 So I'm going to go ahead and introduce our faculty member today, John Wyatt. Amongst many things, John is a professor of management, science, and engineering. He's the director of the energy modeling forum and a senior fellow at the pre-court institute for energy, which is the institute I work for. His specialization is in energy in environmental policy analysis and strategic planning. He has had a very broad influence during his career, including being an active advisor to the United Nations, the European Commission, the US Department of Energy, the US Department of State, and the Environmental Protection Agency. In California, he's been an advisor to the California Air Resources Board, the California Energy Commission, and the California Public Utilities Commission. John was honored in 2007 as a major contributor to the Nobel Peace Prize awarded to the Intergovernmental Panel on Climate Change. Quite an honor. And John's overall goal is to accelerate the use of systems models at all levels, state, federal, etc., to provide the best available information and insights to government and private sector decision makers so that we can move forward in the best ways possible. And in my experience, John is a warm, curious, and caring person. And with that, I give you John. Thank you for that very gracious introduction. And I actually feel most pleased that you added that last part, because I think that's important. I'm known as somebody who recruits grad students and employees based as much on character as technical skills. You guys all have such awesome technical skills. I think for the kind of work we do, it's good to be a good team guy. I'm also a philosophical devotee of the great American philosopher, the late coach John Wooden, if you know that. So I could spend the whole time giving you John Wooden quotes, but I won't do that today. So I'm going to try to share my screen. I'll try to go on. So as Raheela mentioned, my longest standing role, which I've done, probably half my time for the last 30 plus years is to direct the energy modeling forum. I will probably not talk about that explicitly, but show you some results from some recent and pump prime your interest by talking about some ongoing and upcoming studies. I have done several roles for the Precourt Institute for Energy. I helped Jim Sweeney as deputy director run the Precourt Energy Efficiency Center during its 12 or 13 year existence. Most recently, and to put in a plug, I am the faculty coordinator of the Stanford Energy Seminar. You should have gotten a flyer regarding our program for fall quarter or we'll shortly get that. So moving on, I know I only have a 50 minute slot, so I'm going to try to talk for about 30, 35 minutes, but I often dabble on endlessly. So I'm going to greatly economize my normal talk, which is on energy environmental systems analysis and policy. I'm going to spend just a few minutes giving a little background on me, a background on modeling, because of the kind of work I do. I reflect often on why are we doing all this stuff in the first place. I've renewed my interest in the same in the fake news era, if I could call it that. I don't want to get too political in this talk. I am not really here to teach you how to be an advocate. There are better people to do that. I'm here to talk to you about what you might want to do in terms of developing your systems analysis tools that could help you be a advocate on either side of whatever your favorite divide is, but mostly my objective is to try to provide neutral information to all parties. So I'll go through the first three items in about three or four minutes, and then I'm going to focus my talk just to give you a flavor for the kind of work we have been doing, are doing and will do in the future on some basic concepts that we've found useful historically. Some of these for some of you will seem old school, but I will quickly go through some fairly recent examples and then dive into the directions for future research, which hopefully some of you will get actively involved in and or will choose to do so in your time here at Stanford. In terms of what you said, my general interest since my college essays in high school, which was a few decades ago, maybe four or five by now is to use analytic methods, use my skills in math and physics or something to help solve big problems. I focused mostly in what we call in the MS&E department, my home department, policy and strategy domains. My parents accused me of trying to stay in college as long as possible. I started a little bit too late to be a star in the space or a major player, I should say, in the space program and aeronautical engineering and astronautics. I went into engineering management, operations, research and statistics, and then my PhD is actually in management science from a small public university across the bay. I was apropos of the MS&E curriculum or coverage. I did minor in economics with a master's degree equivalent organization theory and finance at that point. I then did a postdoc in public policy and political economy at the Stanford of the East in the Boston area and along the way did four years of summer internships and then the Rand Corporation paid for my dissertation effectively, interestingly on synthetic fuels and my thesis was on public policy, R&D priorities in energy, energy, R&D policies. My Rand spinoff was should the U.S. Air Force have any interest in synthetic fuels at that point? They were thought to be produced not from biofuel sources but from shale oil as it was then called and coal if you can believe that. So just quickly kind of do what I was mostly focused on during my somewhat lengthy career now. I was really focused in from high school through the undergraduate school and a bit beyond on the core national security race to space. Then my Rand group interestingly got into air and water quality. I did an air quality model in LA where Rand is interestingly. By the time I got here I then morphed that into work on energy security, the acid rain problem. I then developed partly because I had been in a business school for part of my graduate work on corporate strategy. We could talk the whole talk about that but I won't. And then I would say now I focus mostly on climate change as