How energy systems are shaped

Kara Miller: I'm Kara Miller.  

Robert Stoner: And I'm Rob Stoner.  

KM: And from the MIT Energy Initiative, this is What if it works?—a podcast looking at the energy solutions for climate change.  

Today we're going to take in the big picture view of how energy systems are shaped and which technologies might win out. But first, just consider how indispensable reliable energy is for us. 

Jessika Trancik: We rely on our energy systems to allow us to do work—physical work—process information, support all economic activity. And we're always trying to, you know, improve that energy system. I mean, it's never been static.  

KM: That’s Jessika Trancik, a professor in MIT's Institute for Data, Systems, and Society.  

JT: If we look back to the beginning of industrialization in the late 1800s, we started to adopt coal-fired electricity. And then over time, used more natural gas, nuclear power, increasingly solar and wind energy. And we also have the fuels for transportation, for industry. So things are always changing. And the economy, the country that can have the best functioning energy system is going to have an advantage.  

KM: That advantage, she says, can play out in a number of ways. It can be an advantage in cost. It can be an advantage of maintaining human health by, let's say, creating cleaner air. It can be an advantage in producing fewer greenhouse gas emissions and helping to reduce climate change.  

JT: It's really difficult to isolate oneself from the evolution of technology that's happening and really changing this energy system, which is such a fundamental part of our economy.  

KM: But of course as policy changes, so does our support for different technologies. Trancik says it's likely that the technologies with the most momentum are just going to keep moving forward in our global marketplace, almost no matter what. 

JT: Solar energy is growing rapidly, wind energy, electric vehicles, batteries, and heat pumps. To step back from those markets and not invest and not try to be part of that transition is risky.  

KM: Part of that risk is missing out on the sorts of jobs that these technologies generate. And those aren't just jobs for scientists, nor are they just jobs for factory workers. More and more, they're a whole different kind of job—a job that brings technology to the customer and optimizes it for them. And in the U.S., where far more people provide services than actually make goods, that is an area for real, potential growth.  

JT: Installing, let's say if we take rooftop solar, installing the hardware and acquiring customers and everything that's not hardware represents about up to 50% of the total cost of these installation projects of the actual delivered service, the solar on your roof. And so there's a lot of room to improve that. There's a lot of room to find new creative ways to bring in energy storage or to, you know, at the utility scale to combine solar and wind energy, you know we're at the commercial scale looking at commercial solar installations. I think there's so much room to innovate in that space.  

KM: But how did a technology like solar get to where it is? How did it become cheaper, better, increasingly ubiquitous? Trancik’s research looks forward, but it also looks back.  

JT: In the 70s and 80s, Japan was a really major player in terms of creating incentives for the growth in markets for solar energy. They subsidized the adoption of solar energy and that grew markets for solar energy in Japan. And then Germany took over as the leader through their feed-in tariffs, which essentially guaranteed a price for solar electricity. That incentivized the growth in solar energy markets. And there were other countries contributing at this time as well with their policies, but I'm just talking about those that were leading in terms of the size of their solar markets through these policy instruments that they were adopting. And then more recently, China took over. And what you saw was that as these different countries were growing their market share through the use of these policies, private companies competed for market share. They invested in this technology, trying to gain the advantage through innovation over other companies. The U.S. also played a very significant role in terms of government policy and investing in research and development funding from government. That contributed a significant amount to where we are now with solar energy costs.  

So overall, we've seen that costs have come down by more than 99%. It's a really rapid cost reduction trend. Costs have been falling by about 10% per year. Lithium-ion batteries are the one other energy technology we've studied that have seen such a rapid cost reduction, but solar energy, it's so much cheaper than it was a couple of decades ago. And when we model this process, we can actually see what's actually driving the improvement at different levels. So we can see at the level of engineering design, how much of the cost reduction came from improving efficiency, how much came from improving different kinds of materials that are used.  

