 Well welcome also for me. I've just come back from Washington so I can tell you that I'm very glad to be here. And I'm just going to talk a little bit about the kinds of things that we worked on at DOE. But I'm happy to answer questions pretty much anytime so this can be a conversation rather than a lecture. But I'll try to give you some examples of how we thought about the portfolio of research that makes sense for the country as a whole. With some examples because I guess it's that many of you in some way or another might intersect with some of these. And as I say, I'm happy to answer questions along the way. If you think about the energy system as a whole, we really want... So any chance we can get rid of the lights on the screen? We're looking for really three big parts of how energy interacts with modern society. One is economic cost matter. Energy is a cost that flows through into absolutely everything else. It's done as a commodity so the margins are small. And so our competitiveness internationally and nationally depends on how well we do that part of it. The environmental side is I think now so firmly established that political ups and downs will be part of this a lot to this. But the climate change is something we really do have to deal with this in this century. There are lots of other impacts I'll talk about those for in a bit. And then national security, ensuring supply and having... Just examples of Houston and Florida here in the last few days illustrate how hard it is to live your modern life if you don't have electricity, for example. Your mic is not turned on. How many engineers does it take to... There we go. I'll let that set better. So if you think about all those people in Florida and Houston that try to go to the grocery store, a grocery store can't actually sell you anything today unless they have electricity because they don't know what anything costs. It's all done with the scanners. And you say, well, okay, but that's just the way things have developed. So assuring supply and making sure that it's done in a way that's diversified and not easily disrupted is important too. So you say, well, okay, wrong direction. I kind of said most of this stuff that particularly on the economic side, it involves jobs on the environmental side. It's partly greenhouse gas emissions and climate change, but it's also very much about health effects. As the fraction of coal, for example, in the energy mix goes down, there are real benefits that come from fewer criteria pollutants and fewer particulates and those kinds of things. And I've already really talked about the security side. So the assigned topic here was innovation. And it's worth spending just a minute in thinking about how innovation happens. And up there in red is the kind of, well, I called it the standard model. I don't know if it's really standard, but people often say, well, especially those who don't want to spend money at the government level, say, well, it's okay to do some science, but if it's of any use to anyone, then industries should do the work and that's it. And so this model is that the university or national lab researcher discovers something cool and then writes a scientific paper, publishes it, and then hands it off and everything goes smoothly. My view is it doesn't really actually work that way, that if you think about penetrating big existing markets for energy supplies and even if you have some interesting new device, well, how about a Tesla, for example? You have to work very hard, one, to iterate enough to have that thing that be substantially better than what you can buy for less money somewhere else requires lots of iteration, lots of design constraints, lots of work to improve the cost aspects of manufacturing. So it's this deeply interconnected process that really requires all kinds of skills and not just the technical skills. So you'll see, as I talk about what DOE has done in the past, at least, it's an attempt to populate a portfolio of ways to approach R&D that will satisfy lots of these pieces. Costs matter, I said this before, but it's hard to say this enough that it really does matter how expensive something is. But costs are not set in stone. I'll show you in a minute that the cost of renewable technologies and batteries and LEDs, for example, have come down dramatically in the past. And then I would say that innovation accelerates when we have good connection between the sciences and engineering and the social science and behavioral side of all of this to the extent that we can create that conversation. The applications can illuminate the need for the science, the science can offer opportunities for the applications and we can figure out how to get all this in place in a way that actually penetrates the markets. So I said that costs are not immutable. This is just an example of that. These plots are typically done, this scale is not shown here, but it's typically done as you double the size of the market costs continue down. And it's because once people start doing these things at scale, they figure out better ways to do them. The solar market here, the solar PV market is a good example of that. The transistor market, the earliest transistors cost a whole lot, but now we can put millions of them in a tiny little space. So all these things driving technologies down, the learning curves is an important part of evolving toward the future. So this is a plot of what the costs, if you went out to buy a wind turbine or sort of distributed PV, that this would be solar cells on your roof or your backyard, utility scales, PV. This is battery cost, this is modeled for production at scale. And LEDs, you can see dramatic reductions over a pretty short time period. Now part of this was, you know, on the PV side, this was a big, big in developing market. Germany put in a feed-in tariff that encouraged lots of Chinese investment, probably a little too much, which led to competition that made it hard for companies in the United States, but overall the prices came down. And you can see the level of reductions, 40% for land-based wind, 50% roughly for distributed solar, and 60% for utility-scale solar. That's