(00) or 600 megawatt projects in the U.S.
So solar is central to two of those three. And as I think we’ll talk about maybe a little bit later, benefited by the third, electrification of vehicles. Where I think we stand in solar now—I think it’s completely accurate to talk about how rapid the pace of change has been. You know, there was a big wave of solar innovation that Charlie was a part of in the early days, in the late ’70s and early ’80s. And then there was continued innovation, but not a whole lot of solar getting built out for a long time, until the 2000s, during which time things started to accelerate pretty drastically. And the result of that has been something that I don’t think everybody recognizes if they’re not on the inside of the solar industry, which is that in a lot of cases today solar is the least-cost resource to build new electricity-generating capacity.
That is part of why solar and wind account for more than half of all the new capacity that is getting built in the U.S. And you’re also seeing solar, wind tenders that are technology agnostic, without any incentives, in other countries globally. So suddenly solar is actually cost effective. Another way to look at that is the SunShot program that Charlie leads, when it was originally started in 2012 set out what was, at the time, supposed to be a really ambitious goal for the cost of solar, which was let’s get to, you know, among other things about a dollar a watt installed for our utility-scale solar projects in the Southwest by 2020. Dollar a watt by 2020 was the goal in 2012. And we’ve seen projects getting built in 2017 that are already a dollar a watt. So we’re hitting those targets and we’re hitting them early, in part because of the SunShot Initiative, in part because of global dynamics and technology innovations.
Solar’s cheap now. But being realistic about it, solar just crossed the threshold of generating 1 percent of all electricity in the U.S. last year. It’s still a negligible share of the total pie. And the reason that that matters, as I’m sure Varun will talk about it as well, is that you don’t get to just stop when you hit a dollar a watt for solar, or any particular benchmark that makes solar competitive today. What solar has going against it is that it’s not dispatchable, it’s not controllable. So the more solar you put on the grid, the more you have projects generating power all at the same time. It becomes less valuable. So it’s called the value deflation effect. It’s not a big deal at 1 percent penetration where we are today, but it starts to become one as you get to 10 percent and 20 percent.
Basically, that means solar needs to keep getting cheaper, and it needs to keep getting cheaper and it needs to keep interacting with the grid better. And finally, you need to sort of—if you’re serious about getting a lot more of it on the grid, you need to start to redesign how you operate the grid and how you design power markets to enable the things that will be complementary with solar to thrive. So it’s a really interesting place now, where sort of the cost question is not done, but we’ve achieved a really important benchmark that was set out decades ago and, again, years ago. And yet, there is still continued need—in fact, maybe more than ever need—for further innovation and further cost reduction within the industry.
MCKIBBEN: Great. Thank you. Varun.
SIVARAM: Thanks, Tracy.
Let me start by saying how excited I am that Charlie and Shayle are both here with us at CFR. Charlie is an absolute pioneer of the solar industry. And I turn to him all the time for guidance on my own research. And Shayle and I have been close collaborators, most recently on a paper where we made the case for why investments in innovation are very important today to avoid solar kind of hitting a wall decades into the future. So that’s where I want to pick up from where Shayle left us.
I argue, and I’m writing a book on this subject, that solar is the single most important energy technology of the 21st century. And by midcentury, solar will need to account for at least a third of the world’s electricity. And that share will need to increase in the second half of the century. Those numbers are enormous. So Shayle just told you that solar crossed the 1 percent generating threshold in the United States. And I think globally solar is on the order of 2 percent of global electricity. We’re talking about 30 percent or more by midcentury, and that’s even taking into account the electrification trend that both Charlie and Shayle brought up, that more and more of our global economy will start to depend on electricity. For example, as more electric vehicles come on the road.
So we’re talking about between one and two orders of magnitude of increase in solar power in order to meet the world’s goals, for example—for example, of limiting global climate change, of improving energy security, of improving energy access and of powering continued economic growth, especially in the developed world—the developing world. So the recent progress that Shayle and Charlie mentioned, in particular the fact that solar is now the least-cost resource of any resource in many parts of the world, that is very promising. And really, something remarkably has happened in the last five, 10 years. The solar industry has graduated from being this niche, cottage industry to being a global powerhouse.
But that isn’t grounds for complacency. I argue instead that we’re actually at this very precarious moment, that if we stop investing in innovation we could see a similar movie play out to what happened when we looked in the 1950s and said, look, nuclear power is very soon going to be too cheap to meter. It will solve all of our power problems. Instead, nuclear power happened to peak in the 1990s as a share of world power production, and has declined ever since. For completely different reasons, the same thing could happen to solar. So I argue that there are three types of innovation that are badly needed and that we need to invest in in order for solar to achieve this potential of 30 percent by 2050 and even more by century end.
The first kind of investment builds off some of the things that Charlie and Shayle just talked about. It’s business model innovation. In particular, what worries me is that the existing sources of capital that have financed solar’s rise to, today, a $100 billion global industry won’t be sufficient to continue to finance its rise in the future. And Bloomberg estimates, for example, that $2 ½ trillion of additional capital is going to be needed by 2040 for solar to get to where it needs to be for the world to meet, or at least stay on track, for its climate goals. That money is not going to come from existing investors. Those existing investors including, for example, private equity firms. They include in the United States tax equity investors. And they include multinational development banks that have done a lot of great work in the developing world.
In the future, however, when we are at the trillion-dollar scale, we will need institutional investors. And institutional investors need liquidity in the securities that they trade. And they can’t be investing on a project-by-project basis. Innovation to package up solar in a way that’s palatable to these large, colossal investors is the first crucial step to scale up solar from where it is today, 1-2 percent, to at least 10 percent or more. But beyond that, Shayle alluded to a dynamic that won’t allow solar to continue growing, even if we get more capital into the mix, because in addition to business model innovation we’re also going to need new technologies.
A theme that I haven’t heard yet, but I want to emphasize, is that the solar industry has coalesced around one technology: silicon. It is an almost homogeneously silicon-dominated industry today. But there are many, many technologies on the horizon. Now, Shayle and I have a bet going on that he thinks I can go the entire panel—
MCKIBBEN: Go ahead. Do it. Do it. (Laughs.)
