Fiber Splicing & Installation

What Splice Loss Should You Expect From a Fusion Splice?

A properly made fusion splice on matched single mode fiber is one of the smallest loss events on the whole route. The pass limit for each splice is set by the project specification and the network owner's acceptance criteria, usually as a bidirectional average, and clean work in a controlled environment lands consistently inside it. When a splice will not pass after two attempts, the problem is almost never the fusion settings.

The real causes sit upstream of the arc. A cleave angle out of tolerance, contamination on the fiber or in the V-grooves, coating residue left by a dull stripper, core geometry differences between two manufacturers, or a cable overtensioned during placement will all show up as loss at the splice point. We find and correct the cause instead of accepting a number that only looks acceptable from one direction.

Direction matters more than most specifications acknowledge. When two fibers with slightly different backscatter characteristics are joined, an OTDR reads elevated loss from one end and a gainer, an apparent negative loss, from the other. Neither reading is the truth. Averaging both directions produces the actual splice loss, which is why backbone acceptance testing should always be bidirectional.

Splice loss is also a budget line, not just a workmanship score. Every splice point, every patch panel, and every mated connector pair spends margin your optics need at turn up. If a design stacks enough splice points onto a long span to crowd the budget, we would rather say so during planning than hand you a passing splice log attached to a failing link.

• Cleave angle outside tolerance, usually from a worn blade or a rushed prep.
• Contamination on the fiber end face, in the V-grooves, or on the electrodes.
• Coating residue left behind by a stripper that needs replacing.
• Mismatched fiber types or core geometry between two cable manufacturers.
• Fiber stressed during placement by overtension, tight bends, or crushing.
• Protection sleeves not fully seated or shrunk unevenly in the heater.

How Your Loss Budget Adds Up Before Anyone Splices

The link budget is decided in design, and splicing either fits inside it or it does not. Fiber attenuation over the route, splice count, connector count, any passive splitters, and a reserve for future repairs all come out of the same number the optics can tolerate between transmitter and receiver.

Single mode fiber attenuates less at 1550 nm than at 1310 nm, so a span can pass at one wavelength and fail at the other for reasons that have nothing to do with distance. 1550 nm is the wavelength that exposes macrobends, because bend loss climbs sharply as wavelength increases. A span that reads clean at 1310 nm and shows an unexplained step at 1550 nm usually has a coil pulled too tight, a cable pinched in a duct, or slack stored on a radius the cable was never rated for.

Splitters dominate the budget on PON distribution. A split ratio consumes a fixed block of loss before the fiber leaves the cabinet, which leaves far less headroom for splice and connector losses downstream. Backbone spans are the opposite case: few components, long distance, and attenuation over kilometers doing most of the spending.

Repair margin is the piece most often left out. Every future cut, whether from an excavator, a vehicle strike on a pole, or a storm, adds a pair of splices and a length of replacement cable. A link built with no headroom passes on day one and fails after its first repair. We keep that in view when we recommend splice locations, and we say so when a design leaves nothing in reserve.

• Fiber attenuation across the full route length at both test wavelengths.
• Loss at every fusion splice, counted from the splice plan rather than estimated.
• Loss at every mated connector pair, including the patch panels at each end.
• Splitter loss where the architecture is PON.
• Reserve margin for future repair splices and replacement cable.

OTDR Testing: What It Proves and What It Misses

An OTDR shows you where events sit along a fiber and roughly what each one costs. It is the right instrument for locating a bad splice, a tight bend, a crushed section, or a break, and it is the only practical way to characterize a long span from one end. It is not a measurement of end to end insertion loss.

Insertion loss comes from a light source and a power meter, one at each end, referenced before the test. That pair of readings is what an optical link actually experiences. An OTDR infers loss from backscatter, which is why two directions disagree on the same splice and why the first and last connectors are difficult to see at all without launch and receive fibers.

Launch and receive fibers are not optional on an acceptance test. Without a launch cord, the near end connector sits inside the instrument dead zone and its loss is invisible. Without a receive cord, the far end connector has no fiber behind it to backscatter, so it cannot be measured either. Those two connectors are the ones most likely to be dirty.

Pulse width is the tradeoff you set before you press start. A short pulse resolves closely spaced events but does not reach far. A long pulse reaches the end of a long span but merges nearby events into one and can hide short sections entirely. We test long spans at more than one pulse width when the splice spacing calls for it, instead of accepting a single trace that averages the detail away.

We support your testing however your acceptance process is written. That can mean we test and deliver the traces, we test alongside your technician, or we make the plant ready and stand by while your team or a third party certifies it. In every case the traces, the splice records, and the fiber assignments have to agree with each other, and we check that they do before anything is sent.