because those markets have really opened up. People are investing now in those kinds of renewables, and by doing it at scale, the costs have come down. LEDs are down by an astonishing 94% over this time period. It wasn't very long ago where you went in the store, you looked at the cost, and you said, I think I'll wait on those. But now the lifecycle cost of an LED here is just much, much, much lower than, and even the initial purchase cost. And I think there's actually a lesson in that. When you're trying to get to scale for something that costs a few dollars, that's different from trying to get to scale for a coal-fired power plant with carbon-captured storage or a nuclear power plant where you might be talking about $7 or $8 billion for the nuke anyway. And you just are going to buy more of the little ones, and so people can invest in that in a way that doesn't require such a big upfront cost. So anyhow, quite significant progress over the last few years. And a trend that says, you know, if we can continue to drive those costs down, then we'll have a competitive economy and one that is well supported. So a question that sometimes asks is, well, why should government do this stuff? Industry in this country and around the world has shown that it can innovate and certainly that's true. It's an important part of it. Particularly in recent years, industry has tended to focus on things that are nearer term, maybe the next five years to make it into something in production. And because margins are low, if you look at the, if you look at kind of industries here, pharmaceuticals, this is R&D as a percent of the sales and pharmaceuticals spend 17%, to get to automobiles it's 2.5%, utilities 0.2% and petroleum products down around a tenth of a percent. So, and if you look at the innovation side of things, there was a big burst of money spent. This is during the Recovery Act, during the recession, the government spent lots of money in these areas. But that has declined. China's spending on clean energy technology has increased dramatically. And I think they're hoping, I think, to capitalize on the fact that the US seems to be pulling back. Clean energy venture funds went through kind of a peak here and then I think the venture community said, gee, we're used to turning these things around in five years and going 10x on our money. That doesn't happen in energy technologies because we have to penetrate a big existing market. So I think the argument here is that there is in fact a role here, that this piece of investment in both the fundamentals and in the early stage commercialization of technologies is something that's not being done as effectively across the industry setting as it could be. So this all matters, of course, as well for being able to meet both the commitments of the Paris Agreement and the real underlying need, which is to deal with climate change. This is a set of, these are economic simulations run by a model called NEMS. And basically what we did here was to take the business as usual EIA estimates of energy use and then add in some cases where we put in a carbon price of $10 a ton or $20 a ton. The advanced technology cases where we took all the DOE goals for improving the cost of solar or any one of the other technologies and said, okay, we meet those whenever we plan to. And that leads to reduction in CO2 emissions. But it turns out you really need to push harder in order to get something like the trajectory to prevent more than two degrees C increase in climate change. So this is just another version of the argument that says, look, we need all the tools we have now plus some more tools that we need to invent in order to get where we need to go on climate change. So he said, okay, fine, I'm convinced now. What do we have to work with? And if there's any one idea I'd like to leave you with here, it is the idea of a portfolio. The thing about research is that there's a kind of endless supply of interesting questions to work on. And, you know, some of those things we're going to make real progress on. And some of them we're going to run into a roadblock either because some other piece of knowledge or design doesn't yet exist and we'll need that. But the thing is you don't know in advance which ones of those things are really going to pay off. So you do what everybody does with their stock portfolios too, which is to invest across the full range knowing that the sum total will produce lots of good stuff even though if you don't know exactly which pathway will be the most productive. So on one side there's the science program. Basic Energy Sciences does lots of the chemistry and material science and subsurface flow work. Nuclear physics, that's kind of fundamentals, high energy physics also fundamentals. This is particle accelerators and astrophysics out in the planet. Biological and environmental research, these are, this is where climate change research plus understanding biological systems and how they all flow into all of this. And fusion, both the International Fusion Reactor and the Domestic Program and then the Advanced Scientific Computing. So that collection of programs under the Office of Science, it's about, it's almost $5 billion. It supports 10 national labs including Slack here on campus and Lawrence Berkeley across the bay. So it's a big program. The applied energy programs, energy efficiency and renewable energy, that's where the energy efficiency side of course is one thing that we could do a lot better in this country. And you saw some of the results of the renewables work. Fossil energy, nuclear energy and something called electricity delivery and energy reliability. This is really the grid and I'll say much more about that now because one of the ways you make the whole system be more resilient is to have a better grid. We used something called energy frontier research centers. These are ways to bring people together. They might be at a university or they might be led by a national lab with university participants. They tend to bring people together working across a variety of disciplines to work on some specific problem, typically about $5 million a year. We also did some