Maintenance and restoration: keeping a plant alive for 25 years

Glass does not wear out, but the things around it do: closures let water in, connectors collect dust, excavators find cables, documentation drifts from reality. A fibre plant that is monitored, inspected and documented keeps its loss budget for decades; one that is left alone fails at the worst moment and takes hours longer to fix because nobody knows where the fibres go. This page describes the preventive maintenance cycle, monitoring options, the spares an operator must hold, the restoration process from alarm to permanent repair, the root causes of outages, and the lifecycle decisions of an ageing plant.

Preventive maintenance cycle

Interval	Task	Why
Continuous	DDM power monitoring with alerts on deltas (Monitoring); RFTS where deployed	catches slow degradation
Monthly	review loss trends per link; check alarms on protection paths (are they still up?)	a dead protection path is an unprotected link
Quarterly	inspect and clean accessible panel ports that will be re-patched; verify labels at changed positions	contamination is the top fault
Annually	walk/drive the route: markers, manhole covers, poles, aerial sag and clearance, vegetation; open a sample of closures for water/rodent damage; cabinet filters and seals	outside plant degrades visibly before it fails
Annually or after works	OTDR every fibre (or a sample plus all fibres near works) at 1310/1550/1625 and compare with the installation trace	new bends, splice degradation, water
3–5 years	full documentation audit: as-built vs field, splice diagrams, spare counts	the restoration time depends on it
On every change	update records, re-test affected fibres, re-baseline DDM	—

Monitoring options

Method	What it sees	Cost	Notes
Transceiver DDM via SNMP/telemetry	Rx/Tx power per link end, trends	none (built in)	resolution ±0.1 dB relative; alert on 2 dB drop from baseline (Management & monitoring)
Remote fibre test system (RFTS)	OTDR at 1625/1650 nm over a WDM coupler on live fibres; locates events automatically	high	carriers, long routes, dark-fibre providers
Dark-fibre monitoring	periodic OTDR on spare fibres of the cable	medium	detects cable-level damage before it hits live fibres
Protection-path supervision	alarms on standby paths, periodic switch tests	low	—
Distributed sensing (DAS/DTS)	vibration and temperature along the fibre — detects digging near the cable	high	new in utilities and pipelines
Environmental sensors	water, door, temperature in cabinets and closures	low	IoT sensors on FDHs

Spares and the restoration kit

Item	Quantity	Notes
Cable of each type on the route	500–2 000 m on reel per region	same fibre count or higher; same type (G.652.D / G.657)
Splice closures	2–3 per type	plus sealing kits, trays, protectors
Mechanical splices and field connectors	24+	temporary restoration
Pigtails and patch cords	24+ of each connector type, both polishes	APC and UPC separate
Fusion splicer, cleaver, OTDR, LSPM, VFL, fibre identifier, inspection scope	1 set per crew	maintained and calibrated
Transceivers	5–10 % per type, minimum 2	tested on the bench (Code database)
Attenuators, couplers, splitters	assorted	—
Consumables	sleeves, IPA, wipes, cleaners, labels	—
Documentation	current as-built, splice diagrams, test baseline — offline copy in the truck	the network may be down

Restoration process

Step	Action	Tools	Target time
1. Detect	alarms: link down, Rx −40 dBm, protection switch; customer reports	NMS, DDM (No link)	minutes
2. Classify	equipment vs fibre: swap module, check both ends' Tx	bench/switch DDM	15 min
3. Localize	OTDR from both ends → distance to break; convert to route position via GIS (fibre length ≠ ground distance: slack, sag, closures)	OTDR, as-built	30–60 min
4. Dispatch and access	crew to the location; permits for road/manhole; safety setup (Safety & handling)	—	1–3 h
5. Temporary restore	mechanical splices or temporary cable jumpered around the damage; restore priority circuits first	restoration kit	+1–2 h
6. Permanent repair	new cable section spliced in with two closures (or one if slack allows); fusion splices; reseal	splicer, closures (Splicing)	days (scheduled)
7. Test	bidirectional OTDR of repaired fibres; insertion loss vs original budget	OTDR, LSPM (Testing)	—
8. Document	updated as-built, splice diagrams, new baseline traces and DDM values; incident report with root cause	—	within days

Restoration adds loss: two extra closures ≈ 4 splices ≈ 0.2 dB plus maybe 50–100 m of cable; check the link budget still has margin (Link budget).

Root causes of fibre outages

Cause	Share (typical outside plant)	Prevention
Excavation / construction ("dig-ups")	40–60 %	route marking, call-before-dig, locate services, depth, warning tape, DAS
Vehicle and equipment strikes (poles, cabinets, low aerial)	10–15 %	clearance, protective posts
Rodents and animals	5–10 %	armoured or rodent-resistant cable, sealed ducts
Weather: ice, wind, flooding, lightning	5–10 %	aerial design margins, closure sealing, grounding
Vandalism and theft (copper thieves cutting hybrid cables)	5 %	dielectric cables, secured cabinets
Component failure: closures, connectors, splices ageing	5–10 %	inspection cycle, quality parts
Human error during works	5–10 %	fibre identifier before cutting, change control, labels

Slow degradation and its signs

Sign	Cause	Action
Rx power down 0.5–1 dB over a year, uniform	connector contamination, closure ingress	clean, inspect, OTDR compare (Plant faults)
Loss step at one OTDR event vs baseline	splice or connector degrading, bend introduced by works	re-splice, re-route
Attenuation slope rising, worst at 1383/1550	water/hydrogen ageing of old fibre	replace section; use 1310 only
Seasonal swings ±1 dB	temperature on aerial/outdoor plant	acceptable within margin; document
Reflectance growing at a mechanical splice	gel drying	replace with fusion

Lifecycle of a plant

Question	Guidance
How long does fibre last?	25–40 years; G.652.D installed since the 2000s is fine for 400G coherent; the coating and cable, not the glass, set the limit
Upgrading capacity on existing fibre	1G → 10G → 100G → 400G on the same G.652 pair; add DWDM when pairs run out; coherent tolerates old fibre's dispersion; check PMD on pre-1995 fibre (Fibre types)
When to replace	water-damaged or hydrogen-aged sections, repeated breaks on a route, insufficient count, G.653 on a DWDM route
Connector wear	500–1 000 matings; panel ports that are re-patched daily need replacement or cassettes
Documentation	the asset with the shortest life — audit every 3–5 years
Optics	10–20 years of laser life; replace on rising bias or falling power, not on age (Tx bias & ageing)

In CodingBox

Two of the maintenance inputs live in module memory: the DDM values that trend and the identity that ties a module to a port in the records. CodingBox reads both on the bench for spares testing and baseline records, so a restoration crew swapping modules at 3 a.m. can tell a tired laser from a broken fibre (Check transceiver, DDM in the app).

The records this cycle depends on — identifiers, labels, splice schedules, trace archives, handover packages: Documentation & labelling; the live side of monitoring — DDM, OSC, OCM, RFTS, alarms and how to correlate module data with fibre data: Monitoring & management.