1.2 km 10G Link: Multimode Bid vs. Single-Mode Reality
A regional contractor is bidding a 1.2 km 10 Gigabit Ethernet link between an administration building and a new warehouse. The route uses existing outdoor duct with three handholes and an entrance conduit at each end, and the customer expects service in six weeks. The estimator priced 12-strand indoor/outdoor OM4 tight-buffered cable at $0.95/ft and 10GBASE-SR SFP+ modules, arguing that multimode is cheaper and that our crew only splices and tests multimode. The crew is two CFOS/C-certified technicians who work almost exclusively at 850 nm plus three helpers whose background is copper only. A single-mode alternate adds roughly $1,400 in cable and $600 in optics, but the crew would need 1310/1550 nm fusion splicing and test capability, and the single-mode cable has a four-week lead time. The duct is already about 40% filled by two other cables, gel-filled cable is prohibited by the facility, and the pull will happen in February at about -5 C.
1. Evaluate the estimator's fiber and optics proposal. What standard distance limit applies, and what physical effect governs the outcome?
Read the worked answer
The multimode proposal fails on distance, not on loss. 10GBASE-SR is specified to 300 m on OM3 and 400 m on OM4 at 850 nm, and this 1.2 km route is roughly three times the OM4 limit. The governing effect is modal dispersion: many modes travel slightly different path lengths and arrive at different times, spreading the pulse until the receiver cannot resolve symbols. The Ethernet distance table is really a bandwidth statement derived from the fiber's effective modal bandwidth (about 4700 MHz*km for OM4) and the source's spectral width; extra loss margin cannot buy back bandwidth. A link like this can pass a perfectly clean loss test and still fail at 10G, which is exactly the trap in the estimator's reasoning. 10GBASE-LR over OS2 at 1310 nm is specified to 10 km, giving roughly 8x margin on length and headroom for future 25G or 100G LR optics. I would reject OM4, accept the ~$2,000 premium, and state in the submittal that acceptance is by both loss and bandwidth-appropriate design. I would also note the duct is already 40% full and gel-filled cable is barred, so the single-mode cable must be dry, low-friction, and sized for that fill.
2. The project manager asks whether the existing crew can deliver the single-mode option in six weeks without new hires. What competency and schedule plan do you give?
Read the worked answer
Single-mode is a different skill set, not just a different cable, so treat it as a competency plan rather than a purchase. Before any production splice, each splicer should demonstrate 1310 nm fusion splicing at an average of 0.10 dB or better and 0.15 dB maximum on a practice drum, plus acceptable cleave angle and a bidirectional OTDR technique using a proper launch fiber. The two CFOS/C techs can be crossed-trained in one day on arc parameters, cleave quality, and connector inspection, but plan a supervisor check on the first ten production splices. Equipment must be verified: the OTDR needs 1310/1550 modules and current calibration, and you need a 1310/1550 light source and power meter plus an inspection scope. Procurement is the real critical path, because a four-week cable lead means ordering at award, not after submittal approval. The helpers stay on pulling and cable management, not termination. As a compression option, price pre-terminated single-mode assemblies, which remove field splicing entirely and cut both schedule and skill risk.
3. What installation precautions does the February pull at -5 C require, given the duct fill and the facility's no-gel rule?
Read the worked answer
Cold-weather pulling changes the mechanics of the cable, so I would not treat this as a routine pull. Many indoor/outdoor tight-buffered constructions have installation temperature ratings near -10 C, so at -5 C the cable is close to its limit: store it indoors or in a heated trailer for at least 24 hours before pulling, keep pull speed steady and low (under about 30 m/min), and avoid shock loads, sharp bends, and kinking at handholes. Confirm the manufacturer's minimum bend radius and enforce it with sheaves and rollers at each handhole, not by hand-feeding around corners. Because gel is prohibited, use dry, low-friction cable with a suitable pulling lubricant verified as compatible with the jacket and with the existing duct, and pre-lubricate on a 1.2 km route. The duct is already 40% full, so calculate the actual fill with the new cable before committing; if it exceeds the allowed fill or the pull tension climbs, stop rather than pulling harder. Use a dynamometer and a breakaway link set to the cable's rated tension, and leave the cable ends sealed to protect the dry water-blocking construction.