When we look at all of this at the global level, what we estimate is that about 60% of that 99% cost decline came from market expansion policies that then kickstarted a lot of private sector innovation. So it was really the private sector that was innovating, but it was incentivized by these government nudges. So there are many different types of policies within that category. That's one way to do it. And then about 30% of the cost decline came from research and development funding.  

And so the bottom line is we can use these kinds of insights from the past to be more strategic going forward about what we're investing in. So these kinds of analyses can also be applied then prospectively, looking forward in time. As listeners may know, the U.S. didn't sustain its lead in solar panel manufacturing. It wasn't able to sustain that lead, but maybe with these kinds of insights, the U.S. or other countries or companies could really find those mechanisms to invest in, you know, at the high level in terms of the policies, but also at the engineering level. What are the key mechanisms at the right time for different kinds of technologies? 

RS: So, let's talk about some of those technologies because, you know, there's a tremendous amount of innovation going on in the United States—not necessarily in solar and wind, although there's still quite a lot going on in both of those as well, but in other areas like fusion, magnetic fusion, inertial fusion, nuclear fission, the development of small modular reactors and advanced reactors. Is it possible that in a funny kind of way that if we make solar and wind too expensive and batteries too expensive, things that we tend to import, does it tilt the field in favor of those other advanced technologies and maybe give us a whole new kind of advantage? Am I just being a mindless optimist or is that even slightly credible to you?  

JT: Some of the technologies that you mentioned are at very different places in terms of their development.  

RS: Yeah, years, and years before we have fusion reactors.  

JT: Yeah, yeah. And I mean, I think that there's still a lot of opportunity to innovate in terms of manufacturing, but also the installation, the deployment, the business models in some of these technologies that are already growing rapidly at the global scale. Solar energy, storage, solar coupled with storage, solar and wind, heat pumps, electric vehicles, electric vehicle batteries, you know, the charging infrastructure for that could be improved. There's a lot of room for innovation in home energy systems. Demand-side management is an area that I'm interested in and that's been around for a long time, but there's many new ways to do it. I think that the U.S. and other countries that are not currently dominating in these industries can still do a lot within these industries that are growing, that are becoming rather established. They can still participate. I also understand Rob, the reason for mentioning some of the examples that you did. So for example, there's a lot of work happening on nuclear-fission-related innovation in the U.S. and, you know, I just think though that it's risky to step away entirely from the technologies that are already expanding. And to not make an effort to take part in that transition.  

You don't necessarily want to try to do what other companies, if you're a company, or what other countries, if you are a country trying to incentivize innovation, you don't necessarily want to do what others are already doing really well, but you can take a part of that transition and you can try to innovate there. And I think you have to do more than simply prevent those products from being affordable in the U.S. because at the end of the day, to be competitive in this space, you don't wanna just think about what is within your national borders. You need to be able to sell to other countries. So the strategy of saying, okay, let's step away from these technologies that are produced by others, let's step away from those industries, and try to only support other ones that are less mature and then sort of hope we can catch up overnight and grow these new industries, simply adopting policy in the U.S. and focusing on the U.S, I think it's not really a winning strategy. It's certainly very risky. I think we have to take that global perspective.  

You mentioned fusion, for example, which is earlier in its development. And certainly that can be part of the portfolio. So the goal you have as a country versus a company is quite different, but both want to be thinking about technology portfolios, how you're investing in technologies. You can't invest in all of the above equally because it's usually not a good strategy. You want to see which ones are maybe a little bit more likely to succeed or a lot more likely to succeed, but then you also want to invest in some technologies where the future’s very uncertain but it could be wildly successful if things go the way you want, if things go well.  

So you wanna have that balanced portfolio. And if you have the resources, you can include riskier technologies, earlier stage technologies. You could include nuclear fusion in that portfolio. But my point is that you also want to continue to keep those markets that are already growing, the technology serving those markets, you wanna keep those in your portfolio as well. You can try to offer a better service. You know, for example, a more robust home energy system, a faster way to complete construction projects, that sort of thing.  

