 This is a little bit of an experimental talk. Over the last six months, I've read about 60,000 pages of footnotes. Did you laugh? But it's pretty close to true. It's not even an academic. Not even an academic. This is starting in roughly... You'll see 1965 through till roughly 1985. That's where all of the interesting... I wanted to understand what we understand about energy. So this is a deeply nerdy talk about energy data and what we actually know. I think hopefully I can surprise everyone here, and you'll all realize that a lot of what we think we know about energy is wrong for very strange reasons that are hidden in footnotes in the 1970s. Of course, what I do for living mostly is not doing energy data, but is starting energy companies. So I started McCartney Power, a wind energy company that's now running under Google X. I have a solar tracking company, Sunfolding, that's doing very well and shipping megawatts. We have an energy storage company that is licensing pressure vessels for hydrogen and CNG into the automotive industry and then a couple of other interesting projects. So if you want to distract me at any moment during this boring talk on energy data, or maybe it's not boring, feel free to distract everyone with questions about other things you might want to know. And we can go there. But let's start with a deep dive into energy data. The Department of Energy actually just funded us for six months to do this data energy review and being the WAGs that we are at other lab, we actually registered this domain name. So our version of the EIA's energy data can now be found at the Department of Energy. Keith Pascoe should get credit for all the lovely graphics that you'll see. So let's go down the rabbit hole. 1965, I think this is great, this is under Lyndon B. Johnson. He commissioned a study restoring the quality of our environment. And what's the signature to that? What is really interesting to me, this is the first mention in a government document of atmospheric carbon dioxide. Some people are nodding that they might recognize this one. So it was one full chapter and pretty much describes what is still true today that carbon dioxide is a problem. So we knew this in 1965 under LBJ. 1971, of course, we're starting to get close to oil crises. In fact, the first one is starting to happen, which comes to the attention of Richard Nixon. Energy was important, and the energy policy was important at this period in American history, including our growing awareness of the environmental consequences of energy production. This is from another president who's being explored for challenges. We don't think of Nixon as great, but in fact Nixon did make all these statements and is the origin of the EPA, the origin of the Department of Energy. It was all signed into being under Carter, but Nixon kicked it all off. He was concerned with a sufficient supply of, and quite importantly here, clean energy, although at that moment that was more about smog than CO2, is essential if we are to sustain healthy economic growth. 1973, of course, we started to see the real energy and oil crisis hit America. The price is amazing. That was probably still a full tank. Now, the thing that existed prior to the Department of Energy was called the Joint Committee on Atomic Energy, and they were tasked with analyzing the national energy dilemma. I love this. They were tasked with developing an energy display system which in less than an hour could give an extremely busy person an understanding of the size and complexity of the national energy dilemma. An extremely busy person here meant a congressman or a senator. Interestingly, Al Gore's dad was on the Joint Committee of Atomic Energy when this was bought into being. This was the title of the publication that came out in 1974, things moving pretty quickly, and they leaned on something called the Sanky Diagrams. If you know what a Sanky Diagram is, raise your hand. If you don't, you're in for torture or a treat. This is often considered the first Sanky Diagram. Sanky Diagrams are useful for looking at energy problems because they are conservative, so it just looks at the flow of energy from one end of a process through to the other, and you can divide it into waste and used energy, hence the conservatism. I love the name of this guy. Captain Matthew Henry Phineas Ryle Sanky, when I proposed to my wife that this should be my son's name, she said no. She also turned down Balthazar. Who knows, my son's name is Huxley, for those who are interested. This is actually the first known Sanky Diagram. This was a French academic called Meinard who charted Napoleon's sort of excursion into Russia. The thickness of this is proportional to the number of Napoleon's troops that he has left. This is also geographic, so it's roughly going up the river. I can't remember the name of the river that Napoleon went up to into Russia or anyone here. You're Stanford, you should know this. Wow, maybe at Harvard. You can see that he's here in Moscow, and even on the return journey, to get a sense of numbers, he gets down to about 10,000 troops remaining when he started with 482,000 troops. Anyway, so that's the very first Sanky flow diagram measured in human lives, which is pretty dark. So this is what the Joint Committee on Atomic Energy actually presented. This was the origins of what you may all know as the Lawrence Livermore Sanky Diagram. This was the first one. And in 1950, I think this is hydroelectricity, natural gas, coal, oil imports, and oil domestic. And they, for no particularly good reason, divided things. They treated electricity generation uniquely, and we're going to come back to that later. And I think that's a problem now that we're trying to rethink our energy system. And then as residential, commercial, industrial, and transportation, they have a thing here called non-energy. We will also come back to that. And I think that was something they did right in this. They then showed that this was 1950, this was 1960, this was 1970, this was 1980. So they actually started to project into the future. I think this was 2000. This was right after the first statement, nuclear energy is so cheap it'll be too cheap to meter. So we had a very optimistic perspective on how energy was going to scale in America. So they then encouraged you to take those last four images. This was this beautiful fold-out book. So you would pull out the images, and then they had this other page with scores in it. And you would build your own three-dimensional model of the future as a congressman. You would put it together. And in less than one hour, you would now