Stephen Lacey: This is The Interchange, conversations about the changing business of energy and cleantech from Greentech Media. I'm Stephen Lacey in Boston. This episode's a little bit different. We're going to be more rapid-fire than usual. We're biting off a bunch of different subjects rather than digging into one specific subject.
We're going to walk through the top 10 trends to watch for in solar this year. That includes tech trends, pricing and market design. With me are my esteemed colleagues, Shayle Kann and MJ Shiao. You know Shayle as my co-host and our VP of research. Hey, Shayle.
Shayle Kann: Hey, Stephen.
Stephen Lacey: MJ Shiao is our director of solar research. Hey, MJ.
MJ Shiao: Hey. How's it going? Thanks for having me.
Stephen Lacey: Good. I think the last time we recorded a conversation we duped our listeners into thinking that Amazon had developed a drone-based solar installation service. I promise that this conversation's going to be a little bit more fact-based.
MJ Shiao: It's nice to talk about real things.
Shayle Kann: I still, once in a while, by the way, somebody sort of comes across our back catalog and finds that one, doesn't realize that we recorded it on April 1st, and gets really mad. Do you get this once in a while?
Stephen Lacey: I have.
Shayle Kann: Yeah.
Stephen Lacey: I've probably gotten three emails in the months after that release.
Shayle Kann: Yeah, happy to have achieved that.
Stephen Lacey: All right, so I promise that everything we say is going to be grounded in truth, or at least the closest we can get to truth in this nutty solar market. We're already almost through January, but it's never too late to issue a top 10 list, right?
The collective brain trust at GTM Research did just that recently, from community solar to competitive auctions. The team drafted up the most important influences in solar in 2017. We're going to quickly walk through all of those. Firstly, MJ, how did your team come up with this specific list?
MJ Shiao: Yeah, so we wanted to look at a cross-section of the important trends across all different parts of the industry. Not just looking at what's happening with installations or just what's happening with technology, but looking at the picture as a whole.
These trends have also sometimes been in the works for a few years, but 2017 really represents a turning point for each of these trends and how the industry has actually evolved. It's not just looking at things that we've always known about solar in terms of it getting cheaper or being deployed more, but really looking at how it is, the way that solar is being deployed, or the way that technology has evolved is taking a big leap in 2017.
Stephen Lacey: OK, so we're taking a big leap in policy in 2017, which is a continuation of what we've seen over the last couple of years. The world is moving away from direct incentives and feed-in tariffs and moving toward competitive auctions. In Latin America this is particularly acute. In the Middle East as well.
Throughout 2016 we saw record-low prices, which we will talk about as part of the second trend. Just in general, why are we seeing so many more competitive auctions versus the traditional long-term, fixed-price feed-in tariffs that were so popular for so long?
MJ Shiao: This is a great question. First it's important to note that there are still major feed-in tariff markets out there. For example, a lot of the Chinese market is still operating under a feed-in tariff style market.
Stephen Lacey: Japan as well.
MJ Shiao: Exactly. As we've seen, feed-in tariffs have always had trouble kind of getting the right level of incentive. What happens is you either have these very big boom cycles or these bust cycles, depending on where the cost of solar is, versus where that feed-in tariff is.
Now what a lot of countries are doing is letting the developers basically set the price for solar through these tenders, getting a low cost but feasible bid for solar energy. One of the things that has made tenders proliferate and be so successful is just the fact that the cost of solar has fallen so much that it can be in a lot of cases competitive and more open energy auctions, as well as renewable-specific auctions as well.
Shayle Kann: Though I think it is important to note that one side effect of competitive auctions relative to feed-in tariff is that competitive auctions at least thus far pretty much always support centralized utility scale solar, not so much distributed solar. It's hard to run a competitive auction for residential solar. You could do it in aggregate portfolios, but as far as I know nobody's done that yet.
What we're seeing is all these new countries pop up and think, "What I need is the cheapest solar," and the cheapest solar is always going to be utility scale. They introduce these competitive auctions, but the downside to that, the extent that you care about distributed solar, is that we're not getting any distributed solar in any of these new markets basically.
Though the move from a feed-in tariff is perhaps more sustainable and results in lower-cost solar build, ultimately it does mean that most of the world is starting with just centralized solar and not distributed.
Stephen Lacey: Right. If you take a look at the long tail of emerging markets, particularly Chile, Argentina, Mexico, Abu Dhabi, where we've seen record-low prices and pretty big auctions, they're all supporting utility-scale solar and really have no residential market to speak of.
