Is the 2-week pace of Bitcoin’s Difficulty Adjustment Too Slow for Smart Grids?
TL;DR—If bitcoin mining and the power grid integrate into a single money-energy system, more frequent bitcoin difficulty adjustments might be in order.
Bitcoin Mining as an Enabler of Decarbonization
Far from being a liability for decarbonization, bitcoin mining can be a key ally of climate action and the growth of renewable energy. As I have explored more in detail in a previous article, there are long-term dynamics tied to the structural evolution of both bitcoin and the energy system pushing in this direction.
There is also a more immediate reason why bitcoin mining and the electric grid could mutually benefit by combining into a functionally integrated Internet of networks. This is that both large mining centers and smaller mining set-ups can easily ramp up or down their electric consumption at a moment’s notice, at the exact rates and for the exact durations that are needed to help balance the grid.
Bitcoin mining is ideally positioned to be a key component of a decarbonized energy system by providing flexibility services to smart electric grids. Flexibility services as demand and supply response are essential to enable the growth of distributed renewable energy and the gradual displacement of traditional, centralized, fossil-fuel power plants.
If the price per kilowatt-hour (kWh) paid by utilities to miners to provide response services to the grid is higher than the average value of mined bitcoin per kWh, it could be advantageous for miners to turn their machines down or off for a time, to make the energy they would have used available for the grid.
When energy demand is high and there is not enough power being produced or stored to satisfy it, bitcoin mining can be ramped down and the energy it would have used can be made available to the grid. When, on the contrary, the energy supply being produced is larger than the demand, surplus electricity can be easily absorbed by bitcoin mining, keeping the grid balanced. Such extreme flexibility, which very few sectors can offer to the grid, is a crucial and valuable service bitcoin mining can offer for the assimilation of ever-larger inputs of renewable energy into electric grids.
Electricity being the largest cost and the key variable for bitcoin mining also incentivizes miners to take a more active role in the sourcing, production and management of energy, beside playing a passive role as consumers. A possible consequence is that there might be increasing convergence between bitcoin miners and energy producers.
In this potential scenario, the two sectors will overlap more and more until they merge into one, with energy producers becoming also bitcoin miners and vice versa. That’s because, just as it makes sense for energy producers and utilities to create their own mining capacity as a hedge against the structural decline of renewable energy prices and as a source of grid flexibility, so does for bitcoin miners to have their own on-the-grid renewable power plants (or even fossil fuel power plants, more unfortunately for climate), which would provide very cheap power for mining and enhance their grid-flexibility role.
In this Internet of networks scenario, with bitcoin mining providing the energy underpinnings of the monetary system, it would also make sense to price energy in bitcoin.
But mining is first and foremost the cornerstone of bitcoin’s security. The electricity used by miners in the proof-of-work consensus process to add a new block to the blockchain provides a deep energy moat protecting the overall integrity of the bitcoin network from outside manipulations.
This possible, emergent double role that bitcoin mining can play leads to a series of considerations about the inevitable trade-offs between mining as the key security system ensuring bitcoin’s integrity and mining as an enabler and flexibility provider for smart grids powered by renewables.
On one hand, if bitcoin mining is to become a key player in providing flexibility services to the electric grid, the degree of flexibility mining can offer is a primary concern for the grid and the community at large. On the other hand, whatever level of flexibility mining guarantees to the grid has the potential to affect the degree of security it provides to the bitcoin network, which is of extreme concern for the bitcoin community.
So, any consideration about the flexibility of bitcoin mining for the grid has to take into account the positive or negative trade-offs such flexibility has vis-à-vis bitcoin mining’s security role. Bitcoin mining as a flexibility provider for the grid needs to find the right fit with bitcoin mining as a security system for the bitcoin network — especially if the same bitcoin mining activities are going to develop into a functionally integrated Internet of networks.
To the extent that bitcoin mining will become part of a functionally integrated Internet of networks, one possible stage where I can see mining’s double role as security system and flexibility provider being tested is bitcoin’s difficulty adjustment mechanism — which in turn has a direct impact on mining as a viable business.
Every Two Weeks vs Near Real-Time
The bitcoin network automatically readjusts the difficulty of adding a valid block to the blockchain more or less every two weeks (more precisely, every 2,016 blocks), so that, no matter the level of the overall computing power deployed by miners to add new blocks, it will always take about 10 minutes, on average, to add each new block.