understand that we didn't know how to get from 1950 to 1990. Roughly what happened in the next two years is we started something called the Department of Energy. The Energy Information Administration was meant to be independent so that it would collect all of the data. That's where we get all of the data from this from. This is the very first assembling of the EIA's data by Lawrence Livermore into this diagram that I think is many people's first introduction to broad-scale energy flows. This is the 3D model. There you go, clear as mud. I can give you one hour with that and you'll really get it. Just to underline how I think this is super interesting to me. So this is 1977 and it is profoundly interesting to me just how pervasive energy and public energy education was at that time. So much so that our favorite Flintstones actually got their own energy special. I have a full copy of this. It's now been taken down off the internet. So actually by 1977 all of the perpetual motion machines that Silicon Valley now invests in were already invented, which is one of the upsides there. But actually it is worth seeing the whole thing, particularly like some of the magnets and the frogs chasing the mosquitoes. So anyway, this is what those flow diagrams look like today. We still use British thermal units, quadrillions of them, which is 10 to the 15 BTUs. This is the EIA's version of that flow diagram. This is as it goes out through Lawrence Livermore where they actually break it down into end-use sectors. So is that everything we can know? This is the annual summary that we get. It comes from data in the annual energy review. It turns out we can know a lot more. So these fuzzy blobs represent all of the data that is collected by the EIA. So the pink one is the manufacturing energy consumption survey. The yellow one is the commercial building energy consumption survey. Orange is residential energy consumption survey. Green is national household transit survey. This down here is the federal energy management program that requires every federal government agency to report their energy use. Pink is an unusual one. This is the global trade project out of Purdue. CS54 is the weekly coal survey. EIA-MER is the monthly energy review. And the EIA-914 is the natural gas and oil industry monthly survey that you have to fill out if you operate one of those plants. So all of that data exists. Let's have a look here. I just put this here for the students who wish to read this and look at all the footnotes later, memorize those URLs. So if we take all of the data points that are in all of those surveys and union them, you now get a flow diagram that looks like this. This is obviously overwhelming as you're sitting there and looking at it. At our website you can go in and zoom in on any particular thing. This will go down to, and I can't believe I know these, one quarter of one quad, which is about one quad. If you think of quads as percents, that's the easiest way because just by chance America uses about a hundred quads. Anyway, one quarter of one percent of US energy flow is in the animal slaughtering and processing industry, somewhere hidden in here in the manufacturing sector. Anyway, we attempted and succeeded in going to about 0.1 percent resolution. That's not accuracy, that's resolution. I could put error bars on any one of these particular data sets if people are interested. But we get to about 0.1 percent resolution. So here's transportation. They break it out by various types of transportation, which you can see here on the left, whether it's freight rail or what type of trucks, what type of vehicles. Obviously the majority is in light trucks. That's F-150s and cars. So this is now SUVs and cars. You can break down what is in sort of rail or shipping-based transportation of food into all of these things like agricultural products. A quarter of a quad is moving wood products around the country. International shipping out of America is 0.6 close to a quad. And then you can break down because of the national household transport survey. You can even determine that 0.66 quads, for some reason they put school and church in the same bucket. 0.66 quads of US energy flow is in people going to school or church. I was at a lecture with the guy who used to run autonomous vehicles for a Tesla. And he said, no, no, no, that data can't be right. And I said, why? Tesla's data doesn't look anything like that. And then you have to sort of remind him that maybe the rest of America isn't the same purchasing demographic as his Tesla drivers. Again, for completeness, National Household Travel Survey and Commodity Flow Survey, just as a footnote, we collected these things the way you used to collect data in the 70s still today, meaning you call people on the telephone and you survey them of where they drove that day. And so I think it's about 150,000 data points in the National Household Transport Survey. I think that survey is plus or minus 25% accurate, maybe worse. Obviously in the future we could in fact use Tesla and Uber and all of their data and in fact nearly any new car ships with GPS. So you can imagine we could make this data much higher resolution, which I think would be great. So that's just a sort of note on where this is. Residential commercial buildings are also interesting. The types of categories we break down residential energy consumption into is single family detached homes, that's the big one, that's the majority. Single family attached apartments and mobile homes. So a half a quad is in mobile homes. So you should take them more significantly as a potential opportunity to improve America's efficiency than I think people do. And then under commercial business they're broken down into things like education so close to a quad of power is keeping the lights on in this establishment as well as all the schools in the nation. 