← Shift Key with Robinson Meyer
Shift Key with Robinson Meyer
How Sun and Wind Become Electricity
00:59:40
Jesse teaches Rob all about where solar and wind energy come from.
Loading summary
Transcript99 lines
- [00:00]
Robinson Meyer
You are listening to ShiftKey Heat Maps weekly podcast about decarbonization and the shift away from fossil fuels. This week, class is in session. ShiftKey Summer School continues. Jesse is teaching us how do solar plants work, how do wind farms work, and where do those technologies come from? It's all the basics that you didn't know you didn't know. And it's all coming up after this. Shift Key is brought to you by the Yale center for Business and the Environment. Do you want to accelerate your career in clean energy? Then it's time to explore online certificate programs from the Yale center for for Business and the Environment. Whether you're designing, policy unlocking, financing, or developing important projects, Yale's online clean energy programs equip you with tangible skills and powerful networks and you can continue working while learning in just five hours a week. Propel your career and make a difference. Learn more about Yale's year long financing and deploying Clean Energy program or their Clean and Equitable Energy Development Program, which is just for five months long by going to CBEY Yale. Edu that's CBEY Yale.
- [01:07]
Jesse Jenkins
Edu.
- [01:14]
Robinson Meyer
Hi, I'm Robinson Meyer, the founding executive editor of Heat Map News, and you are listening to ShiftKey Heat Map's weekly podcast about decarbonization and the shift away from fossil fuels. Last week on the show we talked about the wide ranging and destructive impacts of the one big beautiful bill Trump's reconciliation law that is going to increase carbon emissions and set back the wind, solar and electric vehicle industries in the United States. It is a big deal and we are going to keep talking about it on the show, especially as we get more information about how the Trump administration will implement it and write the rules around it. But first, we want to cover some unfinished business. Jesse Jenkins, my colleague, is, as you hear at the top of almost every show, almost everyone except this one, frankly. He's a professor of energy Systems engineering at Princeton University. As you know, he's a brilliant engineer and a friend and a very good podcast co host, but he's also an excellent teacher. We want to bring you some of that teaching. And so for the past few weeks we've been running a series we call Shift Key Summer School. It's a look at the basics of the electricity system, which is the central system, the central energy system to solving the climate challenge. On our first episode of Chipski Summer School in late June, we discussed how the power grid works and how you should think about the major electricity and power units, what images to keep in your heads when you hear Kilowatt or megawatt or gigawatt. In our second episode, just a few weeks ago, we talked about the history and engineering of fossil fuel and thermal power plants. Coal, nuclear, nuclear, natural gas and a little geothermal. Today, we're going to talk about the history and engineering of our new advanced clean energy technologies, solar, wind and batteries. For some of you this will be a refresher, but for many of you, I think this is going to be full of well needed new information and background about the power systems at the heart of our climate policy and increasingly at the heart of our lives in geopolitics. I'll add, we recorded this before the reconciliation bill passed. So this is a Trump reconciliation bill free zone. Of course, again, in the weeks to come, we'll have new conversations and interviews and analysis of that bill. But for now, you know, here's maybe something more abstract to think about. The background, the engineering, the history of solar, wind and batteries. It's episode three, lesson three of Shift Key summer school and class is now in session.
- [04:00]
Jesse Jenkins
So, Jesse, when last we left this story, we were in the mid-1960s. We have covered all the various forms of thermal generation that exist. And to connect this back to something we've talked about, all these various forms of thermal generation that we talked about last class are what cause large spinning masses to persist on the power grid, which help keep the power grid up as this large continent wide synchronized machine. But there are other types of power generation on the grid. In fact, they are the types of power generation that we find ourselves most frequently occupied with here on Shift Key. And they are wind and solar and they are different from other power sources because of how they interact with the grid. But also, of course, because unlike everything else we've talked about, perhaps the exception of hydro, they are famously renewable sources.
- [04:56]
So yeah, we're not just boiling water.
- [04:58]
Yes, we're not just boiling water anymore or moving water with the help of hydro. So let's talk first about solar, which after all is the first mover in the entire energy system on earth, except for geothermal. Where does solar come from and how does it work?
- [05:17]
Yeah, so it's pretty remarkable when you think about it. You just stick this hunk of silicon out there in the sun and somehow it generates electricity. So what's going on there? So we got to talk a little bit about solid state physics to understand this. There's basically three types of materials that when it comes to conductivity or the ability of the material to carry an electrical charge in order to carry an electrical Charge, you basically need to be able to free up some electrons from the molecular structure of that material and allow them to flow from one molecule to another throughout the material itself. So kind of bouncing from one part of the crystal, you know, one molecule in the crystal to the next molecule in the crystal. And we talk about conductors, which are materials like copper or aluminum, right. That are very good at doing that. We use those for our transmission lines and in our consumer electronics, other things, because they require very little energy, basically none, to free those electrons and allow them to move through the material. Then we have insulators, things that are very difficult to carry an electrical charge because they have a very large, what's known as band gap, which is the amount of energy required to be absorbed to push an electron up in a higher energy state, which kind of frees it to move around and carry an electrical charge. And in the middle, we have semiconductors, which is a basis of all of the computing revolution. We make transistors out of these.
- [06:42]
You are listening to our voices right now via the magic of semiconductors. We are mediated through you because of a whole cosmos of semiconductors.
- [06:53]
That's right. And so they're also really useful, though, for either generating electricity or, in the case of light emitting diodes, generating light. So all of the LEDs and all of the solar panels are also made of semiconductor materials. So for a semiconductor, you have this band gap, which, again, is the amount of energy that has to be absorbed by an electron in order to push it up into what's known as the conduction band. That's a bit of a higher energy state than it naturally hangs out in. So it's excited. Do we say it goes, yay, I've got more energy? And then it wants to run around. Actually, in most materials, it quickly falls back down because you've sort of created a hole in the natural band where it hangs out the natural energy state, which is a bit more negative or a bit more positively charged because that electron moved away. And then you've got the electron hanging out with its negative charge. And they usually want to recombine very quickly. And so in order to turn a semiconductor into something we can use for solar, to produce electricity from solar, we've got to do something a little bit more to the design of the solar cell, but we'll get to that in a minute. So the first idea is we have this band gap, right, which is this extra amount of energy that has to be absorbed to turn the semiconductor into temporarily a conductor.
