WBD492 Audio Transcription

Bitcoin Energy Markets with Shaun Connell

Release date: Friday 22nd April

Note: the following is a transcription of my interview with Shaun Connell. I have reviewed the transcription but if you find any mistakes, please feel free to email me. You can listen to the original recording here.

Shaun Connell is Executive VP of Power at Lancium and an energy trading expert focused on incorporating Bitcoin mining into energy transition services. In this interview, we discuss the fundamentals of energy trading, the need to balance energy supply with demand, and the symbiotic relationship between Bitcoin mining, renewables and the energy grid.

“You can put on this Bitcoin miner attached to your nuclear facility… and since your Bitcoin mining site can be dialled up and dialled down, for how it’s dispatched, you’ve just put a supercharger on your nuclear plant; you can now actually provide flexibility to the grid because you have the ability to ramp your nuke, through the Bitcoin mining.”

— Shaun Connell

Interview Transcription

Peter McCormack: How are you doing, man?

Shaun Connell: I'm good.  How are you, Peter?

Peter McCormack: I'm good.  When Nic Carter says to me I need to talk to somebody, it's an instant invite on the show, instant.  Nic reached out and said, "You've got to talk to this guy, you've really got to get him on", and we had a great dinner the other night and it was great to talk to you.  We're going to talk a lot about energy today, it's highly relevant to Bitcoin and mining, but I think it will be really useful to dig into a bit of your background first.

You were an energy trader, so the insights you get into the market from that, I think's going to be super-useful for anyone listening.  So, do you want to dig into the background?

Shaun Connell: Sure, so first, thanks for having me here, and a lot of gratitude for Nic as well; he's made some great interest for me, so I'm very appreciative, and I've enjoyed getting to know him.  So, my background is that I think I'm one of the few people that's in the Bitcoin space that was energy first.  My experience so far has been that there's a lot of folks in Bitcoin that are hardcore into Bitcoin and they're forced into learning the energy side, because it's such a big part. 

So for me, I started in the energy industry in 2002, and it was just when the energy markets were being deregulated.  So, it was essentially going from vertically integrated utilities to open, competitive markets; so, it was really unknown how these markets would develop.  It kind of feels similar to how Bitcoin is now, which is essentially, how are these markets going to develop?  Are they going to be hashrate markets?  Are there going to be some futures?  Who knows, but how is this industry going to evolve?

For me, it was starting in 2002, and over the next 15 to 20 years, essentially I helped build, and eventually led a trade desk that was responsible for optimising a portfolio of generation assets, as well as proprietary trading in power markets.  So, my real strength and experience is understanding the fundamentals of power markets, so we say, "What happens when you put in 2,000 MWs of wind in a certain area?  What is that going to do to price?  What is that going to do to behaviours of people, and what are they going to do after that wind's put into that market?" so, these price signals drive behaviours.

So for me, back in 2017, I think it was, when I first heard about Bitcoin, it really came down to a friend called me up and said, "Hey, this is Bitcoin".  I said, "That's amazing, tell me about it", and he gave me the digital gold narrative, and I really gripped on to it.  And then I said to him, "Well, how do you buy Bitcoin, or how do you get Bitcoin?"  He says, "Well, you can buy it and it costs $10,000, or you can mine it".  I was like, "Okay, well how much does it cost to mine?" and he said it was, at the time, maybe $200. 

It was crap, there's a big disconnect there, because in commodities markets, over time, prices always converge, so the cost to buy versus just going and buying yourself versus make or buy was so big it was like, "I really need to understand the mining part of this, because if I'm going to be getting interested in Bitcoin, I really have to understand those pieces". 

So, that was my intro to Bitcoin, through the energy side, and then I really needed to dial into how does mining work, so that you know if prices are going to shoot to $200,000, what does that mean for -- what are people getting paid; is it sustainable or not?  Because Bitcoin mining is producing hashes, which is just a commodity.  It's the same thing as power generation.  There's say gas pipes, there's natural gas in, power out; Bitcoin mining is power in, hashing out, so it's just another one of those things.  So, that was my intro to Bitcoin and my background.

Peter McCormack: Okay, can we dig into the trading desk side of things?  I don't understand any of this.  Just explain to me what the desk does, the people on the desk do, what their goals are, their objectives.

Shaun Connell: Sure, yeah.  So the thing about it is, before deregulation ended, let's say you're a power generation company and you have these generation assets and you're guaranteed a rate of return on those assets.  All you've got to do is produce the power and you're going to get a guaranteed, call it 8% or 10%.  So, who you're selling to is pretty much like the utility commissions to say, "We need to add more generation", you build the assets, you get paid.

With deregulation, what happened was that model went away and they said, "We need to have competitive markets where it's not just one supplier supplying the energy to everybody".  They wanted to have competitiveness happen, that generation has to compete with other generation.  So, now again, you're this power company that used to have "rate payers", they didn't have really any choice, and now you're going into this deregulated market, which means that you now essentially own this power generation and you need to sell it.

So, with this deregulation, what happened is that they came up with these new power pool structures, so ERCOT is one of those, PJM, New York ISO, and essentially they're auctions for the power and there's a settlement price.  So, if you did nothing for any type of hedging or forward-selling of this generation, you're going to take whatever the spot price is, the auction clear, for your power.  So, the reason why you might have a have a trade desk is that you can essentially do the eyes closed and say, "I'm going to run my power generator and I'm just going to take whatever the spot price is", or you can find ways to do hedging, so they're going to do long-term power sales. 

You're going to try and find ways either through financial markets, that you can sell their power; you're going to find industrial customers, such as Bitcoin miners, that will take long-term offtakes.  The thing about it is, if you are running a business and you are, say, a public company and you want to have very low volatility, you want to replicate that utility model, you want to know what your costs are and you want to forward-sell your power generation.  So, you don't want to have any kind of exposure to the spot price coming low, is it high; they'd rather just lock this off.

So, as a trade desk, the responsibility is, on one side saying, "How do we now essentially monetise these generation assets that used to be just a rate of return; how are we going to do that?" and it was kind of unknown, and what will the markets look like.  That's your existing generation.  So next you need to say, "Where is there opportunity now for us to create new generation assets?" so think about this in Bitcoin mining, but inverse.  This is where they're saying, "Where can we add power?" and mining is, "Where can we subtract it?"

So they say, "Okay, we need to find out which markets have favourable market design that's going to allow us to put in a new power generation asset, into an area that we have a high degree of certainty that we're going to be able to recover our investment and then make a certain return".  So, by having trade desks that are actively in those markets and really understanding those, you can make better decisions on saying, "Where is the market that is going to give us the best upside return on this new generation asset?" and you want to not have to rely on a consultant to say, "Hey, you should put it here", you need to know yourself.  So, that's kind of an overview.

Peter McCormack: Okay, I'm going to get you to walk through the absolute basics of this, just so I understand.  So, a power generation asset, give me an example of what one might be.

Shaun Connell: Sure.  You can use a coal plant.

Peter McCormack: Okay, so a coal plant, you burn the coal, you create power, and you sell that to, what, the local grid, but maybe the local companies; so, you can sell it to both?

Shaun Connell: Sure.  So think about it's like there's always two transactions that happen: you are going to produce the power, and you are going to inject that power into the grid.  And you've got something that's called your "point of interconnect" and it's a meter that's going to measure, how much did you inject into the grid there?  And then if you're a new market, such as ERCOT, you're going to receive the real-time, or the settlement price for that location; that's your payment.  So, I did nothing and there's this meter; all I've got to do is inject into that meter, and then somebody's going to give me a paycheque for what was produced.

Peter McCormack: And is that a daily price?

Shaun Connell: There's a day-ahead market, where you can sell into the next day, which is like a financial market; and then, there's a real-time market, that can be broken out into five-minute increments.  So, there's a price power for every five minutes of the day.

Peter McCormack: Wow.  We'll just grab the coffee!  Okay, so you can either be day ahead, or five-minute increments, but you've got an idea of the price you're going to be getting; there's, I guess, a standard-ish price that mildly fluctuates and you kind of know your cost, so you kind of know your daily-ish profit on spot?

Shaun Connell: Kind of.  And the closer you are to that day, the better you know.  You don't know, for example, a year out, what is that price going to be on the day a year from now.  But for today, I've got some certainty around what is the demand in Texas and what is the generation that's available to meet that demand.  So, I've got some good indication for what today will look like, because of that mix.

Peter McCormack: But how many potential customers can you have, because for example, if you're producing all this power today and the grid's buying it from you, and maybe there's another local business who wants to buy direct, do you transmit it direct to them, or do you have to do it via the grid to them?

Shaun Connell: I love these questions, this is good!  So, as I said, there's two types of transactions.  So, every power generator at least has one, and that one transaction is that they're generating power and it's injecting the grid at that point.  They're just essentially -- think of the grid as a bunch of, call them like pipelines, and they're dumping something into the pipeline, and they're getting paid at that exact point.

The second transaction that happens is some type of financial transaction, or a sale to somebody else; but the electrons, as soon as you inject those electrons into the grid, you can't transport those exact electrons to anybody.  It's essentially getting mixed in with everything else.  So, even if though, for example, you can be located right beside a hydro facility, within a mile of that facility, and you would think that you're pulling the hydro generation, because I've got a transaction with you here at the hydro generator.  And you want to de-risk, you don't want to just have that spot price of energy. 

You want to say, "Do you know what, I want to sell a fixed price, because I don't want to be exposed to them", so you'll make a transaction and say, "I'm going to sell you my power for $40".  I say, "Okay, I'm going to buy it, you're the hydro facility, I'm a mile down the road.  You're going to inject your power into the grid and you'll get paid for the power that you're injecting, but our transaction is going to cancel out your payment there, and you're going to receive the one I'm getting that's going to be given to you.

Peter McCormack: All right, so I don't actually ever send the power direct?

Shaun Connell: No, you can't.

Peter McCormack: You're not allowed to?

Shaun Connell: It's just impossible.

Peter McCormack: It's impossible to build the infrastructure to do that?

Shaun Connell: Well, the way you could do it is if you're co-located, meaning that you never injected it into the grid.  It's like you have a hot dog and it's loaded up with ketchup, and then you put a little bit of mustard on that.

Peter McCormack: Yeah, sounds good!

Shaun Connell: Then you're a mile down the road and you're like, "Do you know what, I'm just going to take the ketchup from the hot dog, I don't want the mustard"; it's impossible to take it off once it's in, so you can't separate it.

Peter McCormack: Right.  So, it all goes into the grid, but I can do separate transactions?

Shaun Connell: Yeah.  So, the other type of transaction that can happen, and you're starting to see this with interest on Bitcoin miners, and this is pretty standard for other types of customers that would be like industrial customers will do.  There's a power generator here and it's going to get built and I need to have steam and I want to buy energy and we're not even going to connect to the grid.  So, the power generation company is going to say, "This plant is producing power in steam, and I'm just going to send it to you using our own wires", it never hits the grid.

