When Can Bitcoin Mining Count as Circular Energy Policy?

An annual renewable share describes a generation portfolio. It is not a project-level emissions factor and not a finding of surplus. A new constant load may be served by different resources at different hours, require network investment, alter exports, or displace another use of scarce low-carbon electricity. The governing question is what changes because this facility connects and operates? A bounded review should identify how generation or imports respond over time, what connection and upstream capacity the project requires, and which alternative use the allocation is being judged against.

This distinction matters in both directions. A hydro-dominant province cannot assume that every additional megawatt is consequence-free. A more carbon-intensive system cannot assign every project the annual average grid factor as if it were the marginal generator in every hour. Annual averages may describe context. They do not, on their own, measure induced emissions.

The technical ability to switch off is not yet demand response. A grid service requires, at minimum, a defined baseline, a dispatch right, notice and duration rules, telemetry, a performance test, settlement, and consequences for non-performance. Voluntary shutdown when market prices rise may protect a miner’s margin, but it should not be counted as a public reliability service unless the system operator can rely on it under specified conditions.

The relevant metrics are therefore event-specific: megawatts nominated, megawatts actually reduced relative to an approved baseline, response time, duration, availability, rebound after the event, and failures. Revenue received by a facility is not a substitute for these measures, and a payment made under one market design cannot be imported as evidence of value in another.

Nearly all electricity consumed by computing equipment ultimately leaves the facility as heat. That physical fact says little about circular value. The policy unit must be useful heat delivered and substituted, not gross heat produced or even heat captured at the rack.

A credible heat claim must disclose the delivery boundary, supply and return temperatures, hourly and seasonal heat demand, distance to the user, network losses, uptime, the heat source actually displaced, and the auxiliary energy drawn by heat pumps, pumping, and backup. Avoided emissions must then be calculated from that counterfactual, net of auxiliary electricity and losses. Where the recipient would otherwise have used low-carbon electricity, the emissions benefit may be small; where the mine closes before a district-heating asset is repaid, an apparent circular benefit can become stranded infrastructure.

For this reason, a reported percentage such as “heat recovery efficiency” is insufficient without its denominator. It may refer to heat captured inside a cooling loop, heat leaving the facility, heat accepted by a network, or useful heat delivered to customers. Those four quantities differ, and a percentage that does not name which one it measures cannot be read as a delivered-heat figure.

Mining equipment is mobile and revenue is volatile. A province evaluating a long-lived interconnection or heat network should therefore test the durability of the load against the life of the public or shared asset. Connection security, decommissioning obligations, equipment-disposal records, and responsibility for stranded network costs belong inside the admission decision, not after it.

These four tests define the paper’s use of circular. Circularity is a verified relation among a load, the electricity-system counterfactual, a callable service, and an external use of a by-product, rather than a property of a generation pie chart.

Hydro-Québec’s approved 2026 Electricity Rates provide the clearest Canadian example of a mining-specific operational obligation. Rate CB applies to an annual medium- or large-power contract where at least 50 kW of installed capacity is dedicated to cryptographic use applied to blockchains. Service is non-firm. Hydro-Québec may reduce real power demand to 5% of the customer’s reference value for up to 300 hours in a rate year, on two hours' notice. Consumption above that limit during a curtailment period is billed at $1.038/kWh (Hydro-Québec 2026a). Hydro-Québec’s customer page confirms the 50 kW eligibility boundary and identifies the April 1, 2026 rate adjustment (Hydro-Québec 2026c).

Those tariff terms establish a dispatch right and a financial consequence. They do not establish actual event performance. A complete evaluation would still require anonymized event logs showing calls, baselines, delivered reductions, exceptions, and rebound.

The revised rate sits at a different legal stage. In February 2026, Hydro-Québec proposed a revised Rate CB averaging 19.5 cents/kWh (Hydro-Québec 2026b). The Régie opened docket R-4333-2026, which remained active at the source freeze and contained procedural decisions and filings, including revised proposed tariff text filed June 29 (Hydro-Québec Distribution 2026; Régie de l’énergie 2026b). Hydro-Québec had requested a decision in time for a November 1, 2026 effective date (Régie de l’énergie 2026a). As of July 15, however, 19.5 cents was a proposal, rather than the approved current average rate.

The institutional lesson is narrower than “Québec integrates mining.” Québec has specified a customer class, priced unauthorized consumption, and reserved a substantial curtailment right. Whether that architecture produces net value depends on performance and cost data not published in the sources reviewed here.

The Energy Statutes Amendment Act, 2024 received royal assent on May 16, 2024 and added section 21.1 to the Utilities Commission Act. That provision authorizes regulations prohibiting electricity service for cryptocurrency mining indefinitely or for a specified period, and authorizes rates, limits, timing rules, and conditions (British Columbia 2024).

The operative rule is the Cryptocurrency Power Regulation. Its present text, last amended October 20, 2025, prohibits the authority from supplying service to defined new projects, including by accepting specified design deposits or entering specified study agreements (British Columbia 2025). The scope is precise:

Calling this simply a “province-wide ban” loses legally material boundaries. The regulation addresses supply by the authority to projects within its definitions. It does not, by its own words, adjudicate the climate merit of every facility or establish a circularity standard. It allocates electricity through connection eligibility, fixing which new projects the authority may serve.

That distinction also disciplines causal claims. The regulation can be expected to prevent the covered new service relationships. This paper does not claim that it reduced provincial mining electricity by a specified amount, displaced mining to another province, or changed emissions by a specified amount. Establishing any of the three would require connection and facility data from before and after the rule.