KM: So, I mean, I wonder if there's a technology or two that if you are looking forward, now this could be looking at the U.S., it could be thinking about globally, but that you find particularly exciting and that you feel like is a game changer.  

JT: So I think what I'm excited about is actually the fact that we have more tools that we can use to provide this high-quality energy service. You know, one that's cleaner, that causes less negative health impacts, one that's more convenient, one hat hopefully will provide more affordable energy, including to people around the world that don't have access to modern energy services. But even within countries like the U.S., people don't always have access to high quality energy services, reliable energy. So there's just more tools that we can work with.  

So what are some of these additional tools? I've already mentioned a number of times solar and wind energy, but just to come back to that for a moment, one of the nice things about solar and wind, other than the fact that their costs have been coming down, is that they're available in many parts of the world. So you have this resource. If you combine the right amount of solar, the right amount of wind for a particular place, you know, if you're strategic about those percentages, you can get a pretty reliable resource.  

So if we take Scandinavia, for example, this is a region, and actually let's talk about the Nordic countries, which includes a couple more countries. This is a region that has nearly reached complete decarbonization. So almost entirely removed CO₂ emissions from its electric power system. And so what are they using for generating electricity? They are using nuclear energy, they're using hydropower, they're using some biomass, they are using solar, they're using increasingly wind energy. So they've created this low-carbon electricity system. Not every country has those same resources, but my point is that depending on where you are, you can sort of mix and match these different resources.  

So with these additional tools, with these large energy resources, and then also with the ability to use electricity for more services, you know, you can use electricity now for much of your transportation if you want, you can use it for your heating through heat pumps, these give you more options as well for getting rid of carbon emissions for developing that reliable energy system. So I think it's really that combination that I'm excited about, but there's also other aspects I think are encouraging and sort of interesting to see where things go. Demand-side management is one technology that we need to look at. It's been around for a while, but there's potentially the need for much more demand-side management. There is potentially more economic value in demand-side management going forward because you have these fluctuating energy resources like solar and wind. And you have potentially new demands for electricity coming on board. So you want to manage those fluctuations on the demand and supply side.  

So what is demand-side management? It's basically saying that you're shifting demand around in time and space. So you are maybe charging your vehicles, and with demand-side management, you could be incentivized to charge your vehicles during certain times. And we published a paper a couple of years ago looking at simply installing workplace charging. What we see is that simply by installing those electric vehicle chargers at the workplace, the effect can be very significant in terms of smoothing out the peak in electricity demand in the evening, kind of flattening what would otherwise be an evening peak. That's great because you can save on building power plants. You know, by putting your chargers in the right place you may not have to build as many power plants. And you could also use what will otherwise be solar energy that may be wasted on the grid. So that's an example of demand-side management. And in that case, you don't even necessarily have to ask people to change their behavior or charge at certain times. It's really just about putting the physical infrastructure in the right place. You're putting it where people are naturally parking during the days. And so that's one example, and there's other examples of demand-side management too. You could incentivize using a bit less energy at certain times. That's not very popular. But what we're seeing now is that some companies are saying, Oh yeah, you know, if you're willing to pay us that much, because actually the savings to the system can be very significant, they're saying, okay, if you’re willing to pass that much, we're more than willing to make our demand, some portion of our demand flexible. I mentioned heat pumps before, you can use what are called “air source heat pumps” where you're drying from the air so that the air is sort of providing your temperature differential, but you can also do water source heat pump, you can do ground source heat pumps and these water source, ground source heat pumps can be advantageous in certain climates where you have very cold temperatures. So this is another exciting area of innovation. Geothermal overall is an interesting area. There's so many different areas, and so I think there's definitely a need to diversify the portfolio, but still I think it really does make sense to stay involved in those industries that are currently rapidly growing and find ways to innovate within those.  