what Hula had said in my introduction at the international country and state levels. Also I've gotten more interested in not just the mitigation policy reducing emissions but also in multi-sector, multi-region, multi-stressor dynamics which is part and parcel of extreme weather events like hurricanes and droughts and more recently wildfires. I have kept a play through stream through all these interest areas in technology and innovation policy and uncertainty analysis. I would say I am probably a somewhat balanced but if you wanted to label me with a bias I'm kind of a technology optimist. I think that goes back to the 60s where as a probably junior high schooler I thought Kennedy was nuts to announce we were going to land on the moon in that decade and we did. So the next question just to wrap up this preamble is just to get you to think a little bit more broadly. I think often we all students, faculty get kind of into the middle of a debate and lose track of what the heck we're trying to do in the first place. So the question is why do we build these analytic tools to help us make decisions? We need insights and numbers for policy development. It's easy to decide what to do based on political philosophy or it's easy to see that now or kind of religious beliefs I guess I would say. But at some point people are interested because the problems are so massive there's so much uncertainty and so complicated. A little bit of data and analysis goes a long way. I won't take the cheap shot on the COVID epidemiology but you could easily do that probably better than I could because you're more directly affected by it. I'm just hanging out at home right now. So what are the advantages of modeling? We sometimes forget the easy ones like providing consistency and the different thoughts we have about how the systems of issue are evolving and what we can do about them. We try to learn insights, robustness about what policies seem more robust across the wide range of uncertainties and contingencies. We struggle in interdisciplinary work with what principles do we use? Do we use disciplinary principles which sometimes are inconsistent within but certainly in between disciplines? What do we mean by empirical evidence? You see in the COVID debate a lot to talk about science as if data is science it's probably not science and as if the data is all perfect which as we see every day in the evening news it's not quite so perfect. Then I wouldn't spend a lot of time here but I do spend a lot of time in real life working with the global modeling community on issues like how should models be evaluated to get people thinking that they are more credible. First and foremost who decides all these things is that the disciplines themselves the natural academies the lawyers more often than you might like or us. I like the us as long as it's you guys and related people. So back to the storyline here I'm going to quickly go through seven basic systems modeling concepts that I have been useful through the years. The first couple of things I did even in my thesis that have now evolved a lot the middle of the list are things that I did early in my career in the bottom with the exception of the very bottom one are things that I am doing now and hope to do more in the future. The first one is a way to compare different energy technologies in terms of their cost environmental performance this is on the cost side there's this concept called the levelized cost of energy of the levelized cost of electricity so this is a static view pick a place pick a technology try to add together the fuel cost the operating cost and then the tricky part is the capital cost the cost to build the facility that you're using to generate the electricity state and the the the that that's the hard part because those facilities don't last just one year so if you want to do per kilowatt hour you kind of have to amortize if I could use that finance accounting term the the usefulness and life of the facility over the it's full lifetime so you could think of a simplistic way of doing that and say well this power plant be at nuclear or solar last 30 years 50 years let's just take that cost and divide it some of you who are business school um business oriented know that you probably the right way to do this is to assume that you're borrowing the money so either you're borrowing it from lenders or from your own retained earnings or own profits and you should take into account the financial time value of money that you could have invested that in something else so you get a capital recovery factor so-called that boost that capital cost component a bit above what you would get by just dividing by the number of years this is an old diagram just for this I did at the beginning of the global climate and energy project on a project that Jim Smooney around and I were involved in and constructing the original GSAP portfolio working for Lenore and Chris Edwards who are then the directors of GSAP I don't have my full story on this which was an animation which took solar PV and working with the solar research teams here mostly in material science and Kimmy I had an animation that said here's where we are in solar our group thinks we can by 2020 or so which is this year reduce that cost by a factor of 10 I did an update on this for the pre-court advisory council about four or five years ago and the group we were reunited and the group said well now we think a factor of 10 is too wimpy we're going to go another 50% so lots of changes this unfortunately also is at a time in our history where natural gas prices were pretty high so the cost for solar PV in particular a little bit less for solar thermal have come down quite substantially but we have had a fuel cost reduction because of the shale gas boom but now we're seeing in many regions solar PV can compete with not just nuclear and coal but also natural gas not every region but some regions this has led fast forwarding into later on in the talk a issue of grid integration so I think on the margin we can expect to see what we've seen the last five to ten years dramatic increases in solar PV wind and to some extent solar thermal because you can store in thermal form the generation for a longer periods of time but the bits and watts