SIVARAM: —without saying a particular word of a technology that I’m really excited about. So at least in these opening remarks, I’m not going to say it. (Laughter.) But I dare someone to ask me about—
MCKIBBEN: I’m going to ask you. It’s coming. (Laughter.)
SIVARAM: So I argue—actually, Shayle and I both argue that without investment in new technologies, as you put more solar on the grid and it undercuts its own value, it will not be able to get cheap enough to outrun this value deflation effect. What’s more, solar electricity alone cannot power the world. Or, at least, cannot supply the world’s energy needs. That’s because, for example, at least 40 percent of the world’s transportation needs are non-electrifiable. We’re talking about shipping, heavy-duty trucking, aviation, that is very difficult to run off the electricity grid. You instead need liquid fuels. And so I argue for innovation not just into new solar power generating technologies, but also solar technologies that can fuel our other sectors.
Finally, even those two types of innovation won’t be enough. If we really want to get, by the end of this century, to a world that is largely powered off of solar power, we’re going to need systemic innovations. That’s innovation to the way that we consume energy in all forms of human society—not just the power grid, although that’s an important place to start. We need grids that are more flexible, that can store energy, that can put up with solar that you can’t control and dispatch. But on top of that, we’re going to need to be able to power our heat needs, our mobility needs, using solar energy somehow, by harnessing sunlight. We’re going to need to get our food and meet our water needs by using energy from the Sun. All of these couplings to different sectors are going to require systemic innovations that transcend just innovation within solar materials, though they include those innovations as well.
So I’m really excited to talk to you about the many different types of innovation in solar. I hope by the end of this session everyone will go away and know what a band gap is, if you don’t already, because I think that without investment in innovation, the optimism that solar has deservedly earned today will not be justified a decade into the future, that solar could really hit a penetration ceiling. And it is, frankly, humanity’s best hope right now for this century. So with that, Tracy, back to you.
MCKIBBEN: Great. No, thanks, Varun.
I mean, I think as you’ve heard we’ve talked about—particularly, you know, from Charlie’s perspective as well as the market perspective that Shayle provided, and Varun touching a little bit on the innovation—you know, we’ve—as you each know, we’ve seen the sort of swan dive in pricing around solar. At the same time, you know, as the price has gone down—we started at, you know, over $100 a watt and now in some states we have less than a dollar a watt. Insulation has grown as the prices have come down. What I’d like your thoughts on, on going forward, what’s the innovation that’s still needed? Varun, you talked a little bit about that. And I’m going to try to get that word out of you that you committed to not saying tonight. But what—as you look at—we’ve got silicon manufacturing down here in the U.S. We’ve improved that, which has led to some of the price increases.
So I think the focus now, we would agree, is on efficiency and how do you do that. So if you would talk a little bit about, Charlie, from your experience, you know, how do you create that innovation drive? And what’s needed from a scientific perspective? And, Varun, if you could talk a little bit about the kinds of new technologies that are out there that’s going to create these kinds of efficiencies that I think you need to see in the solar development.
GAY: So one of the things that is sort of touchstone for me is a learning curve. Solar and wind are manufactured goods. And manufactured goods follow a learning curve. The learning curve for transistor cost reduction is almost 30 percent. For every cumulative doubling of transistors, costs come down by that 30 percent number. In the case of solar, over the past 40 years that learning curve has basically been for every cumulative doubling costs have come down by 20 percent. As we expand the use of solar, those costs are going to continue to come down through that learning curve, either by new innovations or by economies of scale and efficiencies that are associated with logistics here.
Solar is like a building product. We do work with a semi-conductor material, but I haven’t had to know much about band gaps in my entire career. What’s important here is manufacturing. And it’s a lot like the glass industry. The siting of factories close to end-markets so time—just in time delivery. Logistics costs of moving product, those are the kinds of things that will see factories proliferate around the world as markets establish themselves.
Solar, we’ve been blessed by having markets that are elastic through a span of time. When I initially started, most of the applications were on tops of mountains and places that were difficult to get to, but they were a solid business that we could grow on when the costs were high. The next market that came on our radar was for rural village electrification, where people lived without electricity. I grew up on a farm where we had a wood-burning stove in our kitchen. So I appreciate what it was that was important about solar, to be able to bring to a rural village. That market had a great elasticity to it. And we were able to continue to scale. And with the scaling, costs came down in that predictable way.
Interestingly enough, when we went from servicing mountaintops to servicing rural villages, we realized that we needed to have a sales model. We didn’t have one of those. We looked at Caterpillar, which was providing a capital good in rural, isolated places. And they followed a—basically a two-step distributor-dealer model. We emulated our business model for servicing the customer base that we had internationally around Caterpillar’s model. Today, Caterpillar has teamed with First Solar, for example, putting microgrids in rural villages. Those markets have continued to grow. And the microgrid technology fits into the technology space in urban areas, where most of the business is today.
And so that technology pipeline that Shayle referred or, or was defined at SunShot in 2012, had six-cent-a-kilowatt-hour electricity as a target for 2020. We’re now 90 percent of the way there, with seven cent a kilowatt hour as a reference point today. So when I joined DOE, we upped the ante here to get to 3 cents a kilowatt hour, and do that by 2030. So we have a set of tasks that we pursue related to technology improvements that can accommodate the pressure on cost reduction by improving efficiency. There’s a lot of new materials that are coming in. A-students today go into solar. They aspire to be in solar. When I graduated, A-students became nuclear engineers. I was somewhat of a slacker, and looked at solar as a good place to go.
And today, those folks have great ideas for how to improve the performance of materials. The next thing that’s really important here is to reduce the risk. Now, the risk when I got started was how long a warranty you would put on a product that didn’t have a history of working any more than sitting on a satellite in a hangar in Huntsville. Now there are modules in the field that I built forty years ago that have the same output today. Every morning I wake up, I wish I had the same output potential—(laughter)—as I had forty years ago. But the existence proof is there. It’s possible to build product with a predictable performance out over a long time horizon, twenty, thirty, forty years, or longer. There’s no moving parts here.
