On The Brink with Castle Island - Blake King on How Grids can Accommodate Bitcoin Mining (EP.330)

Episode Date: June 27, 2022

Blake King, power markets engineer at Galaxy, formerly of Oak Ridge national lab and ERCOT, joins us to talk about how changing grids are accommodating an influx of load from Bitcoin miners. In this episode:  Why the grid is undergoing a radical transformation today Blake's background in electrical engineering, power system modeling, and renewable development The origin of Blake's papers on bitcoin mining and the grid Bitcoiner misconceptions about mining? Why Bitcoin is not a battery Why you can't just arbitrarily add load to the grid Why Bitcoin mining does not generally increase energy prices during scarcity or peak load Why rational actors curtail or sell power during scarcity events Are grid planning models correctly estimating the risk of furloughing thermal generation? Is it possible to make the green energy transition as currently envisioned? Why renewable generation requires more transmission How renewable generation costs are increasingly upfront as compared to thermal generation Are bitcoiners overrating the importance of controllable load? Why Bitcoin miners may be over ambitious in terms of modeling ancillary service payments Are ancillary markets going to increase in size with more renewable generation? Economic demand response versus formal ancillary services Different types of ancillary services products Why load resources are able to bid out thermal generation in ancillary services markets How exactly Bitcoin miners get signals to curtail their load during scarcity events Does being a load resource imply 'voluntary' curtailment? Why miners owe it to the grid operator to be transparent in their breakeven price and operating plans How weather in Texas could affect block arrival times in Bitcoin Why the commodification of ASICs would increase the integration of the renewable energy sector and bitcoin mining Why ERCOT hit the pause button on the addition of new mining loads ERCOTs level of sophistication around Bitcoin mining Will miners toggle between Bitcoin, green hydrogen, and other location-agnostic uses of power? Do the economics of green hydrogen make sense relative to BTC mining? Content mentioned in this episode:  Bitcoin Mining and the Grid, Part I and Part II Intro and outro used with permission: Ocarina #968

Hello and welcome back to On the Brink. This is yet another episode of the mining miniseries. Today we're sitting down with Blake King, who is a power markets engineer at Galaxy, formerly of Oak Ridge National Lab, and Urquhart itself, Blake has done it all in the power industry, and he joins us to talk about the intersection of Bitcoin mining and modern grids, which are changing dramatically. Blake's written a fantastic miniseries for brains called Bitcoin Mining and the Grid. He shares a tremendous amount of detail on how miners are interacting with the grid today. We cover a lot in this conversation the transition that a lot of these grids, including Urquat, are undertaking how Bitcoin miners specifically can modulate their load and how they

compare to flexible generation, the ancillary services market, whether curtailment is voluntary or not. And even how the weather in Texas could one day affect the block arrival times in Bitcoin, Blake is a subject matter expert, and I'm really glad I got him on. I think we're going to have to do it again because there's so much to cover and we didn't get to it all. Let's dive right in. Blake King, welcome to the show. Very excited to have you on. Thanks for being with us. Thanks, Nick. Glad to be here. I'm a big fan of yours. Feelings Mutual. I have learned a lot from, your little mini series that you wrote. And, you know, I think our mining miniseries here is all

about interesting new models of mining. So I think you're the perfect guest to contribute. Well, happy to be here. And yeah, it is, it is an interesting time. I'll say especially in power systems, we're at a very sensitive moment in the industry where, you know, the future is really uncertain on what's going to happen in power markets. And so, you know, we can speak confidently about Bitcoin mining's impact, but the future is still unknown and all of these things are really just, like I said, really sensitive to your assumptions and how the future kind of unfolds. Right. I think people that haven't looked into power, not aware of how significant the transition is sort of expected to be or politically mandated. If the sort of plans that are generally laid out

occur, grids are going to look dramatically different. Yeah, it can't be understated. I mean, you speak to people in the industry and it's, I mean, it's a given. Like the renewable transition is happening. It's not necessarily. And you look at interconnection cues. I mean, sorry to jump into the weeds here, but you look at interconnection cues, which are, you know, basically new generation developments that are entering the grid. And it's all renewable. I mean, there might be one, one combined cycle plant, one piker plant, but it's, you know, it's. 90% wind solar storage. And the existing generation fleet is aging.

So you're really looking 20 years down the road and saying, wow, the grid is really going to look different. And it's a given. Yeah. And it's remarkable because wind and solar, it's basically a different product almost than what you're getting. Like, you know, it's all like power, but it's just a completely different. There's so much different dimensionality to it in terms of where it can be produced,

when your discretion in producing it or not, how it overlays on the grid. It's a different product entirely. We could have a three-hour podcast, and people do. I mean, there's people that are passionate about this. There's market design lawyers that go into this, and they have their own thesis that they run with. But, I mean, it's effectively a completely different dynamic, and one in which price isn't necessarily,

price or value aren't the only variables to be optimized for. right we've now injected this carbon-free variable into the market and basically stated okay we need to work towards this and so what that's going to do to power markets that have optimized around other variables you know it's it's really interesting so let's let's start with your background I you know I think that you're one of the most complete voices on the intersection of power and Bitcoin and I think you're going to be a really important one for those that don't know you Tell us about your background who you are and how you came to Bitcoin and sort of your role today. Wow, Nick, making me blush.