With that said, the upside is that you're getting extremely cheap solar, at least the lowest bids are very cheap. We've seen about 1,200 megawatts in the last year bid for between 2.4 cents and 2.9 cents in Chile and Abu Dhabi alone. Really remarkable. You got kind of a perfect storm of equipment costs, solar resources, and low-cost financing to hit those levels.
GTM's looking at potentially 2-cent-per-kilowatt-hour PPAs in 2017. MJ, talk us through the continued low prices in these utility-scale contracts as a result of the tenders.
MJ Shiao: Yeah. To be clear, $20 megawatt-hour, 2 cent a kilowatt-hour PPAs are not going to be anywhere close to the norm in 2017. Even with that in mind, these bids are also typically bidding project prices a few years out. For example, the 2.4 cent a kilowatt-hour PPA bid in that UAE project, that, I believe, is for a 2019 start of construction.
What we're seeing is continued reductions in the costs of solar and the costs of capital for solar. Also an opening up of what folks refer to as the sunbelt, that Equatorial region where you have high solar radiation, you have more one-axis tracking being installed to increase the performance.
As a result you could get to a point sometime this year where we are very, very close to a rational 2 cent per kilowatt-hour PPA bid in one of these big tenders that we've been talking about.
Stephen Lacey: Let's talks about pricing in the module market. Due to oversupply we saw a crash of prices in the second half of 2016. I'm hearing some spot prices at near 30 cents a watt for Chinese-made modules. The question is whether the increase in demand in the second half of 2017 will help moderate prices a bit.
Shayle Kann: Well, let me try to frame this one a bit. Obviously we had a price crash at the end of 2016 and public companies were really starting to suffer. You saw a bunch of major actions taken, especially from the big U.S. players, SunPower had to eat its vegetables, First Solar had to eat its vegetables. A lot of other Chinese manufacturers started shutting down new capacity.
I think that the going wisdom then based on all of that and just the general sentiment in the market is that 2017 is going to be another year of really painful oversupply for these companies and that some of them might go under, similar to what we saw in the 2010 to 2013 timeframe.
Our prediction is that that's not quite going to play out that way, so it's a bit counterintuitive. MJ, do you want to walk through it?
MJ Shiao: Yeah. Last year in 2016 what we saw is a major price reduction of about 40% between the first part of the year and the end of the year. That was driven by demand seasonality. You had huge build-out in countries like China, the U.S., even other secondary countries like the U.K. where demand in installations was very much front-loaded. The back half of the year was somewhat weak in terms of demand in shipments for modules.
In 2017 we see a slightly more balanced demand seasonality throughout the year. You're still going to have a big run-up to the mid-year feed-in tariff cuts in China, and you're also going to have a lot of projects that spilled over from late 2016 into 2017 for U.S. demand as well.
In addition, however, you're still going to have some back-of-the-year runs from India, from countries like China again as they build out...that market. What we're saying is that we probably won't see the same drastic cuts to module prices from the beginning to the end of the year that we saw last year.
In fact, we're starting to see a little bit more stability in spot market pricing for modules and the upstream supply chain, even as we speak, in the early part of this year.
Stephen Lacey: All right. Onto the next subject, which is operations and maintenance, a sector that's becoming increasingly important as we get a lot more PV deployed. A lot of older PV plants now may need repowering, and you have a lot of vertically integrated utility-scale developers that have gotten into the O&M field. Why do we forecast more consolidation is that sector?
MJ Shiao: We're forecasting more concentration, which is not necessarily the same thing as consolidation. What we mean is that large portfolio owners are starting to try to continue to build their portfolios.
Now, some of that does mean there is some consolidation in terms of mergers and acquisitions of third-party portfolios, things that these vertically-integrated developers and asset owners didn't necessarily build themselves. They're trying to reach a certain scale in order to continue to, first of all, lower their cost of performing operations and maintenance and asset management, as well as also trying to diversify geographically and reduce their exposure to any specific country or any specific regulatory scheme.
Stephen Lacey: Do you have a lot of exposure to regulatory changes in that part of the field?
MJ Shiao: Certainly. For example, look at a country like Spain, which retroactively messed around with feed-in tariff levels and what system owners could receive for feed-in tariff levels for the year. That trickles down to the O&M supplier. It dictates how often they're going to be able to fix the system if something goes wrong, whether they should fix the system when it goes wrong, or even if they're going to be able to be paid for the services they provide.
