Martin Green Interview Transcript

[music]

Bruce McCabe: All right, so I'm here at the University of New South Wales and it's so nice to be back in my home country in Australia, and we're going to explore the questions today, how much further can we go with photovoltaics? And I'm very excited to be sitting with the best person in the world to be talking about that subject, because professor Martin Green has been researching and building and improving photovoltaics since the 1970s. [chuckle] And he looks my age so that's quite remarkable to start with. 

So Professor Green, quick introduction, is a professor at the University of New South Wales, also the director of the Australian Centre for Advanced Photovoltaics, which is across multiple universities and research centers as I understand it. Five decades worth of research and contribution in this field and still going. His research teams have held the record for silicon solar cell efficiency for 30 of the last 39 years, which gives you a sense of it. And the PERC solar cell technology that he and his team has introduced is now I believe, embedded in more than 90% of all installed solar on this planet. So that gives you a sense of the footprint and the contribution, the global contribution, from a science and engineering point of view. I can't think of a bigger global impact out of Australia. So it is indeed such a privilege to be having this conversation. Look, umpteen awards, I'm going to put some links on the website to the amazing awards, but many national awards given: The Japan Prize, The Australia Prize and the very prestigious Global Energy Prize, Fellow of the Royal Society. It's a very, very long list, but I've found the person to talk to and have a wonderful conversation. Professor Green, welcome.

Professor Martin Green: Thank you.

BM: And how far can we go? Let's explore this question. How much further can we go with photovoltaics? because it's come so far in 50 years.

MG: Yeah, yeah. So we can go a long way in terms of the amount installed, but also in the performance of the cells.

BM: Performance. Well, let's start with performance. Let's kind of, yeah, get stuck into that because I believe we talk in terms of the amount of energy that falls on the surface from the sun and how much of that can be converted into electrons, basically, for our use. And it's sitting … where at the moment? 

MG: Yeah, well, the present commercial PERC modules are 21.6% is about the best you can do.

BM: So a fifth of the energy falling on a PV surface is converted to usable electrons. Is that the right way of putting it? 

MG: Yes, exactly. Yeah. So we hold the record for overall conversion of sunlight though at 41%. So that...

BM: Wow.

MG: It just sort of shows there's still plenty of scope for improvement, but you've going to do things a little bit differently.

BM: Okay. So in the lab, we've doubled that efficiency, it's just at the commercial level we're sitting at 20%.

MG: That's right, yes.

BM: Wow.

MG: Yes. So there's still a long way to go efficiency-wise and, yeah, it's proving difficult to translate that 41% into a low-cost device, but I think eventually we'll be successful in doing that. So we have the potential for essentially doubling the efficiency from where we are now.

BM: And that's with the silicon-based photovoltaics, is it? Or are we talking about another, are we talking about... Are we jumping already? The Perovskites. I always have so much trouble with that word.

MG: Yeah, yeah. So with the silicon, you're stuck. I think the commercial modules will get up to 25% with silicon eventually, but you're sort of stuck there. But if you can find other materials that have all the good properties of silicon and stack them onto silicon, you can push the efficiency much higher. So in our 41% module, we actually have four cells converting different parts of the solar spectrum.

BM: Oh, wow. Okay. So different... Yeah. Right. Okay. So it's dealing with different wavelengths if you like? 

MG: Yeah, exactly. Yes.

BM: And that, is that four layers? Or how does that look architecturally? [laughter] Draw a mental picture for me.

MG: The final product would just be a stack. And the cell on top responds to the highest energy photons in the sunlight. And then as you go down through the stack, the cell specialized for lower and lower energy photons. So silicon is a very good general-purpose converter because it converts all the visible photons, high energy ones, and then right through into the infrared. But other materials can be more specialized for things like blue photons or green photons and stack them on top of silicon and you can get an improvement in the overall conversion efficiency.

BM: Fantastic. So if we rewind, when you started this journey, what were the efficiencies then? Because it just gives us a sense of how far we've already come to get to 20% at a commercial level and 40% in the lab. [chuckle]

MG: Yeah, when I started in the early '70s, there was really no commercial panels for use here terrestrially, they were all up in space. So the first commercial panels started getting developed in the early '70s and they were about 5% or 6% efficient, but they weren't very durable. The packaging wasn't all that great. So they were only expected to last a year or so out in the field. But that's come on enormously both in terms of the efficiency and the expected durability from the panels.