I appreciate that. So I'm pretty much what you would consider to be a mid-career transmission planning engineer if I had to give a tagline. So I graduated electrical engineering with a power emphasis. There's basically an army of people like me that operate and plan the grids throughout the United States. We usually work in cubicles. and we either do some sort of analysis or planning on how the grid is going to operate, you know, tomorrow next week, next season. We look years out and we say, okay, this is the generation development that's coming online,

this is what's retiring, and how do we make sure that this all works? So I did that at Urquhart for a while, building, planning, steady-state models. Then I did a short stent at Oak Ridge National Lab, still doing this kind of niche power system modeling. And then most recently I worked at RedCard. which is a renewable developer. So there's all these basically companies out there that their business model is develop wind, solar, storage, and either flip it. So sell it to an eventual owner or take ownership of the facility yourself.

And so I worked at res primarily doing their transmission and interconnection and then moved into their, basically their R&D group, which was looking at more interesting models of development. So co-location with hydrogen, co-location with ammonia. You know, me being a Bitcoiner for the past five years, I was pretty much gung-ho on, hey, we should co-develop with Bitcoin mining, which is basically kind of how I got that job in the first place. And so I worked there for the majority of the time. That was whenever I wrote the articles. And now I work at Galaxy Digital Services doing kind of power markets.

Wow, that I feel like your background is just absolutely perfect. It's really a jack-of-all-trades type thing. You know, I can't sell myself too deep because power markets are so diverse that you can have, you know, you can have a power trader that has worked Urquat markets for 30 years and just knows very specifically, like some dimension of the market. And you can have someone that's, you know, done the same thing for planning or for outages. You know, it's just there's such a diverse market, but I kind of have like a little bit of depth. at various different places.

And I think that's kind of helped me, especially on the educational content. Well, if your depth is like, you know, like a submarine or something, and that, you know, like a power market engineer who's worked Urquhart for 30 years is like the bottom of the Mariana trench, my depth is like the top inch of the ocean. So that's me. I'm like a snorkeler.

I can't go too far beneath the surface. So that's why I'm so pumped out of this conversation. You know, there's a few we've done on the mining miniseries that have just taught me so much. And I'm still learning. You know, I don't know about you, but I think the like power markets and like grid marketplace rabbit hole is like five times deeper than the Bitcoin rabbit hole. I don't know if you feel the same way. Yeah, they're definitely similar.

I mean, the power markets is all jargon. Like there's such a specific set of jargon that people use. And there's a lot of confusion over. over physical delivery of power versus just financial settlement. And so you can lose people like right off the bat whenever you say, we're using 100% renewable energy. And then the person is like, but how does that work? You know, you just kind of like bait and switch.

The electron. Yeah, you bait and switch the electron settlement versus just the financial settlement. And then you take that idea and you run with it for 30 years and that's why everyone gets confused. I'm like preemptively annoyed because you have a hard stop in an hour and I think this, This deserves more than an hour.

So you wrote this great paper called Bitcoin Mining in the Grid. Well, there's two blogs. They're published at the Slushful website. First one is Bitcoin Mining in the Grid. Second one is transmission curtailment behind the meter. Basically, I've been obsessed with this stuff for two years, but I feel like you condensed the whole story into these two blog posts. And to me, it's the number one thing I recommend to people now.

They want to understand the dynamics and the interplay between Bitcoin and sort of like the modern grid with your focus is on it. Urquhart is the main case study, which I think is probably the most, the one we have the most information about. So what actually motivated you to write this? So I've been fighting on Twitter these various different fronts for a while, primarily, you know, since I did work at a renewable developer, and just to set the stage, you know, there, there are plenty of, you know, climate activists out there that have very specific, you know, ideas or thoughts around Bitcoin. I was never really of that need. And I was never really of that nature. You know, I worked at a solar developer, but there was, there wasn't really kind of like a

climate philosophy involved in that decision. That was just a, you know, a position that was interesting to me analyzing the grid and performing work. So whenever I see, but whenever I see people on Bitcoin Twitter are very, it's a very divisive concept, I guess, renewable energy. And so I would try and steal man the argument of like, you know, solar in the grid or something like that. I would try and come at it and steal man that argument and get hit with all this all this blowback of you know oh you've opened pandora's box on you know renewables and stuff like that and so i i basically just spent some time trying to disabuse people of like marginal cost you know like what does it mean to be a marginal cost what does it mean whenever it's

like urcot's building renewables you know it's private development you know things like that and i basically wanted to condense all of those kind of repetitive canned replies into an article and i met zach and will zach and will at like a company this event in Denver. They were having some sort of compass mining event. And I met them. And I think they just kind of like we talked about power for a while and got their contact information that way. And then Zach hit me up a month later. It was like, hey, you're interested in writing a blog post about Bitcoin in the grid. We're targeting like six to 800 words. And my first my first reply was like, well, that's not going to work. Like I'm going to I'm going to need, you know, thousands of words.

And I think, you know, that was one of my main, my main fears is that the articles are so long, you know, like I'm going to lose people that aren't, you know, really, really interested in it because it's so long, but I have to do it justice. And I think, you know, I think the first article is something like 20 minutes. And the second one is something like a 30 minute read for the average reader. But I do think they are good sources of information from like a base layer understanding. And, you know, hopefully, hopefully people get like at least, like you said, just an entry level understanding. into how the grids work and just like the beauty of Bitcoin's kind of alignment on the price. No, I love it. I love these articles. I, you know, I've been trying to assemble this sort of expertise from so many different disparate sources, but there's so few people that are, you know,

truly understand the nature of Bitcoin and then also like deeply, deeply understand power grids. That intersection, that Venn diagram is like a half dozen people. There's, there's really a lack of infographics. You know, there's like a lack of, And any sort of educational material that is good is really deep in some sort of marketplace PowerPoint presentation. Like you might find a great graph, but it's, you know, it takes you two hours to find it. So I have questions, but I want to start with because you're like kind of playing this translator

