Governing the power grid

We've talked about the power grid on this podcast a couple of times before. We talked about the history of the power grid and then we talked about how it feels, how blackouts occurs. Well, today on the show we're joined again by Set Blomzak. So Set is a professor of energy policy and economics and international affairs in the Department of Energy and Mineral Engineering at Pennsylvania State University. He's a co-director of the Payne State Center for Energy Law and Policy and he's an external faculty, the Santa Fe institution. So in this episode Set is going to talk about how the power grid is governed. We're going to take every complex physical system and we're going to work out how do you govern that to make it reliable and competitive? Because the really interesting thing is the grid was governed in one way for the first 100 or so years of its life and then we changed it. We changed it to a more competitive model. Now this has done some really good things. It's really allowed the arrival of green energy to be fed into the grid and Set's going to talk about that. But in the end he's going to ask a key question. Have we actually met the grid more reliable and have we actually met it more competitive? This is simplifying complexity. A podcast where we explore the underlying principles of complex systems. Systems that seem to defy our rational view of the world. Like economies, ecologies, or even you or me. I'm forensic engineer Sean Brady and I'll be your host. Set, welcome back. Thank you for having your back Sean. So you've been on before. We've talked about the grid and the history of the grid and how the grid works and then we've also talked about blackouts as well. I'm particularly trying to model blackouts and understand blackouts and cascading failures and we talked about the facts that blackouts follow power laws as well. This time we're going to change tech and we're going to talk about something that you alluded to in the first episode, which is governance, the rules by which we decide to run the grid. What got you interested in governance? Governance is like the social side of electrons. We often think about the power grid as being a complex engineered system and it certainly is. It's an engineer marvel. It has all of these complex physical behaviors. I mean, it's also a very highly social system. It looks the way that it does because people or groups of people make decisions about it. And so the governance of the power grid, which really encompasses who is it that is making decisions about how we plan and operate the grid, how are those decisions made? Whoever's making these decisions, how do they make these decisions? And then the aspect of this, which is really important, is how do these governance decisions influence the way that the grid actually behaves? How do these governance decisions influence the rate at which we can absorb renewable energy onto the grid? How do they influence the cost of electricity that people like you and I have to face? This is an area that I got interested in thinking about this sort of through some work that I had started and have continued through kind of a research group called the RTO governance network. And just to remind your listeners, RTO is a sort of US term of art. It stands for regional transmission organization. The regional transmission organization basically is this creature that was created through the process of electricity deregulation in the United States. And the regional transmission organization is basically in charge of planning and operating the grid over some large area. But just as sort of a reminder, one of the sort of tricky things is that the RTO has this social responsibility to keep the lights on, but it doesn't own anything. It doesn't own power plants, it doesn't own transmission lines. And so it has to come up with mechanisms or incentives to get the owners of the actual physical infrastructure basically to supply services. And so a lot of the ways that the RTOs do that is through markets. So can you just give a very quick version of how decisions used to be made about the grid for the 100 years or before they started the change. How did governance work and who made those decisions? So for nearly a century, the electric utility industry operated on this model where the utility and in the US, that was kind of a mix of publicly owned and privately owned utilities. And I believe almost every other country in the world, the utility was an arm of the state. So the utility for almost 100 years was really in charge of the grid. And they were the ones that would decide kind of which power plants they were going to turn on at different times in order to meet demand. They were the ones who would decide, you know, whether they were going to build a new power plant or a new transmission line. And they basically made these decisions, basically a kind of under the watchful eye of their regulator, who was concerned about the utility providing affordable and reliable electricity service. That's basically the model that the utility industry used for, again, almost a century until the United States and other countries kind of started this process of deregulation and kind of changing the way that the electricity industry operates. But you said that that's almost was like a risk-based way of managing the grid. In other words, the regulators watching utility say, and I want to do XY and Z and the regular might say, well, that the air could compromise reliability. So we're going to be very careful about doing that and think about it. So is it fair to say reliability really was the primary driver? What cost there as well? But reliability was key. I think that is fair. And it was really about, you know, achieving the reliability goals at the lowest possible cost. And so if there was a trade-off between