So in our technology portfolio, we see being able to accommodate the demand growth here in the U.S., by continuing to drive down costs and improve the reliability and serviceability for systems that are deployed. In the wave after that, we do need dispatchability. We do need some form of storage. The grid can accommodate a great deal of solar, as Hawaii has proven out—after modeling was done about a decade ago. The models have actually performed exactly as projected for accommodating upwards of 30 percent solar into the grid that exists. An island that—a series of islands that require diesel fuel to run the generators.
So there’s an innovation pipeline that is part of what we do R&D on in SunShot. There are nearer-term things that we do that relate to reducing risk, like aggregating data from the field around reliability, so that the performance can be projected. And we work with IBM, for example, at Watson, on the ability to take solar forecasting and couple that with demand forecasting. And Watson is basically a learning machine. And that learning machine, applied to optimizing the grid, is where a huge raft of technologies are emerging.
MCKIBBEN: Great. Varun, if you want to talk a little bit about the technology innovation.
SIVARAM: So Charlie gave us this narrative of, look, if you’re a business, you really don’t need to worry about this device. It just kind of sits there. No moving parts. If you’re the business, you really don’t even need to know what the band gap is. What you should focus on is how do you make it more cheaply, because over time you’ll incrementally reduce the cost. And you should also focus on making it last really, really long and giving a great warranty. And I think that indeed is the dominant narrative in the industry. But panels are more fun when there are disagreements, and I want to voice my extreme concern over this narrative.
Because—and by the way, this is completely accurate. For example, many solar firms spend, you know, less than a percent of their revenue on research and development, right? This is no longer a science business. This is a manufacturing at scale business. And that is tremendously worrying to me. One reason it’s worrying is because there is this growing rift between the industry, which is manufacturing these commodities—and by the way, because of the commoditization, these products are now mostly made outside the United States. Solar panels, the vast majority are made in China, Taiwan, and Southeast Asia, not in the United States anymore.
I see this rift growing between the industry and academia. In academia, where I spent some of my time working on solar technologies, we see tremendous progress and new work being done. And it rarely gets translated into new products in the marketplace. And this is a dysfunction that I think is not unique, but is particularly bad in the solar industry. And I think that so long as the solar industry decides to abstract itself away from the core technology and say, we’re just going to focus on making these things last long and making them slightly cheaper, they’re foreclosing really revolutionary routes to not only cheaper and better solar product, but brand-new solar markets that they haven’t dreamed of.
So let me try and give you a flavor of these new solar products. In order to do so, I’ve got to do a little bit of science. So solar photovoltaics, this is converting solar energy into electricity, the name of the game is how can we convert a bunch of incoming light photons, like packets each of which has some energy, into a stream of outgoing electrons, each of which has some energy? I want to take in as much light as possible, and out I want to get as many electrons with as high an energy per electron as possible. It turns out that conventionally there’s this tradeoff you face, because depending on this property of your semiconductor called a band gap, you can either absorb a lot of photons but turn them—and turn them into a lot of electrons, but each electron will only have a little bit of energy. Or, you can absorb very few photons and turn them into few electrons, each with a lot of energy.
In addition, you want to make sure that when you turn your photons into electrons, you can actually get the electrons out of your solar panel. And in order for them to get out of the solar panel without bumping into one another or getting lost—and so coming out at a high energy—you need a very crystalline material. So these are our two kind of parameters that we play with. We pick a semiconductor and we try and tradeoff between how many photons to absorb, and then we try and make it highly crystalline. As it so turns out, silicon has done a pretty good job of this. Silicon absorbs a fair number of photons and puts out electrons at a reasonably high energy. And because we heat it up to 1,000 degrees or more, it’s highly crystalline. That means that the electrons can easily get out of the solar panel without bumping into each other. Terrific.
However, there are new developments in laboratories that kind of shatter this tradeoff. They say, no longer do I have to heat up my materials to 1,000 degrees. And no longer do I have to content myself with just a single semiconductor that trades off between how many photons you get in and how much energy each of the electrons are that you get out. These new materials—I’m just going to say it—
MCKIBBEN: Yes, please.
SIVARAM: —are called—well, one of these new materials is called the perovskite. And perovskites are one of a class of materials. Other sorts of materials that can do this are quantum dots or organic solar cells. These materials can be inkjet printed, like a—you know, like you might print a normal page of typewritten words. And in addition to printing them, they can be flexible. They can be partially transparent. You can make them colorful. And most importantly, you can stack them one on top of each other with different band gaps, so that you capture as much of the solar spectrum as possible and output as many high-energy electrons as you possibly can.
These materials shatter the tradeoffs that we have historically kept, that you could only have a brittle, mottled blue and blue panel, and it would do a certain efficiency level. Moving forward, we can take a perovskite layer and put it right on top of a silicon layer. And that will make a solar panel that’s more efficient than anyone’s ever seen before from the commercial solar industry. Moreover, we might be able to make window coatings. We might be able to coat literally anything—be it a military tent or a car. And very far down the road, we may also be able to harness brand-new solar technologies, for example, using a technology called catalysis to turn incoming sunlight into a liquid fuel.
So there are these horizons for brand-new technologies. I want to push back against the narrative that the name of the game in the solar industry today should just be take the very simple, very well-understood existing technology, and let’s just make it cheaper through scale, and give it a really long, 50-year warranty. That, in my opinion, would be the most disastrous thing that we could do for the future of the solar industry.
KANN: May I—may I just offer one additional word of caution, apart from just crowing about having made Varun say perovskites. (Laughter.)