So, you have 100% certainty that the power that that was generated from, and if it was from a coal plant, you know you're consuming coal; if you are co-located with a windfarm, I'll use an example of saying that there's a 100-MW windfarm, and the power market determines the capacity factor, and what that means is that, what is the capacity factor of a windfarm?  It's about 33%, it's different in different areas.  What that means is that on average, over the course of a year, 33% of the time -- sorry, at a 100-MW windfarm, you'll have an average of 33 MWs.

So, you can be co-located with a windfarm, essentially consuming just the power from the wind and just consuming as produced, when it's producing, and you know it's 100% certain that that came from wind.  But as soon as you connect to the grid, you have any connection, then you lose all the ability to say, "This is 100% wind [or] this is 100% solar"; it's more of a high probability of the mix of electrons where those source from.

Peter McCormack: But if I'm here in Texas and I'm plugged into ERCOT, I can't do a deal and sell energy to a company up in Minnesota, or can I?

Shaun Connell: You can, and so it's just a financial transaction.  So, you're making a financial bet, if you will, on the settlement at some other location.  Then, what you are taking on, if you make the sale, is you are taking on the risk that the price between where you are and the price where they are is not going to expand or be significantly different.  So, there's a lot of risk in that.

Peter McCormack: You see, this is the bit I don't understand, because I understand putting into this grid and the company's connected to it, but I'm not actually putting any energy into -- these electrons are never going to be connected to whatever's up in Minnesota, right?  So, how does that actually work?

Shaun Connell: So, it never gets to Minnesota.

Peter McCormack: It never gets to Minnesota?

Shaun Connell: Never.  So, it's just a financial hedge, where you think that you are --

Peter McCormack: Am I buying energy in Minnesota?  No?

Shaun Connell: So, think about it like this.  There's oil that's -- the contract's called WTI, which is the oil from the United States; and then there's Brent, which is the oil in Europe.  I can be producing oil here, right next door, in Cushing, Oklahoma, this WTI location.  So, I'm going to be producing oil and I need to find a way to say, "I want to hedge my production risk", so I don't want to be just exposed to the spot price, or whatever it is.  But the only person that I can find to trade with me to hedge my risk is somebody that's at Brent.

So, you have 100 barrels of oil, the price of WTI is $100 and the price at Brent is $110, and so you sell them Brent contract, so that the oil never made it there, but what you're betting on is that the difference between WTI and Brent isn't going to expand so much that it's actually going to cost you, because there's just a financial settlement with the transaction that you bet on Brent contract.

Peter McCormack: Right, okay.

Shaun Connell: Energy markets are tricky, right?

Peter McCormack: Yeah.  Well, I'm still not -- do you understand it?

Danny Knowles: Not fully.  It's basically just a financial instrument, rather than being anything to do with giving anyone any energy?

Shaun Connell: That's right.

Peter McCormack: But say the company in Minnesota, I sell them at $40, they pay me for it, they give me X amount; I don't understand how I settle with where the actual power is generated from in Minnesota.  I'm getting these $40 in; where am I actually paying?

Shaun Connell: So, there's a physical market and a financial market.  The physical market is here in ERCOT where you're injecting.  And then, the financial market is that you are somebody in Minnesota and you would choose a location in Minnesota, and you would be betting on some price that's a Minnesota price.  And what's going to happen is that, say that the price in Minnesota -- say both Texas and there were $40 and you said, "Okay, I'm going to do a financial trade at $40".  Come settlement, that settlement is going to settle against the real-time spot price, so that is the settlement index.

Peter McCormack: Oh, okay.  So, it's like a bet, it's just a bet.

Shaun Connell: So, you would run your risk analysis on saying, "Is this bet highly correlated with the location that I'm actually producing at, so that it is actually risk-reducing, not risk-adding?"

Peter McCormack: So, I essentially buy contracts and then I sell contracts?

Shaun Connell: That's right.

Peter McCormack: Yeah, I get it, I think I get it.

Shaun Connell: Again, back to physical and financial, you're always pushing the electrons in the grid, and you're getting a physical price.  There's an index that's selling on that real-time price, and then financial contracts are structured on the settlement of what they think that real-time index price will settle.

Peter McCormack: Okay.  There is one bit I still don't understand, I must sound like a dummy!  So, when you talk about energy, you talk about there's an input of, say, burning coal, and there's an output of electrons.  I understand that if I sell contracts to a company that needs it at $40, I'm essentially getting an input of cash from them.  I don't understand where they're getting their electrons from and how that's being paid for.  That's the connection I don't understand.

Shaun Connell: So, you're getting paid index and then they're receiving the index, because you're going to make a trade that they're going to receive that index and pay you the fixed, because you don't want to take on the index price, you want to take on the fixed.  So, the payments will be like, you inject the power into the grid, they're going to pay you index; and with your counterparty, you're going to settle on saying, "I'm going to give you the index price, because I don't want it.  I'd rather have your fixed price".  So, you're transferring the risk from you to that person on the settlement price.

Peter McCormack: So, am I essentially hedging across multiple markets to give me more stability where I'm producing power?

Shaun Connell: If you have a portfolio of generation assets in, say, Texas, you want to have your hedges as close to the location where your generation is as possible.

Peter McCormack: Right, okay, I see.

Shaun Connell: You don't want to have them spread out and say, "I'm going to do some bets in New York"; you don't want to hedge your Bitcoin exposure with Dogecoin, right!  That's a basis, right.

Peter McCormack: That's fair!

Shaun Connell: But when they're all going up, they seem like they're lockstep, right  But when they split, then there can be a lot of risk; it's very similar.  So, we want to make sure that when you're producing, you're trying to hedge in the location that's as close as possible.

Peter McCormack: So this is why, for example, I had a fixed price on my gas and electric back in the UK that I was still paying a relatively low price when the prices started to spike, because I'd essentially forward-bought at a price from my supplier, and they'd essentially hedged and forward-bought.  And then, when the price of gas started to shoot up, there was a lag between that and when my prices went up, because they've essentially already forward bought.

Shaun Connell: I kind of look at it as the spot price is saying, "Who owns the risk?" because that's unknown.  When it's a spot price, as I mentioned, in most power pools across the US, there's a price every five minutes.  So, somebody owns the risk of whatever that price is and you can say, "I don't want to have that risk, I'd rather have the fixed price".  So, you're going to transfer that risk, because you'll have somebody else that buys that risk that's a financial settlement saying, "You take this risk", so your fixed price energy bill. 

You say, "I don't want the spot price, I don't want to be up and down.  I want to sleep well, so I want you to take that, and I'm going to pay you for that and I'm going to pay you a fixed price, which makes me feel better.  You might win if that price is lower, you might lose, but I don't care; I just don't want to have that risk".

Peter McCormack: So, if someone took on the risk then, when the energy and gas prices started to shoot up, who was the loser in this scenario?

Shaun Connell: So, there could be lots of losers.  If I am a retail electric provider and I've sold you power for $50, because you don't want to have the volatility, so now I'm the provider for you.  So, I'm on the hook, I've taken on your risk, so I've got to make sure that the way I make money is if the actual price that's settled is much lower than what you paid me for.  As a retailer, I have the option of hedging that, so actually going to market and finding a way to cover that; or I can say, "No, actually I'm quite comfortable and I'm going to wear the risk".

So then, imagine a market like Texas, where there are many retailers, big and small, and some of the small retailers didn't hedge.  Then, there could have been some big, major event, like winter storm Uri in 2021, or pretend this is the same kind of market design but in Europe; if they didn't hedge, they're in a lot of trouble and they're likely out of business.  Then, you lost your retail contract, because they're out of business, and you've got to go and find somebody else who's going to now give you a new contract, and it's likely going to be more expensive, because this price has gone up.

Peter McCormack: Well, there were a number of UK energy suppliers that did go up in smoke, didn't they, Danny?  What was it, like eight big ones?

Danny Knowles: Yeah.

Shaun Connell: Does that make sense though?

Peter McCormack: No, it does.  But what actually happened is a number of these energy companies blew up, because before, we just used to have -- energy was nationalised in the country.  We had British Gas essentially.  And then, when the markets became deregulated, anyone can go in and buy and sell that gas in the market and try and create their own efficiencies and, I guess, actively trading.  But a number of these companies, once the prices shot up, they blew up.

Danny Knowles: It was Bulb, the big one that went.

Peter McCormack: Bulb, yeah.  But the UK market is set up in a way that any of the retail customers ends up being rescued.  I'm not sure how it works, but they are protected from this happening.

Shaun Connell: And you kind of link that into something like Bitcoin miners.  To use an example, when the Bitcoin miners are connecting to the grid in Texas, there's always a physical transaction, so they have the choice.  And being an industrial customer, that means a large-scale customer that's deemed to be sophisticated so they can take on the risk of this kind of index pricing, they can connect to the grid and say, "Do you know what, I'm just going to pay the spot price, and I'm not going to do any hedge.  Whatever the real-time price is, that's what I'm going to pay". 

Or they can say, "I'm going to try and enter into a long-term fixed-price hedge with some energy supplier", which is financial so it's hedging their physical consumption, "and I'm going to buy [for example] in Texas, I'm going to buy a five-year 24/7, so round-the-clock, hedge of 10 MWs".  Let me just insert, 1 MWh will charge a 100-KW Tesla battery 10 times, so that's just a reference of what is 1 MWh; so 10 charges of a Tesla battery.  So, they have a choice and say, "I'm going to make this hedge and I'm going to do, say, $35 for my hedge".  So, they've got two transactions: the physical, and then this financial. 

So then, in winter storm Uri, which is kind of happening in Europe right now, where these prices went up, they have the choice now of saying, "I'm a 10-MW miner, I'm just going to turn down my facility, so I'm going to go to zero".  Or, they could provide some ancillary services to the grid, we can touch on that a bit, but assume they just turn offline.  That means that their physical meter that they're going to invoice for is going to be zero.  But they have this hedge.

Peter McCormack: Can they sell that back?

Shaun Connell: They can, or they can let it settle against whatever that real-time price is.  So then, in 2021, that retail price went as high as $9,000 for a few days.  So, you'll hear in the headlines that there are a lot of Bitcoin miners that got paid to turn offline from ERCOT, which is false; ERCOT doesn't pay them to turn off.  It's that the miner made the conscious decision when they hooked up to the grid saying, "Do I want to transfer this risk and have a fixed-price hedge?", because they didn't have to; and if winter storm Uri came through and they just turned off, there'd be no payments to that miner.

But they had this financial hedge, and that hedge settles against the actual real-time price.  So, $35 selling at $9,000 every five minutes for three straight days, so a lot of them did very well because they turned offline.  So, it wasn't that they gave power back to the grid, but they stopped consuming from the grid, so energy that was needed across other customers across Texas, they said, "You guys take the power, we don't need it, it's way above what we get paid to do Bitcoin mining, so we're going to turn off".  Then, they got rewarded for making their financial hedges.  So, those are the two transactions: they turned off; and then they got paid for this hedge, which was kudos to them for having them.

