What G5 covers
G5 Issue 5 was published in 2020. It deals with harmonic voltage distortion — the extent to which equipment connected to the network draws non-sinusoidal current and, in doing so, distorts the voltage waveform for everyone else on that part of the system.
Harmonics come from anything that rectifies or switches: variable-speed drives, EV chargers, inverters, UPS systems, LED lighting at scale, induction heating, arc furnaces. The current a six-pulse drive draws is nothing like a sine wave, and pushed through the source impedance of the network that current produces a distorted voltage.
G5 does not cover voltage fluctuation or flicker — that is P28 — nor voltage unbalance, which is P29. The three standards deal with three distinct ways an installation can degrade supply quality for its neighbours.
Who needs a G5 assessment
In practice, anyone applying to a distribution network operator to connect equipment with significant power-electronic content. The DNO will ask for a harmonic assessment as part of the connection application, and the form that assessment takes depends on how large and how disturbing the equipment is.
The good news is that the standard is deliberately staged so most connections never need a full study. Small equipment that conforms to the relevant IEC product standard clears Stage 1 with little more than a declaration. It is only when equipment is large relative to the strength of the supply that the assessment gets involved.
The one number that decides everything
Almost every G5 assessment turns on the ratio between your equipment's rating and the short-circuit power at the point of common coupling. A strong supply absorbs harmonic current with little voltage distortion; a weak one does not. Get Ssc from your DNO before you start — everything else follows from it.
Planning and compatibility levels
G5 works with the same three-tier EMC concept P28 uses. Compatibility levels are system-wide reference values. Planning levels are what the network operator assesses a new connection against, and sit at or below the compatibility levels. Emission limits are what an individual installation may contribute, derived from the planning level after allowing for the distortion already present.
Planning levels are set per harmonic order and per voltage band (Clause 5.2), because the acceptable distortion at 11 kV is not the acceptable distortion at LV, and the fifth harmonic is not the thirty-seventh. There is also a limit on total harmonic distortion alongside the individual-order limits.
Clause 5.4 adds subharmonic and interharmonic distortion — frequency components that are not integer multiples of 50 Hz — with their own limits.
The assessment stages
G5 uses a staged process, summarised in Clause 6.4. Each stage is more work than the last, and you only descend to the next when the current one does not clear you.
| Stage | What it is | Who does it |
|---|---|---|
| Stage 1 | Screening against product-standard conformance and tabulated equipment allowances, scaled to the short-circuit power at your connection point. | The applicant |
| Stage 2 | Calculation of your installation's harmonic contribution against a reference short-circuit power, then against the actual network, using per-harmonic emission data. | The applicant |
| Stage 3 | A detailed engineering study using measured background distortion and a network model, including resonance. | The network operator |
Stage 1 in detail
Stage 1 has four substages, and which applies depends on your equipment (Clause 7.1).
- Stage 1A — equipment conforming to IEC 61000-3-2. This covers small equipment drawing up to 16 A per phase. Conformance is the assessment.
- Stage 1B — equipment conforming to IEC 61000-3-12, which covers 16–75 A per phase. The standard derives a minimum short-circuit power your connection must have for that conformance to hold. If your actual Ssc exceeds it, you pass.
- Stage 1C — for equipment beyond the product standards, the standard tabulates how much total equipment capacity is permitted at a 10 MVA reference short-circuit power, by technology type, then scales that to your actual supply. There is a separate route for installations mixing technologies.
- Stage 1D — the same idea as 1C but using measured harmonic headroom at your connection point rather than the standard's default assumptions, which generally permits more equipment.
Stage 2 in detail
Stage 2 is where most non-trivial industrial connections end up. It has three substages, working from a standard reference case toward your specific network.
- Stage 2A — tabulated permitted equipment capacity at a 60 MVA reference short-circuit power, by technology, scaled to your connection.
- Stage 2B — assessment using your equipment's declared per-harmonic current emissions rather than a technology category.
- Stage 2C — the full calculation: your per-harmonic emissions against the network impedance at your connection point, compared with the planning levels after allowing for existing background distortion.
Stage 2C is the one that needs real data — a per-harmonic current spectrum for every item of equipment, the network impedance across the harmonic range, and the background distortion already present. It is also the stage that produces the numbers a DNO will actually scrutinise.
Aggregation and diversity
Where several items emit at the same harmonic order, Clause 6.3.3 permits a non-linear aggregation if you can declare operational and locational diversity — the emissions will not all peak together. Without that declaration the standard requires linear addition, which is materially more conservative. Worth understanding before you assume your installation fails.
Stage 3
Stage 3 is a detailed engineering study, carried out by the network operator for a specific connection. It uses measured background distortion and a full network model, and it is where resonance is assessed — the case where network capacitance and inductance combine to amplify a particular harmonic far beyond what a simple impedance calculation would suggest.
Reaching Stage 3 is not a failure. It means your connection is large or unusual enough that the simplified methods cannot fairly assess it. It does mean time and cost, which is why it is worth getting Stages 1 and 2 right first.
Equipment technology matters
G5 treats different converter topologies very differently, because they emit very different harmonic spectra. The categories that recur through the tables are six-pulse rectifiers, twelve-pulse rectifiers, active front ends, single-phase rectifiers, and — added in Issue 5 — soft starters and load controllers conforming to EN 60947-4-2 or -4-3.
The differences are large. An active front end is permitted many times the capacity of a six-pulse rectifier at the same connection point, because it emits far less low-order harmonic current. If your assessment is marginal, the equipment specification is often a more productive place to look than the network.
How G5 and P28 fit together
They answer different questions about the same connection, and a DNO application usually needs both.
P28 asks whether your equipment will make the voltage move — dip when it switches on, flicker when it cycles. G5 asks whether it will distort the waveform. A variable-speed drive can pass P28 comfortably while failing G5, or the reverse: a large direct-on-line motor barely troubles G5 but can breach P28's step voltage limit on its first start.
The practical order is to run the P28 check first, because it is quick and free, then do the harmonic work. If P28 already tells you the connection needs a Stage 2 assessment, the harmonic side almost certainly does too.
What this looks like in practice
For a straightforward connection — a few drives on a reasonably strong supply — Stage 1 may clear you on a declaration and a short-circuit power figure. For a substantial industrial installation, expect to assemble a per-harmonic emission spectrum for every converter, agree the network impedance with your DNO, and work through Stage 2C.
The work is not conceptually difficult, but it is unforgiving: the calculations are repetitive, the tables are numerous, and a units slip propagates silently through to a wrong conclusion. That is precisely why we built a calculator for it rather than maintaining a spreadsheet.
Where this page stops
This is an orientation, not the standard. It paraphrases and cites; it does not reproduce G5's text, tables or equations. For anything you are going to submit, work from your own copy of ENA EREC G5 Issue 5 (2020), available from the Energy Networks Association.