Energy advisory · AI & electricity markets
AI is creating extraordinary economic value and unprecedented demand for electricity. As access to power becomes a critical constraint on AI infrastructure, the consequences are reshaping electricity markets, infrastructure investment, and the economics of data-center development.
Royal Powers Energy, Inc. quantifies those consequences. We created the Compute Heat Rate to measure the electricity price tolerance of AI compute. We apply that framework alongside power-market, infrastructure, and financial analysis to help clients understand where AI-driven electricity demand creates risk, where it creates value, and what to do about it.
The Compute Heat Rate
The Compute Heat Rate, created by Hans Royal, provides a framework for quantifying AI workload electricity price tolerance by relating the economic value of compute to the electricity required to produce it. It answers a question that didn't exist a decade ago: What is electricity worth when it is converted into AI compute?
The answer reveals an extraordinary gap between the economic value created by AI compute and conventional electricity costs. That gap has profound implications for power prices, infrastructure investment, grid flexibility, large-load tariffs, and the allocation of value across the energy and data-center ecosystems.
First published in 2026, the Compute Heat Rate has been cited by PJM Interconnection and the Brattle Group. Full methodology is at SSRN. Current index values are published quarterly at computeheatrate.com.
Visit computeheatrate.comCore question
“What is electricity worth when it is converted into AI compute?”
The Compute Heat Rate makes that value visible and quantifiable, creating a common analytical language for power pricing, infrastructure investment, market exposure, and policy design.
What we do
Our work spans load analysis, market intelligence, infrastructure economics, and strategic advisory. All of it is grounded in primary research and two decades of transactions.
Risk & Exposure
AI data centers represent a new class of electricity demand with unusual scale, growth, and economic value. We analyze where that demand can materially affect electricity, gas, and capacity markets, identify who is exposed, quantify the economic consequences, and help clients act. The CHREX platform quantifies data-center penetration, generation buildout, and price impacts at a local scale per electricity settlement hub.
Flexibility & Speed to Power
When access to electricity is the constraint, time and flexibility acquire economic value. We quantify the value of faster access to power, flexible load, curtailment, interconnection structures, and other mechanisms that can unlock capacity. We help clients understand how that value can be created, allocated, and captured through infrastructure investment and commercial agreements.
Market & Policy Intelligence
Utilities, regulators, grid operators, and policymakers are rewriting the rules governing large loads in real time. Royal Powers Energy provides an economic framework for evaluating large-load tariffs, cost allocation, flexibility requirements, and other policies shaping the next generation of electricity demand.
Who we work with
Our work is designed around decisions involving risk, capital, and contracts. Understanding the changing economics of electricity creates a material advantage.
Compute Heat Rate cited by PJM Interconnection (May 2026) and the Brattle Group
Insights and research
Hans Royal writes on grid economics, load growth, and market structure, with regular CHR updates as inference pricing shifts. Methodology is at SSRN. Current numbers are at computeheatrate.com. What it all means runs on Substack.
Readers include utilities, grid operators, and investors working through the same questions RP Energy advises on.
Read the SubstackGrid Load Growth
Interconnection queue data, ISO load forecasts, and what the numbers say about AI-driven demand growth relative to new generation coming online.
Power Market Mechanics
Forward prices, capacity market dynamics, and how a new, high-tolerance demand class collides with market structures built around industrial loads at a fraction of the price.
CHR Numbers, Updated
Published CHR calculations for current frontier AI models, updated as inference pricing changes. Methodology traces to Royal (2026), published at SSRN.
The firm
Co-Founder, President
Hans Royal spent nearly two decades building businesses across corporate renewables, supply-chain decarbonization, and tax-credit investing. Hundreds of transactions. More than 25,000 MW of wind and solar development worldwide.
He created the Compute Heat Rate, a framework for measuring AI-driven electricity price tolerance, cited by PJM Interconnection in May 2026. It is one of several research contributions behind RP Energy's advisory work.
He writes on AI infrastructure, electricity markets, and energy strategy. His work has appeared in Reuters, the Associated Press, Latitude Media, RTO Insider, Utility Dive, and Data Center Dynamics.
Co-Founder, CEO
John Powers has 23 years in energy markets. He joined Renewable Choice Energy in Boulder, Colorado in 2003 and helped lead the firm through its acquisition by Schneider Electric in 2017, serving as Vice President of Global Renewable Energy and Carbon Advisory. In that role he led a $50–100 million business with 200+ employees supporting deals across six continents.
The team he led advised on roughly 10 percent of all corporate clean energy transactions ever signed: about 25 gigawatts in total, plus more than $2.5 billion in tax credit transfers across supply chain decarbonization programs in North and South America, Europe, India, Japan, Singapore, Vietnam, Australia, and Mexico.
He writes on electricity markets, compute-driven load growth, and greenhouse gas accounting. He holds a degree in mechanical engineering from Duke University and lives in Colorado.
Contact
Whether you are modeling AI load growth, evaluating energy assets, or thinking through long-run power costs, the analysis begins with the numbers. We do that work and advise on what follows.
Boulder, Colorado
