Great Britain Reality
How HMGov's "Clean Power 2030" plan looks, using six years of GB data
Introduction
The ‘decarbonisation’ of any power generation system is complex. When I first wrote about it here and here in early 2022, I knew very little. As I learned more and received useful comments I understood it slightly better and consequently adjusted how I performed the GB analysis and took my developing ideas further. I also wrote about the ‘decarbonisation’ of some other electricity grids (California, Texas, US48, New England, Germany, Australia, and so on), learning more as I’ve gone along.
One of the learnings that surprised me, and one that still seems to surprise many, is this fundamental principle of electricity grids:
Generation = Demand *every* minute of *every* hour of *every* day
Generation is Supply; Demand is Consumption; they must instantaneously balance.
If Generation ≠ Demand, frequency goes higher or lower = *very* bad.
This is why there cannot be ‘surplus renewable power’ on a grid. There could be additional reliable schedulable Demand (which we have to build) and then once built, if surplus ‘renewable’ generation was to become available, we could make use of it. Hence the need, for example, for electrical energy storage as a grid becomes increasingly ‘renewable’. As well or alternatively, there need to be high-capacity interconnectors to adjacent territories so that surpluses can be exported and shortfalls imported to keep that critical balance. Absent additional reliable schedulable Demand or someone else to interconnect with, any potential ‘surplus’ has to be curtailed / constrained. Which means increasingly less benefit from the investment in the ‘renewable’ generation facilities: diminishing returns.
The major problem we all have right now, is that electrical energy storage does not yet exist at anywhere near sufficient scale. Nor is what we *do* have anywhere near cheap enough.
I also think official narratives around the energy ‘transition’ underplay the immense scale and technical complexity of what is being attempted. Too much is presented by experts running opaque models concluding ‘everything will be just fine’. Hence I use publicly-available data to illustrate with real numbers where we *are*, and I use simple arithmetic extrapolations to show where we might *end up*.
In short: will the lights stay on?
HMGov’s Plan to ‘Decarbonise’ the GB Power Grid
The plan, as reported by the National Energy System Operator (NESO) in November 2024, is to:
- increase the *capacity* of Offshore Wind to 50 GW;
- increase the *capacity* of Onshore Wind to 27 GW;
- increase the *capacity* of Solar PV to 47 GW;
plus
- “a four-to-fivefold increase in demand flexibility”; [more de-industrialisation]
- “grid connected battery storage [increased] from 5 GW to over 22 GW” [?? GWh ??]
- “more pumped storage” [NESO doesn’t say how much]
- “along with nuclear plant life extensions”
Also:
“Our work identifies two primary clean power pathways. In addition to the elements outlined above, one pathway successfully builds 50 GW of offshore wind by 2030, but no new dispatchable power from hydrogen or gas with CCS. The other pathway delivers new dispatchable plants (totalling 2.7 GW) and 43 GW offshore wind. Either of these requires a dramatic acceleration in progress compared to anything achieved historically and can only be achieved with a determined focus on pace and a huge collective effort across the industry.” [My bolding. Sounds expensive… VERY expensive.]
Oh, and the NESO plan falls flat on its face unless quite incredible (literally “incredible” as in “totally unbelievable”) extra grid capacity can be built within the timescale, see Figures 1. [My bolding. Again, VERY VERY expensive.]
Figures 1: Grid Constraints, and Capacity Expansions for ‘Success’:
There’s a lot about ‘demand flexibility’ in the NESO report, but zero acknowledgement that to date GB’s reductions in Demand and reducing CO2 emission have been accompanied by de-industrialisation. Correlation or causation? You decide.
There is this, however, on Demand in 2030: “Electrification of transport, heat and industry will help enable delivery of the UK’s carbon targets. Our analysis assumes that sectors electrify at sufficient pace to meet the Nationally Determined Contribution (NDC) emissions target for 2030 and the legally-binding carbon budgets recommended by the Climate Change Committee1. This results in an electricity demand growth of approximately 11% to 287 TWh from today to 2030”.
Demand growth: 11%, i.e. Demand factor 1.11.