Stephen Lacey: OK. Let's go into system pricing. What we're already seeing is that this year fixed-tilt utility scale system pricing may fall to around $1 per watt, if not below that. That's pretty remarkable. What is getting us closer to the dollar-a-watt price in the utility-scale fixed-tilt system sector?
MJ Shiao: Yeah. First of all, these are EPC, turnkey EPC, prices, so in many cases you still need to think about what the developer or what the asset owner's margins are on top of it. However, for the build cost of solar the biggest thing that has happened over the past year is quite simply module prices. When you had 60 cent a watt module prices beginning of last year fall to 30 cents a watt levels, that's a huge, of course, reduction in the overall price for a utility PV system.
That's also being supported by other reductions in the cost stack for a PV system. We're seeing continued competition amongst PV hardware, so foreign inverter companies, new tracker companies that are pricing aggressively within the market as well.
Finally, overall, just more efficiently tackling soft costs by developers and EPCs as well. Thinking about how to optimize labor, how hardware choices affect labor and balance-of-system costs in general for the system, and thinking again about the system costs overall rather than just buying the cheapest components and cheapest balance-of-systems out there.
Shayle Kann: That's the thing about the system price breakdowns that always astounds me when I look at them, even from us. A lot of places you'll see, OK, solar costs x dollars per watt, or $1 per watt. That'll be a single number. It won't be broken down. That doesn't provide too much value. We spent a lot of time over the past few years trying to break that down into as many component parts as possible. We're looking at the module, the inverter, the electrical balance of systems, structural balance of systems, and labor all individually.
You add all that up and you still get to 70% of the total system cost for utility-scale, where the other 30% is this bucket of non-labor soft costs. I get that there's all these other soft costs that you were just talking about, but none of them are that big individually. Yet, in aggregate they add up to 30 cents a watt on a $1 a watt system. It's crazy how big that number is. Do you have a sense of is this million, like, half a cent things, or are there some big chunks within that?
MJ Shiao: The biggest chunk within those non-labor soft costs is basically overhead and margin, right? You hope that companies are making a margin and that they don't reduce that too much so that they can continue to operate. That is a place where companies have looked to when they're reducing costs overall and they're thinking about how do they continue to win projects. They will reduce their margins accordingly in order to be able to bid competitively.
Beyond that, there are a lot of smaller cuts of activity when you look at non-labor soft costs. Things like permitting, things like engineering and design. With all those factors, it's easy to say, "OK, we're going to reduce non-labor soft costs," but it becomes, when you start looking at, OK, specific steps, concrete steps, you can take to reduce things like design and engineering. It becomes much more challenging.
There's a certain case to be made that a lot of these costs, soft costs in particular, are very sticky and very difficult to continue to squeeze out.
Stephen Lacey: OK. That wraps up some of the global trends. Let's talk about the U.S. In the U.S. when we think about utility-scale procurement, we're leaving the era of the RPS and entering the era, or re-entering the era, of PURPA, which I like because I like to say "PURPA" better than "RPS."
Procurement's changing. That's opening up a greater number of states where utility-scale solar is competitive and utilities are procuring projects. Then, also on the residential side, which we have talked about on this show with Nicole Litvak, we're seeing loans start to gain a much bigger market share. Leases and PPAs in residential solar are declining a bit.
Just walk us through what's happening there in terms of procurement on both the utility-scale and residential side.
MJ Shiao: I think the important thing with the utility side isn't necessarily just pointing to PURPA, which if you look at the projects in development right now it makes up about 25% of that pipeline. What's more important is that overall, non-renewable portfolio standard drivers are what is pushing the utility market forward now.
In the past you had big RPSs and, for example, like California, that drove the majority of utility-scale development. We're starting to see that diversify into procurement drivers like PURPA that we mentioned. Certainly corporate offtakers are a big piece of wholesale solar development as well. Community-choice aggregators have been slowly growing. Still a very small percentage, but an important piece, especially in markets like California.
Finally, voluntary procurement, where utilities are looking at the cost of solar and saying, "Hey, this makes sense to build out because it's either cheaper than or a hedge against our traditional generation plants as well." As these drivers diversify, certainly the geographies and the types of systems that are built out will diversify as well, which is important for the overall health of the solar industry and its continued sustainable growth.
On the flip side, in residential, what we're seeing is a change in how customers are buying solar as well. In the past we've seen residential market growth being driven primarily by third-party ownership. We're seeing that start to flip in 2017 to loans and direct purchases, cash purchases by the homeowner.
