Germany Reality
How the German Energiewende (Energy Transition) is going, and how it might proceed
Notes: Substack is warning this post is too long for email. In any case some of the charts are in ‘gallery view’ and are consequently miniaturised. For best legibility, please click the graphic at the top of the email to transfer to the Substack post. Then you should be able to zoom any chart by clicking on it. I think/hope.
Summary
For this post I downloaded data detailing hourly Demand, Generation and Import/Export power flows in Germany over the last 3½ years. I assessed how close (or not) to being 'decarbonised' the Germany electricity grid is currently.
I outline the methodology I use to extrapolate each set of hourly data into the future. And I then assess what multiples of current [Wind + Solar] plus what quantities of long duration energy storage (LDES) might be required to keep Germany's lights on every minute of the day while operating a carbon-free power grid.
Three (3) to four (4) times current [Wind + Solar] fails to consistently keep the lights on. Future CO2-free power shortfalls occur many times throughout each future year. Very large quantities (up to 1,200 GWh; 1,200,000 MWh) of LDES are helpful, but only an immense quantity (~6,000 GWh; ~6,000,000 MWh) of LDES is sufficient with the 2024 data to eliminate power shortfalls within those five months.
If current 2024 [Wind + Solar] were to be increased by a factor of twelve (12) times, there would still be power shortfalls unless LDES with capacity ~630 GWh (630,000 MWh) was available to 'time-shift' power from intervals of surplus to intervals of shortfall.
I *really* think Germany needs to reconsider its ideological ban on Nuclear energy.
Introduction
I have analysed real data from power grids - examples: the UK; California; Texas - and have observed a pattern. Because Supply must closely balance Demand every second of the day, ever more ‘renewable’ power generation *capacity* of uncontrolled sources (Wind and Solar) leads to increasing curtailment absent a route to store or export the surplus.1 For this post I analyse the German power grid to see if the pattern repeats.
In the early phase of deployment of Wind and Solar onto a grid, the quantities of uncontrolled power supplied from them remains within the ‘control band’ of dispatchable (usually fossil-fuelled) generators on that grid.2 Large fuel savings result. Claims are made that this demonstrates the energy transition will be both cheap and easy. In this early phase, Wind and/or Solar generation rarely have to be curtailed, only when a *local* section of the grid cannot cope with the power: an example from the UK being constraints on some Scottish Wind farms because their connectivity to the wider GB grid was limited.
Then in later phases of Wind and Solar deployment, the available total Demand increasingly becomes another and more critical limit3. The UK provided a prime example of this over the early May bank holiday weekend. Demand was low, the sun was high, so we ended up paying countries at the other ends of our interconnectors to take our surplus power. Our system operator’s forecast is that UK constraint payments will cost £3,483 million over the next 24 months; an average of £145 million per month.
So, adding more and more ‘renewable’ generation *capacity* results in greater and greater curtailment, and/or greater and greater exports to neighbouring grids, sometimes at significant cost. Exporting power is fine, but if the neighbouring grids have also increased their ‘renewable’ generation *capacities* to the point they too reach saturation during the same helpful weather conditions, they have to be paid to take power if they can, but if they can’t take it they won’t. Hence increased levels of curtailment tends to reduce the benefits from extra ‘renewables’ investment.
Electrification (heating converted to heat pumps; mobility converted to EVs; etc.), is forecast to increase electrical power Demand everywhere. It is apparent from analysis of real weather-dependent power that long lulls in Wind occur multiple times in a typical year. When lulls coincide with long winter nights, even an impracticably-large multiple of [Wind + Solar] is not sufficient to keep the lights on. This currently means we need to keep much of the legacy fossil-fuelled power generation capacity available and ready to fill in the gaps. This is increasingly costly per MWh of energy they produce because their fixed costs - capital, operations staff, maintenance and inspection, etc. - remain, well, fixed. But that’s the cost currently of energy security within a grid: having to retain sufficient standby capacity of dispatchable power generation to keep the lights on 60/24/365¼ whatever the weather does.
We can accept this fossil cost of energy security and we can anticipate that energy bills will continue to climb because of the need to keep much of the legacy power generation system in place while over-laying ‘renewable’ power generation systems. Or we could avoid the ongoing need for fossil energy in various ways:
- accept that the lights will go out more often at night when the wind drops - not good for vulnerable people, hospitals, industries, or for your society generally, I would suggest;
- get creative with load-following nuclear technology (but not in Germany…?);
- capture surplus ‘renewable’ energy when it is windy and during sunny days, and then feed that captured energy back into the grid when it’s not windy / when it’s dark.
