Subscribe
Share
Share
Embed
Episode 1: Getting the Most from Transmission
Transmission is the backbone of the energy grid, and in this inaugural episode of Frequency Band, hosts Paul Dockery and Becky Robinson explore this critical topic with two leading experts. Rob Gramlich, President of GridStrategies, and Kyri Baker, Research Scientist at Google DeepMind and Associate Professor at the University of Colorado-Boulder, share their unique perspectives.
From optimizing power flows to planning and deploying new infrastructure, this episode dives into the challenges and opportunities shaping the future of transmission. With engaging discussions and thought-provoking insights, it’s the perfect starting point for a podcast series tackling the big questions in energy.
Creators and Guests
What is Frequency Band?
From Fort Peck, Montana to Tijuana, Mexico, the alternating current transmission network in the West oscillates in the narrow band of narrow band 59.97 and 60.02 cycles per second. On Frequency Band, the California ISO will host industry experts to talk about how the physics, economics, and governance of the grid come together to to keep the system in sync.
I'm Paul Dockery.
Becky Robinson:And I'm Becky Robinson.
Paul Dockery:On Frequency Band, Becky and I explore the physics, economics, and governance of the grid from a from a system operator's perspective.
Becky Robinson:Joining Paul and I to talk about getting the most out of the transmission system are two energy experts. First is professor Kyrie Baker. Kyrie is an award winning teacher of power system and grid optimization at the University of Colorado Boulder. Welcome, Kyrie.
Kyri Baker:Thanks for having me.
Becky Robinson:And Rob Gramlich joins Kyrie, Paul, and I. Rob is an award winning public servant, having earned an Exemplar of Public Service Award from FERC and an Achievement Award from EASIC for his contributions to market design and transmission planning. Hi, Rob.
Rob Gramlich:Hi, Becky. Thank you. It's nice of you, and great to be with you.
Paul Dockery:And I feel like I've interacted with both of you on all the social medias over the years, but blew my mind. You have never been on a podcast together before and maybe never like interacted virtually before. Is that right?
Rob Gramlich:That's right. It's great to be with you, Kyrie. But yeah, as we'll get into, we've kind of look at different ends of the system here, but we'll talk about that.
Kyri Baker:Yeah, our backgrounds are very different, but I think that's what's going to make today interesting.
Paul Dockery:That's right. That's right. We're gonna get into, like, short run, long run, understanding the transmission system from different perspectives. But to start that off, I thought we'd start with a ranking game that focuses on power system fundamentals and a game I'm calling Merit Order. So in blind ranking games, you're forced to rank an item before the rest of the list is revealed.
Paul Dockery:From the prep call, it sounds like none of you have participated in a blind ranking game before. So this will all be new. Is that right?
Rob Gramlich:We're very blind.
Paul Dockery:Okay. Well, classic decision making on under uncertainty framing, which is perfect for the power system and for market operators, because you have to figure out how to rank an item before knowing the rest of them. And we're gonna have five items. They're all power system fundamentals to try to kind of get to the real core of understanding transmission and power systems. So for example, one of the fundamentals I could have used are the three P's of transmission planning, permitting and paying for made famous by our own Rob Gramlich.
Paul Dockery:Thank you, Rob.
Rob Gramlich:Would have loved that.
Paul Dockery:And when I introduced it, yeah, it's great. We get to you get credit for it every time it comes up. What I would have done my best Rob Gramlich impersonation and describing it in a few words or less. And then you would have to rank it from one to five. You discuss, discuss, discuss.
Paul Dockery:Any thoughts? Where would you rank it, Rob?
Rob Gramlich:Well, with the three compared to each other?
Paul Dockery:No. Just as one thing, the three p's of transmission compared to like four other things you don't know yet. Where you where would you go?
Rob Gramlich:It's gotta be number one.
Paul Dockery:Okay. And then, like, Kyrie Becky, you have to I have to you have to agree. Like, would you agree with Rob that's number one of all the power system fundamentals that you don't know what is coming up?
Kyri Baker:No. I think it's at least number two or number three just because, like, transmission is nothing without generation. So I think some aspect of generation generating power has to be number one.
Paul Dockery:Okay. Hot take. Where then you got and then Becky, you're gonna have to participate in this as well. It's a consensus driven game. All three of you have to come up with it.
Paul Dockery:Got you got it? Where where do you feel? You supported Becky or Rob in this?
Becky Robinson:Well, it's funny. Like, if you'd asked me first, I'd been like, well, those three things sound pretty fundamental. So does that transmission, so does that put it at at number one? But then I take Kairi's point of, like, how are we framing the box that we're that we're expecting these five things to all fall in. Right?
Becky Robinson:And generation is a key I mean, what's the point of transmission if we don't have generation? So I think it it I think framing helping frame the box will be important to to an informed answer.
Paul Dockery:That's what's so fun about this game. You have no idea how I framed it. You have no idea what I think are power system fundamentals. So it's like, I don't know. So that's decision making under uncertainty.
Paul Dockery:You don't know what the box is. This is like every conversation on power systems. No idea what the box is. That's gonna be the fun part of the game. Three p's is not part of the gameplay today because that was an example.
Paul Dockery:The first power system fundamental I want you all to rank are Kirchhoff's laws. This includes both Kirchhoff's current law, the sum of currents into a node equals the sum of currents out of a node, and Kirchhoff's voltage law, the sum of voltage around any closed electrical circuit is zero. Where do you all wanna rank Kirchhoff's laws?