initiative here alongside the the kind of material science-based battery initiative have been working on better ways to do grid integration one link to what's coming up is part of that problem is you have so much solar PV now or we'll soon have it that you have this duck curve effect where you get too much solar generation say in California midday you can't even use it all and then by the time the sun goes down you're running really short and give it the dynamics of operating the power system that's the big problem so you need batteries or what's called demand response incentives for people to use more electricity in midday unless at the end of the end of the day so moving on in this that so that concept I actually used in my dissertation many moons ago was then observed by a group of people I won't go through the litany that you could put together these pictures of levelized cost in a spatial and temporal diagram of how the energy system evolves using mostly our principles from physics and thermodynamics that kind of trace the possibilities as you go from resources in the ground or in the air through various conversion processes a big one in the US and elsewhere is converting those fuels into electricity to satisfy various end uses the big dogs in the US are obviously in the challenges right now we still use a lot of coal for electric generation and a lot of oil for transportation processes so to get into the MS&E part of this we have a whole group that does optimization and various stripes now you could think of putting together an energy systems model where the objective is to minimize the cost of operating the system including the fact that you might increase costs slash prices to reduce demand you don't want to leave that as a freebie you want to kind of weight that into the calculus you use and you have constraints on satisfying energy demands only using the available resources so satisfied demands only use available resources and convert energy forms and efficiencies of the available technologies a big challenge right now is how will these technologies evolve in the near and long term and alongside that how fast will the new technologies even though one could argue they are superior actually get used and adopted in the marketplace I think we need our business school colleagues to both analyze that problem and put together business plans including finance and marketing to accelerate that so as I said you you can run a energy system model to calculate the reduction in total system cost including the welfare loss from reduced demand and use that calculation to sketch out what's called a marginal cost of emissions called a controller a marginal abatement cost so here I have the level of carbon dioxide emissions and if you don't tax you you don't pay anything to reduce emissions the more tightly you constrain it the faster it goes up in recent years we've seen this curve tilt down substantially probably not despite what some people would like to believe down to completely zero or negative although that would be nice and people are still working on that the problem with doing this solution which is low cost solution is we probably incur a lot of climate damages so I'll spend a few minutes on where we've where we are we have many experts here on campus on this side of the ball the marginal cost of climate impacts there are four basic ways that people have used in IPCC reports and national climate assessments and the modeling community and those are structural models that take physical physical and economic inputs into a structured process-oriented model of a particular sector in a particular region cross-sectional statistical models that look at variations say temperature and precipitation in agriculture across different regions and try to tease out kind of common technology and crop choice and adaptation choices like irrigation across different regions I wouldn't say anything about travel cost models and continued valuations these are ways to evaluate impacts that are not producing market information that one could use in a statistical or econometric approach there are three types of ways to evaluate agricultural impacts apropos the last side is straight statistical approach a process approach or a high bid approach a signature article in the 2009 and the proceedings of the natural academy for the first time showed results that had not seen before they basically said suppose yield growth is a non-leader function of heat so that log yield can be represented this way kind of the typical tricks the smart econometric econometricians use to do these cross-sexual analyses with time-inverting country fixed effects to make a long story short for the first time despite the fact that earlier econometric work said there isn't much effect a very slow decline in major grain categories corn, soybeans and cotton there is only a very gradual effect they saw breakpoints part of it was focusing on the extremes in the temperature distribution another fact was they had data that was now aggregated at the county level as opposed to the state level or big bigger regions within the U.S. so this so-called Schrincker-Roberts paper was signature our own David Lobel in earth system science of the school of earth has worked with M-factive through one of my projects sponsored a comprehensive follow-on to this work which I'll give a little bit of intro into in a second the other way to go about this besides this more aggregated statistical approach where you essentially put a few physical values in with the typical economic drivers and outputs in terms of effects on crop productivity so called is to do a detailed physiological model and we have the whole bio eco community at stanford that does leaf photosynthesis canopy photosynthesis crop growth development and yield here you look at soil conditions soil moisture much more intensively on both carbon balance but also importantly nitrogen balance and probably phosphorus by these days chris field our head of the woods institute here is signature and has done many co2 fertilization and other nutrient fertilization experiments at jasper ridge so you put all that science together in a detailed process model the same slinker robits now working with David Lobel who I've already set up actually said well is a statistical model or a process model better as is usually the case the combination of the two which they are artfully put together does better than each one individually at a half time to really justify that with historical data this is a projection that shows what one might expect for a two to two degrees see rise and a 20% rise in rainfall obviously bigger fear is where you also have decreases in rainfall um so this uh pass forward to a special issue that one of my projects sponsored with this group and many others tried to do a multi model multi um uh model methodology uh uh comparison uh as David had predicted up to one degrees the the results from the statistical and process models are quite similar and then you see a breakpoint you see uh harboring back to the uh slinker robits you see these breakpoints which some people call tipping points now in the climate debate so that's an interesting thing to keep an eye on and makes me less sanguine about saying oh these these will be slow gradual as opposed to more aggressive ones concept five is uh to put the economic calculations that I showed in the energy systems diagram which are basically partial equilibrium we only consider one sector the energy sector and the cost of doing that and that gets traded off against uh aggregate reactions in the rest of the economy to break it down into multi sectors energy and non-energy and then reformulate the objective function to be overall consumer utility or welfare in that way you can both look at structural changes in the economy as it affects energy demand and solutions that seek to change the structure of the economy as well as the energy technologies per se as a remedy for things like climate change and even traditional air pollution uh little footnote here to set up the next slide you can also at the level of consumer utility tradeoff consumer choices between consumption and savings which affects the amount of investment you make which has a long tail of benefit so the Ramsey rule uh means you trade off your immediate gratification from direct consumption or you take your productive output and invest it in capital equipment which has a small but long lived increment to future production possibilities you can also put leisure time in here to get a full picture of consumer welfare uh concept six is over the years we've learned often the hard way that these more stylized uh perfect market uh equilibrium or optimization models leave out important things like consumers and people investing in energy technologies uh don't have microprocess in their brains or Stanford degrees so having promoted energy efficiency in my dissertation i learned the hard way that you couldn't just do the analysis and expect the man on the street main street to immediately do what you say uh that you needed to help them understand what those more detailed studies were saying they weren't about to invest half their available time in learning the data and and how to put it together to make their own decisions we learned that not only energy but water and land markets are far from Perkwik we have high uh degrees of regulation steel in many states in the u.s with regulatory agencies uh preventing a kind of least cost so-called efficient standard we've also learned the hard way covid makes this even more obvious that there are multiple objectives for decision makers across both energy environment and the economy but again we see now in bold relief equity energy access energy poverty and more generally sustainability issues where often the least the most disadvantaged people among us are put in a non-sustainable state in support of us i'll use the term elix i think you know what i mean by that achieving our climate and other objectives so one of my favorite ways i now use in my classes in fact to go from this more um kind of mechanical uh first economic and operations research principle point that which i started and even the engineering principles is a thing that Pam Mattson and colleagues have put into a book based on 20 years of research on a sustainability science approach the reason i like this is you see the normal uh economic way which i've already laid out of thinking about this problem where you have catholic assets and you engage in productive processes to produce good services and then you consume them they immediately say oh no but that is an interesting place to start but we all have objectives that have multi dimensions to them on the one hand at number two in addition to the normal economic model which focuses on manufactured capital as i just described it with the little fly to hand uh we also have a need to keep an eye on human capital what are the abilities of our citizens through education and training to do things natural capital the ability of our unmanaged ecosystems often to provide vital services for us as a society and knowledge capital which is basically ip knowledge about how to do things differently or better are those uh subjects have been studied by economists in in more narrow studies but not brought into the larger models and my personal favorite is social capital which says do you really have the ability given the existing organization of human activity through markets or rule-based allocations to actually deliver the goods that you see coming out of these more narrow models so there is a book on this i highly recommend the final concept which i'll go all the way through the application was actually done believe it or not in the 1990s it's in a book called buying greenhouse insurance published in 1992 that says really climate change is a risk problem it's not what we expect to do it's what we fear could be coming up in the tails of the distributions if i could use that term so then i've done a little bit of a stylized thing suppose you check that normal distribution around possible outcomes across the climate sensitivity which is the amount of temperature change for doubling of co2 which is a key statistic from the climate models and the damage function which is per a two-degree warming what would the reduction in gdp be and basically truncate it take the tail off of that center kind of do the mean of that just tail and then do the mean of the rest of the distribution