‘Charging up’ energy storage would allow curtailment to be reduced by providing reliable / schedulable and additional Demand on the grid while maintaining frequency control. The technologies we currently have for storing energy comprise:
battery energy storage systems (BESSs) - superb for near-instantaneous response both charging and discharging and with system round-trip efficiencies of the order 80%-90%; limited to around 3,200 MWh gross capacity / ± ~900 MW power rating currently, although 8-hour BESSs are in development (note, all being built next to existing or proposed solar farms: maybe they’ve read another of my Substack pieces).
pumped storage (PS) schemes - can have huge storage capacity, for example Snowy 2 in New South Wales will eventually have gross 350 GWh [350,000 MWh] / 175 hours of energy storage and ± ~2,200 MW power rating, but is taking ~10 years to complete and its cost keeps rising ‘to an estimated $12 billion’ AUS [~ 8 billion US Dollars].
Germany already has a number of PS schemes, which is why we can assume current long duration energy storage (LDES) storage capacity in Germany is significant and likely to be growing relatively rapidly as BESSs large and small are built and added to the system.Yet-to-be-developed LDES schemes and technologies4. Some possibilities are listed in the About Storage Innovations 2030 preamble to the US DOE’s Technology Strategy Assessment Findings from Storage Innovations 2030 on Compressed Air Energy Storage July 2023 [pdf]; omitted from the DOE list is liquid air energy storage (LAES); and quite possibly others.
But what multiples of Wind and Solar *capacity* are needed, and how *much* energy storage would make a difference? My analysis gives indications of the scale of the technical requirements for Germany. Note I do NOT attempt to assess the economics of any of the future scenarios.
Germany Data
Germany power grid data going back to 01 Jan 2012 is available from the Tools area of the Agora Energiwende website under Agorameter. My focus is recent data for actual generation from Wind and Solar. I then extrapolate using simple Wind and Solar multipliers to estimate how future power might be generated.
I needed to download two sets of data for each date range:
- Power generation and consumption:
Figure 1: Germany Hourly Demand and Generation (GWh/h):
And to account for cross-border power flows I needed:
- Power price and power export/import.
Figure 2: Separate Germany Net Export / Import (GWh/h):
The Agorameter site allows (after quite a bit of trial and error) download of up to a year at a time of hourly data. I downloaded four sets of data: all 2021; all 2022; all 2023; and for 2024 to 31 May. For context:
- Covid lockdowns may have affected Demand in 2021 - but no significant effect is visible in the data.
- The invasion of Ukraine by Russian forces began on 24 Feb 2022 and
- the Nord Stream gas pipeline explosions of 26 September 2022 combined to irrevocably change the energy landscape for Germany.
- Then nuclear power production in Germany was ended on 15 April 2023.
I believe we need to be transparent when evaluating the ‘energy transition’ for a territory. The magnitudes of the power imports / exports are significant. For Germany to become energy-independent (in case other territories decide in future to keep their power to benefit their own populations) I therefore believe we need to determine German in-country Demand [i.e. net of Import / Export] for each hourly interval. It is then the in-country Demand I increase via the Demand multiplier.
I found an "… estimate of Germany’s power consumption in 2030 …
So 700,000 GWh is my target for each future year extrapolated from actual 2021, 2022 and 2023 data. Incidentally, I hope this level of consumption [Demand] is sufficient to reverse Germany’s de-industrialisation.
To ‘decarbonise’ future Germany, power currently and formerly generated by fossils + Nuclear needs to be replaced by lower-carbon-emitting generation technologies [Wind + Solar PV]. We can numerically estimate how high this energy ‘mountain’ is by summing the various columns of hourly GWh/y in the Agorameter files as summarised in Table 1:
Table 1: Energy Quantities 2021, 2022, 2023, and 2024 to 31May - Existing and Future:
Note that in 2021, Germany generated close to 10% of its predicted future Demand from Nuclear. Germany has since banned Nuclear for what appear to be ideological reasons.5 Demand as recorded by Agora has fallen each year since 2021, but after accounting for imports/exports, Net in-country Demand has stayed fairly constant.