Peter McCormack: It's fascinating.  If you're taking out a longer position, say a five-year position, do you tend to pay more because you're getting a stable price; or do you tend to pay less, because you get the economies of scale of buying five years' worth?

Shaun Connell: There's a couple of parts to that.  So first, there's a credit cost.  So it's saying, you go to the bank and you say, "I want to buy a $20 million home", and they'd be like, "What's your credit?"  If you don't have credit they're like, "Well, why don't you buy a $200,000 and we'll give you some credit for that".  So, the longer the transaction, the higher the quantity, you need to have some type of credit for that, and you can pay for that credit that they could say, "We're going to charge you an extra $3 per MWh, or we're not even going to sign with you; you need to go find somebody like a parent, some other company to sign for you".

So, credit matters and it costs for credit, so it's very challenging for start-up miners to get long-term contracts, because there's a credit cost to that, especially if it's a lot of megawatts.  So, that's the first part: it's credit, and cost is different for everybody.

The second part is that there is a forward curve for power across -- in Texas, there is one primary location, which is the most liquid traded hub that people transact on to make these kinds of hedges.  So, these is a forward curve for this that can go out ten years.  And a forward curve is the same thing as like an oil market.  It is everybody's best guess of the future, and you can make a bet on that and say, "Bullshit, I'm going to sell that", or, "That seems really cheap, I'm going to buy it".  So, this is the standard forward curve.

However, there are really not many natural long-term buyers of power.  So, think about for Bitcoin mining, in producing this commodity which is hashes, and you've got a whole bunch of them because you've got all these miners and you're going to be producing a lot, and you say, "I've got all these hashes and they're going to be worth a fortune, because look at the price of Bitcoin.  You should buy this from me".  Who's your buyer?  Who wants to buy a five-year strip of hashrate that's delivering a certain amount every day for five years out?  There are no real natural buyers. 

Same as power generation, which there are a whole bunch of natural sellers, because we mentioned post-deregulation, so around the year 2000, when they lost their rate payer and had the market, when you sign a residential contract for your house, you're a very small volume compared to a generation facility, and you probably sign for one year, two years; but they've got 50 years of generation.  So, there's no real natural buyers to these large power contracts, not a lot of companies sell them.  So typically, companies will accept less in the future for it, so you can have this pricing curve that goes down, because there is not a lot of liquidity in the market and they'll take less for it, just because there are not a lot of natural buyers, if that makes sense.

Peter McCormack: Well, yeah, because what it's making me think of is interest rates with regards to mortgages.  If you want a two-year fixed rate, you tend to get a lower rate than if you, say, take out a five-year fixed rate, because you might be thinking, "Well, I'd rather have a fixed rate, because I know what it is, I know what I earn and I feel more stable", but you're going to pay a higher interest rate, because the mortgage provider doesn't know where the interest rates will be in three, four, five years.  So, it's a different kind of hedge.

Shaun Connell: Kind of, because sometimes the interest rates in the front of the curves can be lower or higher, based on what behaviour a government will want you to do.  So for example, during COVID, they took the near-term interest rates to near zero, because they want everybody to spend.  And so then, your interest rates in the back were more expensive.  And there's sometimes where they'll want the opposite, and they'll raise those.

Peter McCormack: Okay, I think I'm starting to get it.  It's super-complicated, and there's a lot of people in Bitcoin very interested in the energy markets right now, Bitcoin mining itself, our understanding of the role it plays, not only with securing the Bitcoin Network, but in supporting and helping balance energy grids is a huge topic of conversation; we're going to dig into some of that.  But you as an energy guy coming in, also now understanding Bitcoin, is there any part, any observations you've got of bitcoiners and how they talk about energy, where you think we -- it's hard to say "collectively", but things that people are getting wrong, or things you don't understand what's going on here, or any naivety?

Shaun Connell: So, there is a bit and going back to a shared experience I had in energy markets and where the punchline would be that if Bitcoin price went to $2 million today, there's going to be a lot of problems that occur from that.  And what are all the linkages --

Peter McCormack: Not for us!

Shaun Connell: Yeah, right!  But I'll share a piece of that.  In 2008, I worked for a power generation company, and essentially the gas prices kept on going up, from 2000 to 2008, all the way straight up.  So, anybody that said, "I'm going to hedge", meaning, "I'm going to forward sell", from 2000 to 2008, they lost money, because prices kept on going up, so nobody was hedging.

Then 2008 comes, and so the company that I worked for had a lot of coal generation, and what usually is the marginal unit for a power pool for this being despatched to meet demand is a natural gas unit.  So, we were a coal plant and the price was $40, and natural gas at the time was above $10, I think it was $10 or $12, so power prices were, call it like $120.  So at that time, with the price being $120, because the price is set at the marginal unit and everybody receives that, you're making this really big spread between $120 and $40, so everybody's high-fiving, because we weren't hedged and you think you're geniuses.

Then 2008 happens and then there's this big correction that happens from the Financial Crisis, but the big story was that fracking came out, a new technology that opened up a whole lot of gas.  Gas prices never recovered, and instead of the gas price being up here for the power price going to here, to the $120 market and coal, $40, all of a sudden it dropped down to where coal and natural gas were competing with each other. 

So, the message of the story there is that over time, prices always converge.  Had they stayed at $120, you would have had coal plants saying, "Let's go build a new coal plant, because there's this big spread, so we're going to add a whole bunch -- what are the low-cost resources that we can bring on that's going to actually make this margin?"  So, it didn't last very long and that spread collapsed.

So, for Bitcoin, it's like if prices went to $1 million today, we'll say $1 million --

Peter McCormack: Just zoomed up, yeah.

Shaun Connell: Yeah, so what does that mean?  It means that in Bitcoin mining, a metric that is the measurement of profitability for a miner is the value per exahash.  An exahash is a billion billion hashes.  So, given the current price of Bitcoin, the blocks per day, the number of Bitcoin produced, gives you the revenue for Bitcoin.  Then you take the hashrate and you put it over the course of a day, so how many hashes were there in a day and how much was the revenue.  So, that price is $2.50, ballpark right now.  So, a new-gen miner, and I'm going to try and make it simple.  A new-gen miner is like an S19, so they produce about 120 exahash per MWh, so 120 times $2.50, so what is that?

Peter McCormack: That's $300?

Shaun Connell: Yeah, so right around there.  So, there making really good coin, and their marginal cost of electricity is around $50, so they're happy.  If Bitcoin price went to $1 million, well the value per exahash would shoot through the moon, because of existing hashrate, and each of these miners would be, on average, making about $9,000 per MWh.

Peter McCormack: Sounds great for the miners!

Shaun Connell: It does.

Peter McCormack: It sounds great for the Bitcoin hodlers!

Shaun Connell: Yeah.  So, $9,000 per MWh, what are the behaviours; what are the behaviours of what's going to happen?  So, what you're going to get is, the narrative of Bitcoin mining being flexible, that will turn offline when power is too expensive, everybody's online, they're not turning offline.  So, these miners are not going to be a balancing load, they're just going to be on and it's not going to be fabulous for grids. 

You are now going to have, if you are the owner of -- so, the value chain of mining is that there's the manufacturers of the hardware, there's the capacity to get sites and then there's the miner that's mining.  So, anybody that's actually producing these miners is going to say -- in 2020, at the bottom of the Bitcoin price, you could buy a new S19 mining for something like $3,000.  And then, as soon as prices bumped up, it went to $12,000.  So, if this price jumped up, then these miners would be --

Peter McCormack: $100,000.

Shaun Connell: -- $200,000, whatever it is, because you're plugging in a printing press.  So the point is just, the entire value chain, there would be so many people fighting for, "Where do I capture this margin?"  So an example I shared with you of saying, when power prices were $120 and it came to $40, is prices always converge.  So, either people are going to sell Bitcoin because, and I'll get a lot of hate on this from people saying that you sell your Bitcoin, but for this thing to jump up so high, this is hard to be sustainable.

It's like the example again of oil prices.  Does it make sense that oil can go from $100 right now, and then tomorrow the price was $1,000; it would break things.  All of a sudden, it would have a cascade of people that got margin called as if they had forward sold this, and oil would have problems with people just getting from A to B, but it would be a catastrophe.  But if oil went from $100 to $110 to $150 to $170 to $120 to $130 to $140 to $160, $180, $160, there's a lot of benefit in this step change that happens that doesn't disrupt and break stuff.

So, back to your comments on what do I think are some areas where I think it's a little naïve just to think that we can have these Bitcoin prices go so high so fast, because there's a whole lot of implications and some folks in Bitcoin might argue saying, "Well, it can last, because Bitcoin's a super-scarce asset".  But in the same time, you'd have this massive rush of hardware come to market, you'd have a hard time finding these connections, there could be disruptions across the grid, it could be bad for branding for Bitcoin, and you would be at home plugging in your computer to mine, because it would become economic to mine on your computer again.

Peter McCormack: So, I mean we forget about it, say, going to $1 million; but what you're really saying is fast increases, rapid increases in the price of Bitcoin has knock-on effects within the energy market, as well as knock-on effects elsewhere, but within the energy market specific.  And when we have a narrative of Bitcoin miners being a load balancer, that's okay when the price is stable; but when the price is accelerating, they're not a load balancer, because they're not going to want to switch off.

But can the agreement between the miners and the grid, can they have the agreement that they have to turn off at specific times, it doesn't matter what the price of Bitcoin is, if there is a certain amount of demand of the grid?

Shaun Connell: Yeah, and just to clarify again, similar to cars in that there are cars that go slow, like a Civic, and a Ferrari, there's different speeds, miners have different efficiencies and different revenue attached to those.  So, Bitcoin can act as a load balancer in a very large range.  So, when you use the analogy of $1 million Bitcoin, the same category of $1,000 Crude in a fast, short time.  So, Bitcoin mining can be incredibly flexible on each stage on the way up, and it can even be flexible if you have these big price surges, where it goes $100,000 or $200,000.  But this kind of super-exponential forever, there are some challenges on that.

Now, for Bitcoin being able to be controlled, there are ways.  So, there's different types of load, and we often hear about demand response, and there are very different types of demand response.  So, there's retail demand response, then there is wholesale demand response.  So, retail demand response, you'll hear voluntary, and I'm sharing that wholesale is not voluntary, and I'll give some examples.

To be retail demand response means that you are a miner that is connected to -- you're mining at home, and the utility company has come out with a time-of-use schedule, where it's more expensive for you to consume energy in the evenings and cheaper during the non-evening times.  So, that is demand response.  What they're saying is, "Change your behaviour.  We want you to consume less in the evening peak and more during the day", but that's voluntary, because if you do consume in the peak, you're just going to have to pay more.  So, there are no fines for that.

