Directly buried fibre cable: how much must it withstand crushing and rodents?
7 min read
In January, a directly buried cable stops transmitting at one precise point. The acceptance report, two years old, is faultless: crush test passed, impact passed, attenuation within budget. Nobody, though, can find it written down what level of resistance to rodents and insects that cable was meant to guarantee — because on this point the Recommendation that defines it sets no value at all. It defers to an agreement that, even if ever put in writing, is not sitting on the same shelf as the rest of the acceptance file.
Two tests give you a number. One does not.
ITU-T L.101, Optical fibre cables for directly buried application, edition 3.0, approved on 29 August 2024 by Study Group 15 (formerly L.43 from 2002, then L.101/L.43 from 2015), is the Recommendation for cables meant to sit in direct contact with the ground, with no duct and no tunnel around them. It covers the cable’s mechanical, environmental and electrical characteristics — a different regime from the minimal-infrastructure DSA installation, openly best-effort for rural areas: this one is about the “ordinary” buried cable, the kind that goes into a standard specification.
Clauses 6.2.4 and 6.2.5 treat crushing and impact as certainties, not hypotheticals, occurring “during installation and throughout operational life”. In Annex A, which the text explicitly states “forms an integral part of this Recommendation”, both events carry a number.
Crush (A.3.5, Table A.10): plate-to-plate load for one minute and for ten — unarmoured cable 1.5 kN and 0.75 kN, armoured cable 2.2 kN and 1.1 kN. Attenuation, measured at the end of the long-term load, must show no change.
Impact (A.3.6): flat hammer, minimum face radius 300 mm, three strikes at least 150 mm ± 15 mm apart, energy of 10 J on unarmoured cable and 20 J on armoured cable.
Torsion (A.3.7): two metres of cable, 180° rotation in each direction, five cycles, no change in attenuation afterwards. Repeated bending (A.3.4): mandrel radius equal to half of twenty times the cable’s outer diameter — the same multiple discussed for minimum bend radius — minimum 150 mm, maximum increase 0.15 dB at 1550 nm during the test, no cracking of the armour beyond 5 mm, checked at 5× magnification.
Figures an acceptance test can check one by one, and a specification can demand on record.
The gap sits in the same table.
Then there is Table A.5, “Biotic characteristics”. Two rows: fungus, and rodents/insects. For fungus, the test method points to ITU-T L.161 — which we have not read for this piece: L.101 cites it only as a method reference, without reproducing its content. For rodents and insects, the method column is blank — a dash — and the value reads “as agreed, see 6.4”. Clause 6.4, in turn, says in a single line: “This topic is covered in [ITU-T L.161]”.
No figure, no test method listed by L.101 itself, no minimum threshold. Resistance to rodents and insects — the most-cited reason, in the same document, for fitting armour at all — is the one mechanical characteristic left entirely to agreement between manufacturer and customer, one L.101 does not even require to be put in writing in a verifiable form.
Armour, under clause 7.4, is to be provided “where protection from external damage (e.g., crush, impact, rodents) […] is required”: steel tapes “of various constructions” and, on metal-free cables, aramid yarns or glass-fibre-reinforced strands; steel wire and braid the Recommendation names elsewhere — with the caveat that lightness and flexibility drop once armour is fitted. It is the same reason fibre networks in harsh industrial settings call for dedicated jackets: yet neither armour type, here, carries a declared anti-rodent threshold from the standard itself.
The rest of the picture: freezing, chemicals, high voltage.
Freezing needs no severe winter: clause 6.3.6 notes that “the maximum expansion of frozen water occurs at −2 °C”, and ground holding buried cable is almost always wet. Water freezing inside or around the cable crushes it, it does not just chill it.
On chemical attack (6.3.9) the Recommendation sets no list of substances: it asks that whatever is present in the soil be surveyed, and the sheath chosen accordingly. Under 6.3.10, “mechanical aggression”, it concedes the limit up front: estimating how much a cable will take is difficult — protection comes from sheath thickness and/or armour, but the real risk is set by whoever knows that ground.
Near a high-voltage line, clause 6.5.3 adds a specific hazard: variable soil conductivity generates a voltage gradient that produces leakage currents and dry-band arcing on the jacket — intermittent arcing that damages it over time. The typical fix is a semiconductive over-jacket, or a track-resistant compound. The same territory as the shared trench between power cables and fibre: here the hazard is the electric field, not direct contact.
On temperature, Table 1 sets −30 to +60 °C in operation (a note also accepts as compliant cables tested down to −40 °C, a stricter limit used by many existing specifications) and 0/+50 °C for installation with a PVC sheath, −15 to +50 °C for PE.
What to write into the specification.
- Crush and impact class stated in newtons and joules, per Table A.10 and clause A.3.6 — not “resistant”, a figure.
- Armour type specified by name: steel tape, braid, glass-fibre-reinforced strands or aramid yarns — not left to the manufacturer.
- Rodent and insect resistance put in writing, with an agreed, named test method — not a “per L.161” that just refers you onward again.
- A survey of the chemical substances present in the ground before the sheath is chosen.
- A semiconductive or track-resistant jacket, if the route runs near high voltage.
- A stated temperature range, consistent with Table 1 or with a stricter value agreed separately.
See the service · Talk to an engineer
What we don’t know
Compliance with ITU-T Recommendations is voluntary: the text says so itself, “Compliance with this Recommendation is voluntary”. L.101 has no appendix of its own declared non-integral, unlike other Recommendations in the same series: it has a single Annex A, of tests, explicitly defined as an integral part, plus a bibliography of informative references only (prefixed “b-”). We have not read ITU-T L.161, cited twice by L.101 and never elaborated within it: we do not know whether it holds thresholds L.101 simply does not reproduce, and we neither state nor rule that out. We have not read ITU-T K.29 or K.47 either, cited for lightning. The IEC test methods cited (60794-1-21, 60794-3-11) are paywalled: we report only the values L.101 itself restates in its own tables. We make no claim, either way, about an Italian obligation on burial depth or rodent protection: not checked against an Italian primary source.
Two threads, on this subject
First thread: crush and impact class, armour type, rodent resistance agreed in writing and temperature range become specification lines, each with its own figure — checked at acceptance against the test report for the batch actually laid, not a generic data sheet.
Second thread: the test report, the as-built of the route, successive OTDR traces and splice-reopening records stop being separate archives and become, with CSIDIA, the group’s other company, a single map of the network on which an AI runs the diagnosis: on a duct-free buried cable, where crushing and a rodent bite leave no mark visible from the surface, the diagnosis that matters is the comparison between successive OTDR traces tied to the same run identifier — attenuation creeping up a few tenths of a dB, month after month, before it becomes an outage like the one in January. And the crew closes the fault. Within the client’s perimeter: on-premise, on standalone machines, or a dedicated cloud with a data centre in Italy, always with shared management.
The site visit, at no cost, produces the list of your directly buried, duct-free runs — with, for each one, the declared crush class, the armour type, the agreed rodent-resistance regime and where it is written down. Including the boxes that stay blank: it is yours to keep either way, whether or not we go on to work together.