energy hubs. So one on advanced batteries for example, that these are tend to be a little bigger at about the $25 million a year range. And then we use something called cross cuts. So it turns out that if you want to make the grid better so that you can accommodate intermittent renewables and be resilient when there are hurricanes and it's cyber secure and all those things. Then there are lots of things to do. It might range from power electronics to replace these big expensive and slow to build transformers, active controls on where power goes. So those would be fundamental material science. But it's also the business of understanding the system and managing it and responding on time scales that are too short for humans. So it turns out the expertise for all of that is distributed across the agency. Some of it was in energy efficiency and renewable energy. Some of it is in the fundamental mathematics of all this in the Office of Science and some of it was in the Office of Electricity. So we said, well look, it makes sense for us to try to manage all of this as a single entity recognizing that you still have to work with Congress because those appropriations are handled separately. So there's a little sand in the gears that comes from that part of it. But nevertheless we and we were able to convince the Office of Management and Budget that this was a good way to think about it. There's some more areas listed here. Congress was harder to convince and determined to make sure that we spent the money in the appropriate slots. So it took some effort to manage all that but I still think it's a good way to think about how to work on the problems. ARPA-E, this was an attempt to say, OK, look, there is this period of early sort of, you know, the science of the advanced material for power electronics is now well enough in hand that we can start thinking about products. But there's still a lot of work that has to go into making something that you can actually sell and developing the ability to manufacture and all those kinds of things. And that next stage of the early stage commercialization is kind of where ARPA-E focused. We have, of course, 17 national labs, 13 of them were in the part that I looked after at DOE. Then there are three national security labs and one devoted to cleaning up after nuclear weapons manufacturing. And then finally there was the loan programs office. You know, I pointed out that the cost of many renewables had come down including utility-scale solar. The loan programs office helped finance, with loan guarantees, the first five utility-scale PV systems. After that, the financial community looked at them and go, oh, OK, we see how this works, we can do this. And now the loan programs office doesn't do that anymore. They don't need to. It just helped at that stage of things. So you say, well, OK, now I know you can't read all this probably. But this is a snapshot of where the portfolio was in 2016. Grid stuff, electric power generation, buildings, industry, transportation, all end uses. And fuels were one way to think about it. And then the cross-cuts were down here. The numbers indicate the budget amounts. And the colors up here tell you which part of DOE funded them. So again, we tried to think about the portfolio as a whole as a way to work on pieces that all are important in a well-functioning portfolio. And if you looked here at Stanford, not every box here would be occupied. But there's work going on here funded either by DOE or NSF or lots of other, perhaps even industry, that work on many of these kinds of systems as well. So we have a program that's pretty well across that portfolio. I'll give you a few examples. And I can see I'm low in time here. So I'll stop because I want to be able to answer questions. So on the grid, the old grid was a modest number of big generating resources, radio, all of us sitting out at the edge. The new one is going to be much more resilient in the sense that we'll have microgrids. So in places like Florida, you might be able to bring back on the microgrid around Miami, which I think actually does not yet exist. But in any case, you get the idea that to bring that on even before it can talk to the full grid, can deal with the intermentancy of the fact that the sun doesn't always shine or the wind doesn't always blow. And sometimes it blows too hard. But now you need to manage all of this and you need to do it in a way that is stable. So there are lots and lots of interesting challenges to be thought about here. And one thing that I think we don't do adequately in universities, and actually I think this is kind of a way it needs to be a field by itself, which is managing these very complex interlinked systems. So systems of systems. So the thing about the grid is one system, the transportation system is another, the pipeline system that delivers all those fuels, water. These are all interdependent and linked to each other. So we need to learn how to think about these in a better way. On the energy storage side, at one end there's the battery that goes into your Chevy Volt, or there is all these transportation options. Those tend to be small scale where weight and really does matter. For utility scale, it kind of doesn't matter what it weighs, but you want it to be cheap and last a long time, and be able to store a lot because if you want to. So there's kind of how much power you have to deliver and how long you have to do it, that product is the energy. And there are lots of options here. And that's the good part because there are plenty of ideas to compete and plenty of new battery chemistries that offer possibilities here that will then have to, the marketplace will have to sort out which ones of these work in a particular setting. And sometimes there are ideas that have to come together in an interesting way. So we use turbines everywhere to generate power in airplanes or to generate electric power. The ones for electric power are mostly steam turbines like this where you can see the size of a human. And these are big things. You heat up boil water and shove it through these turbines and as it expands the mass of water flowing, a water vapor