Since April 2026, Tesla has been running a three-tier Supercharger pricing structure across Germany, France, the Netherlands, Sweden, Norway and Denmark, and buried in it is a detail that matters more than the headline kWh rate: the idle fee is no longer flat. Once a car finishes charging, Tesla charges roughly €0.50 per minute if the station is at 50% occupancy or higher, and that fee doubles at full occupancy. Move your car within five minutes of the "charging complete" notification and the fee is waived entirely. Below 50% occupancy, there's no idle fee at all.
That's not a punishment for idling. It's a price that tracks the actual cost of the thing being priced: a blocked bay is worthless as a problem when the station is half empty, and expensive when someone is waiting. Most CPOs in Europe haven't built that logic in, even the ones who've already adopted dynamic energy pricing.
What Tesla actually charges
The structure is simple once you see it: nothing below half-full, roughly €0.50 a minute from 50% occupancy up, and double that once the station is completely full, with a five-minute grace period after the "done" notification before any of it kicks in.
Tesla's Supercharger idle fee in Europe, since April 2026. No fee below 50% station occupancy, roughly €0.50/min from 50% up, doubling to about €1.00/min at full occupancy. A five-minute grace period after the "charging complete" notification applies before the fee starts. Source: Tesla Support.
The low-utilisation paradox
The IEA's Global EV Outlook 2026 puts time-based utilisation of the European public charging network at around 10% in 2025, projected to climb to roughly 15% by 2035. On paper, that sounds like blocked bays shouldn't be a real problem, most of the day, most stations have space.
But averages hide exactly where this matters. A network running at 10% utilisation across 24 hours is not evenly occupied, it's empty most of the time and briefly saturated during a handful of predictable windows: weekday evening commutes, weekend travel corridors, the two hours around a motorway service stop at lunch. Utilisation is low on average and high in the moments a driver actually cares about, which is precisely the pattern occupancy-based pricing is built for and flat idle fees are blind to.
What a flat idle fee gets wrong
A flat idle fee, charged the same at 2am with an empty station as at 6pm with a queue, solves the wrong problem twice. Off-peak, it annoys drivers who aren't actually blocking anyone and adds friction for no operational benefit. At peak, when a bay actually is blocking a waiting driver, a fee set low enough to be tolerable off-peak often isn't high enough to change behaviour when it matters. Either the fee is too aggressive when nobody's waiting, or too weak when somebody is. A flat number can't do both jobs at once, because it's answering a question that changes by the hour: how much is this bay worth right now.
Plenty of European operators still run no idle fee at all. Given the AFIR-driven pressure on transparent, ad-hoc pricing and the general reluctance to add fees that look punitive to drivers, that's an understandable default. But it leaves a lever unused that has nothing to do with squeezing extra revenue out of finished sessions, its actual job is protecting throughput on the ports that are already scarce.
The same input your dynamic pricing already has
Here's the part that should be encouraging rather than daunting for operators already running demand-aware energy pricing: occupancy-based idle fees don't need a new data source. If your platform already adjusts energy price by time of day, location, or real-time demand, you already have a live read on session state and connector occupancy per site. That's the same signal an occupancy-tiered idle fee needs. The energy price answers "how much should this kWh cost given demand." The idle fee answers "how much should this minute of occupying a scarce bay cost given demand." Same underlying question, different unit.
That also means an idle fee doesn't have to be a single site-wide flat number the way most implementations treat it. It can key off the same occupancy tiers a dynamic pricing engine is already tracking: low at low occupancy, meaningfully higher once a site crosses a threshold where someone else is plausibly waiting. Tesla's 50%/100% split with a grace period is one reasonable version of that. It's not the only one, but the shape, tiered to real occupancy rather than flat, is the part worth borrowing regardless of the exact thresholds.
The practical takeaway
If your network runs a flat idle fee, or none, it's worth checking what your busiest sites look like at their actual peak hour, not their daily average. A network at 10% average utilisation can still have specific bays blocked for real money most weekday evenings. The fix isn't necessarily a higher fee. It's a fee that's aware of when blocking a bay actually costs someone something, and when it doesn't.
We built ProxiLink around pricing that responds to real demand rather than a flat schedule. Occupancy-aware idle fees are the same principle applied to the other half of a charging session, what happens after the kWh stop flowing. If you want to see what that would look like on your own network, we're happy to walk through it.
Frequently asked questions
What exactly changed in Tesla's European idle fee since April 2026?
Since April 2026, Tesla's Supercharger idle fee across Germany, France, the Netherlands, Sweden, Norway and Denmark stopped being flat. Once a car finishes charging, it's roughly €0.50 per minute if the station is at 50% occupancy or higher, doubling to about €1.00 at full occupancy. Below 50% occupancy there's no idle fee at all, and a five-minute grace period after the "charging complete" notification applies before any of it kicks in.
Why doesn't a flat idle fee work as well as an occupancy-tiered one?
Because it's trying to answer a question that changes by the hour with a number that doesn't. A flat fee set low enough to be tolerable at 2am with an empty station is usually too weak to change behaviour at 6pm with a queue, and a fee aggressive enough to matter at peak ends up annoying drivers who aren't blocking anyone off-peak. Occupancy-tiered pricing prices the actual cost of the bay being blocked, which is close to zero when the station is empty and real when someone is waiting.
Doesn't Europe's roughly 10% average public charging utilisation mean blocked bays barely matter?
The average hides where it matters. IEA's Global EV Outlook 2026 puts European public charging utilisation at about 10% in 2025, but that's a 24-hour average, not a constant. Networks running at that average are largely empty most hours and briefly saturated during predictable windows like weekday evening commutes or weekend travel corridors, exactly the moments a blocked bay actually costs someone a session.
Do I need new infrastructure to run occupancy-based idle fees if I already do dynamic energy pricing?
No. If a platform already adjusts energy price by time, location or real-time demand, it already has a live read on session state and connector occupancy per site, the same signal an occupancy-tiered idle fee needs. It's the same underlying question, what's this worth right now, applied to a different part of the session: the kWh versus the minute of occupying a scarce bay after charging finishes.