The challenge with retail demand response is, they don't have the telemetry, the equipment to understand what you're consuming in real time, so use examples like this type of demand response would be like a Honda Civic, and there's a speed limit.  They'll say, "We're going to have different speed limits at different times of the day.  We want you to drive faster, drive slower, but at different times of the day to control traffic, here's the speed limit.  Please help us out".  They're not going to know until the very end of the month what you did, because they're going to do the invoicing for your billing and they're going to look back and say, "Oh, actually he did have some behaviour changes".

Now, go to the wholesale side.  Think about the wholesale like a Tesla, two Teslas, one is fully autonomous and one isn't.  So, they have different equipment, they have different capabilities.  So, they both have the ability for a grid operator to have visibility into what they're consuming exactly at what point in time.  So, there's two types of wholesale demand response.

The first one is, "I don't have the flexibility to essentially change my speeds during the day, but you can see exactly what I'm doing during the day.  And if you need my power during the day, you can flip my breaker and turn me off".  That's called an "under-frequency relay" and it's essentially a shunt trip, and it's the grid operator saying, "I'm a wholesale DR, so I'm the grid operator under their control".  And it says, "If something happens, since you're paying me to provide the same types of service as you're asking these generators to provide for backup power, etc, you can trip me offline when you need me".  So, that's one type of DR.

The second one is the Tesla that's fully autonomous.  What that means is that the speed limit, the grid operator is passing back and forth in real time the speed limit, based on the actual real-time needs of the roads here; and it has to follow the instructions of the grid operator, there's no choice.  So you are essentially on autopilot in your Tesla, and if you ever take control of the wheel yourself, you're going to get a really big fine for doing that.  And the fine's going to be so big that you're not going to want to take your hands off the wheel.

So, that's what we're seeing in these mines that are on the grid, is that -- I know we've gotten kind of tricky on this stuff!

Peter McCormack: Yeah!

Shaun Connell: My dad, I've explained his power bill ten times, and he's frustrated often that the cost for the transmission, the wires cost, is higher than the energy cost and he's thinking, "This is just terrible".  And I look at him and I'm like, "Holy shit, your energy is super-cheap!  You don't know that you have the cheapest energy in the country, because you're looking at the bill".  So the point is, I've explained his bill probably ten times, and it's still kind of a hard topic for him to grasp. 

So, I can understand the complexity of these topics, when that's just a simple energy bill for my dad; and now we're talking about demand response and why it's transformational for a grid operator to have the control over these Bitcoin mining facilities, so that they can use them to help manage enhanced reliability of the grid as they bring on more renewables, etc.

Peter McCormack: Well, we'll get into the supplies, but what we're really saying here is that the demand for energy can change very quickly.  I mean, here in Texas, we've been here for the month and we've gone from warm days to cold days.

Danny Knowles: To tornados.

Peter McCormack: To tornados!  But the weather's changed quite a bit and because of that, there are some days where we're putting the heating on and there are some days that we aren't.  You can have a cold snap, where there's more of a demand on the grid for whatever reason.  You also might have, I don't know, tell me if I've got this completely wrong, but there might be a huge football match on, and everyone's going to be at home watching it, the Super Bowl, so there's going to be an increase in demand.  So, there's different things that impact the demand.  But the grid always has to oversupply, otherwise we have blackouts maybe?

Shaun Connell: So, the grid operator, up until recently, has only had control of generation, so coal plants, natural gas plants, hydro, nuclear, etc.  So, they get to control the generation; they have no control over load.  So, their responsibility is that they need to despatch this generation to exactly meet what they think is going to be for load.  So, they don't get to say, "Hey, load, we don't have enough generation, can you just turn down?"

As a grid operator, what they're trying to do is, a grid is a heartbeat, and the heartbeat is 60 Hz per second, and it has to stay there all the time.  So, what the grid operator does is says, "Okay, we've got to manage it, so it's got to be 60 Hz", because if it goes down to, say, 59.2 Hz, just that drop, that's essentially like machines that are connected to the grid will protect themselves because, "Oh, shit, something's happening, I've got to protect myself, so I'm just going to disconnect".  So, it's a cascading effect that essentially gets to a blackout.

So, the grid operator, their job is to do real-time balancing of energy to match demand, that keeps the heartrate at 60 Hz.

Peter McCormack: But what if there's a massive increase in demand all of a sudden?  Is it like going running and the heartbeat goes up to 70, 80?

Shaun Connell: Yeah, so in that example of there's a massive increase in demand, the heartrate will start to go down, because there's not enough generation being --

Peter McCormack: Oh, I see what you're saying.  Yeah, okay.

Shaun Connell: And the opposite's true.  If all of a sudden demand comes offline, it's dinner time and everybody's turning off lights and stuff, all of a sudden it will go up and they're like, "Okay, what we've got to do in that case is we've got to turn our generation down".  So, the grid operator is going to make their best guess of a forecast about what is load going to be today given everything they know, and they know a lot.  They have historical data that says, "Today's a Super Bowl.  How do people behave on a Super Bowl?  Well, before the Super Bowl, this is the behaviour of the load, and then all of a sudden people turn their ovens on". 

So, they've got all that data that they're saying, "Okay, we're going to expect, on the Super Bowl, given the weather for today, given the time of day, who's there --"; they'll probably even take into consideration who's playing.  If Texas is in the Super Bowl, they're going to say there's going to be more people.

Peter McCormack: If it's the Jets, no one gives a fuck!

Shaun Connell: That's right!  I was going to say something about Bedford soccer!

Peter McCormack: Hey, be careful there!

Shaun Connell: Everybody watching, right?

Peter McCormack: All 12 of us!

Shaun Connell: So, they're going to have this information.  They're saying, "I've got control of generation, the heartrate's got to stay here, and I'm going to forecast the load there's going to be.  And essentially, whenever there's an increase of load, I've got to put more generation on; and when there's a decrease…" and that's just on the load side.  On top of that, in the old days, as in the old days in the late 1990s, there was no such thing as wind and solar.  So, this generation stack that they had was extremely flexible, meaning reliable, "I can count on you to show up.  I'm expecting this demand over here, and I need you to be ready to come online when I say, because we've got to match this demand".

Peter McCormack: You know how much coal you need to burn to generate that power.

Shaun Connell: And the coal's on the pile, and the people at the plant, they're ready to go.  Natural gas is the same thing.  All the people are there, all the fuel's there, and it can be despatched round the clock, or have enough flexibility to turn up and turn down.  So, this was late 1990s, generation stack, you have control of the full generation.

So then, renewables started to come out around 2000, 2003, 2004, 2005, and then all of a sudden, there's now a large portion of the grid that can be wind and solar, which is zero-cost energy, which means when it's being produced, you can't say, "Wind, I need you to turn up 3 MW and solar, I need you to turn up another 2 MW", you can't.  So, what you need to do is you need to say, "Here's now my wind and solar and now, not only do I have to forecast the demand, I've got to forecast, is my generation going to show up?"

So, think about a day like here in Texas recently, something like 60% to 70% of generation for the day was wind and solar.  So, on the day, for example, let's call it 50 GW, and of that 50 GW about 70% of that, like 30 GW and change of wind and solar.  So now, if wind and solar starts to drop off here, you need to find something to back it up fast.  So you say, "Okay, I need to call my coal plant, I need to call my natural gas plant.  I need them to come online, because I just lost this generation".  So, does that make sense on this?

Peter McCormack: Yeah, I get that, I totally get that.  But do they have to build in a buffer?  You talk about the heartbeat; how much of a buffer do they have to build in, just in case there's a sudden demand?

Shaun Connell: The punchline on that is, it's exactly matching real-time demand.  However, you'll hear there's losses that happen between the point of generation, all of the way to where it's consumed.

Peter McCormack: Yeah, Harry explained it to us; you lose electrons.

Shaun Connell: That's right, but it's not -- in tiny Bitcoin miners, if they connected the transmission voltage, that means for example generation, transmission lines, transformer, distribution lines, transformers; the closer you are and the less of that transformer you use, the less amounts of losses.  So, if you're transmission-connected, you lose 2% losses.  If you're distribution-connected, it can be something like 7.5%, so there's a big difference.

As a grid operator, bringing that back and saying, if the entire demand across, and we'll use easy maths, is 100 MW, they might be producing 104 MW, 105 MW to make up for all those losses, but they're exactly matching what the demand is at that time.

Peter McCormack: Isn't there a risk therefore that they predict it slightly wrong and there's overdemand; or what happens in that situation?

Shaun Connell: So, you've done a really creative way to introduce ancillary services, so that is insurance.  So, you were talking about, do they overproduce?  What they have is they actually have, it's called "contingency reserves".  So, those are like insurance products.  So, I'm going to use the grid of the past, so in the late 1990s, and as these markets were being deregulated, FERC came out with these -- FERC is the governing body for energy across the United States -- "These are the required four essentially ancillary services that you need to have from generation.

The grid operator goes and procures those from power generators, so I'll walk you through those in a simple way.  There's regulation, and regulation is like imagine the 60 Hz, and they're going to despatch their generation accordingly.  They're going to send them a signal, but sometimes they're off a little bit, so they need to have backup.  They say, "I need to have 3% of what I expect load to be on standby, that I can control it to balance the system".

Peter McCormack: Okay, yeah.

Shaun Connell: So, that's regulation.  So, that's for that insurance product.  The second one is, it's that, "Oh, shit, something bad happened".  It's the contingency reserve.  It's saying, "I lost a nuclear powerplant, it just tripped offline.  It was 800 MW, which is very large, and I need to have something that can step up right away".  So, these ones, they'll call them "10-minute spinning reserve", and they'll call another one, "30-minute spinning reserve", and there's some other names for them; but those are the categories.

Peter McCormack: What are they though; where does that power come from?  What is that generation?

Shaun Connell: So those are power generation resources that are connected to the grid, and say you're a power generator and you're a natural gas plant, and your natural gas plant's 100 MW.  You can either sell energy, which means I'm going to fire up my turbines and I'm going to produce 100 MW; or, you can sell energy and then provide some ancillary services and be on reserve.

Peter McCormack: So, you essentially have the buffer, which is an insurance?

Shaun Connell: So, you give the grid operator insurance.  Because you're saving some of the grid for the energy, but you've sold them this -- it's called "capacity of ancillary services" to the grid operators just in case.

Peter McCormack: Okay, so you're not actually producing it, but you could product it at a click if required?

Shaun Connell: That's right, yeah.  And it's online, it's ready.

Peter McCormack: So, it's interesting, because I had this perception that, say, how much power does Austin use, do we know, or Texas?

Shaun Connell: So, Brad Jones, CEO of ERCOT, mentioned the other day about the Bitcoin mining in Austin, and I think probably I'm going to get this wrong, let's call it 1.5 GW, so 1,500 MW.

Peter McCormack: Okay, say it's 1,500 MW; is that daily?

