The winter storm that hit the United States earlier this week caused Bitcoin miners to reduce their workload and pull significant computing power from the network in a short period of time.
According to the data, between January 23rd and January 25th, hashrate dropped by 40%, approximately 455 EH/s went offline, and block generation temporarily slowed to approximately 12 minutes.
The fact that the steepest decline came from Foundry USA, the largest mining pool with the largest presence in the US, tells us that the decline was driven by restraint.
Why are so many miners able to shut down so quickly? Why did they choose to do so, and what do those choices mean for Bitcoin’s security budget, transaction flows, and the politics of connecting large industrial loads to a power grid that can be stressed in frigid temperatures?
Reduction 101: Miners as flexible loads, not fragile infrastructure
Although the definition of reduction is simple, the reality is somewhat complicated. At the simplest level, miners partially or completely reduce power consumption. This is because power is scarce, it’s expensive, or it’s contractually more valuable to sell the ASIC back to the grid than burn it out.
In the United States, particularly in Texas, the choice has matured into a full-fledged business model. ERCOT has explicitly created mechanisms for “flexible large customers” that can reduce load during peak demand times, citing Bitcoin mining facilities as a core example.
The idea is simple. If the load can be reduced quickly and reliably and repeatedly, grid operators can treat it as a pressure relief valve during severe load conditions.
In a real mining fleet, reductions tend to fall into three buckets.
The first is purely economic. Miners observe a simple spread: the revenue per unit of hash versus the total cost of producing that hash. If real-time electricity prices spike, the cheapest decision is to stop hashing.
This isn’t charity work, and it’s certainly not business ethics. This is just basic unit economics measured minute by minute, especially for miners exposed to wholesale prices.
The second one is already under contract. Some miners have signed demand response agreements where the “off switch” effectively becomes part of the product they sell.
Texas offers multiple ways for flexible loads to participate in reliability programs, and the past few years have generated numerous examples of miners benefiting from reducing contracted power or selling it back to the market during stress events.
The company disclosed that miners can earn money by not consuming electricity when the grid is under stress. In Riot’s August 2023 update, the company split its funding into two buckets. These include $24.2 million in “power credits” (described as power curtailment credits earned by selling contracted power back to ERCOT at market spot prices) and $7.4 million in “demand response credits” tied to participation in ERCOT’s demand response program.
A small, day-to-day version of this appears in nearly all of Riot’s monthly reports. In its November 2025 update, Riot cited an estimated $1 million in electricity curtailment credits and an estimated $1.3 million in demand response credits, noting that these demand response credits come from participation in the ERCOT and MISO programs and that the total credits are offset against the total cost of electricity.
According to Iris Energy’s August 2023 Investor Update, the company’s Texas site has generated approximately $2.3 million in “power sales,” primarily self-regulated power credits under hedging agreements tied to real-time ERCOT prices.
In this setup, mining sites become more like a hybrid of a data center and a power trader than the old mental model of a warehouse that just runs ASICs until they break.
The third is emergency or rules-driven. The state of Texas now expects the largest new loads to be able to accommodate curtailment as a condition for interconnection in a grid emergency scenario, explicitly naming crypto miners and data centers as targets.
This is important because it transforms the savings from a nice-to-have to something built into the operational plan.
What makes this week’s storm a teaching moment is that the incentives are aligned.
Cold snaps increase heating demand, tighten reserve margins and often trigger maintenance alerts. The storm caused severe disruption to the U.S. energy system, with soaring prices and operational strains reported across the region.
Therefore, for miners working on flexible load placement, curtailment is often the cleanest and most rational response to a grid that suddenly has megawatts worth more than terahash.
This is also why the situation at the pool level progresses so quickly. When a U.S.-focused carrier makes a cut, its pool registers it almost immediately. This week’s dampening effect was most easily seen by the visible drop in Foundry’s hashrate and the resulting block slowdown.
Although the network is global, marginal hashrate fluctuations can still be regional if enough capacity is concentrated behind a few carriers and grid structures.
Bitcoin difficulty timer: Why slow blocks are usually a temporary tax
Hashrate shock scares people. Because it maps directly to security. That’s true, but in a very narrow sense. This is because fewer hashes per second means that the cost of a brute force attack on the chain is lower than the peak hash rate.
But the more important operational question is what Bitcoin does when the hash quickly disappears. The answer is that Bitcoin has a built-in rebalancing mechanism with built-in delays.
