From Trust, yet verify

Recently, I came across the blog post titled The Dunkelflaute Disaster: What Happens When Wind Power Goes Silent. That post discusses the phenomenon known as Dunkelflaute, a German term that means “dark doldrums” (periods when there is little to no wind and sunlight, leading to a significant drop in energy production from wind and solar). The post highlights the challenges faced during such events and the importance of having (fossil fuel) backup energy. I generally don’t look much at the comment section, but in this case I did and by doing so found an interesting question.

It started with the comment that “it is impossibly expensive to run two energy systems in parallel”, which then lead to this response:

Have you added up to see if that is true? We’ve been running multiple systems for a long time. Coal with gas, and particularly with gas peakers.
The thing is, wind and sun are free, so the fuel savings are very great, and can offset a lot of costs.

I surely agree that grids have been running multiple systems for a long time. Conventional systems indeed used to run base load combined with peaker power plants, but the commenter seem to suggest that intermittent systems with backup are somehow comparable with base load with peakers because they both run multiple systems. As far as I understand, base load/peaker systems and intermittent/backup systems have their own dynamic and are not one-to-one comparable.

So, how different are those two systems? Let’s compare both.

In an electricity grid, production should always be synchronized with demand. Traditionally, a system of base load and peaker plants were used to achieve this synchronization. In such a system, base load power plants run continuously over extended periods of time, ideally producing roughly the same as the minimum demand. This base load can be nuclear power, but also hydro, coal or gas. When there is more demand than what base load provides, then peaker plants kick in, filling in the gap between demand and base load. This is how it looks like simplified (just as illustration):

Base load plants will continuously produce electricity (red area) and peaker plant production (purple area) kicks in when demand is higher than base load.

This is very simplified. In reality the peak capacity will be higher than just the maximum demand, there will be extra capacity available for unexpected situations. There are also levels in peak capacity (fast reacting and slow reacting backup).

Trying to do the same for the intermittent system plus backup will give something like this (also here, simplified and without safety margins):

In this case, there is no guaranteed production. The intermittent production (the orange area) is between almost zero production (dunkelflaute) and high production (sunny and windy day). The rest needs to be filled in by backup capacity (the light blue area).

Now side by side to get the proportions right:

In the conventional system, base load and peakers work complementary. The total capacity is at least equal to the maximum demand (again, this is simplified, in reality there will be also safety margins etc.). The backup capacity is relatively low because peakers only need to react to changes of demand.

In the intermittent system, the backup system run more or less in parallel with the intermittent power sources. The electricity production by solar and wind is not guaranteed, therefore the backup capacity needs to be larger than in the conventional system. This because backup needs to react to changes of demand as well as to changes of intermittent solar/wind production. This means that more backup capacity is required in the intermittent system, meaning a larger capacity needs to be build, operated and maintained. Depending on the period, the backup will run at a low efficiency (extra fuel consumption). These costs will eventually result in a less economical viable system (then not only solar and wind will require financial support, but also backup systems will also).

I agree that we have run multiple systems for a long time. The sun and wind are free, but the installations to create electricity from sun and wind aren’t (and we need more of them, spread over a larger area). Integration of that intermittent energy into the grid is also not free. Reliables and renewables have different dynamics and different price components. This means that we cannot just take a conventional system, replace some of it by intermittent power sources and then declare “a lot of” cost savings. I don’t see those lowering costs in the real world, despite increasing share of renewables.