Shaun Connell: That's for the exact moment.

Peter McCormack: Okay.  So in my head I was like, well maybe they produce 20% more just in case.  But actually, what you're saying to me is that the network is so advanced, they've got so much good information, that they run it as a really efficient network with a backup of insurance services, just in case they require it.  But it's actually a very efficient network.

Shaun Connell: Very, yeah.

Peter McCormack: Okay.  But there's a bit more variability in it now, because we've got energy coming from renewables, solar, wind, which we know if the wind isn't blowing, you don't have the energy; and if the sun isn't shining, you don't have solar energy being created.  So, you have essentially insurance services for those as well, or the ability to…  Okay.

Let's try and imagine the scenario where it was 100% renewable, because I think that's going to be easier to explain for my next scenario; 100% renewable, but it's obviously not 100% reliable.  What is the scenario for providing power in that scenario where maybe the wind isn't generating enough or the solar isn't generating enough?  Is it with renewables, you do have to have an oversupply, and are we looking at things like batteries to be able to store for these scenarios?

Shaun Connell: So, maybe I'll try and take pieces of that.  So, pretend all the technology that existed was coal plants, gas plants, and the renewables; yeah, so that's the mix.  So then, on this one day, you have all renewables; we're saying the batteries aren't here, demand response isn't here.  On a day where it's all renewables, you need to have some type of resource that's there on standby that can provide for that.  In power grids, what they do is they model for the largest contingency so, "What is the biggest, 'Oh shit'?"  So, they take that and then they take that times two; that's how they model it, so the two worst things happen.

The grid operator will look at the mix and say, "Hey, we've got this large solar facility online and it's a very large amount, so we need to have at least twice the amount of that as backup".  But if you only have wind and solar online, you don't have that ability.  So, what they're going to have to do is say, "Okay, how much do we need to have?  We're now going to have to put the generation online, and the generation, the coal and natural gas has to be online, because we're going to need to be able to call on it, and we're going to curtail the wind and the solar, the excess amount". 

People are learning right now that there's been this big push for low-cost energy, and they're now switching to saying, "Reliable low-cost energy", so reliable is a very big thing.  And so for reliability, which is the responsibility of the grid operator, they don't care about price, it's reliability and they run this auction.  They're going to have this generation that's online that's coal and natural gas, and then they're going to ask the wind and solar to essentially disconnect from the grid, curtail; so just spill the energy, waste the energy.  So, that's kind of that scenario, okay.

Now we'll say that you have a grid that is wind and solar and now you've got, we'll say batteries.  So, do you mind if I do a frame of just --

Peter McCormack: Please do.

Shaun Connell: Okay.  I think that we need to reframe mining. 

Peter McCormack: Okay.

Shaun Connell: And, there seems to be that my observations and part of my opinion is that it's a very easy attack vector to come at it from the energy side.  So, we need to shut it down.  It feels like we are on defence for --

Peter McCormack: What you mean here is that the FUD that comes out is that bitcoiners miners use X amount of energy, the same as Norway uses, and this is a huge waste and bad for the environment?

Shaun Connell: That's right, yeah.  So, we're having to defend it as a good use of energy.  So, that feels like a lot of defence, right.  So, I feel like it would be wise of us to do this as a reframe, and what I'll share of it is around some of the goals that are occurring around decarbonisation, and tying back into instead of defending, "Is this just or not?" let's say that this is actually helping with this goal of decarbonising energy.

So, the IEA has a Net Zero Emissions report saying, "The world is decarbonising energy, and whether people agree with it or disagree with it, it doesn't really matter, because the parade is in motion, it's picking up speed".  So, what countries are doing is that they're making these large pledges of saying, "We're going to be net zero emissions by 2050, and I think it's something like, out of all the emissions in the world, 70% has been pledged as this commitment to net zero emissions.

In that case of having net zero emissions, what needs to happen is you need to electrify everything.  You need to move away from combustion engines in cars and go to EVs, etc, and you're going to do a 3X in your electricity needs, so that's massive.

Peter McCormack: Why; why 3X?

Shaun Connell: So, for the last ten years, electricity demand has been very flat here in North America, but if you're going to electrify everything, that means that what we use right now for fuel in our cars is now going to go into EV, so you need electricity for the EV.  Heat that's in buildings that comes from, say, natural gas is going to be electric.  So, you're creating a massive amount of demand, assuming that you can get that electricity as low-carbon electricity.

Peter McCormack: Okay, I understand, yeah.

Shaun Connell: So, in that scenario, you're going to do a 3X of electricity demand around the world.  Simultaneously, you're going to essentially retire all of your coal plants and fossil fuel generation, which is your reliable generation, and you're going to replace it with a 15X in wind and solar capacity.  What the IEA is saying is that to do this, and you talk about insurance products, it's saying, "You're going to need a hell of a lot more insurance, because you just 3Xd your demand, you've removed your flexible resources from coal and natural gas, you've added wind and solar, so you need to have about 4X more flexibility, resources that can be despatched up and down".  Of this flexibility right now, the majority of this comes from conventional generation, and demand response in batteries right now represents 1%.

Peter McCormack: Wow!

Shaun Connell: In the IEA's 2050 scenario, batteries and demand response are going to have an oversized role where they're going to represent about 50% of all flexibility, with 30% coming from batteries and 20% coming from demand response.  So, the demand response right now needs to do a 20X, something like from 100 GW to 2,100 GW.  So, you kind of step back and say, these countries are making pledges and not everybody knows what that means; what are the implications?  If feels good to say it, but what does that mean?

What it means is that you need to have a tremendous amount more flexibility in saying, "Hey, country, you have pledged net zero emissions.  To do that, they're saying that you need demand response; is that right?"  "Yeah, we need a lot more demand response".  "Good, because we're going to get you the best type of demand response that is super-flexible, that you can call on any time you want, and this is us serving that larger goal of providing flexibility to these grids in transition, that without that flexibility, you can't do it".  So, it's a reframe of saying, instead of making this about, "Bitcoin's a really good use of energy, because right now --"

Peter McCormack: Because it's good money!

Shaun Connell: But also, there's about 12 GW of total global demand, 12 GW of energy being consumed around the clock, ish, for Bitcoin.  And I mentioned about if Bitcoin goes to $1 million, all the people that are going to start mining and all the implications of that.  So, in that future, it's 200 GW or 300 GW, a very big number.  So, in the future, Bitcoin will consume a lot of energy, because everybody sees it as really valuable.  So, wouldn't it be amazing if you could turn that narrative from, "Mining is the sin", to, "Mining's the saviour", right?

Peter McCormack: Well, we talked about this previously.  One of the super-interesting things about this is, mining has been the sin, it's the dirty side of Bitcoin, it's the easiest attack vector.  I mean, there are lots of different attack vectors, but this has been a consistent one, especially in a time where climate change is a huge issue, that people want to attack Bitcoin mining.  But we get to do two things at the same time: we get to actually make Bitcoin part of the decarbonisation, by allowing it to support that flexibility; and at the same time, expand the Bitcoin Network, grow the Bitcoin Network.  So, we get to do the two at the same time.

So I think that reframing is right, I think it just hasn't really landed for people.  I think if somebody is thinking about climate change, they tend to think about it in the most simple terms.  We're burning a lot of fossil fuels and that's putting carbon in the atmosphere; we need to stop doing that.  But none of these people, if I asked any of my friends, none of them are going to understand that, "Oh, great, well we need this massive increase in wind and solar, and perhaps nuclear, we're going to need this massive increase", but they don't understand demand response, they don't understand flexible load, they don't understand how grids work and they don't understand the role Bitcoin mining can play being part of that.

So, it's actually quite a complicated answer.  It makes sense to us, because we spend time talking about it.  But communicating that out, how grids work and what role mining plays, is actually quite complicated, and that's one of the challenges I think we face.

Shaun Connell: It's a huge challenge, and you mentioned this resource mix and that's where it started before it went over here.  And we watched this video earlier and it was with Brad Jones, who's in-term CEO with ERCOT, and it's saying, "What do grid operators think about Bitcoin?"  So, I watched this the other day and it just came out and my jaw dropped, it just absolutely dropped.  It was essentially Brad Jones, who really understands the ERCOT grid of course, he leads ERCOT; and then, what he said in the video where he started off saying, he called it crypto mining but, "Bitcoin mining is a great opportunity for us".

For context on that, ERCOT right now has approximately 10 GW of solar, so their average demand is 45 GW, so they have about 10 GW of solar.  Over the next 18 months, they're going to go to 20 GW.  So, I was on a phone call with Brad Jones, and the Texas Blockchain Council had put this together, and on the call, Brad Jones was saying, "Listen, guys, just so you know, you don't have to sell me on Bitcoin, I don't own anyway, I'm risk averse, but I understand completely how valuable Bitcoin mining can be to this grid", and that's where he shared the stat on, "We have 10 GW now, we're going to 20 GW".

For flexibility, there's four types of flexible resources.  There's your regular generation, there's your storage they can have, which can be hydro in a dam, or it can be batteries; and there can be interconnections, which is like free-flowing electrons between neighbours; and then there's demand-side management, like demand response.

So, in Texas, they don't have interconnections with other systems, they're very unique.  So, ERCOT is an electrical island.  It could very well be an island on its own, it's disconnected from the rest of the grid, aside from some small interties that are just DC ties that are just direct current ties.  So, they don't have any hydro either.  So, most areas, like the Pacific Northwest, has got hydro, there's some up in New York; Texas doesn't have any.  So, you're saying you're going to retire this coal, retire this natural gas, I don't have any interconnections, I don't have hydro; I need batteries and I need demand response.

It was really fascinating to watch Brad talk about Bitcoin mining, because some of his words were that when demand is flexible, he talked about how Texas is a great resource for wind and solar, and there's an abundance of wind overnight.  Bitcoin miners have been locating in these areas and essentially soaking up this excess wind, because there was too much for the grid that it couldn't handle.  So, his words were saying, "Instead of curtailing this great resource, they're able to consume that and to improve the economics for these wind developers so they can build more.  And then when price changes, and it's that point where the Texas customers need it, they're quick to turn down".

So again, this is the CEO of ERCOT, and I'm hopeful that this type of video can be watched by policymakers, because it's saying, if I'm a policymaker and I'm saying, "I'm worried, because there are people telling me that they're worried about Bitcoin mining, it's going to consume their energy and there's going to be blackouts; what do I tell them?"  This is an example of saying -- and we use the example of retail demand response and wholesale demand response.  Wholesale is where the grid operator has control of your load and can have flexibility for sending you signals to turn up, turn down, etc. 

So, the CEO of this grid, that's attracting all of the Bitcoin mining right now, that's catching headlines, that just occurred this winter storm Uri, that you would think that the CEO of this power grid would be like, "Leave.  We need to keep that power for ourselves".  He's viewing it as the same way that I kind of shared in that framework that says, as you integrate higher levels of renewables into your grid, and the sweet spot's about 25%; when you pass 25%, you start to encounter a lot of issues and you start redefining the rules and coming up with new insurance products.  ERCOT's doing that right now, so they're creating some new rules, and they're creating new insurance products for this fast-ramping --

Peter McCormack: And you said that one day, they got to 70% that came from renewables?

