Request an intervention

Operational notes Engineering

F and T are not EI: what a spec that only cites L.32 is missing

7 min read

An old brick wall with a bricked-up arched opening whose outline is still visible, next to a small barred window; black and white photograph, no people.
The wall was closed again, and you can still see where the arch was. The scar does not say which yardstick was used to measure the closing.

We’ve already told what ITU-T Recommendation L.32 says, clause by clause about sealing cable penetrations through fire-sector walls: minimum times, the sealant’s nine conditions, the penetration register. But one line in that 1998 document poses a different problem: what yardstick proves a seal holds? L.32 answers with an American test method and two letters, F and T. In Italy, what an acceptance test checks — and what the fire prevention code classifies — is minutes and a letter that starts with E. Neither of the two documents says how to move from one to the other.

The method L.32 names: American, not harmonized

Clause 2, point 3 of L.32 sets the minimum separation times between fire sectors — which we’ve already covered — and adds how they are to be verified: “These recommended requirements may be tested by the method given in standard ASTM E814 or by other nationally authorized methods.” ASTM E814 is the US test standard for penetration firestop systems, a paid standard: of it we can only report how L.32 names it, not its content.

Point 4, letter i) then requires approval certificates from authorized laboratories on the nominal values F — flame transfer to the unexposed environment — and T — heat transfer to the same environment: two separate values, not a single time-based rating. L.32 does not state the unit F and T are expressed in, and we won’t guess one: what matters is that neither is called EI.

The class an Italian spec has to ask for

Italy’s fire prevention code — Ministry of the Interior decree of 3 August 2015, as subsequently amended (the technical annex was replaced by the decree of 18 October 2019) — gives the subject a precise table. Chapter S.2, Resistenza al fuoco (fire resistance), applies explicitly to «Sistemi di sigillatura di fori passanti e di giunti lineari», systems for sealing through-penetrations and linear joints: the same seals L.32 is about. The standards it calls up are EN 13501-2 and EN 1366-3, 4; the classification is E — integrity — at 15, 30, 45, 60, 90, 120, 180 and 240 minutes, and EI — integrity plus thermal insulation — at the same steps, plus a 20-minute step. There is no F or T anywhere in this table: there is a number of minutes and a letter that says whether, besides stopping flame and smoke, the seal also stops heat toward the unexposed side.

The same chapter states, in S.2.1, that Chapter S.3 on compartmentation is a complement to it — a reference we found worded the same way in the text after 2019: a seal’s E or EI class is not a side detail, it is the ingredient compartmentation is built from.

Two yardsticks, no conversion table

A spec that reads sealing compliant with ITU-T Recommendation L.32 has not written something wrong: the recommendation remains in force, its only edition since 1998. But it has not written what actually gets classified and tested in Italy. L.32 leaves the test method open — “or by other nationally authorized methods” — and in Italy the nationally authorized method runs through EN 1366-3 and EN 1366-4, classified per EN 13501-2, not ASTM E814.

In the source documents we checked there is no table matching an F or T value to an EI class in minutes: the two systems stay parallel. An F90/T60 certificate under ASTM E814 is not, by itself, an EI 90 certificate under EN 13501-2: different tests, from different bodies, with no declared equivalence in the sources we read. That absence is not a technicality: a firestop rated under one system carries a laboratory report built around that system’s own furnace curve and pass/fail criteria, and nothing we checked lets a reader translate it into the other system’s terms. A spec that stops at the recommendation’s name, without writing the European class, leaves the supplier free to pick which yardstick to use — and so the performance.

Letter b): which class has to come back

In the same clause 4, letter b) requires that the sealing system “facilitate the installation of new or replacement cables while maintaining the original safety conditions.” We’ve already written about this as the point that decides whether compartmentation survives the first maintenance job. There is a corollary specific to the class: if the spec asked for EI 90 on a penetration, the original safety conditions mean that, once a cable is added, it has to come back to a verifiable EI 90 under the same system — not a generic resealed properly, nor a different product with no report attached. The 1998 letter b) does not name EI: today’s spec does, or doesn’t.

What to write in the spec

  1. Do not stop at citing L.32: write the required class under the European system — EI, with the number of minutes, for each penetration, distinguishing higher-risk sectors per L.32 itself — transformer rooms, power switchgear rooms.
  2. Ask for the product’s classification report under EN 13501-2, obtained through EN 1366-3 testing for through-penetrations or EN 1366-4 for linear joints — not a commercial datasheet, and not an F/T rating on its own.
  3. If a supplier offers a product certified only under ASTM E814, ask for the equivalent EI classification obtained through European testing: no source we verified authorizes inferring one from a declared F or T value.
  4. When a penetration is reopened for a new cable, require in writing the restoration of the same EI class declared originally, with a new report or a declaration from whoever did the work — not just the seal visibly closed back up.

The point

A seal’s class, like the penetration register we’ve already written about, is not only for the acceptance test on the day the site closes: it stays legible years later, when someone reopens that same opening for a cable that does not exist yet today. It belongs in the same single network map an AI diagnoses and a team acts on, together with CSIDIA, the group’s other company: on-premise on autonomous machines, or a dedicated cloud with data centres in Italy, always with shared management.

Do you need the EI class of your penetrations written into a spec, or have you lost track of which classification report the ones already closed were installed under? Talk to a technician: the site visit is free, and it starts from the next penetration due for inspection, not from zero.

What we don’t know

The technical content of EN 13501-2 and EN 1366-3/4 — paid CEN standards — stays outside this piece: only title, edition, status from the UNI catalogue. The same for ASTM E814: only what L.32 says about it. We read Chapter S.2 in two passes: first from a full copy of the 2015 decree hosted on a technical-legal site (the official Gazette PDF for that supplement carries only its first booklet, without the annex) — there the table is numbered S.2-22. The annex was later replaced by the decree of 18 October 2019 (Official Gazette no. 256/2019, ordinary supplement no. 41, verified by opening the page, not reconstructed by hand); to check the current number we read the same table in an official-origin coordinated text (Direzione centrale per la prevenzione e la sicurezza tecnica, the fire corps’ fire-prevention directorate): same subject, same standards, same E and EI steps, but numbered S.2-19. We have not verified whether a more recent decree touched that table again: that is why, above, we cite no number. We found no codified requirement in either text for a declaration of correct installation after a maintenance job: point 4 above is good specification practice, not a verified legal citation.

Sources