[Music] you might yeah what's fun right great thanks a lot Todd great to be here nice to see such a full room I'm really impressed that you guys show up on a Friday afternoon to talk about a technology that's outside your department but I'm sure pizza doesn't have anything to do with that either so but just to start out my perspective and this goes through my research and teaching and why I chose to write a book my perspective is that I think we can make better choices about the technologies that we're going to need to develop to address energy problems and climate problems if we better understand how technologies have developed in the past and especially if we can understand how successful technologies have developed and solar has been a successful tech for the trajectory that Todd just outlined so that's that's kind of where I'm coming from and really when we're talking about where solar is and having access to cheap clean solar electricity that's helpful but I really think maybe even more helpful is the lessons that we can learn and apply to other technologies and maybe there's even nuclear technologies that can that can pick up something from some of the things that have worked with solar and so that's where I'm going with this Solar story it's what can we is what external validity is there in terms of generalizing this case and think about where it might apply elsewhere so the three points that I'll make in terms of answering this question of how did solar energy get cheap is one that no one country did it no one country persisted and taking the technology from an idea to a commercial product to widespread deployment every country that was ever leading in whatever aspect of the technology was most important at that time gave up their lead after a few years every company that was the most dominant producer of solar panels in the world never stayed like that for more than two or three years it was really a sequence of countries and companies within them contributing something unique and distinct and it was much more instead of a competition it was much more of a relay race of a sequence that really goes in the order of US Japan Germany Australia and China and I'll focus on three of those cases when I talk about the solar case so that's the first part no one country did it the second part is how important international flows of odds call it knowledge international flows of knowledge were in supporting that chain of developments and the value chain that emerged from that it was flows of information moving around the world so you can think of scientific publications and scientific conferences it was maybe non written down know-how that was in people's heads that got moved around the world when people relocated from one country to another and took expertise with them and applied it in some new domain it was machines moving around especially production equipment was crucial in getting the cost of solar to come down and that moved across countries easily and then ultimately it was finance that was moving across countries and the final products being produced and that could be shipped anywhere around the world quite easily so as international flows of knowledge embedded in people in machines and in final products were crucial to catalyzing this global innovation process so that was the second main point and then the third main point this is really about external validity applying it to other technologies is that even though solar was a success it happened way too slowly so that we're thinking about taking the solar model and applying it to something new that we think we might need for energy and climate problems or thing might be useful for those we need to speed that process that I'll outline in a few minutes up by a lot and to one comparison I'll put out there probably like on the order of a factor of four so it took 60 years to go from the first commercial purchase of a solar panel which is by the US Navy to the two cheap solar where we're talking about Power Purchase Agreements $120 per megawatt hour sixty years is too long to apply to other technology so a key contribution that I guess I'm continuing to work on is how do we accelerate this process how do we speed it speed up this development process so that's where I'll get to and let me just you know kind of talk about the story a little bit but let me first start with a little context first I mean I want to acknowledge my research team that I worked with a really interdisciplinary team they brought a lot of expertise a lot of international experience that was really helpful but also made it really fun to work on this project those were great people to work with throughout the research phase of this project the second one and I'll just make this point I make this point in my classes on energy policy it's kind of a central theme of the whole class I start up by trying to tell people that transitioning our energy system to deal with energy problems and climate problems is hard for three reasons so one is that we disagree on what we want from our energy system I think we'd all want an energy system that's cheap clean and reliable but we disagree and have very contentious debates about which one we should prioritize should we prioritize the affordability the environmental impact or something has to do with security and reliability and we've had really hard discussions and choices and people have really different perspectives on what's most important there so my perspective is innovation can help relieve some of those trade-offs and have solutions that achieve all of those in a less contentious format but for now we're often dealing with these trade-offs second if you look historically and if you look at transitions from one primary energy carrier to another this is a hundred and fifty years 160 years there it took on the order of half a century to go from a bio-based economy which is what we had 150 years ago to one on which the Industrial Revolution and a lot of economic growth came from which is one based on taking advantage of the chemical bonds and coal and then once where transportation became huge and oil became the dominant energy carrier but even though you know we knew in the early 1900's that oil is transformative for mobility it didn't become a dominant energy carrier until the mid part of the 20th century and so this slowness of transition probably in large part because how long stuff lasts in the energy sector is really different than other areas but it's something that kind of has an inertial force on how things look and so we might need things to change faster than they have in the past and then third if we think about the climate problem when you put co2 in the atmosphere and this is a really crucial point that I think a lot of people miss and what makes it so different from some of the problems that we've dealt very successfully with in the past environmental problems like taking sulfur and getting rid of sulfur and getting rid of acid rain and dying forests and lakes when you put sulfur in the atmosphere it rains out in a matter of days to week same with particulates that lead to health problems when you put co2 in the atmosphere it stays up there half of it for something like a hundred years so it leads some more kind of inertia more slowness of the system to react so one thing that in making these points I realized I was getting really good at making these points and as a consequence I felt like people are coming away from my classes my talks quite discouraged about us actually making the type of change that could deal with some of the problems that seem real because of the contentiousness of the baits the slowness of the system to react and then the the inherent inertia in the climate system so I started to myself starting to kind of write down and if ultimately closing some of my courses with reasons why I still continue to work on this problem even though I think it is really hard for these reasons so I started putting together a list of reasons to be optimistic about dealing with some of our energy and climate problems and it would be fun to do a whole talk on all these reasons for optimism but I have to say over time this list has gotten longer to me it's also each of the points on it has gotten stronger and think more convincing to myself and to others and then also realize that for other people there's completely different points that they would see as their reasons to continue working and be optimistic about it but I think the number one that I still find most compelling is that even in with climate change 30 years of really going from strong awareness of the problem and politicians dithering about we should do about it in the background we've got technology improving and I'll say in a few minutes it's on a variety of technologies of them getting more efficient of using less materials of reducing the manufacturing costs of performing better of getting starting to get consumers more comfortable with them and so I think that is really underpins a lot of these other elements of reasons to expect we can actually have some of the changes that we need and a lot of it has to do with the technology underneath it and so that's what really informs a lot of my work or motivates it and that's what this book is really about is on this point number one and and why I think there is really reason for optimism so the second point I'll just put up here is just to make the case with a little bit of us most of us data but a little bit of others is to look at how far things have come and it's not so much about where we are now is so low that I'm interested in smore the trajectory it's a slope of this line so this is 2016 dollars per megawatt-hour I guess I don't need to explain what a megawatt hours in this room but in other rooms that I've been in I do say that a megawatt hour is the average residential electricity use for a u.s.