Shaun Connell: That's right.

Peter McCormack: That's incredible really.

Shaun Connell: And so that example is, last year was 27% renewables on average.  So, across the entire year, 27% of all megawatt hours were produced by wind and solar, which means that there some days that were 2%, there were some days that were 55%.  So it swings, but the average was 27%, so that means that there's going to be some days where you have all renewables, and you've got to manage -- we talked about forecasting load.  The power system's now being driven by, "What does my mix look like?  What is wind and solar going to do?  What resources do I need to have online to make sure that I can serve this part over here?"

Peter McCormack: So, where people are worried about climate change, they're worried about the amount of carbon that goes into the atmosphere, and on this podcast I've tried to speak to a range of people.  I've spoken to people who understand the climate and climate scientists, but I've also got Alex Epstein coming on the podcast soon, who's written the book, The Moral Case for Fossil Fuels.  He and Marty Bent, who came on the show, their worry is that if we demonise fossil fuels, we risk destroying the grid, or we risk going into periods where there isn't enough power generation.

We've seen the issues that have happened in Germany recently, where they decommissioned all their nuclear facilities, relying on Russian natural gas to power their grid, I think it's Russian natural gas.  Should people be worried about this, or is this a red herring?  Should we be worried about demonising fossil fuels; and is there a risk that if we try and go too far to renewables that the grid can fail, is more likely to fail?  And in doing so, if the grid does fail, we know that has massive implications on society.

Shaun Connell: The position that I have, it's an "and" statement.  It's not like it's A or B or C or D, it's A, B, C and D, and you find out what the solution is.  So, I think the world should be grateful that there were some countries that were early leaders in doing this energy transition, because he/she who goes first is going to teach you the lessons.  So then, what came out of, say, some of the stuff around Germany you were saying, what do you need to do different that didn't work over there?

Now, there's research that has been published that is becoming very -- a friend of mine grabbing attention, and I've seen a lot of the academic papers written about this, which is just as power systems integrate higher levels of renewables, what is the actual way that you need to incorporate more flexibility through the grid, through different types of market reforms, which is the new rules; and then also, new products.

My punchline on this is that others went first, they're showing us the way, and we can't do this all at once super-fast.  We can't just go like the example where I said, oil going from $100 to $1,000, or Bitcoin going from today's price to $1 million.  It needs to have some type of plan to it.  You don't plan Bitcoin price, but there needs to be a plan on saying, "How do you integrate more renewables?  What do you need more for batteries; what do you need more for demand response?  How do you make it so that you're not dependent on -- there's some risk in your system where you're handcuffed?"

So, the message I'd say to the Marty Bents there is that I'm pro-fossil fuel generation; it works.  We just need to have a plan on how we make this transition.  I'll share a bit about, what's the prize here?  There's a big prize, and it's that solar is semiconductors, which means it's a technology and it's improving.  So, for the first time ever, we are now actually manufacturing energy.  And the solar costs are going at a pace where, I think since 1975, somewhere round there, the learning rate, for every doubling of capacity, what percentage decrease in cost did it have?  Over the past 50 years, or whatever that is, it's averaged something like 30% and for the last decade, it's been 35%.

So, we're now at a really low price of solar, and if we get to the point where the cost of solar becomes near zero, which trends down because it's a technology, and we get to something like a $5 solar price, that just opens up a whole bunch of things we can't even imagine.  What are some of the things you can do with near-zero-cost energy?  I'm not an expert in this space, but hydrogen production all of a sudden becomes extremely economic from renewables, and how that can now be used for hydrogen generation, etc.

You can look at that and say, in these energy transitions, as you're incorporating more solar, and as these Bitcoin miners are essentially partnering and taking to new offtakes with these solar projects that wouldn't have been built otherwise, they're helping to accelerate that capacity doubling, which is lowering this price and getting the point that near zero -- and I'm thinking, call it 50 years out.  So, it's not short term, but imagine what that looks like 50 years from now, and there's zero-cost energy and we've found ways to harness it and to use that for different applications and production, etc.  It's truly remarkable.

Peter McCormack: So, my take from you is that none of this is binary.  It's not, we should just have lots of coal powerplants just so we're protected, and we shouldn't just immediately imagine we're going to transition to renewables.  What is happening is the percent of the mix coming from renewables is growing, and the grids are adapting to that and adapting to a more flexible load.  And they're doing that with the different insurance products, and also by using Bitcoin miners to actually help with that load balancing.  So really, this is just an evolution of the grids.

Shaun Connell: That's right.  And I would say that what's an important start point is, it's reliability first, low-cost energy afterwards, and super-bold font on reliability.  So, how do you make grids reliable in this transition?  What are the levers that you need to have?  Brad Jones talk about this, about what levers need to be available for you to manage this transition?  Then you go to low-cost energy.  So then, on the fossil fuel side you're saying, if you're sacrificing reliability, that you don't have the reliability, don't remove those reliable assets; you want to keep that.

Peter McCormack: Do you think we can get to 100% renewable, and be reliable and low cost?

Shaun Connell: Again, anchoring on the 2050 mark, I've seen what some of these energy mixes look like, and I think they're possible.  It's hard to believe -- it's not going to be in the short term, it's not going to be fast, but in this future that you can have 70% wind and solar, you can have 15% hydro, which is water; and then in this future again of low-cost energy where, remember where aluminium used to cost more than gold and all of a sudden, arc-furnace-invented electricity --

Peter McCormack: Hold on, I didn't even know that.

Shaun Connell: Yeah.

Peter McCormack: Are you talking about aluminium?

Shaun Connell: Aluminium, like cans, and you use for aeroplanes and pop cans and stuff.

Peter McCormack: Yeah, aluminium?  You know we call it aluminium!

Shaun Connell: Soccer, right?!

Peter McCormack: You're good.  Tomato!  Yeah, okay.  Aluminum, I think actually I prefer aluminum, it sounds cooler!

Shaun Connell: Bioluminescence, aluminum?

Peter McCormack: Okay, so aluminium was more expensive to make than gold?  I didn't know this.

Shaun Connell: Yeah, and what do you call them, the forks, spoons?  Utensils were made from aluminium, because it was very prestigious, because it was super-expensive.  And then, that changed when it was really affordable to make aluminium, and then all of a sudden, gold took top spot; it was more scarce.  So, what happens, what does $5 energy do the world?  What are things that could be done that couldn't be done, because they were uneconomic. 

I'm tying back into this IEA view of the future, and I'm using that, because I don't want this to be my perception of the future.  There are reports like IRENA and IEA, and in the future they say 10% of generation-ish comes from hydrogen generation, which means that hydrogen got so cheap along the way that it became economic, that you could actually use that for providing some of this flexibility in the future.

So, you'd have hydrogen generation, you'd have batteries, you'd have demand response, and the conventional generation right now, the coal and the natural gas, it goes from something like 55% to less than 1%.  So, it flippens the demand response in the batteries, and that's the direction that I kind of see.

Peter McCormack: Okay.

Danny Knowles: Without Bitcoin mining and the demand response, do you think it would be possible for them to get there?

Shaun Connell: Yeah, so I think it's harder, I think this accelerates it.  The example is that for 20% to come from demand response, so what are the current applications of demand response, and where do they get a lot of the value for this?  So, I'll give you a couple of examples.  I forget the name of this report I was just recently reading on this, but flexibility from demand response.

The ultimate type of flexibility from demand response is something that is energy intensive on one process, because that way you can just stop and turn that thing, and it's not like an assembly line of things that are all connected.  So, we'll start with steel.  So, steel has an arc furnace, and so there's the melting of steel, there's, I think it's called a steel ladle of some sort, but very electricity intense.  Then it goes downstream and there's ten processes all the way to the end.

So, if a grid operator, for example says, "A nuclear powerplant tripped offline, we need backup power".  Steel or even aluminium, they can provide it, but they can only do it for about two hours, because if they turn everything offline, then all of a sudden it affects what they're doing in their production; so the steel starts to harden.  Here's the chart here that kind of shows it.  So, these are the processes, right.

Peter McCormack: Okay.

Shaun Connell: You see there's the melt shop where it says, "Arc furnace and ladle furnace", and it steps all the way down, so it's an assembly line.

Peter McCormack: Okay, so do these assembly lines run 24 hours?

Shaun Connell: Yeah, so these are industrial facilities, and you often hear in the news that these types of customers get low-cost energy and they usually get better rates, because they're consuming larger volumes and they're helping fill demand in the morning times when there's no demand.

Peter McCormack: Okay.

Shaun Connell: So, if you look at the steel plant and the cement plant, if you need to call on these resources, for the steel plant, you can get almost 96%, which is a lot of peak reduction.  But the challenge is that you can only get that for about two hours, because they've got to essentially get the electricity back on to keep the heat on the arc furnace.

Peter McCormack: Because otherwise, they essentially shut everything down, which is a whole, massive process.  And then to get it all back going, it's…  Right, okay.

Shaun Connell: People and processes and coordination.  So, it's a lot of people involved, a lot of processes involved.

Peter McCormack: A cement plant is for three hours?

Shaun Connell: Yeah, so you can see that there are two energy-intense processes: the quarry and the cement mill.  So, they don't have as much of a concentration at one location, and they have interdependencies all the way down.  So, this example, you can do it for three hours, but you're only getting about 70%.

Peter McCormack: Okay, but the evil Bitcoin mining?

Shaun Connell: Is a single process, very energy-intensive on one process, and it's simply just turn off and for as long as you want.  Here's an example of batteries and Bitcoin mining as a flexible load.  In the future, we need both; it's not A or B, it's both, and there are different use cases that the battery wins.  A Bitcoin mine can't store energy and inject it back in the grid.  If you're having a black start and need to inject it, you can't do that; so, win for the battery.

But if you're in West Texas, and say it's a very windy week, super-windy, and it's so windy that there's not enough demand in that area and all the powerlines moving across West Texas to the major load centres are full.  So, you've got 2,000 extra megawatts around the clock, so there's 2,000 MW being produced every hour that's extra.  If you have a battery -- let's use the example of 1,000 MW.  So, say if you're battery, and pretend it's the biggest battery in the world, it's 1,000 MW, I don't think there's anything close, they're usually smaller; but it's a 1,000-MW battery and it can do four hours.  So, you're going to 1,000 times 4, so 4,000 MWh.

Then, it has to stop.  Then, it can't do anything else for the wind, it can't consume the energy, they'd have to curtail it, because there's nowhere to go.  So, you just had 96 hours and for 92 of those, you had to curtail the wind and for four hours, you could consume with the battery, because the battery had to wait for prices to come back to above zero, or where the wind stops blowing, otherwise when it injects the power in the grid, it's going to make things worse.

