More problems with Miliband’s maths

NESO recently published a report claiming that Mr Miliband’s 2030 Net Zero plans would lead to cost savings. I asked for the underlying calculations under FOI. They were released at the end of last week.

I’ve started going through it, and there is a lot of interest. I’ve already found one major saving that looks as though it’s nothing of the sort.

The savings are presented in the report’s Table 4. The line in question is shown below, and relates to what the NESO calls “the Merit Order Effect”. The original table only gives per-megawatt-hour figures. I’ve added absolute (£ billion) values.

The purported saving, of £2.9 billion (that’s £100 per household), is based on the idea that because renewables are intermittent, they are less valuable, and so earn a lower price in the market than baseload generators. In future, they will earn even less, and so there will be a saving to consumers.

The capture price is how much a renewables generator earns in the market, and the capture rate is that price expressed as a percentage of what a baseload generator earns. Capture prices are irrelevant to CfD generators, because they get a fixed price whatever. But there is still a lot of RO generation which will indeed be subject to this effect.

The principle of the calculation is simple: for each technology, we have the 2023 and 2030 capture prices. We calculate the change in price and then multiply by the amount of generation in 2030 to get the cost saving. But, as you will see, NESO’s seems to have made a bit of a hash of it.

Problem 1

Problem 1 relates to the 2023 capture prices. There is a table in the spreadsheet that shows how they are calculated. The “Combined 2023” column is a weighted average of the two others, giving 3 times as much weight to the second. In other words, this is meant to represent a 12-month period starting from Q4 2022, with an assumption that 2023 prices remain steady from Q1 onwards.

In each case, the renewables are capturing over 90% of the baseload price.

The problem is that what goes forward to the next step of the calculation is inconsistent: it uses the “Combined 2023” figure for baseload, but the Q1 2023 figures for renewables capture prices.

I can’t think why that would be. If I change the £131 figure to £100.1, it increases the capture rates to figures in the 90s, and reduces the apparent saving to £2.3 billion.

Problem 2

Using values based on 2022 and Q1 2023 – the period of the energy price crisis – means they are very high. The graph below shows the day-ahead prices by month, and I’ve marked on the NESO assumption for what offshore wind will capture. Clearly there was much less price to be “captured” during 2024, and it’s not hard to understand why NESO didn’t use a more up-to-date figure. They have clearly been inspired by Labour’s trick of claiming they will deliver cost savings that have already happened.

If I add in a correction to the 2023 baseload price, lowering it to £72 (the average market price during 2024), while keeping the capture rates in the 90s, the overall saving drops to £0.6 billion.

Problem 3

For the closing capture prices, the assumption is that wind will deliver around 60% of the baseload price in 2030, while solar will capture 80%. This seems quite reasonable compared to the 2023 numbers, in the 70s percent, presented before I fixed Problem 1, but looks very ambitious compared to the numbers in the mid-90s afterwards. If I assume instead that we fall to just 80% by 2030, the saving alleged by NESO just about disappears entirely.

This seems to make sense to me. Market prices will rise a little because of the other assumptions in the model, but renewables will capture less value, so there is very little gain to be made.

But, Problem 4

One compensating error is worth mentioning. In calculating the output of these RO windfarms, NESO takes the total offshore fleet as being 14GW. This is correct. It then deducts CfD windfarms, which, as noted above, are irrelevant to this calculation. However, it says these amount to 12.5GW. This is incorrect. This is a list of operational CfD windfarms.

This means there will be more windfarms operating under the renewables obligation and so the saving is proportionately larger. But only marginally so. The revised total is £0.2 billion.

In passing, I will mention another issue. If NESO’s renewables capture prices are correct, then offshore windfarms will only be earning £48/MWh. Older offshore windfarms spend more than that each year on maintenance. That’s fine while they are receiving their subsidies, but as soon as those stop, they will become cash negative and will close. In 2030, that will only cost just under 1GW, but by 2038, the whole offshore RO fleet will have closed, taking the rest of the related savings with it.

Conclusion

As far as I can tell, £2.7 billion of NESO’s alleged £2.9 saving for the Merit Order Effect is spurious. That’s an extra £100 per year you will have to find.