Single-family home and so you can think about how is your monthly electricity bill in terms of these units so it's like dollars per month of electricity so grid electricity is something like 120 dollars a month and that's been pretty stable if we look at the average cost of solar that's been going down over this 20-year period here and then these blue dots are a million data points of US households that have installed solar you can see a lot of dispersion there so some really expensive systems others that are cheaper than grid electricity and then we've also got a larger scale these are utility scale plants in the US and then we've got a few at the bottom there those open circles right here that are utility scale plants in other countries in very sunny locations and this is all took out all subsidies out of this one so if we just zoom in on this point here you can see that some of these plants are down around $20 per megawatt hour those red dots on your right for 2030 this is a an experiment we did in called expert elicitation that we did around 2008 2010 where we found solar experts interviewed them for half an hour to an hour each tried to elicit probability distributions in their head about what the cost of solar would be in 2030 and so we have their highest cost outcome worst case that we call that 90th percentile most likely case and then in a low cost case like at their 10th percentile and now taking those 2010 interviews or expert elicitation z' that's the red dots there and comparing them to Power Purchase Agreements in multiple countries we've now got Power Purchase Agreements in 2018 that are below the most optimistic elicitation from the most optimistic expert for 2030 so people got surprised even those that you would think would have motivational bias to talk about how cheap solar is going to get they didn't expect it would be as cheap as it is and so I think that's interesting that it's beyond expectations what I really want to get at though is come the slow for this trajectory of the the line because Soler's had this rate of improvement that's been more than any other energy technology and so I really want to do is try to understand that the other thing that makes it interesting is it's not the only one it's not completely a case by itself so if you look at solar PV if you look at wind and batteries they've been on similar long-term cost declines though when one goes a little bit longer because we had some expert elicitation about the the future batteries the slope of that line looks quite a bit like solar and potentially with a lot more room to run in terms of power and energy per cost of the of the batteries so there's there's more going on here than just than just solar so I guess the other question I had to answer in this project is why write a book so I've been working on solar data since the beginning of grad school for me and I've been really you know collecting as much data as I could got some pretty nice data sets like this million homes in the u.s.
That we have prices for every single system that went in and you can identify some of the factors and identify some of the cost reductions and I've done a lot of that work with econometrics models and bottom-up engineering cost models but along the way I kind of had this feeling that there was missing there was things missing things that we were hard to measure there were hard to operationalize and turn into variables that you could put into a model and that was what this book project was really about was to identify omitted variables things that seemed like they might be important they hadn't shown up in in some of the quantitative data before so that's why I wanted to try to write this book so it was three questions how did solar become cheap why did it happen as it did in terms of how long it took and then how can we apply that to other other technologies so let me just yeah show you what I found here the first thing I talked about is I talked about some of the solar costs if you go all the way back to the late 1950s that's that first commercial solar sale it's four orders of magnitude of cost reduction so a factor of 10,000 from the first solar cell that was originally produced by Bell Labs and then an entrepreneur called les Hoffmann late 50s started making small ones and then got a contract to the US Navy to power one of the first satellites after Sputnik called Vanguard 1 and so that was the first application of solar was a launch in 1958 and that solar tiny solar cell to just transmit information back to earth costs about three hundred thousand dollars per watt in today's dollars so that's an expensive member or two thousand three hundred thousand dollars per megawatt hours that's your electricity bill if you did your whole household on those little Vanguard one solar satellites so that's where it starts there's a front page of the New York Times the day after Bell Labs made their big breakthrough there that said here's the technology for the future solar could power the Earth's energy system and that was big expectations and really nothing happened for the next twenty years and really the big push in the u.s.
