Request an intervention

Operational notes Engineering

A closure is qualified for five re-entries. On the network, who actually counts them?

6 min read

A bolted flange with a bellows joint on an industrial pipe, shot close up, in black and white
A seal is qualified in the lab over a handful of cycles. Then it is opened and closed for twenty years, and nobody keeps count.

A closure installed twelve years ago in a suburban manhole is opened to connect a new block of flats: one more fibre to terminate, the organiser to reopen, pressure to restore once the work is done. Nobody at the company knows this is the fourth time that closure has been opened — because nobody ever wrote it down anywhere. If it were the sixth, or the eighth, the specification would not know either: it prescribes a product rating, not a counter.

What the standard qualifies, and for how many times.

We have already written here that a closure’s seal is not declared with a rating but qualified under Recommendation ITU-T L.201 (05/2021). The clause that covers re-entries is B.1.13, Re-entries: procedure IEC 61300-2-33, conditions “Ageing between each re-entry: at least one thermal cycle” and “Number of re-entries: 5.”; performance criterion “Tightness after test and after each re-entry.” — checked at every cycle, not only at the last one.

It is a detail that matters more than it looks: it is a type test, run in a laboratory on a sample, over a few weeks. It proves that model of closure can withstand five open-and-reclose cycles under controlled conditions. It says nothing about the specific closure installed in a manhole twelve years ago, nor about how many times it has already been opened before today.

The gap the standard itself flags, and does not cover.

There is a part of L.201 almost nobody reads to the end: Appendix II, titled Ukrainian experience, with an explicit notice at the top: “This appendix does not form an integral part of this Recommendation.” It is informative material, not an ITU-T requirement.

Inside that appendix, at clause II.5.1, sits the only real-field-condition test in the whole document: “1 re-entry and reclose over 1 month in real field conditions.” — one re-entry and reclose, over one month, in real field conditions. One single event, one single month. The same appendix, at clause II.2.3, sets a second useful value: minimum storage bend radius for fibres on the cassette 30 mm, with residual loss allowed below 0.05 dB after storage.

Put together, these two figures tell the problem better than any rating could: the laboratory qualification proves five cycles over a few weeks; the most concrete informative material ITU-T itself attaches validates just one, over one month. An access-network splice point stays in service for twenty years or more. That gap — between what the type test demonstrates and what a network actually lives through — no public standard measures, and no specification asks for it unless it is written in on purpose.

What degrades, re-entry after re-entry.

  • The seal. Every compression and release of an elastomer leaves a permanent deformation — compression set, a property standardised by ISO 815-1:2019, Rubber, vulcanized or thermoplastic — Determination of compression set. L.201 does not quantify it: it only requires the closure to hold after each re-entry of the type test, as already noted above, but it says nothing about how far the original seal can be reused in the field beyond that test.
  • The fibres on the cassette. Every re-entry means handling the slack coils stored in the organiser. The Appendix II value — 30 mm minimum radius, residual loss under 0.05 dB — is lost exactly when a coil is moved to make room for the new splice without rebuilding the same radius: the same physical principle already described for macrobending, here repeated at every intervention instead of once at installation.
  • The maximum number, different from one closure to the next. Where a network mixes products from different suppliers and generations, the qualified number of re-entries may not be the same for every closure. Without a figure per individual closure, “five” becomes an average that no single real closure is certain to meet.

What to write into the specification.

  1. The qualified number of re-entries declared by the manufacturer, per closure model, with reference to ITU-T L.201/IEC 61300-2-33 — not assumed to be the same for every closure on the network.
  2. A re-entry log for each individual closure: date, reason, technician, sealing-test result — cumulative for the whole life of the closure, not only at installation.
  3. A sealing test repeated at every field re-entry, not only at initial handover: the same pressure/gauge protocol already required at commissioning, repeated at every event.
  4. A re-closure kit declared in the catalogue — seals, sealants — with replacement logged at every opening: the original material is not reused beyond what the manufacturer declares.
  5. Storage bend radius checked at every intervention on the organiser, not only the first time the tray is installed.
  6. A written replacement or requalification threshold: once the log reaches the number declared by the manufacturer, an explicit decision — replace the closure or justify a documented exception — not a silent reclose with no trace.

How it is verified at acceptance.

For a closure already in service for years, acceptance is no longer just the sealing test of the day: it is checking the cumulative log against the maximum number declared by the manufacturer, the consistency between that log and the as-built record, and the result of the sealing test from the most recent re-entry — not the original one. A closure approaching its qualified limit without a written decision is a risk the commissioning test, on its own, no longer catches. Only the log does.

The point.

The qualified number of re-entries stops being a forgotten line on a product datasheet and becomes a written threshold, checked at every intervention, with the sealing test and the re-closure kit delivered together with the work — this is compliance applied to a single closure. Every re-entry into every closure then joins the same single network map we keep for the as-built record: not a paper log per manhole, but the data an AI uses to flag which closures are approaching their qualified limit before the next emergency crosses it unnoticed — for an operator, a data centre, an industrial site, a public authority, a healthcare facility or a defence site. Where AI on network data is needed, it runs within the client’s own perimeter — on-premises or in a dedicated cloud with a data centre in Italy — together with CSIDIA, the group’s other company.

Do you have closures that have been in service for years with no log of how many times they have already been opened? Talk to a technician: the site survey is free, and the count can start from the next re-entry, not the first.

Sources