Primarily because, first of all, the declining cost of solar means that a lot more folks have the cash to be able to buy systems upfront. Secondly, the finance options for loans have proliferated as well, such that there are more, not only more loan offerings, but also more attractive loan offerings as well. These loan providers are also partnering with not only the long tail of installers but also some of the major installers like SolarCity, Sunrun as well.
Stephen Lacey: That actually feeds into the next two trends that you outline. One is that increased customer ownership is actually encouraging people to buy top-tier premium-brand solar panels. Just explain why that's happening, and then also the continued surge in module-level power electronics in the residential sector.
MJ Shiao: Sure. What we found in our research, which looks at project-level installations in the U.S., is that there's a strong correlation between long-tail installers as well as direct purchases, and the procurement of premium brand modules. Modules like SunPower or LG that maybe charged a higher price but offer a higher-efficiency product as well.
As a result of more loan options being out there for smaller installers, as well as customers being more involved in the purchase of a system as well, we do see that these premium modules are poised for growth within and being able to hook themselves onto the growth of the solar loan growth as well.
Shayle Kann: Microinverters, DC optimizers, these power-level module electronics have a really significant market share in residential. You see that growing.
MJ Shiao: Yeah, absolutely. What we're seeing is around 70% of residential projects installed today will have module-level power electronics included in them. This primarily is being driven by the National Electric Code, in particular a section that outlines the requirements for rapid shutdown.
Now, most of the country has been under the National Electric Code for 2014 for the past -- at least a couple of years in many states. Maybe a little bit more recently for some states. California has actually been operating under the 2011 code and has finally switched to the 2014 code in January, which, without going into a lot of the wonky details, means that a lot of the systems in California, or new systems in California, will now be built with module-level power electronics or some other form of rapid shutdown in the system as well.
Stephen Lacey: Our final trend to watch relates to community solar in the U.S. We issued a report maybe a year ago and we had to revise some of our projections because of legislative delays, program delays, financing delays. These projects are complicated in nature. They're driven by very specific policies that can be hard to craft. It can take a while for them to actually implement programs.
All of a sudden, we saw this delay of projects, and it looks like in 2017 many of those projects that we thought might be developed last year are going to get developed this year. Is 2017 the year for community solar in the U.S.?
MJ Shiao: Yes, absolutely, in the same way that 2016, where we saw over 200 megawatts of community solar get built out in the U.S. was also the year of community solar for the U.S. as well.
We're building upon a banner year. Certainly a lot of this is being done in the major community solar markets that we think of, legislative markets like Colorado and Minnesota, or places where people are taking advantage of virtual net metering, such as Massachusetts and New York.
The big story here is that utility-led community solar programs are a bigger chunk of the 2016 market and are actually growing quite significantly as well. You have big programs such as Rocky Mountain Power. Also some of the programs in California, that have pushed utilities basically to own, operate and then also take the subscriber acquisition or subscriber management portion of community solar into their fold as well.
This is a big opportunity for solar developers who have not necessarily worked closely with utilities in the past.
Stephen Lacey: Yeah. From 50 megawatts basically in 2015 to 200 megawatts in 2016, a fourfold increase, and then now we're going to see the market double again to about 400 megawatts. Keep your eyes on the community solar market here in the U.S. Any wild cards, gentleman, that we didn't list before we wrap this conversation up?
Shayle Kann: Well, much as I hate to say it, the Tesla solar roof is a bit of a wild card. We've expressed plenty of skepticism about this product and the launch just based on how little we know about it. All we know is this one fancy announcement that Elon Musk made. Now SolarCity is officially a part of Tesla and they're still talking about the solar roof. They're planning to launch it this year.
Look, whether or not it ends up being a big part of their sales this year, or even next year, if they introduce an attractive, economically-competitive, BIPV, building-integrated solar product, then that could be a big deal in the long term.
MJ Shiao: I think policy is also always a wild card on solar. For as much as the industry has moved away from being solely dependent on incentives and specific solar program policies, it is still very much a policy-driven market, whether it's the ITC in the U.S., whether it's planned auction programs in countries like Brazil, or domestic content requirements in places like India, or just the top runner feed-in tariff programs in China.
All of these are very policy-dependent. Any shifts within that policy, any shifts within the regulations regarding solar, means that we could see a drastically different market than what we're expecting at the beginning of the year.
Stephen Lacey: Well, if we see any drastic changes or if Elon Musk rolls out his solar roof with success, or it falls flat on its face, you'll hear about it here on The Interchange. I was joined by MJ Shiao and Shayle Kann. Gentleman, thank you very much. We will catch you next week.