Kyri Baker:Okay. I'm gonna jump in first on this one just because I teach KCL and KVL. And I'm actually you know, you would think this is number one for me, but actually, these laws are only accurate under some scenarios of, like, idealized circuits. So I'm gonna say they're not number one, but they're definitely up there because there are situations where, especially with AC power and non ideal circuits, these laws don't always hold exactly.
Rob Gramlich:This is just bringing back memories of restructuring at FERC in 90s, when it's a lawyer dominated agency, and the laws trump everything, right? But then the economists and engineers would say, well, they don't trump Kirchhoff's laws. They trump whatever Congress does. So it's they're important, right? And kind of loop flow and all the effects of these laws are really important.
Rob Gramlich:So gosh, I don't know. I want to keep some options, for the other ones. So maybe two or three.
Becky Robinson:Two or three. Yeah. That feels I I I like that they're showing up in the top five. I I I mean, Elliot talks about the the physics and economics of the grid, and it is is sort of, you know, two things that we can't, you know, you can't you can't obviate either of those, you know, in sort of the other work you do, right, in sort of how you design the rules, how you design the regulations. I mean, it's it those are things, incentives and and laws of physics that don't go away no matter how much we may try and pull them in different directions.
Becky Robinson:So I I like two or three. Although I'm I feel like Paul's gonna throw some funny things at us. So I don't know. I might is four an option two?
Kyri Baker:Four is not on the table for me. Well,
Becky Robinson:okay. Two or three would
Kyri Baker:be good. Think these these are two fundamental. I mean, they govern, you know, how we decide what what is transmission versus distribution. What does the transformer do? What, How does power get consumed?
Kyri Baker:It's just so fundamental.
Rob Gramlich:Got a binding constraint from the engineer here. How about three?
Kyri Baker:Three? Okay. I might be willing to do so.
Rob Gramlich:I'm just trying to be a peacemaker.
Paul Dockery:Okay. We good on three? So I wanna I wanna follow-up on this. Don't you think like the power balance constraint falls under these two laws? Like, isn't that sort of it implies the power balance constraint that the current flowing in and out of a node is alike.
Paul Dockery:The power at that node has to equal some of load and some of generation. Does that feel right? That's part of what's here?
Kyri Baker:That that's part of it. But, I mean, I think a lot of ISOs have allowable power balance penalties or violations. Right? There's, like, power balance penalty curves that say you could violate power balance. So it's not necessarily as hard maybe as we think.
Paul Dockery:Okay. But Kirchhoff's laws are hard.
Kyri Baker:So they're number I mean
Paul Dockery:Okay. Yes. So if the power balance constraint comes on somewhere else, because maybe it could. Who knows? Who knows?
Paul Dockery:I don't know. It would be lower. You'd put it. Anyway, it's not number two. The power balance constraint is not number two.
Paul Dockery:Number two is Ohm's law. Current flowing through a conductor is directly proportional to the voltage across it. Voltage, equals current times resistance. V equals IR is the standard formulation. Increasing voltage of a conductor allows more current to flow.
Paul Dockery:Adding more resistance at the same voltage causes current to decrease on the transition transmission grid. Ohm's laws why impedance matters and why higher voltage allows for transmitting power longer distance. Where would you put Ohm's law in this list?
Kyri Baker:Okay. Sorry. When I said, like, Kirchhoff's laws don't apply to every single circuit, I think those are closer to applying to every single circuit than Ohm's law is. Ohm's law is for ideal idealized circuits. It's assuming resistance doesn't change across time, which isn't always accurate, and it's assuming this linear relationship between current and voltage, which isn't always the case.
Kyri Baker:Sometimes it's nonlinear. So I think I would put that at at four.
Becky Robinson:I'm gonna trust Kyrie on this because as far as, like, the two two separate laws of things, I defer to your good engineering judgment, Kyrie.
Paul Dockery:But I feel tension here. I feel this fun tension. What are you going to say, Rob?
Rob Gramlich:First of all, inclined to go with Professor Baker on any of the engineering, but like voltage and capacity of delivery is so important to the system and the economies of scale of transmission. So this kind of, I think this kind of gets to that a little bit, right? If we want to deliver more, could you kind of spend a little bit more and get a little more. You can spend a little bit more than that and get way more power. So the economies of scale are a huge issue in my mind, which might move it up in the ranking.
Kyri Baker:That's true. I mean, it's part of the reason we have high voltage transmission because if you bump up v, you can bump down I and lower the losses. So there's also, you know, the grid was kind of constructed based off of Ohm's law from that respect.
Paul Dockery:So it may not be as true of physics in in, like, the most precise, but is it actually more fundamental to how we think about the grid? There's tension here. Four, two. What do you I I don't know. What do you think, Becky?
Becky Robinson:Well, I like how you're framing that out, Paul, because because right. Is is there sort of more yeah. Is that more impactful, right, from a planning and and, you know, how we actually see the grid at all these different voltages on the system and sort of figuring out what is the the, you know, the optimal pieces to bump up versus, you know, what what's okay to be low voltage. I don't know. Two or four?
Rob Gramlich:I mean, I vote two. Kyrie, what do you think?
Kyri Baker:I I think I'm gonna oh, that's tough. I think I'm still gonna vote four.
Paul Dockery:Oh, you get to be the tiebreaker, Becky. Where are you gonna go?
Becky Robinson:Well, I okay. Well, I I feel like I disagree with professor Baker at my peril, but I I am I'm leaning towards maybe two here because it's it's, it's not that it's right. Maybe it's it's less of, you know, a 100% truth from the engineering perspective, but but it's impacting decisions, and it is it is giving that optionality in how we think about building out the transmission system. So I'm gonna go with two. Okay.