BM: Did you always see that sort of potential headroom to improve the efficiencies? Like way back then, was it clear to you from a science perspective that you'd get there one day? Or was it completely unknown territory? 

MG: Yeah, it's, I guess, the thinking was that 20% was about the best you're ever going to do with a solar converter. So that was the general thinking back then. And there were some quite good cells made in the early '70s. So '74 there was a big jump in performance. One of the US labs, COMSAT, which was set up to improve satellite technology, they looked at solar cells, what they could do to improve them and they bumped the efficiency up somewhere between 16% and 17% by present standards for terrestrial conversion. And, yeah, so that was the benchmark when we started. And the thinking was, with a lot of work we might get up to 20%.

BM: Yeah. Brilliant. Brilliant. So here we are. Okay. So 40% in the lab, and Perovskites keep coming up in this field. Now, my understanding, so I want you to deepen my understanding, but my understanding is they have potential to sort of get to maybe 60%, theoretically. And they're cheaper and there's all sorts of good things, but they're just not very stable. And that's the real problem. We can't make the solar cells last very long yet.

MG: Yes, yes. That's the real problem with Perovskites. So Perovskite is a well-known mineral, been known since the... At least the 19th century, maybe even the 18th. But these are very specialized Perovskites and the atoms in them are very heavy and very sloppy. But so it makes the material very forgiving of defects having these big, heavy, sloppy atoms there. But it also makes them very susceptible to degradation. So silicon is one of the most stable and reliable materials. Well, in terms of solar converters, it's the most stable material that you can use to make a solar converter from, whereas Perovskites are the least stable of any solar cell that's got over 20% conversion efficiency.

BM: So I can get impressive numbers, but it's a very hard problem from a physics or a chemistry point of view to get the stability.

MG: Yes. So there's the hope that that can be solved and progress is being made year by year, but it's sort of linear type of progress, whereas what's needed is exponential progress in the stability. So it needs a major breakthrough of some kind to get to the stability levels that you need to go commercial.

BM: That's already a really strong insight because if we look at this world and all the people listening to this, we're all trying to build a better future. So it helps us to simplify, I guess, we need to concentrate very heavily on silicon because that's what we've got. It works. It'll take time, which you don't really have to perfect Perovskite technology and do large installations on it. So we shouldn't really be spending too much... As consumers, we shouldn't be spending mind share on that, should we? 

MG: Well, there's always the hope that someone will make that breakthrough in stability. So no one's given up on that hope yet, that it needs a breakthrough of some kind in, to my view, to get it to the stability levels that could match silicon. because some of the silicon modules now are being warranted for 40 years and to reach that type of stability with Perovskites, there's a long way from where we are at the moment.

BM: A very long way. I believe it's something around a third of a year now, is kind of a maximum you'd get out Perovskites?

MG: Yeah. So that's the best I've seen for an efficient Perovskite. If you field an inefficient one that already has some degradation built into it, you can do a bit better. But for a state-of-the-art Perovskite device getting more than a few months field life after it would be the record for a Perovskite. Well, the interest in Perovskite is not in that material itself, but just that it's ideal for stacking onto silicon. So the strong interest is in a stack of a Perovskite cell on top of silicon. And the efficiency of that now is, I think it's 32.5 is the most recent record in efficiency. But the best one that's been fielded is about 25% efficient and it lasted about two months in the field. So, yeah, we've got a long way to go.

But the promise is still there and there's at least 20,000 researchers worldwide working on Perovskites, probably more because it's quite a simple material to set up the facilities for making and everything. So it's really an attractive research area because it doesn't require much of a cost barrier to get involved and so it's ideal. Chemists all over the world are working on as well as engineers and scientists from across the whole spectrum can get involved with the technology and are doing so.

BM: Yep. Fantastic. So the good news is though, if we even stick with silicon in that story, there's that headroom. We just have to learn how to build them cheaper, mass produce. But in the lab we're sitting at 40% efficiency, so there's plenty of headroom available to us as we then... We've going to get into the cost side. You want to jump in? Go on.

MG: Yeah. Yeah. The 40% is actually made with four cells and one of them is silicon, and the other three are different materials.

BM: Okay. Yes, please elaborate on that, yeah.