What the difficulty adjustment mechanism does is push bitcoin miners to dial their energy consumption up or down to reach a given hashing power output in new coins. If the adjustment mechanism makes it more difficult to hit the target for creating a valid block, miners must spend more energy crunching numbers until they hit the target. If the adjustment mechanism makes it less difficult to hit the target for creating a valid block, miners can spend less energy to have the same chances as before to hit the target and maintain their rate of output in new coins. The difficulty adjustment mechanism also incentives mining hardware makers and miners to develop and adopt more and more energy-efficient equipment, consuming less and less energy for any given hashrate.
As mentioned in the same article I referred to before, the mining industry mirrors the power sector in some interesting ways. For instance, just as utilities and renewable energy producers have to deal with the volatility of sources like solar and wind energy — being progressively resolved through storage and a flexible grid — miners must deal with the volatility of bitcoin’s price and the difficulty adjustment in finding a valid block.
Both types of volatilities, at least in the short to medium term, call for management of risk that so far has largely escaped both sectors, if for different reasons.
“Renewable power companies have a long way to go to fully understand and integrate trading and price volatility into their operations. At the moment, renewable power companies rarely face spot price risks as they operate under government-guaranteed contracts”, former Glencore trader and billionaire Alex Beard recently told Reuters.
Beard’s team was trading as much as 7% of the world’s oil, but now the former trader is investing into large electric storage infrastructure to support renewables, in yet another sign of the way the wind is blowing for the global energy industry. But as renewable output grows and government price-support expires, the renewable energy industry will face a steep learning curve to understand how to hedge price volatility better, according to Beard.
The bitcoin mining industry might face an even steeper learning curve since it’s a sector that, unlike energy, started only about a decade ago as a cottage industry. Only recently, mining operations have started organizing and operating in a more structured way, with some of them becoming public companies and starting to use more sophisticated financial instruments to hedge their risks and run their business. This corporatization of the bitcoin mining business pushes mining closer to more traditional commodity enterprises.
“Similar to how capital markets developed for traditional commodities producers, it’s a natural next step for bitcoin mining to have its own auxiliary instruments to help miners better manage risks”, says Leo Zhang, founder of Anicca Research and an industry leader of hash power financialization, to Bitcoin Magazine.
Among such risk-management tools, there could be the creation of a hashrate exchange where buyers could purchase hashrate from miners and miners hedge difficulty adjustment volatility by trading derivatives on the future value of hashrate. There could also be both shorter-term and longer-term markets, with spot prices, wholesale prices and so on.
This starts to closely resemble traditional, present-day energy markets and markets for other commodities.
So, what do bitcoin’s difficulty adjustment mechanism and its two-week schedule have to do with risk management, volatility and the financialization of hash power?
Within the power industry, the shift from centralized electricity generation powered by fossil fuels and dispatched through one-way grids to decentralized renewable energy generation dispatched through multi-way, flexible smart grids is causing a change of paradigm in the way electricity supply and demand interact. This shift is also causing an important change in how power is priced: from largely regulated and demand-unresponsive prices based at the most on two or three times-of-use tariffs to something closer to real-time demand and prices of power.
This will dramatically increase power-price volatility in the near to medium term.
In the medium to long term, utility-scale power storage will be much more developed than today and it will take some of the edge off renewable energy’s volatility and highly responsive prices. Even so, real-time demand and near-real-time prices of power, along with much more granular weather forecasting, power-price forecasting, energy-flow forecasting, and the deluge of data generated by the faster interaction of traditional utilities with countless energy consumers-producers, renewable energy communities, virtual power plants, storage facilities, electric vehicle-to-grid setups and other configurations will make the power system much more complex than it is today.
So, how is the two-week schedule of bitcoin’s difficulty adjustment mechanism going to fit into this world based on real-time demand and near-real-time prices of power? The answer, of course, is nobody knows.