Kyri Baker:Let me have one more argument before you solidify the two. So I think Ohm's law alone isn't enough for what we're talking about. We still need, like, the power laws. Right? Power equals current times voltage.
Kyri Baker:So without v equals I r mixed with p equals I v, we're not gonna have the things we're talking about with high voltage transmission transformers stepping up the voltage because we've been talking about current, but we're really talking about power. Right?
Rob Gramlich:Alright. I've been outgunned. I'll I'll go with four.
Kyri Baker:You don't have to change your mind. I'm not I'm okay if you two outvote me.
Becky Robinson:Well, Paul, I I mean, are those power laws other things on your list? I feel like we need a sneak peek of.
Paul Dockery:You don't know. That's the fun part of the game. You don't know. You don't know. You don't know what the boxes are.
Paul Dockery:Those coming next? If those if those come next, would you put it at two? Who knows? That's decision making under uncertainty. These are the things we have to do on the power system.
Becky Robinson:Well, Kyrie, what do you think?
Kyri Baker:I guess, actually, even if it was one of the other ones, I guess I would still rate Ohm's Law above it. So maybe I'm willing to say I'm okay with two.
Becky Robinson:Okay. I'll take that as I'll take that as we're a bumpy way we got to Yes.
Paul Dockery:We've got we're going with two. We're going with we're going with Kirchhoff's Law is number three, Ohm's Law is number two. Next up, we have open access. FERC orders 888889, the landmark 1996 rulings that prompted wholesale competition through open access, nondiscriminatory transmission service by public utilities mandated transparency and transmission operations and establish the foundational principles that transmission service must be just reasonable, non unduly discriminatory or preferential. Where do you rank open access?
Rob Gramlich:I mean, I gotta jump in. This was my first job out of college. I had a tiny bit part as a GS seven government employee writing order eight eighty eight. And I think it is, it was transformational. And I think competitive generation markets have been tremendously beneficial for consumers.
Rob Gramlich:So I would rank that pretty high.
Paul Dockery:But where? Is it one? Are you arguing for one Rob?
Rob Gramlich:Yeah. I guess I would.
Becky Robinson:These are the rules of the game. I mean, because number four doesn't feel high, but I I feel like this is a a hard question. As a former FERC staffer, I feel this is a hard question. Kyrie, what do you think? How do you situate, yeah, FERC open access against these engineering laws?
Kyri Baker:I mean, my instinct is is to rank it lower. I mean, I I certainly think it's important, but not as important as, like, the laws that govern electricity. Like, I can't say it's more important
Becky Robinson:than Laws of man versus laws laws of people versus laws of science.
Paul Dockery:This is the fundamental argument on the transmission system. Where do you put the physics, economics, and governance of the grid? How do they overlap and relate? How do you prioritize one over the other? Oh, this is so much fun.
Paul Dockery:Who knew this game was gonna be so fun? What do you think? Are you arguing for one Rob or not? Are you arguing for one?
Rob Gramlich:I mean, yeah, but I mean, Kairi just trumped me with, I mean, the physical laws of power kind of do trump everything else. Like, you could do a variety of governance structures and economic institutional arrangements. But you can't do a variety of physical laws for electric power. Those are constant across all countries and all time.
Paul Dockery:Okay. You're arguing for four. Where are going, Becky?
Becky Robinson:I feel like the argument for the argument for four is not that it's not important. Right? I mean, I think it's sorry. The argument for one. The argument for one is, you know, often we have, not me, but like often with our engineering caps on, we say, well, this is just how it is.
Becky Robinson:This is how it has to be. And it can be really transformational when someone comes along or some institution comes along and says, you know, we're just going to think about this fundamentally differently than we have in the past. And and and I think open access was that. Right? I mean, was it was it was a huge deal.
Becky Robinson:Right? And it it demanded that, you know, that entities that own these power lines, no longer assume that they're theirs to use for the, you know, that that they other people have to be able to use those despite the fact that, you know, as an entity, you know, you built those, your customers paid for those. You know, that's that was a you know, it's it's hard to it's hard to understate how fundamental a shift that was.
Paul Dockery:I'm not participating in this, but I do think to, like, articulate the the number one case and reinforce what Becky said, You can have an electric circuit that operates a little little widget, but to make it work across the country, you need enabling structures, and open access is one of those enabling structures to let people participate. They make it they take it from a widget and make it a system. They make it societal, like in scale and beneficial to all of us. Not just, oh, I can the physics work, but it unlocks more when you do with the with politics, with people, with with the institutions that make it broader than just electric circuits.
Kyri Baker:That's a strong argument and it makes sense. But I still can't get over the that wouldn't exist without the others. Like, you know, what layers of dependencies do we have here? Although, yeah, again, Kirchhoff's laws, Ohm's law, these are just our understanding of physics. Even though it's been confirmed trillions of times over, it still is our understanding and our models that we've come up with as humans.
Kyri Baker:So both of them are sort of human created structures to some degree. It's just one builds on the other. So I would still argue for four, I think.
Paul Dockery:Okay. Gotta get to consensus. Where are at, Rob?
Rob Gramlich:I can go with that. Now we should have put the physical laws as one and two. But, yeah, let's let's go with four for now and see what's left.
Paul Dockery:Feeling good, Becky?
Becky Robinson:Yep. Yep. We can I there's just so much uncertainty about what's coming next?
Paul Dockery:That's exactly right. Okay. Number the the fourth item. Right now we have Kirchhoff's three, Ohms two, Open Access four. The next item on the list is bid based security constrained economic dispatch with locational marginal prices.
