Shedding Light on Electricity Blackouts, with Severin Borenstein

Notable Quotes

• The power grid is not prepared for climate change: “One of the things that is happening—and we’ve seen this both in California, where we had an unprecedented heat wave, and in Texas, where we’re seeing an extreme cold—is that more extreme weather is creating larger problems in the system supplying the electricity. As we’ve seen both in California and Texas—not just on the demand side, but also on the supply side—[weather has] created this imbalance, and as a result, [has led to] controlled, rolling outages.” (5:13)
• Intermittent renewables require careful management: “We have to make a distinction between intermittency that is unpredictable ups and downs in renewables, and [intermittency prompted by] timing, which is that we know when the sun is going to set … Neither of those are problems that can’t be managed, but I think the state has not been precise enough about managing the timing of the sunset and resources we’ll need … This in no way suggests you can’t run a system with a large share of renewables, but it does suggest that you have to plan for it in a very granular way.” (12:51)
• Policy reforms can make power outages less likely: “The problem is that the way we’ve drawn the requirements doesn’t actually match the realities of production … We need to have more of a pay-for-performance system, so that when you actually show up in those critical moments, you get rewarded—but if you don’t show up, you get penalized. I think we’re going to move toward that—certainly after what happened in California last summer. I suspect [policymakers] will also be reevaluating in Texas a move to say, ‘Okay, we need to have an idea of exactly what resources are going to show up and how they’re going to show up, minute by minute.’” (20:58)

Top of the Stack

• “An empirical analysis of the potential for market power in California’s electricity industry” by Severin Borenstein and James Bushnell
• “Measuring Market Inefficiencies in California’s Restructured Wholesale Electricity Market” by Severin Borenstein, James B. Bushnell, and Frank A. Wolak
• “Capacity Markets at a Crossroads” by James Bushnell, Michaela Flagg, and Erin Mansur
• Under a White Sky by Elizabeth Kolbert
• The Box by Marc Levinson
• The Bet by Paul Sabin

The Full Transcript

Daniel Raimi: Hello and welcome to Resources Radio, a weekly podcast from Resources for the Future. I'm your host, Daniel Raimi.

Today we talk with Severin Borenstein, Professor at the University of California at Berkeley's Haas School of Business, faculty director of the Energy Institute at Haas, and member of the Board of Governors of the California Independent System Operator. As the state of Texas struggles to keep the lights on due to extreme cold, I'll ask Severin about lessons learned from California's blackouts during the summer of 2020. We'll talk about the cause of the outages, the role of renewables, and market reforms that could help reduce the risk of blackouts in the future. We'll also focus on how California's experience can help Texas and other regional electricity networks plan for a future with more renewable power. Stay with us.

Severin Borenstein from the University of California at Berkeley. Thank you so much for joining us today on Resources Radio.

Severin Borenstein: Thanks for having me on.

Daniel Raimi: Severin, we're going to talk today about electricity, which is very much in the news as everyone is thinking about what's been happening in Texas over the last week or so. Before we dive into electricity—and we're going to be focusing on California and the west coast of the country—can you tell us a little bit about how you got interested in working on energy and environmental issues?

Severin Borenstein: My history is sort of long and winding. When I was an undergraduate, I got very interested in economics and ended up writing my senior thesis on airline deregulation. Then, I was working on airline deregulation in Washington, DC in 1978, which was very exciting. From that, I started working on price discrimination, something airlines are good at. One day, I was at a gas station and noticed—this was in the 1980s—that they were selling leaded and unleaded gasoline for the same price. I went and talked to the owner about why and started studying gasoline markets. From there, I started studying oil markets. Then in the mid-1990s, I was made Director of the University of California Energy Institute based on my oil and gasoline work. That happened to be the year that California began its deregulation of electricity.

Within six months I had to go up to the state legislature and testify about electricity restructuring. That of course led into work on market power and electricity markets and the California electricity crisis in 2000, 2001. Then from there actually, the markets got very quiet in the mid-2000s just as California launched a very serious program to address climate change, which you've talked about on the podcast. I started working on cap-and-trade markets and environmental issues around energy. That's where I am these days.

Daniel Raimi: Fantastic. Your expertise spans so many different areas that it's really great to have you on the show. I'm frankly surprised that we haven't had you on before. As I said in the introduction, we're going to be talking about electricity and power outages today. It's at the top of everyone's mind. We're recording this on February 17th, as millions of people in Texas and elsewhere have been without power for quite some time. To get us started, can you help us understand the basics of what types of power outages can occur and what the drivers of those power outages are?