By default, difficulty adjustment’s two-week schedule would be yet another, slower, variable for artificial intelligence and machine learning algorithms to factor in with all the other forecasts and variables. Power trading, as well as hashrate trading, will be eaten by software, i.e., they won’t be the realm of traditional trading desks anymore, but they will happen in milliseconds, through proprietary software by specialized providers whose job is neither power-producing nor bitcoin mining, adding an extra layer to the nascent energy-money stack — such is the result of the corporatization, financialization, digitalization and commoditization of both electric power and hash power.
Near Real-Time Difficulty Adjustments?
One basic question from a systemic point of view is: if smart grids will be necessary to manage decentralized power generation with real-time demand and near-real-time pricing, and if bitcoin mining is going to be at the core of this decentralized energy-money system, wouldn’t this new Internet of networks be more responsive to market dynamics, more suited to its emerging function and an overall better fit for the distributed nature of both bitcoin and renewable energy if the pace of difficulty adjustments were more frequent? Shouldn’t the pace of difficulty adjustments mirror or approach the real-time frequency of distributed power generation, pricing, demand and supply, as well as the real-time frequency of bitcoin mining and transactions? After all, bitcoin’s price is set 24/7 on a real-time basis.
The canonic answer is that “adjusting [the difficulty] every block (or faster in general) increases user’s vulnerability to isolating attacks”, according to software developer, Blockstream co-founder and Taproot lead Greg Maxwell, aka nullc, on Reddit.
Maxwell explains in detail in his post the types of vulnerabilities that more frequent difficulty adjustments would open up for bitcoin. Recapping his position, he concludes:
Difficult is fundamentally a security mechanism and like others there are lots of ways to change it that would improve things in typical cases — but it’s the behavior in malicious and rare cases that really define its performance. Any kind of security measure is going to be worse on average than some less secure alternatives…
In any case, I dunno that I’d argue against a well-studied change (though I’m not convinced that we even know what we don’t know about this subject) including a correctly implemented change to more frequent updating (but maybe not every block).
For bitcoin, the negative trade-offs of speeding up the pace of difficulty adjustments are real — even if only on rare occasions and not in your typical day-to-day functioning. But it’s exactly during black-swan events that one wants a resilient system that is going to keep ticking no matter what. Bitcoin as a monetary base-layer needs to be secure.
Could bitcoin as a constituent part of an Internet of networks benefit from more frequent difficulty adjustments, without compromising the network security of bitcoin as a monetary base-layer? Could faster difficulty adjustments help miners better manage risk and hence help them become more resilient businesses, strengthening, in turn, their security role for the bitcoin network itself? If real-time adjustments or even adjustments for every block are out of the question, what other quicker frequency would make sense for the difficulty adjustment mechanism? And what would the positive trade-offs be for the grid and bitcoin as power infrastructure?
I am not able to address these issues from a developer or a network security point of view. I think it would be interesting to ponder them in the context of bitcoin’s evolving roles as both monetary and energy infrastructure, and I hope someone will. From the point of view of power production, power trading and risk-management for both bitcoin miners and power producers, I think it might be useful to explore what a higher frequency of difficulty adjustments could mean.
Flexibility in Electricity Markets
The electric grid works in real-time, with demand and supply requiring to be balanced in real-time, but at the electricity market level, we are still some time away from real-time exchanges. Across the European spot markets, the vast majority of trading is done in the “Day-Ahead” market through hourly products, meaning that traders buy and sell power the day ahead for the day after in one-hour increments. The dispatch periods for which the physical delivery of electricity is traded on the market refer to 60 minutes of physical electricity delivery.
On the “Intraday market”, participants trade continuously, 24 hours a day, with delivery on the same day. As soon as a buy-order and sell-order match, the trade is executed. Electricity can be traded up to 5 minutes before delivery and also through hourly, half-hourly or quarter-hourly contracts. More and more products with such lower time granularity are being introduced. This allows for a higher level of flexibility, so traders use the Intraday market to finetune the trades done in the Day-Ahead market, making last minute adjustments to balance their positions closer to real time.
Increasing the frequency of price signals in electricity markets improves operational flexibility for the grid in the short term, while optimizing investments in flexible generation capacity and enabling higher shares of renewable energy in the power system in the long term.
The idea here is that the better power prices reflect system conditions closer to real-time, the better the flexibility incentives for the system and the power market efficiency and resiliency. The nearer the closing time for a power transaction is to its physical delivery, the better market players can match power demand and generation, leading to more accurate scheduling of power generators, as explained by the International Renewable Energy Agency (IRENA).
