Not Too Slow, Not Too Expensive: The Goldilocks Zone of Ultrarapid EV Charging
The Osprey hub in Brentwood - The maths suggests this is all the power you'll ever need
One of my dissenting opinions is that improvements in EV charge rate are now irrelevant in practice. I get why people are excited about the technical achievements of wanging electrons into batteries at crazy high C-ratings, in the same way I liked high numbers in categories in Top Trumps as a kid. But when it comes to using En Route Charging, I have two requirements that establish my "Goldilocks Zone":
Get enough range in the time I’m conveniently stopped;
Minimise the cost of the charge event.
The first provides some qualification, but the latter is the kicker.
1. How much charge do we want in an en route stop?
First off, let’s define an “en route stop”. If we’re driving multiple hours, a stop is likely to involve getting some food and drink and dispensing with the residuals of the food and drink you had previously. Plus the benefit of the walk to do each. Do this at a place with suitable facilities, and I’d say 15 mins is probably about the minimum stop, I can probably cope with 30 mins if I’m gonna order food and eat in. NB: edge cases for v short stops discussed below.
Ask a DC operator and you’ll learn that average rapid charge events usually come in around 30kWh, or ~100 miles. That makes sense, it’s enough to cover another ~1.5 hours’ of mixed road driving, without leaving you empty at the end. And since you probably left with enough to cover ~4 hours’ driving, this gets you towards the limit of what most will bother driving vs using alternative transport. At the upper end, I’m sure I have needed to get another 200 miles in a mid-long journey hit, so ~60kWh.
So I want the correct amount of range to be added for my purposes in 10-35 mins, typically around 15 mins for 100 miles, say 30 for 200 miles. I don’t want less, so I have to wait, but I don’t want more, where I am billed at a premium rate for miles I could have put in cheaper at Home/Work/Destination Charging options. Or worse, if you remotely stop your charge and leave the car for anything >5 mins you are likely to (quite reasonably) be billed an overstay fee to encourage you to vacate the charger to give someone else a turn. Sadly finishing too early isn’t just a problem in your private life.
In practice, there’s a limit to how fast you actually want this charge to be. Sure, it could be so quick I just sit in the car and then drive off to park away from the charger, but this is actively less convenient, and - critically - likely much more expensive.
2. How much am I willing to pay?
That’s easy: As little as I can, within the bounds of convenience for my trip.
I am not going to spend ~4-8 hours at a services to charge inexpensively at 7kW. I want to get this done and not make my journey much longer, if longer at all. This means I’m after >100kW peak charging, for sure. So we’re into pricey infrastructure, and this is never going to be super cheap. But that’s fine, I don’t use it much.
To maximise charge rate, we must maximise power in the charger, which means maximising the size of the grid connection. Grid connections are pricey, high powered switch gear is pricey, energy & grid costs for high powered connections are pricey. And the bigger they are the more pricey they get.
A lot of drivers view high DC charging prices with suspicion. Knowing what I do about the challenges and costs they all face in the early market, I think we should be grateful to the firms for taking on the risk at all.
Interestingly, the industry has discovered that within the current rapid charging range (50-350kW), these infrastructure costs dominate so heavily that hardware differences become marginal - which is why UK networks price them identically. But push beyond to genuinely extreme speeds like 1MW, and you'd break into a new tier of infrastructure costs entirely.
Calculating the optimum charge rate for typical en route stops
So, we’ve shown we don’t really want about 100 miles, roughly 30kWh, in about 15 minutes. At most, we'd maybe need twice that, ~200 miles (~60kWh), for a much longer drive, in which case 30 mins is probably fine, but we might want to keep it to 15 mins if poss.
If we’re stopping en route, that implies we’re in the bottom half of our battery’s charging curve we’re likely to accept a good chunk of our EV’s max charge rate. So we should be able to accept a nice >100kW power level for much of our charge.
30 kWh / 0.25 hrs = 120kW average
60kWh / 0.5 hrs = 120kW average
So for typical use, 120kW.
Let’s say we want to get the 200 miles in the 15 minute gap:
60kWh / 0.25 hr = 240kW
For longer journeys, 240kW.
So our Goldilocks Zone emerges: not too slow (under 100kW would leave us waiting), not too expensive (over 350kW would break the infrastructure economics), but just right at 120-240kW.
Happy news! Depending on our EV, we can do that on existing DC chargers, either 150kW or 350kW devices. UK networks price these identically despite 350kW costing more to install - the hardware costs more and you get fewer bays per grid connection. But operators absorb this difference, recognising that, overall, the real-world performance gain is too small to justify complexity in pricing.
