Pumped-Storage Station

2D twin of the 3D pumped-storage lab: two reservoirs, a turbine, a pump and a battery serving a settlement, driven by the identical physics file.

Presets

Playback

Plant, demand and weather

Turbine power is P = rho g Q H_net eta with H_net reduced by pipe loss K Q squared. Pumping lifts water back up at 88 % of the static head and costs rho g Q H / eta electrical power. The battery covers the difference between generation, pumping and the settlement demand.

Model checks

The science

What is modelled

An upper reservoir feeds a turbine through a penstock into a lower reservoir; a pump can lift water back. Catchment rainfall and a base stream add inflow. A small battery balances generation, pumping and the settlement demand, and any imbalance appears as grid import or export. Reservoir volumes are tracked exactly and overflow is spilled.

Equations

  • Turbine power: Pt = ρ g Qt Hn ηt, Hn = Hlevel - K (Qt² + 0.55 Qp²).
  • Pump electrical power: Pp = ρ g Qp Hp / ηp.
  • Volume balance: dVup/dt = Qin + Qp - Qt - Qrunoff.
  • Outlet reference: Q = Cd A √(2 g h).

Checks run in the page

  • Turbine and pump power against rho g Q H.
  • Water conservation over an hour.
  • Single-step reservoir balance.
  • Torricelli outlet flow and battery arithmetic.

Honest simplifications

Constant efficiencies, a quadratic pipe loss, no water-hammer, no turbine curve, a head that varies only linearly with the upper level and no electrical network model. It teaches the energy and water bookkeeping of pumped storage, not plant design.

What to try
  • Press Pump at night to refill the upper reservoir while demand is low.
  • Press Peak demand and see how long the stored water lasts.
  • Press Storm refill and watch rain fill both reservoirs until water is spilled.
Upper and lower reservoir fill
Generation, pumping and demand
Battery energy
Generation against turbine flow
Made with 🤖 ChatGPT