0.6 quads in lodgings, that's hotels, motels. Food sales, so just retail food sales is a quarter of a percent of US energy flow. Food service, actually I'm not sure if I know the difference. I've forgotten that footnote between service and sales for food is another half a quad. Inpatient healthcare hospitals is half a quad and you can see more of these. Religious and worship is 0.17 quads. Again these are done as surveys of building owners. So I think they probably also are quarter, you know, plus or minus 25 to 50 percent on accuracy. What we'd really like to do is disaggregation of metadata so that you can understand what is flat screen TVs, what is refrigerators. That's not how it's done today. People check boxes whether they own a refrigerator or not and then the government applies a gross estimate. So it would be nice to be able to improve that in the future although I still think we can know an awful lot today. US government, so Department of Transportation, Postal Service, etc. Nearly all of them, you have to remember for these and also if anyone has questions in line today you should just ask them in line, I'll try and answer them. This is really just the electricity bill, the PG&E bill, so your electricity and gas for the government and their gasoline bills. So Department of Defense is the giant one, 0.73 quads. About a half a quad is just jet fuel for the Air Force. This does not include the manufacturing of all the military equipment or the housing of all of the service people. That is elsewhere, considered elsewhere in the economy. Manufacturing industry, I find this one one of the most fascinating. They go into really quite remarkable detail. So light stock, 0.3 quads, construction of buildings, a quarter of a quad. You can see all of these things. We'll come back and have a look in greater detail in a moment. Sugar manufacturing alone, you wonder why we're all overweight, is 0.1%. Ethyl alcohol, 0.3. Manufacturing energy consumption survey is where we get that from. Again, for some reason this is one of the most frustrating data sets. Certain industries have told the government that we don't want you looking at our data because it's a trade secret. So you can't disaggregate the steel industry or the chemical industry which are three of the biggest in any useful ways because they don't want it's considered a trade secret who's making the most nitrogen based products, for example. I used to think it was because we didn't want to know who was making munitions but it's actually because they don't want to know who's making fertilizer. Is that everything? So interestingly, when you go on like a data dive like this, it takes you six months often to find the things that aren't there that you're trying to find everything but then you find the things that aren't even measured. So street lighting, parking lots, airports, traffic signals and billboards are not residential buildings, they're not commercial buildings, they're not industrial, they're not transportation. So there's actually a few things that fall outside of all of our surveys. Street lighting is significant enough to actually make our 0.1%, so is parking lots are pretty close, which I find interesting. Also interesting that doesn't get measured in the data is wastewater treatment and water treatments as clean water and dirty water and then municipal lighting, which is everything else, which is 0.2. So those things aren't measured but you can, well they're not in any of the traditional surveys but if you're trying for a comprehensive view you need to add those. Other things we can know that aren't obvious to people looking at the gross data is embodied energy of net import. So this is a study done at Purdue under the Global Trade Analysis Project. 11 quads of energy, this is not considered in the 100 quads that US users are imported in the embodied energy of American products. This is the countries where they come from. So Russia, then you raise your eyebrows, why? China, then you might ask who is Asia without China and why Africa? Roughly Western Europe has no natural gas, but Western Europe wants to make clean products. So Western Europe uses Russian natural gas to make your oudies and your box wagons and your IKEA furniture and so that's how America is importing tons and tons and tons, in fact three quads of Russian energy. China is mostly coal making all the things that are made in China. Asia without China is Australian coal, burned in China to make your products and Africa is South African coal burned in China to make your products. A little bit of Nigerian oil, but grossly that is South Africa and Australia are supplying China with energy and Russia supplying Western Europe and we import 10 quads. So what's interesting, in 1974 Nixon launched Project Independence which is how do you make America energy independent. The concern was that 10% of US energy flow was coming from foreign oil. We now have 10% of energy flow coming from foreign goods and I think in a little sense this is, you know, this explains the Trump administration's frustration with trade, right? This is an enormous expense to the US economy because this is a proxy for trade imbalance. Here's a flow diagram from the world, WCI, somebody here will remember that, World Something Institute. Sorry, they took all of the data from the Purdue data and you can see how the energy flows here and becomes a net flow into America for the embodied energy. Other interesting things you can know that is sort of more than you can know, you believed you can know is we know by state all of the coal production in the US. We can also know the natural gas production by state and we can also know the petroleum production by state down below and the petroleum net imports up above. For those who are interested, I always thought this is counter-intuitive. The top four oil origins of oil for the US, Canada, Saudi Arabia, Venezuela, and Mexico in that order. So all these oil wars we are fighting are really not for much oil at all. What is interesting is they took all of the domestic production data for all three fossil fuels and in case you're interested, 87% of oil is produced in red states, 89% of natural gas is produced in red states and 88% of coal. So that helped me calm myself a little bit about the politics of America. It actually really helps you understand. You look at particularly states like Wyoming and North Dakota, that becomes nearly the entire economy of those states. You want to understand why they really fight the fight to keep burning these things. It is their economy. You could flip that on its head and make it a good news. The reason those states have all the energy is because they're the giant states. So if you map solar and wind resources, they also win that on giant margins. So we could have that positive story, but instead we're stuck in negative politics trying to protect the incomes that we have. Anyway, what do you do? That looks terrible. Sorry. We'll come back to that. This is what happens when you put all of that data into one thing. This was just a funny thing where we tried to take all of the quote-unquote wastes that the EIA divides it into. So red accumulates waste through the economy. Blue is useful energy services. I actually think those definitions by the EIA and by Lawrence Livermore are really not very useful in the world that we want to inhabit or