- [08:06]
Can I ask a question?
- [08:07]
Yep.
- [08:08]
Is that a physical thing that's happening in the substance, or is that a metaphor that we use to discuss a kind of quantum state that is more complex than what we're saying?
- [08:19]
It's measured in electron volts. So it is the amount of energy that needs to be absorbed to change the energy state in which the electron tends to exist. This is the quantum nature of it.
- [08:30]
Right.
- [08:30]
It's moving around in an uncertain quantum manner, but it tends to exist within certain bands of energy. Those bands are kind of features of the molecular structure of different materials, and then they're crystal and structure that they form when you have a whole hunk of silicon or other materials. And so the band gap is a physical thing. It's the amount of additional energy required to push an electron from its outermost band, kind of the highest energy state that it wants to naturally exist in, up into a higher energy band. And it can do that under a couple of conditions. One, you can provide an external current, right? So that's what we're doing in a lot of cases, like a light emitting diode, we're sticking electricity in there, then that's going to shoot out some of its extra energy in the form of a photon to create light. But it also works the other direction. The semiconductor material can absorb heat, or a phonon, that's the sort of heat part of the electromagnetic spectrum, or a photon, that's the sort of visible light part of the electromagnetic spectrum. And so this photon, which is sort of like a packet of energy flying through space, will collide with the molecules in the silicon in this semiconductor material. And sometimes it passes right through, and sometimes it hits the molecules in a way that allows the electron to absorb some of that energy. And if the energy in the photon is larger than the band gap, then that electron will get excited and it'll jump up to a higher energy state, and it will be available to move around within the crystalline structure of that material. If the. If the photonic energy is above the band gap, which it usually is, usually not exactly equal, then any of that extra energy is wasted as heat and not able to be absorbed by the electron. And if the energy is too little, then it might hit it and just turn entirely into heat, or pass through the material entirely and not excite the electronic. And so what we want to select semiconductor materials that have a band gap that is low enough that it frequently generates excited electrons, but not too low that we're wasting lots of the energy available in the incoming photon and heating up our materials to Levels that make them less efficient. So there's some material selection piece of this that tends to make silicon an ideal material for solar pv. As a semiconductor, the only question I.
- [10:55]
Have is the band gap. Basically, like electrons leap out of the crystalline structure and then are available to move around, or is it that they like. The band gap is kind of like a. Is the bands, as in there's like various bands that electrons want to live in?
- [11:08]
Yeah, exactly. So the idea is there's various bands the electrons want to live in. And in their natural state, they cluster close to the nucleus of the atom. And because there's another atom of that material, that silicon atom, not too far away in the kind of crystalline structure there, the outer levels of those atoms want to push away from each other. Right. Because they're both negatively charged, they've got the electrons on the outer side, so they're pushed away from each other. So the electrons tend to cluster around the individual atoms. And unless they have a high enough energy state that they move further away from the nucleus, they're not able to pass from atom to atom within the crystalline structure, which is what's needed to carry a conduction or to conduct a current through the structure.
- [11:54]
Does that mean. And perhaps we're getting here. So when electricity moves through a solar panel, because it's being hit by photons from the sun, it's electrons passing across the surface of that panel that's being barraged by photons. There's electrons moving through the crystal crystalline structure to form a current. Are they moving conga line style, as indirect, or are they moving as a fire brigade, as in ac?
- [12:22]
Yeah, they're moving. Unless you do something special, they just go in random directions, and you don't actually get an organized current. So if you just stick a hunk of silicon out in the sun and you attach some copper to either side of it, it will not create a direct current because the electrons want to just go in a random direction.
- [12:43]
Yeah. I mean, it's just a crazy energy reaction.
- [12:46]
Yeah. And so the key to turning a semiconductor material into a solar panel is that we're actually going to sandwich a few layers together in a way that tries to order the flow of those excited electrons in a given direction. Not all of them, but a majority of them. So that they're headed in consistently in one direction, usually towards the front or back of the solar panel, where you have conductors that then they enter into and that drives the external circuit. So in order to do that, we have to specially design the solar cells through what's known as doping to create a photodiode or a gate, which basically preferentially pushes the electrons that get close to that gate in one direction. And we do that by taking a little bit of silicon, by using silicon as our core material, but then taking a little bit of additional material from nearby in the periodic table, where they have one more electron or one less electron than silicon. So we use materials like phosphorus or arsenic on one side, and we use materials like aluminum or gallium on the other side. And by sandwiching these, putting these thin layers of these materials that have a different amount of electrons next to each other and then sandwiching them together, we create polarity between those materials that helps push electrons to get close to them in a consistent direction, Rather than just randomly going everywhere. Because every atom is a silicon atom with the same number of electrons. So that we basically have two slabs of silicon with a little bit of these two dopants of different types on either edge of them. And then we sandwich them together. And then we slap some conductors. First we put some anti reflective coating in glass, usually on the top, and then we slap some conductors on the front and back. And now we have a solar cell which can actually conduct power in a direct current. We take a bunch of those solar cells and we stick them all together in series. And now we have a solar module. And then we can cover that module in glass or whatever to make it rigid and more durable. We can slap it on some aluminum around the outside to give it some shape. And then we can stick it on some poles or on the roof, and we've got a solar system.
- [15:04]
I want to talk in a second about how we connect these solar systems to the power grid. But first I want to ask, is there a reason that we use silicon here, that it's such a good semiconductor? I guess it's like carbon. It has four covalent bonds. It's like the element directly below carbon. Carbon. When we think about why lithium's in batteries, it's like the lightest metal. It's the third element. It's the lightest metal. And so it's good for battery. But why is silicon what we use in solar panels and. Or for that matter, in computer chips?
