0
0
Solar and wind now generate electricity more cheaply than coal in most of the world — the levelized cost of solar has dropped over 90 percent since 2010, wind by over 70 percent. The economic argument for renewables has essentially been won. And yet grids still depend heavily on fossil fuels during the hours when the sun is not shining and the wind is not blowing. The reason is storage. Grid-scale battery technology has moved from experimental to commercially deployed with remarkable speed, and the implications for how power systems operate are significant.
Why Storage Is the Central Problem in Renewable Energy
Electricity must be consumed almost exactly when it is produced. Traditional power plants — coal, gas, nuclear — can be turned up or down to match demand at any moment. Renewable sources cannot. Solar peaks midday regardless of demand; wind often blows hardest at night or in seasons that don’t align with consumption. The grid’s balancing act becomes far harder when the generation mix cannot be dispatched on command.
This mismatch creates what grid operators call the duck curve: solar peaks in the afternoon, then drops sharply just as evening demand surges. Gas peaker plants have traditionally filled that gap, but they are expensive to run, polluting, and uneconomical to maintain for the limited hours per year they are actually needed. Battery storage offers a direct alternative — capture excess solar generation during the day and release it in the evening when demand peaks.
How Grid-Scale Battery Systems Actually Work
The dominant technology today is lithium-ion — the same chemistry as electric vehicles, packaged in large modular units connected to the grid through power electronics that respond to system conditions in milliseconds. A typical installation consists of rows of battery cabinets, each containing hundreds of cells, managed by systems that monitor temperature, state of charge, and health across thousands of units simultaneously.
What makes these systems particularly valuable is speed. A gas turbine takes minutes to ramp from cold; a battery system reaches full output in under 200 milliseconds — fast enough to catch frequency drops caused by a generator tripping offline. This capability was previously only available from spinning generators with physical inertia. Battery systems replicate it electronically, and in several markets have already largely displaced gas peakers in the frequency regulation market.
| Technology | Response Time | Typical Duration | Best Use Case |
| Lithium-ion (Li-ion) | <200 ms | 2–4 hours | Peak shifting, frequency response |
| Flow batteries | Seconds | 4–12+ hours | Long-duration storage |
| Pumped hydro | Minutes | Days to weeks | Seasonal storage |
| Compressed air (CAES) | Minutes | Hours to days | Large-scale bulk storage |
Beyond lithium-ion, flow batteries and iron-air systems are advancing toward commercial scale for long-duration applications — ten or more hours of discharge — enabling grids to manage multi-day periods of low renewable generation rather than just evening peaks.
What the Numbers Look Like Now and Where They Are Heading
Global grid-scale storage capacity crossed 300 GWh in 2024, up from under 30 GWh in 2020. Costs have fallen from around $1,500 per kilowatt-hour in 2010 to below $150 in leading markets today, with further declines expected as manufacturing scales. The US, China, and the EU account for most deployments, driven by a mix of economics and policy.
Storage is now being built without subsidy in some markets, replacing gas peakers that would otherwise require expensive capacity payments. In California, battery storage regularly covers the evening demand ramp that once needed gas, and on favourable days the grid runs on near-100-percent renewables for sustained periods.
Key developments shaping the next phase of grid-scale storage:
- Sodium-ion batteries entering commercial production, offering lower cost and avoiding lithium supply constraints
- Co-location of storage with solar and wind farms becoming standard practice in new renewable project design
- Virtual power plants aggregating thousands of distributed batteries — including in electric vehicles — into grid-controllable resources
- Regulatory frameworks in most major markets now explicitly recognizing storage as a grid asset class, enabling new revenue streams
- Long-duration storage pilot projects moving toward commercial deployment, targeting ten to one hundred hours of capacity
The grid of 2035 will look fundamentally different from today’s. Cheap renewable generation combined with affordable storage resolves the central technical objection that has slowed decarbonization for decades. The engineering challenge is real, but it is no longer a reason to doubt that a grid powered predominantly by wind, solar, and storage is achievable. The pieces exist. The question is the pace of assembly.