design for. And we're going to come to that. That's going to be one of the main points of the talk here. Okay, sorry. Something went wrong. Okay, imagine that this is electricity. This is hydroelectricity. The question came to me, what is primary energy for hydroelectricity? Is it the surface area of the dam, the modulo, the depth? Is it the rainfall in the catchment area? How do you, for coal or oil or natural gas, primary energy is easy? It's tons of coal, it's millions of cubic feet of natural gas, and it's barrels of oil. What is primary energy for hydroelectricity? And this concerned me long enough that I found the very first instance, I think it was in about 1972, where they had to address the question of what is primary energy for hydroelectricity. And they decided that the primary energy for hydroelectricity will be the amount of fossil fuel plant that you need to bring online if there is a drought. So what you do is you take the efficiency of all, the average fleet efficiency of all of the fossil plants, which is about, still the day, about 38%, and you reverse engineer the primary energy from hydro from the delivered electricity from all the hydro plants. What does that mean? Is that we invent one and a half quads of energy that never existed as part of our 100 quads that we use. Let's call that a ghost quad. It was never even produced. That becomes important because we also treated nuclear like that. So, you know, pub quiz for the undergraduates here. What is primary energy in nuclear? Is it equals MC squared? No. Is it energy density of uranium? No. Is it uranium density modulo? How well you can extract energy in a light? Water reactor? No. It really is what's called the heat rate, which is a proxy for how efficient is the steam plant post the creation of heat by the nuclear reactions. That's about 37, 38% also. What does this mean? You take the amount of electricity delivered by nuclear, and then you penalize it by the inefficiency of coal and natural gas, and then you come up with an amount. We actually deliver 2.8 little more quads if you want to use those units of nuclear electricity to the grid, but we report 8.7 quads of primary nuclear energy. We treat wind and tidal the same way we produce wind and solar. We treat the same way we treat hydroelectricity, and we treat geothermal even worse. The interesting thing about that is there's eight or nine quads that never existed in our 100 quads. Everyone's celebrate, breathe a sigh of relief, we're all working on renewal energy. We just made the problem 10% easier by definition. Other weird things when you go into the details of this data, why do we treat electricity different to gasoline? Electricity is this weird interim between primary energy and the various sectors. You can think of it as produced and useful energy, but interestingly for gasoline we don't have an interim which is, we have oil over here and we use gasoline here or diesel, but we don't have an interim. If you go and have a look in the details, petroleum refineries use an amazing six quads to turn oil into gasoline. That's considered a 100% useful use of energy. So that energy in the Lawrence Livermore flow diagram goes to energy services, not to waste energy. You really should think of that as well as probably oil and natural gas extraction close to two quads and pipeline fuel for natural gas. So this is the energy to pump natural gas through the 4.4 million miles of pipeline we have. You should think of them sort of as wastes in that. What does that really mean? Your 20-mile-per-gallon car is about 15 miles per gallon. And I think we should treat all of those similarly that we do to electricity. So I'm going to wrap it all up in the end. But now we've got all of this data. We can do some fun things with the power of the internet and search. We can now tag all of the 500 data points that the EA gives you. And we can even put operands on those tags. So what I've done here is I've taken any single data point that is related to agriculture or food. Sorry, not the big agriculture. I mean what winds up as food and collate all of those for me. And then if we zoom in on the detail of that, you get a pretty interesting perspective on what is food. So about 13 quads of all our energy goes to food. People traditionally think food is agriculture which is fertilizers and pesticides there. That's about a quad. There's online farm equipment diesel. It's about three-quarters of a quad. But actually bigger than those things. So people driving to work in the food industry if you use the survey data from National Household Travel Survey. If you work in the food industry, that commuting is equivalent to the amount of diesel we use on farm. And there's an equivalent amount roughly of energy of us going to the supermarket to buy food. Similarly, things that are hidden is the enormous electricity generation losses required to make the electricity for refrigeration at the top. Refrigeration is about a quad. About a quad in cooking as well. So we just cross-referenced this to a recent study by someone in academia and we got the same number, about 12 quads. I think ours is a little more accurate. But this is what happens if you do it the old-fashioned way. All right, another interesting question you could ask. Is changing consumer behavior enough? I'm very frustrated with the, if we all buy a Tesla, use stainless steel water bottles and get LEDs that the problem solved. This is pretty untrue. Can we figure out how untrue this is? So I tried to, instead of having the abstract buckets of manufacturing and residential and commercial, why don't we try to express all of the flows in terms of things we understand as consumers? You can see I've rearranged all those flows up there. What do they look like? Interestingly, your only direct relationship where you pay a dollar basically for a unit of energy is through your purchasing of electricity and natural gas in your home. That's about 18% of the country's energy flows. About 18% is you or we as consumers directly purchasing gasoline for cars and SUVs. About 1% is us purchasing gas for flying somewhere or there's a little bit of energy we would purchase to run our motorboats and ATVs. About 37% of all US energy flow, you as a consumer have a pretty one-to-one relationship with your dollar purchasing electricity. This idea that we're going to solve the problem by making the right purchasing decisions to me is not very true because the other 63% is embodied in the infrastructure of our lives and sort of the way society has been built. So 4% of our energy is in keeping retail open to purchase our stuff. About 12% is for retail for services. That means education in hospitals and massages, I guess. 