Elephant in the Grid
Not as bullish on the bitcoin mega-miners.
Much hype has been shilled the last couple years around the potential for bitcoin miners to improve the capital and operational efficiency of utility companies and our electrical grids. As a hardcore bitcoiner I am as excited about this as anybody, but in my usual contrarian fashion I would like to point out that that there may be an elephant in the room here.
Before I dive into things, let me clarify why mining is good for the grid for those of you new to the subject matter.
Bitcoin mining is a reliable, portable and flexible demand on energy. Mining computers can respond to electricity supply constraints rapidly and release or increase their power consumption when the market demands it. By giving the mining computer a price signal, its logic can rapidly change its computational load to fit market needs within seconds. This is in stark difference to a traditional energy consumer, say a manufacturing facility.
For example, at night when there is a trough in energy demand but supply from unreliable sources like wind may be high, the bitcoin miners can eat up that excess generated power by increasing their computational load. Conversely, if it’s day time and energy demand by the grid is high but the clouds have covered all of the solar farms, bitcoin miners can rapidly turn down computational load and free up power for everybody else.
Few people would argue that running a mining computer is more important than the home or business keeping the lights on. Subjective value dictates that the consumer in most need stays in business and that consumer won’t usually be the mining computer, which is happy sitting idle on a shelf.
The Folly of the Mega-Mines
If you follow the bitcoin mining space you would no doubt by now have routinely seen Riot Blockchain, Marathon Digital, Core Scientific and other other large mining farms market themselves as gift to the grid here in North America. I myself am a cheerleader of these projects, they are absolutely great for the grid, but there’s more nuance here. A few elephants in the grid, if you would allow me.
The big ones:
There are more energy efficient applications of mining on the grid that can achieve the same objective as large, centrally owned and centrally located mega-mines as computational load response.
Computation load variation in response to market price signals can be achieved equally both with single large, central datacenters as with multiple, smaller datacenters of the same total load.
Electrical grids are currently inefficient precisely because so much capital has been wasted installing unreliable energy generation like wind and solar instead of maintaining and upgrading reliable generation. Bitcoin mining, as a load response, does not fix this.
Large bitcoin mines are a security threat and regulatory honeypot.
Since Riot Blockchain is the biggest fish in the pond with a single facility plan exceeding 750 megawatts in one location, let me pick on them to illustrate the points above.
Unnecessary transmission losses
Riot’s ‘Whinstone’ facility in Rockdale, Texas currently pulls substantial power from a large grid interconnect:
Riot taps directly into this interconnection to draw 300 MW of that juice, which powers 120,000 high-speed mining computers stacked in racks 30 feet high in three narrow buildings, each longer than two football fields. Construction is under way to expand to 750 MW, with 130,000 more machines to be installed by the end of 2022
A grid interconnect is, economically speaking, a great place to install one massive mega-mine, but I would argue that it is an inferior application compared to simply distributing many smaller bitcoin mines at each upstream generating station instead.
The closer a load is to the source, the less transmission and distribution losses that load will incur with drawing the electricity. I don’t have a specific number on how much energy waste in transmission to the example Whinstone facility may be, but some quick googling estimate between 15-22% of the total power generated on utilities is lost in transmission and distribution to the end user. If we take the average of 18% losses, at 750 megawatts that would represent over 135 megawatts of power losses just to feed one bitcoin mining site.
A lot of this power could potentially be saved and turned into bitcoin had the mining computers been distributed behind the meter at each power plant, instead of being lost in thousands of miles of copper, voltage transformation and more.
Now I’ve simplified this argument of course to make a point, but distributed mines do not typically have the same economies of scale from a construction and execution standpoint, so the big mining farm in one location will be capitally cheaper up front relative to being split up across many different locations.
In my estimation the Whinstone facility is superbly engineered and appears to be extremely capital efficient in terms of capital cost per power capacity deployed ($/kW), maybe the best in North America. However, does the lower up front cost justify the potential ongoing energy losses compared to parking the same computational load at each generation facility feeding it? Considering energy cost continues to trend as the most important economic metric in bitcoin mining over the long run, I have my doubts.
I could write a whole article on pros and cons of large centralized mines vs small distributed mines, but I’ll leave that for a future piece. For example I didn’t compare the costs relating to legal contracts and execution risks of centralized vs distributed facilities, the regulatory risks, financing risks etcetera.
