Solar, EV & Renewables: A Practical Guide to PV Strings and Charge Points

Matching a PV array to an inverter means clearing three genuinely separate checks, each governed by a different design variable. The number of modules wired in series (Ns) determines the string's voltage — too many modules in series and the string voltage can exceed the inverter's absolute maximum DC voltage rating, or fall outside its MPPT operating window. The module's own short-circuit current (Isc) determines the string's design current — independent of how many modules are in series — which has to stay within the inverter's maximum DC input current rating for that MPPT channel. And the total array power relative to the inverter's AC power rating (the DC:AC ratio) is a separate sizing decision entirely. A design can clear any one or two of these checks while failing the third, which is exactly why each is checked independently rather than assuming one 'looks about right' implies the others are fine.

It's counterintuitive at first, but PV module voltage rises as cell temperature falls — silicon PV cells have a well-established negative temperature coefficient for open-circuit voltage. That means the highest string voltage a system will ever see happens on the coldest day it's exposed to, not the hottest — which is exactly why the maximum string voltage check (against the inverter's absolute maximum DC voltage rating) uses the site's coldest expected temperature, while the low end of the MPPT operating window is instead checked at the hottest expected temperature, where voltage sags the most. Checking only one temperature extreme, or checking both against the same temperature, misses exactly the condition each specific check is designed to catch.

It's common and often economically sensible to size a PV array's total DC capacity somewhat above the inverter's AC power rating — a DC:AC ratio above 1.0 — because an inverter only actually 'clips' (loses potential output) during the relatively few hours of peak irradiance when the array would otherwise exceed the inverter's AC rating, while a larger array captures more usable energy during the much larger number of lower-irradiance hours that make up most of a year's generation. Pushed too far, though, oversizing starts trading away increasing amounts of that peak-hour energy to clipping for diminishing additional energy capture elsewhere — which is why DC:AC ratio has a sensible working range (a very rough industry rule of thumb is often cited around 1.1-1.3) rather than an open-ended 'more is better' relationship.

Most electrical design parameters are things you size or derate for. A TN-C-S (PME) earthing system and an EV charge point are a different kind of problem: IEC 60364-7-722 doesn't ask for a derating factor here, it prohibits a PEN (combined protective-and-neutral) conductor from being extended into the final circuit feeding the charge point at all, full stop. The concern is that a PEN conductor fault upstream can put dangerous voltage onto a property's earth terminal — a recognized but generally low-consequence risk for an ordinary socket circuit, but a much more serious one at an EV charge point where a person may be in simultaneous contact with damp ground and a large metal vehicle chassis. A design on a TN-C-S supply has to either convert the final circuit to TN-S, fit open-PEN detection, or use a local TT-style earth electrode — there's no 'derated but acceptable' middle option the way there often is elsewhere in electrical design.

Electric vehicle onboard charging equipment (the rectifier and power electronics that convert incoming AC to the DC the battery actually needs) can, under certain fault conditions, produce a smooth DC residual current — and a standard Type AC or even Type A residual current device (RCD), designed primarily around AC and pulsed-DC residual current, can have its trip mechanism 'blinded' by a sufficient smooth DC component, failing to trip when it should. This is why EV charging circuits need either a Type B RCD (which detects AC, pulsed DC, and smooth DC residual current all in one device) or a Type A RCD combined with a separate residual DC detecting device (RDC-DD, per IEC 62955) that specifically watches for smooth DC current and trips before it can blind the upstream Type A device — together treated as providing equivalent protection to a Type B RCD on its own.

For a single EV charge point, sizing is straightforward — the circuit is sized for that point's own full rated current. At a multi-point site, it's tempting to assume not every charge point will draw full rated current simultaneously and apply some diversity factor to reduce the feeder size — but IEC 60364-7-722 specifically ties that permission to having an actual Load Management System (LMS) in place, one that actively monitors and limits total site demand in real time. Without an LMS, the code requires designing as if every point could draw full rated current at the same time, precisely because there's no active control mechanism actually guaranteeing otherwise — diversity isn't a general statistical convenience here, it's conditional on the control system that makes the diversity assumption trustworthy in the first place.