Ask most public charging operators what’s holding back their growth in 2026 and the answer used to be permits, hardware lead times, or finding good sites. It isn’t anymore. According to Driivz’s 2025 State of EV Charging Network Operators Report, 46% of operators now name energy constraints as their single biggest challenge, and more than 90% expect grid capacity to limit growth over the next twelve months. Every operator surveyed expects some impact from it: 82% anticipate moderate constraints, 10% expect significant ones.
That’s a different problem than the ones the industry has spent years solving. You can install a charger faster than you can get a grid connection upgraded.
Why this changes the planning conversation
For a while, expanding a charging site meant asking how many chargers fit on the concrete. Now the real question is how much load the existing grid connection can carry before it needs a physical upgrade, and a physical upgrade means a queue: a request to the network operator, an engineering assessment, sometimes a transformer swap, and a wait that doesn’t move on your schedule no matter how urgent your business case is.
That reshapes what growth looks like for an operator. Adding chargers to a site that’s already near its connection limit doesn’t add capacity, it just means more hardware competing for the same finite amount of power. Two DC chargers sharing a connection that can only support one running at full power isn’t a growth story, it’s a queue with extra steps.
The expensive fix, and the cheap one
There are really only two ways to deal with a connection that’s running out of headroom. Pay for more capacity, or use the capacity you already have more deliberately.
Paying for more capacity is the obvious answer and sometimes the only real one. It’s also slow and capital-intensive, and it doesn’t help you this quarter while you wait for it.
Using existing capacity more deliberately is the cheaper, faster lever, and it’s mostly a pricing and scheduling problem rather than an engineering one. If a site’s connection can comfortably support three simultaneous full-power sessions but struggles at five, the goal isn’t to block the fourth and fifth car. It’s to make plugging in at that moment less attractive than plugging in twenty minutes later, without anyone noticing they were nudged.
Illustrative: price rises as a site’s occupancy approaches its connection ceiling, and eases off once headroom opens up again. Same demand-aware logic as in our full breakdown of the five pricing models, applied here to physical grid headroom instead of occupancy for revenue.
What load-aware pricing actually does here
This is a different job for dynamic pricing than the one most operators think about first. It isn’t about charging more when demand is high to capture margin. It’s about using price as a signal that keeps a site’s actual power draw inside what the connection can handle, so the connection itself doesn’t become the reason growth stalls.
The mechanics look like the demand-shifting logic operators already use for demand charges: price rises as a site approaches its load ceiling, and eases off once headroom opens up again. The difference is what triggers it. A demand-charge model reacts to a monthly billing peak. A connection-aware model reacts to the physical limit of what’s plugged into the wall right now, in something close to real time.
Done well, this buys an operator real runway. It doesn’t replace a grid upgrade where one is genuinely needed, but it can be the difference between adding two more chargers to a site this year and having a project stall in a network operator’s queue for the next one.
Why this is worth planning for now, not later
Grid capacity constraints aren’t a temporary side effect of a busy quarter. They’re the predictable result of charging networks scaling faster than the grid infrastructure underneath them, and that gap tends to widen before it narrows. An operator whose pricing can already shape load on a per-connection basis has a tool the operators still waiting on their next transformer upgrade don’t.
We wrote just yesterday about why a tariff that just tracks the wholesale price belongs in a membership discount rather than the base tariff. This is the same starting point from a different angle: price isn’t only a revenue tool, it’s also a tool for managing physical limits, as long as you point it at the right problem.
If growth at your sites is starting to bump into what the connection can carry rather than what the market wants, pricing is worth a look before the upgrade request goes in, not after.
If you run or price public chargers and want to compare notes on how you’re managing connection headroom today, happy to talk operator to operator.
Frequently asked questions
Why is grid capacity now the biggest challenge for charging operators?
According to Driivz’s 2025 State of EV Charging Network Operators Report, 46% of operators name energy constraints their biggest challenge, and more than 90% expect grid capacity to limit growth over the next year. Charging networks are scaling faster than the grid infrastructure underneath them, so the connection itself is increasingly the first limit, ahead of market demand.
Why does a grid connection upgrade take so long?
A physical upgrade requires a request to the network operator, an engineering assessment, and sometimes a transformer swap. That timeline depends on the network operator’s schedule, not the operator’s urgency, and can delay an expansion for months.
How does pricing help with a constrained grid connection?
By using price as a signal that spreads sessions out instead of concentrating them, a site’s actual power draw stays inside what the connection can handle. Price rises slightly as the site approaches its load ceiling and eases off once headroom opens up, so nobody needs to be blocked outright.
Does this replace a real grid connection upgrade?
No. Where an upgrade is genuinely needed, it’s still needed. Demand-aware pricing buys time and headroom within the existing connection, and can be the difference between growing this year and having a project stall in the network operator’s queue.