6% is the production and distribution of energy. That was what I mentioned before in terms of oil refineries, etc. About 20% of our energy is in the production of all of our stuff. 10% is in the production of all of our imported stuff. 10% is the transportation of all of our stuff, the movement of our stuff. About water, I don't see any energy used to transport water. It's huge. It's not huge. It's small. I showed you a little bit of the treatment of it. In California, they say 20% of all our electricity is for water. That's an enormously generous definition of water. I think in that study, they even consider freezing your ice cubes as water energy. I think it's closer to 1%, but I can say that it's very difficult to pin down. And you should be aware of the headlines. Other environmental headlines you should be aware of. It's much, much better to eat salad than it is to eat meat. If you look at those studies, the denominator they use is kilograms of food, not calories. It's kilograms of lettuce compared to kilograms of beef. I don't know if you've eaten the lettuce leaf. Beef is diesel. Beef is diesel. It's like 40 megajoules a kilogram. Lettuce is a few megajoules a kilogram. The study still shows that being a vegetarian is better, but it's about five times less better than the headlines have tried to teach us. Anyway, so I think this gives you a different perspective on how do you solve the problem. Yes, we need to make good purchasing decisions, but what we really need to do is consider how we have built our society and how do you embody good energy decisions in our everyday lives. I mean, we don't want to have to be confronted at every supermarket experience with the moral dilemma of which can of salmon to buy. You would like to have had rules and society build an infrastructure that makes it easy to make the right decisions in the first place. So, quickly, dollars to quads. If you're a Stanford, so you're all more interested in money than science. Okay, that was low blow. I was actually joking with my friend on the way down here. I think Stanford on a 20-year timeframe is going to kill us because somehow your students earn more money and they like the school better so they donate more back. At MIT, we make the wrong life decisions and we hold resentments for unknown reasons. So on a 20-year timeframe, Stanford's going to kill it. Anyway, so you can basically do a conversion of these quads whether you're talking wholesale crude oil at $30 a barrel or retail gasoline. You can roughly get these conversion factors for energy prices. I actually think the enormously exciting thing that's happening in our historical moment right now is wholesale retail energy price arbitrage. This is where all the giant opportunities are. If you put your electricity, so one quad of renewables today is about $16 billion a quad. That's a lot cheaper than gasoline if you can get it to the electric cars. In fact, 2017 was the first year that the AAA showed the total cost of ownership for electric vehicles as the lowest in class. So I think this is good news and I think it's there. We can come back to this point at the end if you're interested. I think it's really interesting. Okay, here's a moment of optimism. Are you ready? You're usually probably going to have depressing lectures here. So I challenged myself to finish with some real optimism and some snarky commentary about the misunderstanding of the word efficiency, not just in academia but in government and the public and private sectors. All right, memorize this picture. Seriously, what we're doing here is, I don't know if you know, in the first Senki diagram I showed you, the blanket assumption for transportation is that all of those cars run at 20% efficiency. And that's how that waste stream is generated in the LLML transportation graph. Let's debate how true the 20% efficiency is for a car from gasoline to motion. There's similar weird, gross... If you think the inaccuracies in the data is unreasonable, it's way more unreasonable in what we define as energy services and what we define as waste. So this is Saul Griffith's unabashedly political Let's Save the World remapping of the US Senki diagram where instead of using the word efficiency, we're just going to say, let's not do anything stupid. Let's just do smart things. So now I'm going to do something really risky, which is use... Absolutely. Actually, I don't even care about humanity. I thought deeply about this recently for reasons. Of course. Wait a second. Here we go. I think you have to... If you're really going to talk about energy and climate change and why we're doing what we're doing, I think you really need to look at yourself, why you're doing what you're doing. I grew up in Australia. My mother was an artist, and she painted an awful lot of Australia. Does anyone know Sir Joseph Banks? He was on Captain Cook's Voyages of Discovery and sort of documented all of the natural world through the Pacific. My mother is like a modern Sir Joseph Banks without the Sir and other things. So every holiday I had for my childhood was going to some pristine, amazing place and looking at unbelievable flora and fauna. And I love the natural world. So I'm maybe a John Muir-esque environmentalist. I love the natural world because it is beautiful. I think you have to have... I think my argument for solving climate change, saving humanity is just an aesthetic argument. I want to live in a world that's more beautiful. Yes, we could carve off half of humanity and send them to Mars, but that's going to be a miserable existence. Excuse my French. Those people who wish to die on Mars, go for it. But that's not for me. I really like sea turtles. I really like the Wandering Albatross and I've sat on cliffs and watched them. And so my biggest concern about climate change is that we're going to make the world just a lesser place for us all to be. And less in a true aesthetic sense, less beautiful, less interesting. And that's why I would like to solve climate change. All right. Anyway, so let's have a look at this. That was my sub-story over. The first thing I do is I remove all of the false quads. So as I go through the diagram, I just get rid of the stupid shit. So these are the quads that don't even exist. Then the next thing is I'm just going to assume that those three big red bars, they are the thermoelectric losses from coal, oil, and natural gas. So let's just, oh no, two of them are, no, the top two are thermoelectric losses from coal and natural gas. The third one down there