Introduction

Bitcoin mining demand response refers to a miner's decision to reduce or stop electricity consumption in response to grid conditions, a power-purchase agreement (PPA) provision, or a formal demand-response program, rather than continuing to mine. The economic question is not whether a curtailment payment exists, but whether incremental compensation and genuinely avoided costs exceed the mining revenue forgone and any additional costs caused by curtailment. This comparison depends heavily on the specific power contract, the demand-response program's rules, and the mining pool's payout method, so it cannot be generalized across regions or contracts. This article sets out a framework for making that comparison and the data each input requires.

Compare the two operating scenarios

For the same event and its associated settlement effects, compare continuing to mine with curtailing:

Net benefit of curtailment = incremental compensation + avoided costs − forgone mining revenue − additional curtailment costs.

Here, forgone mining revenue means expected income after pool fees but before site electricity and hosting costs. Incremental compensation is the difference in payments or credits between the two scenarios, not necessarily the full amount reported by a program. Avoided costs are expenses payable when continuing to mine but not when curtailing; additional curtailment costs are expenses incurred only because the site curtails. Count each item once. Costs that remain identical in both scenarios cancel out.

A positive result means curtailment has the higher estimated economic return under the stated assumptions. It does not override an existing obligation to respond to dispatch instructions.

For a hypothetical voluntary event, suppose a site reduces its load by a constant 1 MW for two hours, reducing consumption by 2 MWh. Assume it receives a separate $120/MWh performance payment, avoids $50/MWh in variable electricity costs, forgoes $300 in mining revenue after pool fees, and incurs $20 in restart costs. There are no other payment differences, penalties, or clawbacks.

The net benefit of curtailment is $240 + $100 − $300 − $20 = $20. Unchanged fixed hosting fees cancel out. If the $240 credit already included the $100 electricity saving, that saving could not be added again. These figures illustrate the calculation; they are not market rates or a profitability forecast.

Identify program rules

Before estimating any dollar figure, a miner needs to confirm what kind of settlement it is actually eligible for, because different mechanisms are not interchangeable.

In ERCOT, for example, a Load Resource is a load registered to provide ancillary services or energy in the form of demand response, and ERCOT maintains separate qualification procedures for Controllable Load Resources and Non-Controllable Load Resources (ERCOT). Participation typically requires registration, qualification testing, and compliance with performance requirements specific to the program. Other markets and bilateral PPAs may instead offer a credit for unused contracted power, which is a separate mechanism from a grid operator's ancillary-service payment.

A miner should document, for each applicable program or contract clause:

• the trigger conditions for a curtailment event (price threshold, grid operator notice, or contractual option);
• whether compensation is based on verified performance, availability, or a fixed credit for unused power;
• the settlement basis, including the rate per MWh for energy or the capacity rate per MW with its applicable hour, day, or other contractual period, plus the invoicing or payment timeline;
• any payment reduction, clawback, or penalty for non-performance if the miner fails to curtail when called; and
• whether the electricity contract is take-or-pay, which affects whether the miner actually avoids energy cost by curtailing.

Separate the decision to enroll or commit capacity from the evaluation of an event after that commitment. ERCOT states that Load Resources scheduled or selected in its Day-Ahead Ancillary Service Market may receive capacity payments regardless of whether they are dispatched in real time (ERCOT). Such a payment is not automatically additional revenue from a particular curtailment event. Compare what would actually be retained or lost in each scenario, and account for any clawback only once. Once a miner has committed to providing a service, the applicable dispatch and performance obligations still govern its response.

These rules vary by jurisdiction, utility, and individual contract, so a curtailment plan developed for one site should not be assumed to apply to another without reviewing the governing documents.

Estimate foregone mining revenue

The second input is the mining revenue the operation would likely have earned had it continued operating during the curtailment interval. This estimate should be built from pool-side and network-side data rather than assumed from nameplate hashrate.

The pool's payout method changes how this estimate should be constructed. Mining pools calculate a miner's contribution using shares, which are proof-of-work solutions that meet a pool-assigned target; this pool target is numerically higher, and therefore easier to meet, than the Bitcoin network's block target, so shares are easier to produce than network-valid blocks (Bitcoin Developer Guide). A share does not by itself guarantee that the pool has found a block.