one and the other, you know, reliability wins almost every time. But the utility, you did have this mandate from the regulator to meet that level of reliability in the most affordable way possible. So then the change came. And again, we briefly touched on this in our first episode, very quickly, what drove the change and what did the early changes start to look like? Or the early experiments in change start to look like what really started to drive the need for some kind of reform of regulation in the electricity kind of was largely economics. As we discussed in an earlier episode, a lot of concerns around large electricity users, right, kind of particularly industry, that electricity rates and energy costs were becoming basically uncompetitive, you know, where you had industry that had to increasingly compete on a global basis that electricity costs were too high. The utilities didn't have the right incentives to keep costs down despite the mandate from the regulator. And that their customers were being asked to shoulder the economic burden of bad utility investments. You know, we had talked before about the cost overruns at nuclear power plants. And that's kind of one very prominent example of this. And so that was really kind of what started the push for reform. And a big part of that reform, you know, was really trying to replace, you know, some of that heavy hand of the regulator with the invisible hand of competition. And so the idea was that rather than having the utility be it basically be in charge of everything, right, be in charge of all the power plants and all the transmission lines and everything, was that we had this kind of robust enough transmission grid that we could allow the owners of power plants in different locations to basically to compete with the utility to sell power using the transmission grid as kind of a platform for competition. But for these markets to be successful, you needed a couple of things. One is you needed that transmission grid to be sufficiently robust to be able to move power around easily in a kind of a bunch of different ways depending on where the economics would drive you. Okay, you needed enough competitors, right? You needed enough power plants and different power plant owners so you didn't have a monopoly. And then the sort of other critical thing that you needed was you needed that platform for competition, the transmission grid to be operated totally independently of the power generation. Okay, and so this sort of old model where the utility owned both the network and the power plants on the network, you know, the concern was that it was that basically the utility would restrict access to the network to favor its own power plants and not those of its competitors. The utility would erect all of these barriers to entry and, you know, other non-competitive stuff. And the solution to that, which was adopted with some gusto in the United States, but United States is not the only place, was really to break apart the utility, not in terms of breaking it into a bunch of smaller utilities, but in terms of breaking it into like a transmission company that would own the wires and maybe a separate a corporate generation company that would own some of the power plants, and you would basically set up these independent organizations, right, called the Regional Transmission Organizations in the United States, whose job it would be to basically to run the transmission grid that was actually owned by the transmission utilities. So these utilities own the transmission lines, and then basically it's up to the Regional Transmission Organization to do the planning to basically be like the traffic controller for the power grid, and the Regional Transmission Organization was supposed to basically act in an economically and technologically non-discriminatory way. It was supposed to say, "Whose ever power plants are the cheapest? I'm going to run those." Whoever can build a new power plant in the places that we need it at the most competitive price, they're going to get to build it. It's not necessarily the utility, it's whoever is going to be the most competitive. So how in that model do you satisfy yourself that you're going to have reliability? This is where it starts to get complicated, right, because the Regional Transmission Organization it has a regulator and it has to make sure that the grid is reliable. But remember, it doesn't own anything. And so it basically has to come up with mechanisms to keep the grid reliable. And in the spirit of harnessing the benefits of competition, a lot of those mechanisms are market-based. And so it's going to be a really hot day and electricity demand is going to get really high. The Regional Transmission Organization, this grid operator basically has to let the market price rise in order to provide a financial incentive for more power plants to say, "Okay, I will be ready to generate electricity." If the Regional Transmission Organization looks ahead a certain number of years and says, "Oh, well, we think that electricity demand in some area is going to be higher in a few years," it has to use some kind of a market mechanism to provide incentives to build new power plants. Do you want to talk about some of the different models or versions of this that were tried and how they worked out? So sort of one of these things where these Regional Markets, they actually have a lot of similarities. And they sort of have a lot of similarities because in these initial days of electricity deregulation, there are different kinds of market models that were tried and some of them failed rather spectacularly. So before you and I had talked about the great failure in California. And so the model in California was not replicated anywhere because it was such a big failure. And that was a model where you had the grid operator running the