MG: Yeah. So you, like, we tried to... You need cells that respond to the different colors in sunlight and silicon, as I said, responds across the whole spectrum of colors.

BM: You did, yeah.

MG: But you can do things to silicon to make it respond just for blue photons for example. So initially we tried to do that. So by quantum confinement in the silicon, you can increase its response threshold to a higher energy. So we tried that initially and after a lot of work in that area, we decided that we had to give up on that. 

So we are now looking at synthesizing what I call an artificial silicon. So if you combine elements from the periodic table in the right proportions to get the same chemical valence as you have with silicon, you can synthesize other semiconductor materials that have different properties from silicon. So the classic example is Gallium-Arsenide which is used in high-frequency electronics and in a lot of optoelectronic devices. So it's a group 3 and a group 5 element in the old periodic table, which averages out as group 4 which is where silicon sits. So you want to combine elements from the periodic table so that when you average out the valence you get a valency of four.

BM: Got it.

MG: And so we've been exploring the periodic table to try and find materials that average out at a valency of four, but while involving materials like silicon that are abundant and non-toxic and stable. And so far we haven't found one that, and I guess the other important thing is give good efficiency. So we found one material that's stable and non-toxic and abundant, which is a combination of Copper, Zinc, Tin, and Sulfur. So CZTS it's called because just the initial letters of each of those elements. But we hold the world record for efficiency with that material, but it's around 11%. So we have to get up to 20% before it's worth putting into a stack with silicon. So there's other materials that we're looking at, and Perovskites don't satisfy that average valency of four, they get their semi-conducting properties in a different way, so that you're not necessarily stuck with these synthetic silicon materials, there's other materials that you can be exploring as well.

BM: And so far with the synthetic combinations, the synthetic silicon materials, you're not hitting grand stability issues or there's no roadblocks here we're talking about. It's mainly about the economics of building them, I guess.

MG: It's the efficiency that's a real stopper at the moment. So getting up to that 20% efficiency is quite difficult. I think those are the eight materials that have demonstrated more than 20% efficiency in Photovoltaics, and all of those involve toxic or scarce materials excepting the Perovskites and Silicon. They're the only two that... Well, the Perovskites have a lot of Lead in them, so they don't have a complete bill of health. They have about 30% Lead in their composition, it's one of their heavy elements involved in forming this Perovskites material.

BM: Later on in this conversation, I want to get on to that, sort of some circular economy questions about PVs as well, because it's all about supply chains as well when we look at the future, isn't it? It just strikes me, it's quite funny, I had a conversation, one the last ones we did was in Chicago at the Argonne Institute, a guy called George Crabtree, he heads up energy storage. And almost word for word similar conversation in that they're going through a periodic table with battery technologies, and they've going to look at the pros and cons not only of the physics or the chemistry, but the supply, the toxicity, the stability and so forth. And again, it's about headroom. Can we double the energy density with just new combinations at the anode and cathode and that sort of thing. So it's very similar.

MG: Yeah, I think in Photovoltaics, the silicon's good enough. The International Energy Agency says that the silicon panels are producing the lowest cost electricity in history and so on. And there's still plenty of cost reduction with silicon, but just as a scientist, you don't feel you've really done your job until you push the technology to the absolute limits. So that's why we're chasing this 40% with a low-cost practical cell structure.

BM: Let's now switch to the cost side because one other fascinating thing is there's been almost a beautiful logarithmic curve in reduction of cost or cost efficiency, if you like, overall, but part of that is cost. And part of that is ‘learning,’ and I think that's the right terminology in the industry, as we ‘learn’ through volumes of manufacturing to do it better and more efficiently. There seems to have been a pretty consistent... I don't know, is it 20% reduction cost for the doubling of the installations? Or the manufacture of PVs? We've been on this lovely curve, haven't we? 

MG: Yes. So if you average from the early '70s when the first terrestrial cells started being marketed to the present, you get about a 20% learning rate, so every doubling in accumulated production...

BM: Accumulated production? 

MG: Yeah. You get down to 80% of the price before that doubling. But more recently, up until the mid-2020s from about 2012 to 2020, that 8-year period, it jumped to 40% learning rate.

BM: Wow. Okay.

