Skip to content
Electrify Europe

Methodology & sources

How to read the numbers in this dashboard: definitions, data sources, and known limitations.

Definitions

Final energy consumption
Energy delivered to the end consumer for actual use (heating, mobility, industrial processes...), after transformation and transport, as opposed to primary energy (extracted before transformation). This is the measure used for every figure on this site.
Electrification rate
Share of electricity in the final energy consumption of a sector or sub-sector, as a percentage: electricity consumption divided by total consumption, across all energy carriers.
Sector / sub-sector
This site distinguishes 5 sectors (residential, tertiary, industry, transport, agriculture). The electrification rate and consumption by sector are Eurostat's direct figures (harmonised EU source, see below). Detail by sub-sector comes from separate SDES sources, with a different vintage and classification: energy end-uses (heating, hot water...) for residential/tertiary (CEREN), transport modes (energy balance) and the farming branch (PEFA). It may therefore not sum exactly to the sector total. Industry is deliberately left without sub-sectors: the only breakdown available (PEFA, NACE branches) has a scope too far removed from the Eurostat total to be shown without being misleading.
Charging session
A connection between a vehicle and a charge point, from plug-in to unplug: it carries a start timestamp, a duration, an amount of energy delivered, and a success status. It is the basic unit of the Charging pages.
Location type
Category of the location where the charge point sits. The Charging pages distinguish five: business (private use), business (open to the public), local authority, multi-unit residential, and heavy goods vehicles. Each session is linked to one of them.

Conversion factor

1 Mtoe (million tonnes of oil equivalent) = 11.63 TWh. The unit selector at the top of the page applies this conversion to every figure and chart on the site.

Sources

Two kinds of data

The dashboard combines two ways of bringing data in, without ever fabricating any:

Automated open-data feeds
Most modules: data is pulled directly from a public API (Eurostat, Enedis, GRDF, IRVE, DiDo/SDES, ADEME, Agence ORE, Ecolab) and re-aggregated automatically on every update.
Assisted extraction from public PDFs
Where no API exists (monthly registrations by energy, heavy-truck market structure, Avere/Gireve barometers, CCFA/CSIAM orders), figures are extracted from official PDF publications, with the original kept for verification.

How the figures are computed and selected

A few deliberately simple, verifiable computation choices:

Electric share
Always a ratio (electric vehicles ÷ total), recomputed from raw counts rather than taken from an already-rounded percentage, so it stays independent of the source's unit.
Aggregation at the source
Very large datasets (IRVE charge points, the 12-million-row DPE database, municipal consumption) are never pulled row by row: they are aggregated server-side, and only small pre-reduced totals are published.
Selection & scope
Every figure keeps its source's scope: 5.1-tonne heavy-truck threshold, tonnage × energy limited to goods categories (N2/N3), France and EU27 rows for the European comparison. An unpublished value stays empty, never estimated.
Derived values, flagged
When a source only publishes a cumulative figure, the missing monthly detail is reconstructed by exact difference (never invented) and flagged as derived.

Official targets (PPE3)

The “target” trajectories overlaid on the homepage and the Industry page come from the Décret n° 2026-76 du 12 février 2026 relatif à la 3ᵉ programmation pluriannuelle de l'énergie (PPE3) (published in the Official Journal on February 13, 2026). The decree sets the share of electricity in final energy consumption at 34% in 2030 and 38% in 2035 (up from 27% in 2023), and, for the first time, a numerical trajectory for industry: from 30% today to 47% in 2035.

This “30%” starting point for industry, cited in the decree, does not exactly match the Eurostat electrification rate shown on this site for 2024 (35.7%): this is likely a methodological or scope difference between the PPE's energy accounting and the Eurostat breakdown (nrg_bal_c) used here. Both values are shown without being merged.

Electric vehicle charging

The Charging pages rely on the Observatoire national de la donnée de recharge (Advenir × ADEME) (32.9 million sessions, 2016-07 → 2026-06). Bornes financées par les programmes Advenir et ADEME (~150 k points de charge), pas l'intégralité du parc de recharge français : les niveaux et les parts entre typologies ne sont pas directement représentatifs de l'ensemble du parc.

24-hour charging curve
For each session, the energy delivered is spread pro rata across every hour of the day it spans (a session straddling two hours feeds both, proportionally). The average power shown for a given hour is the energy attributed to it divided by the cumulative active charging time within that hour, calculated only over sessions that actually delivered energy. This is therefore an average power “at the plug”, smoothed over the vehicle's entire dwell time, not a peak power. Even heavy goods vehicles therefore show modest values, since many charge slowly overnight at a depot.
Reliability and failed sessions
The success rate is the share of completed sessions out of the total, by location type and by month. Failed sessions are never discarded: they form a dimension of their own (reliability) and are counted in the energy breakdown; they are only excluded from the average-power denominator, since no energy was delivered.
Assigning sessions to a month
A session's month is derived from its actual start date, never from the source file's name: a file dated “2022-08” may contain sessions that started anywhere from June to August.

