Two distinct architectures in high-power DC charging
High-power DC charging infrastructure can be configured in two fundamentally different ways. In the first — the distributed architecture — each physical charger unit is a self-contained assembly containing its own AC-to-DC conversion electronics, connected to the building electrical system via a conventional AC supply circuit. The AC cable is sized and designed in the same way as any other AC load, and the scope boundary is clear: the electrical designer is responsible for the AC supply from the distribution board to the charger input terminals.
The second architecture — increasingly common at high-power sites — separates the AC-to-DC conversion into a central power unit (sometimes called a power conversion system or rectifier stack) connected to multiple satellite dispenser units via DC cables. The satellite units contain the user interface, the cable management, and the vehicle connector, but do no power conversion themselves. All the energy flows from the power unit, across the DC bus, to each satellite. This arrangement offers flexibility — power can be dynamically shared between satellites based on demand — but it introduces a category of cable that sits outside the conventional scope of an AC electrical design.
The sizing challenge: what the nameplate doesn't tell you
In a centralised architecture, sizing the DC cables between the power unit and each satellite requires more than dividing the nameplate output power by the DC bus voltage to obtain a current. Several factors modify the effective current demand.
Load diversity. A power unit rated at, say, 600 kW may feed six satellites each rated at 150 kW — a total nameplate demand of 900 kW. The product is designed on the assumption that not all satellites will be at maximum output simultaneously, and the power unit's maximum aggregate output is constrained to 600 kW. The DC cables from the power unit to each satellite must be rated for the maximum current that satellite can draw, but the supply-side cables to the power unit need only be rated for the aggregate maximum — typically a lower figure than the sum of satellite ratings.
Duty cycle and cyclic rating. Rapid charging sessions have a defined duration. A vehicle is unlikely to draw maximum power continuously for the full cable rating period. IEC 60287-2-1 provides a method for calculating cyclic current ratings, which can legitimately exceed the continuous rating where the load profile includes periods of reduced or zero loading. Applying this method where it is appropriate can allow a more economic cable selection — but only where the load profile is well characterised and the thermal mass of the cable and surrounding installation is sufficient to absorb the cyclic heating.
Voltage drop. At the high direct currents involved in DC satellite cable runs — 500 A or more is not unusual — even modest cable resistances produce significant voltage drop. Most DC chargers have a specified minimum input voltage below which they will reduce power output or shut down. Long cable runs may require uprating the conductor cross-section beyond the thermal requirement to maintain acceptable voltage at the satellite input.
Manufacturer design responsibility — a scope boundary that matters
For most proprietary centralised DC charging systems, the manufacturer holds design responsibility for the DC satellite cables — they specify the permissible conductor cross-sections, the maximum cable run lengths, the permitted installation methods, and in some cases they supply the cables as part of the system package. The consultant's scope is typically limited to the AC supply from the building distribution system to the power unit input terminals.
This scope boundary is not always clearly communicated at project outset, and it can create problems in both directions. A consultant who attempts to size the DC cables without reference to the manufacturer's requirements may produce a design that is technically reasonable but incompatible with the manufacturer's system — voiding warranty or creating interface issues on site. Conversely, a project where the consultant assumes the manufacturer will handle the DC cable design, and the manufacturer assumes the consultant will do so, ends up with nobody having designed those cables at all — a situation that tends to be discovered at the point of installation.
The correct approach is to confirm the scope boundary explicitly and in writing at the earliest opportunity — before detailed design begins. The manufacturer should be asked to provide their cable specification, maximum run lengths, and any constraints on installation method. Where the manufacturer provides this information and confirms design responsibility, the consultant should document the interface clearly in the design file. Where the manufacturer declines design responsibility or provides insufficient information, the consultant will need to perform a full design — using IEC 60287 analysis and detailed voltage drop calculations — based on the best available system data.