French Power Demand
Weather & Multi-Energy Transition Case Study
A public-data market intelligence case exploring how demand, weather sensitivity, electrification, flexibility and new energy vectors interact across the French energy transition.
Built as a disciplined market-intelligence workflow: data ingestion → source health → demand baseline → HDD weather normalisation → forecast skeleton → transition scenarios → multi-energy vector framework → executive storytelling.
Executive KPIs
Scope and rigor of the case study in six numbers.
Exploratory winter-demand sensitivity using HDD base 18 and 46 aligned observations.
Index-point divergence across transition pathways, showing how assumptions reshape exposure.
Highest 2040 multi-energy index, driven by electrolysis, industrial load and e-fuels.
Accelerated electrification sits close to the hydrogen pathway in 2040 exposure.
Short weather-demand sample: enough to structure the method, not enough to overclaim.
Electricity, methane/LNG, biomethane, hydrogen, e-fuels and biomass mapped together.
The 11-step analytical chain
From governance and public data to weather normalisation, scenario logic, multi-energy exposure and publication-ready claim boundaries.
Scope & claim governance
ValidatedDefine the public-data scope and prevent unsupported forecast, trading or internal-model claims.
Public data foundation
ValidatedMap available demand, weather and metadata sources and document access limits.
Public data ingestion
ValidatedUse RTE/ODRÉ éCO2mix demand, Open-Meteo weather and INSEE metadata context.
Demand baseline KPIs
ValidatedStructure hourly and daily demand patterns before adding scenarios.
HDD weather normalisation
ValidatedEstimate a short-window weather sensitivity using HDD base 18.
Forecast skeleton
ValidatedBuild scenario architecture without claiming validated operational forecast accuracy.
Medium-term scenario drivers
ValidatedCompare 2030, 2035 and 2040 transition pressure across transparent pathways.
Multi-energy vector framework
ValidatedConnect electricity with methane/LNG, biomethane, hydrogen, e-fuels and biomass exposure.
Executive dashboard packaging
ValidatedTranslate technical outputs into a readable market-intelligence narrative.
Communication pack
ValidatedExplain assumptions, caveats, key results and reproducibility clearly.
Public-readiness audit
ValidatedCheck old-scope contamination, overclaiming, placeholders and publication risk.
Key result · Weather
Weather first: separating heating volatility from structural demand
Before reading demand growth as electrification, industrial load or flexibility, the case isolates the weather-driven component using HDD base 18.
Aligned RTE/Open-Meteo winter sample · 46 observations
Signal check
Material weather beta, limited standalone fit.
Weather driver
HDD base 18
winter heating proxy
Thermal sensitivity
+1.4 GW/HDD
1,404.58 MW/HDD
Correlation
0.3027
observed weather-demand fit
Evidence base
46 obs
aligned winter sample
Why it matters
The important signal is the combination of a large weather sensitivity and a modest correlation: weather clearly matters, but it does not explain demand variation on its own.
That is precisely why this layer is useful. It reduces the risk of reading cold-spell volatility as structural demand growth.
In the analytical chain, this filter sits upstream of electrification, industrial-load and flexibility scenarios.
Boundary conditions
Exploratory single-driver layer from the analysis repository. A production-grade specification would add calendar effects, hourly controls, seasonality, holidays, activity signals, non-linear temperature effects and fuel-switching controls.
Key result · Forecast
Forecast skeleton: separating demand pressure before making claims
This layer is not an official demand forecast. It shows the architecture needed to separate baseline demand, electrification pressure, efficiency response and cold-weather stress before scenario interpretation.
MW · current sensitivity layer · scenario averages
Scenario spread
3.72 GW
3,717.74 MW between low and high cases
Weather stress
+2.81 GW
versus reference skeleton
Electrification pressure
+1.51 GW
versus reference skeleton
Efficiency relief
−0.91 GW
versus reference skeleton
Market read
Sensitivity architecture before forecast claims.
The useful signal is not the absolute forecast level. It is the separation between structural pressure, weather stress and efficiency response.
In this skeleton, cold-weather stress creates the highest demand pressure, while efficiency response is the only case below the reference anchor.
The decision value is methodological: this layer shows what should be separated before moving to a production-grade demand model or scenario interpretation.
Boundary
Public-data forecast skeleton only. It should be read as a sensitivity and model-architecture layer, not as an official demand outlook, operational forecast interval or trading recommendation.
