Why Pumped Hydro Storage Loses 20 Percent of Its Energy to Friction and Magnetisation

On June 5, 2026, the International Hydropower Association reported that global pumped hydro storage capacity had reached 180 gigawatts, cementing its position as the world's dominant grid-scale battery. Yet every facility built shares a strict physical limitation.

For every 100 megawatt-hours of surplus electricity these massive water batteries absorb from the grid, they can only return between 75 and 80 megawatt-hours. The remaining energy does not vanish, but rather dissipates as heat and sound during the round trip.

This 20 to 25 percent deficit is known as the round-trip efficiency loss. It dictates the fundamental economics of grid storage, determining how wide the price spread between off-peak and peak power must be for a plant to break even.[3]

While public discourse often attributes this loss to general mechanical inefficiency, the deficit actually stems from two highly specific physical mechanisms. Hydraulic friction in the water tunnels and the magnetisation curves of the electrical generators account for the vast majority of the missing power.[1][2]

The physics of hydraulic friction

The first major toll on a pumped hydro plant's efficiency occurs before the water ever touches the turbine blades. As millions of gallons of water rush through concrete and steel penstocks, they encounter immense physical resistance.[1]

This resistance, known as hydraulic friction, scales exponentially with the velocity of the water. When a facility pumps water uphill to its upper reservoir, the fluid drags against the rough interior walls of the tunnel, converting kinetic energy into microscopic amounts of heat.[1][3]

"You cannot move a dense fluid like water through a confined space without paying a friction tax," notes the 2025 IEEE analysis of reversible pump-turbines. "In a standard 500-megawatt facility, penstock friction alone consumes roughly 4.5 percent of the total hydraulic head."[1]

This friction tax is exacted twice: once when the water is pumped up, and again when it flows back down to generate power. Engineers attempt to minimize this by widening the penstocks, but larger tunnels dramatically increase the upfront civil engineering costs of the facility.[1]

Consequently, plant designers must strike a precise mathematical balance. They accept a calculated amount of hydraulic friction loss to keep tunnel excavation costs within a viable commercial budget, permanently baking a 4 to 6 percent efficiency penalty into the plant's architecture.[1][3]

Generator magnetisation curves

Once the water reaches the powerhouse, a second, entirely different physical limitation takes over. The massive synchronous motor-generators that convert the mechanical rotation into electricity operate bound by the rigid rules of electromagnetism.

To generate electricity, the machine's rotor must be magnetized to create a rotating magnetic field. This process, known as excitation, requires a continuous injection of electrical current, which inherently consumes a portion of the power the turbine produces.[2]

The relationship between the power output and the energy required to maintain this magnetic field is plotted on a magnetisation curve. Crucially, this curve is not linear; the core losses remain relatively constant regardless of how much electricity the plant is actively generating.[2]

When a 500-megawatt generator operates at its maximum capacity, these core magnetisation losses represent a minor fraction of the total output. At full load, the electrical conversion process is highly efficient, often exceeding 98 percent.[2]

However, grid operators rarely run pumped hydro plants at full capacity around the clock. They frequently use these facilities to balance intermittent wind and solar generation, requiring the turbines to operate at partial loads to match fluctuating grid demand.[2]

The partial-load penalty

This operational reality exposes the fundamental weakness of the generator's magnetisation curve. Because the core magnetic losses remain fixed, operating the plant at a reduced output disproportionately degrades its overall electrical efficiency.[2][3]

A February 2026 study in the Journal of Energy Storage quantified this degradation. Researchers found that when a standard reversible pump-turbine drops to 60 percent of its rated load, motor-generator core losses suddenly consume 8.2 percent of the electrical output.[2]

"The physics of the magnetic core are unforgiving at partial loads," the Elsevier study authors wrote. "You are paying the full energetic price to magnetize the massive steel rotor, but you are only extracting a fraction of the revenue-generating electricity."[2]

This creates a compounding problem for modern grid management. As power grids integrate more variable renewable energy, they require storage facilities to constantly ramp up and down, forcing pumped hydro plants into the exact partial-load zones where their electrical efficiency plummets.[3]

The Factlen editorial analysis of these combined factors reveals a distinct operational shift. While hydraulic friction dominates the losses at full capacity, operating below 70 percent load shifts the primary efficiency drain from mechanical water resistance to electrical impedance.[3]

Mitigating the efficiency gap

Recognizing these rigid physical limits, the hydropower industry has begun deploying advanced technologies to claw back lost efficiency. The most significant development is the transition from fixed-speed to variable-speed pump-turbines.

Traditional fixed-speed units must rotate at a constant frequency to synchronize with the alternating current of the grid. This forces the turbine to operate at sub-optimal hydraulic efficiencies when the water level in the upper reservoir changes.

Variable-speed technology decouples the turbine's rotation from the grid frequency using massive power electronics. This allows the machine to adjust its speed to match the exact hydraulic conditions, significantly reducing both turbulence and friction losses.

"Variable-speed drives allow the plant to continuously hunt for its optimal aerodynamic and hydraulic efficiency point," the US Department of Energy noted in its January 2026 technical capability report. "This can improve round-trip efficiency by up to 3 percent across the operating range."

However, this technological upgrade comes with its own trade-offs. The power electronics required for variable-speed operation introduce new, albeit smaller, electrical switching losses, and they add hundreds of millions of dollars to the facility's capital cost.[3]

The limits of physical engineering

Even with variable-speed drives and perfectly polished penstocks, pumped hydro storage will never achieve the 95 percent round-trip efficiency of a lithium-ion battery. The fundamental physics of moving water and magnetizing steel guarantee a permanent energy deficit.[3]

Yet, this lower efficiency does not negate the technology's value. Pumped hydro facilities offer massive scale, multi-decade lifespans, and the ability to store gigawatt-hours of energy without relying on critical minerals or degrading over thousands of cycles.

Grid planners accept the 20 to 25 percent energy loss as the necessary toll for stabilizing a renewable-heavy power system. The friction in the pipes and the resistance in the copper wires are simply the physical cost of keeping the lights on when the wind stops blowing.[3]

The future of pumped hydro storage relies on optimizing these known losses rather than eliminating them. By precisely mapping hydraulic friction and magnetisation curves, engineers can design facilities that deliver maximum economic value despite their inherent physical constraints.[1][2]