MG: So it was going to 60% of the initial price for every doubling in accumulated volume. But in 2020, we had a few issues crop up, including Coronavirus, which wasn't probably the most important one, but it did have some impact, which meant that we're in a new regime now. So in the past, we were sort of demand limited in terms of the market dynamics, so can the market support more production or not, was the big question. Whereas now, we're supply limited, very much with things like the Ukraine invasion and so on. Everyone wants to install solar panels and the industry is expanding very rapidly, but not as rapidly as demand.

So prices have sort of stabilized, or well since the mid-2020s, they've gone up about 30% instead of before 2020, the prices were going down 20% a year. So with this cost reduction, you could either think of it in terms of a learning curve, the cost reduction per accumulated volume, or you can think of it as just an annual decrease.

BM: Yes.

MG: And it turns out if the market is growing exponentially, which it has been, that the two are equivalent. So each year between 2016 and 2020, the costs were reducing by 20% each year that the wholesale price of a panel was reducing by 20% each year, year upon year. But since the mid-2020s, the price has actually gone up 30% instead of going down 40% each year...

BM: Yeah, and we...

MG: So a bit of a new era now.

BM: Yeah. And I guess, yes, neither of us can predict the economics of the geopolitics, they're the imponderables in this. But from a science perspective and engineering perspective, if we look long-term, there's still plenty of learning room, isn't there? To... Obviously, we've got bumps in the road like we have now economically, so from the manufacturing side, we should be able to reduce the cost, the question is how the markets work when the demand goes exponentially higher and that bounces out, but from a manufacturing point of view, we should be getting more efficient, shouldn't we? 

MG: Yeah. Yes. So a lot of the cost reduction has improved efficiency, so we think the introduction with PERC sort of halved the cost from what they would otherwise been.

BM: Really? Okay.

MG: That was largely due to the improved efficiency, although there's other functionalities that PERC allowed to be introduced into the use of the panels. But the other improvement comes through manufacturing volume and all the equipment manufacturers are now working on, equipment that they're going to market in 2 years time and they know it's going to be bigger and higher throughput and all that kind of thing to be able to sell it. So, just the size of the equipment, the throughput and so on, and is going up every year and there's a roadmap that lays out expectations of what equipment suppliers are expected to meet in their future equipment, and no use selling or developing equipment that's not going to meet those expectations. So there's a bit of a juggernaut underway, and the cost are just going to reduce as more and more solar cells are made regardless of the technology, just through this increased throughput and refining the manufacturing processes and everything, simplifying things, finding a better way of doing things.

BM: Yeah. Yeah. So there's plenty of room there. One of the things that, it's slightly a cheeky question, but you look at the learning curves for wind and solar, and they're much steeper, they've been historically much steeper in solar and there seems to be much more headroom in solar. I think wind... Solar 20% reduction with doubling of... What was the right term? 

MG: Accumulated.

BM: Accumulated.

MG: Production volume.

BM: Production volume. Wind is sort of half that, 12%-ish. And I look at the future and we've got definitely a huge role for wind in that future, but I can't help but thinking solar has to be the dominant, by far the dominant form of renewable energy production in say the 10 to 20 year timeframe because of that.

MG: Yes.

BM: It's just going to be so much cheaper.

[chuckle]

MG: Yeah. It is going to be cheaper. And yeah, the two are fairly complimentary though, so that solar and wind together are better than either individually because the wind tends just to blow at night and then winter and the solar's not great in either of those periods. So that reduces the amount of storage that you're going to need to adjust for the intermittency of the combination. So, yeah, there's more solar capacity installed now than wind, so it's overtaken wind in terms of the total cumulated volume produced now. But it does have the lower cost potential. So wind, you can make the wind generators bigger and so on, which is happening. [chuckle]

BM: Yeah.

MG: But yeah, the solar, it's a bit more like micro-electronics. The cell costs can be really reduced to a fraction of what they are now in the fullness of time.