- [15:32]
So again, the ideal here is something with the kind of like a sweet spot for the band gap that's not too low or too high. If it's too high, then most of the photons will have insufficient energy to excite the electron and get something flowing. If it's too low, then you end up with a lot of excited electrons that are quickly lost back into recombination and give away their energy as heat, rather than being able to flow through the material. And so it turns out that there's kind of a ideal bandgap range of about 1.15 to 1.35 electron volts, sort of in this middle range where you have a set of materials that just happen to have that kind of property. Silicon is one. Cadmium telluride, which we also use in some solar cells, is another in that space. Gallium arsenide is another. So there's a handful of these materials that we have used in different types of solar cells that happen to be in the middle of that range. Perovskites are another kind of new class of materials that are on the edge of that. And we can kind of tune, we can layer another layer of this material on top of silicon to absorb more of the spectrum. So that's one way to boost efficiency is actually that multiple semiconductor materials stacked in a row to make sure that you're absorbing more and more of the light spectrum. But the key is finding a material with that ideal band gap range. And silicon happens to fall right in the middle of that.
- [16:59]
And I would that you anticipated my next question, because I was going to say that sometimes you hear about perovskite as the next big innovation in solar panels. I think, interestingly, it has a. Maybe this is wrong, but my sense as a reporter is that it occupies a kind of nuclear fusion like space where it's perennially on the frontier, but has never really been able to break through.
- [17:25]
Yeah, perovskites are interesting because they're tunable to some degree. You can design the molecules in different ways to basically tune the band gap and have different qualities for absorption of the incoming light. They also have been an area where we've seen very rapid progress in the maximum efficiencies achieved in laboratory settings and test settings for perovskite based cells. So if you're a material scientist, a great way to land on the COVID of science, or, you know, nature, has been to make a new perovskite cell that beats the previous record for the most efficient cell. So there's been a lot of research activity directed here. The challenge with them is that they naturally are not very durable, and in fact, they photo degrade. So as they hang out in the sun, they break down unless you add additional materials to them. And as you add those additional materials, they get less efficient. So there's a big challenge of manufacturability and durability for perovskites. And that's been the kind of key limiter to bringing them to market as a practical solar cell. But we are starting to see some first tier manufacturers come to market with tandem cells, or hybrid cells that basically layer a perovskite on top of a silicon cell, where that perovskite is tuned for a different chunk of the spectrum with a higher band gap where you put that on top and it will do a better job of absorbing the higher elect strength photons. Then the lower strength ones that are low below the band gap for that perovskite will fly right through the perovskite and then will collide with the silicon, which has a lower band gap and will absorb another chunk of those photons. And so you can get a higher combined efficiency from those cells. So the kind of maximum efficiency we've achieved for a silicon PV cell, a crystalline silicon PV cell, it's just pure silicon, is on the order of 28% and that's up only gradually from about 24%, which was reached back in 1995. So we've kind of haven't seen a huge pace of improvement for a perovskite cell. They've gone from a couple percent efficient in 2010 to 30% efficient in the best case today, just in the last 15 years, which is exciting when we start to layer these together. It's possible to get multi junction cells that can be even as high as 40% efficient by combining kind of a couple of materials that on their own might be only 30% efficient.
- [19:47]
Much like sticking together a Rankine cycle.
- [19:53]
And a Brighton cycle.
- [19:54]
That's right, Brighton cycle, exactly. Is the cct.
- [19:57]
Tandem cells are the combined cycles of solar panels. That's a good way to think about it. Yep. And it's interesting to note that the first, I think we mentioned this in the last show. The first solar cells were actually built way back in 1884 by this guy Charles Fritz in New York City. And they were selenium, which is a different material that sort of doesn't have an optimal band gap. And of course they didn't really know what they're doing. And the glass wasn't ideal and it's reflective, whatever it was only about 2% efficient. So, you know, we've come a long way from 2% efficient to 25 to 30% efficient today. But that's getting very close to that 30% efficiency. You know, 28% efficiency for a silicon solar cell is getting very close to its maximum theoretical efficiency, given its material qualities. And so the only way to continue to push efficiency up further, further from that will be with additional materials like perovskites.
- [20:48]
I'll refrain from doing too much economic history here, but the history of silicon semiconductors and silicon solar panels is like very closely linked. Both basically emerge from working groups at Bell Labs. And I believe the story of silicon solar panels is that they leave a semiconductor in the sun or they leave it in light and notice that the power levels are crazy on it. And from thence the silicon solar panel is born. And then of course, they're both deployed in initial applications within NASA.
- [21:19]
Yeah, that's right.
- [21:21]
Solar panels on the side, and solar and semiconductors.
- [21:23]
Yep. I'll give a great shout out here to Greg Nemet, a professor at University of Wisconsin who's written a book, How Solar Got Cheap, that kind of goes through this. But what's remarkable solar is it's like 99% less costly than it was in the 1960s or 70s. Right. It's not that it's come down in price by 10% or, you know, 50%. It's not even 90%. It's come down by 90% over the last like 10 years. It's come down by 99% since originally invented. And so when the first solar cells came out, the only application that could possibly justify the expense was sticking them on satellites in space, where you needed some way to generate some electricity up there. And the batteries of the day didn't last very long if you just flew a battery up there. And so the solution was to use solar panels. And NASA was willing to pay what it took to do that. And so that kind of jump started the industry starting to develop and improve on the material science, develop better modules and better designs. Eventually it got to the point where maybe you want to put one on an offshore oil rig or some other very off grid location where it's very expensive to get energy from there, maybe it starts to get cheaper and you can think about whether we want to put it on homes or other applications.
- [22:37]
And I think there was an initial divide in the solar industry too, between whether this was going to be a utility scale technology or distributed technology. And so it was housed in these. In the 1970s, it was housed in these conglomerates that were energy conglomerates, such as Exxon at the time. And a lot of the leadership there was thinking about it as a utility scale technology and then saw it as being farther from market than it was. And it turned out that another good place to stick solar panels was consumer electronics. So if you're a millennial and you use pocket calculators from the 90s or the late 80s, you'll remember there was a tiny solar panel on that. The idea that you could stick a solar panel on a calculator was a major advance and a major innovation in deployment, let's say, of solar panels, and therefore in commercialization and bringing down the cost of solar panels made by.