role. So when you say when you talk to Bitcoiners, Bitcoiners have strong feelings about energy, right? So as you say, a lot of them just like reject the like, like, climate crisis or like the notion of like a renewable transition. They think yes, gee, is like a attempt to, you know, politicize power, which like, in fairness, maybe it is. But and then, you know, you've got like environmentalists and they just like think all Bitcoin mining is evil and, you know, like just awful and and, you know, using renewables isn't good enough. So like, what are the biggest misconceptions you have that you see in the Bitcoin, uh, focus discussion of

power? Like what did Bitcoiners get wrong? about the way mining works and interacts with the grid. Well, I would say, you know, the low-hanging fruit here is the whole Bitcoin as a battery kind of discussion that I think rubbed a lot of energy people the wrong way. And this goes back to what I was saying about how, you know, power is its own industry. Like there's a Bitcoin Twitter, there's an energy Twitter, and they're having their own fights and their own really strong philosophies on moving forward. There's people that basically believe we need to stop.

all fossil fuel generation yesterday. And then there's and then there's a whole, there's a whole continuum of those people basically all the way through to people that say, oh, you know, carbon is bad. We should try to mitigate that. But, you know, we shouldn't really give an inch of reliability concern for that. And I'm really concerned about the reliability issue, right? So there's a whole swath of people like that. And then whenever, whenever Bitcoiners try and come and say, we're adding load to the system, period. We can add 15 gigawatts of load to the system as fast as possible, period. And it's good for the grid, period.

Like, whenever you use those things, there's like nine footnotes after that, like all deserving of their own paragraph long legalese asteris. That's like, you know, not necessarily true. I think, okay, I've got a good answer for you now. I think what Bitcoiners get wrong about the grid is. is that the grid is actually a very sophisticated complex machine that takes a lot of work and like a lot of concerns about reliability to get right. It's not an infinite bus that you can just plug an infinite amount of load to and sustain.

And I think that's going to be real prescient in Urquhart now where basically everyone's been coming to Irkot. But Erkot, the grid operator is looking at these numbers like, hey, we can't actually handle all this. Like we literally cannot handle all this load, but we have a mandate to serve it. So how are we going to do that? There's some sort of legal mandate, right? Like you have to at least do a study or something like that, right? Well, well, Urquhart's duty is to reliably operate the grid.

Right. So how, so if they can't, if they can't perform that mandate, then they have to take action. And you know, that that happens with generators, like a West Texas generic transmission constraints. All these transmission constraints that Urquat has to impose is to reliably operate the grid. Same thing happens on the load side. You can't just add hundreds of megawatts of load to a situation in the grid where it can't handle it. And I think a lot of miners think that the grid is just this infinite source of power that you can take down. And that's true if it's a small, you know, if you're just a one one household plugging in, but if you're trying to do this at scale,

it's not possible. You're just obfuscating the complexity behind it. army of engineers. Yeah, I mean, the numbers we're talking about in terms of Bitcoin miners trying to add load to the Urquhart grid or like gargantuan. I don't know, I don't know what they are, but we're like maybe more than 10 gigawatts and an Urquod is like 80 total, right? Exactly, exactly. So for context, the Urquhart peak load, I think, is something like 76 gigawatts. And that's peak, right? So a normal average high load day, maybe like 50 or 60 gigawatt. and Urquot put out on the large flexible load task force that they have confident numbers of something like 17 gigawatts over the next 60 years.

So you're talking about, I mean, that's a lot. Like that's a lot. Because all of Bitcoin is 15 gigawatts, give or take. Yeah, I mean, you can have tons of discussions on is there even that many ASICs in the world? You know, like, how are we doing? But from Erkot's perspective, you know, they have to believe you. They can't, they can't not believe you. they have to approve or disapprove of these requests.

And this is block loads, right? These are loads that want to be on 24-7. These aren't your traditional load profiles where, you know, there's kind of a peak in the morning and then a peak in the evening. So, you know, talk about throwing a wrench in a grid that's already pretty sensitive to solar, wind, storage, and aging generation fleet of coal and natural gas, you know, things like that. To ask the symmetrical question, like when you talk to, like, regular folks,

about Bitcoin mining, what would you say and energy? Like, what would you say they're a misconception? I have an answer to this, but I'm curious what you'd say their main misconception is about energy and the way Bitcoin consumes it. The biggest misconception that I disabuse others of is that Bitcoin aids to scarcity pricing, which is like most new generation Bitcoin miners are curtailing somewhere between $100 and $200 right now. Now, they're not exactly perfect market participant. which means that they're not exactly turning off right at that moment.

But the idea that a Bitcoin miner is going to add to your all-time peak load is just not true. I mean, they're so flexible in that regard, you know, not even talking about ramping or ramping your load, but just turning off. So that's kind of like the main argument that I see that's most frustrating, I guess, is people saying, oh, they're adding to the problem of the grid. And it's like, well, when there's scarcity problems at the grid, the price is thousands of dollars. Any Bitcoin miner that's any sort of rational actor does not want to be paying thousands of dollars. They're curtailing and they're selling their power at that real-time price.