Paul Dockery:So this is the formulation of nodal markets in The US that ensures the physics and economics of the grid are aligned. Locational marginal prices from bid based security constrained economic dispatch reveal the price of serving the next increment of load at every bus in the network as an output of solving an optimization problem. Where do you all rank spot pricing of electricity?
Becky Robinson:I'm glad we have a number one spot open still.
Rob Gramlich:Okay. Are you suggesting it's number one?
Kyri Baker:I have mixed feelings about this one because
Becky Robinson:Well, I want to hear from y'all.
Kyri Baker:The security constraint economic dispatch problem assumes like linear models of power flow. And so the LMPs are not exactly the cost that would physically take to deliver power to a specific location. So I think I have a bit of qualms with it just because it's not a 100%
Paul Dockery:that's a part the way we formulate security constraints. It's a modeling decision now. Concept is not limited by that modeling decision, though. I understand that we are limited by computational intractability of the problem, but this formulation isn't. You could if you had the right math in it, it still like merges the physics and economics and concept in theory.
Paul Dockery:I shouldn't be intervening. I just find this to be really interesting.
Kyri Baker:Location pricing has incentivized so much. And I think Rob, you have more to say about that, but it's important, but I'm just hesitant to rank it above everything we've said so far.
Rob Gramlich:Well, I'm a huge fan. Was holding back a little bit to see what you both would say. To me, it's an extremely efficient way to operate a regional power system and regional power systems are extremely efficient ways to operate electricity. It is the de facto standard market design for the country and really for the world, even though not everybody has adopted it. I think everybody acknowledges that it's the efficient way to go, but for some of the politics and winners and losers of shifting from what Europe's doing to that, I think they acknowledge there.
Rob Gramlich:I'm just recently back from China, I was amazed that, you know, people like, you know, I was in a room of 100 people, and they all know about LNP and FTRs and Bill Hogan and Schweppi and the spot pricing of power original paper back in the 80s and all of that. And they view that as the model. And it was kind of actually, in a trip otherwise, it was really depressing about US versus China's like energy infrastructure development. It was one thing we could be proud of here that we actually developed this and it's the model for the world.
Becky Robinson:Yeah, I think of the title for episode.
Rob Gramlich:Cold enough Spot Pricing of Electricity. That's great, Paul.
Becky Robinson:In arm's reach. No, I love it.
Rob Gramlich:I didn't mean to cut you off.
Becky Robinson:Yeah. Thinking about getting the most out of the transmission system, I mean, it feels like that is not the only tool, but is the overall framework that we have to think about how we optimize over the transmission system. And and agree with Rob. It's been so successful, and it it is it continues to be so relevant and, you know, it creates the right incentives. It, you know, it's it's doing a lot of work.
Becky Robinson:And it just seems to I guess it's it's so I'll repeat, it's very fundamental. I mean, that seems like the point to me. That's the argument for number one.
Kyri Baker:Yeah. Okay. I think after listening to both of you and I also thought about it a bit more, I mean, it really is it encapsulates some form of Ohm's law, Kirchhoff's laws, the constraints of the problem. But it also helps incentivize economic development in ways that, yeah, countries that just use merit order dispatch or something without any constraints don't have. And it improves reliability.
Kyri Baker:I'm actually okay with with number one, I think for this.
Rob Gramlich:Oh, alright. I'll take it. That's great.
Paul Dockery:Sticking in the number one. Well, it's we only have one or five. You you it's last choice. Okay. You're good with one.
Paul Dockery:We're going with one. Spot pricing electricity number one. Okay. That leaves number five for Maxwell's equations. These
Rob Gramlich:are
Paul Dockery:the four laws that govern the physics of fields and waves. They explain how electric and magnetic fields are created, how they interact, and how they propagate through space. Most importantly, for the podcast named Frequency Band, it explains the magnetic coupling of all the synchronous generators on the grid. This magnetic coupling combined with system inertia and rotating mass is what provides the electric system with the inertia it needs to keep it in the narrow band of a frequency between fifty nine point nine seven and sixty point o two hertz across the western interconnection. But you left ranking the magnetic coupling of the grid for five.
Paul Dockery:How do you feel about that?
Kyri Baker:Super important for the history of the grid back in the war of the currents. It's like Tesla versus Edison. Part of the reason we went with AC power was because there's this magnetic coupling, and we can use transformers to step up the voltage efficiently. I think it's super important, but I actually feel comfortable with putting it at five. Like, it could be that we had a big we have a big network of HVDC lines at some point in, you know, a couple hundred years from now that is replacing the AC grid.
Kyri Baker:Probably unlikely, but possible. So I think that magnetic coupling is maybe less important.
Paul Dockery:Okay. So any other comments, Rob, Becky, on Maxwell's equations?
Rob Gramlich:Yeah, no disrespect to Maxwell. But yeah, no, it's also fundamental physical law. But I was worried you're gonna say mother or apple pie, we're the last one and then we had to put them fifth and you know, this feels easier than putting those fifth.
Paul Dockery:Yeah. Okay.
Becky Robinson:Yeah. If Kyrie is good, I'm good. Yes.
Paul Dockery:Okay. Well, our merit order then, our merit order is one, spot pricing of electricity. Two, Ohm's law, v equals IR. Kirchhoff's laws is number three. The sum of currents into the node equals the sum of currents out of the node, and the sum of voltage around a closed loop is zero.
Paul Dockery:Four is open access. FERC orders 888889. And lastly, we have the magnetic coupling of the grid as formulated by Maxwell's equations. I didn't include Joule's laws. We did talk about along the way.
