TN-S at the service head… PME in the street? Why both can be “true” (and when it matters)

You check the service head and it looks like TN-S: earth taken from the sheath/earth terminal, no PME label, nothing obvious.

Then you’re asked to supply a metal outbuilding, outdoor equipment, or fit an EV charger — and you hear:

“Treat it as TN-C-S (PME).”

That can feel like a contradiction. It isn’t, once you separate two ideas:

If you’re in Colchester or North Essex and need help with earthing, outbuildings, or EV charger considerations, bookings are via Gray Logic Electrical.

• What the supply presents as at the property (what you can actually see and test at the origin)
• What the upstream network might behave like under a rare but important condition: loss of the neutral/PEN reference

This post explains why that distinction exists, why it often doesn’t change a normal EICR outcome, and why it does change how we treat exported earth outdoors.

1) Why “TN-S at the intake” doesn’t always mean “TN-S in the street”

Distribution networks evolve. Older areas were built one way, then repaired, diverted, reinforced, extended — sometimes many times over decades.

A very common outcome is a hybrid network: parts of the LV system are one arrangement, other parts have been altered using modern methods (often PME-enabled sections). Your cut-out can still present TN-S at the property even if upstream sections are no longer “textbook TN-S all the way back”.

That’s the entire point: The intake can present TN-S, while the upstream network — in a reference-loss scenario — behaves closer to PME/hybrid.

(If you want the formal wording and how DNOs describe this, see the TECH NOTE panel.)

2) Why DNOs and EV guidance often say “treat it as TN-C-S”

This isn’t “because the DNO couldn’t be bothered” — it’s because you can’t guarantee the upstream topology and it can change in the future.

For normal indoor circuits, that distinction rarely changes what we do (because we’re verifying the installation’s protective measures and condition). But for outdoor touch risk, it becomes much more important.

EV charging is the clearest example, because it is designed around outdoor simultaneous contact (vehicle body + ground), and BS 7671 section 722 is built on that reality.

3) The EICR confusion: “It looks TN-S, Ze is 0.55 Ω… so why am I being told PME?”

This is the scenario that triggers online arguments:

• Intake presents as TN-S
• Ze measures something like 0.55 Ω
• Someone says: “If we treat it as PME, shouldn’t that be ‘too high’?”

The practical answer is:

• An EICR is a verification report. We record what’s observed at the origin, we measure and verify, and we code based on risk and evidence.
• “Treat it as PME” is usually design caution for outdoor touch risk, not a demand to rewrite the intake label on every certificate.

So the sensible position is often: Record what you can evidence at the origin for the purposes of inspection and testing, but design/advise conservatively where outdoor touch risk exists.

4) Where it does matter: exported earth to outdoor metalwork

This is the key point most people miss.

Indoors: you usually have an equipotential zone

Bonding + CPCs are intended to reduce dangerous touch voltages between things you can touch at the same time.

Outdoors: you can lose that protection

When you export earth to a metal outbuilding, metal-clad shed, steel frame workshop, or outdoor equipment, you create a realistic scenario where someone can touch:

• metalwork connected to the installation’s earthing system and
• true earth (wet ground, concrete, barefoot conditions)

That’s where a “reference loss” type event matters.

And this is why you’ll hear: “Treat it like PME.”

Not because “TN-S is fake”, but because the consequence outdoors is higher and the uncertainty upstream is outside the installer’s control.

5) “If the neutral/PEN drops, does all metalwork become live?”

In a PME/hybrid reference-loss scenario, the installation earthing system can rise in potential relative to true earth.

Two practical clarifications:

• Indoors, with correct bonding, the worst “touch between two different references” scenarios are less likely because simultaneously accessible metalwork is kept closer together (potential-wise).
• Outdoors, the ground is the reference. If you can touch exported/bonded metalwork while standing on true earth, the touch voltage risk becomes more meaningful.

So the scary sentence (“all metal becomes live”) is directionally describing the same mechanism, but the real hazard is: touching outdoor metalwork referenced to the installation earth while you are referenced to true earth.

6) If it’s so serious, why don’t we see people dying all the time?

Because you usually need a stack of conditions at once:

• a reference-loss event upstream
• meaningful loading/current at that moment
• a person in a high-exposure contact scenario (often outdoors)
• low enough body/ground resistance to permit dangerous current
• enough time for harm to occur

Serious outcomes appear rare in domestic settings.

Illustrative “1 in 300 million” (hypothetical, not an official statistic)

You’ll sometimes see people quote a number like “1 in 300 million”.

There isn’t a neat public dataset that isolates fatalities specifically attributable to open-PEN/reference-loss as a clean official count, so any probability number needs to be clearly labelled as illustrative.

Here’s what an illustrative calculation looks like, just to show why you can end up with a headline that small:

• Assume (hypothetically) 2 fatalities in 10 years attributable specifically to open-PEN/reference-loss conditions
• Assume a population of ~70,000,000
• Person-years ≈ 70,000,000 × 10 = 700,000,000
• Risk ≈ 2 / 700,000,000 ≈ 2.86×10⁻⁹ per person-year
• That’s roughly 1 in 350 million person-years (i.e. “on the order of 1 in 300 million”)

Again: the weak point is the fatality count — without a published dataset that isolates those cases, this stays illustrative.

7) Best practice: what we actually do when outdoor metalwork is involved

This is where I stop arguing about labels and start controlling risk.

A) Separate “EICR recording” from “outdoor design caution”

• EICR / verification: record what’s observed at the origin, measure and verify protective measures and bonding, code based on risk and evidence.
• Outdoor work: treat it as higher-consequence and choose protective measures accordingly.

B) For EV charging: follow Section 722

EV charging has explicit requirements because it’s an outdoor touch-risk scenario. The right solution is the one that is compliant for that specific installation.

C) For metal outbuildings/exported earth: design for the scenario

The right approach depends on the site, but the decision is driven by questions like:

• Is the outbuilding metal clad / steel frame / significant touchable metal?
• Are people likely to be well-earthed (wet ground, bare feet, concrete slab)?
• Are you exporting earth in a way that creates unintended parallel paths?
• Is the protective arrangement robust if the supply reference is lost upstream?

There isn’t one universal wiring diagram that suits every outbuilding — which is why good electricians look at the site and document the reasoning.

D) If something genuinely doesn’t add up, that’s when enquiry is justified

If readings are unstable, behaviour is odd, or the intake arrangement appears unsafe or inconsistent, enquiry becomes justified and the proper information route exists under ESQCR.

The bottom line

• “TN-S” at the service head can be a fair description of the intake and the upstream network can still be hybrid/PME-enabled in a reference-loss scenario.
• This often doesn’t change a typical EICR outcome.
• Where it becomes a genuine safety/design issue is exported earth outdoors, especially metal outbuildings and EV charging.