And that's what a rational actor does. That's a great one. I mean, I guess it depends on whether it's a grid where you have that real-time spot pricing or a more regulated one. But yeah, that's a good one. And even then those regulated entities probably have some sort of curtailment agreement, you know, where they can just ask you to curtail and give you more favorable economics during the off-peak. Or at least I would hope so. Although I think using ERCOT as a case study is interesting because people might think that's sort of typical of power grids,

but like ERCOT's sort of quite unique. That's a good point. Yeah, there aren't that many liquid markets. I mean, there are liquid markets, but, you know, there are certainly non, or there are certainly regulated utilities. where it's much more rigid of an environment. So I haven't even gone to the questions I wanted to ask you yet. This is what I just thought of. So in terms of like, because you're bursting onto the scene now, right?

And I expect you to be like an important voice in this. So where would you situate yourself? You said that there's people that think we should sort of turn off all thermal generation yesterday. And then there's people that just like reject solar and wind. Like on the scale of sort of Alex Epstein to Eric Holthouse, where would you put yourself? Well, I certainly don't think we should give an inch on reliability. I think we shouldn't give an inch on reliability concerns. And I think that our planning models are fundamentally not useful in terms of understanding our risk in turning off reliable thermal generation.

So for example, let's say Urquot, let's say some coal facilities in Erkot want to retire, right? You can't just do that. You have to give the operator notice. The operator has to study the model and say, okay, we think that there isn't a reliability concern for you retiring. You can retire. If not, they have to issue an RMR, which is a reliability must run. So they say, you can't retire.

You have to run. The problem is the planning models that we use nowadays, I don't think are capturing the full, you know, stochastic attributes of the grid at the scale. of renewable penetration. So I would say that I'm not willing to give an inch on the reliability side, but I also think there is a steel man for carbon having a net negative or net, you know,

questionable impact on our world. I don't know to what extent we should. I think that's a pretty intuitive argument to make, but I'm not sure at what level we should impose restrictions on society in order to curb that. and I'm politically fearful of anything. Like I've had some discussions with more liberal, progressive bitpointers about this where I basically say, you know, I understand the argument.

It's pretty intuitive to me, but I'm pretty much ready to disregard any sophisticated climate models that try and tell me that the world is rising at a certain amount of temperature. Because I just don't believe. I think there's a precision error with those models. And they come back to me and they ask me a question, which is, well, what level of confidence would you believe? You know, like what level of modeling would get you there? And honestly, I'm not sure that any level of modeling. Right. And so that really gets me at a rock and a hard place of like, intuitively, it makes sense to me that burning fossil fuels might have some impact on the globe.

But I don't really feel confident in any of the ways that we're able to measure that and then subsequently impose restrictions onto the climate. So I'm not really sure what to do with those two, you know, cognitive, dissonant things. Yeah. Not to mention the fact that it's so politically convenient to rely on these models for basically other agendas, you know. Exactly. Yeah. If you expected sincerity from that camp, you'd expect to see an embrace of nuclear, for instance. And to me, that's sort of like the tell, you know, that like it's actually more a political expedience oftentimes than a sincere concern for the climate, you know. Yeah, the nuclear. And it's also a question of pragmatism versus.

I don't know what the opposite of pragmatism is. You're probably better at that words. Yeah, idealism for that. Like, it seems so easy for Bitcoiners to fall onto nuclear as like, oh, well, we need to hit up a nuclear side's grid. But you look at recent examples of nuclear development, super expensive, super long timelines. It's not like we're going to all of a sudden shift direction and build hundreds of small nuclear facilities across the United States.

Like, that might score points on an idealistic conversation. today if you say we should build nuclear. But in terms of actually getting megawatts on the grid, it's not going to happen. Not in any sort of expediency. So, I mean, like in your view, like there's a, I think there's like an established view basically, you know, which is political orthodoxy almost now, which is we sort of have to transition to like wind, solar and batteries. And that can, you know, so there's the normative aspect, but there's also sort of the descriptive aspect, which is like, we believe, that this can actually be sufficient to power a grid. It's like a full decarbonization on an aggressive timeline.

And you see these ideas are actually put into place in places like Germany and California and stuff like that. Obviously, those experiments seem to be faltering. But what's your view on that? Like, do you believe this can be tractable? There can be a grid powered by intermittent renewables plus batteries, plus maybe some, you know, long-term seasonal storage in the form of hydrogen or something like that. Do I, right, do I think it's possible? Yes, I think it's possible. I think it would be

prohibitively expensive. I think it would be insanely expensive to the point where it would be, at least at the current pace that the technology is progressing. It would be prohibitively expensive, especially to residents, you know, to build the transmission necessary to make up for, you know, the fact that the solar profiles are all the same when profiles are all the same. When profiles are all same. You'd have to build a lot of storage. You know, the storage has some sort of refresh rate that's, you know, in the single decade timeline. You know, you're already getting degradation of storage after 10 years. So you'd be constantly refreshing gigawatt hours of storage in the form of commodity pricing, constantly refreshing panels, you know, things like that nature. Now,

is it technically possible? Yes, I think it would be prohibitively expensive and inefficient in terms of like renewable development, right? Like if you're a solar developer, or a wind developer, your project needs to pencil. And for your project to pencil, you need to be making money most of the time. That doesn't work if you're, you know, cannibalizing your own hours. Like if everyone in Ercot is all having the same solar profile, no one's getting paid. They're all trying to sell at the same time, right? Yeah.

Correct. Yeah. People don't understand the correlation thing. But yeah, when I learned about that. The other thing is transmission. It's like, you know, energy is not globally fungible. That's a misconception people have. and you need this high voltage transmission.