Severin Borenstein: Most power outages that we experience as residential and small commercial consumers are distribution line failures. The two most common reasons are a tree actually falling on or touching a distribution line, or a transformer on a distribution line blowing. Those are the cylindrical devices that are up on the power poles. When they get overloaded, they blow and knock out power. That's what we mostly experience.

Here in California, we've experienced different varieties of that over the last few years where they've actually been intentionally shutting off power to distribution and even some transmission lines where they were worried that those lines could start fires. We had what are called “public safety power shutoffs.” The kind that we experienced in California in August and has been going on in Texas is extremely rare. These are system shortages or sometimes shortages in local power areas where there just isn't enough electricity available.

This isn't a problem with transmitting the power. There's just too much demand, not enough supply. Those situations occur extremely rarely, typically less than once a decade, but when they do, they get a lot of attention. One of the things that is happening—and we've seen this both in California, where we had an unprecedented heat wave, and in Texas, where we're seeing an extreme cold is more extreme weather—which is creating larger problems on the system in supplying the electricity. As we've seen both in California and Texas, not just on the demand side, but also on the supply side, it has created this imbalance, and as a result, what are called controlled, rolling outages.

Those are done to avoid the more extreme outcome which we saw in 2003 in the Northeast, and at times before, where they don't get a handle on it. They don't cut demand fast enough, and the whole system overloads. When that happens, all the generators start shutting off in order to protect themselves from the frequency variation on the system. Then you get what's called a “cascading outage” where you can get whole states or larger systems having their entire power grid shut down. To avoid that, grid operators start requiring shutdowns in smaller areas in order to keep control of the system.

Daniel Raimi: Great. That's super helpful. We're going to focus today, as you said, on what happened in California in mid-August of 2020 when there were widespread rotating outages over a few days. The sequence of events that caused those delays, I've come to understand, is extremely complicated. We're not going to try to get into every detail, but can you give us a basic overview of what happened?

Severin Borenstein: First, I have to note that these really were not widespread rotating outages. It's interesting that these get so much attention even though they actually affected a small number of people relative to the whole system. The first day of the two outages reduced the total demand by about 2 percent on the system; the second day reduced it by about 1 percent. That's still a lot of people losing their power. For purposes of comparison, typically a California household loses power a little over once a year due to distribution line failures. These outages would raise that typical outage rate from around 1 to around 1.05 for an average household. This was not an enormous electricity event, nothing by the way like the public safety power shutoffs we've had where the power companies have been intentionally shutting off power lines in order to reduce fire risk, which have affected far more people.

Nonetheless, they are a real indicator of some of the problems that we're seeing around the country and around the world. In California, fundamentally there was a record hot weather and high demand, and we saw the highest usage probably California has ever seen. I say “probably” because we don't exactly know because there's so much behind-the-meter solar now that we have to sort of guess at how much in total customers were using. But based on how much behind-the-meter solar there is and what it was probably producing, these were probably the highest consumption days that we have ever seen in California. Demand was very high. We were keeping up with it as a system, though it was getting very tight. Even a day ahead, it looked like it would be very tight. Then the sun started to set.

Unfortunately, we were very dependent on solar at that point. When the sun sets, not just the grid solar starts to shut down, but the demand from customers starts to go up because that behind-the-meter solar, which was showing up as less demand from customers, also started to disappear. Now we were still set and looked like the system would make it through, but in one case on the first of these two outage days, we suddenly lost a power plant, a gas-fired generator. That forced us further down into reserves. I should note, it's not that the system runs out of power, because you don't want to get to that point. When you start to get close, the system operator has to say, “well, we don't have enough contingency coverage in case we lost another power plant or a transmission line. We need to institute rolling blackouts right now.”

We dipped below those required operating reserves, and when we did, the system had to call for rotating outages. Now people have talked about the role of gas-fired power plants, and we did lose a couple, although not more than you would typically expect in extreme heat and of renewables. On the second day of these outages, we did have a somewhat unusual wind event where wind power picked up a lot in the hour before the rotating outages occurred and then it dropped very suddenly. That was part of what pushed it over the edge. When I heard people talk about, “Well, it was the failure of wind power,” that's sort of like talking about the last batter in the bottom of the ninth striking out and saying, “Well, he lost the game.” It got to that point, and we were needing every resource.