The point is that energy transmission losses could be avoided if the miners were distributed closer to the source of power, these big facilities are just not that exciting to me for this reason.
All miners can flex their load
While the marketing arms of mega-miners are hard at work promoting that they are best suited to do load response, the reality is anyone can do it just as well as them.
For years, ERCOT has struggled with fluctuating energy prices and sporadic service, which is why it strikes deals with flexible energy buyers, like crypto miners. Through established “demand response” programs, ERCOT will actually pay major industrial users to cut power.
All a mining computer needs is a price signal from the electricity retailer and through an automated algorithm the computation load can vary from 0 to 100% of its capacity in response to price signals in seconds. Even the lowly home miner can do this if their electricity retailer gave them a real time electricity price signal and they programmed an advanced firmware, such as BraiinsOS, to respond to it. As long as the retailer could specifically bill the user for the fair applied value of the energy used then it should be relatively straight forward to execute.
The main thing is the electricity retailer would need to offer the home miner a load shedding program and be able to signal price and measure consumption real time via a smart power meter IOT device. The Retailer may have to serve and receive information to and from the home miner who signs up to the program.
There could be two typical scenarios:
Generation surplus scenario
If a electricity retailer has an oversupply of generation then:
The Retailer lowers the cost of electricity to the market/user to encourage loading
The home miner ASIC firmware polls the Retailer’s server for real time pricing information. It receives signal that power is cheap due to generation surplus. ASICs automatically overclock to a power level and hashrate that the user mapped out relative to the electricity price.
The result is more power consumption by the flexible load and therefore more revenue / profit is created for both the miner and the the grid’s generation network than otherwise would be. Profits can be reinvested into making the grid more robust.
Some generation plants will save costs as well for not having to ramp down, especially in lieu of highly intermittent and unpredictable wind and solar generation. Intermittency is generally costly, especially if it is unpredictable. For example having unnecessary manpower on standby.
By eliminating these costs the increased grid efficiency yields lower aggregate energy costs for ALL users on the grid.
Generation shortage scenario
If a electricity retailer has a overdemand of generation on its grid then:
The Retailer increases the cost of electricity to the market/user to encourage shedding
The home miner ASIC firmware polls the Retailer’s server for real time pricing information. It receives signal that energy is expensive due to supply shortage. ASIC automatically underclocks to a lower power level and hashrate that the user mapped out relative to the electricity price.
The lower energy consumption frees up generation for more critical energy users. Grid brownouts, blackouts or generation overloads are avoided.
Significant costs are avoided from preventing business interruption across the economy.
In either scenario the electricity retailer would measure the energy consumed through it’s smart power meter and bill the miner equaling the product between the variable rate and energy used.
This is just an example indicating how trivial a load shedding program can be made to be, though of course this is not exhaustive. I am not even suggesting that home miners would be better off doing this. Like I said earlier the flexible load should be closest to the source to minimize transmission losses and maximize grid efficiency.
Right next to the source is the most optimal location for a bitcoin mine, and why not? Bitcoin mining is a location independent market so anyone can flexibly mine with the right tools.
Bitcoin doesn’t fix unreliable generation
In normal times, when the world hasn’t lost their minds from fiat abuse and carbon hysteria, energy grids in the first world worked a lot more reliably than they do today. Beyond the occasional upset condition like rare winter storms that catch a normally warm climate like Texas unawares, power grids were generally reliable due to coal, nuclear, hydro and natural gas making up the bulk of the reliable energy supply.
For the engineers designing these power distribution systems the main challenge has traditionally been balancing reliable supply with an unreliable and unpredictable demand load. Weather of course is highly unpredictable which constantly throws huge curveballs at power generation planners every day as temperature swings directly correlate to variation in power consumption.
Today the problem has compounded further with the subsidization of “renewable” (unreliable) energy like wind and solar farms. Regions that went overboard on unreliable generation projects like Europe are now dealing with insanely high reliable fuel prices as they scramble for scarce supplies. Recently even in California rolling blackouts were a problem. We’re talking about one of the richest states in one of the richest countries in the world. Talk about pathetic.
Bitcoin proponents may say bitcoin fixes unreliable generation by having computational flexible load, but this is really just a band-aid slapped on top of a severed foot.
Yes, flexible bitcoin miners help consume the oversupply of energy and they also help ease the pain from the undersupply of energy to the consumer who needs it to live and do business, but it does NOT fix unreliable generation and a resultant crippled grid.