Concerning both dispatch periods and transaction closing times, increased time granularity in power markets helps the grid forecast real-time operations with better accuracy. Shorter market time units allow more fine-tuning of wholesale power prices, which can trickle down to retail prices, also incentivizing flexibility through demand response services. With flexibility services valued on short-term markets, investments in more flexible generation capacity are also incentivized.
Flexibility in Hashrate Markets
Hashrate markets are nowhere near as developed as power markets, but, in principle, many concepts and dynamics that apply to power markets should be transferable and could arguably be relevant also for hashrate markets. Adjusting what needs to be adjusted, a unit of electric power output (a kilowatt-hour or kWh) and a unit of hashpower output (a kilo-hash per second or kH/s) and their multiples are not set apart by many degrees of separation. In the most simple terms, both outputs are physical, fungible representations of work, allowing the completion of certain tasks — electric power is multipurpose and can power an infinity of functions, hash power is single-purpose and is dedicated to powering what author and bitcoin “evangelist” Andreas Antonopoulos has called the “Internet of money”.
They exist in a Russian doll type of set-up. Electric power contains, i.e., it provides the space and the base-layer for hash power to execute its work, which in turn provides the base-layer for the blockchain to carry out its function, which in turn provides the base layer for other decentralized, trust-minimized, consensus-based operations. As the inner layers of a Russian doll contribute to its overall stability and sturdiness, so the layered configuration of “power” and the work it embodies can contribute to the overall stability and resilience of the Internet of network matryoshka.
Increasing the frequency of price signals for the main cost input — energy — in hash power markets would improve the overall operational flexibility and economic viability of hash power market participants in the short term, while optimizing investments in hash power capacity and enabling more crypto-miners to remain profitable in the long term.
One difference between the energy world and the crypto world is that the constituent components of the power grid are centralized organizations, which also tend to be fewer in any given jurisdiction than miners in the whole bitcoin network. The grid constituent components are Transmission System Operators (TSOs, transporting high voltage power over long distances) and Distribution System Operators (DSOs, dispatching medium and low voltage power locally). TSOs and DSOs represent the conjunction point, the middleman, between the supply side (energy producers) and the demand side (energy consumers).
In traditional energy systems, there are very few producers and very many consumers. The relatively small number of producers makes it easier for the grid to assess the supply side of the equation. As a middleman, the power grid should have as a clear a picture as any of the supply side of electricity — the grid is quite centralized, consisting of one or very few TSOs plus a few dozen DSOs, with a high degree of coordination between the two layers, making the subjects interacting with the grid very few on the supply side. Furthermore, the bulk of trades are done only intermittently in Day-Ahed merkets, exchanging not very refined, chunky hourly packages. And yet, notwithstanding this relatively “simple” setup, the grid will have to increasingly rely on more granular price signals to have better visibility of the system and accommodate the growth of renewable energy prosumers.
As the number of renewable energy generation systems increases, participants that are both power producers and consumers, or prosumers, also increase, making the assessment of the supply and demand side more difficult for the grid. Price signals that reflect as closely as possible the actual ebbs and flows of power demand and supply make it easier for TSOs and DSOs to manage complexity and plan for the future through more granular data points. Hence the necessity of a more granular timing in power market transactions, which would affect the operational and risk management of hash power in turn.
This higher degree of complexity could indicate that a nascent hash power market — where miners are almost by definition decentralized, uncoordinated and more numerous than TSOs and DSOs — might benefit to an even greater extent than electric power markets from more granular power-price signals.
As miners have no direct visibility of other miners’ operations, more frequent power price-signals would provide more flexibility in managing miners’ operations and risks, allowing more responsive hash power price-signals as well on nascent hashrate markets.
This dynamic could imply the need for a more frequent retargeting of bitcoin’s difficulty adjustment mechanism.