When the wind dropped in that case, we had a problem. The last point is that usually in California when these sorts of extreme heat events occur it can bring in a lot of power from the rest of the West, and we do. California imports power all the time and is very dependent on trading power with the rest of the West and on net importing. Unfortunately, this was an unprecedented heat wave across the West. California was trying to get power from the rest of the western United States but there was less available than usual. The actual outages did not occur at the time of peak demand. This is going to be important in thinking about planning in the future. They actually occurred at the time of what's called the “net peak.”

When you net out the renewables, particularly solar, you have what's left over. The problem occurred not at the very highest demand of the day but a little while later when demand had gone down a little but solar power had dropped a lot as the sun was setting. That's when the system really started dropping too low into its reserves to maintain operation.

Daniel Raimi: That's great. You touched on this already, at least a little bit, but I'm sure our audience is aware that California gets a large and increasing share of its electricity from wind and solar. About 28 percent of the power produced within the state was from wind and solar in 2019. Wind and solar are intermittent. To what extent did the intermittency contribute to this problem?

Severin Borenstein: Well, we have to make a distinction between intermittency that is unpredictable ups and downs in renewables, and timing, which is, we know when the sun is going to set. Both of these played some role, though we did, as I mentioned, have that wind event where wind power ramped up and then rapidly ramped down. Of course, we ran into the problem after the sunset, which is a completely predictable event. Neither of those are problems that can't be managed but I think in both cases and particularly in the case of the sun setting, the way the state has been managing it has not been precise enough about managing the timing of the sunset and resources we'll need. That is something that showed up. This in no way suggests you can't run a system with a large share of renewables, but it does suggest that you have to plan for it in a very granular way.

Daniel Raimi: Great. I know we're going to come back to that in just a moment. Before we speak to that planning issue, there's one question that I wanted to ask which is in informal conversations with fellow energy nerd friends. I've heard some of them speculate that certain generators in California switched off or ramped down their power production with the intent of increasing prices, sort of driving a price spike that would enable them to gain higher profits even though their production would be a little bit lower because those prices would really spike. Do you know if there's any merit to those claims?

Severin Borenstein: That's something one worries about. In fact, Jim Bushnell, my longtime co-author, and I have been working on market power and electricity markets since the mid 1990s. We wrote a paper before the California electricity crisis warning of this problem, and then during that crisis, pointing it out. I'm not shy about pointing out market power issues, but the best evidence is that's not what was going on here. In fact, the independent department of market monitoring, which is technically part of CAISO but operates completely independently has done their own study. They found that prices were no higher than they would have been under a scenario with no exercise of market power. Essentially, there is no evidence that the prices were higher than they would have been.

The very high prices we saw were a result of true scarcity. That's how the market should work when a commodity gets scarce, the price should go up. We do need to worry about whether that's due to somebody reducing supply, but in this case, there's no evidence that that was the cause.

Daniel Raimi: Great. Severin, you just used an acronym that we haven't defined yet. I realized now it would be useful for our audience to tell them what does CAISO stand for and what does CAISO do?

Severin Borenstein: CAISO is the California Independent System Operator. I'm on the Board of Governors of it. I should say that nothing I'm saying represents their views. This is the grid operator. There's one in Texas. The listeners have probably been hearing about ERCOT, the Electric Reliability Commission of Texas or something of that sort.

Daniel Raimi: I want to say it's “council” and not “commission.

Severin Borenstein: Council of Texas, that's probably right. These are the operators who are actually making sure in real time there is enough power flowing into the system to exactly match the amount of power being demanded on the system. In California, that's called CAISO, the California Independent System Operator. They don't actually operate generation, but they are the traffic cop that makes sure everything stays in balance. They also run the markets where the power is traded a day ahead and in real time.

Daniel Raimi: Excellent. That's really, really helpful. The next couple of questions I want to ask you are sort of turning from what caused the problem, which is what we've been talking about, to this question of how do we address this problem going forward. One of the important terms that I've learned from reading your work and watching presentations you've made is this issue of resource adequacy requirements. Can you tell us how California's resource adequacy requirements work? How might they contribute to the problem? Then, how might they be amended to address this problem in the future?

Severin Borenstein: This is a big area of difference across the country and across the world. You need a traffic cop, making sure power flows in to match power flow being taken out. But do you need to plan in advance to make sure that there is enough capacity on the system so that during those peak times you can have enough power? California does; it does a resource adequacy planning process. Texas, by the way, doesn't. Texas basically leaves it to all the load serving entities to make sure they have enough generation to meet their demand. What we do in California is we go to all the entities that serve customers—what are called load serving entities—and say, “if you have so much demand to serve, then you have to make sure that you've contracted with enough supply to meet that.”