Slack Storage Decisions You Live With for Years

Slack is the difference between a two hour repair and a two day one. When a cable is cut, the crew needs enough fiber on both sides to reach a splice trailer or a controlled work area, plus enough to make the splice without stressing the cable. Slack trimmed short to tidy a handhole is the reason a future outage runs long.

Where slack goes depends on the plant. Underground, it coils in a vault or handhole large enough to hold the coil at a radius the cable tolerates, secured so it does not rest in standing water or shift under a lid. Aerial, it stores on snowshoes or storage brackets sized to the cable, mounted where a bucket can reach it and where the span will not sag into a clearance problem.

Bend radius governs every coil. Cables carry a minimum bend radius for installation under load and a smaller one for long term static storage, and both come from the manufacturer, not from what happens to fit the enclosure. A coil forced into an undersized vault can read fine on the installation trace and slowly get worse. That is the failure that surfaces at 1550 nm months later.

We also place slack where it will be useful. That means extra length at road crossings and bridge attachments, at points where the cable transitions between aerial and underground, near equipment that will be replaced during the life of the plant, and at any location where a locate ticket is likely to bring a machine close to the cable. Slack in the right place is cheap. Slack in the wrong place is just weight.

• Coil diameter held above the cable long term bend radius, not sized to the enclosure.
• Enough length at each splice point to reach a splice trailer or other controlled work area.
• Additional slack at crossings, aerial to underground transitions, and future equipment sites.
• Aerial slack on correctly sized storage brackets, clear of the clearance envelope.
• Underground slack secured off the floor of a vault that will hold water seasonally.

What Good Enclosure Workmanship Looks Like

An enclosure either stays sealed for decades or it becomes the place your network fails. The difference is prep work you can see the moment the lid comes off: jackets stripped to the correct length, strength members anchored to the bracket instead of to a tray, buffer tubes routed on a radius rather than folded into place, and every tube landed in the order the splice plan calls for.

Sealing is a system, not a gasket. Gel sealed closures need clean, correctly prepared jackets and even compression across the seal. Heat shrink entries need the jacket abraded and cleaned, the sleeve centered, and heat applied evenly until adhesive flows all the way around the entry. Unused ports get plugged, not taped. Where the closure supports it, we pressure test before it goes into the ground or up onto the strand.

Bonding and grounding get skipped more often than any other step. Armored cable carries a metallic member that has to be bonded and grounded per the design, and ungrounded armor near a Florida lightning strike is a path toward your equipment. We bond it, and the closeout shows where.

Tray discipline is what makes the next technician's job possible. Fibers routed with their excess length parked in the tray, protection sleeves seated in their holders, tubes labeled, and trays loaded so the one you need is not buried under three others. A closure that has to be dismantled to reach one fiber turns a small repair into a risk to every other circuit in the case.

Labeling closes the loop. Cable identification at every entry, tray and buffer tube identification that matches the splice matrix, and enclosure identification that matches the as built. If the label in the field and the record in the file disagree, the record is worthless the first time someone works that plant at night.

Placing the Cable: Blowing, Pulling, and Aerial Handoffs

Splicing is only as good as the cable it joins, and most fiber damage happens during placement. Overtension, a bend taken around a sharp corner, or a jacket dragged across a broken duct edge produces attenuation no splice technique can recover. The placement method should be chosen from duct condition, route length, and cable design, not from whatever equipment happens to be on the truck.

Blowing, also called jetting, floats the cable on airflow so tension stays low across long continuous runs. It performs well in clean, proofed duct and microduct, and it can place a run in a single shot where pulling would need intermediate access points and the splices that come with them. It also depends completely on duct integrity, which is why we verify duct before a blow rather than during one.

Pulling is the answer for shorter sections, congested duct that already holds cable, transitions through vaults, and any run where a pull line is already in place. It demands monitored tension, the correct swivel and grip setup, sheaves and quadrant blocks at every direction change, and lubricant matched to both the jacket and the duct material.

Aerial placement adds its own handoffs. Cable is lashed to strand, then transitions into a closure on the pole or drops into a riser and continues underground. Those transition points are where slack, bend radius, and grounding decisions all land at once, and they deserve to be planned rather than improvised in the field. We build both the aerial and the underground side, so the transition is designed by the people who have to splice it.

Splicing in Florida Conditions

Humidity and heat are the working conditions here, not the exception. Fusion splicing wants a controlled environment, so production splicing happens inside an enclosed work area, such as a splice trailer or tent, with power, a clean bench, and protection from heat and humidity. Splicing on a tailgate on an August afternoon produces exactly the loss numbers you would expect from it.