- [23:23]
Because it expanded the market.
- [23:24]
Because it expanded the market. Exactly. And that was able to eke out some of the cost reductions in the 80s. But let's not. And of course, the story, just to skip forward, the story of the past 15 years in solar has been that despite this whole universe of theoretical efficiency improvements that you can make to the kind of classic solar cell, and that American researchers and American companies got very interested in coming right out of the global financial crisis, all the real life improvements have come from Chinese manufacturers just absolutely eking cents and half cents over time out of the manufacturing process and getting it. So it's so cheap to build a traditional cell while at the same time improving and deploying their own efficiency upgrades that we've been able to expand the world in solar cells and we haven't needed some of the more basic science upgrades to their efficiency that maybe were once anticipated.
- [24:14]
Yeah, I mean, if you look at what a solar cell looked like in 2005, there were still pretty thick hunks of silicon. And what the Chinese got really good at was kind of incrementally improving the efficiency of those cells with just some tweaks in material design, anti reflective coatings, and other smart ways to have the silicon absorb more of the incoming light and just getting really good at manufacturing them. So thinner and thinner, and thinner and thinner cuts of the silicon crystal, larger diameter cells. So you need less cells overall to create a module. My favorite word from the solar industry is kerf. Kerf is silicon sawdust. So it's the leftover silicon that you get when you cut through a crystal of silicon in order to make a wafer. They got really good at recycling that kerf and not wasting anything. And they started to do these in massive automated factories, which brought down the cost of production. The only US company that's kept up with this is First Solar, which is the only surviving thin film solar company, which took a different route to getting past that big hunk of what was then expensive silicon, and then said it circa 2005, 2006, which was to do a kind of roll to roll printing strategy of creating a kind of more flexible material that it can spit out and print onto a conductive backing and they have managed to stay relatively competitive in the global market. But a lot of other startups that were pursuing more exotic ways or alternative materials or other strategies around the late 2000s, which are trying to leapfrog ahead of the silicon cell, failed to do so as the solar industry in China just steadily, like a freight train, kept making these cells cheaper and cheaper over time.
- [25:57]
Robinson Meyer
Do you want to build critical skills that are transforming the clean energy sector? Then discover the Yale Clean and Equitable Energy Development Online Certificate program from our sponsor, the Yale center for Business and the Environment. This fully online 5 month program is designed for working professionals who want to build projects that serve both people and the planet. You'll dive into energy fundamentals, project finance and key phases of development and operations while learning how to center community engagement. In just five hours a week, you'll learn from leading experts, develop practical skills and gain gain actionable insights. By engaging in video lectures, live sessions and a real world group, Project Yale ensures you have the expertise to advance clean energy solutions that benefit communities and drive economic resilience. Are you ready to elevate your career and make a difference? Then visit cbey Yale. Edu to learn more and enroll in the Clean and Equitable Energy Development Program. That's CBey Yale. Edu or check out the show notes for more.
- [27:08]
Jesse Jenkins
So let's talk about the power grid. So we've talked about solar cells in the abstract. Now, solar cells is something that current moves across. But how does power go from, how do electrons go from moving across the solar cell to being something that you can deploy either on a rooftop or a power grid and get useful electricity from?
- [27:29]
So, yeah, so we've got these cells which are little flat squares. And what we do to create a module is we stick a bunch of those next to each other in a module that might be, you don't know, a meter tall by a half meter wide. And we're going to connect the conductors on each of those in series to the next cell down the line. And so again, this is a direct current. So we're just creating a path for that electricity to flow across the module. And then we collect all of the current that's flowing across the various cells on the module and we gather those across a bunch of modules and we send that electricity down a larger conductor away from the array. So we go from cells at high of 28% efficiency to a module that has some losses from the wiring and from maybe soiling on the panels. Or reflection from the glass up front that maybe brings your maximum efficiency down to something like 25% of the incoming energy. And then we're going to combine all those circuits up, we're going to run them through some protection devices that will trip off if current or voltage gets too high. And then we run them through what's known as an inverter, because if you want to connect it to the grid, we need alternating current power, like the kind that's produced from the thermal power plants we talked about last episode, not direct current power. And so this inverter is a piece of power electronics. Got some more silicon in there that basically flips back and forth the polarity of the current every 60th of a second to match the frequency of the grid, and then can convert that direct current into alternating current. That also has some losses involved. It's not perfectly efficient, but the best inverters are 95 to 98% efficient. So it's a relatively small loss at that point. Some are, I guess, can be less efficient than that. But at the end of the day, something like 7 to 10% losses from the module to the. The power lines occur during that system, that path from the module to the grid.
- [29:27]
Okay, so last question, which is that it really matters where you put a solar panel. And also I have the sense maybe this is wrong, that you do get these efficiency gains by having utility scale solar. Is that true? Why is that? Why is solar better in some places than others?
- [29:42]
Yeah, so I'll take that first question. Why is solar better in some places than others? The sun is the answer.
- [29:48]
No.
- [29:49]
So different places. Of course, a solar panel in the dark is not going to produce any electricity. It needs to absorb electrons. And so you want to do a couple things. One is ideally, as the sun travels across the sky and changes in its position in the sky over different seasons, ideally, you want to try to keep the panel pointed 90 degrees at the incoming sunlight, like a plant that would move and track the sun. So if I stick a solar panel on my roof, it's not moving around. So I want to pick a kind of fixed angle that over the course of the year, might produce the most electricity or maybe the most valuable electricity for thinking about the time of day that's most valuable to produce power. So traditionally, we've just put them facing south, tilted at the latitude that you're at. So if you're at 25 degrees north, you tilt your panels at 25 degrees. That happens to be the ideal to produce the most electricity over the course of the year. But if you're doing a utility scale system or you're on a farm and you've got the space to do this, you can also install a tracking solar array, and those can either track in one dimension or two dimensions to try to follow the sun more precisely. And that allows you to absorb more energy over the course of the day. So the silicon cell itself, the module, is still the same efficiency, but you're collecting more of the incoming light by keeping it ideally pointed. And that increases what's known as the capacity factor, which is basically the percentage of the time that you're producing, on average, your maximum capacity. Right. So 30% of the time, or I'm producing over the course of a day, 30% of what I could produce if I was cranking things out 24 hours a day, then I'd have a 30% capacity factor. That's different from the efficiency of the cell, which has to do with how I convert the incoming solar light into electricity.