Was on nuclear power in the 1950s and the 1960s and there was huge deployment that happened after that push then we have 1973 October Arab oil embargo the Arab oil States refused to sell oil to the US to Japan to the UK the Netherlands in Canada and as a result those countries but especially us and even more so Japan decided to get really serious about finding ways to produce energy domestically and that launched project independence that was President Nixon's plan to have the u.s. import no foreign oil by the end of the decade by 1980 and that wasn't achieved but there were a lot of improvements that happen but there's a lot of funding that went into R&D and a lot of it went into nuclear a lot of it went into finding ways to turn coal into synthetic fuels and then a little bit went into solar but for solar that little bit was gigantic it was like a billion dollars of R&D that went in in the late 70s into the early 1980s and he had a couple things that came out of that so technical improvement so efficiency went way up there was it entrained a law people so people moved from the space program the Apollo program that had just finished up and we're starting to work and taking some of their skills to photovoltaics and people came from other industries as well and one of the most important people there was a guy named Paul macaque he came from Texas Instruments we're in the 60s he'd been working on building calculators and they had developed this idea and we're measuring it that as they built more and more calculators the cost per unit were coming down and there was kind of a idea behind this of the learning curve that as you build more units you find ways to spread fixed costs over more and more units or you make incremental improvements or you get experience that leads to developing new ways of doing the manufacturing and so Paul macaque said why couldn't we do this for solar and so he came up with this graph here and this is me kind of replicating his graph all the data except for one point and all the axes are from Paul macaque from 1975 and so what he did he just took the costs from 1958 to 1974 and applied what he called a learning rate so a learning rate is what happens to costs when you double cumulative capacity produced so he said with calculators and the other things Texas Instruments had built it's between a 10 and 30 percent cost reduction for every doubling and 20% is a good kind of mid-range number to use and so he just forecasted that into the future using those early 15 years of PV production data he had one interesting way that he thought about it too is this calculation here and so here he was saying this is the extra investment it would take to get solar down to what he thought would be grid parity so where we wouldn't need to subsidize it anymore and so in a way you can maybe think of it as a triangle that starts at whatever number you think is grid parity and goes up to here and you need to subsidize that amount above the regular electricity price in order to move capacity to the right on this axis and that costs would follow and then I just plotted 2018 on it and you can see Mecox projection is pretty prescient it really he didn't take that much data but there's something about this learning curve that did seem to work really well for calculators and it's worked really well for solar I mean well in terms of being pretty accurate prediction of what happened and then there's also a real policy implication that comes out of this idea is that you know maybe what you do is you get companies to pay for this or maybe you get the public sector government that's subsidies to do this so that eventually you don't need to subsidize it anymore I think that's one of the most important ways to think of subsidies for solar the subsidies aren't really trying to get us the benefits of clean energy the subsidies are trying to get us to the right on this learning curve to get the cost down so we don't need to subsidize it anymore and if the costs keep going down then you start having social benefits say this is great electricity that start paying back the subsidies and remember this is a log scale so a small triangle here is worth a lot of public benefit and this is the subsidy that you need and so if the one country that really took this idea and ran with it was Germany in the early 2000s and they did a subsidy program that added up to about 200 billion dollars which is a lot it's not you know a giant country but it's pretty in line with with make Hawks estimates so if you take nineteen seventy five dollars and account for inflation that 10 to 20 billion dollars looks like something like seventy billion dollars in today's dollars so maybe the Germans paid too much and there's a lot of people think that we could have gotten the same outcomes from the Germany subsidy program with maybe half that much subsidies there's something about this projection that I really like because it got a lot of a lot of things right I'll come back to the German case in just a minute so going back to this overall picture we've got project independence project sunshine the other thing that was an important development and this is one of the things I think was really lost in the soul Solar story was there was a policy in nineteen seven US federal government called the public utilities regulatory Policy Act and it was just a way to say that small-scale producers of electricity should be allowed to put power onto the grid and should be compensated for it by utilities and then also that utilities needed to compensate those producers at the utilities avoided costs so what it would have cost them to produce that power separately and so it's kind of an arcane federal law from the 1970s that gets mostly forgotten about but it starts to get implemented in the 1980s in California and really for the wind power sector so we start seeing it implemented in California and it's it's really hard for these small-scale producers to negotiate with the large utilities and so what the Public Utilities Commission in California did is come up with standardized contracts so it said here's the legal language it's all right here we'll come up with a schedule of rates for the next ten years and if you qualify for being able to commit to provide reliable power to the grid you can get this reimbursement rate it was like 12 cents per kilowatt hour so a pretty generous rate and you're guaranteed that rate for ten years and that became known as the interim standard offer contract number four that's that data point there from 1985 and it was dramatic not for solar solar was still too expensive but it was for wind so about two billion dollars of Wall Street investments went into California wind farms in kind of three mountain passes in California in their early net into mid-1980s because they were getting this almost a guaranteed a guaranteed return for ten years on the electricity they produced and people around the world saw that that you could catalyze investment by giving certainty to the producers and the Germans especially saw that so that was a policy innovation that the Germans later pick up I'll talk about that in a minute the next one are talking about is Japan and Japan keeps the solar industry alive in the 1980s and 1990s the u.s.