Afternoon storms decide the sequence of the day. An open closure, an exposed cable end, and a bucket in the air are all the wrong places to be when a cell moves through. We stage work so an opened closure can be sealed before the weather turns, and so an exposed cable end is capped rather than left overnight.

Water is the other constant. A high water table and heavy seasonal rain mean vaults and handholes hold water for part of the year, and any closure placed in one has to be rated and sealed on the assumption that it will be submerged. Coastal routes add salt to the same problem, which shows on hardware, brackets, and pedestal bases long before it shows on the fiber.

Lightning is not a rare event on this coast. Bonding and grounding of armored cable, enclosures, and pedestal hardware is treated as installation work with a documented result rather than as a formality. That matters most where fiber shares a route with power and where a copper element runs alongside it.

Which Fiber Service Does Your Project Actually Need?

Most projects need more than one of these services, and the split depends on what is already built. A route with conduit in the ground and no cable needs placement first, then splicing, then termination. A route where cable is already placed and coiled at each structure needs splicing, enclosures, and testing. A cut cable needs all of it at once, compressed into a restoration window.

Placement method is the first fork. Blowing suits long, clean, empty duct and microduct, where low tension and fewer splice points pay off. Pulling suits shorter sections, vault transitions, occupied duct, and conduit that already holds a pull line. Aerial cable is lashed to strand, which is covered under our aerial construction work and handed off to splicing at the pole closures.

Splicing and enclosure work usually travel together, because the closure is where the splices live. Termination is the last step on each end of a span, where field fiber becomes a connector that equipment can plug into. Testing runs through all of it, and the acceptance method should be agreed before the first reel is staged.

• Fusion splicing: joining fibers at closures, panels, and repair points, with a recorded loss for every fiber.
• Fiber blowing: jetting cable into proofed duct or microduct over long runs at low tension.
• Fiber pulling: winch placement through conduit with monitored tension and correct rigging.
• Fiber termination: pigtail splicing and connectorization at panels, cabinets, huts, and demarcation points.
• Splice enclosure installation: closure prep, sealing, bonding, mounting, slack, and labeling.

How Does Splicing Fit Into the Rest of an Outside Plant Build?

Splicing is the step that turns separate cable sections into a working network, so it sits near the end of the build and inherits every decision made before it. Bore depth, conduit routing, handhole placement, aerial attachment heights, and slack left at each structure all show up at the splice bench. A splice crew working behind a drilling or aerial crew it has never spoken to is where schedule slips start.

We build the drilling, conduit, aerial, and fiber pieces ourselves, which means the people who place a handhole know the closure that will sit in it and the slack it needs. That is not a guarantee every structure ends up where the plan put it. It does mean the handoff between placement and splicing is a conversation inside one company rather than a dispute between two.

Sequencing matters for testing as well. Spans can be tested as soon as they are spliced, before the next section is complete, so a problem in one segment shows up while the crew that built it is still nearby. Waiting to test the whole route at the end compresses every fix into the last week of the schedule, right when the owner wants to turn up service.

The splice plan is the document that ties it all together. It should exist, match the cable that was actually ordered, and be approved before splicing starts. When it changes mid project, which happens, the revision should be dated and issued to the crew in writing so the field record and the design record stay the same document.

What You Receive at Closeout

A splice job is not finished when the last fusion is made. It is finished when your records let someone who was never on site find the splice point, open the right tray, and identify the right fiber. That package is part of the work, not an extra line item at the end of it.

The splice record is the core document: cable to cable fiber assignments, tray and position, the loss reading for each fiber, and the technician and date. It is the file your NOC opens at two in the morning, so it has to match the labels in the field exactly.

Test results travel with it. Bidirectional OTDR traces at the wavelengths your specification calls for, delivered in native trace format so your engineers can reopen them and in a readable summary for the project file, plus light source and power meter insertion loss results where your acceptance process requires them.

The rest is location and condition evidence: enclosure photographs before and after closing, coordinates or stationing for every splice point and vault, slack storage locations and lengths, as built redlines showing what was actually placed, and daily production reporting through the job so the closeout is a compilation rather than a reconstruction.

• Splice matrix with per fiber loss values, tray positions, and cable identifiers.
• Bidirectional OTDR traces in native format plus a readable test summary.
• Light source and power meter insertion loss results where specified.
• Enclosure and slack storage photographs, before and after closing.
• Coordinates or stationing for every splice point, vault, and slack coil.
• As built redlines and daily production reports compiled through the job.