which is kind of like an expected value decision maker might do and then look at two cases one is what we've normally done historically is clairvoyance meaning we learn the true state of the climate system now and in the future immediately and we act accordingly as opposed to we're really flying by and for a number of decades about what the climate sensitivity and the damage functions actually are to make a long story short we see a couple of interesting things in this kind of a stylized picture we do see even in the optimistic or expected value case we do we do get an indication in a cost benefit sense of substantial emission reductions we do see if we're we already know we're in the more dire scenario for sure we would immediately reduce emissions why we're we're kind of moving closer to that now as you see in the news every day but more importantly in the risk management case we see that the optimal strategy given that we don't know the what the future will bring on either the mitigation or adaptation side or damage side the what's indicated is some hedging that I can use the parlor term from finance I did real options in finance and in fact the hedging is non-linear so this is more than 10 percent of the way from the happy face scenario to the more dire scenario so that is something I'll come back to at the end of this talk so I'll now in a few minutes go through some applications cost benefit analysis I've already given the full introduction I just display these are a little bit dated Bill Nordhaus is I'll come back to him in a second playing off of a two-degree target against a baseline scenario uh against a two-time co2 historical pre-industrial co2 emissions and his optimal which was then 3.4 it's actually a little bit under three now which isn't too far from two but far enough to simulate some debate I wish we were doing this optimal scenario now but we're not at the global scale here are the corresponding prices I will remind you probably my oldest and most productive non-stanford college Bill Nordhaus actually did win the Nobel Prize at economics not the peace prize as I was involved in for integrated climate science into long-run macro economic analysis I welcome questions on that in our EMF studies which he draws on to aggregate the marginal abatement cost uses in our in his model we actually have done a sequence of studies on us climate policy I'm going to do us and then global we've looked at these kind of efficient policies using either a uniform carbon or co2 tax to two equivalent tax or a cap and trade system we can trace out through many scenarios a what the economists call an efficient frontier interestingly we don't seem to be able to do those policies in the u.s now historically or now and probably in the future so that interesting to look at is the kind of specific policies and measures that do seem to get more salience at all levels from local to state to national and global regulations on fuel efficiency standards for automobiles restricting coal use this particular scenario bands new coal power plants with a renewable portfolio standard confining the two the steve chu jeff bingaman wrinkle which was a clean energy standard which was like an rps but you give small credits for a nuclear and solar interestingly we see these are all over the place we wish in terms of cumulative emission reductions we always want to get more of those for a lower net present value total discounted cost between 2015 and 2050 we want to be over on this side that's why we think of this as the efficient frontier we see for this particular model with this particular metric in this particular technology scenario how things search out that we keep the technology scenario in the metric we see that in different models these are all kind of name brand models often used in policy model they're all over the place so one take home is some of these individual policies and measures get us pretty close to efficient frontier more surprisingly some actually are underneath it I won't go into why that is in any detail but even the most hardcore economists now understand that because the economy isn't a perfect perfectly competitive economy in this case has pre-existing taxes on many goods and services they shouldn't be so sanguine thinking that this frontier is the true and only righteous solution quickly I'll go through at the global scale what we were asked to do by the national and global climate negotiators is take our pre-Copenhagen round of model comparisons and put in a more complete set of technology options and more realistic policy alternatives so this is a kind of dense diagram that says if all supply technologies are allowed to progress in terms with the market economics included in the models with no restrictions this is the cost range you get for a 450 scenario which is probably about 3.3 or 3.4 degrees equivalent on expected science here's what happens if you restrict the individual technology options this doesn't look too bad you're kind of under 1% of just kind of a GDP for this type of scenario but if you go for 450 which is more like two degrees C you get a much wider range so this is a wister diagram showing the kind of average value from the different models for the individual scenarios this is the full range this is the 25 to 75 range not a problem with these distribution just the range we also saw in this somewhat controversial that relative to the all in or all the above strategy we are particularly sensitive to not having carbon capture and sequestration or biofuels I would say this hides somewhat the chapter I wrote underneath this which showed that energy efficiency has huge benefits and if we were a little bit more aggressive about that we could reduce our dependence on these two technologies and underlying this renewables are quite handy in any scenario back to uncertainty a interesting what I thought was a hokey approach is the MIT approach which took a different version of the sequential decision maker uncertainty and said even for modest climate control there are really two benefits one is the mean temperature reduction is likely to be less and also the higher end of the distribution even for modest controls this is a 675 part per million nowadays of course we're more fearful