ViaBTC's current documentation illustrates why payout method matters for estimating a loss. Under PPS+, the block-reward component is calculated using shares, current network difficulty, and a 4% fee, and is settled hourly; the transaction-fee component is instead distributed using PPLNS logic, subject to a 2% fee, after a pool-found block reaches six confirmations, based on the miner's share of pool hashrate over the preceding five difficulty rounds. Under PPLNS, both the block-reward and transaction-fee components follow this five-round, post-confirmation allocation method with a 2% fee (ViaBTC Help Center).

The practical implication: a PPS+ miner can reasonably forecast the block-reward portion of lost revenue using recent hourly settlement data for the same hashrate, labeled clearly as an estimate. The transaction-fee portion, and any PPLNS-based earnings, should not be attributed to the exact curtailment hour, because the relevant reward depends on blocks the pool actually finds and on a rolling difficulty-round window. A more defensible approach is to measure the effect over a longer period that fully captures the affected PPLNS window, or to present the shorter-window figure explicitly as an estimate with stated assumptions.

Inputs needed for this estimate include: the device's online/offline timestamps and measured hashrate before the event; valid and rejected shares from pool records; current Bitcoin network difficulty; the pool's documented fee and payout method; and a BTC/USD price with its valuation timestamp if a fiat comparison is required. Pool-estimated hashrate should not be used to infer device efficiency (e.g., J/TH), since that calculation requires measured power input and hashrate over matching windows.

Include restart and hosting costs

A complete comparison must distinguish costs that change between the two operating scenarios from costs that remain the same, without re-counting costs already reflected elsewhere.

Relevant cost items include:

• Avoided electricity cost — the variable energy cost not incurred while offline, which may or may not be separate from a power credit depending on the contract (see below);
• Restart costs — any additional energy draw, labor, or equipment wear associated with bringing ASICs back online;
• Hosting or facility fees — fixed hosting charges that typically continue regardless of curtailment and cancel out when identical in both scenarios; include only any actual difference in charges;
• Non-performance penalties and clawbacks — assign these to the scenario in which they arise. A penalty for continuing to mine when called is a cost of non-performance, not an additional cost of compliant curtailment. Do not also count a lost payment as a separate clawback if it is already reflected in the compensation difference; and
• Pool payout effects — include the effect of a reduced contribution to a PPLNS window within forgone mining revenue, not as a second cost. Only a separately documented reward-eligibility rule would justify an additional adjustment, and only if its effect is not already included in the revenue estimate.

The most consequential accounting risk is double counting between a power credit and avoided electricity cost. If a demand-response program pays a separate, verified amount for load reduction, that payment and the avoided energy cost may be legitimately additive. But if a reported credit already reflects the value of unused contracted power — as in a PPA buy-back arrangement — adding avoided electricity cost on top of that credit can overstate the benefit of curtailing. Riot Platforms' 2025 Form 10-K states that its power curtailment credits represented both sales of unused power under its PPAs and participation in ERCOT and MISO demand-response programs, two mechanisms that should be modeled separately rather than combined by default (Riot Platforms 2025 Form 10-K). Riot reported $56.7 million in such credits for the year ended December 31, 2025, compared with $33.7 million for 2024 — a year-over-year difference that demonstrates how materially these figures can move with grid and market conditions, though the filing does not disclose an event-by-event breakdown suitable for a payment-versus-lost-revenue ratio.

Stress-test settlement terms

Before relying on a curtailment estimate for planning, a miner should test it against variation in the inputs that drive both sides of the comparison.

On the power-market side, settlement value depends on the applicable program's price formation, which can differ by season, time of day, and system condition. On the mining side, expected revenue depends on hashprice and network difficulty, both of which change independently of any single operator's decisions. Academic work using Texas market data has found that Bitcoin-mining load response weakens when hashprice is higher, implying that a curtailment threshold calibrated under one hashprice environment may not hold when mining revenue conditions change; this finding comes from a preprint and should be read as research rather than settled industry consensus (arXiv preprint).

