⛰️ Pumped Hydro — pump vs generate economics

Each asset has two sides: generator units and pump units, telemetered under separate IDs. We show how much each side moved per month, the average price each side saw, the realized spread between them, and the wholesale net revenue (generation earnings minus pumping cost).

The generation ÷ pumping ratio is NOT an efficiency for every asset. Tumut 3 sits on a river and Shoalhaven has catchment inflows, so they generate far more than they pump — the ratio mixes storage cycling with natural inflow. Only a pure pumped-storage plant (Wivenhoe, Kidston) reads as a round-trip proxy (~0.7–0.8), and reservoir levels still blur short windows.

Wholesale energy only: regional spot price × as-dispatched output. No FCAS, contracts, or water-value accounting — the same lens as the rest of the site. Data: AEMO NEMWeb dispatch SCADA and prices.

Price matching: average prices are volume-weighted over the intervals that have a matching 5-minute price record; unpriced energy is excluded from both sides of the division, never booked at $0. Net revenue is shown only when ≥98% of the month's energy is price-matched and the pump side is telemetered — an unknown pumping cost is not a zero cost. SCADA MW is an instantaneous 5-minute snapshot converted to energy at MW÷12.