A DC charger presents itself to the grid as a large non-linear load. It does not draw current in proportion to the voltage waveform; it switches, rectifies and regulates, and the current it pulls has a shape of its own. Everything that follows in this subject is a consequence of that shape.
The vocabulary around it is small but frequently misused. Total harmonic distortion and total harmonic distortion of current are different measurements of different things. Power factor has two definitions that give different numbers for the same equipment. Knowing which is which turns a datasheet line item into a specification you can actually verify.
Why a Rectifier Draws a Non-Sinusoidal Current
A simple rectifier conducts only when the incoming voltage exceeds the voltage on its DC side. Current flows in short pulses around the peaks of the supply waveform rather than smoothly across the whole cycle. Those pulses contain the energy the load needs, and they also contain frequencies that are multiples of the supply frequency. Those multiples are the harmonics.
Modern chargers add a correction stage in front of the rectifier. It shapes the input current to follow the voltage waveform, which reduces the harmonics substantially. This is why the subject matters less than it did for older equipment and why it still matters: the correction is imperfect, and it has to hold across the whole load range rather than only at the design point.
Where correction is poorly executed or defeated at light load, the harmonics return. A site running many units at low output can therefore show worse current distortion than the same site running fewer units at high output, which is counter-intuitive and frequently missed in commissioning.
THD and THDi Are Not the Same Number
Total harmonic distortion, written without a subscript, usually refers to voltage. It describes how far the supply voltage waveform has departed from a pure sine. Total harmonic distortion of current, written with an i, describes the shape of the current the equipment draws.
The two are related but not interchangeable, and the relationship depends on the impedance of the supply. A charger with high current distortion connected to a stiff supply produces only a small voltage distortion. The same charger on a weak supply, at the end of a long cable, produces a much larger one. This is why a product can meet its current distortion figure and still cause a voltage problem on a particular site.
The practical rule is that the equipment specification describes what the equipment does, and the site survey describes what the site experiences. A quotation that lists only one of the two numbers is answering half the question.
Two Definitions of Power Factor
Power factor is real power divided by apparent power. When both voltage and current are sinusoidal, that ratio equals the cosine of the phase angle between them, and the two definitions coincide. When the current is distorted, they diverge.
Displacement power factor considers only the fundamental components and ignores distortion. True power factor, sometimes called total power factor, includes the distortion and is therefore lower. A unit can report a displacement power factor very close to unity while its true power factor is meaningfully lower, and the difference is carried by harmonic currents that still have to flow through the supply.
This distinction matters at the meter and at the transformer. The utility bills on true power factor, whether or not it names it that way. A specification sheet quoting displacement power factor is not wrong, but it is answering a different question from the one the bill asks.
What Harmonics Do to Transformers and Neutral Conductors
Harmonic currents produce additional losses in a transformer beyond the ones it was designed for. Eddy currents in the core and skin effect in the windings both increase with frequency, so a distorted current heats a transformer more than a sinusoidal current of the same root-mean-square value. The practical consequence is derating: a transformer serving a heavily distorting load cannot deliver its nameplate capacity without exceeding its temperature limits.
The neutral conductor is the second problem and the more surprising one. In a balanced three-phase system with linear loads, the fundamental currents cancel in the neutral and it carries almost nothing. Triplen harmonics — the third, ninth and fifteenth — do not cancel. They add in the neutral, which can end up carrying more current than any single phase conductor was expected to handle.
On a charger site this is rarely the binding constraint, because chargers are usually three-phase loads with effective correction. It becomes relevant when single-phase AC points are installed in quantity on a shared distribution board, which is exactly what workplace and residential car parks do.
The third effect is on the supply voltage itself, and through it on everything else on site. Distorted voltage causes additional heating in motors, can disturb sensitive equipment and can cause protective relays to operate unexpectedly. It is the reason a neighbouring tenant can complain about a charger installation that appears to be working correctly.
When a Site Should Re-Measure
A measurement taken at commissioning describes the site as it was built. Five situations justify taking another one.
When chargers are added, because the site's distortion scales with the number of rectifiers connected. When the supply is changed, because a different transformer or a longer cable run changes the impedance and therefore the voltage distortion for the same current distortion. When other equipment on site begins to misbehave, because that is the symptom the building notices first. When the utility raises a question at the connection point, which usually means a measurement is being requested rather than offered. And when a site's utilisation pattern shifts, since light-load operation can produce different distortion than the design case.
The measurement itself has a requirement that is often overlooked: it has to be taken at a known load. A reading taken with one charger running at low output says almost nothing about the site at peak. The useful record states the number of units in session, their combined output and the point in the installation where the instrument was connected.
What to Put in a Specification
Three details convert a marketing figure into a verifiable requirement.
First, state the load point. Distortion and power factor both vary with output, so a figure without a stated load is not a specification. Second, state which power factor definition is meant, displacement or true. Third, state whether the current distortion figure is measured at the terminals of a single unit or at the site's point of connection, because those are different measurements with different values.
Products in this class typically quote their figures at rated load, which is the condition under which the correction stage is working hardest and the numbers look best. A DC unit quoting a power factor at or above 0.99 and a current distortion at or below five percent at full load is describing a well-corrected design; the 120 and 180 kW stations and the 480 kW hub host both quote their figures at full load for that reason.
The question worth adding to a tender is what happens below full load. A correction stage that holds its performance across the range is a different design from one that meets the number at the top and degrades underneath, and the difference shows up on sites where vehicles spend most of a session at partial power.
The equipment figure describes the equipment. The site survey describes the site. A power quality argument that quotes one as if it were the other is where most of these discussions go wrong.
For most installations none of this becomes a problem. Correction stages are effective, and a well-designed site with adequate transformer capacity absorbs the residual distortion without difficulty. The exceptions are concentrated in three places: large sites with many units on one transformer, sites with substantial single-phase AC load on shared boards, and sites at the end of a long or weak supply where the impedance amplifies what the chargers produce.
Those three cases are worth identifying before the design is fixed, because the remedies — a larger transformer, active filtering, a different distribution arrangement — are all cheaper to plan than to retrofit. The measurement protocol and the figures to request are summarised in the technical documentation on the download page.
Asked to justify power quality at a connection point?
Send the transformer rating, the number of units planned and the length of the run. Which of the three effects is likely to bind follows from those three.
Ask a specification question