Just like Texas is a great case study there. It's hard to build. It's kind of expensive, but it's also hard because it's hard to build giant lines that go in a straight line across thousands of miles because zoning and because of property rights, basically. So you have to sort of go full China if you want to do that. Yeah, I mean, we're talking millions of dollars per mile. And we're talking years.

And here's the thing is as West Texas is an example, even if you build transmission lines, you can still have problems getting the power back. There's still stability issues. Like there's a whole myriad of concerns here with short circuit current where wind and solar don't provide enough, you know, spinning mass to actually maintain stability on these areas. So, you know, there's basically a whole slew of problems once we look in this bag of solutions that are still going to need technical solutions, still going to need more money to be thrown. You know, it's like, okay, we build a lot more transmission.

Okay, now we need synchronous condensers. Now we need to tweak the inverters on the solar. And this is what I'm saying. I think it's technically possible kind of because I'm an optimist on the technology. But it's expensive. And we're talking like 10 times, 100 times residential rates. Wow. Yeah.

So it's like the political popularity will, the rubber will hit the road at some point where people realize what this is going to cost, basically. I think you're having that reckoning in Europe right now, you know, in certain blue states in this country right now. Yeah, especially as more cost goes toward the fixed infrastructure and less on the marginal cost of fuel. You know, there's a good guy I follow on Twitter called Duncan Campbell. Yeah, he's smart.

Yeah, super smart guy. He's part of like the DER Task Force, which I think is a funny little group of people that are all interested in this stuff. And he put out a good paper on how the majority of the cost is moving from, you know, fuel from like the marginal cost of production, which has typically been coal or natural gas and over towards the fixed infrastructure side. So we're locking in costs more. They're up front, right? Those are up front. And then you depreciate them. Yeah. So that's renewables are always seen as like, well, you have no marginal cost to produce. But the flip side of that is it's up front.

Exactly. But that's an, that is an interesting characteristic. And we're probably going to get into that with regards to Bitcoin is, you know, if you can pay all your, a majority of your upfront cost and kind of remove your sensitivities to fuel cost that has some benefit, you know, in terms of locking in a forward price for your energy. So now does that scale to a society? I don't think so. So you, you know, an energy pragmatist, I think, and with, with reasons. So I want to talk about controllable load.

This is one thing where like maybe Bitcoin has went too far in touting it. I don't know. Curious for your view on that. Do you think that the narratives around demand response and controlable load in Bitcoin are overcooked? Are they undercooked? Are they the correct amount of cooked? I think it's interesting. You know, I mean, the control-belode argument is the same thing that storage can do, right? Storage can also do these things. The main problems that I've seen is that Bitcoin miners have been baking in these

payments into their models in terms of their power price. And they've been assuming they've been assuming that they can take these payments as part of their general strategy. I think once you actually get down into it, it's actually very hard to control your load. And it's especially hard to control your load within margin after receiving an instruction from a grid operator to then just subsidize your power cost in terms of also cycling your machines. So now you may have taken a $10 to $20 per megawatt hour, which would be, what is that, one to two cents a kilowatt hour subsidy. But you've got all this software overhead now. You're cycling your machines, and you're also kind of cannibalizing the market. So you're also bidding this service into an auction in the day ahead market, which is then,

in terms of demand response, ancillary service payments, that is also going to cannibalize itself and also competing with storage. So I think because it's all, like you're saying, this is like a finite amount of sort of insurance the grid wants to buy, and then that's divvied up between storage providers, other load resources and to the extent everybody starts buying this, selling it at the same time, they're all competing with each other for that fixed payment. That's right, correct. And there is an argument to be made that maybe ancillary services will become a bigger

market once we have more renewable energy. I believe that. I mean, you have to buy more insurance. I think Sean Connell said this on my podcast. If you put a lot of homes in like a floodplain, you're going to have to start buying a lot more insurance. Actually, maybe the insurance won't sell you insurance. at that point.

Maybe it's about analogy. But like, you know, you're going to have to buy a lot more insurance if renewables are a big part of the grid because your grid's just less reliable at that point. There's a good, there's a good, you know, secondary thesis here, which is that if, if the Bitcoin miners provide this demand response as a central part of their strategy, but they also clearly create more net generation onto the grid, right? So you're raising your total capacity on the grid and you're paying for that with load that is completely curtailable, then that's a steel man.

Then you can say, look, we've added reserve margin and we're also willing to curtail very easily. If you're just adding load to the grid but not adding any additional generation, you're kind of just cannibalizing your own self on the on the dispatchability side. One thing that I never fully understood is, so there's a bunch of different ancillary service products. And as you say, it's like intense amounts of jargon, like unreal amounts. So like control below this, like I think demand response is sort of the general term. And then you have sort of like different, is that fair to say and like control below be a subset of that? Yes. So and I hope you don't put me on the spot here. So demand demand response generally,

the way that I frame demand response, right, is basically you're receiving a signal from a grid operator. that you respond to and you're on the demand side. Do you include like economic sort of self curtailment, spontaneous or organic curtailment in that? Yeah, I would. I would. And I guess that, you know, completely negates my exact definition before

because it's not exactly a signal from the grid operator at that point. Not, I don't. Well, it's an economic signal. And they kind of make the market, right? They influence the market, right? Right. Exactly. And that's, that's interestingly two sides of the coin,

because most of the time, if you are participating in demand response, you have sold your service, you know, forward. So you might sell your service today, tomorrow, and then tomorrow you need to be online such that you can respond. The problem is you cannot, you can't now respond to the price signal. Like, you now have to be a taker. So there's kind of an interesting predicament there, where it's like if you want to self-cortale and sell your power at the real-time price, you cannot have.

sold your services in the day ahead market, which I think was actually one of the questions that you sent me earlier that we were going to talk about. As I understand it, Urquot, because I think Urquat has like a pretty well-developed ancillary services market. And I believe one of the reasons for that is because it's pretty islanded as a grid. And so they can't rely on intertize. Is that, is that a good reason? Ooh, the reason for Erkot's ancillary service market. I don't know, that's kind of a loaded question. I mean, it's also an energy-only market. So with the energy-only market, you kind of have to have more firepower to deal with real-time scarcity, I'd say. So that might account for some of the bigger non-spin responsive reserve services.

