CHR Framework | CHR Index | Research | Commentary | Contact
CHR Framework

The Compute Heat Rate

A framework for measuring the electricity price thresholds associated with AI compute workloads.
Hans Royal
Originator, Compute Heat Rate Framework
Current CHR Index  |  Contact

What is the Compute Heat Rate?

The Compute Heat Rate translates the economics of an AI workload into an electricity price threshold expressed in dollars per megawatt-hour. It provides a demand-side counterpart to familiar supply-side measures such as the gas heat rate and cost of new entry.

Compute Heat Rate (CHR)
The long-run, full-cost electricity price threshold associated with workload economics, including non-electricity costs and a required return treatment. CHR is most relevant to build, investment, and siting decisions.
Dispatch Compute Heat Rate (CHR-D)
The dispatch-focused electricity price threshold at which an already-deployed workload becomes economically indifferent between continuing to operate and curtailing. CHR-D excludes sunk capital costs and required return and deducts only costs avoidable upon curtailment.

The distinction matters because a facility can face one threshold when deciding whether to build and a different threshold when deciding whether to continue operating equipment that is already deployed.

How CHR is calculated

CHRw = (Rw − Cnon-elec) / (1 + m)
Long-run Compute Heat Rate for workload type w.

Rw is the economic revenue or value attributable to a workload per MWh. Cnon-elec represents the applicable non-electricity costs, and m is the required return treatment. The quarterly Index documents the current inputs, workload treatment, weighting, and source vintages.

Illustrative calculation, not a current Index value

Assume a hypothetical workload produces $12,000 of revenue per MWh, carries $2,000 per MWh of non-electricity costs, and uses a 25% required return treatment. The mechanics are: ($12,000 − $2,000) / 1.25 = $8,000/MWh. These rounded inputs exist only to demonstrate the formula. See the CHR Index for current inputs and reference results.

CHR-D uses the same economic discipline but a dispatch-specific cost boundary. It excludes sunk capital and required return and subtracts only costs that can actually be avoided when the workload curtails.

Why these measures matter

Electricity markets have long measured the costs at which generators enter, dispatch, or retire. CHR and CHR-D provide comparable information about a fast-growing demand class. They help distinguish long-run investment economics from short-run operating behavior and make workload-specific price tolerance visible in the same units used by wholesale markets.

Applications

Power markets and planning

CHR and CHR-D can inform load forecasting, scarcity analysis, resource adequacy, market design, and the treatment of flexible or firm data center demand.

Data center strategy

The framework can support siting, power procurement, infrastructure planning, dispatch policy, and evaluation of grid-connected and behind-the-meter configurations.

Risk and investment analysis

Expressing workload economics in $/MWh makes it easier to compare compute demand with generation costs, forward electricity prices, industrial demand thresholds, and hedging decisions.

Latest CHR Index

Q3 2026

The current quarterly publication reports blended long-run CHR and dispatch-focused CHR-D, explains the methodology used for the edition, and preserves prior quarterly releases as historical records.

View the Q3 2026 CHR Index →

Formal Research

The formal foundation for the framework is Hans Royal’s paper, The Compute Heat Rate: Quantifying AI-Driven Electricity Price Tolerance and Its Implications for Wholesale Market Repricing (February 28, 2026).

Visit the Research library →   Read on SSRN

Selected Commentary

Ongoing analysis applies the CHR framework to market developments, technology changes, and energy policy. These pieces are commentary rather than quarterly Index releases.

View all commentary and subscribe →

Canonical Citation

Royal, Hans, The Compute Heat Rate: Quantifying AI-Driven Electricity Price Tolerance and Its Implications for Wholesale Market Repricing (February 28, 2026). Available at SSRN: https://doi.org/10.2139/ssrn.6322318

The SSRN record was revised June 4, 2026. The revision date is publication history, not a replacement for the canonical citation date above.

Connect with us

Questions about the Compute Heat Rate, the Index methodology, research applications, or commercial use? Send us a note.