Paul Dockery:I think that was fun. Any takeaways before we get into what I think is a fun conversation that we've already started unpacking a little bit?
Rob Gramlich:Fun balance of physical versus institutional versus, you know
Paul Dockery:That's right.
Rob Gramlich:Policy issues. Hard to hard to say which is more important.
Paul Dockery:I think it did reveal along the way. Becky, you want to say something?
Becky Robinson:Well, I just was gonna say I think it's interesting that we our inclination was not to name anything number five. You know, like, five was the last spot to fill. Right? Because you're naming things that all feel important. And so I'm curious, and when we do this game more, like, is there gonna be something you throw out?
Becky Robinson:We're like, oh, clearly, that's that is the least important of the things that you could possibly launch out of us. But I just interesting aspect of the game.
Paul Dockery:I'm actually curious for next time, like, what one did I leave out that maybe we should have covered? Any of the I I left out we should have covered, Kyrie?
Kyri Baker:I think those were pretty good. Yeah.
Paul Dockery:Love the endorsement. I love the endorsement of my list making. Okay. We're gonna get into the difference between, like, short run and long run. So I think, Kyrie, you come at this problem from the physics of the grid and how to get the most after kind of the configuration of the system is set and how to unlock it.
Paul Dockery:Rob, I think of you coming at this this question of transmission in the grid from the long run. How do you set up your system, make sure you're investing in the right transmission resources so you can get the most out of a broad widely connected grid. I think those are really interesting perspectives. It came through a little bit in the game where we're thinking about it from, hey. What's the setup question?
Paul Dockery:What's the big the big structural things that we need to balance against the near term short term optimal stuff? That's what I wanna get to in in really the heart of this, which is getting you all to kind of navigate that space together. I'll start with you, Kyrie. What do you wish Rob would think more about when he was advocating for a big, big transmission grid and thinking about, how the short run impacts the long run?
Kyri Baker:I think just talking from, like, an optimization perspective, you're going to get a totally different optimal solution if you assume, like, a simplified operational model of the grid. So I think right now, these two areas are decoupled because it's computationally challenging to say like, I'm going to model every single five minute power flow in my fifty year planning. But with new tools, it's potentially possible. So I think I would encourage both of us to talk a little bit more about how our communities can work together and incorporate higher fidelity operations into planning. Because I think they do impact multimillion dollar decisions pretty strongly.
Paul Dockery:Where are you coming at from, Rob? What do you wish Kyrie took more into account? And can you do that? Can you do higher fidelity? What would that help you with?
Rob Gramlich:Yeah, well, certainly it's important to like from the perspective of squeezing as much as you can out of the existing network, all the short term optimization work is really important to Kyrie and many other people do and that all the ISOs and RTOs a lot on with their software. And we just ranked, you know, bid based security constraint, economic dispatch as our number one. So like hard to say those aren't important. But I do always tend to get back to the greater importance of getting the infrastructure right, the platform, the big transmission lines that are the platform for everything else that happens. And you can think about California going back, you know, during the energy crisis twenty five years ago, would it have been better to accelerate the real time day ahead, you know, market with LMP and, you know, mixed energy programming that was later added and all those optimization things?
Rob Gramlich:Or would it have been a better use of everybody's time to get PATH 15 built ASAP? And to me, like some of those transmission investments proved out billions, if not tens of billions of dollars of benefit. So that was built in 2003 and it did help, but getting that infrastructure right, I think tends to be important. And that's another thing where sometimes short team people, I don't want to attribute this to Kyrie, but sometimes you can just look at production cost modeling and say, Oh, well, there's not much transmission value. But that always tends to ignore reality.
Rob Gramlich:Like real congestion always seems to be much greater than prospective forecast congestion just because models kind of think the system's always going to work well. And California energy crisis was a great example of like, things can just blow out extraordinarily, and transmission is that insurance policy where you can't say exactly where and when, but it provides so many options under various scenarios.
Becky Robinson:So Kyrie, I like both of these points. Like, Kyrie, I like both where you and Robert coming from on this. And I guess, like in thinking about the tension between, like, taking more granularity of the the power system modeling into account when we're doing these transmission planning conversations. I mean, it seems like one of the huge challenges, what's the Yogi Berra quote? It's tough to make predictions, especially about the future.
Becky Robinson:There's so much uncertainty in what the future is gonna look like, right, with a changing resource mix, with, potentially different you
Paul Dockery:know,
Becky Robinson:demand changes in different ways at different locations, flows potentially changing directions, right, in terms of future use compared to today's use. So it seems like one of the arguments I could imagine for people saying, no, we don't need more granularity in in these transmission planning models is just if I if I add that more granularity, there there might be more that I'm getting wrong in terms of, like, how things can play out differently in the future. So how do you like, what what's your pushback there? Like, why would you say despite uncertainty, it's still really relevant to take more of that into account?
Kyri Baker:No. Those are great arguments because there's, like, a concept of overfitting. Right? Like, at some point, you do not need that granular of estimates thirty years from now because they're going to be wrong. And so my argument against that is, well, I think it depends on the situation, but with fast enough models, you could simulate like millions of hypothetical situations and just get, you know, a distribution of potential solutions.
Kyri Baker:You could do very, very crazy things like remove, you know, 50 lines that you would otherwise think have zero probability of being removed because you can't predict the future. Run this on the cloud for one month, something crazy, simulate every single possibility known to mankind and get yeah, like I was saying, a distribution of possible outputs. So that is a possibility. Whether or not it's worth doing that is another question. Maybe for a shorter term, long run simulations, it would be like ten years, fifteen years, but maybe not for for things longer than that.