Bitcoin simply does not make the wind blow or the sun shine when we need it most.
So what does fix unreliable generation? Not installing it, for one. But in all seriousness, only efficient energy storage can fix this problem and to date nobody has invented anything that is even remotely economically feasible, let alone ethically sourced. Until that day comes we are stuck with the reality that we need more reliable generation.
Are mega-mines good or bad for Bitcoin?
Well, it depends.
The short answer is yes they are a net good but they can also be used to censor subsets of the bitcoin market, so we should be wary of these risks.
Bitcoin’s success is derived from its ability to resist censorship which in turn depends on the ability of its users to maintain their privacy. Privacy both for merchants using bitcoin in trade and for miners confirming transactions and keeping the entire decentralized network in sync with each other.
While large bitcoin mines themselves don’t put individual user privacy at risk per se, the problem lies in that these facilities are impossible to stay private from prying eyes and are therefore easily regulated (co-opted) at low cost.
Compare how easy it would be for some authority in Texas to slap some arbitrary compliance rule on Riot’s Whinstone facility compared to trying to enforce the same rule on the same 750 megawatts worth of individual miners scattered all across the Texas grid. At 3.3 kW per ASIC and one ASIC per person, that’s 225,000 individual people a regulator would have to enforce the same compliance rule, whether that’s a mining license, permit, tax, royalty or ban. Good luck finding all those people let alone enforcing it, doing so would be extremely costly compared to enforcing one single facility and owner.
What if Whinstone was built in a country like present-day Venezuela, a country that recently has been known to nationalize entire industries? What if Maduro confiscated the entire 750 megawatt mine and started using it in collusion with partner countries to censor the entire network using 51% attacks?
There are tons of reasons why mega-mines can be used against the bitcoin network with as little as the stroke of a pen.
Fundamentally bitcoin is designed to challenge the State and more specifically it is an attack on the State’s money printing golden goose. It does not seem too farfetched to me that the powers that are interested in maintaining their fiat dominance will band together to attack the bitcoin network in some way.
Censorship and regulatory moats.
As we have seen in China, a regulator can swoop in any time and absolutely disrupt everything. Everything they know about anyway. Luckily China did not nationalize their bitcoin mines and instead kicked them out of the country, which just caused them to decentralize over the globe. I don’t know if they will make that same mistake again next time.
It isn’t surprising to me to see most of the mega-mining corporations in North America constantly pandering to regulators and the detestable ESG social-justice-warrior-financiers, lobbying state representatives and pretending to be mining on “clean” energy when all they are doing is buying access to market themselves as clean.
Marketing teams “green-labelling” themselves because they are buying social credit tokens called ‘Carbon Credits’, thereby supporting a dangerous and oppressive carbon accounting system. Only well financed businesses can even play this game, so when the large miners play it they effectively are trying to build themselves regulatory moats that keep competition at bay.
And how can we blame them? The system has made it profitable for large miners to legitimize and pander to regulators, even when the rules they are tabling are ineffective and pointless. These corporations certainly aren’t evil, they are just playing by the rules of the the game that were written by people who print money. For me I am worried it will usher in rules that will hurt the little guys, which seems to always be the case.
For these reasons and more I am not so bullish on what I’m seeing from these mega-miners.
Maybe I’m just paranoid?
Between the OFAC initiatives we’ve seen from mining pools and the brown-nosing fiat mega-miners, I feel like its only a matter of time before things get ugly on the regulatory front in the bitcoin mining space. It’s almost as if it is the core business plan for some of these mega-miners to lobby for regulations that put them at a competitive advantage…
Things are getting weird quickly, my friends. Just last week the Canadian government threatened the leaders of Kraken and Coinbase bitcoin exchanges while also illegally shutting down Canadian’s bank accounts for donating to peaceful protestors. They’re trying to blacklist cryptocurrency based donations as well.
Despite all of this scary stuff I am actually ultra optimistic on the future of bitcoin mining and the continued rise of decentralized mining. It is destined.
It is Satoshi’s Vision.
I plan on writing all about ‘The Future of Bitcoin’ in a future piece, so if you liked this commentary please subscribe below and share with your friends.
@coinheated on Twitter for inspiring the power metering commentary
@GideonOPowell on Twitter for triggering me enough to clarify some things :P
Thanks for reading my Bullish Thoughts! Subscribe for free below