I know about the reg up or heck down. So there's like four, I'd say right, there's like four different ancillary services that generation assets and load resources can sell into. Am I getting that right? All right. So here's the quiz, right? So we've got regulation up, regulation. down responsive reserve and non-spin, I think.

If you're putting me on the spot here. I think that sounds right. These are all different time frames, right? And I know we're running through this pretty quickly, so I apologize to listeners. You're going to have to replay this if you're interested in it. But there's regulation up and regulation down, which is basically responding super fast seconds. And that is you controlling your output to either raise frequency up or raise frequency down. then there's responsive reserve which is a little bit slower so this is like minutes so this basically says

okay you know maybe there's a faster load ramp than when we expected or maybe solar is ramping down faster than we expected and so you need to be able to respond on like a minutes time frame to change your output to help us out and then there's the non-spin which is you know in traditional terms is when a generator goes from not spinning at all to spinning really quickly and so that that means, you know, that's on also like a 10, 30 minute time scale. So those, those are in my head. Okay. The insulator service markets that you could bid. And these were generally designed for generators, in particular, like thermal generators, like gas or even oil plants, right? Right. Like petroleum that can spin up and down on short notice and like they're very carbon emitting,

but they're the most reliable and sort of like they can react really quickly. Is that right? Correct. Yeah, as opposed, the good juxtaposition here is coal or nuclear, which can't really ramp up or down and are even willing to submit negative bids such that they don't get turned off, right? That would be the opposite. Yeah, nuclear is like the classic. It's like a staircase. Like maybe they're off for maintenance once every quarter or something, but they just don't change their output. That's right. Yeah, a good heuristic for nuclear is that, a nuclear's base output is 100%, and it actually costs them money to use control rods. So they would rather pay some amount of money versus use control rods.

That's like my favorite kind of rule of thumb. So I guess like when now you have these load resources, which are like sources of demand that can modulate their demand up and down, now all of a sudden they can also sell into the ancillary services market, which is kind of like a new and interesting thing. alongside storage, I guess, as well, right? Totally. And in the broader context, they're also bidding out the thermals that could be relying on these kinds of sources of revenue as well to keep up their business.

So it's an interesting wedging out of all the services that thermals used to provide. Because it costs a Bitcoin miner, maybe less in terms of opportunity cost to dial down their consumption, then it would cost a thermal generator spin up. Is that correct? That would be the assumption, and that would go into the economic modeling of bidding into the ancillary services. How many dollars per megawatt

are you willing to sell this service in the day ahead for? And so I guess the assumption would be that, yes, it would be cheaper for the Bitcoin miner to do that, especially if they can count on not actually being deployed a majority of the time. They're just selling the capacity. Because that's the interesting thing is, because you know,

you think about the miners, a synthetic power plant. Their idle state is that they're mining Bitcoin, right? They're drawing power. They're mining Bitcoin. Now a standby generation asset, their idle state is that they're doing nothing. And then when they're called upon, they have to spin up. Is that you're shaking your head. Am I getting this wrong? Well, it depends. It depends on the type of plant. Like if you, if you are, right, they're totally different diverse business models here. Like there's, you know, the term base load. Like if you are a base load power plant, you want to have high capacity. factor you want to be on all the time you're willing to bid negative because you don't

want to get turned off because then it costs money to be idle to be cold and then restart so there's business models like that and there's everything in between where some combined cycle plants model you know some percent capacity factor where if they're below that capacitor factor they actually start to not make money anymore and it makes sense for them to retire this is why you see this is how you see thermals retire they're just not being dispatched enough for it to make economic sense for them. And then you have things like peekers, which are like pretty much what you're saying. They're willing to sit idle 90% of the time because they can just fire up during the 10% of

their time and make their revenue targets. So I guess my question, that's something I've never quite understood for load resources that are participating here and they're curtailing. When they're participating in a formal program where they get the order to curtail or diminish their load, is that a physical switch that the, like, ERCOD or the grid operator can switch? or is it they just send them an email or something and ask them to turn down? So I think this might be in regards to the specific precise orders. This might be a little bit outside my wheelhouse. But I believe it's through something called a qualified scheduling entity,

which is more like your agent that interacts with the operator. So the operator, Ercot, would tell your queasy, hey, we need you to do this. And then the queasy is the one that actually, you know, submits the command or the signal to do the thing. And then that would be basically Ercot tracks your output. So Erkot's meter tracks what you are doing in response to their signal. And if you don't act within a timely manner, within some predefined set criteria of, you know, scale of your output changing and time of your output changing, then you can receive fines. Yeah, so there's intermediation, I guess, for a start. So you're not directly talking to Urquat.

And so this is part of the debate because people talk about this being voluntary. What you're describing it is like you as a load resource, you qualify for these programs. You establish that you're sort of able to accommodate the requests. And then when a request comes in, you pretty much have to sort of comply with it, right? Correct. Yeah, there's a clear juxtaposition here that I want to highlight, which is that this is a very formal program right here. Like load resources of very specific terms, capital L, capital R. And you can participate in specific markets like this.