BM: Very exciting. And the other thing that excites me about solar a lot, I get very excited about where you can put it, because obviously the most efficient way of doing this is the grid scale solar farms, especially in a country like Australia, which is blessed with some good real estate and lots of sunlight, but we also put it on our rooftops. And then there's all these innovative ways of thinking about where we can put this stuff. Now, I know there are tradeoffs in all of these, but I remember going way back, somebody asked me on stage, I was talking about the future and they asked me about solar roadways. And I remember laughing, almost ridiculing the idea on the basis that yes, you could do it, but the expense of proofing this stuff against 18 wheelers thundering down the roads, I could already see all the problems, how your mind jumps to the problem. And then in 2016 I saw my first solar roadway in Normandy, in France. And there's others rolling out and everyone's kind of learning, well, we can put them in footpaths and we can put them in other low-wear places. And it's almost diabolically crazy when you first look at it, but actually we are finding new ways to roll out photovoltaics in unusual locations, aren't we? 

MG: Yeah. No, that's right. So, one great strength of solar is modularity. You can have a solar system powering a digital watch, or you can have some of the big systems now, multiple gigawatts, like bigger than a big coal or nuclear plant. So, the whole, it can address that whole range and very flexible in where it can be deployed. So, when I started in the '70s the main applications were in navigational aids and then the old Telecom Australia got interested in powering telecommunications in outback areas of Australia.

BM: Remote outback...

MG: Yeah. So that was the only market. But that... The fact that solar was economic for those markets even back then meant there's been this continually growing market for the solar as the costs have become lower and lower, just opens up the range of applications than you can use it for. But even when it was very expensive, there were some applications that were in fact economic.

BM: Yeah. Yeah. And lately I'm seeing quite a lot around, a lot of activity around, we'd call it photovoltaic glass I guess, but transparent thin film stuff that can be applied to buildings.

MG: Yes.

BM: Any thoughts on how far we can go with that? 

MG: Yeah. Yeah. No, you can certainly have a window that's transparent that'll give you a reasonable efficiency of conversion. Building integration has never proceeded as well as expected. So I remember giving a paper with Harry Seidler in an architectural conference in the 90's, because I was trying to build up architectural use of solar because if you're paying hundreds of dollars a square meter for a bit of marble or something, why not pay that for a solar panel and have something really trendy in your building? So, that was Harry's idea as well. Some of his more controversial projects, he thought he could neutralize some of the criticism by having them very renewable energy focused. [chuckle] So he came and visited me to discuss what they could be doing architecturally with solar. But I'm not sure where I was going with that but maybe if you put me back on track. [chuckle]

BM: Just how far we can go with that idea of transparent solar laminating into the surfaces of windows and buildings, but buildings, let's start with buildings. Yeah.

MG: Yeah. It hasn't taken off as well as expected. Like Elon Musk, just about everything he's touched has turned to gold but he tried to commercialize solar roof tiles...

BM: That's right. Yeah.

MG: And you don't hear much about that these days, and he's about the 100th person to try that without any success. [laughter] So yeah, it tends to be... I think for buildings, architects like having things done customized, whereas the cost reductions in solar have been through pumping things out in a standardized fashion. So I think that's part of the problem. So, if you make a green solar module, the architect, "I want a red one, or... " Yeah, so...

[laughter]

BM: Right. Okay. Of course.

MG: "Can you make it 2 square meters and so... "

BM: There's the aesthetics. It's always the aesthetics.

MG: Or, "1 want a 2.5," or something like that. So I think there's that customization that is involved in buildings that is one of the disadvantages. But solar roof tiles, I guess you can standardize on those, but it's been an application that sounds really attractive but people have had difficulty making the product stick. So there are a lot of solar systems installed in building facades and so on, but like this building where you've got some architectural features based on solar on it, but they were customized and cost a bundle to get fabricated as a customized item.

BM: Yeah, so it's the cost, isn't it? Yeah. And there's a building going in in Melbourne which is nine storeys, and I think it's going to be fully skinned in photovoltaics.

MG: Yes.

BM: And I don't think they've started construction, but they're about to, but it's $40 million. There's a payback period on it, and they're looking at it, the zero energy footprint building and all this sort of stuff, which is very exciting, but it is a high capital outlay.

MG: Yeah. This building [the one we are sitting in] has quite a nice architectural feature on the roof in the forms of some waves, but some of the wave is made up of standardized panels. So if you can use the standardized panels, you've got a completely different structure, cost structure than if you want a customized and architectural product.