KANN: Which is, we have a tendency—and it’s a totally reasonable tendency. But we as an industry, and certainly anybody who’s not spending all day thinking about solar, have a tendency to talk about the panel, and assume that that’s what matters—that’s all that matters, right? And certainly that is what generates the energy when you install solar. The problem with that thinking is that the reality is that even today, because panel costs have fallen as far as they have, the panel represents a minority share of the total cost to put solar on a roof or in a field. You might have—so, let’s just take that dollar a watt. So this is a utility scale installation. This’ll be, you know, on the ground somewhere out in the Mohave Desert, right? It costs you a dollar a watt to build. You can buy panels today for 40 cents a watt or less, meaning 35, 40 percent of the total cost of your system is actually coming from the panel. The rest is coming from a bunch of different things.
There’s other hardware. There’s a(n) inverter that converts the DC electricity the panel outputs to the AC electricity that’s on the grid. There’s a racking system that props the solar panel up. It may be tracking. So it may track the Sun either on one axis or two. There’s cabling. There’s wiring. There’s a combiner box. There’s electrical equipment. Then there’s also a whole variety of soft costs, the things that aren’t actually part of the system but you do have to do in order to get one of these things built. So there’s system design, there’s engineering, there’s financing costs. If you’re talking about solar that goes on somebody’s rooftop, there’s a customer acquisition cost, which is a huge cost in solar right now.
And the problem with all of this stuff is that no other individual cost component adds up to nearly the cost of the panel. So if you want to talk about how to reduce the cost of solar, it’s easy to say, let’s focus on the panel. But the reality today is, let’s just say you had a 20 percent reduction in solar prices. And with a 35 cent a watt panel, you’re down to 30 cents a watt. You’ve achieved 5 cents per watt in total cost savings. That’s 5 percent of total system costs. That’s not going to get you there, certainly not to where we’re talking about needing to be by midcentury.
And so the problem is, you actually need to focus on innovating and reducing costs in every single element of that cost stack, many of which come from pennies at a time or less than pennies at a time. You’re operationally optimizing. You are focusing in system design. You’re figuring out better ways to acquire customers. All of these things have to happen in order for solar to get cheap enough that it can be 30, 40 percent economically on the grid. It can’t just be a focus on panel technology, as important as that is.
MCKIBBEN: Great. Thanks, Shayle.
So that points me to—I’d like your thoughts on, as you look at—particularly focusing on the sort of soft cost issues and how you address those, but not leaving out, you know, future—you know, innovation and where the technology is going. I’d like your thoughts on sort of where are the opportunities, you know, for businesses, and how—you know, whether it’s a policy perspective, that, Charlie, you may be able to provide us some information on, and how do we support these opportunities and investments going forward. And then, Shayle and Varun, if you would talk a little bit about, you know, how do we encourage capital. Where is it going to come from? How do we, you know, get businesses and investors more comfortable with what’s happening in solar technology in light of a little uncertainty around the policy dynamics?
GAY: Yes. So to speak to policy, solar is a phenomenal job creator. Today there are about 260,000 jobs in solar. That’s about the same as the total number of jobs across the coal industry, for example. Of those jobs, those 260,000, 167,000 jobs are associated with installation and development of projects—part of the soft costs that Shayle was referencing here. Those are good-paying jobs. Those average across the U.S. in the range of $20 to $25 an hour. We need a big pipeline to keep the trained and accredited folks in place to be able to install systems well and to maintain them well.
The next lever here in policy is basically infrastructure. We have infrastructure in our electric grid. We can use that existing infrastructure better by combining it with the information that we have available. One vignette of an example here is to get DC electricity converted into AC electricity, a unit called an inverter is installed with a PV system. And that inverter actually can operate 24 by 7. It can bring value to the transmission system operator and allow for reduction of spinning reserves to keep the grid voltage and frequency stable. There are value propositions and distributed energy resources that go with solar, because solar fits not just on the distribution but also on the transmission side of the grid here.
And businesses that combine information and the ability to access that information to better use electricity when you need it—it’s like Uber for electricity here. Putting the electrons where you want them, or planning for tomorrow based on weather forecasts—when I was growing up, we were lucky if we knew what the weather tomorrow might be. And now when I turn on the TV it’s 10 days from now here’s sort of what’s going to happen. A lot of that capability of projecting what the demand is going to look like maps to the weather patterns and changes. And the probability improves as we get closer in time to that need.
So the ability to create jobs and the ability to provide stable infrastructure, we will have a more robust and resilient grid by having solar in the grid. We have fewer vulnerable big generators that might be attacked. And we can embed in that network, by combining the information with this infrastructure, a lot better cybersecurity than we have today. So those are the moving elements—energy security here, jobs, and infrastructure.
MCKIBBEN: Great. Shayle.
KANN: Well, so I guess first on the capital question, I’ll disagree with Varun for the sake of argument here, which is that I don’t—I don’t think capital is actually going to be a big challenge for solar over the next century. I think that you get the cost right, and capital will follow. There is a need to do all the things that Varun is talking about. We don’t have a lot of institutional investors getting involved in solar yet. I think that’s largely because the industry hasn’t been at scale yet. And so you have a hard time aggregating up enough dollars for an institutional investor to invest to get them to start caring, especially if they have a learning curve to learn about it. So that’s a challenge.
We also got a challenge of sort of do you trust the performance of these assets? And it really helps that you have things that are operating for 40 years now that you can point to. So there’s work to be done there. I just think that work is happening. And I think that as costs continue to fall, that’ll naturally—you know, money will follow where there’s an opportunity. But where I think that there are business opportunities, sort of today and for the next little while, it’s not dissimilar to what Charlie’s saying. I think you can think of solar in a vacuum, which is how do we get all—how do we get more solar built, right?
And most of the businesses that have developed thus far have been focused around that question—either how do we get solar built on households, are you going to have SolarCity, which was probably the biggest success story in that sort of residential solar world. IPO, followed by acquisition by Tesla. How do you get solar built on, you know, GreenFields, utility-scale stuff? This is companies like First Solar and SunPower. Or just how do you build cheaper panels, and there’s a whole panoply of companies there.