We're going to count different sorts of suppliers in different ways. Natural gas-fired power plants we think will be available 90 percent of the time, and so we're going to count it that much. Then, we do nuclear plants and so forth. With a solar plant, well, we know that that only works when the sun shines, so you only get this amount of credit. Demand response also counts as resource adequacy. If you have a certain amount of demand reduction that you can count on, then we're going to count that too. We're not just talking about the huge investor-owned utilities but also some smaller providers, such as what we call community choice aggregators. These are the cities and towns that have become their own retail providers in California. Each of them has a resource adequacy requirement. They add up what they've contracted for in order to meet those requirements.

There are a couple problems with what California has done. The most important one is that we have been using rule of thumb approaches to estimate what you estimate will be needed in ways that don't really match the power that you have to have hour by hour and minute by minute. The most important one probably is the way we credit solar. Solar is a great resource and we can do a lot with it. But when the sun is not shining, if you don't have storage as well, you're not producing. What we've been doing is crediting solar for its average production over a four or five hour window. It meets that requirement. The load serving entities count it and say, “see, we've contracted for it.” Of course it's producing huge amounts in the first couple hours of that window at the end of the day, and then nothing at the end of that window. Yes, on average, it's producing the right amount, but average isn't what counts when you're balancing the system.

The first thing we need to do is we need to have a much more granular minute by minute approach to resource adequacy. I want to be clear, I am not blaming these load serving entities who have contracted for the solar. They have read the rules and followed the rules. They are not the ones who make the whole system balanced. They are told, “you've got to do the following things” and they've met those requirements by and large. Obviously, sometimes people don't exactly do it. The problem is the way we've drawn the requirements don't actually match the realities of production. We have similar problems with demand response— that demand response is credited in ways that when push comes to shove, we're not actually getting what the credit counts for.

Now, the demand response providers will say, “well, we're producing what we said we would.” That is true in some sense, but the credits that are being received for the demand response don't actually match the value it's bringing to balance the system. Then we just have the general problem with generators showing up. They get paid for providing resource adequacy, which means they have to bid into the market. In some cases they don't and in some cases they are bidding in, but they don't actually have the power to provide. We need to have more of a pay-for-performance system so that when you actually show up in those critical moments, you get rewarded. But if you don't show up, you get penalized. I think we're going to move towards that, certainly after what happened in California last summer. I suspect they will also be reevaluating in Texas a move to say, “okay, we need to have an idea of exactly what's going to show up and how it's going to show up minute by minute.”

Daniel Raimi: Great. To do those types of reforms, I imagine there would be new modeling that would be required. Some additional research as a kind of first step to making those changes.

Severin Borenstein: Well, actually the irony is we do this modeling. The California Independent System Operator does it; the California Energy Commission does it. The standards that have been set and the crediting for resource adequacy is not reflecting the best modeling. The modeling is much more refined and granular. It's just using a rule of thumb approach to that modeling, which back when we ran the system entirely with thermal and nuclear, it pretty much worked because most of those differences were random outages. We had fuel running into the generators, and sometimes the generator would switch off, but it wasn't happening in a correlated way. That's clearly not true now with wind and solar. In Texas, we've seen it's not true with natural gas and even coal-fired power plants where extreme cold can cause a disproportionate share of the generators to shut down and then suddenly you've got just not enough power coming into the system.

Daniel Raimi: So interesting. You've already given us some lessons and recommendations to think about going forward, but are there other key lessons that you would draw from the California experience? Maybe what the connections, similarities, or differences might be from the tactics experience, particularly with regard to this issue of incorporating more renewables into the system.

Severin Borenstein: I have to first say, as we did have what I would say was a relatively small outage in California, I think there is certainly a chance it could have been a lot worse. There was a pandemic going on and we don't know how high demand would have been if there weren't a pandemic on those days, but we wouldn't have had any more resources. Likewise, in Texas, it's unclear how the pandemic has affected it. We do need to jump on this. We have to recognize that although no generation is perfectly reliable, each generation type is unreliable in a different way. We need to recognize that the sort of outages that occur at gas plants are not the sort of outages or performance failures that occur in solar or in wind or in hydro, as we saw in Brazil a few years ago where they had a drought. Each of those has to be modeled differently and accounted for differently.

That means a much more granular approach to that. I think we've also learned that we need to take a new approach to demand response where we're really asking what was the counterfactual demand that they would have required so that we're not crediting demand response with reductions that would have occurred anyway. That's generally the problem when you start paying a customer to use less of something; it's always a question of, less from what baseline? And do you have a difference between the true counterfactual usage and their actual usage? This is really also changing as markets. We have more small participants. Back when we had big utilities, they sort of took care of these issues themselves.