BM: Yeah. And the other one is vehicles. Now there's lots of people working on skinning vehicles, and I look in the bottom of this building, the beautiful solar powered vehicle you've got here and you've just broken a record, I believe, the students at this school in getting a car to go 1000 kilometers in a single charge, a single run. It's been using the cells, right? But you look at now they're curved, beautiful cells there in the body work of the car. Other people say, "Well, we can do printable solar cells to laminate into the skin of a car, perhaps." I think the University of Newcastle, your colleagues there have been... There's a bit of a publicity, they're running a Tesla around Australia and every day they stop and roll out a roll-able photovoltaic panel, which is cheap to produce but apparently won't last very long, to actually get a car around Australia. So what about automotive and embedding PVs in cars? How far can we go? 

MG: Yeah, yeah. No, I think it's a great idea. So, it hasn't progressed as quickly as what I would've hoped, but there are some manufacturers that will sell you a car that's fully coded with solar panels now that there's a lot more practical where it's basically a normal car. But you can get 30 kilometers or more from the solar panels alone if you install them on the car.

BM: Oh, I just see... Because I just want to park at the supermarket car park and leave it there and then come back and have more fuel than when I arrived. [laughter] I love that though.

MG: Yeah, I think it's a great idea as a range extender because the solar cells are a lot lighter than extra battery. So it seems to me that's a no-brainer as a range extender. And also you have range anxiety if you've run out of electricity in an electric car, you've got to call in the road service or whatever to save you. But if you had the solar cells, say, you might just leave the car parked in the sun and go and have a coffee or a meal or something and come back and you'd have enough electricity to limp along to where you could charge it.

BM: That would be marvelous.

MG: So I always see those as two advantages. And then while you're parked, you could run the air conditioning if it was a sunny day without worrying about flattening your batteries doing it. So yeah, there's many attributes. I'm surprised even with conventional vehicles in Australia, hopping into a car that's been parked out in the sun on a summer day is one of the worst experiences in life, I think.

BM: Yes. [laughter]

MG: But if you had the solar, just a small solar panel on the roof, you can keep the car cool while even parking in the sun. So I was surprised that that application didn't take off. So some of the car companies looked at it, but apparently there wasn't a market for cars like that in the car future.

BM: That's a lovely metaphor because the... Just imagining jumping into a car on a blistering summer day in Australia. It is absolutely one of those horrible experiences and yet if you just flip it around, how much energy are we talking about? That is an enormous amount of energy that builds up.

MG: Yes.

BM: And is striking in that car. Just in terms of the heat that hits you before you think about anything else, so it's wasted.

MG: Yes. Yeah. So the idea was you just suck cool air into the car from underneath the car, which was shadowed. So that was the ideas the car companies were working on. But I think Mazda used to have a solar panel that they used to sell with their leather option to prolong the life of their leather upholstery with the cars that had leather upholstery. But it didn't really catch on to the extent that I thought. But I think with electric vehicles as they become more prevalent, as a range extender and the other features it can maintain there.

BM: Yeah. So it's all about cost trade-offs as well and how much to put it. What about... because these thin cells, do they have a life issue? I believe they don't last as long, they might break down. So is that a consideration here as we start to think about laminating PVs into, well, the back of that laptop, for example, as well as my vehicle. Some devices only last a few years, so perhaps it's perfect.

MG: Yeah. The thin film cells need to have more protection because for various reasons, they're more moisture sensitive, so they require more careful encapsulation than silicone, which you don't seal the silicon modules up to prevent moisture getting in, those cells can handle a bit of moisture. In fact, it's preferred that the modules aren't sealed so things can get out of the module as well as getting in. But with the thin films, you generally have to seal the modules up and try to make them water tight because they are more sensitive to moisture for various reasons. So yeah. But I think you can overcome those issues. I don't think there's any inherent reason that you can't make a thin film module that lasts for 40 years as well.

BM: Yeah. So let's just get into installation rates now. Globally, we have a short amount of time to really make a huge transition on this planet. People like Mark Jacobson at Stanford, who I've spoken to, amazing. They can sit there and say, "We CAN do this, we can do the whole planet on renewables. We can get to 78 Terrawatts," which is what we kind of need to get to by 2050.

MG: Yes.

BM: When we look at the current run rate, it's too low. We need to increase radically the amount of installation solar. So where are we at? What sort of Gigawattage are we installing? Are we on 300 a year at the moment, or 250 in solar or? 

MG: Yeah. Yeah. So the figures haven't settled down for 2022 yet, although 260 gigawatts is looking like the number that a few have subscribed to.