Now, what all these companies are starting to look at, and where I think the real opportunity lies in the next wave, is combining solar with everything that’s going to be adjacent to it. So this is combining solar with energy storage, so that you can make solar controllable. Combining solar with low-control at the customer premises, so that a customer’s electricity demand can be managed better to align with when the solar is generating. These things are interesting. And today, for the most part, there’s not a great economic value proposition for them. But the way that we are heading in how electricity works in the U.S., partially as a result of all the solar that we’re building, is that your electricity bill is going to change.
You’re in New York, so it probably already is changing. But throughout the rest of the country, what’s going to start to happen is your electricity cost is going to be dictated more and more and by the time when you consume and the location that you hold on the grid, because we’re trying to make electricity pricing more cost-reflective. Right now, it’s really fixed. It’s a blunt instrument. That’s been sort of OK until you start putting solar and energy storage and other things in various places in the grid. Now you want to reflect—you want that to receive the value that it has. And you want other customers not to pay for it.
So the result is going to be a much more dynamic landscape for customers. And I think there is plenty of opportunity to figure out how to either provide something to customers that sits in the background and they don’t have to think about all the time, but actually optimizes their energy so that they’re paying as little as possible. And if they have a preference for clean electricity, that they’re actually using the solar power that they are generating. Or, providing something to the utility or the grid operator. There’s a need there for all sorts of data and information and actual management of the electricity system that comes with putting all these new assets on the grid. And utilities are really still just starting to figure out how to even gain visibility into all these assets, let alone control them to the extent that that is something that they’re going to start to do. So everything that’s adjacent to solar seem to me to be where the real opportunity lies for the next decade or so.
MCKIBBEN: Great. Varun, quickly, if you want to—
SIVARAM: Sure. Quickly, on capital, I think our disagreement is a distinction without a difference. I agree that lots of innovation is happening to try and, for example, make solar more liquid and easily investable. But attempts to do so, so far, sometimes have failed spectacularly. For—
KANN: SunEdison being—
SIVARAM: SunEdison being a good example. So we need to get there. And I hope capital will follow the opportunity. I’ll make the point that in the developing world, there are real opportunities to take advantage of these adjacencies. How can you combine solar with these other needs? So one promising project I recently saw in southern India was to pair solar—not solar photovoltaic panels but concentrated solar; this is solar mirrors that heat up water—to pair that with food refrigeration and water distillation. These are huge needs. And this returns of my theme of systemic innovation. How can we couple the food and water sectors, especially in the developing world, with solar? That’s a good way to do it. And that’s a way, by the way, that doesn’t use photovoltaics.
MCKIBBEN: Great, thanks. So at this time, we’d like to invite members to join in the conversation with questions. I want to remind everyone that this is on the record. Sorry about that, Charlie. Probably should have said that at the beginning. If you’ll wait for the microphone to come to you and speak directly into it, also if you would please state your name and your affiliation. We ask that you limit yourself to one question, and not a bundling of questions, so that we can get as many as possible. So if you’ll wait for the microphone to come to you.
Q: My name is Larry Bridwell, and I teach international business at Pace University.
And I’ve assigned students an interview with Bill Gates on the future of sustainable development. And by the way, people in their 20s are quite interested in this subject, as opposed to those of us who are older. And Bill Gates has recommended that research on sustainable energy be tripled from what he says is currently $6 billion a year to 18 billion (dollars). So since you work for the Department of Energy, I would like you to comment on how feasible this is, both from a political will to triple from 6 billion (dollars) to 18 billion (dollars). And could the Department of Energy, particularly under the incoming secretary of energy, be able to manage this increase from 6 billion (dollars) to 18 billion (dollars)?
GAY: I’m sure that Bill won’t have a problem finding people to take his money. (Laughter.) Sort of the underpinnings here around how much investment is realistic for a given amount of time, and the number of people in the field capable of intelligently spending that money. Roughly speaking, today, the last time I looked, there are about 60(,000) to 70,000 people engaged in research. Maybe at Pace University you may even have a problem that articulates exactly what that number is. And a lot of folks aren’t necessarily in America. A lot of the population centers in India and China represent incredible talent pools of innovation. Now that solar sits at the big-boys table when it comes to planning electricity, I think that there are a very large number of ways to have a competition of ideas around what additional funding could provide. And I think that funding can span the gamut across this value chain that we’ve been talking about here, from the basic materials through deployment and innovations in finance as well.
It would be an incredible accelerator if that funding were released. And I would hope that prudent stewardship of those funds, tracking how well they’re spent, and essentially speeding the cycles of learning here so that we make our mistakes faster. Most of my career—the mistakes I’ve made in the first two years of my career, I’ve just survived in my career by remembering what those few mistakes were and sidestepping them when they came along. So the world is very different now, because of those A-students entering this space with lots of ideas. Maybe they’re perovskite and new material ideas. There are fabulous innovations happening in power electronics. So silicon carbide, gallium nitride, bet gallium oxide—actually, I do need to know band gap now and then—(laughter)—but those materials have tremendous room for improvement and have relevance to both distributed energy sources and electric vehicle adoption, or other applications which go with increasing electrification. So my overarching tendency here would be with that money look at that full value chain, and look at the adjacents that we’ve discussed here, and look globally.
MCKIBBEN: Great. Right here.
Q: Evan Michelson, The Sloan Foundation.
So, Shayle, early on you mentioned the issue about intermittency. And to me, that means that you have to think about storage. But my sense is that industry hasn’t kept pace with the development in solar, at least perhaps not until recently. How do you see this issue of storage connecting with solar in terms of a way to smooth out the intermittency, which is one of the biggest challenges of the sector?
KANN: Yeah. That’s a good question. And, I mean, the collective eye of the solar industry over the past few years has turned squarely toward energy storage. So even if you believe that it hadn’t kept pace historically, it’s certainly catching up now. Storage has a few things going for it, right? So I guess, first, just to contextualize, I think you’re right. You know, long term there are—there are multiple potential answers to the intermittency of solar, if you get a lot of solar on the grid. One is changing demand patterns, right, so have customers start to use electricity and have it align more with when the Sun is out. A second is using fast-acting natural gas resources, for example. Flexible capacity from those things. So storage isn’t your only solution, but long-term it’s got to be a good part of the picture.
