Modelling the impact of the closure of the last nukes in Germany

Hey gang.

This is topical but it seemed useful to ask, now I know a bit more about tools like PyPSA and so on.

What is the expected impact on hourly or annual carbon intensity of electricity in Germany, of shutting down the last three nuclear plants over the next 24 months vs keeping them open, in the context of an accellerated buildout of renewables?

I’m asking this as I see tweets like this , arguing that the shortfall in now power likely to be made up by the expanded deployment of renewables by Prof Stefan Rahmsdorf:

The actual electricity generation (not just capacity) by the planned renewables addition next year alone in Germany corresponds to the output of 3 nuclear plants. In 2025, more than 4. By 2030 it will correspond to a total of 30 nuclear plants.

Link to Stefan Rahmsdorf tweet, with accompanying charts.

But this doesn’t tell the full story, as we know the generation profiles are totally different.

I can see this thread in this forum which gives some pointers, but these all seem to be before the Russian invasion of Ukraine, and I don’t think they reflect the accellerated policy response we have see over the last 12 months or so:

A couple of bits of relevant information first:

  • the now shuttered German nuclear reactors ran on fuel rods produced in Russia

  • nuclear reactors and fuel rods are tightly bound — indeed, finding another source of nuclear fuel (in light of Ukraine), conducting the proofing tests, and gaining official certification would have been a long, arduous, and ultimately uncertain process

  • the 4 months prolongation of the nuclear phase‑out past year‑end 2022 had a very limited effect on the system at large because no new fuel stocks were sourced (as one network operator commented, the effect was in the same league as recent improvements to their transmission line line loading algorithms)

  • the German plant operators, to my understanding, are more than happy with the nuclear shutdown and are looking forward to integrating and operationalizing a different tranche of technologies going forward

  • Germany has still not resolved the issue of final storage for high and low‑level nuclear waste — despite huge efforts to find solutions

The real decision about a nuclear prolongation should have been made some years back — in which case, the lignite phase‑out could have been brought forward. That said, such a move would have reduced pressure on the much needed expansion of renewables. Moreover, such dynamics are very difficult to anticipate and capture numerically. Fast forward to today and the question of a further extension of operations makes no sense at this juncture (despite media and social media interest in that proposition).

Regarding modeling any prolongation from extant policy, there would need to have been relatively sensible scenarios placed on the table, each with their own backstories — and these would have been quite hard to formulate and defend in the past, I suspect. Plus the difficulties in identifying and characterizing the likely effects and counter‑effects at play, as mentioned in the previous paragraph.


The various energy systems frameworks (GenX, PyPSA, TIMES, and so on) are now being applied to new‑build nuclear in a number of countries of course. But that is a very different question. And GenX has also been used to model nuclear fusion and determine the kind of operational and cost requirements that that technology would need to exhibit in order to be useful:

  • Schwartz, Jacob A, Wilson Ricks, Egemen Kolemen, and Jesse D Jenkins (16 March 2023). “The value of fusion energy to a decarbonized United States electric grid”. Joule. ISSN 2542-4351. doi:10.1016/j.joule.2023.02.006. In press. :closed_access: