LSZH vs PVC: Which Sheath Should You Specify?

The real difference is in what comes off when it burns

PVC is roughly half chlorine by mass. That chlorine is what makes PVC naturally flame retardant, and it is also the source of the problem: when PVC decomposes in a fire it releases hydrogen chloride gas, which meets the moisture in the air and in people's lungs and becomes hydrochloric acid. This is a materials question, not a manufacturing quality one. A PVC sheath made to the standard behaves exactly this way.

LSZH compounds replace the halogenated polymer with a polyolefin base and achieve flame retardance with a mineral filler, usually aluminium trihydroxide or magnesium hydroxide, which releases water of crystallisation when it is heated. The result is a sheath that produces very little acid gas and much less smoke, at the cost of a material that is stiffer, slower to process, and more expensive per kilogram. The filler loading is the reason: a compound carrying 60 % mineral filler is not a cheap compound.

Halogen acid gas: the two numbers in IEC 60754

Halogen-free is a measured property, not a marketing word, and it is measured in two parts. IEC 60754-1 determines the amount of halogen acid gas evolved when the material is burned, expressed as a percentage by mass of hydrogen chloride; a halogen-free material evolves less than 0.5 %. IEC 60754-2 takes the gas, bubbles it through water, and measures what that water has become: pH and conductivity. A halogen-free result requires a pH of at least 4.3 and a conductivity of no more than 10 µS/mm.

The reason it is two tests and not one is that the second one is the one that predicts the damage. Concentrated HCl arriving on a printed circuit board, a relay contact or a steel structure in the presence of condensation is what destroys equipment after a fire, and it does that some distance from the fire itself. A data hall that survives a small electrical fire and then loses every server in the building six months later to corrosion is a real failure mode, and it is the argument for LSZH in data centres that has nothing to do with evacuation.

Smoke density: IEC 61034 and the 60 % line

Smoke is measured in a closed 3 m cube, per IEC 61034-2, by how much light gets through it. The acceptance line for a low smoke cable is a light transmittance of at least 60 %. It sounds like a modest number until you translate it into what people experience: in a corridor, visibility is what determines whether anyone finds the exit, and the difference between 60 % and 15 % transmittance is the difference between seeing the exit sign and not seeing it.

There is a separate reason smoke matters in buildings that are still safe to leave. Escape routes in a tall building, a metro station or a hospital corridor are enclosed, and in an enclosed space smoke travels faster than people and further than the fire. Cable is rarely the only combustible material in a fire, but it is often the one that runs continuously through the escape route, distributing smoke along the length of it.

Toxicity: the part that actually kills people

Acid gas and smoke are both proxies for the thing that matters, which is whether the people in the building can breathe and get out. Most fire deaths are caused by inhaling fire effluent rather than by burns, and the principal toxicant in most fires is carbon monoxide, which is produced by burning any organic material, LSZH included. No cable is non-toxic in a fully developed fire.

The differences that do exist sit on top of that. Chlorinated materials can produce hydrogen chloride and, under some fire conditions, chlorinated dioxins and furans. Halogen-free materials still produce carbon monoxide, and their fire effluent can be irritant. The honest engineering position is that LSZH reduces two specific hazards - acidity and smoke - and that it does so in the conditions where evacuation is still possible, which is precisely the phase of a fire that the cable standards are written for.

Flame spread is a separate property, tested separately

Flame retardance is about whether a cable carries a flame along its length once a fire has reached it, and it is tested on its own. IEC 60332-1-2 applies a small flame to a single vertical sample. The bunched tests of IEC 60332-3 apply a much larger flame to a bundle, and they are graded by how much combustible material is in the bundle and for how long: category A uses 7 litres of non-metallic material per metre of sample for 40 minutes, category B uses 3.5 litres per metre for 40 minutes, category C uses 1.5 litres per metre for 20 minutes and category D uses 0.5 litres per metre for 20 minutes. Category A is the most demanding, and a project that specifies C when the installation needs B has specified the wrong thing.

Two cables can therefore be identical on smoke and opposite on flame spread. In practice most LSZH compounds are formulated to pass a bunched category as well, because the mineral filler suppresses flame, but the pass has to be evidenced: ask for the test report for the category, not for the material name.

What LSZH costs you, in money and on the tools

The compound is more expensive, because the mineral filler loading that gives the flame performance is high, and it is a filler that costs money rather than a cheap extender. Processing is slower too: the compound is stiffer and more sensitive to temperature, and the extrusion line is often run at a lower output. Both costs end up in the price per metre.

On site the differences are practical rather than financial. An LSZH sheath is stiffer and less elastic, so it is harder to pull through a tight bend and less forgiving of rough handling in cold weather. It strips less cleanly than PVC, and because the filler makes it more abrasive, it takes more effort and blunts tools faster. It also tends to mark and scuff more visibly on a finished installation. None of these are reasons not to use it where it is required; they are reasons to tell the installation crew what to expect.

Where each is customary or required by building type

The decision is rarely left to the engineer in an occupied building, because the occupancy type drives it. Enclosed spaces with many people, limited exits, or people who cannot walk out unaided are the ones where the standards and the clients converge on LSZH. The other driver is equipment: any building where a cable fire would destroy something far more expensive than the cable, which is the data centre case.

PVC remains the correct choice where the installation is outdoors or in open industrial space, where the smoke from a cable fire disperses instead of filling a corridor, and where the local code does not require otherwise. It is also still the default for many industrial control and power applications where the fire performance requirement is about flame spread alone, and a flame-retardant PVC passes that test economically.

How to specify it: the CPR classes and the RFQ wording

In the European Union the choice is not optional for cables placed on the market for use in construction works. Regulation 305/2011, the Construction Products Regulation, requires a power or control cable intended for permanent installation in buildings to be CE marked under the harmonised standard EN 50575, with a declared reaction-to-fire class and a Declaration of Performance. The class is measured on the complete cable, not on a sample of the sheath.

The classes run from the best performing downward: Aca, B1ca, B2ca, Cca, Dca, Eca and Fca, and they are normally written with three additional classifications - smoke s1, s1a, s1b, s2 or s3, flaming droplets d0, d1 or d2, and acidity a1, a2 or a3. A typical power cable declaration is Dca-s2,d2,a1 or Cca-s1b,d1,a1. The smoke classes require light transmittance, with s1 at least 60 % and s1a at least 80 %; the acidity classes require conductivity below 2.5 µS/mm for a1 and below 10 µS/mm for a2.

• Name the properties separately: low smoke, zero halogen, and the flame spread category
• For the EU market, name the CPR class and ask for the Declaration of Performance with the CE marking
• For other markets, name the test methods: IEC 60754-1 and 60754-2, IEC 61034-2, and IEC 60332-1 or 60332-3 with its category
• State whether the insulation, the bedding and the outer sheath must all be LSZH, not only the sheath
• If the circuit must keep working during a fire, that is a different requirement entirely and it is covered in the fire resistant article in this series