Really kind of divests from solar in the early 1980s it cuts R&D funding over two or three years from 1981 to nine in 84 by almost 80% so it gets down about 20% of the highest level which is 1980 to 1981 and so the research frontier goes elsewhere it especially goes to Japan a lot of the people move around and start attracting interest in Japan electronics companies in Japan thinking of Panasonic and sharp start getting interested in solar they had some experience with it from this R&D program and started doing it in in funny ways they started putting tiny solar panels into watches and calculators to distinguish their products a tiny market makes no difference to global energy supply but it got large companies familiar with the technology and they started seeing that there was a way to differentiate their product and start to use it a little bit and then all the sudden there's industrial support for what Paul macaque would have liked which is a subsidy program and that's the rooftop program in Japan from 1995 to 2004 so about 200,000 homes install solar on Japanese roofs it's the first time that we give a subsidy to consumers so you get about half off a half value rebate for your solar system and to see people actually embrace it so 200,000 customers decided to take that on and get 50% off discount on their on their solar system so that was a policy innovation the other innovation that Japan did and this also fits with this mayhawk learning curve idea is that they weren't going to continue that subsidy forever they said it's 50 percent in year 1 it's going to go down to 0 percent in year 10 and it's going to go down on a predictable schedule each year and so that the subsidy will go away over time with the idea that we wouldn't need to subsidize overtime because there's a technology gets better and gets to scale the cost would go down and so that was another policy innovation and so the Germans take that up next so I'll do in my own little picture for the Germans here so and I've got the sequence on the top of the slide here but just to really simplify it if you had to if I just just take a break to do the overall picture here if you think about three countries making contributions I think that explains a lot the us-korea the technology the Germans created a market and the Chinese made it cheap and it was in that sequence and it made sense to be in that sequence and each of those places added as distinct set of efforts and capabilities that really led us to the solar today so I just want to highlight the next two parts of that story the German part and the in the Chinese part so the German part first of all and this is really coming from political science and my policy work a policy window opened up so in 1998 the Green Party became part of the ruling coalition for Germany and the Green Party had been advocating for solar for two decades had a few experiments in German cities and all of a sudden they had their opportunity and so it was hung Josef fell on the bottom there who is the head of the Green Party and a social democratic leader named Hermann Scheer on the top there and together Hans Yosef on the bottom wrote the policy and Hermann Scheer sold the policy and when this it was only a few years that they had this coalition with the Green Party in power they passed the subsidy program and it was the writing was an important part of it they it's a process of policy diffusion so they took this idea from the u.s.
From the California implementation of purpo these guarantee contracts they saw what happened there so they said okay we'll give above market prices and the Germans gave way above market prices like 52 cents per kilowatt-hour and the California was giving 10-year contracts the Germans gave that for 20 years so it was an amazing deal for anyone to jump in to that market and as a consequence people were borrowing money at like 2% because it was such a guaranteed return for 20 years there so they took the idea from the u.s. of these guaranteed contracts and they took the idea from Japan of subsidies direct to consumers and a declining rebate schedule and that led to the passage in 2000 of a quotes called a feed tariff that was really strengthened in 2004 and then really created a demand so that's what we sometimes called demand poll so these are installations in Germany and you can see there's a little bit of rise in the early 2000s but with that policy change in 2004 the German market increased by a factor of four so 300% in one year and rose after that and stayed high for like five or six years and all of a sudden solar was taken seriously as a business opportunity and there was a lot of kind of stability and credibility because of the way the policy was designed with these 20-year contracts and so what that really led to what later became the Germans called their gift to the world and this is a little bit ironic in that you know the Germans spent 200 billion dollars on these subsidies eventually all the men almost all the manufacturing happens in China and people are kind of like well why did you do this Germany and so there's a few different ways to answer that question one is that the equipment suppliers in Germany have done very well supplying to the Chinese and the consumers of electricity I've done very well buying solar cheap solar panels from China but also it's our gift to the world and so this was this idea of kind of Germany being a eco innovator and leader but when I say the gift to the world I think of it a little bit differently it's they created this market that was so large and had so much credibility behind it that it led to investment and it wasn't just lots of people saying this is a no-brainer to spend ten thousand euros and put a solar panel on our house and get a large return over 20 years there was that investment but the really important investment was that manufacturers and even further up the supply chain equipment suppliers to manufacturers people that were producing specialized equipment like this wire saw here that would saw silicon ingots really thinly so you get many many solar wafers out of the same amount of silicon they could now start producing equipment specifically for the solar market because it was big enough to devote R&D to devote engineering resources to instead of what they had been doing was just take cast-off secondhand equipment from the semiconductor and computer industry and repurpose that for solar and their sir a lot of efficiencies and cost reducing opportunities that came out about using equipment just for solar because you didn't have the need for purity and the concerns that are really important in the semiconductor industry when you're just trying to convert sunlight into electricity so that led to all this investment and these machines that came up so that's the German story and so I really think Germany's gift to the world is this creation of a market its stimulated investment in China and then in the u.s.
In Switzerland and other places that were making these machines that the Chinese then bought so that brings me to the Chinese part of the story oh and this is just some of the if you say why did it get cheap so plants got bigger the efficiency doubled over this period here silicon prices went way down wafers got wider and they got thinner so thinner because of those wire saws and then yields increased as well so all these changes were happening as we're at this time of increasing demand and then the so the China story and so if the u.s.