market. You had a lot more distributed decision making. So basically the grid operator would essentially ask power plant owners to give the grid operator information about where they wanted to generate electricity and at what price and then the grid operator would kind of mash all of these things up and kind of try to figure out who is going to generate electricity. And that didn't work very well because it turned out to be really easy to manipulate. So fundamentally you're not saying that the price of electricity would go up and down and they're incentivized to provide or are back off depending on where that is. It was sort of like trying to run a collaborative arrangement where you were asking the parties to tell you what they wanted to do and then you were trying to sort of make that all fit together in a way that worked. Yeah, a little bit. I mean, it was a lot more like the California model was a lot more like what you would see in like a commodities market like trading futures for natural gas or something like that, right? Where you did see buyers and sellers responding to prices as you know, on a hot day demand is going to be high. And so the price is going to go up. But you know, you really sort of relied on them kind of coming together in a very decentralized way. It's maybe sort of easier to understand and kind of in contrast to the model that was basically developed in my home state of Pennsylvania, which turned out to be much more successful and has largely been copied by most of these kind of other electricity markets, which is basically that the grid operator instead of basically trying to run like a stock exchange for electricity holds auctions. And so if it's going to be a hot day tomorrow, the grid operator says it's going to be a hot day demand is going to be high to the power plant operators out there, you know, make me an offer to supply electricity tomorrow. And all the power plant owners will submit offers and then those offers are like bids in an auction. And then it's the grid operator that decides who wins the auction. Technically the term we use is who clears the auction. So who clears the auction and is scheduled to produce electricity the next day and who doesn't clear the auction. And presumably it's the cheapest power plant. Is it that simple or there's more to it than that? That's basically the idea is that on days when demand is low, it's really only going to be the cheapest power plants that are going to operate and the market price is going to be really low. On days when demand is really high, you're going to need those more expensive power plants. And so the market price is going to rise. What happens if you don't deliver the next day? Yes. So there is a they're not getting the market price for whatever electricity you were going to sell, right? There are penalties for basically non-compliance because then the grid operator basically has to scramble to go find more electricity. I'm just really keen to understand California then and why it was different. So what was the fundamental difference between that and California? The fundamental difference was basically the degree of control that the grid operator had over the market outcome. So in the California model, basically the grid operator would basically take kind of the individual production preferences of all of the power plants and try to aggregate them together in some least cost way. And that's very different than the grid operator holding like a very centralized auction where there's basically there's one pot. Everybody throws their offer into the pot and then it's the grid operator who decides, okay, you're going to produce this much. There's like a very centralized single point of kind of determination of who's going to produce electricity. As opposed to like all of the individual suppliers making decisions and then the grid operator tries to patch them together. Where did the California one go wrong in that sense? What did failure look like? Some people just simply not deliver or did they change their mind or what did failure look like in there? I mean, it looked like a lot of things. So failure sort of arose because this market, which was very complicated to administer, once you sort of figure out how it worked, it became very easy to manipulate because another thing about the California market was that it was sort of decentralized to the point where it was a little bit divorced from like the actual physical system. And this is maybe easiest to illustrate through an example, okay? So the other thing is that the state of the market is that the state of the market is so there was basically a company playing in the California electricity market called Enron. There became very famous. We talked about this previously for manipulating the market. So what Enron did was they said that they wanted to produce a lot of electricity from this power plant and move it over this little rinky dink power line. It's sort of like a little bit like trying to put 10 pounds in a two pound bag. They said we're going to generate a lot of electricity and we're going to move it over this little rinky dink line to some electricity customer. And the California grid operator said, if you do that, then you're going to overload the line. And Enron said, yes, we will. And then you get to pay us to not overload the line. I can see what you mean by divorce from reality. Right? And so, you know, where we're sort of divorced from this sort of way in which it was divorced from reality was that there was no rule in the California market that said you could not basically that said you cannot schedule more electricity on this line than we say that it can carry. You can't put more pounds in the bag than we say that you can fit in the bag. And this was turned out to be one of the other advantages of the very sort of centralized