is all of the energy cost of producing fossil fuels. So let's assume we're going to do completely solar, wind, nuclear, future. There'll be some other bit players, some biofuels. But let's just treat them all, treat it as though we're going to go with those three. That's good enough for now. So we eliminate basically the dumb burning of fossil fuels, the dumb inefficiency of using gasoline to run a car that should be electric, and then all of the production and distribution of natural gas and oil. You get rid of a huge chunk. Now I have to start remembering stuff. So we'll go over here. The other assumptions here are, I annotated it, here we go, this is good. Now I don't have to remember. Here's the other stupid things. Okay, electrify all cars. I just used the assumption that the electric car at 400 watt-hours per mile is about one-third the energy cost of gasoline and diesel. That was a good enough assumption. I cross-checked against just taking all the vehicle miles traveled. So a Tesla, when you're driving it like me, it gets about 400 watt-hours per mile. If you drive it like my mother, about 300 watt-hours per mile. So even if you use 400 watt-hours per mile and you multiply it by all vehicle miles, you end up with about this number. So electrification of all cars gives you about a close to five quad discount. Same five quads if we get electrify F-150s and light trucks. Freight truck electrification would give you close to three quads. Obviously this relies on an awful lot of good batteries, but I believe everyone here is working on it. Or some of you, hopefully. Fossil mining equipment. So there's a quarter of a quad of dozers mining your coal. So that goes away. There's a quarter of a quad in... coal is half of freight rail in the US. So it's a quarter of a quad in moving coal from mine to where we burn it. Three and a half quads in petroleum refineries goes away because we don't need them anymore. We don't need oil and gas extraction, which is 1.8 close to two. These four and a half quads we count is energy materials that goes into products. This is the coal, the natural gas, the oil that becomes plastic, becomes industrial products. In some respects this is sequestering energy, so we'll remove that from the energy picture because it's not producing carbon, they become products. And then the last two are maybe the most unreasonable. We go to high efficiency or high coefficient of performance heat pumps for oil refrigeration. We're using a COP of four to five here, and we apply that to all HVAC and all refrigeration. We'll save five and a half quads there, and then we do all lighting by LED lighting. And we'll save another 1.3 quads over the one quad we've already saved with LED lighting. Just anecdotally, this is great, the two biggest improvements in our flow diagrams in the last 20 years have been LEDs and... Cold water enzymes for washing your clothes. We saved about a quad going from having to do hot water cycles for all clothes washing to cold water enzymes that we evolved using biotechnology. I think that's fabulous. If none of you here are working on the following problem and I wish you all take it on tomorrow, roughly four to five quads is in turning trees into wood, paper, cardboard, and newsprint, about a quad each in paper, newsprint, and cardboard. Most of that energy is separating the cellulose from the lignin. That rhymes with a problem like making a cold water enzyme so that we don't have to boil the lignin off, get to work. And then we get another two or three quads off here. That's great. It's not deployed. Anyway, I think it's important. I'm glad that you know about it. It's not deployed yet, but neither are electric cars at this scale. So what happens if you do that? What does the economy look like? Energy services now. There's still some thermodynamic waste, but this is kind of Carnot unavoidable waste. And energy services, you add these up. But you need about 40 quads to run the U.S. economy, not the hundred. So I just made the problem about two and a half times easier. Not by doing anything efficient. You're still driving the same size car. You're still driving. Americans don't have to compromise their lifestyle. I see you all shaking your head. I mean, obviously I would love Americans to compromise their lifestyle being Australian. But although you are using my coal to make your plastic toys, but it's about 40 quads. Now, I think we would be having a much better energy policy debate in this country if we were more realistic about with visions of what the future could be and we were more cautious about the differences between efficiency and efficiency. So starting roughly with Carter, people were confusing wearing sweaters with the thermodynamic efficiency of your steam plants because both were described by the word efficiency. There were attempts to get away from efficiency as a difficult word. For example, everyone who remembers negawats. So negawats was how we were going to negatively get watts from somehow. But I'm not sure that either of those are good. I still don't have a better word than either. But if we just, I call it, just don't do the dumb things, it looks like the problem is a lot easier to solve. This impacts, if you do a very coarse, so I now know other weird things, there's 8.8 million lane miles of American roadways. Each lane is considered 10 feet wide. So you can multiply that out. I can't remember the number. And it gives you the square meters of road. If you do high density solar on there and get 30 watts per square meter of all of that land area, you need three sets of American roads to make 100 quads. So everyone's like, imagine this 200 mile by 200 mile square in the desert of Arizona. That's not how we're going to solve the problem. What does it really look like to solve this problem? It probably looks more like having, well, I think it's hard to imagine three solar roads next to every road and car park you've ever driven on. But we just made the problem 2.5 times easier, so maybe it's one roadway. In fact, if you get to the 40 quads, it looks like 100% penetration on all domestic roofs and all car parks. And then you just with solar. Obviously, if we add nuclear and wind and others to the mix, but it starts to feel more manageable than a giant new piece of infrastructure that we have to build to cover surface area. We've already built enough things pointing up. So we probably should cover them first. I've probably spoken enough. I don't know if I have any more slides. I can go back and look. Okay, three questions and a quick story. This is the other good news. Does anyone... I didn't hear