Battery and pumped-hydro storage: power, energy and losses
Storage moves electricity between hours. To understand what it can do for a grid, look at both its power rating and how much energy it holds.
A 1 GW battery cannot necessarily supply 1 GW all night. Its stored energy, efficiency and connection limits determine how much reaches the grid.
GW is power; GWh is energy
A 1 GW power rating describes a rate of charging or discharging. A 4 GWh energy capacity describes the stock the device can hold. Delivering 1 GW for one hour supplies 1 GWh.
Dividing 4 GWh by 1 GW gives a nominal four-hour duration. In PowerLab, however, the displayed stored energy is internal energy and discharge has losses. A full battery therefore delivers slightly less than four hours at 1 GW. This convention is explicit so energy does not appear from nowhere.
PowerLab batteries have four hours of energy capacity per GW and 90% round-trip efficiency. Pumped hydro has eight hours per GW and 80% round-trip efficiency. These are game parameters, not universal specifications or product recommendations.
A complete charge–discharge calculation
The game applies the square root of round-trip efficiency at each stage. For a battery, each stage is about 94.87% efficient.
× √0.90 = 0.9487 GWh stored
× √0.90 = 0.90 GWh returned
The remaining 0.10 GWh is lost across the cycle. The storage unit cannot charge and discharge in the same hour. A full unit cannot accept more energy, and an empty one cannot supply output.
Charging also consumes transmission capacity when the electricity comes from elsewhere. A large battery behind a congested line may be difficult to fill before the evening peak. Location matters alongside duration.
Hydro generation is different from pumped storage
The U.S. Department of Energy describes pumped storage as storing energy by moving water to a higher reservoir and generating when it flows back down. It consumes electricity to charge, much like a battery.
PowerLab separates this from its hydro plants. The game’s hydro plants are run-of-river generators whose available output varies with a synthetic water resource. Pumped hydro stores previously generated electricity and starts empty when built. Both use fixed game siting eligibility.
Building pumped hydro does not create a new primary energy source. If it charges from coal generation, those emissions are recorded during charging. Discharge does not erase them or count them a second time.
When should you charge or discharge?
Start with a concrete need: enough stored energy to cover a forecast evening gap, within the storage unit’s power limit. Check where the charging electricity will come from and whether the connecting lines can carry it.
PowerLab offers Auto, Charge, Discharge and Hold. Auto uses a simple daytime charging and evening-to-morning discharge schedule, with a low-cost charging threshold. It is not a perfect trading strategy. Manual charging can draw on more expensive thermal supply, which may increase costs and emissions.
A cheaper charging hour does not by itself establish that an investment is profitable. Construction cost, operating expenses, efficiency losses and how often storage is used all matter. The accelerated game is not a project-finance assessment.
Put the idea to work.
Choose the Carry energy into the night challenge. It starts with an empty 1 GW / 4 GWh battery. Charge it, inspect the stored-energy meter, and deliver at least 2 GWh after sunset while maintaining service.
Try storage dispatch →Common questions
Does a battery generate new energy?
No. It returns some of the energy used to charge it; round-trip losses mean less energy comes out than goes in.
Is pumped hydro the same as a hydro power plant?
No. In PowerLab, run-of-river hydro produces electricity from water flow, while pumped hydro stores electricity by pumping water uphill.
Sources & model boundary
Examples and calculations: PowerLab’s illustrative model, checked against the local simulation. Background reading: U.S. Department of Energy: pumped storage hydropower.
This guide is educational. Grid Operator’s network, economics and construction are simplified game assumptions; Market Lab’s uniform-price model is a separate one-hour experiment.