model, which was that centralized auctioneer could basically by itself optimize power generation over the entire system. Because it knew which power lines could carry a lot, which ones were little rinky dink power lines. And after it got all of the bids from the suppliers, it was a lot of money. It would basically optimize those bids so that you met demand reliability and didn't overload the system. Because it still vests all of this control in the centralized grid operator, it's a little bit less of a free market than the California model, but it turned out to work a lot better. And is that the start of the standard model we see all over the world now before doing it? Again, different regions have different ways of doing things, but kind of by and large, this model of having kind of a very centralized grid operator that really runs the market and optimizes the system, that by and large has been copied very widely. What's interesting in what you say and said is that, or what's coming from it in my head, is that you've got a complex system and a complicated system in terms of the grid from a physical perspective. And one of the things we talk about complex systems is no central controller, but this isn't essentially an attempt to put some form of top down control to achieve reliability. Is that fair? Yeah, I think that is fair. And it's both to achieve reliability and to make sure that the outcome is as economical as possible. And you're getting at kind of one of the big questions that comes up in how we run the power grid, which is becoming a very pressing question as we get more and more distributed power generation sources, you know, people putting solar panels on their roofs and neighborhood solar panels and people potentially selling electricity to the grid from their electric cars. There is this very fundamental question, which we haven't resolved yet, is to how decentralized can we really make control of this system? On paper, it sounds like maybe we could, but in practice, that's been very difficult. What do you mean by that, Seth? Why do we mean by it looks? So I can see by adding more green, essentially, you start to decentralize the production. Why does that look good on paper, but then in reality, it's problematic? Well, I mean, on paper, you can sort of theoretically see how such a system leads to some equilibrium. The challenge in sort of porting that to the power grid is that the grid itself is kind of a constantly dynamic thing. And there isn't one equilibrium. So, you know, we can optimize for some operating point. And then immediately we're going to have to change it again, because the grid is like it's, you know, demand is constantly changing. It's a really dynamic system. And being able to get decentralized agents, like decentralized decision makers, to consider all of the ramifications of their actions on this highly complex dynamic system. I mean, again, on paper, there are conditions under which you can kind of mathematically prove that this works. In reality, that has been much more difficult. It's in part because, you know, I mean, the grid has all of these emergent behaviors that are like really hard to predict. In our last episode, we talked about blackouts and we talked about influence maps. You basically say in there that you can get more decentralized, yes, but because you're not an equilibrium, you still need some entity. There's someone who's got some measure of control over the grid to be able to say, look, I know if you really important influence maps here, and I know that I have to pull a couple of levers to keep those, to manage those influence maps. So we're going to get into cascading failures and all sorts of issues. Yeah, we are still at the point where in the real world, we need some centralized agent operator that has that kind of whole system visibility. Again, in the real world, we haven't figured out a way to tell like, say, the owner of an electric car, like you really shouldn't put all this electricity into the grid right now because it's going to like mess with things like a hundred miles away. And we talk a lot about self-organizing systems. Does that mean that the power grid just still has a fragility to it that makes it not quite self-organizing? It's just vulnerable. I think the way that the grid is currently configured, I mean, I think that's basically the case. How would we change it? It's a good question. I think that there's sort of a really fundamental question, which I don't have the answer to, as to how much you can change it. And part of that kind of has to do with the fact that there are like undesirable versus undesirable equilibrium states in the power grid. And you would sort of basically, you'd need kind of some mechanism for very rapid information exchange, right? So the person with the electric vehicle knows that if they do something, they're going to disrupt the system a hundred miles away. But you also, you also need us basically kind of a system of like penalties and rewards. Kind of all of these distributed agents can see and understand is basically going to be able to capture all of these different potential effects on the grid. And the second is that, you know, to have these systems of penalties and rewards be aligned with what is the ultimate system goal, which even in the deregulated environment, the ultimate goal is still reliability. And yeah, very simple question to ask, but I suspect a very difficult question to answer. Has deregulation worked? Has it achieved what the objective of it was? You're right. That's a very complicated question. I think that deregulation has done some things well. And in some cases has not lived up to its promises, right? So one of the promises of deregulation was that these competitive forces would sort of really radically drive energy costs down through basically through competition. And