- [31:29]
Yeah. Capacity factor is like how much you're using.
- [31:31]
How much sun am I harvesting effectively? Yeah. And so then the other thing, of course, that helps is putting it at a place that's sunnier. Right. In addition to pointing it at the sun, you need to have the sun in the first place. And so if you go from a cloudy northern latitude to a sunny southern latitude, you're going to get more production. That variation isn't as large as you might think, though, from the best site in, say, Arizona or New Mexico to the worst, you know, 10th percentile sites in, like, Maine, northern Maine, or Portland, Oregon, where I grew up, where it's very cloudy. That difference in solar insulation or the kind of solar resource potential is only about a factor of two. So I get about twice as much solar output from a kind of ideally placed panel in Arizona as I do in Portland, Oregon or Portland, Maine. That's a lot. But it's not, you know, we can find much better resources much closer to Portland, Maine and Portland, Oregon. Right. And so this is why it doesn't really make sense to build a giant solar farm in Arizona and then send all that power everywhere else in the country, because the transmission lines are so expensive that and the efficiency gain is not that huge. It doesn't make sense to send power that far away. It might make sense to put my solar panel on the east side of the Cascade Mountains and send them to Portland, Oregon, but not to go all the way to Arizona, because the variation in solar potential is much more gradual across different locations and doesn't span quite as much of a range as wind power, which we can talk about.
- [33:02]
I was going to say this idea that solar only varies by it sounds like about 100% its efficiency.
- [33:10]
Oh, sorry, it's capacity vector or resource.
- [33:12]
Yeah, I suspect, in fact, I suspect from previous conversations that this is going to be an important tool that comes back later. This idea that solar only really varies by 100% in its resource potential, that Arizona solar is only twice as good as main solar is going to be really important after we talk about wind.
- [33:31]
Yeah, but just, I just want to pause on that for a second though, because this is why a place like Germany, which became a leader in solar pv, still makes sense as a place to put solar. Like, you know, Germany's not a sunny place comparable to Massachusetts or Oregon or something like that and its potential. And of course they paid a high premium to be early adopters of solar when it was expensive. That's partly why they're paying so much. Still today, states like California that were also early movers in the US are paying off a lot of solar panels that were built 15 years ago that when they were a lot more expensive. But if you think about it, if I were to build a solar array in Arizona eight years ago and then build the same type of array in Germany today, because this cost of solar has come down by more than 50%, the solar array in Germany is as cost effective today as the very best solar array in Arizona six or eight years ago. So, you know, there's just cost trade off. If you know, Yes, I have 50% less output, but if the panel costs half as much and I just build twice as many, I can still get a useful, economically viable amount of energy out of these panels, even in northern climates. So solar obviously makes the most sense where it's sunniest. But because solar panels have gotten so cheap, it even makes sense in the United Kingdom or in Germany or in the Willamette Valley of Oregon. At this point, you just have to be clear about how much energy they're going to produce per panel and how that affects the economics.
- [34:56]
They should be growing Pinot noir in the Willamette Valley. I don't think they should be putting solar panels. I'm joking. Okay, so that's solar. But in the middle of the 20th century, we began to harness at utility scale another source of energy too, one that has become intermittently more and less important in our kind of renewables mix. So where does wind energy come from and how do we get grid scale electricity from the wind? Because on the one hand, it seems like, oh, wind. It's a big turning turbine. We should just have it turn at 60 Hz and hook it up to the grid. Is that how it works?
- [35:34]
Yeah, it's not quite how it works. So let's start with where the wind comes from, which is easy. I learned this early on from Calvin and Hobbes when I was a kid. It comes from the trees sneezing. That's the answer.
- [35:43]
That's the answer.
- [35:44]
I think we're done with the episode. No, of course it doesn't come from the tree sneezing. It comes primarily from two things. The wind comes from differential heating of the Earth's surface by the sun. That's the main thing. And also from the fact that the Earth is spinning around. But if it wasn't for the differential heating of the Earth, it would just be a steady swirl that went in the opposite direction of the way the Earth was spinning. But because we're tilted, the Earth is tilted on its axis, which creates the seasons. And of course, because we have night and day cycles and also because different surfaces of the Earth are more or less reflective and more or less heat absorbent. You get, yeah, there are clouds, there are forests that are dark versus light. There's the oceans, which are very dark and absorbent. Different areas of the Earth's surface will get warmer and or colder faster as that sun comes and goes. And those differences in heat will cause convection in the atmosphere. And that convection moves around the air. And that's what we call the wind. Right. The wind is just the atmosphere moving around. So differences in temperature and pressure that result from this differential heating of the atmosphere drive the wind patterns. And there are some very macro scale patterns like you probably read about the trade winds and the colonial era, when these dominant wind patterns in the Age of sail that had a determinative factor in the geography of colonial expansion and things like that. Those are these sort of larger macro scale patterns. The jet stream is another one that has a very high elevations, but we all encounter when we fly east west across the United States and plays a big role in kind of driving macro scale weather patterns across the continent. But there are also more local scale patterns like anybody who lives next to the ocean or a great lake knows that the ocean has a lot more thermal mass than the land, so it changes temperature more slowly. So when the sun comes up in a hot summer day, it's a lot warmer inland than than it is out over the Pacific Ocean. And then at night, the opposite happens. Right. It Cools off more over land than it does over the ocean. Similar things happen around mountain ranges or valley passes where, again, you've got cold air up high and warmer air down below, and they want to change places. And so you can get these sort of convective patterns. Or again, where I grew up in Oregon, the Columbia Gorge is this sort of wind tunnel. That is where the Columbia river cuts through the Cascade Mountain range, which otherwise creates a pretty big barrier between the western portions of Oregon and Washington and the eastern portions, which are much more high desert environments that get warmer in the summer and colder in the winter. And so that generates this sort of wind tunnel effect that makes Hood river