A Hypothetically Real Case
Bitcoin mining’s difficulty recently underwent its largest drop in history, after two things happened in a matter of a few days: Tesla’s CEO Elon Musk said on May 12 the car company would not accept bitcoin anymore due to mining’s excessive reliance on fossil fuels; while on May 18 China confirmed and tightened its bitcoin ban for local miners and banks. In the wake of these two events, the bitcoin hashrate dropped by about 67% in the six weeks between May 12 and June 27, when it fell to a two-year low. The difficulty decreased four times in the two months between May 13 and July 17, lowering the estimated number of hashes required to mine a block by about 45%, with a single-adjustment drop of about 28% on July 3rd, according to its predetermined two-week pace.
After about 5 months from those two events as of this writing, the hashrate has rebounded, but not totally recovered yet, being still about 20% lower than its May level, as miners are still relocating from China to friendlier jurisdictions in what has been defined as “the great migration”. Difficulty has also rebounded, by about 45% from its July low, but it also remains about 20% lower than its pre-Elon-announcement and pre-China-ban level.
Bitcoin price reactions to these two events were immediate and repeated, with valuations dropping by more than 47% from May 12 to July 20. From its July low, bitcoin’s price has now more than doubled (107%), and it’s about 5% higher than its pre-May 12 level as of this writing.
This whole chain of events has put to the test and proved the resiliency of the bitcoin network, as the blockchain clock kept ticking every 10 minutes or so without a glitch, thanks to the overall hash power still carrying out its work outside of China.
But, what would have happened if instead of two weeks the pace of the adjustment mechanism were, say, one hour, one day, or one week? This hypothetical real case is interesting because it allows us to consider the movements of hash power and difficulty as if under a magnifying glass, amplifying over a short period of time what might usually happen over longer periods.
On May 12, the day of Elon Musk’s announcement, regardless of any possible headwinds for bitcoin, mining difficulty actually increased to its historical high, in response to the fact that in the previous period miners had put more hash power to work. After the Musk-China announcements, the first downward difficulty adjustment happened on May 29. That means that for about 17 days after the two events started unfolding — as bitcoin’s price was falling by almost 39% and the hashrate was declining by 19% — the difficulty was still higher than it had ever been before.
For those 17 days, miners were forced to run up the steepest hill they had ever encountered in the race to find the next valid block, i.e., spending more energy per hash than they had ever had, at a time when they were collectively weakened in their hash power, i.e., producing a lower number of hashes per second. Less hash power multiplied by more energy expenditure might seem like a good compromise for the overall security of the network, leaving the energy moat protecting bitcoin sufficiently well-guarded. But the security moat around bitcoin could have been more effective if hash power had increased in response to a more rapid decrease in difficulty and energy expenditures.
If the difficulty had decreased more rapidly in response to the first fall in the hashrate, the hashrate drop could have slowed down and possibly reversed sooner, providing a higher level of security for the blockchain.
Looking at miners as an aggregate who provides a communal service securing the bitcoin network, out-of-sync difficulty and hashrate, mismatched demand and supply price-signals don’t necessarily assure better network security nor better company operations and returns, nor better energy management and balancing of the grid, if that were already a factor.
If the pace of bitcoin’s difficulty adjustment had been one day or one week, price-signal delays would have been less severe, miners’ hash-power supply would have been more in sync with network difficulty levels, arguably increasing the overall efficiency of the money-energy system without compromising its security and resilience — possibly striking an overall better result.
The same would be true also for periods when prices and power trends are directed upwards. One way or another, with wide time discrepancies between price signals and operational responses someone always leaves money on the table, money that could be better spent assuring what each participant wants and needs to do.
If it ain’t broken, why fix it? It might make complete sense to do absolutely nothing and not invent problems that don’t exist. But even when things could have gone on without modifying anything, a very prudent and conservative system as bitcoin has undergone slow, careful, meditated changes.
Could the difficulty adjustment mechanism undergo one such change? The integrity of money and bitcoin is at stake. But also the viability of modern energy systems, and with it, the integrity of climate and our planet, are at stake. There could be positive trade-offs between bitcoin’s roles, which would benefit both bitcoin and our planet.
To the extent that bitcoin and energy integrate into an Internet of networks, a more frequent adjustment of difficulty might become more relevant, as well as necessary and beneficial.
I am not able to go much further than this with this analysis, so I hope someone else will, exploring if it would still make sense, in Greg Maxwell’s words, to “argue against a well-studied change… including a correctly implemented change to more frequent updating” of bitcoin’s difficulty adjustment mechanism.