flowing over the wings in the turbine drive it. If you do this with a fluid that's denser, for instance supercritical CO2, it's denser so the mass flow over the wing which determines the power generation doesn't have to, the wing doesn't have to be as big. So you can generate the same kind of power with a much smaller turbine here and only a single side instead of two sides. And you say, well great, why don't we do this? Well of course there are interesting materials, challenges and you have to figure out how to burn a fuel with say pure oxygen but recycle CO2 to keep from melting the turbine and keeping the flame alive. There's lots of engineering, lots of seals issues with CO2. So there's lots to do but if we did this, this could really revolutionize both the CO2 capture side and the size of power plants that are going forward which should reduce their cost. What about photovoltaics? Well there are really lots of opportunities here. As I said there have been clear improvements in both the materials and then all the costs that go around putting one of these together and generated and right now the 2020 goal for the SunShot program is six cents a kilowatt hour levelized costs. So toward the end of the time I was in Washington we decided we really had to update the goal because we're going to get there. So we're right at seven cents now and we're going to beat it before 2020. So you go, okay, let's figure out what to do. Well it turned out this actually was an interesting challenge because the Secretary of Energy was a guy named Ernie Moniz, MIT professor before he went back to DOE and he'd spent a lot of time thinking about all the pieces of levelized costs. So it took several meetings to convince him that we really had multiple pathways where we could get here and we're really kind of shooting at cutting those that cost in half again. That's a low enough cost if you look at the competing cost that you can think about another thermodynamic transformation. So it costs you an efficiency that's less than 100% every time you go from, say, sunlight to electricity and then if you then make, I don't know, hydrogen or some other fuel after that, you know, there's an efficiency cascade down there. So getting the cost down really offers lots more opportunities for diversifying other parts. Building efficiency, gosh, we've shown here at Stanford that there's a huge opportunity to do better there and three quarters of the electricity and about 40% of all the energy we use in this country ends up in a building somewhere. So this stuff actually matters. And I imagine you're hearing more about some of the kinds of things that these buildings here, these newer buildings, really bring to the party. So on and on, I'm going to skip through here the last few slides. Transportation side, there's lots of opportunities and we can power them with fuel cells or with batteries. Advanced manufacturing offers some new ways to assemble things that require less in the way of materials. I think this example, so this is a bracket that goes somewhere in an airplane. This is, if you just do it in the traditional way of machining it, this weighs a kilogram and it takes eight times as much material as you get when you're done. And if you do this with the additive manufacturing of making a powder and assembling it into the finished part, you can reduce the weight by 60% and the conventional bracket uses three times as much energy to fly it around in the air. So now one thing you'd like to know is whether this part is as good as the other one. And so that's, I'm going to skip forward just one more. Well, I'll say that in a minute. Now in the fundamental science side, in some places we'd like to capture CO2 and put it somewhere else beside the atmosphere. There's a whole host of interesting compounds being investigated, but this is still very much at the basic science side. Can we do this at all? We'll worry about the cost side, but there are really exciting opportunities there. And then we have the basics of really understanding how all the physical processes that work in catalysis or materials or any of the kinds of things we use for energy. I mentioned the, so if you look here, this is a little turbine blade. That was made by advanced manufacturing, this powder assembly idea. And that picture was taken at the Spalation Neutron Source at Oak Ridge National Lab where they were able to image the residual stresses still in that blade. Now those of us who fly around in airplanes would like that blade to hold together. That's a useful property of turbine blades. And using all the fundamental scientific tools for characterizing materials, you can actually show that this is an adequate way to do that. And of course there are many other science applications for these kinds of facilities as well. So the picture I hope I've, well, and then I just need to say, underlying all of this is advanced computing. DOE is now working on the next big quantum leap in computing speed and architecture. It's our ability to design turbines to reduce the drag on trucks. This little exercise deployed immediately into the truck business because it reduced drag very quickly. And then what that just says is that there's a huge opportunity space for clean energy innovation. And both for individual technologies and for systems, we should work across a portfolio. It needs to be fully stocked across primary energy resources and conversion technologies and systems and time scales with an emphasis on efficiency everywhere that the enabling science is absolutely critical to our ability to do this in a big way of going forward. And grad students in Stanford have a very important role to play. And so you guys actually get to do all this cool stuff. So have at it. Thank you very much. Great. We've got time for a couple of questions. Come back over here. So as a scientist, you just kind of came to talk about economics. A lot of us are going to spend a lot of time to laugh over the next couple of years. One question for you is what are the best ways to get to know the other side of things that isn't just in a classroom setting? Yeah, well, so one important thing you could do is