AI compute / Electricity markets / Power investment
Understanding what AI demand means for electricity markets, grid flexibility and the investment decisions ahead.
Explore the library
Featured research
Open access. No signup required.
Why flexible data centers are a critical first step toward protecting electricity customers, but may not be enough. John Powers examines ERCOT wholesale repricing and a practical approach to ratepayer protection.
Read researchFree download · 24 pages · No signup required
Foundational research
Created by Hans Royal, the Compute Heat Rate relates the economic value of AI compute to the electricity required to produce it. Explore the current published CHR Index and visit computeheatrate.com for the framework and ongoing research.
Go directly to the published Index for current reference values, assumptions and methodological context.
View the current CHR IndexFollow the continuing research on AI workload economics, electricity demand and the value of flexibility.
Explore computeheatrate.comIn collaboration
Trading flexibility for faster grid access can create value for data-center developers and the power system. The economics depend on which workloads curtail and how much capacity gains earlier access.
Read at Occam Edge Discuss this research / Request a briefingCombining CHR demand economics with CREO’s Marginal firm Capacity Cost supply framework offers a way to examine the implications of AI electricity demand for clean energy investment.
Read the paper at CREO (PDF) Discuss this research / Request a briefingSubscribe to research
Subscribe for free to Compute Heat Rate on Substack for new research and commentary. Reading and downloading research never requires a subscription.
Optional. Signup takes place on Substack, under its privacy policy.
For sponsors, utilities and developers looking at a specific market or project.
Request a briefing