Water in the closure: ITU-T L.315 makes it visible to the OTDR

An underground fibre cable is protected from water by design. The recommendation’s Summary says so plainly: “Widely used underground optical fibre cables employ water-blocking materials and are maintenance free in regard to water penetration. However, water penetrated into closures/cabinets would increase the risk of significant degradation to the optical fibres and/or connectors.” The closure is not the cable: water that gets inside one stays a risk to fibres and connectors. ITU-T Recommendation L.315, approved in March 2018, sets out how to make that water visible at a distance, using the OTDR that is already on the network: not one more sensor to read, but one more event on a trace that gets looked at anyway.

Why the closure, not the cable

A closure’s seal depends on several factors at once. Clause 6.1: even a closure designed to be sealed can let water in, “because it depends on the level of protection afforded to the products, the immersion period, pressure and the quality with which they are assembled in the field.”

The first three variables are not in the installer’s hands. The last one is: this is not a product defect, it is a workmanship variable — the same sealing claim on a closure we have covered — governed by a specification with an acceptance check. Clause 6.1 classifies the activity itself: “Water detection is defined as a function of surveillance and testing maintenance activity that is categorized in preventative maintenance” — preventative maintenance, not repair.

What the system has to do

The same clause sets out four requirements:

“it should be conducted in conjunction with the optical time domain reflectometry (OTDR) testing using a maintenance support, monitoring and testing system”
“it should use an inactive optical fibre (not carrying a communication signal) dedicated for maintenance use as a monitor line”
“it should be capable of monitoring multiple sensors in the longitudinal direction of an optical fibre link within the dynamic range of OTDR testing”
“it should be capable of locating splice closures/cabinets where water penetration has occurred. The locating resolution should be higher than the minimum installation distance of the water sensors in an optical fibre link.”

The second point is a design consequence: a dark fibre has to go into every cable, or the system cannot be added later without reopening it.

How the sensor works

Two operating principles, clause 6.2.1.1: the IL type “changes insertion loss to a monitor line when water surrounds the sensor”; the RL type does the same with return loss. Two ways to fit it, clause 6.2.1.2: external, “externally attached to a dedicated optical fibre, such as a bender mechanism”; pigtail, “spliced to a dedicated optical fibre.”

Two physical requirements: “The water sensor should be small enough to be accommodated in fibre organizing trays in closures/cabinets”; “The water sensor should not contain conductive materials. The constituent materials should have sufficient durability in underground environmental conditions.” In a chamber, one more piece of metal is one more risk.

A real sensor, in the informative Appendix I: an absorbent material swells on contact with water and bends the fibre, raising the loss. And it is “insensitive to water vapor” — condensation does not trip it. Insertion loss before immersion 0.01 dB, an increase above 2.0 dB at 1550 nm, hold “more than 2400 hours”.

The numbers that go into a specification

Table 1 is what ends up in the specification:

operating temperature: 0 °C to 60 °C;
threshold: IL type ≥ 2 dB, RL type ≥ 25 dB;
response time: ≤ 24 hours;
retention after detection: latching.

Four notes matter as much as the numbers: contaminated water may not freeze exactly at 0 °C; thresholds hold at 1550 nm ± 20 nm; response time runs from the moment a sufficient amount of water surrounds the sensor; and latching holds the state, for a period set by the operator’s maintenance policy.

Clause 6.2.3 adds three design rules: “The IL/RL change should not reach the threshold value before water gets in contact with the sensor. After sufficient amount of water surrounds the sensor, the IL/RL change should exceed their thresholds within the response time and maintain their states (latching). The IL/RL threshold should be designed to be clearly distinguishable from normal events on an OTDR trace such as splices and/or connectors.”

That last rule is the one that matters on site: a sensor that reads like a splice in an OTDR report, or that falls inside a dead zone, is worth nothing.

How we check it

At acceptance we verify that the maintenance fibre genuinely exists, is dark, and is the one declared in the design — not just any spare fibre. We check that the sensor sits in the organiser tray, not left hanging off a loose loop of fibre, and that the declared type, IL or RL, matches what was actually installed. We compare its signature on the OTDR trace against neighbouring splices, to confirm the two stay distinguishable. And we record the position in the as-built, with the correct chainage: without that, a water alarm arrives with no way of knowing where to go.

See the service · Talk to an engineer

How often it gets tested

Clause 6.3 fixes the test wavelength: “A wavelength of 1550 nm ± 20 nm or longer is used.” And the cycle: “The periodic test cycle should be less than the sensor hold time.” — to be weighed against mechanical risk from exposure and mean repair time.

The urgency has a number, from Appendix II: “In general, the failure probability of optical fibre in water is more than ten times greater than that in dry air.” Table II.1, assuming a 10-day repair time: a 5.0 × 10⁻⁶ failure probability is reached in 735 days, cycle ≤ 725; 1.0 × 10⁻⁶ in 102 days, cycle ≤ 92; 5.0 × 10⁻⁷ in 49 days, cycle ≤ 39.

What the numbers mean, clause II.2: “For example, considering a failure probability of less than 10−6, which is comparable to the reliability of cable portions, a period of less than 100 days is required for the maintenance.” This is where the two ends of the document meet: the 2400 hours of hold time in Appendix I are exactly one hundred days, and clause 6.3 wants the test cycle shorter than the hold. The sensor remembers long enough — provided somebody interrogates the fibre before it stops remembering.

Appendix II states its own limits: “This appendix does not form an integral part of this Recommendation” — the example’s parameters (a 3-metre fibre, a 20 mm bend radius) are a method, not a prescription. L.315 itself is an international technical recommendation, not a law: compliance is voluntary.

What to write into the specification, what to check at acceptance

In the specification:

a dedicated, dark maintenance fibre, for every cable and every span;
sensor type, IL or RL, and the required threshold;
response time and hold time;
no conductive materials;
housing in the organiser tray;
test wavelength, test cycle and who runs it.

At acceptance:

presence and identification of the maintenance fibre;
a reference trace with the sensor already installed, so the event is known beforehand;
the event distinguishable from neighbouring splices;
the chainage recorded in the as-built.

Two threads, applied

Threshold, response time, hold time, test wavelength and test cycle become verifiable lines in the specification and acceptance checkpoints, with the record a client can exhibit.

OTDR traces, measurements, sensor positions and as-built records stop being scattered files: together with CSIDIA, the group’s other company, they become a single map of the network — which closure, on which span, what history of alarms — on which an AI runs the diagnosis and the crew closes the fault. A water alarm in a known closure, with the correct chainage, is a targeted callout rather than a guess: the same logic behind the continuous optical monitoring and the as-built records already covered. Within the client’s own perimeter: on-premise, with no deep integration, or a dedicated cloud with a data centre in Italy, always with shared management.

Do you have underground closures with no dedicated maintenance fibre? Talk to an engineer: the site visit costs nothing, and the dark fibre gets designed in before the cable is laid, not after.