enroll for Energy at Stanford because all around you are people who are thinking about this from the economic side as well. And there's actually an improved both the biz school and the engineering economic systems in the School of Engineering. You think about work on all this kind of stuff. So getting to know your fellow students, I mean, one of the things that we hope for in having a group like this one together is that you'll meet some people and your paths will cross in ways that nobody ever foresees as well. It's an opportunity for conversation and some of you are going to be involved in startup companies that will have to think about this. What are we going to sell early enough to survive through the period of getting going? And so nobody knows everything. So you have to identify talented people around them and work with them to put their assorted skills together. And I will say that it's a good thing to be polite because you never know when your paths are going to cross. So it turned out that the director of the National Science Foundation, Frans Cordoba, and we were classmates in college. And who would have guessed we'd find ourselves in charge of $18 billion a year of federal research money. Neither one of us would have guessed that. And I can tell you our friends would have been even more astonished. But there you are. Speaking on the budget, when you had that slide up, I noticed that there was a lot of money being spent on the computer scale. And the order of hundreds of millions would seem like a lot relative to, essentially, every other department. So I guess my question for you is, to what extent are people within the DOE making decisions on what money is getting spent there versus political pressures to the bottom line? You ask a very good question. So how is the budget put together? And it's a complicated process. What we did inside the agency was to try to figure out how to balance the portfolio. One thing about nuclear stuff is that it's expensive. And it's mostly done because it's so expensive, it's mostly not done in... The big facilities are not done in industry. So the DOE provides some facilities and you have to support those and make them be secure. But in working through trying to balance all of that, then we produced a draft of a budget. It goes off to the White House, to the Office of Management and Budget, and they fiddled around with it. Mostly what they did was to take a billion dollars out of fossil and nuclear and put it in energy efficiency and renewable energy. Then it went to Congress where they undid that and moved things around some more. Now, we tried very hard to think about not only what does it make sense from a technical standpoint, but also what can we get through Congress? We know that there are people there who are going to try to protect individual programs and so you try to do the balancing and make an argument based on that. Every president's budget has a certain element of fiction to it. This current one might have more than any more of that. I hope it has more of that because it was a disastrous recommendation. But it is a complicated process and the Congress sets the budget, not the administration. The government was trying to change. Did the DOE have to change the way it operated when it came to proposals and making budgets and all? So did the DOE have to change the way it operated in the Office of Management given the current steps of the government? You know, it's a little hard to tell. The DOE, the top leadership of DOE is still pretty sparse. And it's my impression that most of that budget, the budget recommendation from the White House actually came from the Office of Management and Budget, not so much from the agency. But I'm obviously not there, so I'm looking in from the outside. And I think from listening to the hearings on all this that the congressional part of this is going to be different. The mid-year budget extension didn't buy the recommendations from the White House, so I think that will be the case. But the truth is, I don't know for sure how they assembled the budget. Two more? Here and here. How concerned is the DOE about the cybersecurity So that is a very good question. The question is about cybersecurity, and the answer is very. The place where it intersected most with what we do was the combination of the grid and cybersecurity. There is a big internal effort on this and a lot of coordination with industry because mostly the grid in the United States is not government-owned. It's mostly owned by some combination of utilities and regional operators. So I can tell you that there's a lot of work going on and they really, the essential control systems are not internet-based. They have their own separate ones, which doesn't make them perfect, but makes it harder. And then the other part of the cybersecurity side is the nuclear weapons and all of that, and there's a big effort on that side as well. Government's not as good at protecting personal information. I just got the fourth notice that my information had been shared with the Russians and the Chinese and whatever. So there you are. Some parts of it were not so good at. So last question. Thank you. Outside of this room, you've spoken about the carbon tax and how you feel it should be revenue neutral. You could just elaborate on that. Yeah, so a carbon tax is, you know, I think fundamentally we have to have some way of recognizing that the energy technologies, including the fossil ones, have some kinds of impacts. And so a carbon tax, most economists agree that a carbon price set by a tax is the simplest one to collect and has fewest ways to game the system, although you can do the same thing with cap and trade and so on, so there's not any single one way to do this. The revenue neutral part is just a way to make this not be a big drag on the economy and to help build the support that it takes to get this done. So if all of us, those of us who do well in saving energy would get a little money back and those that want to use a lot more would pay more and not get money back overall. So it's just a way to try to make the system work. Great. Please join me in thanking them once again.