The Alberta materials reviewed here do not support describing the province as a regulatory void. The Alberta Utilities Commission states that, effective April 25, 2022, all new power plants, including power plants used for cryptocurrency mining, require AUC approval, with requirements varying by generating capacity (Alberta Utilities Commission 2026). This is particularly important for mines powered by on-site generation rather than an ordinary grid connection.

For transmission-connected facilities, the Alberta Electric System Operator issued its Guide to AESO Connection Requirements for Transmission-Connected Data Centres on June 12, 2026. The guide’s definition expressly includes digital-asset mining. AESO will review transmission-connected data centres regardless of size to determine applicable requirements and mitigations. Where the guide conflicts with ISO rules, Alberta Reliability Standards, or project-specific functional specifications, those instruments remain the authoritative controls (Alberta Electric System Operator 2026). The guide addresses matters such as modelling, ramping, disturbance response, monitoring, and real-time coordination. It is technical connection governance, not a crypto-specific tariff or a statutory guarantee of access.

Alberta’s generation context is relevant but must be dated correctly. The Canada Energy Regulator reports that the province’s 2022 electricity-generation intensity was 470 g CO₂e/kWh and notes that coal-fired generation was phased out by June 2024 (Canada Energy Regulator 2026). The 470 g value predates that change. It is therefore a historical annual average, not a 2026 factor and not a marginal emissions estimate for a proposed mine. The earlier paper’s conversion of a constant 1 MW load into a single annual emissions number is withdrawn.

The supported comparison is institutional. Alberta uses general generation approvals and data-centre connection requirements where Québec uses a mining-specific tariff and British Columbia uses a defined connection prohibition. Whether Alberta’s framework is more or less socially valuable cannot be inferred from the absence of a matching instrument, and answering it requires project outcomes.

Manitoba’s temporary directions are history, not the current rule. The Requests for Service Regulation, M.R. 68/2025, has been in force since September 1, 2025. Section 4 provides that Manitoba Hydro must deny a request for service if the proposed use of power is directly or indirectly related to mining a cryptocurrency (Manitoba 2025). This is not a pause scheduled to expire on April 30, 2026. It is an operative ineligibility rule unless the regulation is amended or displaced.

Two later Acts require separate treatment. The Manitoba Hydro Amendment Act, S.M. 2026, c. 17, was assented to on June 1, 2026. If proclaimed, it would permit Manitoba Hydro, subject to Lieutenant Governor in Council approval, to establish a cryptocurrency curtailable-power program, including criteria for eligibility, frequency, duration, magnitude, notice, and application to existing customers (Manitoba 2026b). The Manitoba Hydro Amendment and Tax Administration and Miscellaneous Taxes Amendment Act, S.M. 2026, c. 32, also assented to on June 1, would create separate customer classes and an intensive-use levy framework (Manitoba 2026c).

Neither Act was in force at the source freeze. The current consolidated Manitoba Hydro Act lists S.M. 2026, c. 17 and Part 1 of c. 32 as “not yet proclaimed” (Manitoba 2026a). Royal assent makes each an enacted statute; it does not activate provisions expressly dependent on proclamation. The current analysis must therefore rest on M.R. 68/2025 while treating the two Acts as possible future architecture.

Manitoba is analytically useful precisely because it shows why labels fail. “Moratorium,” “ban,” “curtailable program,” and “levy” describe different instruments at different legal stages. Combining them into a single posture obscures the decision actually facing a customer today.

The applicant should fund the studies and assets attributable to its connection, including credible upstream reinforcement and an exit-security mechanism proportionate to stranded-cost risk. The regulator or utility should publish the allocation method, not confidential commercial terms. Preferential rates should require an explicit finding about who bears residual costs.

Any claimed flexibility should be contractual. The agreement should define the baseline, nominated capacity, notice, event duration and frequency, telemetry, measurement interval, rebound limits, settlement, audit rights, and non-performance consequences. Annual disclosure should report both availability and delivered response. A facility should receive no grid-service credit for uneventful technical potential.

Project assessment should state the electricity counterfactual. It should use the finest supported temporal data, and report contractual renewable procurement separately from physical system effects, with additionality, delivery, and matching rules disclosed. An annual provincial average may remain a sensitivity case, but it should not be the sole estimate.

Heat credit should accrue only at the delivery boundary. The ledger should record gross heat captured, auxiliary electricity, distribution losses, useful heat accepted by the recipient, temperature, downtime, and displaced fuel or electricity. Results should be reported monthly so that summer rejection is not hidden inside an annual percentage. Capital support should depend on an independent feasibility study and a durable offtake arrangement.

Operators should disclose equipment turnover, reuse, resale, recycling, and final disposition. They should also maintain financial security for site closure and identify ownership of heat and interconnection assets if operations cease. Those obligations determine whether a short-lived computational load leaves long-lived public liabilities.

The public record should distinguish capacity from energy, capability from performance, captured heat from delivered heat, and gross from net avoided emissions. Verification should be independent of the operator and use a stable methodology. Commercial confidentiality can protect customer-specific information without eliminating aggregated event and performance reporting.

The paper deliberately sets no universal 1 MW or 5 MW Canadian trigger, no fixed renewable percentage, and no mandatory national heat-recovery share. Those numbers require a jurisdiction-specific record about administrative burden, system risk, network topology, and feasible heat users. Precision without such a record is presentation, not analysis.