Key result · Scenarios
Medium-term transition scenarios: reading demand dispersion, not a single forecast
Four indexed narratives, with 2026 = 100, compare reference transition, accelerated electrification, efficiency and flexibility response, and hydrogen-linked industrial pressure.
Scenario spread
6.7 pts
in
2030
Scenario spread
15.5 pts
in
2035
Scenario spread
22 pts
in
2040
Four indexed transition narratives · medium-term horizon
Reference
110.4
2040 endpoint · +10.4 pts vs 2026
Electrification
128.6
2040 endpoint · +28.6 pts vs 2026
Efficiency
109.4
2040 endpoint · +9.4 pts vs 2026
Hydrogen
131.4
2040 endpoint · +31.4 pts vs 2026
Market read
Dispersion becomes the signal.
The important signal is not one central demand path. It is the wideningscenario dispersionbetween 2030, 2035 and 2040.
By 2040, hydrogen-linked industrial growth reaches the highest endpoint, while accelerated electrification remains close behind. Efficiency and flexibility moderate pressure rather than remove it.
For market access, this creates a practical question: which customers, segments and load shapes are exposed to structural demand pressure, and which can be managed through flexibility or efficiency.
Decision use
This section converts transition narratives into an indexed scenario frame that can feed portfolio stress tests, commercial segmentation and medium-term market monitoring.
It should be read as a public-data scenario layer, not as an official forecast, load forecast submission or trading signal.
Boundary
Indexed scenario narratives depend on assumptions. A production model would require sector-level load drivers, calibrated adoption curves, calendar controls, weather normalisation and explicit uncertainty bands.
Key result · Multi-energy
Beyond electricity: mapping cross-vector transition exposure
This layer does not forecast commodity prices. It maps how electricity, molecules, industrial decarbonisation and flexibility can interact across transition scenarios.
Electricity
Core demand vector affected by EVs, heat pumps, industry and electrolysis.
Methane / LNG
Residual flexibility and security-of-supply vector during transition.
Biomethane
Renewable gas substitution and molecule decarbonisation vector.
Hydrogen
Electrolysis-linked vector connecting power demand with industrial decarbonisation.
E-fuels
Electricity-linked molecule pathway connected to hydrogen and synthesis routes.
Biomass
Dispatchable renewable and industrial decarbonisation support vector.
Higher score = stronger transition pressure across energy vectors
Market read
Cross-vector pressure, not price forecasting.
Highest exposure
Hydrogen industry growth
2040 index 131.4
Exposure spread
24.2 pts
Highest vs lowest composite pathway
The key signal is the concentration of pressure across electricity, hydrogen and e-fuels. Hydrogen industry growth creates the highest composite exposure by 2040.
For market access, this turns demand risk into a cross-vector question: which customer segments are exposed to electricity demand, molecule substitution, industrial decarbonisation and flexibility constraints at the same time?
2040 transition index
Composite across all six vectors
Boundary
Composite scores are an exploratory scenario layer. They do not represent commodity price forecasts, official demand forecasts or trading signals. A production model would require calibrated weights, sector-level demand drivers and explicit uncertainty bands.
Analytical capabilities demonstrated
The value is the discipline of turning public energy signals into bounded market interpretation.
Public demand, weather and metadata sources are documented before any modelling layer.
Hourly and daily demand baselines provide the starting point for interpreting exposure.
HDD base 18 isolates heating-related demand variation before structural interpretation.
Scenario skeleton compares sensitivity bands without presenting an official forecast.
2030, 2035 and 2040 indexes compare electrification, efficiency, flexibility and hydrogen pressure.
Power, gas/LNG, biomethane, hydrogen, e-fuels and biomass are read as connected transition vectors.
The case turns technical outputs into market takeaways, caveats and decision-support language.
The analytical workflow is reproducible, versioned and structured for public review.
The market read
- 1Weather first, structure second: demand changes should not be interpreted before isolating weather sensitivity.
- 2Electrification raises the demand ceiling, but efficiency and flexibility can materially offset gross pressure.
- 3Hydrogen and e-fuels reconnect molecule decarbonisation with electricity-demand exposure.
- 4Methane/LNG is treated as residual flexibility and security-of-supply, not the main structural growth vector.
- 5The case demonstrates analytical architecture and assumption discipline, not an official market forecast.
Methodology discipline
This distinction is intentional: the project demonstrates analytical architecture and assumption discipline, not an official market forecast.
Methodology, evidence and boundaries
Explore the workflow, key results, scenario logic and claim boundaries behind the public-data case.