A useful historical illustration is Riot's August 2023 Texas heat-event disclosure, in which the company reported curtailing more than 95% of its load during peak-demand periods, produced 333 BTC that month, and estimated $31.7 million in combined power and demand-response credits (Riot Platforms August 2023 operations update). This shows that curtailment-related proceeds can be substantial at scale, but the filing does not disclose the BTC production the company would have achieved absent curtailment, so the $31.7 million figure cannot be read as proof that the payment exceeded forgone mining income for that specific month.

Stress-testing should also examine the sensitivity of the comparison to BTC price movement during the valuation window, since mining revenue is typically BTC-denominated while most settlement payments are fiat-denominated.

Make a curtailment checklist

The following items can help structure an event-level evaluation; they describe one practical approach rather than a mandatory industry procedure.

1. Event interval — start and end time of the curtailment window.
2. Curtailed load and energy — record the load reduction in MW, its duration, and the resulting energy reduction in MWh against the applicable baseline. Use metering data where available; for constant reductions, MWh equals MW multiplied by hours.
3. Settlement basis — program type, rate, applicable settlement period, and whether payment is based on availability, performance, or a fixed power credit. Record the payment difference between the scenarios rather than assuming all compensation is event-dependent.
4. Mining revenue forecast or counterfactual — expected revenue after pool fees and before site electricity and hosting costs, the pool payout method, block-reward versus transaction-fee treatment, and the measurement window used.
5. Avoided electricity cost — confirmed as additive to, or already embedded within, the settlement amount.
6. Cost differences — restart costs and any changes in hosting fees, penalties, or clawbacks, assigned to the scenario that incurs them. Exclude unchanged fixed costs from the difference.
7. BTC/USD valuation timestamp — the price source and time used for any fiat comparison.
8. Net result and stated assumptions — incremental compensation plus avoided costs, minus forgone mining revenue and additional curtailment costs, with each input's source and any limitations noted explicitly.

Conclusion

Participating in demand response is a power-market and contractual decision as much as an operational one, and its economics are specific to the miner's jurisdiction, power contract, and pool payout method. For a voluntary event, curtailment becomes more favorable as incremental grid compensation and avoided costs rise relative to forgone mining revenue and additional curtailment costs. Higher expected mining revenue after pool fees makes curtailment less attractive, all else equal. Existing service commitments must still be honored. A sound evaluation keeps curtailment payments, avoided power costs, and mining revenue as distinct, clearly sourced inputs, measured over a consistent window, rather than treating any single public disclosure as a universal benchmark.

FAQ

Is a curtailment payment the same as mining revenue?

No. Demand-response compensation may pay for availability or verified load reduction, while mining revenue is BTC-denominated income credited by a pool for contributed work. Track them separately and compare only the amounts that change between continuing to mine and curtailing.

Can every mining operation participate in demand response?

No. Eligibility depends on the grid operator's program rules, the miner's power contract, and jurisdiction-specific qualification requirements, such as ERCOT's Load Resource registration process. A miner should verify current program rules and contract terms before assuming participation is available.

Does curtailing always avoid the full electricity cost for that period?

Not necessarily. Under a take-or-pay electricity contract, the miner may still owe contracted costs regardless of consumption, so the avoided-cost assumption should be checked against the specific contract.

How should a PPLNS miner estimate lost revenue from a short curtailment?

Because PPLNS allocates rewards based on a miner's share of pool hashrate over several preceding difficulty rounds and depends on blocks the pool actually finds, a short event's effect is better measured over a longer, aligned period rather than attributed to the exact curtailment hour.

References

1. Riot Platforms, Inc. Form 10-K for the year ended December 31, 2025.
2. Riot Platforms, Inc. August 2023 Operations Update.
3. ERCOT. "Load Resource Participation in the ERCOT Markets."
4. Bitcoin Developer Guide. "Mining."
5. ViaBTC Help Center. "How Are Profits Calculated?"
6. Majumder, Subir. "Hashprice moderates the electricity demand response of Bitcoin miners." arXiv preprint, version 2, June 3, 2026.