Utilities, ISOs, and grid operators

Framing

Investor-owned utilities, ISO and RTO planners, and IPPs are facing the first sustained load-growth surge in two decades, and the demand signal is concentrated, lumpy, and politically charged. PJM's 15-year forecast revisions, ERCOT's capacity-rationing debates, and the Northern Virginia data center backlog all trace back to the same underlying question: how much of the announced AI capex actually becomes interconnected, energized load, and on what timeline.

What this consultancy contributes is the public-data math behind that question. Interconnect queue parsing across the seven major ISOs. Behind-the-meter decomposition. PPA pricing curves. Demand response value at the cluster level. Cross-checks against EIA, FERC, NERC, and Lawrence Berkeley National Lab interconnect studies. The work is intentionally modular: a single load-growth memo for an internal forecasting team, a multi-region PPA market read, or a rolling retainer for resource adequacy planning.

The product is honest numbers, not advocacy. Utilities buy this work to defend internal forecasts against analyst critiques and to brief commissions and customers.

Decisions we inform

• Sizing data center contribution to system-wide load forecasts.
• Pricing PPA offers in tightening capacity markets across PJM, ERCOT, and CAISO.
• Estimating realistic energization timelines across announced data center pipelines.
• Quantifying behind-the-meter and SMR roadmap impact on net-load forecasts.
• Building demand response and flexible-load value cases for hyperscaler customers.
• Briefing public utility commissions and intervenors with defensible exposure numbers.

Named offerings

Sample questions

• Our 15-year load forecast assumes 7 gigawatts of net new data center load. What does honest queue-position math say is realistic, and on what timeline?
• We are pricing a 20-year PPA for a hyperscale customer in our service area. Where does the market actually clear given announced renewable pipelines through 2030?
• How much of the West Texas data center backlog is energized by 2028 under realistic transmission build assumptions, and where does that leave reserve margin?
• Our PUC is hearing a large-load tariff case next quarter. What does the cost-of-service math look like under the announced data center pipeline?
• What demand response value can we credibly offer a colocation customer if their workload mix is mostly inference?