And I think you’re right that we hadn’t seen as much action in energy storage historically as we had in solar, in part because if you’re thinking about energy storage primarily as a resource to manage renewable energy intermittency it wasn’t really needed, right, because we didn’t have that much solar or wind, for that matter, on the grid yet. So energy storage sort of naturally follows solar. The benefit for energy storage is that that’s not all it can do, right? Battery storage even specifically. It can both do other things for the grid—so it can provide, you know, peak capacity reduction. It can provide ancillary services on the grid. There are other value streams that energy storage can attain. And batteries share a value chain largely with the production of electric vehicles. So you have a completely other—a separate industry that is developing and driving those costs down.
So while it is true that batteries are still, I guess, on a relatively basis, expensive relative to solar, they are currently following a cost trajectory that looks very similar to the one that solar saw a decade ago which, if it continues, will result in energy storage being totally cost effective for most of these renewable intermittency applications within the timeframe you would need in order to get a lot more solar on the grid.
GAY: I’d like to build on Shayle’s comment around these parallel trajectories of technology, the backing up from the market and the application space to looking at today in the U.S. we invest perhaps on the order of $180 million a year in the batteries associated with transportation and electric vehicles. We also invest heavily in advanced manufacturing. A lot of those are roll-to-roll kinds of concepts, which naturally lend themselves to advanced batteries, lithium-based technologies where there’s essentially a foil of copper and a foil of aluminum. It’s more complex than that, but they need to be rolled together.
And those technologies, the advantage—there’s a parallel here between how they scale and how solar has scaled. That is getting uniformity and reproducibility over large areas. The batteries use in electric vehicle are for delivery of power, for getting torque to the wheels. The lithium-based chemistry for motive power is different than what you’d ideally have for stationary battery storage. But I would argue that simply let’s change to lithium titanate for the storage use. We figured out how to get the reproducibility of the scaling in the electric vehicle space. And we learned how to get the costs of the tools down from this advanced manufacturing work.
So I think that we’ll follow a learning curve and there’s markets. So the important point here I wanted to make is having markets that will absorb these products help us move down that learning curve faster. And there’s technologies that innovators can bring to bear that can follow that.
MCKIBBEN: Varun, I think you wanted to add?
SIVARAM: Yeah, but very quickly. Charlie just brought up, basically the ratio of energy to power density in your lithium-ion batteries is going to stay fixed. But that may not be exactly what you want in your power sector, depending on what your—for example, you may want a lot more energy needs—energy storage than power discharge capacity. So there was a cute study that came out of NREL that said, look, depending on how long you need to store the power for, is it a good idea to pair PV plus batteries—lithium-ion batteries, or take solar thermal technology, which kind of comes with inbuilt storage—thermal storage on site, from heating either water or molten salts.
And it finds that, I think, the longer your storage requirement, the more economic al the solar thermal technology is compared to PV. So even though today it really looks like on a per watt basis PV is way cheaper than solar thermal, depending on what your storage requirements are, solar thermal could actually end up on top when you incorporate the cost of storage. I’ll finally say that—to add to Shayle’s battery—no pun intended—of different approaches for storage, grid expansion is a good one. The bigger your balancing area is, the less storage you need, because the likelier it is that the Sun is shining somewhere and someone’s turning on their light somewhere else.
MCKIBBEN: Great. In the back.
Q: Hi. Riju, Blackstone.
Varun, you started to touch upon some of the problems, particularly in developing markets. And I think you guys quoted the stat before, that the panel itself is maybe 35 to 40 percent of total installed costs. Could you maybe give us a little bit more information on how large does the auxiliary cost become in international markets? And what is the difference between maybe things you’d call adjacencies and then other problems I would maybe call peripheries, that are even harder to control, whether it’s, you know, state-owned utilities that can’t pay for the solar power you’re generating. How much of that becomes an insurmountable problem for all of us solar aficionados?
SIVARAM: Terrific question. Do you mind if I start?
SIVARAM: So I know the Indian market best, so I’ll answer from that context. In India, many of the costs that are not the cost of the panel, so the balance of system costs, are considerably cheaper than in other places. Your labor costs go down. Oftentimes your equipment costs go down. And Shayle probably has real numbers for this. I don’t off the top of my head. So one lesson is your cost stack looks different where you are. And a bias toward analyzing developed country markets can lead you down the wrong path when you analyze economics in, for example, India.
Second, you talk about adjacencies. And I think the point I’d make is that these adjacencies can be opportunities and they can also be hindrances. And one can turn into the other. And so we talk about hindrances. In India, utilities tend to be insolvent—or, many of them are insolvent, and hence it’s very risky to sign a contract with a utility to off-take your power, because you don’t know if they’ll be able to keep up their end of the bargain. That raises the cost of capital.
Another adjacency is that—sorry, an adjacency that could be turned into an opportunity is in 2012 the biggest blackout in human history was caused by too many irrigation pumps being turned on and causing strain on the grid during a drought in the middle of summer, right? This caused an enormous blackout. And this could be alleviated if you had solar-powered irrigation pumps. So that’s an opportunity for solar to kind of alleviate one of these adjacencies. So there is no, like, simple through-line here. The point is, you really need to do your contact-specific analysis. Solar looks very different in developing country markets than it does in the developed world.
KANN: Yeah, I don’t have a ton to add. I think that’s right. But, I mean, you know, one thing to think about, so it costs less to install solar in India. But depending on what your value proposition is, if your entire financial calculus is based on selling power to a utility that is insolvent, or even in the case of something that’s put on a rooftop, based on feeding power back into the grid and getting credited for that, net metering being the term we use for that in the U.S. You can’t finance that, generally. So your cost of capital goes up, or the capital just isn’t there.