Peter McCormack: Okay.

Shaun Connell: And then, insert second is you have a 1,000-MW Bitcoin mining load that has flexibility.  So, in this example, 96 hours, and so that mine could consume 1,000 MW every single hour, all the way to the end.  So, they've just consumed 96,000 MWh versus the battery, which was 4,000 MWh, so now relate those in payments.  There were 96,000 MWh that received payment for that energy, and in the first one there were 4,000 MWh.  So, this is an example where it can kind of soak up the excess wind that would have just been curtailed otherwise.  And curtailed means wasted.

Peter McCormack: But when the Bitcoin miners turn off, and they can do it instantly, is there a period of time where it becomes -- because, I don't understand the full commercial relationship between the miners and, say, the grid or those who are generating the energy.  But if you turned round to them and said, "We need you to turn off for a week", that's a lot of time they're not mining.  Is the commercial relationship set up in a way whereby they still are being paid whilst they're turning their machines down; or, are they getting just such a low rate when they're on that they can turn off?  What's the whole commercial relationship here?

Shaun Connell: So, they're similar to a power generator.  ERCOT's not paying them to turn off, it's a signal.  The price signal is saying, "Turn off".  So for the example we shared before, like a power generator can either produce energy or backup power, in the energy market, the way that the example of ERCOT would tell a Bitcoin miner to turn offline is saying, "As of today, the breakeven point for an S19 miner is $250.  The breakeven point for an S9 miner is $80". 

So, if the price came out in real time and it was $120, the S9 miner says, "Too rich for my blood, I'm off", because I'm not going to pay you more than for what I'm going to get in return; it doesn't make sense.  It's like you give me $10, and I give you something that's not beneficial, I give you $5; it's not of value to you.  But the S19 miner's going to stay online, because it's still economic for that miner to be online.  If the price goes above $250, the miner's going to be like, "Okay, you can have this power too", because why would I consume something that costs more?

This example of energy side, where they're actually following an instruction from ERCOT, and they have the price of energy that they have, and just on the energy side they're going to react and say, "This is above my cost".

Peter McCormack: Right, so ERCOT changes the price as the demand changes?

Shaun Connell: That's right.  And price is to drive a behaviour.

Peter McCormack: Yeah, okay.  And so therefore, the miners can't forward-buy the energy for two years at a fixed price, because they wouldn't be able to load-balance it, and that's the way the contract works?

Shaun Connell: You're interesting; you're injecting complexity now!

Peter McCormack: Yeah, I know.  I was just saying, if I was the miner, I would want -- if I've got to switch off, I might be able to switch off for a day or two days.  If I had to switch off for, say, a month, I'm losing money as a mine.

Shaun Connell: So, remember the hedge, so the example that the miner says, "I'm going to buy a fixed price"?  So this example, where the miner's online and all of a sudden, the prices are above my breakeven, I'm like, "This is awesome", because I'm actually getting paid the real-time price.

Peter McCormack: With my miner switched off?  Okay.

Shaun Connell: That's a lightbulb moment, that's good!

Peter McCormack: Well, close, I'm close.  I think what's going to happen is, we're going to have a little chat afterwards and I'm going to be like, "Do I fully get this?" but I think I do.  I mean, look, it's fascinating.  Do you see a scenario where the grid itself may choose to start mining, or there's a merger between the generation companies and miners, or miners become generators, or generators become miners; or even the suppliers, Shell say in the UK, that's who I buy my energy off, they also become miners?  Where do you see this going?

We've seen it in that some exchanges, Bitcoin exchanges are starting to look like banks; and some banks are looking like they might start becoming Bitcoin exchanges.  Do you seen anything happening between miners and the energy market?

Danny Knowles: There was news today of Exxon mining.

Peter McCormack: Okay, what's that?  Can you show us?

Danny Knowles: Yeah, I can pull it up. 

Peter McCormack: Danny might already have an answer for us on this!  Okay, what's this?  So for people listening, "Exxon Weighs Taking Gas-to-Bitcoin Pilot to Four Countries".  Okay, "Exxon Mobil Corp is running a pilot program, using excess natural gas that would otherwise be burned off from North Dakota oil wells", oh, this is flaring, yeah, "to power cryptocurrency [no, to power Bitcoin] mining operations and is considering doing the same at other sites around the globe, according to people familiar with the matter.

"'The oil giant has an agreement with Crusoe Energy Systems Inc to take gas from an oil well pad in the Bakken shale basin to power mobile generators used to run Bitcoin mining servers on site', said the people, who asked not to be named, because the information isn't public.  The pilot project, which launched in January 2021 and expanded in July…", etc.  Anything else interesting there, Danny?

Danny Knowles: I think you've basically got most of it.

Peter McCormack: Look, keep going, "We continuously evaluate emerging technologies aimed at reducing flaring volumes across our operations.  Exxon expects to meet the World Bank's call to end routine flaring by 2030".  So, are they actually doing the mining themselves?  They're not, are they; it doesn't look like it?  But the point being is, they're seeing this now, and that's Exxon, who in the 1970s, their scientists realised that putting carbon in the atmosphere would increase the temperature of the Earth.  Thank you, Nathaniel Rich.  Pissed a few people off there.

So, back to my question, do you see any of these mergers?

Shaun Connell: Yes, and people are realising that Bitcoin mining is energy, it's all energy.  So, think about, you run a power generation company and the way it works for you, we talked about like you inject your power into the grid, and you receive what they give you; you don't get to have any say in it.  It's just whatever the price is, it is.  So imagine all of a sudden, and this is for the first time ever, there's a second customer that will buy as much as you can produce, for whatever the volume, the right size, whatever you want, and pretend this is something coming to you, and you have a say in what the price is. 

So, this is either, "I'm going to have somebody come to my site who's a miner, and I'm going to have an option", or, "I'm going to say I'm going to have mining on my site and I'm going to choose.  I'm going to say, 'Which customer is willing to give me more for my energy?'"  So, I'll use an example of in Ontario, Canada; that's my focus area for power trading, they have a lot of nuclear power.  So they've got, I think it's like 12 nuclear powerplants, they've got a whole bunch of hydro and they've got a bunch of wind, and there's something that's called "surplus baseload generation", and that's an announcement from the grid operator saying, "We've got way too much generation that we don't have enough demand and export capacity, that we actually have to do something to fix this". 

What they do is, they actually have the nuclear plants ramp down for six hours.  So, nukes don't ramp, they're not designed to ramp, so they had to come up with a process to essentially take some nukes offline, ramp them down slowly, and then keep them at the same, and then they ramp them back up.  Now, imagine there's an option, and there's two benefits on this one, saying, "I'm a nuclear powerplant and instead of despatching down", and there could be some payments that they get in Ontario that don't show up, but pretend that the choice is zero dollars for ramping down, it's going to say, "I'm an 800-MW powerplant, I'm going to put on 400 MW of Bitcoin mining". 

Then the scenario, instead of ramping down you say, "I'm just going to push my power to this mining facility".  So, now what you've done is you've just created a second customer, that you weren't just forced to say, whatever the grid's going to take, you're going to take.  And there's a second part.  So, we talked about what types of resources are flexible, and so flexibility is usually from fossil fuels for despatching up and down.  But you can put on this Bitcoin miner attached to your nuclear facility, and you can send it power.  And so, since your Bitcoin mining site can be dialled up and dialled down for how it's despatched, you've just put a supercharger on your nuclear plant that you can now actually provide flexibility to the grid, because you have the ability to ramp your nuke through the Bitcoin mining.

Peter McCormack: So, I wonder if they will build out their own facilities, or we will see mergers of companies?  That will be interesting.  Okay, potential flaw in the whole thing, because it's brilliant, we have a massive expansion of renewables, massive expansion of Bitcoin miners connecting to the grid, providing that flexible load balancing.  But say the Bitcoin price crashes to a really low price, it gets to the point where some of these miners can't actually profitably mine, do we then lose the load balance?  Does the Bitcoin price have to stay at a certain price to make this work?

Shaun Connell: So, there's a lot of -- it's healthy for these kinds of pullbacks to happen.  In power generation again, whenever load drops, the most uneconomic generator comes offline.  So, it's really healthy for us to have these essentially drawdown times, where you find out which of the miners are uneconomic and shouldn't be essentially online anymore, because they're using an inefficient kind of miner.

So, think about how, in the early days, people were mining on their laptop, and essentially nobody can mine on their laptop because it's not profitable, there's no way, so that's forever gone.  So then in 2017, some of these miners at the time were making, this S9 miner was making $450 an hour and in 2020, it went all the way down to $27, $28.  So, what happened was, some of the miners came off the network, which was very healthy, and you find out which ones are efficient, and which ones are swimming naked.  Then there's this push-up that can happen in the next push.

So, even in the absolute worst-case scenario, let's use the far tails, Bitcoin mining goes to zero.  I use the example in Ontario that they would have to export the power out of the grid and pay people to take it, because they had to find a home for it.  So, I remember conversations at the time, "Why don't we just find some toasters".  Instead of selling it for $0, you just find toasters that you can just switch onto the zero and collect zero, as opposed to having to pay somebody to take the power.

Bitcoin mining, in an absolutely tail event, is really fancy toasters that you can use for managing this despatching.  So, on the insurance products --

Peter McCormack: So, even if they're mining Bitcoin at zero, they're still getting paid to take the energy?

Shaun Connell: Because they're providing that flexibility, because they can be despatched up and down.

Peter McCormack: That is a lightbulb moment, fuck!  Okay.

Shaun Connell: And as we get bigger, so in 2017, there's maybe a couple of thousand megawatts, so 2 GW, so now we're at 12 GW, so now there's more flexibility.  Imagine four years later, and we're at, make up a number, like 60 GW.  So, those things are all providing flexibility.  So, the more it grows, the more that this kind of flexibility can be provided in other systems that have this need as they grow more renewables into the system.

Peter McCormack: So, there's this careful dance though over time of building out the renewable infrastructure, enough ASICs being available to also be built, and the network and price will grow with that as well; there's this dance between it all.  But it feels like the Bitcoin Network will grow with the growth of investment in the renewable infrastructure; it's kind of a little dance that's going to happen.  Wild!

Shaun Connell: And if you can think about it as well, in the summer of 2020, you could buy 1 MW worth of S9s, which is like 700 miners; you could buy 1 MW for $14,000.  And then, at the time, an S19 miner was something like $2 million or $1.5 million or something.  So, you think about it, just to get your return on capital from that S19 miner, which is now $3.5 million, requires you to get between $50 to $100 back every single hour.  But that S9 miner, that was considered almost like e-waste, scrap, these wind facilities can do a one-time payment, like the infrastructure required to have some miner racks, and have the right to have an electrical infrastructure. 

So now, they can actually buy these old-gen miners when there's a downturn and say, "We'll take those", because it's not a very big capex.