based on the last couple of IPC assessments about this range so we still have a lot of work to do but if we had done what Bill Nordhaus said we'd already have taken most of this part of the distribution off the table but unfortunately we haven't done that so just to wrap up here I'd like to go into future directions I've tried to set these up we are just finishing a study on biofuels with carbon capture and sequestration but better sustainability considerations largely informed by the ecology groups including Chris Fields group at the Woods Institute we are looking more closely at the juxtaposition of climate targets and sustainable development goals out at the UN we are looking at ultra low greenhouse gas emission scenarios with greatly accelerated energy efficiency I already mentioned I'm starting a new energy modeling for study on higher electrification scenarios for North America the idea there is you decarbonize electricity and then you electrify all the end-use sectors we are looking at more far out things like director capture and solar radiation management and on methods we're doing integrated risk analysis we're focusing on equity I've already said these things my own personal project in this area is largely focused on energy water and land dynamics for the data scientists among you we're now trying to think about how to combine traditional energy systems that impacts model with data science concepts this is a plug for my own project with a group of 20 collaborators at different universities called program on coupled human and earth sciences and this is really for the climate impacts managers and adapters to look at the regional and sectoral scale because we can't really change the climate signal too rapidly and we're struggling with hurricanes fires and the like so to end here's my snide pushback to the people who try to put us into a narrow box and saying you're really not doing things that are good for society it's just your crazy left-wing extreme perspective that in taking this more holistic approach that we are now able to do more than before but still not very perfectly we are looking at a wide range of sustainability dimensions so that's all I had I left maybe 10 10 minutes or a little bit more for questions so do we have any thinkers in the audience I'll stop sharing at this point wow we were only supposed to have 15 people it looks like we exceeded our audience side by 10 so that's good so any questions comments on anything policies modeling Stanford courses majors career paths and so on we had a question from Benjamin early on Benjamin why don't you unmute yourself and ask your question for first of all thank you so much for this conversation it's been really interesting my question was related to a couple of points you run up across the presentation around institutional barriers and also around the sort of difference in capital costs or different forms of energy sources I remember seeing on one slide I looked like the solar and wind infrastructure investment required significantly more sort of capital upfront as opposed to you know that being balanced out by operating costs down the pipeline and I was just wondering if you could maybe speak to you know what some of those barriers are that like might prevent you know either a private investor or a public utility from wanting to invest in that upfront capital and how you sort of get past that okay so that's an interesting question both historically and prospectively we do have the sustainable finance initiative here which I urge you to take a look at which does this more systematically but to answer your question so a big problem we observed in the energy efficiency sector work is most consumers don't have easy access to capital so they are this is called the first cost aversion so this means you don't want to put 15 or 20k down for a solar array on your roof because you're feeling like you don't have easy access to capital markets now there are two things you can do one is the institutional one you can have a third party come in like for solar and basically say well if that's a problem we will take care of it for you we can pool your interest together we can take title to the solar rays on this so this was alongside the technology advance a huge stimulus for for solar so the way that worked is they said we will offer you a reduction in the cost the traditional utility would cause would charge you for electricity and you won't have to either pay that upfront cost because we'll pay it for you and essentially charge you a favorable financing rate to recoup that and still keep you under the utility rates so that was a major step forward but at a bigger more systemic level once you get out of households the same kind of things impact industrial consumers in the industry and power sector and that's really where the sustainable finance initiative is focusing on a flaw with the models which are trying to work on as many of those assume that this capital recovery factor is the same for all technologies in our regions so even this is a kind of institutional behavior thing even corporations fear that an untested technology will look good and then it won't work out as planned for various reasons I don't have time to get into and they'll be left holding the bag so there's a lot of work now and I just talked conveniently to Tom Heller the leader of the sustainable finance initiative he and I'm trying to do this from the modeling side but he's out there the front lines working with Wall Street but also the the development banks in China India Brazil etc etc and this plays out differently in every world region so where we ended up which I'm trying to get the modeling people to do is you have to be very specific about which acts for you finance people which assets you're talking about and whose portfolios you're trying to manage he's come up with a problem at the sovereign level at the country level is if you go for one of these aggressive this is a new problem which does provide an institutional constraint if you buy the story that we need to go heavy renewables and you set up the system so that your investors in your domain can do that and the whole thing craters we get kind of a big short scenario in the real estate market I've there