Today, storage is not widely used. Despite moderate renewable energy integration, places like Denmark and Germany use very little storage. In countries like Uruguay, even higher levels of renewables are used with little storage.
To understand, we need to look at how the grid works.
A system operator balances generation vs demand. Operators look at day-ahead demand curves. The expected vs actual demand can be plotted. You can see plots of day-ahead and hour-ahead demand here:
The operator plans the day-ahead generation to match predicted demand and holds some in reserve for unplanned outages. The generation is a mix of sources. If the sources are variable renewables like wind or solar, the prediction is based on forecasts. For now, in the majority of the world, flexible sources exist to compensate for variable renewables. Why? Take a look at that graph. Reserves are higher than demand. The total available generation must be higher than the annual peak demand in summer. That means there are a large number of power plants idle most of the year. That’s why up to 40% renewables can be accommodated without any major changes.
We need to plan a major shift in energy infrastructure to renewables. What do we need to increase renewables integration? We need finance sources and approval streamlining. We need to reduce solar BOS costs.
We need better solar and wind forecasting to help match load and reduce reserves. Increased NREL funding can speed implementation of improved, taller wind turbines. Increases in transmission can reduce curtailment and allow connection from high-resource areas to load centers. Improved grid practices like “Energy Imbalance Markets” can allow neighboring ISOs to share power and resources more effectively and closer to real time, in 15-minute updates.
But what about those arguments about how solar and wind are not available for long times? Let’s suppose we assume a very crude model of demand vs generation. Assume there are 2 weeks with no solar or wind on average everywhere. About 4% of generation. Now assume those two weeks are powered by the mothballed and shutdown FF power plants. How much carbon emissions is that? That is an overly simplified and certainly incorrect scenario. But it illustrates the point. Existing FF power plants can be used as reserves at low capacity factor. We wouldn’t need 100% FF for those two weeks, either. We have dispatchable renewables like hydro, geothermal, and biomass.
Just as we do today, we can hold some wind and solar in reserve, as overcapacity. Added dispatchable renewables like geothermal, hydro, and biomass combine to meet demand. When you look at it that way, storage is useful, but not critical.
The vast majority of Tesla storage will be used for utilities to replace gas peakers.
It will be at least a decade before meaningful amounts of grid storage will be needed to implement renewables. I think we need to shift the conversation from what energy will look like in 35 years, to what it takes to integrate renewables right now.
Demand management and transmission will be used to integrate renewables before storage becomes a bottleneck, for example, so advancing those markets and associated infrastructure is important now.