BM: For solar? 

MG: Yeah. So some are still saying 300, over 300 and Bloomberg is saying 260 or 268. So I think it'll settle down somewhere around the 260 sort of level.

BM: Yep. And it's been growing rapidly.

MG: Yeah. So it's expected certainly to be over 300 next year, so a third of a Terawatt. And I'm expecting it to get to a Terawatt a year sometime in the 2020s. So that's quite significant because you start taking big chunks...

BM: Big chunks.

MG: Out of your CO2 emissions if you're displacing coal from electricity generation or oil from transport, which are the two easiest greenhouse gas emissions to address.

BM: So whether we need to get to probably over 2 Terrawatts a year at some point, if we're going to really make that 78 by 2050, aren't we? Or we just go up? 

MG: Yes. Yes. So, yeah, I think we'll get to that in the 2030s for sure.

BM: And from a point of view of minerals and things, can we supply that production capability? Are there any limits to that? Or is it all doable, and...

MG: Yeah. So with silicon, there's really no constraint. So the biggest issue at the moment is silver. So the industry's using, I think the latest figure is 15% of the world supply of silver goes into the solar panels.

BM: Interesting.

MG: But you can do things to control the amount of silver that you use, which has been done over the last decade. Once we got over 10% of the silver use, people started paying attention to how much silver was used in the panels, before then it wasn't really an issue. But our university has licensed technology, two types of technology that use no silver or much, like 100 times less silver than the standard panel but used copper instead. So we lasted technology in BP Solar in 1985 that they used right through the '90s up to 2006 as their premium product. And that had very little silver in there, mainly copper was a conductor, and those modules were as reliable or more reliable than the silver-based modules. So it sort of shows that you can do it, but there's a risk in changing from silver. Copper is much more reactive within the silicon. If it gets into the silicon cell itself, it can start doing nasty things which the silver doesn't.

MG: So there's a risk of failure down the track that you can't pick up in accelerated testing and so under unusual circumstances. So higher risk in changing anything in a module, particularly when you're giving 40-year warranties, you want to stick with what you know. So that's a big barrier to changing from silver. So everyone's just gone the route of trying to minimize the amount of silver through redesigning the way you contact the cells, basically. But copper is an alternative. And then aluminum, we're starting a project where we're starting to look at the potential of aluminum replacing silver or copper. So copper is best, is easiest if you're electroplated, and that's what BP was doing with the technology we licensed to them. But electroplating is a much messier commercial process than the screen printing that's now used with the silver. So if you can screen print the metalization, it's a much cleaner process than electroplating.

BM: So, a related question to this, is how much can we recover? Because recyclability seems to be, especially from the naysayers, it's always what I get thrown at me, "These things, they're hard to recycle," which they are presently, aren't they? because they're laminated and all this sort of stuff. But what are your thoughts on that? Yeah, and is the school here working on recycling? 

MG: Yeah, so we have several groups within the school and university working on recycling, so it's a very hot topic. But one of my colleagues from ANU did a calculation: if we supplied all Australia's primary energy with solar photovoltaics, so that's everything, you'd add 40% to the glass waste stream, the present Australian glass waste stream if you did that. So it's, you'd rather not do that of course, but it is not like orders of magnitude difference. It's something that could be handled quite conveniently. So the main material in the panel is the glass cover sheet and there's probably not a lot of value in that, in that there's a lot of glass available for recycling. The glass used in the solar panels is low iron, so it might be able to be recycled for producing further low iron glass that might add some more value to it. The frames are aluminum, although some manufacturers are now making them from steel, but the aluminum frames are very easy to recycle. You just rip them off and they're ready to go.

BM: Yes, yes, yes.

MG: But the other materials, like the silver in the modules has some value that, decreasing value as manufacturers reduce the amount used, and also it's very hard to get out, so I don't think it can be recovered economically. And the silicon manufacturers are pushing to ever higher efficiency in the cells, so they need better and better quality silicons. So the silicon from the panels would just be regarded as technical grade silicon rather than getting any value for the purification that it's experienced in its lifetime. So it's a sort of crummy silicon basically. So it's only worth a couple of dollars a kilogram rather than the present clean silicon that's used in photovoltaics is something, well, it's been as low as $5 a kilogram, but it's selling now for about $27 a kilogram. But if you're only getting $2 a kilogram for the silicon, it's very hard to make recycling that an economic proposition.