Made the technology Germans made a market a Chinese made it cheap and this is how it happened and it really is two phases it's one the early 2000s and then the second phase is after 2009 and in that early period its entrepreneurial it's really the Wild West it's a few scrappy innovators pulling some things together and scaling up and be very successful with that after 2009 the central government gets much more heavily involved it sees the industry as a winner and it champions it and funds it a lot but it didn't really fund up much at all before 2009 so this early period the scrappy startup I think one of the most kind of revealing anecdotes or kind of stories that makes the case for this international flows of knowledge was this group in Australia that started hiring Chinese students and it was in the early 1980s Deng Xiaoping had this program for a thousand Chinese students to go overseas and come back to China to see what they would learn and to see how they could make use of what was going on in the rest of the war and one of those thousands went to the University of New South Wales in Australia and he had some experience with micro electronics and he was hired by the guy in the left near their name Martin green and he had good experience with that student he said he was quite well trained and he could contribute to the lab and he hired three more students over the next 10 years and then in the early 1990s he hired another student and that's a guy on the right in the picture there and in 1994 he brought that student to China to explore the possibilities of setting up production there and that student then was a translator for the trip that was that was only role he had and they came away from that trip just saying there's no possible way you could manufacture in China there was no infrastructure there's no supply chain there's no interest by any of the cities and setting up manufacturing there there's certainly no investment to do it and so they left in after that trip in 1994 and they went to India to set up manufacturing of solar in India and that didn't work either so then they went back to Australia and set up a company called Pacific solar in in Australia about five years later that translator started talking to his Chinese friends back home who were saying you know things are starting to change here I should take another look and so in 2000 he went back and started talking to cities about setting up manufacturing and still it was pretty hard to get something to happen but some one thing that had changed is that cities now could retain some of the tax revenues from manufacturers and other companies that were located in their cities and so all the sudden the cities were starting to have incentives to attract investment and to set up manufacturing and so they were able to get five million dollars five million dollars to set up a tiny manufacturing line and so this guy this young donkey who's in the picture there travelled around the world bought secondhand equipment brought it back to China and set up a manufacturing line and this was just around the time that the German market was starting to take off and so he said you know we can sell to the Germans and he tried to sell the panel's to the Germans and the Germans were pretty skeptical they were very you know proud their own engineering process quite skeptical of the Chinese Chinese quality of those panels but she had credibility because even working in this University of New South Wales and they were achieving world record efficiencies on the cells and so the Germans really respected that and saw that he had the technical credentials to make something that would be reliable and and technically sophisticated and so he started selling to the German market and started this process of iteratively scaling up in by 2005 that company that he started I'm creating the name of it right now oh man what is that name of that it'll come back to me oh my god I've since so much time researching this company anyway okay so he starts this company it goes public on New York Stock Exchange raises 300 million dollars and so all of a sudden Solar is a legitimate company so this just to put this in perspective again there's international flows of knowledge and and money in this case so it's us pension funds that are taking their retirement investments putting into the stock market that stock market is going into these Chinese solar companies including this first one here the Chinese are taking that money so it's four hundred million dollars taken to China and using that to buy equipment from the u.s.
From Switzerland and from Germany taking that equipment stalling it in China and then producing panels that they sell to Germany later to Italy to Spain to California and then ultimately China itself becomes the biggest solar market and so a few things that happen with that so we've got used equipment coming from the US we've got pension funds and then we are selling to Germany so these international combinations of knowledge are so crucial to this whole to this whole story and this guy's young ground she who was a translator on that 1994 trip became the richest person in China from 2005 to 2007 and then they ended up buying an Italian company that some fraudulent loans and the whole thing went under and so by the time I talked to him about a year and a half ago he he didn't want to meet with me he needed to be a little bit quieter about about his location so but one thing that even though his story didn't end very well the solar industry really legitimized itself by this pathway especially being able to sell to a market like Germany and be able to attract investment from the yes those two things really legitimized it so that by the time we have the global financial crisis in 2009 and 2010 the central government's willing to put twenty maybe thirty billion dollars of low-cost loans into Chinese solar companies to keep them afloat when these Western subsidy programs were ending because of the global financial crisis in 2009 and 2010 and then eventually the Chinese themselves start subsidizing solar and have their own feed-in tariffs in them by 2013 China is the largest market in the world for solar and then just to two aspects of the Chinese case that are important to appreciate is that in the 2000 to 2008 period Chinese labor cost advantage was very helpful it was so much cheaper than trying to do production in Germany and some other places like the u.s.