left holding the bag and get blamed for all that so there's this kind of reverse thing now just identifying that as an issue and a constraint doesn't mean that there aren't creative ways to work around that and that's where a lot of the attention on the finance side is going right now so that was a great question thanks and we have a quick question here are the emf 33 and 37 labels are those labs or classes I know believe it or not this will make me seem very old they're the number of the studies we done so I was a grad student during emf one we've now done we're starting on our 37th modeling or comparison project we did emf 34 on net and north american energy integration because of the current political situation I parked 35 and 36 in japan and europe and they are doing asian region to global scenarios in the japan instance and sustainability scenarios and trade scenarios in the european thing so that in fact the german and japanese governments are paying for that and borrowing our copyright it just means I have to get up early in the morning and participate in their steering committee events so it is literally emf 37 is our 37th energy modeling forum study believe it or not great and we have a question from my yank my yank would you like to unmute yourself and ask your question great thank thank you very much for the lecture professor very very encompassing of a broad topic you touched on some of the behavioral constraints that can impact modeling and how how we should be looking at problems to that we face can you just talk a little bit about how behavioral economics and modeling of that comes into play in potentially climate change but also in the energy sector for example in terms of grid balancing when energy sources are becoming more distributed in nature as we go forward with renewables or behavioral economics in terms of how people make decisions in the face of uncertainty when they're basically trying to figure out for example at what level to set carbon taxes on or how to set variable tariffs for renewable energy supply that kind of thing yeah this is a very hot research topic and there are many students at all levels from freshman to that to graduating PhD students working on this 20 years ago all the top economics department said behavior is a second order of fact we're not going to waste faculty slots on behavioral economics now everyone wants to hire their third or fourth month I would say a personal story Amy Levens and I came up about the same time and in my thesis I promoted energy efficiency as our first best solution to some of these same issues which existed in a different form back then he was more promotional about that so he's much more famous than I am and I think the thing that Amy and I both missed I already alluded to is we thought once the information we we understood the information it would be easy to transfer it to normal people and they would act accordingly behaviorally economics came in and said not so fast you really need to think about who these people are and what their ability to you know take in information and use it to decide on even what's best for them in terms of their objectives let alone societal objectives so that leads to a bunch of constraints and opportunity that don't work in the within the price system these are generically known and behavioral economics is behavioral nudges so these would be information programs we will make the information more accessible and my personal favorite uh demonstration programs we will subsidize people heavily to go out in your neighborhood or in your utility region to actually do these things including your home utility we'll just oversubsidize them so they don't even have to think about doing it they will do it and therefore demonstrate that it can be done and how it can be done I would say when most consumers and even utilities uh uh consumers industrial if the if the government says to do something they don't necessarily believe that it's going to work out so well and if the regulators say they may even be more um uh suspicious of that but if you actually constrate subsidized demonstration projects on the ground they can do that so I'm seeing that more and more now you kind of did a fast one because if you go from the consumer level either in the industry or consumer into grid operations you have a whole another set of issues like I mentioned demand response my most advanced PhD student is working on demand response that's getting people used to bidding into the market when you get the duck curve in the afternoon can we pay you a subsidy in terms of lower overall rates for the right to turn you off you know turn your refrigerator off for one hour but not two or turn off your AC remotely or upon issuing an instruction so I think that's where a lot of the action is now that you'll see around bits bits and watch and storage action so on two of the big initiatives going on thank you fantastic thank you yeah um bits and what's did a has been doing a digital grid webinar series and the last one in that series was on transactive energy which is very interesting you can find the recording of that on our website yeah and I believe well he'll actually runs the seminar series is that correct I do I ran the yeah so you have an expert on what they've done you the website is great but you could even contact with he'll I'm sure she would be willing to tell you who's talked to who they have in the in the drawing on the drawing boards now for future seminars yeah I for those of you who don't know I'm the program manager for the bits and what's initiative all right we've got another question from question would you like to unmute yourself and ask your question and just time check we have seven minutes left so let's try to be concise sure thank you so much for the wonderful talk professor I just wanted to ask you towards the end you alluded to efforts underway to integrate data science and techniques to systems modeling could you please explain a little bit about like what some of these efforts look like okay so I'll give you my own personal take on that which is one of many probably it really is so historically the energy environmental modeling community has been very structure-oriented so as I presented a lot of work on