How do we understand decarbonisation progress when emissions depend on the weather?

Power system decarbonisation is often assessed by emission figures, but how much of that progress is real, and how much is just weather-driven fluctuation?

The transition to clean power systems comes with both ambitious goals and significant challenges. National Grid ESO recently suggested an interesting definition of a “clean power system”: “By 2030, clean sources produce at least as much power as Great Britain consumes in total, and unabated gas provides less than 5% of generation in a typical weather year.”  

A power system with 95% clean energy and 5% natural gas would have an estimated carbon intensity of 20 gCO₂/kWh (considering ~400 gCO₂/kWh for gas-fired power plants and 0 for clean energy, ignoring lifecycle emissions). This is very low compared to global electricity averages (~480 gCO₂/kWh) and even lower than the EU average (~240 gCO₂/kWh in 2023).1 2

Power system decarbonization is often assessed by such emission figures, but how much of that progress is real, and how much is just weather-driven fluctuations?

A mild winter or a windy year can temporarily lower emissions, while a cold, still year can spike them. This variability isn’t just an annual effect—it can change emissions significantly even over short periods, such as a three-week span:

Figure 1. CO2 intensity of electricity consumption. Hourly readings over the last 20 days, overall mean represented by a blue asterisk. 3

Does this mean we need to analyse multiple years or periods to get a reliable measure of emission intensity and its progress? Can power system modelling help in this?  

To distinguish structural decarbonisation progress from weather-driven fluctuations, we use power system models that incorporate 33 historical weather years (1983–2015).

The emission levels in two key scenarios from our German Power System study for WePlanet Dach, recently published —“Nuclear” (technology-neutral) and “VRE100” (excluding nuclear)—are shown in the attached figure.

Figure 2. The distribution of yearly direct emissions as percentage of 1990’s level for power sector, from German power systems under “Nuclear” and “VRE100” scenarios. Each boxplot represents the range of outcomes over 33 weather years with boxes showing the 25th and 75th percentiles. Median values are shown with grey lines and mean values with green triangles. The whiskers extend to minimum and maximum values excluding extreme outliners.

On average, emissions in the Nuclear scenario drop to 2.6% of 1990 levels, while the VRE100 scenario reaches 8.4%. However, the impact of weather variability in VRE100 is striking: in extreme years, emissions peak above 12% and dip below 5%, resulting in a ±4% variation. For Germany, this translates to an annual fluctuation of approximately 25 Mt CO₂eq—a significant margin.

As countries advance their decarbonisation efforts, the increasing weather dependency of power systems demands a critical approach to interpreting emission reductions. Power system modelling plays a key role in understanding these dynamics, but future challenges lie in ensuring a representative accounting of evolving climate patterns.

Sources


1 https://ourworldindata.org/grapher/carbon-intensity-electricity?tab=chart&region=Europe

2 https://ember-energy.org/latest-insights/european-electricity-review-2024/eu-electricity-trends/

3 https://www.linkedin.com/posts/grant-chalmers_rstats-ggplot2-gganimate-activity-7253550287859134464-uw0t/?utm_source=share&utm_medium=member_desktop

LinkedInTwitter
© Quantified carbon limited 2025