At the time and so that was a big advantage but by 2008-9 because that investment in equipment and machines that was really developing or response to the German market it was almost completely automated and so the labor cost advantage was not really serious it was really about how do we get autumn a Shinto go well and to integrate our manufacturing lines and have supply chains that are very tight with lots of information flowing very quickly so that was a big part of the Chinese story was lots and lots of automation and the other part is that top graph there lots and lots of competition so when the US was dominating solar in the 1990s we had companies with 30% of the world market when Japan was in the early 2000s that's sharp with 25 to 30 percent of the world market but even with the Chinese companies becoming giant and scaling up and even than this with this original company whose name I finally remember called Sun tech none of them ever got above 10% so it was extremely competitive and so margins in the industry go way down that was another big reason why cost came down as well so this is kind of you know China's gift to the world then is this cheap electricity and the potential to have you know 30 or 40 percent or 50 percent of electricity by mid-century from solar ok so and then this is just a picture because it's actually pretty hard to get into these places of solar manufacturing in China and this is to make my point about automation like how many people do you see in this picture it's a few reflections of people gawking at the machines doing their things here it's really about pulling these machines together optimizing them and they used a lot of know-how from Applied Materials in the US and some German companies to come in and string those together and set up their production lines but now they're doing that on their own and they've also they're making some of these most of these machines on their own as well so just to now with the last few minutes here to make an effort towards external validity so how can we kind of come up from all these stories and anecdotes and history and think about it in kind of a stylized reduced form way that we might apply to others and the first thing I think is helpful to think about is is these kind of three stages is creating a technology building a market and then getting the costs out and in terms of creating a technology Einstein played a key role here so his nobel prize was for a paper on the photoelectric effect in the scientists at Bell Labs that came up with the first efficient solar cell used his idea of activation energy as they were working on making these PN junctions on the original solar cells we had our D that was crucial throughout this time but changing the focus so looking at configurations and design but then looking at different materials and our D on manufacturing itself that was really important throughout the time and then knowledge spill over so it wasn't all confined to individuals it wasn't confined to companies and it certainly wasn't confined to countries either it moved around and that that's bed things up so they're on the creating a technology side those have been important components in terms of creating demand or creating a market creating reasons for an investment one is that that's a picture of Sputnik they're these niche markets solar had the benefit of niche markets all the way along from 1958 and the first satellite cell then it was oil rigs than it was people that wanted to be off-grid then it was the calculators and the toys and all along the way they were small but growing high willingness to pay but shrinking over time niche markets and you didn't need policy all the way along to make the technology continued if you could find places like up in orbit where energy was really expensive and that solar had some unique capabilities that it could be applied to a second part of the building a market story is the modular scale so that's something that really makes these niche markets work is you don't have to only have gigantic solar you can have tiny solar so the think of the cell in a calculator and then think of the largest solar power plant in the world it's probably the one being built in Egypt right now and it's about a factor of a billion from that calculator to this plant being built in Egypt right now and almost every scale in between has been used to satisfy some need in some market and so there's lots of different ways that you could find niches when you had that ability to have lots of different scales and then policy support certainly policies been crucial but when I say robust it's I guess it's got two meanings robust in that the German policy was very generous and so it's strong robust but robust also have this has this meaning of you know if you if you take away one factor or one thing changes you still have that incentive and that was the case because we had lots of different countries providing policies for solar so when Spain went into the financial crisis and cut off all its subsidies immediately the industry didn't fall apart because you had California building up its California Solar Initiative and then China created its own and so there was never one dominant market except for the German market for a few years that was completely dependent on a subsidy and thus completely dependent on a political coalition hand-holding or an election going a certain way it really created expectations that would be a demand because it was coming from multiple policies and then in terms of making a cheap there's the learning-by-doing aspect and that's been certainly helpful one of the things I got from interviewing people and talking about them setting up their companies was that they didn't go big immediately they scaled up over time they built one line got it to work a little bit built the second line bigger maybe with a different technology there was some small incremental improvements and so this ability to not get locked into one choice but be able to gradually make improvements along the way has been crucial for solar in general but for the manufacturing part especially and then a delayed system integration so now you know we think about the challenges associated with having an intermittent resource on the grid because it's not sunny all the time and there's clouds and thunderstorms and things like that and those are serious challenges that we have to address but you know solar developed for 50 or 60 years without having to encounter those problems and now it does but there's a lot of momentum behind it a lot of things have been worked up by the time it has to do with system integration and compare that to a large-scale maybe a nuclear power plant or carbon capture the large-scale plant we're building some of the first ones of those they have to get system integration right right away on a multi-billion dollar investment so that is an advantage that that Solar had as well okay so is it okay to go for a few more minutes okay I just want to put a couple of ideas up there for how we could use PV as a model and the first thing I'd just say briefly is it's not a model for everything and one thing I've started to work with is sort of thinking about solar PV as kind of one distinct area of climate relevant technologies that's got the characteristic of being having a strong technology component it has this iterative so that when I say iterative there's been about two billion solar panels have been produced in the history of the industry so lots of chances to you changed things learn and then disruptive this idea that by compromising on certain attributes in the case of solar efficiency you could get gains in other areas so in terms of costs of production so the Chinese were very savvy in producing cells are about two percentage points less and efficient so like 15 