The real cost of business mobility in Q3 2026

Saying "electric is cheaper" is both true and not enough. There are several prices per kWh depending on the use case: the regulated residential tariff (about €0.193/kWh on the base option, €0.158/kWh off-peak for a 9 kVA meter) and the professional tariff (€0.317/kWh incl. VAT, small-consumer band, source Eurostat), the one a fleet pays when it charges on its own site (depot, company premises). That last one is what underpins the headline figure of €5.38/100 km. None of these tariffs is just called "the electric cost": each maps to a real use case. Hence the table below, built on ADEME reference consumption figures (5.8 L/100 km for diesel, 17.0 kWh/100 km for electric).

The reading is clear: an employee who charges their company car at home, off-peak, spends €33.5/month on electricity for 1,250 km, against €136/month on diesel for the same distance. That is the gap finance teams find line by line in expense claims, and it is what pushes towards flat-rate reimbursement of home charging rather than paying for fuel top-ups. And public charging? The barometer does not (yet) publish a roaming scenario: the "pricing" field in the IRVE database is free text and sparsely filled, too noisy to derive a consolidated €/kWh from. We do not paper over it with a proxy: we wait for a sourced series.

Cost per kilometre only tells part of the story. Total cost of ownership (TCO) adds maintenance, insurance and depreciation to it. At 15,000 km/year, the barometer publishes a measurable sub-total: energy (measured each quarter by the pipeline) + maintenance + insurance (sourced annual parameters). It comes out at €2,887/year for combustion against €1,993/year for electric, a substantial gap that widens with mileage.

Why talk about a partial TCO? Out of honesty. Depreciation is published only as a rate (6.8%/year for combustion, 11.4%/year for electric): turning it into euros is waiting on sourcing a reference vehicle price, currently a TODO. The lease payment, meanwhile, is structural (it replaces purchase) and is not included in the sum. We do not fill the gap with false precision: we publish what is measured and we name what is missing. That is what makes the barometer citable by the press and by AI engines.

This is the barometer's recurring indicator, tracked from one edition to the next: the breakeven point. It answers a simple question: at what kWh price does electric stop being cheaper than combustion (on energy alone)? In Q3 2026, the threshold sits at €0.640/kWh. The actual professional kWh price is €0.317/kWh, leaving a comfortable margin of €0.323/kWh before parity. Verdict: electric stays cheaper, and by a wide margin.

This indicator has one virtue: it turns a debate that is often ideological into a quantified, verifiable threshold. If the professional electricity price were to double, we would only just approach the parity threshold. Below that, the gap holds. The barometer will publish this breakeven point every quarter: it is the one that will say, with the data to back it up, when the switch becomes arithmetically contestable.

The IRVE database lists, as of Q3 2026, 160,138 charge points across 63,946 stations and an installed capacity of 9,537.7 MW. The figure is robust, on one condition: the data has been cleaned. The consolidated database is denormalised (one row per charge point, with the station attributes repeated), which forces us to deduplicate on the charge-point identifier, otherwise the same terminal gets counted several times: 67,094 duplicates removed out of 227,232 rows carrying an identifier, for 160,138 unique points.

Above all, the puissance_nominale field is riddled with data-entry errors: values in watts instead of kilowatts (above 4,000 kW, we divide by 1,000) and absurd power ratings (above 400 kW, excluded). Once cleaned, the installed base is consistent: the typical terminal sits at 22 kW (median), the profile of a network mostly made up of 7 to 22 kW AC terminals. The average, for its part, reaches nearly 60 kW: it is pulled upwards by the tail of fast DC terminals (≥150 kW). On capacity, it is the median you should quote, not the average.

A regional ranking of electric "readiness" is not published by halves. To mean anything, it rests on three components: the energy price, the density of terminals relative to the regional vehicle fleet (never to population or land area, which mechanically favour Île-de-France) and the growth of the network. In Q3, the founding edition, two are still missing: growth (one edition = no N vs N-1 comparison) and the regional climate penalty (Météo-France temperatures not yet aggregated).

Density, for its part, is now sourced: the passenger-car and LCV fleet by region comes from the SDES (fleet as at 1 January 2025). But publishing a ranking today, stripped of its growth component, would amount to crowning the region with the cheapest diesel, regardless of its terminal network, a precise and misleading league table. The barometer prefers to wait for Q4 and its three legs. An honest ranking beats a misleading one.

Several indicators remain unavailable in Q3 2026, and we state it plainly rather than papering over them. The winter cost premium (electricity consumes more in deep cold) is waiting on the aggregation of Météo-France regional temperatures. The cold-penalty curve, for its part, is already sourced. New registrations of company vehicles (to track the shift to electric in real time) are waiting on a PFA import. Failing that, the barometer already publishes the share of powertrains in the standing fleet (Diesel 48.3%, Petrol 40.4%, Electric 2.9%, source SDES). Lastly, the LOM countdown is waiting on the legal milestones (to be transcribed from Légifrance, not from memory). Gaps, but the sources are identified. That is how a reference tool gets built, edition after edition.

The barometer is reproducible: a Python pipeline (stdlib only) collects the public data, processes it and consolidates ten indicators in five commands. The full dataset for this edition is downloadable in JSON and CSV. The methodology documents the sources, assumptions and versioned parameters. Greenway Index for this founding edition: 100.0 (base 100). Future editions will read as a change from this base.

To cite: Smadja, A. (2026). Greenway Business Mobility Barometer, Q3 2026. Greenway. Data extracted 3 July 2026.