Now when you have many load serving entities, each of them is abiding by the rules they face, but those rules really need to be precisely right. You can't count on the judgment of each market participant to say, “yeah, I need to do something a little different from the rules because the system needs it.” When each market participant is just a tiny piece of the system, they are going to follow the rules. We got to make sure we get those rules right.

Daniel Raimi: That's so interesting. Of course, all of this is going to become more complex at least in some ways, for the reasons that you mentioned at the beginning of our conversation, which is more extreme weather events largely associated with climate change.

Severin Borenstein: Both more extreme weather events and, I hope, greater reliance on non-fossil generation, which has a different set of reliability issues than fossil generation does. We need to make sure that we design a system that is resilient to that.

Daniel Raimi: Severin, this has been such a fascinating conversation. I should tell our audience, we didn't plan on having a power outage podcast the week of the great power outages in Texas. This is a little bit of unfortunate serendipity, I would say. Let's close it out now with a question that we ask all of our guests which is, what is on the top of your literal or metaphorical reading stack? Something that you've read or watched or heard related to the environment or energy, even if tangentially, that you'd recommend to folks. I'll start with a book that is on its way to me currently. I haven't picked it up yet, but it is Elizabeth Kolbert's new book called Under a White Sky.

I'd listened to a podcast interview she did recently with Ezra Klein in which they talked about the book. It's just all about the ways that humans are intervening in natural systems in new ways, including gene editing technologies, solar geoengineering, and all sorts of other very interesting and very fraught applications of technologies in many cases trying to solve environmental problems that we've created in the past with new technologies. Seems like a fascinating book and she's an amazing writer. I'm really looking forward to picking that up. How about you, Severin, what's on the top of your stack?

Severin Borenstein: Well, there are a couple of things I've read recently. One I've re-read recently. One is called The Box by Marc Levinson, which it sounds like an incredibly boring topic. It's about containerized cargo, but wow, when you read it, you realized how the innovation of containerized shipping completely changed trade in the world and completely changed the world economy. I was just blown away by the depth of the importance of this one innovation and not a great technological innovation. Certainly nothing like getting a vaccine in 10 months but a pretty basic idea that completely changed the world economy. I recommend that. The other is more directly environmental, and it's a book called The Bet by Paul Sabin. It is about a famous bet between Paul Ehrlich and Julian Simon, that many of your listeners have heard about, about the price of a set of commodities 10 years in the future.

What the book is really about is the whole resource catastrophe movement that grew up in the late 1970s that said we were going to run out of resources by the 1990s. It just essentially extrapolated the usage and availability of resources and estimated that we'll run out of all sorts of commodities, and that there would be starvation and food shortages and so forth. That obviously didn't happen. It's sort of an important lesson in the environmental movement. When I read it, it certainly didn't give me sympathy for climate change deniers, but it did sort of bring home the idea that simply being experts is not enough. We need to actually communicate the risk in a way that is really salient to the population as a whole, because otherwise, there really is a suspicion of experts.

It's not even political motivation. They just get an idea and run with it. That's what happened with the resource catastrophe movement. It's something that changed my understanding of how we need to communicate the climate crisis. As an economist, appreciation for people who are very focused on just the communication part. Setting aside the science, setting aside the markets and the economics, just communicating is so critical to getting people to understand what's going on.

Daniel Raimi: Absolutely. Well, I really appreciate both of those recommendations. They both sound fantastic. I now really want to get both of those books too. The Bet and The Box, very nice concise titles. We'll have links to them in our show notes. Once again, Severin Borenstein from UC Berkeley. Thank you so much for coming on the show today and teaching us all about power outages.

Severin Borenstein: Well, thanks a lot for having me on.

Daniel Raimi: You've been listening to Resources Radio. Learn how to support Resources for the Future at rff.org/support. If you have a minute, we'd really appreciate you leaving us a rating or comment on your podcast platform of choice. Also, feel free to send us your suggestions for future episodes. Resources Radio is a podcast from Resources for the Future. RFF is an independent nonprofit research institution in Washington, D.C. Our mission is to improve environmental energy and natural resource decisions through impartial economic research and policy engagement.

The views expressed on this podcast are solely those of the podcast guests and may differ from those of RFF experts, its officers, or its directors. RFF does not take positions on specific legislative proposals. Resources Radio is produced by Elizabeth Wason, with music by me, Daniel Raimi. Join us next week for another episode.