This is often put in contrast with the presumed reliability and stability of good old baseload power plants like coal or nuclear. These tend to produce largely the same amount of electricity day in day out - as long as they don't have to be shut down for maintainance or due to adverse weather conditions.
Given that electricity demand is variable, how useful is constant production when the goal is for production to match load at all times? And how much worse would "Flatterstrom" be in comparison?
Using again the thought experiment from this previous post, we consider the current hourly load over a year for Europe's two largest electricity markets, France and Germany. Then we assume two generation profiles which in total matches exactly the total yearly: a fixed, constant production or an optimized hybrid mix of wind and solar production. This means that for each hour of the year, there is likely an imbalance between the current actual load and these hypothetical power sources - either a surplus or a shortfall.
The chart above shows how much of the produced energy does not properly match the load - which is about 20% for the wind/solar hybrid and about 7-8% for a fictitious constant power source. When it comes to seasonal imbalance, wind & solar might even be a better match to the seasonal demand variability in central Europe (more in winter than in summer) compared to a fixed rate generation.
Both of these base scenarios require some level of additional flexibility to balance out the imbalances and either way, it is likely that in a future electricity system, flexibility will become a key factor: the ability to either produce when there is a shortfall or to consume when there is a surplus.
For example, if in Germany we had enough flexibility to balance out supply and demand within each day, the total amount of unmatched load in a year would drop from about 20% to 15% for the wind and solar scenario or to about 7.5 %to 5% for the constant generation scenario.
Presumably it is much easier and cheaper to shift demand or store energy for a short time than for a long time. Hence the chart above also shows how the cumulative need for flexibility breaks down over multiple time-scales - from a few hours to a few days all the way to monthly or seasonal imbalances.
An alternative to flexible demand or storage would be overbuilding and curtailment. This means to build more capacity than is needed to cover the total demand for the year and then throw away some of it when it is not needed. While curtailment is easy to do for wind and solar, it might be harder for thermal power plants even though there is a lot of focus on running coal plants on a more flexible schedule.
Starting to consider cost, battery storage systems are already competitive in shifting cheap surplus energy by a few hours, which is why they are being deployed around the world in large numbers. As batteries continue to get cheaper, their applicable time range might increase slightly. However a doubling of storage time would require for prices to be cut in half.
While batteries are starting to successfully compete with gas power plants for intra-day flexibility, gas and coal remain by far the cheapest option for flexibility over longer timescales unless there would be a significantly higher cost for releasing CO2 into the atmosphere.
Based on this back of the envelope estimation, it should be reasonably feasible to carry 80-90% of the load with wind, solar and short-term battery storage alone. Even for the less than optimal climate conditions of central Europe.
Or in other words, on current momentum of deployment, we could end up with a power grid based to 90% on wind, water sun and batteries, leaving a hard to displace 10% of coal and gas based generation.