The phone call I get most often from electrical contractors goes like this: “The drawing says ESE arrester. Which model — 2500, 4500 or 6000?” The answer is never a preference; it is a calculation, and the model number is the first input. In this product family the number encodes the triggering-advance class of the head — the ΔT value, in the 25 / 45 / 60 microsecond classes, that the NF C 17-102 radius formula converts into metres of protected coverage. An ESE2500 is the compact-duty head for modest heights and tight footprints; an ESE4500 is the middle class that covers most commercial buildings; an ESE6000 is the long-reach head for tall masts, wide footprints or the stricter protection levels. Choose by rule of thumb and you either leave the roof’s far corners exposed or overspend on coverage the building cannot use. This guide gives you the selection logic exactly as our export desk runs it: what the number means, how the radius formula turns it into coverage, a model-by-building-class table, and a calculator you can run on your own project figures.

The Model Number Is the Spec: What 2500, 4500 and 6000 Actually Encode

Buyers sometimes assume the numbers are marketing tiers — good, better, best. They are not; they are engineering classes. The number tracks the head’s triggering-time advantage ΔT: how many microseconds earlier than a passive reference rod the head launches its upward streamer, measured in a high-voltage laboratory. Because the protected radius grows directly with that advance, the model number is effectively a radius class. This is also why the ΔT test report matters more than the catalogue photo: two heads that look identical can be a class apart, and the class is what your coverage calculation uses. When we quote a model, the ΔT class and its test documentation travel with the quotation, and I advise every contractor to demand the same from any supplier — the habit is spelled out in our guide to the certification documents behind ESE sourcing.

One clarification I add to every first conversation: the class numbers follow the NF C 17-102 family convention (25 / 45 / 60 μs tiers), and the exact certified ΔT of a given unit is whatever its laboratory report states — the report, not the label, is what goes into a compliant design file. If your project auditor asks for the value, the report answers them; the model name is just the short code that gets everyone to the right row of the table.

What Does the Radius Formula Say About Your Building?

The NF C 17-102 family radius formula, for a mast standing h metres above the protected surface, is R_p = √(2rh − h² + Δ(2r + Δ)), where r is the rolling-sphere radius for the required protection level and Δ is the triggering advance expressed in metres (a 25 μs class contributes Δ = 25 m, 45 μs gives 45 m, and 60 μs gives 60 m). Three variables, three consequences. Raising the mast lifts the radius but with diminishing returns as h approaches r. Tightening the protection level shrinks r and shrinks every radius in the project. And raising the ΔT class adds coverage at any fixed mast height — which is exactly the lever the ESE2500/4500/6000 choice controls. I run this formula on the client’s actual numbers in every quotation, because the interactions are strong enough that intuition alone routinely picks the wrong model.

The air-terminal head shared across the ESE2500/4500/6000 family — the triggering device inside differs by class, and its ΔT report is what the model number stands for.

Matching Models to Building Classes: The Selection Table We Quote From

Formula in hand, here is the practical mapping our desk uses as the first pass on contractor inquiries — the same logic, compressed into the table we quote from before the detailed drawing review.

Notice the last row — it is the one distributors thank us for later. Retrofit projects arrive with the mast height already capped by the existing structure, the parapet or the architect’s patience, and in those jobs the model class is the only variable left to buy coverage with. Because the formula trades height against ΔT, stepping from a 4500 to a 6000 at a fixed mast height recovers a meaningful slice of radius — often the difference between one mast and two. The reverse discipline matters too: if the table says a 2500 covers the building, quoting a 6000 is overselling, and we will tell you so on the quotation itself.

Where Does the Protection Level Enter the Decision?

The protection level is the quiet variable that contractors underestimate. IEC 62305-1 frames four protection levels with rolling-sphere radii of 20, 30, 45 and 60 metres, and the level comes from the risk analysis of IEC 62305-3 territory — what the building contains, who it shelters, and what a strike would cost in consequences rather than in repairs. Level I is the strict class for high-consequence facilities; Level IV covers ordinary structures. The level enters the radius formula through r, so two identical buildings specified at different levels need different coverage from the same mast. I ask for the project’s required level before I quote a model, because a model that satisfies Level III can leave a Level II requirement uncovered at the same height — and “the drawing said ESE” is not a level.

If the project documentation does not state a level, treat that as a question to raise, not a gap to paper over. Our team will compute from the stated level when you have one and flag the assumption when you do not; either way the calculation note goes into the file with the model recommendation, so the consultant can see the basis rather than take the model on faith.

A cage-style ESE terminal from our range: head geometry varies by series, but the selection logic — class, height, level — is identical across the family.

Height, Footprint and One-Mast Coverage: How Do You Know One Head Is Enough?

The test is simple and ruthless: draw the protected circle on the roof plan. If the computed radius R_p from the mast position reaches every corner of the footprint with margin, one head is enough; if a corner falls outside, the options are a taller mast, a higher model class, or a second position — in that order of preference for most flat commercial roofs. Because corners are always the exposed geometry, I ask contractors for the diagonal half-width of the roof, not the area: the corner distance from the mast is the number the radius has to beat. On L-shaped or stepped roofs the same test runs per zone, and a hybrid with conventional air terminals on secondary blocks is a legitimate answer — our comparison of ESE versus conventional Franklin rods covers when each technology earns its place.

Two practical reminders from jobs that went smoothly. First, wind and fixing: a tall mast with a head at the top is a cantilever in a storm, so mast design, base fixing and the head’s wind rating belong in the same submittal as the radius calculation — our air terminals are built for exposed service, but the mast is a structural item your installer owns. Second, the earth: every joule the head intercepts still travels down the conductors into an earth termination that must meet its resistance target, as the component and earth requirements of IEC 62561-2 reflect. The best-selected model on a failed earth is a failed system; our grounding materials line exists precisely so that both halves of the chain come from one accountable supplier.

Part of our air-terminal range: ESE heads and conventional rods ship side by side, and mixed designs are a normal outcome of the radius test rather than an exception.

Which Documents Should Come With Each Model?

Whatever model lands on the drawing, I tell buyers to ask for the same paper set, because the class claim is only as good as its file: the triggering-advance (ΔT) test report from an accredited high-voltage laboratory, measured per IEC 62305-3 and the NF C 17-102 test protocol; the material and plating declaration for the head; the installation manual with torque and mounting details; and the type-test certificates for the down-conductor and termination hardware that complete the system under IEC 62561-2. In my experience, a supplier who sends all four without being chased is a supplier whose factory audits go smoothly.

Because the ΔT figure is the single number that separates the three models, I never accept a brochure value on its own — the accredited test report, not the label, is what goes into a compliant design file.

Frequently Asked Questions