Kyri Baker:So measuring where the uncertainty starts to stop paying off with the benefit it provides.
Becky Robinson:Yeah, totally. And Rob, is your take like, Path 15 might have showed up in an awful lot of those, you know, huge amount of runs. Right? And then you kind of get more comfortable that like, you know, in a wide variety of scenarios, a lot of different things happening, there's still a lot of value with this big interregional transmission.
Rob Gramlich:Yeah, I think that's right. I mean, Kari is right. Like if she were doing this in 1998, she could have done a variety of scenarios that, you know, led to the low hydro year and all the like things and like, wow, prices really blow out. And that could have popped up in a prospective model at that time. I just tend to find that, I don't know, maybe engineers are conservative, and it's hard to kind of show that some wild scenario is really possible.
Rob Gramlich:So they tend not to include those, which then leads to a narrow and small estimate of benefits relative to what it really could be.
Kyri Baker:And some of those scenarios are just impossible to predict. It's like twenty years ago, we wouldn't have been like, Oh, the transformer architecture for large language models was invented, and now there's all these massive GW loads. You know, there's some things that are just impossible to model. So I totally see the need for simplified modeling in order to make a decision that's 90% of the way there. But I do want to make the argument that, you know, computationally, a lot of these short term models are getting very fast.
Kyri Baker:So they could be used in more of the planning scenarios.
Paul Dockery:One of the the when I enter into these conversations, I often think there's this mismatch between what people expect the model to tell you versus what it tells you and kind of what from a policy perspective need the information they need versus the information they get. Often when I hear this conversation, people think about, you know, the congestion cost, and how that transmission benefits is is like, I'm gonna get paid for that transmission and recover it through changes in spot price. But ultimately, the transmission just like changes the counterfactual and your benefits are just counterfactuals. And it seems like there's often like a mismatch of like how the money flows versus how, what your, the information you're provided when modeling. Rob, does that resonate with you?
Paul Dockery:Because I feel like often in these policy discussions and trying to optimize the short long run, we think about the short run-in a way that probably isn't real.
Rob Gramlich:Yeah, well, does in a couple of ways resonate with me. Number one, there's always been this confusion about restructured markets and whether that's the full LMP type market or something close to it. It was never intended by the original developers to be the signal that led to the right amount of transmission investment. Like the MIT folks and the Bill Hogan and the Andy Ott at PJM and the other people who like implemented and developed the theory and practice of LNP never said, Oh, that's going to produce the right amount of transmission. So that is often confused.
Rob Gramlich:I think those folks would all acknowledge that you actually need a robust transmission planning process in addition that looks at all of the public goods and externality aspects of transmission and plan the right amount. And then the other aspect that resonates with me is that just the amount of precision, because when you're doing long term system planning, whether it's transmission and generation, have to do this synthesis of different models like power flow, production cost, capacity planning, you know, all these things. And, you know, that's way more art than science. And so it's very hard to kind of say there's any precision to the final result and then to prove it to regulators and skeptical stakeholders. So I just think it's, know, we have to keep that in mind and recognize that there's going to have to be some sort of some art and also I think some leadership by transmission planners to like get all the stakeholders aligned around like, look, we're never going go know for certain, but we can tell under all these scenarios that we're gonna need this infrastructure.
Paul Dockery:Kyrie, do you mind jumping in there? Like what comes out of these models that you think, how would you frame these models to policymakers and what that actually is telling you versus what may get confused out of these models?
Kyri Baker:Well, I think it would be a mistake to take the output of the models as an absolute ground truth because the models have things like, what's the resistance of this power line? We never know that with a 100% certainty, for example. So we're never gonna get the flows a 100% right. Even, you know, if we were running an optimization today, things in the grid physically change. They're temperature dependent.
Kyri Baker:They're wind speed dependent. So the models have a lot of noise. So taking their outputs with a grain of salt and sort of saying, okay, here's a trend I'm observing rather than the price is going to be x at this date is an important thing to keep in mind. And another thing that people don't really think about, and I think a study came out maybe out of Jesse Jenkins' group, is the optimization solver that you're using behind the scenes that solves like these mixed integer problems. The type of solver will actually dictate the solution.
Kyri Baker:You can run the exact same problem with two different pieces of software. And it might say, oh, you need to you know, this is a one in this simulation and this is a zero in this other simulation. So it's actually quite specific not just to the problem you're solving, but also the software that ends up solving it. And sometimes these decisions are, yeah, the cost is, oh, the overall cost is comparable, but you're seeing completely separate physical decisions being made that are almost equivalent in terms of cost. So taking it with a grain of salt, understanding that, you know, the software isn't magic, and that it will sometimes provide things that look like garbage, and understanding there's a high level of uncertainty.
Kyri Baker:So use it as an indicator, something that helps, but not as a ground truth.
Paul Dockery:Becky, anything? Where
Rob Gramlich:do wanna
Paul Dockery:go next?
Becky Robinson:I'm just I the question in my mind then is, is it is it good enough for informing a beneficiary's pays approach, as to the who pays for transmission? Right? Because we and not, you know, some there's different methods. I don't know that that I don't know if there's any region that is saying we are going to cost allocate based on these model results. But at the same time, it does feel like it informs, you know, like, is it worthwhile to do this and and who is going to benefit from this line?
Becky Robinson:But but right. Knowing that that's all you know, it is not predicting the future. It is, simulating some possible outcomes. You know, how do we what is the right role for that to play in terms of then thinking about how, right. If it's worth getting built, who's it who's that value, accruing to?
