Hard break or cable cut

Likely causes

• Third-party dig-in

  most common

  Every fiber in the sheath dead at the same distance. Trace ends in a high reflection or drops straight into noise. Fresh excavation, a new bore, or a locate ticket nobody called, at the distance you calculated.

• Crush or shear at a stress point

  common

  Distance lands on a duct entry, a hanger, a handhole where a lid went down on the cable, or a pole line somebody worked last week. Often not a clean break: a non-reflective end instead of a reflective one.

• Aerial damage: storm, limb, vehicle, or rodent

  common

  Aerial span, visible strand or lashing damage, chew marks at a mid-span or a slack coil. Rodents favour slack storage and anything resting on a structure.

• Break inside a closure or a tray

  less common

  Distance lands on a known splice point and only the fibers in one buffer tube are out. Somebody re-entered that closure recently, or the tray was overfilled.

• Not a break at all

  occasional

  A shattered connector or a severe kink can end a trace with no usable return beyond it, and a couple of fibers may still pass. Look at the signature and at whether the whole sheath is really down before you mobilise a splice crew.

• Wrong distance, not a wrong cable

  occasional

  The location on the ground shows nothing disturbed, and the distance lands between two features rather than on one. That is an index-of-refraction or a slack-allowance error, not a mystery.

What to bring

• OTDR with launch cord and the cable's IOR
• clip-on live fiber identifier
• optical power meter
• VFL
• toneable duct or cable locator
• as-built, splice schematic and cable data sheet
• fusion splicer, closures and splice sleeves
• restoration cable and slack storage hardware

Steps

1. 1

   Step 1: Shoot from both ends and bracket it

   Distance from end A plus distance from end B should add up to the optical length of the route. If it doesn't, your IOR or your slack allowance is wrong and your location is wrong with it. Two shots also tell you whether the break is one point or two.

   If that doesn’t do it

   Only one end is reachable: take the single distance, but treat it as an estimate until the ground confirms it.

2. 2

   Step 2: Set the index of refraction from the cable data and add up every coil

   IOR is a per-cable figure from the manufacturer's data sheet, not a global default. Then count slack: coils at every splice point, pole, handhole and building entry routinely add tens of metres each. Add them off the as-built before you convert optical distance into a place on the ground. The OTDR gives you a distance along the fiber, not a distance along the street.

   If that doesn’t do it

   No IOR and no reliable as-built: shoot to a known feature you can physically identify, then work the unknown distance as an offset from it.

3. 3

   Step 3: Read the break's signature before you call it

   A reflection then noise is a clean break or a shattered end. A non-reflective drop into noise is crush, or a bend severe enough to kill the return signal. A trace that simply ends with no end-of-fiber reflection can also be a fiber terminating in gel or a mechanical splice. Check several fibers: if some still pass, you have damage, not a cut.

   If that doesn’t do it

   Signature says bend rather than break and some fibers still pass: the bend sequence. If the fault comes and goes with the weather, the water-ingress sequence.

4. 4

   Step 4: Identify the fiber and prove it dark before anything is cut

   This is the step that separates a repair from an outage report. Clip-on identifier over every buffer tube you are about to open, and a power meter on every fiber you can access. Confirm the cable and the tube by identifier and by tone or trace, never by colour code and print alone, because the tube you want is very often not the one the record says. Cutting a live buffer tube in a working sheath is service-affecting and career-affecting.

   If that doesn’t do it

   You cannot positively identify it: stop. Get the far end to dark the fibers, or get written release from the owner of every circuit in that sheath before a cutter comes out of the bag.

5. 5

   Step 5: Confirm the location on the ground before you break pavement

   VFL reaches a short way and is worth trying at an accessible structure. Use a toneable duct locator or the cable's locating element for the route itself, call the locate ticket for everything else in the trench, and daylight carefully. Confirm what you find matches the signature you measured.

   If that doesn’t do it

   Nothing at the calculated location: re-check slack and IOR, shoot from the other end, and walk the route for fresh disturbance between the two bracketing distances.

6. 6

   Step 6: Restore with the method the job accepts

   Splice in a section of the same fiber grade and count, with enough slack stored at both new closures for one more repair. Two closures and a section beats one stretched splice every time. Log both new closures, the section length and the new slack on the as-built, or the next crew is working blind.

   If that doesn’t do it

   No matching cable available: splicing a different grade in is a documented deviation, not a field decision. Get it in writing and record the fiber change at both closures, because it will show up as a gainer for somebody later.

7. 7

   Step 7: Test and document before you demobilise

   Bidirectional OTDR through the new splices and Tier 1 end to end against the budget. Save the traces with the new distances, and note the restored slack. Circuits come back on somebody else's schedule, so leave proof that the plant is right.

   If that doesn’t do it

   New splices read high: the splice-loss sequence. Total loss high with clean splices: the loss-budget sequence.

More Fiber Optic guides

• Link fails the loss budget
• Dirty endface that won't clean up
• Duplex link dark: polarity reversed
• Bend loss: loss climbs at the longer wavelength
• Link errors because there is too much light

All 54 guides