There's also, on the other side of this, there's just playing in the forwards power market as like an energy arbitrage or I guess if you say. So if you're a Bitcoin miner, you can buy power forward at $50 a megawatt hour for the next five years. And then basically you either use that power as 50 bucks or you can turn off and you can sell the power that you already bought for $50 at your load zone price. That's a separate thing, right? You don't really. Depending on the conditions on the day, that price could be way higher. Exactly.

So during scarcity event where your load zone is $2,000, right? If you bought power for $50 and you can sell it for $2,000 now, why would you be mining Bitcoin at $100? You know, you wouldn't. And so that's a separate thing. You don't necessarily have to communicate with ERCOT, but this is one of the problems that I talk about in my article, which is that if you have everyone doing that at the same time, that's bad for grid stability because you're toggling power up and down on like hundreds of megawatt scale.

And a grid's not designed to do that. It's not, you know, it wasn't built with those engineering requirements to be able to handle huge loads that can swing up and down like that. And that's one of the arguments that I make for steel manning this and making sure that it's a good resource for the grid is having to be transparent. So you want to formalize these entities, these loads, which are responsive to conditions on the grid. Is that what you're saying you want to bring them under the umbrella of these formal programs? Not necessarily. I think what is necessary is more of like a transparency in how you operate.

So for generators, right, you have to submit something. ahead of time called a current operating plan, which is basically you telling Urquot, hey, over the next week, I plan on being online today, offline the next day, online that day, and then Ercot takes that plan and then they facilitate with that plan how they wanna dispatch you.

Orcott doesn't actually order you to be on or off. I'm thinking something similar to that where here's our current operating plan. This is our break-even price. That might be something that is transparent enough to help the grid out without giving away too much of the secret sauce, even though in terms of, Pat, and I could keep on going with the corollaries,

I'll have to stop myself. But giving away your break-even price is like a big transparency issue with Urquot, because as a generator, that's your secret sauce is what your fuel cost is, basically. So just to summarize, because this confused me for a while, the notion of economic dispatch versus like formal enrollment

in like a DR program. So to summer, that miners are responsible to like grid scarcity of power prices in like two ways. Like one is just by being active in the market and turning off when you your mining is unprofitable relative to selling power back to the grid, which is a function of like the hash price, like price of Bitcoin, price of energy and things like that and difficulty. And then the second thing is you can literally enroll in a program. Then you just kind of outsource that decision to the grid operator and they sort of ask you to turn off at specific points. And so,

So that's kind of two ways that mining is like not necessarily contributing to scarcity at these peak times. Is that fair to say? Yes, it's fair to say. And the jargon that people use for that option one would be power trading. You would theoretically be trading the power that you bought forward. And that would be the universe of that world, whereas the other world would be demand response. Okay. And here's a weird interesting question.

to the extent that a lot of hash rate ends up in Texas. I mean, I think there's probably just south of a gigawad of Bitcoin mining in Texas right now, and then maybe in a few years' time, it'll be a couple. Do you expect that hot days in Texas actually ends up affecting block arrival time in Bitcoin? You know, it's funny. I've had this conversation numerous times. Really? That's such a...

That's like the most niche question of all time. It is. But, you know, I mean, your brain relatively ends up. here because the question is if it gets hot and all the miners curtail are we actually going to see a noticeable yeah i don't you think so i think so i think so and i think i think that'll be more of an attribute of like the 2040 grid as as we get more bitcoin vertical integration into renewables that basically needs some sort of put for their energy i think that's where you're really going to see a noticeable effect i think we're still kind of in the in the noise range now where you might be able to

to make a claim using some data, some cherry-picked data, but I don't necessarily think we're there yet. Theoretically, yes, I think, and I think there will be a correlation between congestion and block times. Congestion on a power market and congestion in the Bitcoin now. I love how, you know, eventually, and I actually do think this is going to happen as like miners start using a lot of solar and things like that in the southwest of America maybe, you know, in a single geographical location, not one time zone. I kind of agree. I think you're going to get a lot of mining during mid-morning to like bit afternoon when it's very sunny and solar is cheap and people aren't using that much power maybe and then you know if you get enough miners located there then you get

blocks arriving faster basically yeah the the commodification of bitcoin mining machines would be huge here like if we if we could get a6 to decouple from their opportunity cost and go back to just the raw materials that would be a huge boon for this because we would maybe drop the utilization rate percentage requirement. Right. But yeah, I completely agree. Yeah, and that's an interesting dynamic that people haven't really commented on much is, as A6 become cheaper and they become like a lesser portion of a minor spend,

then your ability to, you know, exploit maybe more renewable and intermittent sources of power becomes greater. Correct. Or more niche uses. Like, like, yeah, like DC clipped energy on. solar farms right like a like a solar farm might have a 1.3 DC to AC ratio so that 100 megawatts solar farm is actually producing 130 megawatts of DC power but that 30 megawatts never escapes the inverters so if there's a way to get on the DC side

you're talking about tons of wasted energy but it's so expensive to do a DC to DC conversion and then you're only getting power at a certain amount of time so like use cases like this that would just spring up of securing wasted energy are still kind of prohibited just based on how expensive these machines are. So I wanted, I have three big topics. I want to talk about controllable load, monetizing renewables and then mining behind the meter, which you talk about a lot in your blog post here. We managed to only hit one of those topics. So I think we are going to have to do this again. I mean, to the extent you're willing. I want to end with some general comments here. So like in