years ago this is going from about 16% to 14% but you could get the cost down by about half by making those compromises and so that what those are characteristics of solar that I think you could maybe apply to other technologies then there's other technology who might need for climate change that are really different like storing carbon in soils or trees that's really low-tech and distributed large scale where system integration seems to be the central challenge so like bioenergy with carbon capture and there we need other models so maybe we need to learn from oil refineries and chemical plants for those maybe the Green Revolution is helpful for understanding how we can get more carbon in soils and there's general-purpose technologies that will must be certainly helpful for dealing with climate change and we've developed general purpose technology successfully before so I really work on now is this first category how do we how do we learn from solar so let me put up a few points here one is that and this I've kind of gone through this story but you know we now have several countries where we're talking about close to 10% of our electricity from solar it's not capacity that's electricity and close to 2% of world electricity is coming from solar and so that looks pretty dramatic over this 10-year period here but if you look at my time frame I think this is you know way way too slow if we're gonna say we're starting in 1958 in getting to 10% of electricity by say 2020 that's way too slow and so part of the challenge here is how do we speed this up and so as an example I've been looking at direct air capture where you take co2 that's already in the air this is address this issue of co2 staying up there really long time and removing it pressurizing it and then sticking it underground so taking it out of the atmosphere directly so I presented this to the direct air capture companies they didn't like this picture but here's what it looks like so if you say it took something like sixty years here to go from the first commercial to low-cost and then we're talking about maybe 30 40 percent solar by 2040 let's take your direct air capture plants this is the first one that went online in 2017 in Switzerland so this is taking ambient co2 at 0.04 percent concentration and absorbing it pressurizing it sending it four hundred meters behind that plant in the upper right hand corner you can see a little picture of a greenhouse there and so that's who's purchasing this co2 they're using it to raise the parts-per-million in their greenhouse from about four hundred to six or seven hundred in increasing the yields on their tomatoes so 2017 that's a real plant there's a real customer so if you apply the solar timeline to that they're low cost in twenty seventy six and their widespread adoption by the end of the century that's not on the on the scale that we need to do with climate change these are some integrated assessment model results for how we would meet a two degree temperature target and if you look at those we need to start be deploying these negative emissions these removal technologies by 2030 in a serious way and scaling it up to billions of tons by 2050 maybe five to ten billion tonnes a year so just to put that in perspective it's like taking out of the atmosphere about a quarter of what we put into the atmosphere now by 2050 and so they need to do something if they're gonna contribute to this way faster than solar and just to put that in these simple terms here they need to speed things up by like a factor of a factor of four and so what so my research is trying to do now is how could we oh yeah and this I was looking at scale up for the fastest solar companies here so log scale here these are Japanese the sharp German sq cells and Sun Tech and Jenko are Chinese and you can see how fast they scaled up their production if one of these companies climb Marx is one of them had a goal to remove one percent of the world's co2 emissions by 2025 they would have to scale up about twice as fast as these solar companies so it's not out of the question but it would take serious commitment and investment and the type of activity that we've been seeing in solar but more to do even 1% of emissions so and this is I'll just stop with this one here part of what I've been trying to work on is how do we get that factor for acceleration so we can learn from solar we can say what happened there but how can we actually do it faster and so this is just some initial ideas of things we could do to speed things up not for solar but for direct air capture or you know whatever other technology might think important so I'm thinking of it in the three rows here so on technology push so creating new knowledge on knowledge flows on disseminating knowledge and then on demand pull creating markets so the R&D I talked about already having a train for true trained workforce sciences scientists and engineers is crucial here public procurement I didn't talk about very much but that was a big part of the u.s.
Effort in 1970s was for the government to start buying solar panels that led to companies actually having to produce for a real customer there were technical characteristics about reliability efficiency that they had to meet and so that really got the companies become from kind of small scale batch producers to something more a little bit approaching Industrial on knowledge flows so codifying knowledge so taking it out of people's heads and putting it into maybe it's science when use reports and papers but probably more importantly data sets that people have access to and can make improvements on and understand what's happening knowledge spill overs I talked about that so knowledge going from person to another company and other country to another and the global mobility that I've talked about quite a bit already in terms of people machines know-how finance and then final products and then in terms of creating markets so the robust markets part is really important this disruptive production so being able to compromise on attributes that consumers don't think are that are important in order to address issues that they do think are really important and that might be something like cost for solar for electric vehicles it might be something like convenience or safety costs as well and then the final one I'll put there's political economy and that is to make the point that every country that had strong effort to create markets for solar and subsidize solar eventually got strong resistance from competitors to solar or other entities that stood to lose from from widespread use of inexpensive solar and in every country Germany Japan US and China those efforts have been successful in kind of blocking the progress of solar and probably the best example where this has actually been resolved was in Germany where they passed that strong policy that was in place for seven or eight years in the big way and about two hundred million dollar two hundred billion dollars of investment goes in or public investment but they made a big compromise they said that large electricity users so energy intensive industries don't need to pay for the subsidy only households need to pay for the subsidy and you know there's a lot of concern about fairness and ability to pay but that was a way to make sure that those large users didn't stop the policy I think if we wanted things to go quickly we might need to make those types of politically savvy moves in order to make things go quicker because a lot of these entities stand to lose from some of these developments and and can effectively slow them down in a lot of cases we don't really have time we need to be moving faster on these things so I'll stop there and happy to take any questions either now or after [Applause] so how does the solar panels trim that's the same oh yeah