Back to Blog

French market Q1 - a perfect storm?

Eylert Ellefsen
Archived blog post. This blog post has been transferred from our previous blogging platform. Links and images may not work as intended.

Since the start of November, the French power balance has been limited due to low nuclear availability. Given how strongly France is now connected to the GB market, we have seen spot prices close to 300 €/MWh - about 50 €/MWh higher than the German market.

As the market continues to be dominated by strong gas and CO2 prices, the French Forward prices for January and February are significantly higher than the latest spot prices, indicating a high positive risk premium in a strained power market.

In this blog, EQ takes a closer look into the fundamental power balance numbers for France in Q1-2022, focusing on the outlook for nuclear generation and consumption uncertainty.

Nuclear generation: end of 2021 and Q1-22

In the chart below you see the weekly nuclear production for 2021 and the weekly max-min curves for 2017-2021. By week 49-50 this year, we see that the nuclear production is about 42GW, while the average production figure for 2017-20 is about 46GW.

By week 49-50 2021, around 14 nuclear plants are scheduled to be offline, although more capacity is expected to come online towards the new year.

The current maintenance plan (based on REMIT data) indicates that only 4 plants will be offline during January, which is a huge improvement from December. However, the REMIT data also shows 9 plants offline for February, which makes the market situation more sensitive.

We have compared nuclear production for previous years and the EDF announcement for 2022 in the table below:

In the rest of our study, EQ will use the upper level at 360TWh as an indication for 2022, which is close to the 2021 level.

As the REMIT data for nuclear availability normally reduces before the delivery date, we have made an assumption of the final availability during Q1. In the table below we have calculated the average monthly production for Q1-22 based on statistics from the previous years. For example, by mid-December 2020 the REMIT number for Jan-2021 was 53.6GW, whereas the actual production came out at 49.8 GW. This results in a production factor of 93%.

We see from both the REMIT numbers and production data that availability will be rather weak during Feb-22, while Jan-22 seems to be rather strong.

French power balance data for Q1-22

The table below shows a monthly power balance overview for Q1-22 compared to Q1-21 (as well as Nov-21 as the latest month). Notice that the Jan-22 numbers are based on latest 45-days forecast, whilst both the February and March 2022 numbers are based on long-term normals.

We see from the numbers that Feb-22 comes out with an export balance of -0.5GW, close to the November balance. By comparison, the January export balance is stronger due to the increased availability of nuclear generation.

How the market is trading Q1-22 products during Q4-22

The chart below shows how the January & February-22 power contracts have been traded since the beginning of October. This shows that despite lower nuclear availability, the Feb-22 contract has been traded either lower than or equal to the Jan-22 contract.

This might be explained by low liquidity on the Feb-22 contract, but it will be interesting to see how this develops as we move towards the New Year.

The chart also includes the average spot price for the first 14 days of December – from which we can see a huge gap between the traded monthly contracts and the spot level. It is noteworthy that the spot level is rather high - partly because of the low nuclear availability recently - displaying a strong indication of the high-risk premium in the market right now (Dec 14th).

Consumption uncertainty

Low temperatures are likely an even higher risk factor for consumption estimates than the level of nuclear availability, as the upside risk caused by cold weather is very strong.

As an illustration of the climatic variation band, we have downloaded the climatic scenarios for January and February for the period 1980-2020 on weekly resolution (baseload).

The chart below shows the maximum and minimum weekly consumption levels for weeks 1-8 across the 41 years selected. It is clear that for the years 1985-87, the maximum weekly levels for January-February were very high, close to 85-90 GW. This is 15-20 GW higher than the long-term normal of about 71 GW. In more recent history, 2012 also stands out as an extreme year (85GW).

Whilst many conclusions can be drawn from this data, the most important thing to highlight is that there is a 50% chance of weekly consumption levels exceeding normal levels by 5GW or more for at least one week during January-February 2022.

Should this happen, we would expect to see prices spike as a result, particularly during February due to the low nuclear availability.

Final words

Our analysis shows a very stressed Forward market for France in Q1 2022, with monthly contracts being traded far above spot market price levels. Given the conclusions we are able to draw from our exploration of the data, we consider it likely that price spikes will occur during some weeks in either January or February, due to a combination of strong consumption & limited nuclear availability.

More from the Blog

Good and bad weather stations

Hugo Birkelund
Hugo Birkelund

Removing “bad” weather stations substantially improves the quality and stability of power consumption forecast systems.

Read Story

The Background of Energy Quantified

We are three power market professionals, with 40 years of combined experience in pioneering, building and maintaining well-known analysis platforms. Make no mistake, we feel privileged to have worked with so many exceptional and smart people.

Read Story

What is Energy Quantified?

Hugo Birkelund
Hugo Birkelund

Energy Quantified was founded in September 2017 with the aim of creating the most innovative data and analysis platform for the power markets. Starting with a blank canvas, the portal will combine recent advances in machine learning algorithms with proven fundamental models.

Read Story

Ready to try Energy Quantified?

No payment or credit card required.
Would you rather like a personal demo? Book a demo