6S Teknoloji · 09.11.2026 · 6 min read
Adding a few electric vehicle charging stations to a car park looks simple on paper; the difficulty starts in the hours when the stations run at the same time. Every AC charging point takes a share of the existing connection capacity of the building, and that demand can coincide with hours when the building load is already high. Load management is the way to keep total charging power within the connection capacity. The common language between the stations and the central software (CSMS, Charging Station Management System) is OCPP, the Open Charge Point Protocol. This article covers the OCPP versions, how charging profiles work and the errors met most often in the field.
At most sites, the sum of the rated power of all stations is larger than what the connection point can carry. Without load management two options remain: size the connection for the worst case, or accept the risk of the main breaker tripping. A managed system instead calculates the current available for charging in real time. For a three-phase connection the basic relationship is:
I_charge = (S_connection − S_building − S_margin) / (√3 × U_LL)
Here S_connection is the permitted apparent power, S_building is the instantaneous load excluding charging, S_margin is a safety margin for measurement delay, and U_LL is the line-to-line voltage. The resulting current is shared among active charging sessions. There is a physical lower limit: under IEC 61851-1 the lowest charging current that can be signalled through the control pilot is 6 A. If I_charge / N drops below that value, some sessions should be queued instead of splitting the current further.
In both versions the station initiates the connection: it opens an outbound WebSocket to the CSMS address and announces the subprotocol it wants during the handshake. As a result, no inbound port has to be opened on the network where the station sits.
| Topic | OCPP 1.6-J | OCPP 2.0.1 |
|---|---|---|
| Subprotocol | ocpp1.6 | ocpp2.0.1 |
| Session messages | StartTransaction / StopTransaction | TransactionEvent (Started, Updated, Ended) |
| Remote start | RemoteStartTransaction | RequestStartTransaction |
| Addressing | connectorId | evseId and connectorId |
| Configuration | GetConfiguration / ChangeConfiguration keys | Device model, GetVariables / SetVariables |
| Security | Security extension added later | Security profiles built into the standard |
The first security profile (Security Profile 1) is HTTP Basic authentication without TLS; the second performs the same authentication over TLS, and the third uses a client certificate on the station side. An unencrypted connection is acceptable only on a trusted, closed network.
In OCPP a power or current limit is sent with the SetChargingProfile message. A profile is a schedule stored on the station, and it keeps being applied when the connection drops. To understand a profile, look at these fields:
ChargePointMaxProfile (ChargingStationMaxProfile in 2.0.1) caps the whole station; TxDefaultProfile is the default for new sessions; TxProfile applies only to one specific active session.chargingRateUnit can be amperes (A) or watts (W). A limit in amperes applies per phase.chargingSchedulePeriod list defines the limit that applies from the startPeriod second onward.The station applies the most restrictive of the profiles valid at the same time. A GetCompositeSchedule request is the most reliable way to see which limit is actually in effect.
Static load management gives each station or group of stations a fixed upper limit. It is easy to implement and needs no extra metering, but because it is based on the moment of peak building load, most of the spare capacity stays unused. Dynamic load management reads the measurement at the connection point continuously, recalculates the available current with the formula above and updates the profiles periodically.
The reliability of a dynamic system depends on three things: frequent and uninterrupted building measurements, confirmation that the station accepted each profile update, and a safe default limit that stays on the station when communication fails. Vehicles may not follow a new limit instantly, so the safety margin should be wide enough to cover measurement and response delays. A vehicle charging on a single phase loads only one phase of a three-phase station; tracking current per phase makes that imbalance visible.
%20; if the server does not decode it, the station never matches its registration.Energy.Active.Import.Register). To see power, current and voltage, the sampled measurand list must be configured on the station.Rejected response.On the server side, 6S Teknoloji includes a CSMS service that speaks OCPP 1.6-J and 2.0.1; the station opens an outbound WebSocket and the matching version is selected from the subprotocol. Stations that are not registered and active on the platform, and stations that fail Security Profile 1 (HTTP Basic) authentication, are rejected at connection time; the password is stored as a bcrypt hash, and when a certificate is configured the connection runs over TLS (wss). Energy, active and reactive power, power factor, frequency, and per-phase voltage and current from MeterValues are unit-normalized and written to the electrical readings of the platform; unrecognized measurands are kept in a separate field. Charging sessions are recorded, and in OCPP 1.6 session energy is calculated from the difference between the start and stop meter values. The last known state of a reconnecting station is preserved.
On the load side, 6S Teknoloji today sends a fixed upper limit in amperes to a station through an API that requires administrator rights: a SetChargingProfile carrying a single-period, Absolute TxDefaultProfile. Remote start and stop commands travel the same path. Dynamic load balancing that updates the limit automatically based on building or transformer load is not part of 6S Teknoloji; the limit value is the one chosen by the operations team.
Related Content
Explore the system with your own data in a demo session.