that's that's a good question I think that I think it's faster in computer memory because with computer memory you can do miniaturization you can do lots and lots of transistors in the same amount of space with Sol there's only so much sunshine hitting a panel and so you can never get above something like 30% efficiency anyway so you can reduce the manufacturing cost the material cost the thickness but you don't have this ability to Jam more and more photosensitive material into the same amount of space so that's that's one thing as made computers go go faster but there are a lot of similarities in terms of the iterations and the scale-up and even the material is the same so makes it harder for things to compete but you know there's places where you know this is in Chile and Abu Dhabi and so in Ann Arbor you have probably a third of the sunlight so that triples the costs in terms of dollars per megawatt hour so that doesn't look as good here so that's these are sunny places that makes it better it's even in Abu Dhabi it's not sunny all the time and so you do need access to low-cost storage so that that could be something that you have to add to this as well but that cost is coming down too and so you know maybe as storage looks like that if you add it on and then you've got space – okay Bobby has a lot of space Chile has a lot of space maybe you could in an arbor but there's a lot of other places where especially large mega cities where the energy density of energy consumption exceeds the space that you have for solar radiation even if you could take advantage of all of it or 20% of it so yeah so there are reasons for other things you know have niches to compete but it does make it harder to compete because not only these costs low but they continue to go down and if they're supplemented by stores it's low and continues to go down it's a moving target in terms of competing with it there's one yes you take the how much it cost to install the system discounted at a discount rate over a certain number of years so they're guaranteed for 20 years we use 20 years and for household systems we use the a little bit higher than a mortgage interest rate so you can calculate the amortize or the levelized annual cost that way and then if you take how much sunshine there is in each location so there's numbers for Ann Arbor and you know the efficiency of the cell you can figure out how many megawatt hours you can produce in a year and so you divide the levelized cost by the megawatt hours that you produce and that's our estimate here yeah yeah I mean I don't think it makes a big difference because that recovery factor kind of approaches the interest rate once you get after 10 or 15 years anyway so it wouldn't make a huge difference in terms of levelized cost yeah can't yeah right yeah there's no battery right so this is just saying it doesn't matter when you produce it we're gonna count those so there's certainly electricity valued at different levels for different places in different times and so that's not included it's just cost no value yeah as thing where the power companies are not so favorable they say cost them more money our son and we under therefore they're charging is more so could you say something about the advantages or disadvantages of trying to purchase power and a sunny part of the country offset our news here and how is that real or is that a shell game yeah I mean it can be real you just have to make sure that if you're you know paying for clean power that's being produced somewhere else that it's not just making all the other power more Brown so but there's ways that you know you can make sure that that's that's happening yeah but it does you know raise the issue about how much renewables should be happening in a certain place and whether we should take advantage of building renewables in places that are sunnier or windier and then do we kind of account for that financially with some kind of compensation or do we do it physically which means building transmission and you know that that can be a challenge but you know there's been a lot of efforts we have to to change a lot of our thinking I think I've been having students coming into my office for the last three years asking about this and why we don't have solar on all the buildings and I keep saying good question ask the Chancellor and she told me they should be asking you me so I had to come up with a different answer and then I said okay well let's look into all the state regulations because they're saying yeah they say we're not allowed to make our own power and we couldn't find anything on that and so the final thing the last thing I heard last week is that the university has a power purchase agreement where we get power for three cents per kilowatt hour so it's $30 per megawatt hour so that's here you know most my house has to pay this so if you're only having to pay this and you're not in Abu Dhabi but you're in Ann Arbor or Madison you know the numbers are up higher and so that's a big challenge is how do you how do you get the university to say okay we don't want this sweetheart three cent per kilowatt-hour deal anymore but we care about sustainability so much having cheap solar helps and having to get cheaper helps but yeah it's a real real challenge and so I think we're going to need not just solar but lots of lots of solutions to deal with that one yeah the people are receiving subsidies but they're not accounted counted here so the way the subsidies work is you pay fifteen thousand dollars for your system and that's what I count here and then if you get your thirty percent investment tax credit back you get that four thousand dollars out for your taxes but that doesn't show up here this is unsubsidized but people are purchasing these things knowing that they will get a subsidy okay yeah yeah well that's one thing about the the learning curve idea is that you know it's a log-log scale and so if you plot it on a linear scale it would look like it drops a lot and then it kind of flattens out like it's approaching almost almost approaching an asymptote so that is kind of what we're seeing with solar but if you plot it on a log-log scale it actually looks pretty straight I mean this is one version of it this is the other is that what you're okay it's so not cost but adoption yeah yeah I'm sorry I missed your question then oh the one yeah yeah the worldwide adoption of solar oh that one yes sorry that one mm-hmm yeah I mean these are market shares so it's not a total so you know oil has been growing and growing I guess it's been flat the last 10 or 15 years but you know there is but in terms of share yeah they do get to some level and then stop but you know most even the most ambitious people like people are most optimistic about solar see it getting into 30 or 40 percent maybe 50 percent it's not no one's not many people are talking about it being a hundred percent because of some of the things we talked about it's not sunny anywhere sometimes you have density maybe you don't have enough storage maybe there's even material constraints oh yeah I mean but but these are tremendously large numbers to get up to 50 percent I mean the one that's been fastest of any of them is nuclear to get up to something like 10 percent in this really short period here and so these are gigantic transitions even though they look like they hit a wall yeah yeah and that's a really big part of these other technologies of starting to look at too is it's something that's not been a big issue for solar but and it hasn't seemed to be a big issue for electric vehicles but could for some of these other ones like this direct air capture where you're taking co2 and storing on the ground people may not want to have lots and lots of pressurized co2 under the ground so we'll have to see about that one [Applause]