BM: Yep, yep.

MG: So maybe grinding it up and using it for roads or something like that is probably the type of use that you can make of other...

BM: It's all about the layers of... Yeah, what you take back, what's worth recycling and all the economics. The other one that comes up, which I have a lot of fun with, is real estate, is people think it's going to cover the country, but actually it doesn't take very much real estate to generate an awful lot of energy. And I think one number, Saul Griffith came up with the numbers, one third of one percent of the Australian land area needs to be covered. So a tiny little square, and of course it will be distributed, and we can power the entire economy off that.

MG: Yes.

BM: Which is fascinating.

MG: I think that actually might be the figure for the whole world actually. I think it's a third of a percent of the...

BM: Really? Yeah.

MG: Of the land surface area of the earth would be required to displace all the world's primary energy, like it's even less than Australia because...

BM: It is amazing, isn't it? 

MG: There is so much land and so much sun...

BM: So it's just a non-issue. And what I have a lot of fun with it, and Mark Jacobson helped with this, but you look at the land area taken up by the fossil fuel industry, which we get back, and it's a multiple of the requirement for installs of all renewables globally! So you actually... We're going to get back a whole lot of real estate by making this transition! Because all those mines, pipelines, refineries, storage tanks in fossil fuels actually occupy an enormous amount of real estate, gas fields, so anyway, yeah, yeah.

MG: Yeah, the other thing is you can put it on buildings, like there are some countries that have a high energy intensity and small land area like the UK and Japan as examples where you do have an issue with the land availability. But the UK offshore wind is probably the way to go there rather than photovoltaics, although photovoltaics again will complement the wind.

BM: Haha. Well, I think we've covered all the areas I wanted to cover and how far we can go. Is there anything we didn't cover that you think more people should know about or they have the misunderstandings over out there? 

MG: Well, I think it's taken a while for people to realize that solar has gone from what it used to be as a high-cost option to a low-cost option.

BM: The lowest, isn't it? 

MG: So it took a took a while for that to sink in, that now bodies like the International Energy Agency and the Intergovernmental Panel on Climate Change, they used to be quite pessimistic about the role renewables could play in the future because they're going to be too expensive. But now that all's been turned on its head.

BM: It's unbelievable, isn't it? 

MG: Yeah.

BM: And it doesn't really matter where you are in the world. I know it varies by geography because of sunlight and so forth, but all over the world it's more efficient, isn't it? It's more efficient.

MG: Yeah, it's... Yeah, sunlight is fairly uniformly distributed worldwide compared to wind, for example. But in the high latitudes you do have problems with winter availability and so on. So if you've got good wind there as well, that's then a match made in heaven. But...

BM: Yeah, absolutely. A bit of investment in overcapacity as well. So we have a bit of extra solar, a bit of extra wind, and now we can think about all sorts of flexibility.

MG: Yeah, apparently going to large overcapacity doesn't add much to the cost of the solar generation. So I just saw a summary from a paper on that this morning, but they were talking about 80% curtailment of the solar adding like a small percentage to the levelized cost of electricity produced. So you could curtail 80% of the solar produced so you had enough in poor generation periods, and still it would be an economic proposition. So I think that's the way things will go. There'll be a lot of curtailment, which will mean there's a lot of free electricity available for someone if you can find a use for it.

BM: Yes, but industrially, there are so many uses for that power, whether we store it or use it industrially. I love the idea that a country like Australia can capture so much of that energy in its exports by doing more processing of its metals here, for example.

MG: Yes.

BM: It's sort of, it's just very obvious if we've got that excess capacity, economically, it has enormous value.

MG: Yes, yeah. I think the mining industry's getting on board with the potential of renewables as well. So I think it's... It may well happen.

BM: I saw a line somewhere that said 90 minutes of all the solar energy falling on this planet, just 90 minutes worth, is enough to power all of human needs for the entire year. Isn't it interesting? It just gives us a sense of what's there for us. And you are the man that's about how to capture a little bit more of a slice with every passing year … So we've run out of time. Thank you so much for spending it with me, a real privilege for me. I wish you every bit of luck and all the support in your work and thank you for shedding light on all this, Professor Martin Green.

MG: Thank you.

[music]