KM: Sounds like instead of a specific thing like wind, you're kind of excited that there's a new way of maybe looking at grids and making them work with all the tools in the toolbox.  

JT: Yeah, exactly, grids and just using electricity for more energy services, you know, that's something that we talked a little bit ago about these trends. And that's a trend that I think is also difficult to stop because there's just an inherent advantage to being able to plug in and power something. You know, if we were charging the batteries here on our devices using a generator, and we had to think about going and refueling or getting fuel for that, it's just inherently less convenient than if you can just plug into a grid, an electric power grid. It's cleaner, it's easier. So I think that is a change that I'm excited about and I think it's gonna continue. It's one that's difficult to stop that transition.  

RS: So where does the momentum or inertia come from in these trends? And a lot of what you do in your research relates to trends and how things are going to be in the long-term. A lot of that comes from cost, but how much of it comes from trying to capture foreign markets, trying to head off climate change, trying to create new industries, trying to employ people. I mean, you know, there are many other things that ultimately bring innovation to the market besides the brilliance of the initial invention. I think you call those elements of the structure of practice in some of the things you've written. So the question is where does the momentum come from? What incentivizes people and what do you think is preserved as we go through this maybe short period of change, maybe a longer period of change where we're trying to repatriate manufacturing?  

JT: At the end of the day, it's about people working and getting paid to innovate in these areas, right? That's where it starts; it starts with people. And if we look back over the last 40 years, there's been a lot of innovation in batteries and solar energy, wind energy. Before that, as I mentioned at the outset, we've had a lot of improvements. The first coal-fired power plant started out at 3% efficient in 1882, I think, in the U.S. and that efficiency has increased substantially to above 30% efficient on average, I believe, in the U.S. right now. And it can be higher than that.  

So, you know, it's like, where did these changes come from though, right? What actually has worked? And I think we'd be hard pressed actually to find a historical example of where tariffs or sort of preventing the flow of knowledge embodied in technology has actually worked in terms of benefiting new industries, new job creation, technological innovation. I mean, there may be an example, but one isn't coming to my mind based on the work I've done.  

And so what has worked is to invest in these changing industries, these changing technologies, right? You had Denmark that adopted a lot of wind and their kind of homegrown industries benefited and innovated and took a lead in terms of both wind turbine manufacturing as well as wind installations. And China of course has also since then grown and really dominates in that area as well. But, you know, Denmark's a tiny country and they're still playing a major role.  

So the examples that come to mind of these strategies in terms of growing industries, innovating, growing jobs have really been about investing in change and seeing what innovation comes out. You know, there are signatures in the data of where things are going. The future is never perfectly predictable, but I think ignoring the signals in the data and sort of trying to step back, trying to reverse these processes of technological change, is more likely to result in actually not staying at the forefront. So, this isn't exactly what you were saying, Rob, but I think it relates to it. And I do agree, we want to diversify our investment portfolios. And invest in a range of technologies, but we also want to place bets based on the information we can get in data. And actually technology change is not as unpredictable as some people think.  

KM: You have created data models that sort of look ahead at, you know, what the rate of adoption might be, how fast we can go, that sort of thing, give me an example of what you're seeing and how you think about like what those models may be telling you.  

JT: There's a couple of things we learn, I mean, one is that, you know, when we started looking at a range of technologies, including energy technologies, what we saw was that actually Moore's Law is not just for computers. So we found and published a paper in 2013 showing that many technologies follow exponential improvement trends. And that simple models like Moore's Law and something called Wright's Law actually have some predictive power in terms of giving us some information about how technologies are gonna change in the future. Those predictions are never perfect and you always want to estimate the uncertainty associated with those predictions, but it's a good starting point. So that's one thing, is that you can use these data trends to get a sense of how things are changing.  