terms of ERCOT, like they hit the pause button on, you know, they just, I think they hit the pause button. They're doing kind of like, okay, let's slow down. Let's like reassess here. Like people talk about Brad Jones interim CEO has said some favorable things. What do you make of ERCOT's sort of current attitude to Bitcoin mining? Like how sophisticated are they about it? I think, I think they are. And by they, so we've got to be specific here. So, so ERCOT, the people that work there, right, they are primarily. concerned with reliability. It's their job to keep the grid up and planning and operating to that end. There's also just the other market participants in ERCOT who are worried about it, other loads,

other generators. I think in general, the people that I've interacted with at ERCOT are optimistic about the Bitcoin mining that's coming in. But they're also cautiously worried at the amount. because the sheer amount of megawatts coming in is scary. Like in terms of what it's going to do to power prices, what it's going to do to the stability of the system depending on how concentrated each megawatt is. And also considering how they don't have any real guardrails on the power trading. Like we just talked about the power trading aspect of turning your system off.

What if you add a 500 megawatt Bitcoin mine that you can just be, you know, effectively flip a switch and turn the entire thing off in minutes. That's something that's never before been seen. But there's load that concentrated coming online into Urquod. And I think they are, you know, cautious about what that can do to their system if it's not transparently, you know, coordinated with the grid operator. I think that's where we're at. And you say they've hit the pause switch to be, to be specific what they've done.

So basically, when you, when you submit a load interconnection request, They basically grant it. They say, okay, you want to interconnect 100 megawatts here. We'll do a study and make sure it looks good and then we'll say, okay. But what if they do a study and they say, oh, you want to interconnect 300 megawatts, that causes violations, right? Like there's not enough capacity to serve you. What they can then do is grant you interim status.

They say, okay, you want to do 300. We can give you 50 and then we're going to have to rebuild this transmission line. and then we can give you the 300, but we don't know how long that'll take. And that's pretty much what they're doing now. They've hit pause on just the cart blanche, and they've moved into more of a slow, slow response, which I think is, you know, I mean, in fulfilling their mandate, it's probably what they need to do. So I know you have to hop here, so we're going to have to leave the remainder of the questions for future conversation, but which I definitely want to have.

My last question, given your prior work, you know, you mentioned, I think, hydrogen and ammonia, and things like that. I'm actually seeing some miners indicating that they want to not just mine Bitcoin, but also do these kinds of things. Like, you know, you build out a high energy infrastructure on maybe like these stranded or underutilized energy assets. Maybe you mine Bitcoin. Maybe you do something else, right? Maybe you do some other location agnostic industrial load. Is that compelling to you? Or do the numbers work? Is green hydrogen? Could it be, could Bitcoin miners end up doing half Bitcoin, half hydrogen? It's compelling from a narrative standpoint.

I think, you know, there's a very strong narrative on the renewable side of doing green hydrogen because it also aids in the energy transition. Right. It's like the green hydrogen is going to replace. Because it's storage and it's a fuel, right? Yeah, exactly. It's an open-ended thing. Now, in terms of the price, in terms of the economics of it, I don't think you're going to see

sizably profitable green hydrogen without some sort of priming of the pump from the government, whether that takes the form of some hydrogen production tax credit similar to how they primed the pump for wind or some sort of ITC, how they prime the pump for storage. The economics on paper for Bitcoin pencil out the gate. And so this is kind of the interesting scenario is that you might see miners or other market participants kind of try to round out their portfolio of services to include green hydrogen, green, you know, ammonia, things like that. But then when you actually look at the financials, you know, it's like 95% Bitcoin mining and then like maybe 3% of a pilot project for hydrogen

just for the narrative. You know, I think there's a little bit of headline risk associated with saying that you're a Bitcoin miner and any sort of diffusion of that risk to kind of put on the energy transition hat is going to be at least part of the system. Yeah. I think like what is the total amount of like green hydrogen base load globally. It's like less than a gig a watt maybe. It's not a lot. Yeah, I'm not really up to date on all the colors of the hydrogen, but I think a lot of the hydrogen, there's like, you know,

blue hydrogen, black hydrogen is all sorts. Oh, I didn't know. Yeah, there is. I think blue hydrogen is the one that I'm thinking of, which is basically you're still using the steam methane reformation process, but you're buying wrecks. Oh, that's cheating. That doesn't count.

You're netting. I know, it's cheating. And that's another part of this whole industry that's interesting is like we've basically already skinned the cat on I'm using renewable energy because people have been saying they're using renewable energy for like 10 years. All they've just been doing is buying the carbon offsets. So now whenever you're actually using renewable energy, you have to be like, no, but I'm actually using it. Like I'm not just buying wrecks, right? But we've already kind of, we've already kind of burned that hole. But yeah, the actual green hydrogen is pretty low.

And that's because it's not exactly competitive with the other hydrogen. It's like double and triple the cost. So the question is, do I mine, do I, you know, do I buy electrolyzers? Do I do electrolysis at pressure and then store hydrogen and then ship it to an offtaker that's bought hydrogen 15 years forward at double what it could for the competitive hydrogen? Or do I just plug in some A6 and start mining Bitcoin off the gate? Right. Blake, this has been fantastic, incredibly edifying. We're going to have to do it again because I, you know, I want to talk about the behind the meter stuff as well.

Yeah, there's so much here, Nick. It's great to talk about with you. Thanks for being flexible on the schedule. It's taken us a while to find the date, but I appreciate it a lot. Okay, hopefully the next one's easier to schedule. Thanks again, man. Talk to you later.