When we plot recorded power flows for the ~3½ years - see Figure 3 - we see there are wild spikes in ‘renewable’ generation (green = Wind Total; yellow = Solar) across Germany. The implication is that instantaneous future in-country Demand (black) will not be met every time instantaneous future [Wind + Solar] falls into its many dips.
Figure 3: Existing DE 2021, 2022, 2023, and 2024 to date:
I derived the ‘Future’ part of Table 1 as follows:
In each of the three whole years, I used a nominal overbuild factor of 20% (because I’ve been assured that 20% is ‘optimal’6):
- nominal [Extra Solar + Wind] Factor7 = [Conventional to be displaced] / [Total Solar + Wind (existing)] * [Future in-country Demand Multiplier] * [1 + Overbuild factor X%], then
- allow for [Expected curtailment] = 1-(1/(1-[overbuild factor, 20%])) = 25%
- use Goal Seek to adjust the [Overbuild factor X%] so as to arrive at [Total in-country Demand forecast] = 700,000 [GWh]
- part-year 2024 factors are set equal to whole-year 2023 factors
- future Biomass and Hydro are set the same as recorded Biomass and Hydro
- I assume for this analysis that any grid constraints (local or national) would be eliminated to facilitate increased Demand and greater *capacities* of [Solar and Wind]. This is a BIG assumption given the multiplication factors of 3 to 4… or more.
The factors in the ‘Future’ part of Table 1 work for the overall electrical energy balances in each of the three years. But we need to plot - Figure 4 - the hourly results of [Generation - Demand] to see if [instantaneous future power Supply] satisfies [instantaneous future Demand]. Oh look! it very often doesn’t.
Figure 4: Future DE based on real 2021, 2022, 2023, & 2024 data, with [Solar + Wind] factors as noted:
The results are clear: an awful lot of red instantaneous Shortfalls are evident throughout the ~3½ years. This is not a solution to keeping Germany’s lights on while the country goes fully ‘green’. Meanwhile the soaring peaks of the [potential] Surplus generation which has to be curtailed sometimes reach 100 GWh/h and higher. If future Germany is to make use of any of that Surplus the grid would have to be massively reinforced to be able to cope.
Germany currently has significant pumped storage capacity plus both commercial and private battery storage. As an illustration of the Future based on 2024 data - see Figure 5 - I assumed total current LDES capacity of 100 GWh = 100,000 MWh, with overall round-trip efficiency (RTE) = 75%.
Figure 5: Future DE based on real 2024 data, [Solar + Wind] factor 3.034, and with 100 GWh of LDES:
Note my methodology as plotted only credits LDES with power to ‘charge up’ when there is surplus ‘green’ power, per Table 2 logic:
Table 2: ‘Green’ LDES Logic:
1 IF there is surplus [future Wind + future Solar]
2 AND there is unused energy storage capacity in future LDES
3 THEN that future LDES can be scheduled to receive and store ‘green’ energy (thereby reducing the amount of energy that has to be constrained / curtailed)
4 UNTIL (at maximum) the future LDES becomes ‘full’
SO
5 WHEN there is a shortfall of [future Wind + future Solar]
6 AND there is ‘charged-up’ future ‘green’ LDES
7 THEN we could get back up to 75% [1] of the energy stored in the LDES (thereby
reducing the shortfall of energy) by scheduling the LDES as additional generation
8 UNTIL (at minimum) the LDES becomes ‘empty’ [2]
[1] based on average round-trip efficiency (RTE) of pumped hydro storage; battery storage RTE is likely better, but the RTE of other technologies may be worse
[2] The operating range of LDES needs to remain within constraints on e.g. depth of discharge (DOD) which will probably depend on its technology. I’ll work with net storage capacity and assume we remain within such constraints, whatever the technology mix.
I assume no limits on charge / discharge cycles or other aspects of LDES operation: if there is capacity to be charged, or stored ‘green’ energy to be discharged, then my analysis uses it.
Of course no power grid can separate ‘green’ electrons from lignite black ones. Table 2 logic defines the energy accounting / grid operator capacity dispatching controls necessary to ensure ‘green’ power remains clear ‘green’. The economic incentives to blur the lines will be immense.
Zooming in on the numbers in Figure 5:
Future [potential] Surplus, GWh: 64,346
Future Unmitigated Shortfall, GWh: -66,504
These tell us that the future [Solar + Wind] factor is not sufficiently large for that set of weather data. The Surplus needs to be much greater than the absolute value of the Shortfall to allow for LDES losses.