The other thing that we've looked at is ways in which you can identify specific aspects of technologies, what we call mechanisms of innovation to invest in, that may bring about the effect that you're looking for. So you can actually model a technology, look at how all of its different interconnected parts might be able to change. And then use that information to inform what bets you're placing. Should you invest in greater efficiency or using less materials? Or even at a higher level, is it more about research and development investment? Or is it more about scaling up manufacturing? Or at the policy level, should we invest in research? Should governments invest in research for this particular technology? Or is it more about stimulating markets and competition in the private sector among companies that are gonna improve these technologies.  

So it really doesn't have to be entirely guesswork. And that's also where I come back to, regardless of your politics, it's like, let's take this information that we have about how technologies change, which ones are changing, trying to pull back from electric vehicles, trying to pull back from solar, wind, from energy storage. It's again, it's just risky if you look at what the data is telling us. I'm not saying you can't slow it down or make things more noisy, more volatile, but I just don't think it's gonna be a successful strategy in the medium- to long-term.  

KM: I wonder how you're feeling about the future, and we've talked about a lot of things. Your work has this sort of predictive element to it. What do you think when you think about the present, but also think about what's ahead? And I wonder what your level of optimism is.  

JT: Yeah, I mean, I think I focus less on am I optimistic or pessimistic, but more on like, what is it that we want to try to achieve with the future of technology. Sometimes it's about understanding what is it that people need. We touched briefly on electric vehicles, which is something that I've worked a lot on, so maybe I can give an example from that, you know, there one of the questions was, okay, well, given how people are driving, how they're moving around in their vehicles, what would they want their batteries to do for them? How good do batteries need to be? Where would they what the chargers to be, and actually what we found was that, you know despite the fact that everybody's driving differently—some people drive much more, some people drive much less, and if you wanna cut your carbon emissions, best thing to do is to walk and next best is bike and then public transit and so forth—but if you look at how people are driving and despite the fact that people have very different patterns in how they go about their day, what we find is that there are ways in which you can really shape the technology toward meeting all of these people's needs.  

There are key locations where you wanna put chargers. And if you invest in chargers in those locations, you're gonna meet the needs of maybe 80% of people. If you randomly put charters in shopping malls and other public locations, you might meet the needs of about 5% of the population. And that's just one example. That's just electric vehicles.  

We saw that, okay, if you improve batteries by this much, that helps, but there's what's called a heavy tail. Everybody has these high energy days. So there you need to put chargers, fast chargers in certain locations along highways and long rural roads to make sure people can charge when they need to. So that's just one example. But my point here is that we can also study how it is that people are using technology, what it is they need, and then another way to shape innovation is to direct development toward those goals.  

RS: So, Jessika, how do you think about, you know, going from early-stage inventions to deployed inventions at scale and what that process looks like and what do we lose with tariffs? Do we lose any of that? Does it change?  

JT: Yeah, I mean, I will say that with the innovations successes that we've seen, you know, solar energy costs coming down by more than 99%, wind energy costs falling exponentially as well, battery costs improving, you know what you see is that part of what drove these technologies forward was competition at the global scale and companies competing with one another to gain market share globally. That's a larger market that's gonna drive faster progress. So I think that's really important.  

The other thing is that, you know, if we think about what does it take for a new technology to get invented, to see innovation, which then involves the adoption or the use of that technology, it normally takes, especially in this space of energy, what we've seen is that it takes some early government policy interventions and some investment in research, but also some incentives for companies to grow markets and for consumers to purchase these technologies. And then you kick off this process of competition. And like I said, we've seen that happening at the global scale.  

And the interesting thing that we see from our research is that you can actually be pretty strategic, I think, in what mechanisms you invest in in this process. So what do I mean by that? What I mean is that you can take a technology, like say solar cells, and say, okay, if I improve these materials and this, you know, improve the efficiency by this amount, how much might the costs come down? And then from there, you can say, okay, are we at a point now with this technology where we should invest more in research or more in scaling up manufacturing to get economies of scale? And then from there, you can actually go all the way up to a policy level and say, okay, how much do we wanna balance our investments and in growing markets and stimulating private sector competition versus investing in government-funded research? So this can be a pretty deliberate process. There's always uncertainty, but I do think that there's ways to actually be strategic, be deliberate in how these investment decisions are being made even more so and much more so than we've done in the past.  

RS: Sort of really interesting points, you know, with globalization comes global scale. And so, yeah, much more intense competition, comparative advantage for some, who wouldn't otherwise be able to provide these things within their own markets, but also the scale of market and the opportunity to achieve economies of scale. Those things we lose if we deglobalize or withdraw and it would be very hard for us to bring new technologies to market solely within the domestic market of the United States. They'd be inferior technologies, they'd be more expensive for a number of reasons.  

JT: That's a potential outcome and a very scary one.  

KM: You have done research looking at how we take sort of high-level, very broad targets and goals for the climate and then translate them into performance targets for specific technologies. What did you find out in doing that work?  

JT: Well, so two examples, I think, were some of the early examples of this work, and then we've extended it to other cases. But one was the case of stationary energy storage, because there was a widespread recognition that we wanted to bring, we, meaning people working on energy storage wanted to bring the cost down of energy storage—you know, like batteries, for example, that are stationary, so they don't have to be lightweight, they can have higher volume—there was a widespread recognition that cost should come down, everybody knew that, but the question was by how much should the cost come down in order for this technology to really make a difference? And so that was one question that we posed.  

And, you now, we started looking at an application of storage where you're using storage to do what's called price arbitrage. So you're selling electricity when the price is high and then you're storing that solar and wind energy when the price is lower. But we were able to arrive at a cost target for storage. And then we ultimately looked at, okay, if you wanna be able to meet all demand with solar and wind or just 95% or 90%, what should the cost of that energy storage be? And there's always uncertainty, of course, but we were able to zero in on a quantitative target that then has been helpful in comparing different technologies coming out of startups, for example, or more mature companies that are trying to bring their costs down. So at the time we found that you wanna see a further, if you really wanna get to really cheap energy storage that can support very large amounts of solar and wind and even getting close to a completely solar- and wind-dominated power grid, you wanna bring battery costs down an additional maybe 80 to 90% relative to where they are now, which could be doable. I mean, it remains to be seen.  

So that was one example. And then the other one was looking at electric vehicles and saying, where would you put, well, first of all, how much do you want to improve the energy density of electric vehicles? And then where would put chargers, electric vehicle chargers? What power should they have? That involved looking at how people are driving around in their cars, where they're stopping, for how long. We were able to identify some locations that were predictably really high payoff in terms of supporting people's need for charging without causing delays, without causing them any inconvenience.  

KM: Really understanding who your consumer is and what they need, because if you don't meet those needs, then… 

JT: Right. And I think with the stationary storage, it came down to patterns in the solar and wind resources. What were the fluctuations in those resources? What were the characteristics, the characteristic patterns there? That's what ultimately defines that storage cost target.  

RS: Sounds obvious when you say it, but, you know, people really don't do these analyses. And very often we're aiming for a target, which is, you know, merely cheaper than it is now, but it may be taking a direction that's still completely untenable. I won't name technologies, but I mean, some that are near and dear to my heart have that tendency.  

KM: Jessika, thanks so much. This is great.  

JT: Yeah, thanks so much. This was great.  

KM: What if it works? is a production of the MIT Energy Initiative. If you like the show, please leave us a review or invite a friend to listen. And remember to subscribe on Apple Podcasts, Spotify, or wherever you get your podcasts. You can find an archive of every episode, all of our show notes and a lot more at energy.mit.edu/podcasts and you can learn more about the work of the Energy initiative and the energy transition at energy.mit.edu. Our original podcast artwork is by Zeitler Design. Special thanks to all the people at MITEI and MIT who make this show possible. I’m Kara Miller. 

KM: Thanks for listening.