Of course there’s ‘Green Jobs’ [pdf. 82 pages of waffle] and ‘Leading the World’ [importing everything from coal-powered China]; all, not to avoid ‘Global Boiling’ [hyperbole alert] but to “enable delivery of the UK’s carbon targets”.
Anyway, assuming herculean achievements / spending across all the changes necessary to make this work over the next 6 years: will it?
How does NESO reach its conclusion that things would be fine and dandy if we follow one of its ‘Future Pathways’? Why, it runs its PLEXOS model:
Figures 2: NESO Modelling (Annex 4):
I searched in Annex 4 for the word “valid” as in “validation” or “validity”… nope. Similarly, zero instances of “verif” in Annex 4. Also, zero in the main body of the report. Out of interest, I searched those same terms in the text of Carbon Brief’s 13Dec2024 report on Clean power 2030 and similarly drew a blank.
But, trust the experts!
Ah, no. Not after reading the report from Module 1 of the UK Covid-19 Inquiry and its recommendation to guard against #GroupThink. I couldn’t find any evidence in the NESO report nor its annexes that they have taken that recommendation on board.
So I searched online for “PLEXOS validation” and found many hits… a quick scan of which left me none the wiser as to how PLEXOS’s weather-dependent generation predictions have been validated / verified.
An example: SEM PLEXOS Model Validation (2021-2029) and Backcast dated 18 November 2021 which was “Prepared for the Commission for Regulation of Utilities of Ireland and the Utility Regulator of Northern Ireland” - so not UK DESNZ. It explicitly states “Information furnished by others, upon which all or portions of this report are based, is believed to be reliable but has not been verified.” and “NERA also notes that certain analysis were out of scope of this assignment:” … “Independently verify system data provided by the TSOs (such as wind profiles and load forecasts)”.
Like those in Figures 3, I guess. Yes, the NESO Clean Power 2030 Report does include four sets of 7-days ‘modelled’ weather-dependent generation profiles. One week in winter, one in Spring, one in Summer and one in Autumn. I leave it up to readers to judge how representative (or otherwise) these are by comparison with the six years of data presented in Figure 4 through Figure 9 below.
Figures 3: NESO Weather-Dependent 7-Day Profiles:
I adjusted the height x width dimensions of my screengrabs to be the same, to allow side-by-side comparisons. NESO report Figure 16 is “Modelled 7-day hourly generation profiles for 2023”. NESO report Figure 17 is “2030 Further Flex and Renewables” which corresponds most closely to this post’s basis. NESO report Figure 18 is “2030 New Dispatch” which involves CCS and Hydrogen, gods help us. Figure 16 profile looks significantly different. Figures 17 and 18 look like they used the same profile. Why the difference? No idea.
So, the question for me remains. Even if a miracle was to occur and the NESO ‘plan’ for lots of additional *capacity* of On- and Off-shore Wind and Solar etc. were to come to fruition… would that keep the lights on?
Because the only thing that links all that extra *capacity* with satisfying 2030 Demand 60/24/365 is whatever is in the PLEXOS model.
Alternatively, I’ll have a go.
GB Data Sources
Gridwatch is the excellent and very convenient data aggregator for power flows throughout Great Britain2. Data3 has been recorded there every five minutes since May 2011, and can be downloaded in .csv format and thereafter manipulated in Excel.
The totals of each year’s powers (from Wind, from Solar, and for Demand) as recorded in Gridwatch do not agree with the official statistics presented in the Digest of UK Energy Statistics (DUKES) 6.2 compiled by HMGov (see Tables 1 & 2). I therefore apply ‘adjustment factors’ to the Gridwatch power flows to bring the totals more closely in line with DUKES.4
Table 1: DUKES 6.2; 2019 - 2024: Wind & Solar *Capacities* and GWhs Generated:
Note in Table 1 that the total generated by Wind in 2022 was 80,542 GWh (80.5 TWh) while in 2023 that had increased to 82,309 MWh (82.3 TWh) - a 2.2% uplift in power produced from (30,163/28,761)-1 = 4.9% increase in Wind *capacity*.