Bitcoin targets one block approximately every 10 minutes, but does not continually adjust the difficulty. The difficulty is adjusted every 2,016 blocks based on the time it took to mine the last 2,016 blocks.
This structure creates a short-term “storm tax.” If many miners stopped today, blocks would be slower today. However, the difficulty doesn’t immediately drop to compensate. The network simply produces blocks more slowly until enough blocks have passed for the next adjustment to reprice the work.
This week you’ll see it happening in real time. CoinWarz’s difficulty dashboard showed that the network was running slower than the 10-minute goal, and the average block time during the tracked window was above the goal.
That was the real experience of that lag when block generation stretched to about 12 minutes. The number of blocks per hour was reduced, confirmations were slower on average, and memory pools could become bloated when transaction demand was steady.
However, slow blocks do not “break” Bitcoin, they are simply Bitcoin imposing a time cost on users and miners due to sudden changes in hash supply.
If the shock quickly wears off and miners come back online as prices normalize and grid stress eases, there may not be a need to adjust network difficulty. If the shock persists, use the following adjustments to lower the difficulty and bring the block timing closer to your goal.
Fee markets sometimes behave in ways that are confusing to casual observers. If demand is steady, a short period of slow blocks can increase fee pressure, but if the memory pool is not tight to begin with and demand is weak, it may pass quietly.
The more important point here is that Bitcoin’s design assumes that mining power is opportunistic and sometimes temporary. Difficulty adjustment is the protocol’s way of accepting that reality without turning every local infrastructure event into a system failure.
A stress-testing winter storm: Uri, Elliott, and what 2026 will bring
This is not the first time winter weather has affected Bitcoin. What has changed is the size of the U.S. footprint and how miners have been integrated into grid programs.
Let’s start with Winter Storm Uri of February 2021, a modern-day reference point for trauma on the Texas grid. While Uri triggered a historic demand surge, power generation failures on various types of fuel caused widespread power outages and political liquidation.
At the time, large-scale Bitcoin mining wasn’t very tied to Texas’ reliability plans. The industry in the state was small, and the concept of “miners as flexible loads” was almost entirely theoretical. This is very different from today’s setup, where reductions are easier to adjust and much more common.
Uri is important to this story because it sets the political backdrop. After such a crisis, new users of large-scale electricity are judged based on a simple question: Will they make the next emergency better or worse?
Then jump to Winter Storm Elliot for December 2022. This episode directly parallels this week’s hashrate pattern. Galaxy’s 2022 mining report describes Elliott as the moment when miners reduced hashrate by as much as 100EH, about 40% of the network hashrate at the time, in an effort to stabilize the grid.
Other academic and policy debates cite arguments of similar magnitude, supporting that Elliott was a major suppression event rather than a hashrate crash.
It’s a clear comparison because Eliot shows two things at the same time. First, large-scale miners can suddenly and extensively shut down during periods of extreme cold. Second, when miners build restrictions into their commercial relationships, those blockades become legible and, in some cases, expected.
What will 2026 add? This adds the reality that “flexible workloads” are no longer primarily about miners, but about a broader class of gigantic compute workloads.
The U.S. Energy Information Administration described Texas as a center of rapid growth in electricity demand, identifying data centers and cryptocurrency mining as key drivers, and pointing to ERCOT’s task force-style oversight of large loads.
This is important because grid policies will change once flexible loads are no longer niche. As AI data centers and other compute-intensive facilities compete for the same interconnection capacity and the same public patience, miners lose the ability to claim they are a special case.
These become one category in a broader debate about who gets the power first in times of stress and who pays for the grid upgrades needed to serve everyone.
Bloomberg’s coverage of the storm points in the same direction, discussing how large industrial loads, including crypto mines and data centers, reduced power usage during the event, and how ERCOT’s demand forecast changed as the situation evolved.
This framing in the mainstream media is a reminder that the next decade of US mining will be told not only through Bitcoin price cycles, but also through grid governance.
Therefore, this week’s hashrate drop is best read as a preview. The US share in mining remains large, and as computing loads continue to increase, we will continue to experience short-term network slowdowns like this due to weather events. The protocol can handle them. The political environment is less forgiving.
Bitcoin’s difficulty timer allows the chain to make reductions viable, and flexible load economics make reductions profitable for miners. An open question is whether regulators and residents will accept the deal. In other words, a big new baggage that promises to leave if asked in exchange for the right to be plugged in for the rest of the day.
(Tag translation) Bitcoin
