RGGI Carbon Leakage Monitor

Status: Offline for methodology audit and fixes.

An external audit by independent AI reviewers (OpenAI GPT-4o, Anthropic Claude Sonnet 4.5, Google Gemini 2.5) flagged several methodological flaws in the v0.1 model published earlier today, including:

• Treating ISO-NE states (CT, MA, ME, NH, RI, VT) as PJM compliance states
• Greedy merit-order without PJM transmission constraints / LMP differentiation
• Heat-rate derivation biased toward high-load operating samples
• Calibration asymmetry between with-RGGI and without-RGGI scenarios

The dashboard is being rebuilt to address these issues. We will republish only when the corrected analysis can stand independent review.

Built by Alpha Inception.