Known limitations

  • The electrification rate, total consumption, and the historical series by sector all come directly from Eurostat (nrg_bal_c), regenerated automatically whenever the data is updated: the reference year shown (2024) is the latest one available for France at generation time, not a value fixed by hand.
  • The flow diagram (Sankey) is now rebuilt from Eurostat (nrg_bal_c), just like the sector layer: same source and vintage as the sector electrification rate. Only the breakdown by end-use sub-sector remains sourced from SDES/CEREN and updated manually: Eurostat stops at the sector level, it doesn't go down to end-use. This layer therefore has neither the same vintage nor the same method. The sum of a sector's sub-sectors, or the rate recalculated from them, can differ noticeably from the Eurostat sector figure shown: two independent sources are presented side by side. Each sub-sector displays its actual vintage in its badge (“{year} data · SDES/CEREN”): residential 2024, tertiary 2020, transport 2024, agriculture 2023 (different source workbooks).
  • The Charging pages only cover charge points funded by the Advenir and ADEME programmes (~150,000 charge points), not France's entire charging infrastructure: the levels and the shares between location types are not directly representative of charging across France as a whole.
  • The “France (weighted average)” series compared against the PPE3 targets doesn't exist as such in the source data: it is calculated by weighting each sector's rate by its consumption in the reference year (2024), with these weights applied identically across the whole series, even though each sector's actual weight shifts slightly from year to year.
  • The Electricity consumption (Enedis) and gas (GRDF) pages cover the distribution network (~95% of consumption), excluding large industrial sites connected directly to the transmission network (RTE HTB / GRTgaz): the “industry” share is therefore under-estimated compared with full national balances. On the Enedis side, the series starts in 2018: a change in segmentation makes earlier years non-comparable, so they are excluded.
  • The charging point fleet (IRVE) register only includes charge points that are open to the public and declared under the IRVE scheme: neither the private fleet (homes, closed businesses) nor undeclared charge points are counted.
  • The carbon content of electricity (éCO2mix) page shows an RTE estimate (emission factors by generation source), not a full life-cycle analysis.
  • The Housing energy ratings (DPE) page only covers homes that have undergone an energy assessment (sale, rental, new build), not an exhaustive census of the housing stock, with a likely bias toward homes currently on the market. For readability, the charts on that page group ADEME’s ~15 heating-energy labels into 6 families (electricity, natural gas, heating oil/LPG/propane, wood, district heating and cooling, other). Shares add up within a family, and the grouping is reflected in the “View data” tables.
  • The vehicle registrations data is published by SDES with an annual vintage (no monthly granularity), and the data centers figures only cover centers using more than 1 GWh per year.
  • The electric-vehicle fleet is shown from two complementary angles: the stock as of 1 January (annual SDES source, all segments: cars, light commercial vehicles, trucks, buses & coaches) and a quarterly view of plug-in cars (Agence ORE, passenger cars only). The two definitions are close but not identical. The departmental breakdown of the fleet stays on passenger cars.
  • The private-owner / business split of the fleet reflects the status of the REGISTERED KEEPER (SDES column STATUT_UTILISATEUR), not the vehicle's actual use: a van owned by a leasing firm and driven by a private individual counts as business-held, and conversely for a sole trader.
  • Passenger-car and van orders (CCFA/CSIAM barometer) are a leading indicator: an order book, distinct from vehicles actually registered.
  • The France vs Europe comparison uses ACEA aggregates (France and EU27 rows); the breakdown by energy has only been published since 2023.
  • The tonnage × energy breakdown of new heavy trucks comes from the European Environment Agency's CO₂ micro-data (goods categories N2/N3): latest complete year 2022, as this source lags by about two years.
  • Displayed charging prices come from the pricing declared in the IRVE file (filled in by a subset of charge points); the home benchmark (regulated CRE tariff) and public-charging benchmark (CLCV survey) are external reference points cited with their source.

Emission factors

Supporting calculation building block (not a dedicated page): a small selection of emission factors (gCO2e per kWh) by energy carrier, taken from ADEME Base Carbone® (version V23.6). It serves as a reusable building block for modules that convert energy consumption into emissions (e.g. éCO2mix, avoided-emissions calculations).

  • Électricité - mix moyen France continentale0.052 kgCO2e/kWh
  • Gaz naturel - combustion0.243 kgCO2e/kWh PCI
  • Fioul domestique - combustion0.325 kgCO2e/kWh PCI
  • Fioul lourd - combustion0.325 kgCO2e/kWh PCI
  • GPL / Propane - combustion0.271 kgCO2e/kWh PCI
  • Houille (charbon) - combustion0.377 kgCO2e/kWh PCI
  • Bois bûche (20% humidité) - combustion0.603 kgCO2e/kWh PCI
  • Bois granulé - combustion en chaudière0.027 kgCO2e/kWh
  • Bois granulés - combustion en poêle à bois0.032 kgCO2e/kWh

Sélection restreinte aux vecteurs pertinents pour ce dashboard (électricité, combustibles fossiles et bois utilisés en chauffage). Chaque valeur est la ligne "total" renvoyée par l'API pour l'élément (combustion + amont déjà sommés côté ADEME quand l'élément est décomposé) : aucune sommation n'est effectuée par ce script. Les réseaux de chaleur ont été exclus : la Base Carbone ne publie qu'une valeur par réseau nommé (des centaines d'entrées très hétérogènes), pas de moyenne nationale unique exploitable ici.

Back to home