The term "smart charging" covers everything from a timer to a grid-services aggregation product. For an operator with a sanctioned load and more connectors than it comfortably supports, it means one specific thing: instructing chargers to draw less than their rating so that the total stays within a limit.
What OCPP gives you
A charging profile is a schedule of limits, expressed in amps or watts, with a purpose and a stack level. The purposes matter because they compose:
| Purpose | Applies to | Typical use |
|---|---|---|
| ChargePointMaxProfile | The whole charge point | A hard ceiling the site must never exceed |
| TxDefaultProfile | Every future transaction | The default limit a connector runs at |
| TxProfile | One specific transaction | Slowing or boosting an individual session |
Stack level resolves conflicts: higher level wins. The composite schedule — what the charger believes it will actually enforce over the next period — can be requested back, and requesting it is the only way to verify that your instruction was understood rather than merely accepted.
Static versus dynamic
Static allocation divides the available capacity by the number of connectors and sets a fixed default. It is simple, it never over-draws, and it wastes a great deal of capacity — with eight connectors and two in use, six shares of the supply sit idle.
Dynamic allocation redistributes among active sessions as vehicles arrive and leave. It uses the connection far better and introduces two problems: the control loop has to be fast enough to matter and slow enough not to oscillate, and it has to fail safe when a charger stops responding.
- Reallocate on session start and stop, not on a fixed short timer. Vehicles do not change their draw that often.
- Rate-limit changes to an individual session. A limit that moves every thirty seconds confuses some vehicles into stopping.
- On losing contact with a charger, assume it is drawing its full profile until told otherwise. Optimism here trips the main breaker.
- Respect a per-connector minimum. Below roughly six amps most vehicles simply stop, and a session that stops is worse than a slow one.
Where it stops working
Three limits are worth knowing before designing around this. First, the vehicle decides: you can instruct a lower ceiling, but you cannot make a car draw more than it wants. Second, many chargers apply changes only at coarse granularity or only at session start, whatever the profile says. Third, on a site with a real physical constraint, the platform is not the last line of defence — the protection is, and it must be sized as if load management does not exist.
The economics
The reason to do any of this is the sanction. Raising sanctioned load is expensive and slow; demand charges on an upgraded connection are permanent. Load management converts that capital problem into a software one, and on a site with an overnight duty cycle it frequently makes a second and third charger essentially free of grid cost.