Table 2: DUKES 6.2; 2019 - 2024: Wind & Solar Percent Contributions and GB Generation Totals:
Note in Table 2 how the Total Electricity Generation in 2022 was 324,901 GWh (324.9 TWh) while 2023’s total fell 10% to 292,682 GWh (292.7 TWh). By and large, interconnectors made up the difference. That Total Electricity Generation is the denominator when calculating percent renewables. The denominator falls 10% and lo, the percent renewables rise 10%-ish as a consequence of the arithmetic, not because of a triumph of ‘renewable’ tech.
I copied the values from DUKES 6.2 into my Excel tool as shown in Table 3. I then copied this tool and loaded each copy with a whole year of Gridwatch data (2019; 2020; 2021; 2022; 2023; 2024) i.e. a separate file5 for each year’s manipulations.
Table 3: DUKES 6.2 Installed *Capacities* and Resulting Generation:
The whole-year Gridwatch files each comprise around 105,000 rows of data recorded at 5-minute intervals = 5/60th hours. It is simple to use Excel to calculate the arithmetic sum of each column of data, for example the column of Wind power flows recorded in MW. Multiply that arithmetic sum by 5/60 gives the ‘raw’ total energy in MWh. Perform these summations for Wind, Solar, Demand and net Import / Export, to arrive at the ‘raw’ totals for each year shown in Table 4.
Table 4: DUKES 6.3 Total Generation; Gridwatch ‘Raw’ Totals; and Estimation of Future Load Factors:
Comparing that ‘raw’ total to the official DUKES 6.2 total for each parameter and year, gives the relevant adjustment factors shown in Table 5.
I give some credit for future improvements in the technologies in my estimations of the effective generation capability from NESO’s new *capacities*:
Onshore Wind: the five-year average load factor is 26.09 (applied to the existing fleet). I allow 5 points improvement so new Onshore Wind load factor = 31.09.
Offshore Wind: the five-year average load factor is 40.83 (applied to the existing fleet). I allow 10 points improvement so new Offshore Wind load factor = 50.83.
The Wind Multiplier for a given year is then the weighted average of old and new On- and Offshore load factors times the increased total Wind *capacity*.Solar PV: the five-year average load factor is 10.356 (applied to the existing fleet). I allow 1 point improvement so new Solar PV load factor = 11.35.
The Solar Multiplier for a given year is then the weighted average of old and new Solar PV load factors times the increased total Solar PV *capacity*.
Table 5: Adjustment Factors; 2019, 2020, 2021, 2022, 2023 and 2024:
The lower end of Table 5 shows my attempt to enumerate the electricity storage we have in GB currently. The end-of-2024 battery number comes from my I’m Sorry they Haven’t a Clue post. The pumped hydro storage MW & MWh numbers are my best attempt to compile the data I can find7. For convenience I’ve counted battery storage as being ‘long duration energy storage’ (LDES) in the analysis that follows.
In Figure 4 through Figure 9 I present for each year:
- top row: charts showing the year’s Adjusted Recorded Gridwatch power flows;
- bottom row: charts showing the Extrapolated Future power flows using the adjustment factors and multipliers shown in Table 5.
Figures 4: Extrapolations Based on 2019 Gridwatch data:
Figures 5: Extrapolations Based on 2020 Gridwatch data:
Figures 6: Extrapolations Based on 2021 Gridwatch data:
Figures 7: Extrapolations Based on 2022 Gridwatch data:
Figures 8: Extrapolations Based on 2023 Gridwatch data:
Figures 9: Extrapolations Based on 2024 Gridwatch data:
Note: in all of the ‘future’ charts that include the action of LDES, my algorithm only charges the energy storage when there is surplus ‘renewable’ power to be used, and only discharges the LDES when there is a Shortfall of ‘renewable’ power. I understand this is not how energy storage is operated. However, my algorithm’s mode of operation minimises CO2 generation, because if energy storage is charged when there is no surplus ‘renewable’ power it’s likely parasitic on fossil-generated power.