Increasing the future [Solar + Wind] factor to 4.00 (i.e. by a further factor 1.318) gives more-promising results:
Future [potential] Surplus, GWh: 129,282
Future Unmitigated Shortfall, GWh: -45,166
We can add various quantities of LDES capacity as shown in Figure 6 to see what effect that has:
Figure 6: Future DE, real 2024 data, [Solar + Wind] factor 4.000 - with 200, 400, 800, or 1,200 GWh of LDES:
All that is still nowhere near enough to consistently keep the lights on 60/24/365¼.
Figure 7: Future DE, real 2024 data, [Solar + Wind] factor 4.000 - Increasing LDES until Shortfall is Eliminated:
The alternative to installing quite daunting quantities of LDES energy storage, is to install quite daunting quantities more [Wind + Solar].
Figure 8: Future DE, real 2024 data, Increasing [Solar + Wind] factor and varying LDES to Eliminate Shortfall:
Conclusions
Germany, your recent 3½ years of data tells me your Energiewende (Energy Transition) has a long, long, long way to go before you achieve ‘decarbonisation’ of just your power generation sector.
Germany, your recent 3½ years of data did not appear to include anything you might classify as a ‘dunkelflaute’: it was unremarkable variable weather-dependent generation. You have even farther to go before you achieve ‘decarbonisation’ of your power generation sector in worst-case conditions.
Germany, if you really, really want to ‘decarbonise’ your power generation and avoid bankrupting yourself, you might want to allow near-zero-CO2 Nuclear power back into your energy mix.
Copyright © 2024 Chris S Bond
Disclaimer: Opinions expressed are solely my own.
This material is not peer-reviewed.
I am against #GroupThink.
Your feedback via polite factual comments / reasoned arguments welcome.
In the UK it’s referred to as “constraint”; in the US it’s referred to as “curtailment”; interchangeable terms for when the grid operator tells a generator to reduce or stop their power output.
In the case of Norway the bulk controllable power generation is from hydro.
Other technical limits e.g. spinning reserve also apply, such that curtailment of ‘renewables’ begins below the exact Supply = Demand level to allow a safety margin. For the GB grid that margin seems to be approximately 2,000 MW.
The US DOE in its ‘lift-off study’ of May 2023 defines:
‘Short duration’ = 0-4 hours [e.g. Battery Energy Storage Systems]
‘Inter-day LDES’ = 10–36 hours; RTE 69% ‘today’, 75% target by 2030;
‘Multi-day/Week LDES’ = 36–160 hours; RTE 45% ‘today’, 55-60% target by 2030’
Seasonal shifting’ = 160+ hours [e.g. electrolysis to produce hydrogen]
LDES technologies do NOT currently exist at scale.
The decision to shut down all Germany’s nuclear power plants seems to have been provoked by the Fukushima disaster. Which was caused by a tsunami in a seismically very active part of the planet. Germany has little seismicity and virtually zero tsunami exposure. Meanwhile Germany cemented its reliance on Putin-gas. Go figure.
Quite how one arrives at an ‘optimum’ without knowing costs of future unknown LDES technologies, I’m not sure. But I’ve rolled with it, then adjusted it anyway for technical energy balance reasons.
To be explicitly clear: future Total Wind’ each hour = actual recorded Total Wind multiplied by the [Extra Solar + Wind] Factor; future Solar’ each hour = actual recorded Solar multiplied by the [Extra Solar + Wind] Factor; then compare [future Total Wind’ + future Solar’] against [future in-country Demand] - [Biomass + Hydro] for that same hour. Etc.
Really interesting article - again! Reading the link to the conversation on LINKEDIN regarding the cost of other countries taking our excess energy, it seems that there is always an awful lot of talk about coping with excess energy but nothing much really happens - in the UK anyway? We just seem to build more and more wind and solar. There's storage using thermal salt, hydrogen, chemical batteries, compressed air, liquid air etc. but never a mention of cost or viability. PS works well but expensive to build and controversial regarding the environment. There's talk of Coire Glas PS but not sure if I'll live to see it built. Nuclear seems essential as baseload or spinning reserve and since Germany closed its reactors it has been burning lignite! Green energy is touted as cheaper than FF but the cost of variability in output never seems to me to be factored in. Are LDES in the form of Li-ion batteries the answer in terms of efficiency and cost?