enough. That's great news. So I'm going to give you one more set of great news. I want you all to ask yourselves these questions. When do you think the auto loan was invented? When do you think the home loan was invented and when did we move to a five-day work week from a six-day work week? Think about in what decade when. I'll give you 15 seconds. Think about that. So the auto loan was invented roughly in the 20s. Ford had sold so many cars. He'd run out of rich people who could pay cash. He offered people a lay-by so you could go and give him cash but not get a car. This wasn't very satisfying to people so GM invented the auto loan. This was only about 20 years after usury... usury which is the sort of... you shouldn't get interest from learning people money. This is nearly all religions have this as a philosophy. That was still sort of the aversion to credit persisted all the way through to the beginning of the 20th century. And we had pawn shops just after the industrial revolution and they were the first things that we sort of accepted having credit and then we used the financial terms of pawn shops to model the auto loan on and that's what made GM big and took market share for Ford. The home loan was invented in the 30s, 40s really took off late 40s and 50s with a return from World War II. It was modeled on the credit system that we invented for auto loans. When do we move to a five day work week? So there was a textile mill in upstate New York that had roughly 50% Christian and Jewish employees. They fought about whether to have Sabbath or Sunday off. The factory owners said monkey. You can have both days off. This was in the 50s and by the 60s because that was a unionized shop that had sort of like a cancer spread across the world. Now imagine the world that we live in without those three things happening. I posit we all care about hardware and energy and making stuff. These were probably more important for what the world looks like today than any of the technologies, I think. That's a lead in to... I'll just leave this abstract slide here and just tell you this romance novel. This is how we're going to save the world. The average US citizen needs 10,000 watts of constant power. That's what 100 quads per year translate divided by 330 million. Anyway, remember 10,000 watts per person constant load. I just showed you they only need 4,000 and their lifestyle is just going to be the same, right? Because we only need 40 quads, not 100 quads. So we need 4,000 watts, not 10,000 watts. The average American house has 2.6 people. So the Bureau of the Census describes a household as a thing and that thing has on average 2.5 or 2.6 people in it. So your average household in the US now needs 2.5 times 4,000. That's 10,000 watts constant power to provide for all of the energy uses in their life. Not just their electricity and gas bill, but everything in the US economy. The cost of solar today, you all know. In fact, I have a solar company and we build trackers and we're in the horrible business of trying to make any money selling solar energy at a dollar a watt installed. Capacity factor is about 0.2, so really that means $5 a watt, right? 10,000 watts for your whole household, $5 a watt. So that's 50,000 bucks for the solar cell to amortize all of your environmental guilt. What was inherent in doing that is I had to give you a better refrigerator and better white goods. Let's call that $5,000. I have to put a $10,000 battery in your basement to roughly give you two or three days. Storage. I'm going to magically solve the long-term storage problem. Incidentally, I think that's the hardest of all the problems now. I also have to buy your house two electric vehicles. My mother-in-law has a Chevy Bolt. We have a Fiat 500E. You can now get those for $35,000. So for $50,000 for the solar cell, $10,000 for the battery, $5,000. I've had $65,000 so far. $5,000 for your washing machine and better dryer. And then $70,000 for your cars. What are we up to? Let's call it $150,000. I think I rounded it up. It was $135,000. So we've got $150,000. If I could give you the same terms as this and amortize those things over 25 years, I think you all know that a 25-year mortgage on $150,000 is pretty low monthly payments. In fact, it's under $300. If I take the current American household and I look at how much they spend, the average monthly payments on all energy paid for the car, the natural gas and the electricity is about $400 a month. So it's cheaper today to go completely clean for the American consumer than not. Now, can anyone tell me what trick I played? Interest rate? Interest? No, I used 4%. 4% as the mortgage interest rate. Now, that may be low for the car loan that I did, but it's reasonable for the house loan if I could do it on a mortgage. The trick I really did was I used wholesale electricity at $0.03, whereas we're all paying $12 to $20. And I used wholesale electricity versus retail gasoline. So I actually think the giant opportunities, as well as being in the paper pulp industry, are in the price arbitrage. Industry. And what did we do spectacularly well in the last 20 years? What did the internet really do? It did price arbitrage between wholesale and retail. This is what gave us Amazon. This is what gave us the other one. That was Alibaba. These things are massive wholesale retail arbitrage. I think there's been tiny efforts to do this, but they don't draw the boundary large enough. So we do have your mortgage or solar cells for your home. Their mistake is doing the most expensive solar, which is rooftop instead of trying to do wholesale industrial installations. And their other mistake is they're not trying to get the benefit from your whole energy expenditures. They're just trying to do your electricity. So what I said to you may not be exactly true today, but it doesn't take much imagination. It's close enough to true today that it's either true next year or five years or 10 years out. And I think that's unbelievably good news for us. Because I think you can build an argument to any partisan politics that it will be economically advantageous for America to follow some plan roughly like this. And with that leading heart liberal conclusion, I now open myself to his sailing questions. Thanks, Mark. Well, that was both entertaining and eye-opening. I think you'd have to agree. We have time for a few questions now only. Any good? I'm sure they're all out. Yeah, I figured. Let's just do these in order. One, two, three. Oh, sorry. OK. How much could we cut down electricity usage by veterans or thermal usage by better insulation? Would suddenly cut down air conditioning and also cut down... I think that is a spectacularly good question and I care about it enormously because we've just started a company working on exactly that problem. I don't think I can give you a reasonable answer because the reasonable answer relies upon so many questions as to what size house are we talking about? Will you downsize the house or is it really just thickening the insulation and sealing the cracks? You've got to account for how much ventilation you're going to lose. You're going to lose some energy because you'll have to ventilate these highly sealed things, et cetera, et cetera. I'd bullpark the answer at 10 quads but I'd say plus or minus 10 quads on that as a response to that question and the plus or minus is really around sociological issues. How quickly are we going to change the vernacular of the built environment? I think I would love to live in... Everyone remembers Bucky Fuller's Dimaxian House. It was elevated, highly insulated, had all of these great features. It would have given you that 10 quad discount if we could roll out Bucky Fuller's Dimaxian House everywhere in America instantly. But people like their gable droos, et cetera, et cetera. So I think the sociological headwind of that makes that question very difficult to answer. I could give you a technical answer that is useless because I think the sociological, cultural aspects of that are really the heavy... the inertia in that system. Okay. Do you have a point for the technical answer? Well, technical answer. We know how to build zero-net energy houses in most environments in the world. Zero-net energy. And so I think by existence-proof, that's true. These houses tend to be about 1,200 square feet, the zero-energy houses. They're doing... They're lying to themselves about the embodied energy in those houses, the people who do the analysis, a few other things. But let's call it roughly 10 quad. This is a great question. It's another area, but we have to find a way to see the question of carbon out of the atmosphere and take it down and... Did I mention I love trees and gardens? Yes. We're going to take a lot more than just trees. We have to put it in the geological storage somehow. Either we can sink it at the bottom of the ocean or we can somehow put it in the ground again. And I wonder if you could put it in these boxes. But that little problem... I put... I have enormous concerns that we won't do it at all. And that the best we can do is stopping deforestation and doing a lot of reforestation. And I think that would be the kind of world I would like to live in is where we do both of those things. We have a crazy idea on how you make carbon negative natural gas. And we've written proposals to agencies that will hopefully fund that. And I've seen... But I honestly haven't seen... The carbon sequestration problem... You take carbon, you add oxygen, it's four times bigger. So all of these ideas are going to shove it down the holes that comes up. The problem is the volume doesn't work out. There's not enough holes that we've pulled it out of to stuff it back in. And it's going to leak. So I kind of think we have to stop... We have to have an honest conversation with ourselves, put that in the unbelievably unlikely and very, very difficult bucket and try to solve all the other problems. Because I haven't seen reasonable proposals of something that I believe is scalable and works that's better than just doing reforestation. Let's go back up here. So one of the big things that can carry momentum is price signals, but it's kind of tricky in the energy industry. We just don't see the price signals doing quite what maybe they ought to right now. Do you think that the price signals could be carried down to the consumer manufacturer level, or do you think that better price signals at the top of the line in the energy usage you showed of all manufacturing would be enough to... I'm going to use your wonderfully insightful, incredible question to answer two slightly different questions because that's too hard. I would love to see a price on carbon. It's got to be 100 times anything that has ever traded before. I think it's really unlikely that the markets are going to be able to figure out how to do it, to make a difference. More important price signal is PG&E's lawsuit over starting a whole bunch of fires last year. And this I think is the hope. My fantasy is PG&E goes bankrupt this year. I happened to be involved in a completely different lawsuit that was arbitrated by a judge this year, and I asked him what he did in his day job and his day job is the PG&E suit. There's about $100 billion in personal property and life claims. The market cap of PG&E is like 30, 40 billion hours trash. It's very difficult to imagine without the state of California propping them up that PG&E won't go bankrupt in the course of that process. They should. At the same time, if any of you have been watching the sort of compensation in Sacramento, not CalCEF, who's the organization that's leading the charge on 80% will be community-based energy? It's PCPUC White Paper. That says that 85% will be part of the $25.20 issue. So PCPUC, California Public Utilities Commission, at least believes that 85% of energy can be community-based, not giant PG&E-based. That would flip from 80-20 to 20-80, that ratio. So your fantasy right now, if we're playing fantasy football for the world, is that next year PG&E goes bankrupt. PCPUC succeeds with this because we as technologists make rooftop solar on electric cars and all of the other options available at reasonable prices. And that becomes the model for the future of the world. And we sell our solar at $0.03. $0.03 a kilowatt hour is roughly what a dollar or what translates to. So then you say, well, why are we paying $0.12 to $0.20 as consumers for the solar that we are now producing at $0.03? This is like five substations between you and these fields. Each one of those substations is adding about $0.01. That's how we get it. We go up voltage and then we go down. Anyway, each order of magnitude of voltage regulation costs you a penny on your bill and there's a $0.03 in billing. So the more we do it at the community level with giant car parks covered with solar and on your rooftop and all the other things, that's so close to your meter there's only going to be one local substation. So I'm actually much more interested in that, those pricing signals than the larger carbon tax or carbon whatever is going to solve the problem. So that's a complete dodge from your actual question. That's an end though on that positive note and thank you so much last time.