Becky Robinson:Right? And who should be willing to to pony up money, right, to to pay to help pay for the new transmission facilities?
Kyri Baker:Yeah. And one thing that I like to think about is, you know, the fact that mathematically optimal does not equal socially optimal. There's some things that we just can't model with equations that we would be unable to truly capture the value or cost of with math. And so that's also another thing to keep in mind is some things just can't be modeled. So we can't fix everything with more granular modeling or optimizers or software.
Paul Dockery:Mathematically optimal is not socially optimal. That's a merch right there. I love that. Rob, this seems like your sweet spot of like, how do you balance the mathematical solutions, the finding a bunch of different alternatives and then beneficiary pays and permitting and how do you actually get things deployed on the grid?
Rob Gramlich:Yeah, well, beneficiary pays is a great, you know, sort of fairness principle, and fits well within the Federal Power Act. But yeah, it's hard to determine beneficiaries and they change over time, I think we do have to recognize that look at, you know, look at path 15, not the sorry, the Pacific Intertie, you know, originally built to send excess hydropower down into Southern California, but now it's going switching back direction every day, right? It's sending excess solar north and so many lines around the country are that way where the flows are not at all what they were originally estimated to be when the original analysis would have been done. I think we do have to kind of acknowledge the benefits to the whole network when we're doing that analysis.
Becky Robinson:Right, acknowledge that there are some things that will remain unknown.
Paul Dockery:And we might have to have flexible frameworks that, that allow for changes in conditions. You know, one of those flexible flexible frameworks is mid based security constrained economic dispatch with locational marginal prices. Any anything you wanna close out on with with LMPs and nodal markets? And how I'll just in my perspective, these are like the the the tools we use to make the physics and economics work together.
Kyri Baker:Yeah. LMPs are actually a mathematically beautiful thing. They pop out from the optimization problem. They're not a construct that we've come up with. They're actually like variables from the optimization itself.
Kyri Baker:And I think that's awesome. Like, it's not a it's something that reflects the physical realities of the grid to the to the extent that we're approximating the physics. But the power grid is really this beautiful merging of economics, physics, and and other things, and you don't see many other areas that are like that. So I just I love being in this area for that reason. It it's something where you can benefit society, but keep it technical, I guess.
Rob Gramlich:Well, I'll echo the appreciation for LNP. And of course, so those prices are changing by node, but also by hour, right? So over time and space, and so very efficient sending the right signals. But we also need to kind of keep in mind that that doesn't mean consumers are always exposed, you know, 100% to these unpredictable and volatile prices. There are ways to hedge both the transmission costs and the power costs over time, and regulators need to focus on that, but it can absolutely be done.
Rob Gramlich:But we just need to make sure that we should, you know, operate the system on a day to day, hour to hour basis as efficiently as possible, because any inefficiency is definitely being paid by consumers every day. So, you know, there's a lot of room for improvement across like the entire US West right now in terms of just the day to day economic dispatch that we could, you know, bring those savings to consumers.
Becky Robinson:Yes. Yes. EDAM. We're so excited about our extended day ahead market and and so excited that of getting that platform up and running to be to be a way of getting more out of the transmission system than we do today and getting getting more optimal dispatch and delivering benefits to to customers across the West. So a great point there, Rob, and thank you for the platform to to, to celebrate our impending EDEM launch on May 1.
Paul Dockery:And this, you know, podcast is meant to talk about the issues, like, at the at the frontier of the sector and the frontier of electricity markets. And we we talked about, yeah, there's advancements where of extending the markets and and keeping to propagate them in the Western interconnection. But I'm also curious what you all see on the horizon for elect like, spot pricing electricity and these power system dynamics. So, you know, spot pricing electricity, I held up the book earlier. That was nineteen seventies, I think, is when Schweppi first wrote about the spot price in electricity, mixed in your mixed integer linear programming started getting used for unit commitment, and I I I don't know.
Paul Dockery:The nineties, I believe. What's the next frontier of, like, computationally, Kyrie, you've mentioned we have more computational power. What's the next step in being able to merge the physics and economics better?
Kyri Baker:I mean, one good next step would be to, yeah, improve the fidelity of the models that are inside the security constraint economic dispatch, just having better estimates of like, what's the real line capacity, you know, dynamic line ratings, or what's the impedance of, you know, better estimates of the state of the power grid, better estimates of the impedance of transformers, stuff like that. Improved data collection will help us improve the dispatch model and ultimately save consumers money. So I think better use of data and higher fidelity physical representations is probably one of the more important next steps.
Paul Dockery:How much can you squeeze out though? I mean, we're already doing and we're using approximations, linearization of these problems. The California ISO does do a little bit we don't do simple linearization. We do, I think, a more computationally accurate problem solution
Kyri Baker:Yeah. Where we solve AC power Right. Yeah. You you don't use
Paul Dockery:the We DC power iteratively resolve and rebaseline an AC power. But but what's But what value is there? What is the thing you're getting more efficiently and the outcomes that change when you have a better fidelity in your optimization?
Kyri Baker:So in areas that use, for example, the DC approximation where you assume lossless lines, all the voltages are the same, no reactive powers, or you have a model for the losses. In our simulations, we've shown that you can save, depending on the network, two to 12% less generation if you switch to a higher fidelity model. So that's just swapping out the software that does the dispatch because it has a better estimate of where the line flows are actually going. So you're ahead of time saying, okay, I'm going to get it closer to right. I'm going to turn on the power plants at the level that's going to actually satisfy demand, satisfy line flow constraints, and I'm not going to have to rely on reserves or slack bus generation as much or these other ancillary services, which lowers generation, it lowers losses, it lowers cost.
