We See The Storm Before It Hits

A physics-based thermal digital twin that corrects ERCOT's day-ahead load forecast using spatial heat modeling, humidity dynamics, and urban thermal inertia — cell by cell across Houston.

20×20 Spatial GridDay-Ahead Forecast CorrectionERCOT Delta Prediction

THE PROBLEM

ERCOT's Forecast Error Is Structured, Not Random

Humidity Blindspot

ERCOT models temperature, not wet-bulb heat index or evaporative cooling load

Thermal Inertia

Cities absorb heat differently based on land use. Concrete, asphalt, and green space respond on different timescales

Urban Heat Islands

Dense commercial zones run 8–12°F hotter than surrounding suburbs. ERCOT treats Houston as one thermal mass

THE SOLUTION

A Spatial Physics Engine Built For The Grid

Houston is modeled as a 20×20 spatial grid— 400 independent thermal cells, each with its own heat balance, land-use classification, and response curve to incoming weather.

Every cell runs a heat balance equation accounting for solar gain, longwave radiation, convective transfer, and stored thermal energy. Land use — concrete, asphalt, vegetation, water — determines how fast each cell heats and cools.

Humidity correction layers wet-bulb temperature and evaporative cooling demand on top of dry-bulb readings, capturing the load ERCOT misses when humidity spikes without temperature moving.

Thermal inertia models how urban surfaces store and release heat over hours — so the grid sees load before the thermometer does, and after the sun sets.

Two Outputs. One Edge.

Corrected Load Forecast

A refined MW prediction that adjusts ERCOT's day-ahead forecast using physics-based thermal modeling

Forecast Error Predictor

Predicts when ERCOT will under or over-forecast and by how much, before it happens

Who Buys The Signal

Energy Trading Desks

Trade the ERCOT delta before the market reprices day-ahead load

Grid Operators & Utilities

Anticipate peak demand shifts driven by urban heat and humidity spikes

DOE & Federal Regulators

Validate grid resilience models with ground-truth thermal physics