One of the sort of likely outcroppings of that as a result, is that those adjacencies, some of the ancillary technologies, like energy storage, are in some ways coming quicker than in some of those countries, because if you can self-consume all of the solar on your rooftop, then you don’t need a contract with the utility. And the other benefit that you’ve got there is that you generally have less onerous interconnection requirement. So that’s both a soft cost that we pay for in the U.S., and it’s time consuming.
So to the extent the battery costs fall enough and that we can optimize these systems, there’s a case to be made that we’ll start to see more leapfrogging in those countries, just because of some of the challenges with the traditional economic model these projects have.
GAY: I would say this is the most important adjacency—a smartphone. So what happens now in developing areas, as they’re scaling solar, is being able to make payments weekly or monthly or quarterly, using the cash banking system on a smartphone. And those package systems, that could provide water, lighting, fan, radio, TV, they are able to reach those markets initially, without waiting for wires. So the smartphone is dramatically changing the environment for rural populations.
And since I’ve been in solar, the population of our planet has more than doubled. The number of people living in poverty, I recently read, is actually starting to decline. And part of that is with energy efficiency that’s enabled more efficient lighting to go together with solar, for example. More efficient motor control systems. And that extends all the way up through that microgrid example I was citing, about Caterpillar working in rural areas. This same software can sync up a gen set with more generation on those wires. And it can work in Palo Alto or Lesotho.
MCKIBBEN: Great. Right here, on this side.
Q: So, Paul Richards from Columbia University.
So coming closer to home, how do you utilities companies in the United States need to change to benefit from new opportunities here? And one example, with my suburban house I’m having an installation now on the roof. And I’m told that I could have one three times larger, there’s enough roofage there, but I’m not allowed to do that because I can’t—I have to have an appropriate relationship with my utility company. So I’m limited in what installation I can make.
KANN: I’ll take a stab at it, and then anyone else can jump in. I mean, you’re getting at one of the big central questions that pertains to this decentralization trend in particular, which is that it is—it is causing all sorts of hardship within the boardrooms of utilities thinking about what their role is going to be in the future, and their relationship with the customer. There are lots of ways in which, you know, some utilities within the U.S. have tried to just head this off at the pass and stop the sort of growth of solar, in particular, and some other things. I think that’s not the dominant problem.
The dominant problem is the regulatory model under which utilities operate. Generally speaking, if you’re a utility in the United States, the way that you make money is you invest—you don’t make more money typically by selling more electrons. That’s actually not how you make money. The way you make money is by investing in capital infrastructure, and then you earn a regulated rate of return on any capital expenditure than you have. So your incentive still is to invest in more stuff, more hardware on the grid. That is what you can earn your money on.
The problem is, everybody is installing all these things on their premises, generally speaking they’re reducing the load for the utility, which reduces the utility’s ability to justify new capital expenditure. What you want is a model that treats the utility as a market operator, as an enabler, to ensure that we still have reliability and that we have low-cost electricity, but doesn’t put them in a position where every time somebody puts solar on their roof it hurts the utility economically. And there is a lot of thinking being done around this. New York is home to the sort of prototypical initiative around this. It’s called Reforming the Energy Vision. It’s being led by the governor and the state utility regulator. You also have some really innovative stuff going on in California, Minnesota, Texas, a few other places. So we’re in the early days there. But I think part of this whole transition is going to be a complete rethinking of the role the utility has and, as a result, the way the utility makes money.
SIVARAM: But let me say to that, so it is certainly the case that the current utility business model incentivizes a utility to put more centralized capital in the ground, and at the very least they should be indifferent between centralized capital infrastructure and another model of delivering electricity—for example, operational expenditures for distributed energy resources—fine. However, an end goal is not to have more distributed energy resources. That’s an intermediate step. The end goal is to have the cheapest energy in the cleanest way, and perhaps to increase consumer choice—although that’s by no means a unanimous goal among folks in this sector. So in your example, you’ve got more roof space. Should you put more solar panels up? The utility system’s not letting you do it. It’s not at all clear, from a societal point of view, that you should be putting more solar panels on your roof. In fact, panel for panel, you probably should be putting them in a centralized farm out in the desert. That’s the much cheaper way of generating solar electricity.
KANN: And it’s open for debate.
KANN: It depends on where he is on the grid, and—
SIVARAM: Absolutely true. Absolutely true. But you look at MIT’s Utility of the Future Report that recently came out, and they concluded that in a preponderance of cases, the centralized model tends to beat out the decentralized model. There are some very congested areas of the network where it does make sense to have a decentralized solar panel. But those areas are few and far in between. And if you want a heuristic, it is that centralized solar tends to be more economically than decentralized solar. Do you disagree with me there?
KANN: No, I think generally that’s true, though I—you know, I wouldn’t just brush over the fact that there are significant—there’s another good study in California, from Berkeley, just looking at, like, places on the grid where installing solar causes significant economic benefit, right, because you can reduce the need to upgrade a transformer, or something like that. So those places exist, but I agree they’re not everywhere.
SIVARAM: And so the ultimate solution that MIT supports is a solution—which is kind of fantasy land, but geeks like it—it’s all power is priced identically if it is injected or withdrawn at the exact same node in the electricity grid at the same time. That is kind of the golden rule of electricity pricing. We are very far from that. We’re far from that because of technology reasons. We actually don’t have very good visibility into every node of the grid, so it’s hard to have granular pricing. We’re not there for political reasons. Consumers may not like very variable and fluctuating prices. And we’re not there because of kind of path-dependent inertia reasons. So we’re very far away from this idealized state, and frankly I don’t think we’ll ever get there. We’ll end up somewhere in the middle. The goal is to make that intermediate resting point the best we can make it.
MCKIBBEN: Great. I think we had a question over here.
Q: Hi. John Watts. I’m with World Policy.
I wanted to ask a question along the lines of which we were just into. And you raised a question of how can we, and how can—it struck me in a career that started in the ’60s in electricity and water, and still investing in that area, that almost—it’s perhaps, if you can call it an industry, one that has the most unwise setting of the price. And central stations being a great example. It’s one of the biggest things we do that is all priced outside of the usual market pricing phenomenon. So I guess my interest would be, how do we break down cells like we do in most of the things we—but I really have a question, what do you think about the one thing you haven’t mentioned, which I think’s very fascinating, is—well, I don’t know whether I should say it—but fusion. And what do you think about that?