Kyri Baker:So the savings there are, I think, pretty substantial just from swapping out the model.
Paul Dockery:That's awesome. That's a frontier of power system optimization. I'm glad you're exploring, Kyrie. I think you are. I think it's fair to say that's what you are enthusiastic about, I believe.
Paul Dockery:Rob, so what's the frontier of like, the long run transmission? Is it really just about building, just build transmission? We know where to build it, just build it.
Rob Gramlich:Yeah, for sure. But I was thinking about the systems too, and with due appreciation for all of the improvements over time that, you know, went into the short term market operation. And again, mixed energy programming, LMP, both of which I think PJM first put in place, which I feel I need to say as a former PJM employee, and my former RTO is getting bashed all the time, that those things led to massive efficiencies and savings for consumers, and that should be appreciated. And PJM couldn't have even operated a power pool half as large as it is if it hadn't done some of those mixed energy programming. But longer term, I feel like it's a little bit more of kind of like the institutional issue of like sometimes when the power systems get scarce and tight, the operators still and this is not, this is like, I think all the ISOs and RTOs, they still kind of override the system, right?
Rob Gramlich:And they put excessive reserves on when things get tight. And then these actions are not reflected in market prices, and that results in potentially inefficient dispatch, but certainly longer term inefficient price signals for investment. And I think that's where the RGOs and ISOs probably should focus most is, and maybe there'll be less conservative, if they get better modeling and maybe AI and automation and the ability to process hundreds of scenarios super quickly with fast computing power will enable them to put more trust in what's happening so they don't have to override it manually. But that's where I think of is what was important for like driving investment.
Paul Dockery:Awesome. Becky, you wanna make it all make sense before we close it out?
Becky Robinson:I don't know that I can say it better than our guests have. Fascinating stuff and lots of good work to come. I mean, I guess I take from your question, Paul. I don't know the answer to your question of what's on the horizon for these things, but I I like that that there's plenty there, it sounds like, in terms of, you know, the continued work and continued, you know, seeking out more efficiencies. And it's an exciting business to be in.
Paul Dockery:It is. It's an exciting time to be in our business. I appreciate all of your participation in this. I really enjoy Wonky Conversations. To to, like, commemorate your participation in Frequency Band, I I am I'm giving you a souvenir in the form of a participation trophy.
Paul Dockery:First, Kyrie, I'm gonna convey to you the Frequency Band Synchroscope Award for exceptional contributions to synchronizing perspectives. Telescopes, microscopes, oscilloscopes, kaleidoscopes, and synchroscopes all shift your point of view to help you see things that are otherwise hard to perceive. Synchroscopes are used on the grid, to reveal whether systems are in sync and operating at the same frequency. Thank you for helping us synchronize perspectives today. Any speech you wanna give to close out to close out the episode?
Kyri Baker:No. Thank you, Paul. Thank you, Becky. And so happy to do this with you, Rob. I think I learned a lot just from the past hour.
Kyri Baker:So, yeah, thanks everybody.
Paul Dockery:Awesome. Keri, I do hope you feel seen, heard, valued, and appreciated. Rob, I'm happy to convey to you the Frequency Band Breaker Control Switch award for your help revealing the steps we need to take as we close in on the future. Breaker control switches are where people intervene to connect or disconnect systems. Your insight help us connect the short run with the long run.
Paul Dockery:Thank you, Rob. Any acceptance speech or closing thoughts?
Rob Gramlich:No. Just thanks for asking Kyrie the tough engineering questions. It's great to be with you and also emphasize, let's not forget the three p's even though we put them to the side, planned permits and pay for some transmission like you and your colleagues at Kaiser have been doing pretty consistently. So thanks for the opportunity to be here.
Paul Dockery:Thank you. I hope you feel seen, heard, valued, and appreciated. Becky, I don't have an award for you, but I still hope you feel seen, heard, valued, appreciated. I do. Any closing thoughts?
Becky Robinson:I do. No. Just thank you to both, Rob and Kyrie. Great getting to talk with you today. And, I took some notes because, yeah, I totally agree.
Becky Robinson:It's great to get to to hear what's on your mind and and learn some things from that. So thank you.
Paul Dockery:And thank you to our listeners. While you aren't seen or heard, you are valued and appreciated. Freak keep Frequency Band together by sharing with other energy enthusiasts like us. Until next time. Stay in sync.
Paul Dockery:I don't know. We gotta work on the closing here. I don't know. I don't feel like I've landed the closing line. Stay in sync, everybody.
Paul Dockery:Stay synchronized.
Rob Gramlich:I like it.
Paul Dockery:Keep synchronized. Love it, Rob. Thank you very much.
Becky Robinson:Frequency Band is a production of the California ISO. It is produced and directed by Paul Dockery with writing by Paul Dockery and Becky Robinson.
Paul Dockery:It is mixed, edited, and published by Paul Kolodich. Jamie Ackman is its editor in chief. Its executive producers include Crystal Ball, Jacob Mays, Nicole Hughes, Aaron Bloom, Deborah Smith, Monica Gaddis, and Pam Sporbord.
Becky Robinson:The views expressed during today's recording are our own and not the official views of California ISO or the official views of the organization of the guests appearing on Frequency Band. Any aggregation, quotation, or references to opinions shared in today's episode should be ascribed to the individual participants and not their respective organizations or Frequency Band. You can find additional information in the show notes of today's episode.
Paul Dockery:Frequency Band, celebrating the walkie charm of electricity markets.
Becky Robinson:Frequency Band, keeping the beat at 60 hertz.