Note also, the magnitudes of the power flows in all the ‘future’ charts:
- peak Wind ~ 70,000 MW (70 GW); minimum ~0 MW;
- peak Solar ~ 40,000 MW (40 GW); minimum 0 MW every night;
- peak Demand ~ 60,000 MW (60 GW); minimum ~25,000 MW mid-summer
Conclusions:
Based on six years of publicly-available data, the NESO Clean Power 2030 ‘plan’ looks rather shaky. From my analysis, annual shortfalls i.e. likely unabated fossil generation quantities needed to keep the lights on, based on the indicated year’s actual weather records, are:
2019: 86 TWh (9.8 GW x 8,760h)
2020: 74 TWh (8.4 GW x 8,760h)
2021: 79 TWh (9.0 GW x 8,760h)
2022: 107 TWh (12.2 GW x 8,760h)
2023: 60 TWh (6.8 GW x 8,760h)
2024: 52 TWh [affected by Gridwatch data freeze]
I think it’s dubious that GB will really find these magnitudes of electricity savings through controlled demand flexibility: far more likely is continuing de-industrialisation through power rationing. The power magnitudes are far too high for interconnectors with neighbouring countries, even if they were willing to enter into such agreements.
P.S. I won’t attempt to critique NESO’s cost assumptions, except to say if this 2018 material is typical of the basis for their decision-making, we’re paddle-less up a well-known creek.
Copyright © 2025 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.
Reminder: none of us voted for the Climate Change Committee or for the legally-binding carbon budgets the CCC has devised or for the ludicrously-tight timescale over which they are being applied. This is all being inflicted on us. The UK Parliament is sovereign, which means it can repeal or amend any domestic legislation, including this.
I’m very pleased to have discovered (as I’ve been gathering data for this post) that the Gridwatch Admins have ensured that the datafiles for each year (2019-2024) included in this post are completely consistent even as, for example, new interconnectors have been added over recent years. Please note, Australia’s Open Electricity.
The data recording froze in Gridwatch from 2024-05-31 22.55 to 2024-07-08 16.25. This simply means that the reconciliation with DUKES will not be possible for whole-year 2024.
Each Excel file with extrapolations, plots, annotations etc. is 57-60 MB in size, hence one file per year.
NESO / DUKES / DNZ please note:
For combined On- + Offshore Wind, the five-year average load factor is 32.56%.
For Solar PV, the five-year average load factor is 10.35%.
DUKES 6.2 Note 6 still states: “Hydro, wind and solar PV capacity are de-rated to account for intermittency, by factors of 0.365, 0.43 and 0.17 respectively…”
0.43 = 43% = 32% overstatement of your ‘de-rated’ reported numbers for Wind
0.17 = 17% = 64% overstatement of your ‘de-rated’ reported numbers for Solar PV
The number of hours each pumped storage scheme could deliver full power depends, rather obviously, on them being fully ‘charged up’ to start with. Whether they could beneficially deliver at those powers also depends on there being sufficient grid capacity between them and the Demand(s) they are trying to supply at the time. Ffestiniog and Dinorwyg are in Wales; Foyers and Cruachan are in Scotland; there are big grid constraints in the way, see Figures 1. By the same logic, whether they could be ‘charged up’ also depends on sufficient grid capacity existing between them and whatever resource is generating the power to be stored.
Interesting thought provoking post. NESO 2030 plan is plausible as a solution but entirely undeliverable and the only way to get anywhere to close to it means importing even more high value equipment as the UK makes pretty limited amount of kit now. We can make blades at scale but thats about it the rest of it comes from largely Denmark, Germany and China so its them that will get the green jobs not the UK. Its utterly disingenuous of Millibrain to say there will a UK green jobs bonanza there wont be and its about time the unions called him out for that. It gets worse as deindustrialisation has largely been responsible for the UKs reduced CO2 and we haven't helped the global challenge one bit and probably worsened it actually by buying from countries with higher fossil fuel usage let alone lower environmental standards.
Personally i see a push/pull going on in NESO between the engineers that know what it takes to keep the lights on vs the NZ evangelists with engineers winning round one and at least getting acknowledgement that gas will be needed for the foreseeable future in a standby capacity. Im not anti NZ but i don't see the urgency of 2030 given the rest of the globe already dwarfs the CO2 we emit so for me a slower plan that keeps costs down and allows UK manufacturing to be restored should be the way forward.
Renewables grid penetration will never exceed around 45% because without unaffordable, unsourceable storage, it will always need coal, gas and nuclear to back it up for those windless, sunless periods - global stilling is increasing reducing further wind farm CF’s in future years and solar above the 45th parallel, is next to useless for grid scale