KANN: You get the fusion. (Laughter.)
SIVARAM: Lots of physicists have gotten in lots of trouble for commenting on fusion. So I won’t say much other than there are many—there’s a broad spectrum of technology priorities. Many of them deserve to be funded. And recently there have been a few startups that have made more progress than ever before. For example, the length of confinement of a fusion reaction—through actually very different mechanisms—that’s exciting. How far is fusion? Fusion’s always been 50 years away—whether it was 50 years ago or now. (Laughter.) So I have no idea where fusion’s going to go, and I can’t really comment.
KANN: I can take a stab at, I think, the other question, which has to do with pricing. And we were getting there to some extent, but I would note, within electricity it’s not all this uniform sort of basic mallet of a pricing scheme. That is generally how it works for customers. But we have pretty vibrant wholesale power markets that do price on a locational and a time-dependent basis. So the way that the big central station generators get paid, generally, actually is a pretty well-functioning market. So I think a lot of the work that’s being done now, thanks, in part, to technology that’s enabling us to start doing this, is looking at what has worked on a wholesale side for electricity and trying to apply it to the retail side.
MCKIBBEN: Right here.
Q: Hi. Teryn Norris from S&P Global.
A lot of folks in this room are probably wondering what’s going to happen with tax reform right now. And, of course, the solar sector’s highly dependent on the investment tax credit. I was wondering if you could give, maybe Shayle in particular, your outlook on what could happen to the tax credits, and what you see occurring to the U.S. solar sector once the tax credit does (fade down ?) starting in 2020.
KANN: So you’ve got—you’ve got two—at least two separate issues, as it pertains to sort of the world of tax today and solar. One is the Federal Investment Tax Credit, which you’re talking about. So this is the primary, really the sole major federal incentive for solar in the U.S. It’s a 30 percent tax credit today. It was scheduled to expire at the end of 2016. And in a tax-extenders bill compromise, bipartisan legislation at the end of 2015, it was extended for five years, but actually effectively for seven years. It starts to step down after 2019, drops to 10 percent. But given the rules about what it takes to qualify, in effect you’re going to be able to build solar until about 2023 with a meaningful tax credit, potentially with a 30 percent tax credit that whole time, under the status quo.
So we got this extension. There is one question, will the incoming administration and Congress in particular, will they roll that back early? Nobody knows the answer to that question. What I will say solar has going for it is, one, inertia. This tax credit was extended a year ago. It was bipartisan. There’s no reason to think that that would change now. It hasn’t been targeted, as far as we can tell, by anybody who’s incoming in the administration. So, you know, it could get caught up in a broader comprehensive tax reform, but I think the odds are—betting odds are generally against it, and certainly the solar industry and the advocates in the solar industry are going to be fighting it as hard as they can.
So that, I think—that’s the black swan event. Would have a big impact, but seems less likely. What seems more likely is corporate tax reform, right? And corporate tax reform has a different impact on solar, because the other sort of—call it an incentive if you’d like—but the other way that the tax policy benefits solar is accelerated depreciation. You can depreciate solar assets over five years. That cause a meaningful economic benefit from solar projects. And if you reduce the corporate tax burden, it can have an impact on how much you can depreciate from those assets. It actually ends up somewhat being a wash in most of the modeling, because if you have a lower corporate tax rate you can’t depreciate as much, but your cash flows from the project in the later years are taxed at a lower rate. So that doesn’t have a big impact.
That leaves one final thing. I guess I should have said three. Which is that solar is very dependent on tax equity. The way that we get these things built is at least 30 percent, and generally up to 50 percent, of the capital stack for one of these projects comes from somebody who has a big tax burden and can invest and monetize those tax credits from the investment tax credit. That’s a limited pool of investors. It’s something like 15 total tax equity investors have been active in solar in the U.S. historically, you reduce their tax burden. There’s still more than enough tax appetite out there to meet the needs of solar, but it remains to be seen whether they allocate that tax appetite a little bit differently. They could allocate it more toward low-income housing, historic retrofits, and other tax credits. So there’s some risk that you end up with a capital crunch, particularly around tax equity, if corporate tax rates are reduced.
MCKIBBEN: OK, last question. Right here.
Q: OK. I wanted to go back to—
MCKIBBEN: Can you introduce yourself?
Q: David Sunderwirth, Bank of New York Mellon.
Quick question about batteries, to go back to that for a second. One of the problems with fossil fuels are the direct and indirect waste streams—you know, when you’re talking about carbon and things like that. Certainly you don’t seem to have that problem with wind and solar as much. But what about batteries? I’m kind of curious about what you do with the battery when it’s ended its useful life? And my understanding is—limited as it is—that some of the stuff that goes into making a battery is pretty toxic stuff. So maybe you could just comment on that.
KANN: I can’t speak to the toxicity question. I haven’t heard that coming up a whole lot as it pertains to energy storage. I will say, one thing that’s interesting with sort of end of useful life is with electric vehicle batteries. When they end their useful life, when it doesn’t make sense to use it because it doesn’t have as much torque to power the car anymore, you can turn it into a stationary battery. So you still can use it on the grid. So there’s a second life and some business models being developed around that, of taking old electric vehicle batteries and using them to provide various services on the grid. So they should have a much longer lifetime than you imagine them having just in a car. But I don’t know about the sort of waste stream question.
SIVARAM: And then I’ll—now that the cat’s out of the bag, there’s a terrific study out of MIT showing how you can repurpose lead-acid batteries from cars to build perovskite solar panels. (Laughter.) So.
KANN: Two drinks.
MCKIBBEN: Second time you’ve mentioned that word.
SIVARAM: Yeah, two drinks.
MCKIBBEN: OK, so we’re going to conclude now. I want to thank everyone for attending. Have a good rest of your afternoon.