the wind, windsurfing and sailboarding capital of the United States, but also creates a big alley where you have high average wind speeds, where we've built a lot of wind farms over time. So because of this sort of local topology and differences in the local patterns, we get very different wind speeds over relatively smaller distances. Right. You go from one side of a mountain range to the other, or closer to the wind tunnel or closer to the ocean or further away, you can get very different wind patterns. And then there are some big macro scale ones. Like, the reason that the middle of the country is so windy is because it's very flat. There's nothing blocking the wind from moving on. More macro scales. And so as the kind of global, the macro scale weather patterns over North America change, they drive these huge flows of wind from Canada to the Gulf coast or vice versa. So it's a combination of these sort of macro scale weather fronts, the kind of things we might hear about, oh, a high pressure front settling over North Dakota wherever, these kinds of macroscale weather fronts, and then the more microscale local topologies and differences between the ocean or the lake and the land, or the east side of the mountains and the west side of the mountains that drive the local wind speeds to be higher or lower. What's interesting here is again, it's because it's the sun that drives the differences in heating. The best wind speed sites and the worst wind speed sites also are about a factor of two different in terms of their average wind speed. So it kind of makes sense. It's proportionate with the available energy provided by the sunlight. And you get areas that are very windy and areas that are pretty calm. But the difference as well is sort of on the order of 100% difference. However, when we talk about how to convert that wind, that moving atmosphere, that wind into electricity, the key difference here is that we use a wind turbine to do that. It's like a big pinwheel or the rudder on a ship. Right. The atmosphere acts like a fluid, much like water flowing. And so, just like we can turn a hydroelectric turbine with flowing water, we can also turn a giant airfoil, which is what a wind turbine blade is with the movement of the atmosphere. And so we stick this big pinwheel out in the flowing wind, and we design the blades of that turbine to act like an airplane, where it generates lift and drag. Either lift and. Or drag can then be converted to rotational force to turn the turbine around. And what's interesting is that the way that converts from wind speed to wind power is proportionate to the wind speed cubed, rather than proportionate to the wind speed directly. Right. So for solar pv, you get twice as much sun, you get twice as much electricity output from the solar panel, but for wind power, you get twice as much wind speed. You get eight times as much potential wind power at the turbine. So that leads to a much bigger difference between a good wind site and a bad wind site compared to a good solar site and a bad solar site, which might only be a factor of two. Difference. The other scaling factors, just to think about the power, the power is proportionate to also to the area that you absorb or you interact with. So the diameter of the wind turbine. And that's partly why you get bigger and bigger turbines.
- [41:39]
Listeners can't see it, but I'm sticking out my arms.
- [41:42]
Yes, exactly. And it's also proportionate to the density of the air, which is much less of a factor. But that means that areas where it's cooler have a higher density and. Or lower elevations have a higher density, so you get slightly more output from that, too.
- [41:56]
That's really interesting. I want to return to this theme, but first, I want to also say what's actually happening in a wind turbine. So you have these giant arms that go across. They're like giant airplane wings, basically.
- [42:07]
Yeah. Which can be, at this point, like as long as a football field. 90 to 120 meters. Huge blades. And those are made of, typically, some kind of fiberglass or other composite material that's very light and can be shaped into the airfoil shape that you want stretched over some kind of lightweight, but more. More structurally sound internal structure, like balsa wood or carbon fiber or something like that.
- [42:32]
And those all connect to a central pinwheel.
- [42:37]
Yep. And so that's called the hub.
- [42:40]
And what happens in the hub?
- [42:41]
Yeah, yeah. So that hub is where all the blades connect the hub and the Blade together is sometimes called the rotor. That's the part that spins around. Then that is connected to what's called the nacelle, which is where you have the gearbox and the generator. So because the wind speed is changing a lot and it's moving at a much slower pace with a high torque, then you want to run the actual generator. There are some gears in there, usually, which then convert. Change the speed at which the rotational motion is proceeding, and then that's hooked up just like it is in a thermal power plant to some magnets that spin around some copper coil, or vice versa. So you actually are creating alternating current within a wind farm. It's just not synchronized to the rotation or to the frequency of the grid at whatever.
- [43:30]
The frequency of the wind.
- [43:31]
Yeah, exactly. It's going up and down. It's too difficult to design a broad gearing ratio that would allow you to convert any rotational wind speed into 60 hertz. That just doesn't work. And so what you have to do is then run it, collect all the local current that's generated by your array of wind farms, and pass that through some power electronics as well. So basically similar to an inverter to connect to the grid and change the frequency Back to the 60 Hz or 50 Hz frequency of the grid itself. And of course, that nacelle sticks on top of a tower. That tower does a couple things. It has to support the weight of the nacelle, which is quite heavy. Right. You've got gears and copper and magnets in the top of that thing. So it's very heavy. And it has to be high enough to keep the wind blades from hitting the ground. And the higher you get, the higher the wind speed you encounter as well, because of less surface drag, which slows down the wind speed. And so also, we're starting to build taller and taller turbines to access better wind speeds. The wind speed goes up roughly to the 1/7th power of your elevation change. So it's not a huge difference. But you do get more power the higher you get off the ground as well.
- [44:41]
I also want to define a term. I just want to leap it and define a term. A nacelle is just what you call any kind of box or container or enclosure that's attached to an aircraft wing or a turbine meant to protect machinery. It's literally just a name for that thing.
- [44:57]
The housing for the gearbox and generator.
- [45:00]
Exactly. It's what those things at the end of the wings on a, you know, 737.
- [45:05]
Right. Well, that's. Yeah. Or the housing for the jet engine. For the jet engine, yeah, exactly.
- [45:10]
Yep. The giant glowing things that come out of the Enterprise and start.
- [45:14]
I was going to say, that's when I first. That's when I first learned the word nacelle is. As a young nerd watching Star Trek. Yes.
- [45:21]
Those are the warp nacelles, right?
- [45:23]
Warp nacelles, yes. I'm glad you brought that up. I didn't have to. Thank you.