Edge cases
Ultra-high mileage drivers
Yes, there are people who feel the need for more power. A neighbour of mine does a ~23 hour drive in one go to Scandinavia and back once a year with a relative taking turns driving (albeit with some ferry action in there somewhere). And the internet is positively HEAVING with grumpy middle aged men who CLAIM they drive insane distances weekly, whilst towing their pet 14 tonne boulder or something (spoiler alert - they really don't, but let's pretend anyway).
But if you're building an infrastructure for mass utilisation, these drivers are so irregular that it is unwise to use them to sway investment decisions. And so some people may have to take a ~15 mins longer break every 4 hours of their journey? Seems like… a good thing, no?
The Tesla exception
And yes, I know the Tesla Supercharger network is way cheaper, as low as 32-36p/kWh for members off-peak, but that's based on a unique business model that allows them to shift cost allocation - using car sales to subsidise charging infrastructure. It's not representative of sustainable standalone infrastructure economics.
A paradigm shift back towards a refuelling model?
Of course, if we can charge at ~1MW, we could totally shift the charging paradigm. ~5 minute charging stops replicates the petrol refuelling model. This could encourage smaller batteries in EVs, making the vehicles cheaper. This model would mean very high utilisation of chargers, amortising hardware and some grid costs. There’s much to like here.
BUT - someone has to show me the path to making charging at this power level remotely affordable, even with said amortisation. Ultrarapid charging is currently up to ~10x more expensive than the most affordable Home Charging. It can be ~2x the price of 7kW Destination Charging. There are moves afoot to try to address some of the underlying factors that cause this deficit - grid cost structures, VAT discrepancy vs Home Charging etc. But I can’t see how increasing the power ~3x is going to make it more affordable than today’s 350kW chargers, it is likely it would actually be more expensive, potentially >£1/kWh.
At which point, who is going to want to use it? In which case, who is going to fund its rollout?
And on the smaller batteries and higher charging rates case, higher C-ratings allowing e.g. 30kWh city cars to charge at 350kW is just fine thanks. I don’t need it so much at the top end of battery sizes.
Trucks, HGVs, heavy plant vehicles
There is a place where MW charging speeds are actually essential, i.e. for big, high utilisation, heavy vehicles. If we want to electrify long distance trucks then we will need to provide some very high powered charging. These vehicles consume ~2-3kWh per mile compared to a car's ~0.3kWh, meaning they need ~200-300kWh to cover the same distance as a car would on 30kWh. Their legally mandated 45-minute driver breaks must accommodate full charging sessions, not just top-ups. This isn't an edge case - it's the core use case for MW charging. I totally agree we need this infrastructure.
But I suggest we save these mammoth connections for mammoth vehicles, rather than overpowering car charging infrastructure for a problem that doesn't exist outside the fevered imagination of middle aged, diesel driving chaps on the internet.
Conclusion
There is currently a renewed focus in the EV world on ever faster charging times being achieved by EVs, notably (but not exclusively) in/from China. And these should be celebrated as the technical feats that they are.
But here's the crucial distinction: faster charging isn't free progress. Unlike semiconductors or battery energy density/cost reductions where we get more for less, every jump in charging speed demands real infrastructure costs - beefier grid connections, expensive switchgear, active cooling. The constraint has long been on the grid side, not the vehicle. These costs land directly on your charging price.
The market has already spoken: UK networks don't even bother pricing 350kW differently from 150kW because the real-world benefit is negligible. Yet push to 1MW and watch that ~70p/kWh become £1+/kWh. For what - saving 10 minutes you probably wanted for coffee anyway? Most drivers will take the cheaper rate and the free sandwich on net, thanks.
Some argue ultra-fast charging might reassure petrol drivers that EV refuelling can match their expectations. Perhaps. But - check the company name! - I remain deeply sceptical of its mass market value to actual EV drivers. We've already found our Goldilocks Zone. It's 120-240kW, it works, and anything faster is solving a problem that doesn't really exist where it counts - in motorway service areas.
P.S. Autonomous Electric Vehicles (AEVs)
Some may raise the oft purported future of vast AEV fleets seeing to our every mobility need, removing the human frailties discussed above. And, while I have my own dissenting views on those too (one for another article), it’s a reasonable point for discussion.
In this scenario the aim of the AEV fleet operator will be to minimise the time taken to charge in order to optimise utilisation of their asset. So it might be worth paying more for an uber quick charge vs the loss of potential earnings from the AEV. But even here, there’s a sweet spot, even if we consider peak demand only. You will still have a point where the additional cost (i.e. above the rate you could get charging slowly overnight during low demand) to charge gets higher per minute than the potential earnings loss from being stopped. Maybe that’s higher than 150kW, or 350kW, but there will be an upper bound, and you will program your fleet to avoid charging above that price.
