Fiber & Outside Plant

How It Works & Technology Overview

Outside plant (OSP) fiber is civil construction that terminates in glass: getting cable between buildings, across campuses, or along rights-of-way, then splicing and proving it against a loss budget. Three placement regimes govern unit economics. Existing pathway (usable conduit, innerduct, or aerial strand): pull labor plus splicing — the best case, near premise-work predictability. Aerial new-build: pole attachment agreements with the pole owner, make-ready engineering (moving existing attachments, on the utility's schedule and price), then strand and lash. Underground new-build: trenching or directional boring — highest cost and highest uncertainty per foot, because subsurface conditions are only ever probabilistically known. Fusion splicing joins glass at ~0.02–0.05 dB per splice; single-fiber and ribbon (mass fusion, 12/24 strands per cycle) machines have very different throughput, so cable construction determines splice economics. Acceptance is bidirectional OTDR characterization plus power-meter loss testing against the electronics vendor's end-to-end budget — "it lights up" is not a test standard. Legal prerequisites shape everything: 811 utility locates (public), private locates for campus-owned utilities (the public ticket does not cover them), permits for public right-of-way, and railroad/water crossings that run on their own multi-month timelines.

Common Deployment Scenarios

• Campus interconnect: building-to-building links on private property; private locates, existing duct opportunism, splice cases in vaults or building entrances.
• Aerial route new-build: pole attachments along a corridor; agreement + make-ready timeline dominates the schedule.
• Underground new-build: directional bore under roads/parking, trench across soft ground; surface restoration to the owner's standard is its own scope.
• Ring/redundancy builds: diverse-path campus rings; the second path is only diverse if it doesn't share the same conduit bank, bridge, or pole line.
• Emergency restoration: cut repair (dig-in, storm damage) — premium rates, temporary splice-through, then scheduled permanent repair.

Products, Vendors & Characteristics

• Cable: Corning, OFS, Prysmian, Superior Essex — loose-tube gel-free dominates OSP; ribbon for high counts; armored where rodents or direct-bury demand it. Single-mode (OS2) is the default; strand counts should be sized for the 20-year need, not the RFP.
• Splicing equipment: Fujikura, Sumitomo, INNO fusion splicers; single-fiber vs. ribbon machines are different tools and different labor math.
• Closures/hardware: Corning, CommScope, Preformed Line Products splice cases; slack storage (30–50 ft loops at each splice point) is cheap during construction and priceless later.
• Testing: EXFO/VIAVI OTDRs and loss test sets; deliverables are bidirectional traces with per-splice loss documented per strand.
• Construction: directional boring rigs, trenchers, micro-trenching (shallow, pavement-cut — fast but jurisdiction-dependent); conduit standard is HDPE with innerduct/microduct for future pulls.

What Makes It Hard (Gotchas & Failure Modes)

• The ground lies: locate paint shows registered utilities, not reality; unmarked private utilities (irrigation, site lighting, old abandoned duct) cause the strikes — private locates on campuses, always.
• Rock: a bore that hits granite triples in cost; in known-rock geography the contract's rock clause decides who eats it — its absence is the mistake.
• Existing conduit is a rumor: collapsed, silted, or packed full; prove it with a mandrel/fish before pricing pull-through rates.
• Make-ready is someone else's project: pole owners perform it on their schedule at their price; carry it as calendar and pass-through cost, not as our labor.
• Crossings: railroad and water/DOT crossings are their own permits measured in seasons — route around them when remotely possible.
• Restoration expectations: "restore to match existing" on decorative surfaces is a different number than hydroseed; photograph pre-condition, price restoration per surface type.
• Weather/seasonality: winter mud cuts placement productivity ~30% vs. dry-season rates; name the season in the estimate assumptions.
• Unconfirmed loss budgets: splicing to a budget nobody verified with the electronics vendor risks failing acceptance against a spec never seen.

Effort & Estimation Patterns

• Estimate per splice point, not per splice: case prep, cable entry, and closure run 2–3 hours per splice point before the first fusion; each single-fiber fusion then runs 15–20 minutes prepped; ribbon/mass fusion lands 12–24 strands per cycle and changes the math.
• OTDR characterization & documentation: ~a day per segment pair for proper bidirectional test-and-report at moderate strand counts.
• Underground placement: per-foot from ground truth (soil, surface, congestion) — never from a state average; bore setup/teardown is a fixed half-day-plus, making short bores disproportionately expensive per foot.
• Aerial: predictable per-pole labor for strand/lash once attachments exist; the agreement + make-ready timeline is calendar risk, not hours.
• Route survey before commitment: walk or drive every foot; desktop estimates from aerial imagery are hypotheses. Survey time: 1–2 days per mile in developed areas including documentation.
• Slack & expansion: 30–50 ft slack loops per splice point and spare capacity in every closure — trivial cost now, enormous cost later.
• Restoration: 40% dig / 60% putting the world back is a fair planning heuristic on finished surfaces — price it per surface type as its own line.
• Contract structure: separate our labor commitments from third-party calendars (pole owners, permits, railroads) or their delays become our liquidated damages.