- [45:27]
It's the only. Only association I have with the word nacelle.
- [45:31]
Same before a winter.
- [45:33]
I would be remiss. My dad, when he listens to this episode, will be thrilled.
- [45:36]
Robinson Meyer
So you have all of these wind.
- [45:38]
Jesse Jenkins
Turbines do basically, then all these wind turbines with their own generators. They're wired to a central inverter, and the inverter smooths out all the different AC coming from them. I mean, I realize I'm simplifying, but they. They smooth out all the AC and then it generates 60 hertz. It comes out the back.
- [45:55]
Yeah, yeah. It syncs it up to the grid. I think it's worth talking about a couple of the sort of economics drivers here too, for wind turbines. One thing first is to note that wind turbines have this thing called the power curve, which describes the power output as a function of wind speed. And as I mentioned earlier, that starts to go up at the wind speed cubed. So that's a very steep curve. However, as the wind gets stronger and stronger, you have to deal with the structural stress of this very high power wind that's coming at you. And so in order to avoid having to over engineer your turbine to handle these very high wind speeds that only happen occasionally, they start to feather or change the angle of the blades at a certain wind speed. So as the wind speed goes up to this higher level, say 15 meters per second or 14 meters per second, they start to turn the blades of the rotor to reduce the drag, and eventually. So that kind of makes the curve start to go back the other way, like an S curve. And eventually they get to their maximum rated power, which is sort of the maximum that their generator is sized to produce. And at that point, they continually feather it to shed any excess energy as the wind speed goes up further. And so there's this sort of flat part of the curve from maybe 14 to 25 meters per second, where it's producing the same output across any wind speed by consistently changing the blades forward or back in order to adjust the amount of energy that's absorbed as the wind speed changes. And then there's this crucial cutoff speed, which is the speed where we don't want to interact with the wind anymore because our turbine might fall apart. That's something like 25 meters per second or somewhere in that range. At that point they turn the blades to be completely perpendicular to the wind. And now it generates no lift and drag and the wind just goes right past it. And so for any of you who are sailors out there, that's what we call being in irons, where you point your sail right into the wind and it depowers. And that's when you can raise and lower your sail or stop your boat from going anywhere. Otherwise it's at any other angle, it's going to start generating some lift and pushing your boat along.
- [47:59]
What's the ideal direction you want wind turbines to be? Close reach, close haul?
- [48:03]
That's a good question. I don't know what the. Yeah, I don't know how they're designed as airfoils.
- [48:07]
Yeah.
- [48:08]
So this is why you get Star.
- [48:10]
Trek, the master and commander in this.
- [48:11]
Yeah, exactly. So there's a few trends that have made wind turbines cheaper and cheaper over time too. And it's a very different story than solar panels. Right. Solar panels, it's incremental gains in efficiency and it's mostly improvements in the manufacturing process. There are some improvements in manufacturing for wind turbines too. There's learning by doing and experience there. But it's actually the design and size of the wind turbines themselves where we've seen really big gains in cost. So the wind turbines that were built in the 1990s were rated at 10 to 50 kilowatts size. They're very small diameter blades, maybe only 17 meters rotor diameter. And what we've seen since then is basically a steady increase in the size of the rotors and also the height of the turbines. So that today the standard onshore turbines are probably on the order of 5,000 kilowatts or 5 megawatts and are on the order of 120 to 140 meter diameter. And offshore, where we're not constrained by the logistics of having to move these giant long turbine blades down roads and up mountains and things like that, under power lines, we build just absolutely gargantuan turbines that are rated in the kind of tens of megawatt scale now and are 150 plus meter diameter. So each blade is 70 or 80 meters long and they're sweeping this enormous area and they have to be then elevated to a height of something like 120 to 140 meters. So that's for comparison, that's like, as tall as the Chrysler building in New York City. The middle of the rotor is the top of the Chrysler Building in New York City for one of these giant 12 megawatt offshore wind turbines that we're building right now. And in a lot of the offshore wind farms, there's a simple economics reason for why you want to just keep building a bigger turbine. And that's because the area of wind that you can absorb with your rotor goes up with the blade length squared, right? So remember, the area of a circle is PI r squared. Well, the R is the length of your blade. The radius of that circle is the length of your blade. And so what that means is, again, you double the length of your blade, and your swept area goes up by a factor of four. And so even if it's a bit more expensive to build a blade that's twice as long, maybe it's twice as expensive, you still get a lot more power for your buck because you're getting this swept area going up with the blade length squared. And the total power absorbed then goes up proportionally to that swept area. So this kind of simple fact that you can get four times as much power by doubling the size of the blade length has driven this sort of steady race to build bigger and bigger and bigger turbines.
- [50:57]
I want to talk briefly about the history of this, because I think with solar, there's this neat industrial history where the first selenium cell is invented in the 19th century. Bell Labs gives us something like the modern silicon cell that has then undergone improvements over time. With wind is a little more complicated because, like, we've had windmills since Egypt. I mean, the idea of a windmill, the idea of harvesting energy from the wind, is like, actually one of the first places that we harvested energy. So in some ways, there's less of a kind of clear economic history in the same way that there is with solar. You don't have the same kind of charismatic objects like satellites going up into space. You don't have pocket calculators. However, there are interesting facts about where the modern wind industry comes from. So my sense is that the modern wind industry comes in some ways out of Denmark's and Dutch companies kind of reaction to the energy crisis, that after the energy crisis hits in the 1970s, you have Denmark, Netherlands, and northern Germany, with their historic shipbuilding industry, respond to the disappearance of cheap energy from the global economy by saying, okay, well, let's specialize in making wind turbines. And because they already have the shipbuilding convert, a lot of the industrial expertise and even equipment they already have into making turbines. Also there's this phenomenon that happens in the solar industry where Germany in the 2000s with its feed in tariff is able to solicit a lot of demand for solar production. That then happens in China. And even before there's a Chinese industrial policy to make solar panels, there's Chinese entrepreneurs who are looking at the boom in demand that's coming from Germany and being like, oh, hey, look, we need to meet this. And that gives rise to the Chinese, the modern Chinese solar industry. There's a version of that for wind, but it's actually California is the locus of demand. California, in response to the energy Crisis in the 80s and 90s, installs, decides they're going to pay a lot to basically bring wind energy to scale. It's Danish companies and European companies that respond to that and are where the production happens and then the installation happens and in California. So in some ways, a lot of the economic policies that we're more familiar with, that you're more likely to hear about with solar, were first pioneered with wind. But in some ways the policies that turned solar into the absolute juggernaut of global energy production that it is today were pioneered in the wind industry.