structural models much less data driven I actually have talked to the people who do cancer modeling and they're almost exclusively natural history data driven in that level I wrote as early as 15 years ago that it was good that we were using these first structural principles but we were not updating even our initial conditions rapidly enough so this data play right now is to try to take advantage of both sides as we have more proliferous more disaggregated sensor driven data sources that we didn't have before and so my and some people think that that's all you need now at the operational level like grid stability what's called unit committed modeling even capacity planning model you could see substituting a pure data science machine learning type approach for a more structural model but I would argue as you go out you get structural shapes that make it harder to detect structure because it's not evident in the historical data sets this is my own personal view I've used the same argument historically for econometricians who don't want to put much structural change prospectively in there personally and you could probably already tell this I think the way forward lies in the interregnum between the two so I call this data simulation we ought to get and I'm working with a group at the national renewable energy lab who has a big data science group and some of the best traditional utility modeling people and they're trying to bring those together and I'm trying to get a little research group with one of the people while he works with liang min and the directors and some of our local alumni some of which I've worked with over many years at places like deep mind you may not know there's never deep mind the company based in the uk that beat the go masters and now the big online computer games they actually have a division run by one of our alumni's in mountain view that's called deep mind energy so we're trying to get the data science people at stanford deep mind and enrol to compare notes lots of nda's and non-disclosures but I think a lot of interests are enjoying ish on the stanford side trying to do public good work so if you're interested in that I'd be happy to talk further about it I do feel like we stand to gain the most again by constructing hybrid architectures which takes advantage of the relative strengths of the data science approach and the structural approach and avoids the limitations of on both sides right thanks for that wonderful answer I will definitely reach out to you from John Foy John would you like to unmute yourself and ask your question and then a couple questions in the chat we're going to run out of time but I've saved those questions I'll give them to John and let him give you feedback at a later point great well great thank you both for the for the talk and the presentation um mine might be quick um but I had a process question so if you have a particular area of interest or a question on uh work that's been done to date what's the best way to do that okay I'll do this in two parts we do have a new program under diana graga I don't know if they met diana which is supposed to be providing resources on who's doing what and what courses are available diana was pleasingly for me a postdoc with us in the precord energy efficiency center and now runs the energy resources course thing so she has a project I would say if you want more information on energy systems model or so-called integrated assessment models the global kind of different types I would uh I would nominate me because that's what I'm most known for so I actually I didn't have time to talk about this run for now 14 years a consortium of almost 60 global modeling centers where we try to coordinate amongst the different regions and different disciplines the global scale work and the peaches project that I mentioned is a usdoe thing one of four multi-sector dynamic projects that are kind of state-of-the-art ish there are a few other people I can tell you about but I'm probably at this point one of the best people to consult about who's doing what type of work I do lean on people like wehila and liang and the other kind of initiative directors and the precord and woods directors to keep me more or less up to date but it's so you know I would say 20 years ago I actually have a talk on this Jim Sweeney and I and a few other people knew everything about what everybody was doing and now that's even impossible for the precord and woods directors so we need to work as a as a as a team yeah thank thank you very much John I'd say the good news and the bad news is that people are working on these problems in a vast array of universities and companies and so the work is being done in lots of different places and that's good news and it's bad news because we don't always know what everybody else is doing I've made that point even on even on campus in in California it gets harder in the US it gets harder so my my pitch in this I actually talked about this precord advisory board meeting and this is my first believe it or not my first priority we could do bigger stuff together if we just had better communication we don't want top down this is not it would be antithetical to Stanford to have a lot of top-down direction but there's a huge benefit and the kind of work you all are interested in and knowing what each other is doing because well organically it's very silicon valious I know well organically you guys if you see that we'll say oh there's a huge gap here I'll start a company or a new research program get a PhD get an MBA go to law school you know we have great examples of all the above I am on the EIPER exec comms so any of you people who are in MBA join program I'm actually helping run their capstone class this this quarter in fact all right well thank you so much John that's I'll give John a hand thank you so much for you for sharing your masterful knowledge with us today and all of you students I'm supposed to remind you to go off to your enroads team meetings now and John also you are going to help out with the enroads yeah yeah and apropos of the times let me end with my people have probably figured out my background so I this is for you guys to go and do the vibranium equivalent magic out there in the real world thanks