KL.Grid is K Labs’ proprietary platform for studying power markets. It reproduces a market’s dispatch logic from the bottom up — every plant, its declared variable cost and its technical limits — and clears the system marginal price hour by hour, 8,760 hours a year.

It runs three ways. A lookback replays a historical period from realized data to reproduce and validate the published price. A forecast optimises unit commitment and reservoir operation to produce the single most-likely forward price path. A stochastic mode runs that forecast hundreds of times under correlated climate, hydrology, fuel and demand shocks — plus forced-unit outages — to produce percentile price tubes (P5–P95).

v3.0 is now nodal: the country can be split into zones, each with its own load and weather, connected by transmission with real capacity limits — so KL.Grid produces a price per zone (LMP), plus batteries, load-dependent heat rates and hydro cascades. It stays single-node whenever you want it.

In hydro-dominated systems its signature output is the endogenous value of water — the opportunity cost of stored reservoir energy, computed by the model from market conditions and the prevailing ENSO regime, not assumed. Reservoirs then bid at that value, so hydro competes in merit order at its true forward-looking cost.

Every run opens in the interactive Grapher: a browser dashboard to explore price and demand, the generation mix, the hour-by-hour merit-order stack, reservoir trajectories, any single unit, and — for stochastic runs — the full price tube and outcome distributions, at hourly, daily or monthly resolution.

Read the launch note here: KL.Grid is live!