- [53:29]
Yeah, the wind industry is like a decade or 15 years ahead of the solar industry, kind of in all aspects, just because of when it took off commercially. And yeah, you're right, it was primarily fueled by oil for their power sector in the 1970s. And so when the energy crises hit in the late 70s, as they go into the 1980s, they're looking around for what else can we do? And they decide we're going to go into wind power. We're going to a bunch of windy rocky islands and shores here and we have, as you said, the industrial capacity to do it. And so we're going to start designing wind turbines. And then both Denmark and then little bit later, Germany in the 2000s, California in the 1990s created these niche markets that allowed these companies to scale. Vestas is the global Danish leader. I have a Vestas wind turbine behind me, a LEGO set, Vestas wind turbine, because Lego is also a Danish company. So if any of you have watched us on our YouTube channel, you've seen my three foot tall Lego wind turbine. Vestas was the leader there coming out of Denmark. There was also actually a US company as well called Zond that also built turbines here in the us they eventually went bankrupt in sort of some of the policy booms and busts, and they were actually bought by Enron. Right. Famous for its power trading shenanigans in the California energy crisis when Enron went bust, that was then bought by ge. And so GE Wind is the kind of US leader in this space and it continues to produce a large number of the wind turbines in the market. So Vestas and GE were the early leaders. Now there's a broader range, but there was some US manufacturing here too, to give the Yanks some credit in the story as well.
- [55:03]
So wind, like solar, as we've been saying, connects to the power grid via inverters, which are power electronics that tell me if this is wrong, but they sense basically what the current 60 Hz pattern is on the grid, and then they sync the disparate electrical signals they're getting into that 60 hertz. There's a sense, right, that, and we talked about this in the Spanish blackout episode, that inverter power isn't quite as responsive as maybe what historically has been thermal generation or velocity. Can you just say why is that briefly? Is it because there's not the automatic pull through energy that is happening when you have a big generator spinning? It's because you have to sense that the change has happened and then accommodate it within the inverter?
- [55:45]
Yeah, that's right. There's no physical inertia going on because you don't have that big spinning hunk of mass that's going to speed up or slow down or carry some of that kinetic energy. When the load on the grid changes, however, you can design the inverters to respond very quickly. It's electronic, so it goes at the speed of light. Right. So in the speed of computing. And so you could design power electronics that instead of just trying to sync to what's going on in the grid, actually try to control or contribute to the frequency or voltage on the grid. This is an area where basically the power electronics are there. What we need are the control strategies that make all of those millions of distributed inverters all sync up in a way that, that work together rather than fight against each other and can help stabilize the grid rather than simply respond to oscillations or shocks in the grid.
- [56:36]
So everything we've been talking about today in terms of renewables goes back to the sun.
- [56:39]
Right.
- [56:40]
Whether it's solar itself or the wind, which is just the air moving around because of the sun, mostly. But there is another source of renewable electricity on Earth, renewable energy on Earth, and that's the Earth's core. That's geothermal. And there are a new set of companies which we're not going to talk about in this series, but we've talked about in other episodes that are trying to harness that geothermal power and put it on the grid. Ever or fervo. Do those companies, does their generation techniques produce something that's more recognizable to us as a thermal generation technique? Do they produce inertia?
- [57:12]
They boil water.
- [57:13]
Are they. They boil water? They don't.
- [57:14]
Well, actually they boil some kind of other lower boiling point working fluid typically because they don't produce as high temperatures. And so they use a working fluid that is usually some organic compound that boils at a lower temperature and so can generate more energy from that low temperature heat. But basically same thing, Boil a liquid, spin a turbine.
- [57:33]
That's the basics. Those aren't inverter technologies, in other words.
- [57:36]
Those are not. Yeah, not as traditionally constituted. They're just another Rankine cycle. Actually, they're called organic Rankine cycles because they use these organic working fluids.
- [57:43]
But there are Rankin cycle, another Rankin cycle. Let's conclude by zooming out. So we've talked about solar and wind, from their humble origins to their economic boom, to how they put electricity on the grid. Where do they stand in the US mix today? Like how far have we come from the mid-1950s, when the first solar cell was invented in Bell Labs, to now?
- [58:03]
Yeah, we've come pretty far. Just a decade ago. Right. In 2005, wind power was 0.4% of U.S. electricity. And solar utility scale, solar was 1, 100th of a percent.01%. So basically nothing. Now wind is over 10%, about 11%. Closing in on 12% of our U.S. electricity supply. That's a lot more than hydropower now and about half as much as we get from nuclear power. Over half of what we get from nuclear power and solar, including distributed, smaller scale solar on rooftops and parking lots elsewhere, is about 7% of U.S. electricity. And both of them are growing fast. Solar in particular. If you look at what's been added to the US grid over the last several years and what's in the queue to be added going forward, it's like 98% of that is wind, solar and batteries. Batteries don't contribute to the supply, but of the supply side, it's almost all solar and wind in the queue. And so it won't be long before they're a much higher share. They're already combined, closing in on the contributions to our grid from coal or nuclear, each of which provide around 19 or 20% of US electricity. And together wind and solar are about 19% now. So come a long way.
- [59:16]
We'll have to leave it there. That does it for today's session of Shift Key Summer School. We'll be back next week with a new episode. Thank you so much, as always, for listening. Shift Key is a production of Heat Map News. Our editors are Ghislaine Goodman and Nico Loricella. Multimedia editing and audio engineering is by Jacob Lambert and by Nick Woodbury. Our music, as always, is by Adam Komalow. Thank you so much for listening and see you next week.