that I don't think there's a lot of evidence that that has happened. On the other hand, I don't think is a lot of evidence that deregulation has really worsened reliability. So we've been able to maintain a reliable system. And we have kind of have replaced the utilities level of control with these markets that are, you know, we're supposed to deliver these competitive benefits. But my own view is that I'm not totally sure that they have, right? So what exactly did we gain? I think, honestly, I think, I mean, in terms of the, to go back to grid governance, how we make decisions about kind of what the grid looks like, I think one of the most kind of actually effective things about deregulation was rethinking this idea of the transmission grid as a platform and kind of basically having this regional transmission organization be a non-discriminatory operator of this platform. And I say that because one of the things about these deregulated electricity markets is that they have actually opened the door to a much broader array of market participants than we had under the utility system. And I think it has created this environment where there aren't barriers put up to different kinds of technologies connecting to the power grid. I'm going to say that and then I'm going to step back on just a little bit. I think a contributing factor to the very rapid rise in wind and solar adoption that we've seen in some areas of the US grade in particular is due in part to public policy measures, right? Incentives to promote that. But I think part of that success also lies in our rethinking of how we run the transmission grid because now it's running a much less discriminatory fashion and you don't have as much a single utility sort of trying to get in the way of these new technologies. Now I don't want to push that too far because within these deregulated markets, right, markets have rules and the rules of these markets vary from location to location, right? The basic idea behind the market looks the same in my state of Pennsylvania as a dozen California or New York or wherever. But there are different rules that kind of govern how these markets work, who is basically kind of how new players connect to these markets, like how they enter and start selling into these markets. And so these markets have rules and in some of these markets, the rules are basically made by what we call incumbents, by basically those market actors that are on the grid right now. And so naturally those market actors are going to be afraid of new competitors, right, or new types of competitors. And so in some cases, we've seen these sort of processes of basically developing the rules for the market. We've seen examples where those are really, really accepting of new technologies and somewhere they have really inhibited new technologies, right? So my colleague Stephanie Lennhart, who is a professor at Boise State University, has this really nice paper where she sort of compares how different markets have adapted to energy storage, like big batteries. there are real differences in not just in how well different markets have been able to integrate storage, but also in how they come up with rules for how energy storage should participate. Because a battery is different than a power plant. And there are going to be different rules for how a battery connects and how it participates in a market. Rules are presumed to have sort of yet to be really worked out and tested. And some places they are still being worked out. But in some places they have been worked out. And so in a place like California where these incumbents have a lot less power. So basically figure out what the market rules should be. If there's been a lot of success at integrating energy storage in my state of Pennsylvania, where those incumbents have a lot more power in determining what the market rules are, it hasn't been as fast. And it's not necessarily because energy storage has less value in Pennsylvania. Or technologically different. Or is technologically different, right? It's valuable in California. It's valuable in Pennsylvania. But the process of coming up with these rules has just progressed very differently in California. That's interesting, isn't it? The rules of the market have an impact. And I mean, this shouldn't surprise us on how technology is actually used and how it functions within that market. Yeah, and I mean, this is very important in the sort of in the deregulated marketplace, because it's not just about the rules of who can connect and who can. But remember that these markets create and destroy value for different technologies. It's not just about who gets to connect to the grid and who doesn't. And so the rules of the market are going to influence how valuable or not valuable different technologies are. And we've seen that kind of play out in the market rules in certainly different areas of the United States, right? There are some markets that value energy storage more highly, right? There are some markets that value renewables more highly. And so these kind of these rules and how they're made turn out to be sort of really important for what the sort really what the future of our reliable low carbon affordable magical power. So, this is going to be just like set. Thank you very much for being on the show. It's been an absolute pleasure talking to you. Thank you. Thank you, Sean. Thanks for listening to Simplifying Complexity. When we look at the key concepts of complexity science with expert minds from across the world. Concepts like emergence, self-organization, adaptation, networks, scaling, tipping points and much more. This podcast was produced by Brady Aeword and Wadland Creative. To make sure you don't miss an episode, be sure to subscribe to or follow the show in your podcast app. I'm Sean Brady and I'll see you in our next episode.