Peter McCormack: They're toasters.

Shaun Connell: Right, and it's better than me selling to the grid for -$10, I'll take $0 or $12 or $20, or whatever it is; it's obviously much higher than that.  So currently, Bitcoin miner is the buyer of first resort, because it's paying more than the grid.  But in the future, as more mining gets built out, the network gets a lot bigger, so you get to 200 GW, mining can act as this buyer of last resort.  So, instead of injecting power to the grid at say -$5, -$10, here's Bitcoin saying, "$20 or $15".  So, it's almost like, and this is a bit of mind-trip, it's setting a floor for energy around the world of saying, "Anybody can put this up and there's always a buyer".

If you think about it, if someone puts up a generation resource somewhere, it's got to be grid connected; there's no customer.  But there's now always a customer.

Peter McCormack: This is what Harry said, didn't he?

Danny Knowles: Yeah, exactly.

Peter McCormack: Yeah, interesting.  I do this occasionally.  I sometimes feel like Jeremy has a question.  I knew it!

Jeremy: One concern that's starting to pop up in my mind is on the political front.  I can see, is there a potential here where you mention two different types of -- there's a lot of terminology flying around here, but the one where it's retail driven and then they have to cut off their machines that are less profitable.  But then, you mentioned another one where the grid operator has a direct --

Peter McCormack: Wholesale.

Jeremy: Right.  Is there a potential here where it seems like that would be the approach that the grid and political parties might want, more so than the retail one, so that they can say, "No power, we need this, we're not going to leave it up to these banks, or whatever, who are running these mine farms"; is there a potential here where there's a political angle where they force some sort of bill, or something like that, to force people to not be retail and to be wholesale?

Shaun Connell: So right now, there's no power that policymakers have over load on what's just and what's not and you can't control it.  So, the message that I'm trying to get out is that Brad Jones, ERCOT's CEO, is saying that these can be a great asset for these grids in transition to say that, as you add in more renewables into your grid, you're losing control of part of your generation.

The grid in the past, loads, you can call them "dumb loads", so they need to be upgraded to smart loads, and saying, "You're going to lose control of your generation here, so you need to off-balance and gain control of load".  And this comes into the insurance part, and the importance of market signals, in saying that as grids go through these transitions, there's more insurance products that are needed to manage this new problem, because there was never this solar rent problem in 1995; that's come up now.

So, the market signal is saying, "There's this high-valued insurance product that is open to generation, open to anybody.  It's an open, transparent market".  So the miners say, "Oh, man, this is great, because as a miner" I talked about like a generator that says, "I'm going to sell (A) energy, or (B) insurance.  I can't do both, because I have to turn down to sell the insurance".  So, these controllable loads have a very unfair advantage that it's not in A or B, it's going to say, "I'm going to consume energy and simultaneously, I'm going to sell you this insurance product, because my opportunity cost is near zero, because my behaviour's not different.  I wasn't limited on how much energy I consume, whereas the power generator is limited on this".  So, my opportunity cost is zero, whereas the opportunity cost for a generator is saying, "What is the value of selling energy versus ancillaries; which one's to my benefit?"

So, a Bitcoin miner has the lowest opportunity cost to beat out all generation, which is providing a lower cost to consumers.  So, it's really self-serving in the sense that, for Bitcoin miners, I believe that as time passes, as they're getting bigger and having larger impacts, it's more obvious on where to go; because as a Bitcoin miner, you have an objective function.  It's like, "I've got to minimise my all-in delivered energy costs, so that's made up of my cost of energy, my cost of transmission, my cost of distribution.  How do I minimise those?"  The perfect market is one that says, "I can buy low-cost energy and I can get paid for providing a high-priced insurance product, because I'm going to buy this and simultaneously sell this, which means that my energy price is A minus B". 

Those market signals are drawing those miners to say, "You need help, because I can see the price, so I'm going to move to you because you're giving me the right market signal".  And, as you get more renewables, that signal's going to get bigger, and so they're going to naturally drive to these areas because of the markets that are created.

Peter McCormack: Where are the holes in this that you see?  Where are the risks, or the potential flaws in this?

Shaun Connell: In what way?

Peter McCormack: In that, considering Bitcoin miners as part of the network.  I mean, the biggest risk I see is actually regulatory, it's people like Elizabeth Warren; that is a risk.  But is there anything I'm not aware of?

Shaun Connell: It's kind of awareness of saying there's miners that are within the control of the grid operator, and miners outside.  And so, there's a lot of -- and they're both good.  But there's concern that just having visibility about, this is voluntary on how they can behave for some part, but just the value of being able to provide this service to these grid operators to manage these grids, I want to share with you that something I get worried about was that if nuclear power was invented today, it would be the magic pill.  It would be like, "Holy shit, this is it.  We've found this great technology, this baseload, can provide this energy, it's amazing".  But there's this perception of nuclear that's tied to bad events that happened, nuclear bombs, whatever it is.  But it's really hard to rebrand nuclear. 

So, my concern and where I'm trying to present the information to folks is that, this is the opportunity to brand mining before somebody else brands it for us; because if somebody else brands it, like Elizabeth Warren, we're in trouble.  But if somebody like Brad Jones brands it and we all start learning about this and instead of, for the past few years, I've heard a lot of folks say, "Bitcoin's good for renewables", the challenge is they don't really know how to explain it. 

So now, there seems to be a way to explain it, and then evidence, like saying, "During the ERCOT winter storm Uri, they turned offline.  And ERCOT, when there's a plant tripped offline, they provided primary frequency response, which means the heartrate dropped and they automatically dropped in seconds, and then recovered after this heartrate slip".  There's now stories of finding out how it's helpful, so we just need to be able to share that message, because if Bitcoin mining gets branded poorly, it's going to be a hard, long road to recover from that branding.

Peter McCormack: We're in my area now, brand and marketing!  We should think about this, Danny.  Bitcoin mining is the most important tool to combat climate change.

Danny Knowles: That's a title right there.

Peter McCormack: Get that in the FT; that would be good.  You know who we should try to get a conversation with again, is Katharine Hayhoe, get someone like her to understand this.

Shaun Connell: Who's she?

Peter McCormack: She's a climate scientist.  We interviewed her for my other podcast.  I used to have this other podcast, Defiance, which we stopped, but we interviewed her a couple of years ago.  She's in Lubbock, Texas, and we wanted to get her back on the show.  Her profile's growing.  She was on, was it Bill Meyer's show?

Danny Knowles: I think so, yeah.

Peter McCormack: Yeah, so her profile's growing and to get somebody like her or Michael Mann, one of the leading speakers on climate change, to listen to this and understand this, that would be helpful, because that would be a leading voice within that community saying, "Listen, this is a tool we can use to actually --" following Troy Cross's thesis as well, "to expand this, to expand the investment in renewables".  How much time have you spent looking at Troy's thesis?

Shaun Connell: So, Troy and I have spoken a few times.

Peter McCormack: Okay, great.

Shaun Connell: I think that it's, yeah, I'm supportive of what he's proposing and again, back to like it's not an "or" statement, it's saying, "Let's try a whole bunch of things and we'll see which one takes".  But it's not delimiting.

Peter McCormack: Yeah, I didn't totally buy his thesis when he talked about, "Everybody should mine the percentage of the Bitcoin Network they own"; I think that's a tough sell.  But what I did buy is that there are people who want to offset, and there are ESG budgets.  And whatever anyone thinks, ESG is not going away, the brand is out there.  If you can divert ESG budgets into Bitcoin mining, which supports and grows the investment in renewables, then that is just a natural win/win.

Shaun Connell: So, here's the challenge I have on what makes flexible Bitcoin mining under the control grid operators different.  Normal loads will have, they call them "REC obligations".  Renewable Energy Certificates are, you have to do something to show you're coming from all clean energy, or something.  So, here's where I have a challenge for that, is that when you flip over to your supporting grid operator and you're helping with load balancing and providing insurance, you're providing a service.  So, you're consuming, but you're giving.

The comparison is a battery.  So, a battery is on the grid to help to say, "There's a lot of renewables here.  We're going to inject and we're going to put that out later on [or] provide some insurance along the way".  But when you think about a battery, are you asking, "What are they doing with the energy when the battery puts it back in the grid; who's consuming that?"  Is that good cause?  Should we make the battery do something with that?  Or, for every megawatt that the battery puts backs in the grid, it consumes a little more than 1 MW, so should they buy RECs?  Should they show that they're buying green energy?

The default is, "No, they're important.  This energy transition requires a large amount of batteries", and it does.  But should they be in the same category as somebody that's just a load only and just taking?  So, it's back again in that a Bitcoin mining load that's just consuming and not giving back, not providing flexibility.  Okay, there's some type of standard, or some expectation that needs to be supporting it.  But another one that's actually providing to the grid and providing this flexibility, well I think that's different.  I don't know that it is, but I see it more as a battery than just a load-only that should have the same kind of requirements on green standards, if that makes sense.

Peter McCormack: Okay, that's fair.  Is there anything that I've not asked you about that you wish I had that we've not covered yet?  There's a lot to take in here!

Shaun Connell: I don't think so.  Well, I'm sure I'll think of some later, but yeah, it's been a lot.

Peter McCormack: Yeah, look, I never really listen back to my shows, just because I naturally hate my own voice and think I sound like a moron.

Shaun Connell: Yeah, I'm not going to watch this!

Peter McCormack: Yeah!  But I think I'm going to need to listen back to this one, there's a lot to take in.  But you think we've covered it all?

Shaun Connell: I think so.

Peter McCormack: All right.  If people want to follow what you're doing or find out more, where do they get hold of you, and do you actually want to hear from anyone?

Shaun Connell: Sure, yeah, I'll try it out.  So, my handle's @ShaunEnergy on Twitter and my DMs are open.  And I'd love to hear from folks, and I'm getting a lot of DMs about people with interests in doing academic papers.  It feels like there's a lot of people trying to arm policymakers with good, factual information.  Yeah, so @ShaunEnergy is across most of my socials.

Peter McCormack: Yeah, do you know Margot Paez?

Shaun Connell: Yeah.

Peter McCormack: So, we're getting her funded to do her paper on ERCOT, what's actually happening with the miners, which I think will be a super-useful document, just to send out to any other grid operator to realise.

Shaun Connell: Yeah, I'm writing one with Nic right now and talking about flexible Bitcoin loads as well.

Peter McCormack: Amazing.  When will that be done?

Shaun Connell: We think at the end of the month.  Yeah, so coming out soon.

Peter McCormack: I know what Danny's thinking!  Okay, this was amazing.  Thanks, Shaun, really appreciate you coming in to do this, a lot to take in.  I didn't understand it all, I know I didn't, and thank you for walking me through it, but yeah, this is great.  Good luck with everything you do, and I think we'll probably be doing this again sometime in the future.

Shaun Connell: Thanks very much.

Peter McCormack: Cheers.