High-Tension Pull Through a Reused Innerduct

No - stop the pull.

• 520 lbf against a 600 lbf short-term limit leaves only 13% headroom, and the limit applies to the peak, not the average.
• Tension at a bend rises roughly exponentially with wrap angle and friction (capstan effect). The rise you are seeing at 140 m is the signature of friction building through the sweeps already passed; the remaining 40 m of route plus at least one more sweep can spike the peak past 600 lbf faster than the capstan operator can react.
• Exceeding short-term tension does not always break the cable. It stretches fiber (excess attenuation, and often loss that appears months later) and can neck the jacket. The failure is silent until test.

Corrective actions:

1. Stop, hold the load, and relieve tension; never let the capstan slip back and whip the cable.
2. Add approved lubricant at the mid-route access point/vault, not only at the entry.
3. Back-feed: have personnel feed cable forward at the entry while the capstan takes up slack, so you are not dragging dead weight through the sweeps.
4. Consider pulling from the opposite direction, or adding an intermediate pull point with a figure-eight and a second pull.
5. Verify the pull sheave/roller at the duct mouth keeps the cable above minimum bend radius during the pull (roughly 10x OD, ~4.3 in here) and that nothing is rubbing a conduit edge.
6. Log the peak tension on the pull record; if it ever exceeded 600 lbf before you noticed, OTDR the whole cable before accepting it.

Conclusion: mechanical damage at a sweep caused by sidewall friction under high tension, not a bend-loss issue.

Reasoning:

• A macrobend at 1310 nm in single-mode typically shows small loss; 0.9 dB at a single point is consistent with jacket/coating damage or a kink, and the scuffing at the same distance corroborates it.
• The damage location (118 m) is inside a sweep, which matches a tension/lubrication problem rather than a manufacturing defect.

Confirm and act:

1. Test bidirectionally - a one-direction reading can be inflated or deflated by backscatter mismatch. Average the two directions to get the true splice/event loss.
2. Inspect the jacket at 118 m physically. If the jacket is breached, moisture protection is compromised on an OSP-rated loose-tube cable, so the affected section must be removed, not just re-tested.
3. Remedy: cut out the damaged section and splice in a closure with proper slack storage, or replace the cable if more than roughly 10% of fibers are affected or if the damage is under a duct you cannot open.
4. Re-test all 24 fibers and record the repair in the as-built documentation.
5. Fix the cause before the next pull: lubricate the full route, add feed assistance, verify sweep alignment, and monitor tension continuously rather than at one gauge.

Fill check (necessary but not sufficient):

• Cable area = pi x (0.215 in)^2 = 0.145 in^2
• Innerduct area = pi x (0.625 in)^2 = 1.227 in^2
• Two cables = 0.290 in^2 = 23.6% fill - under a typical 40% design limit, so fill alone says go.

But fill is not the whole answer:

• Jam ratio (innerduct ID / cable OD) = 1.25 / 0.43 = 2.9, which sits inside the 2.8-3.1 band where equal-diameter cables are prone to jamming in bends - and this route has three 90-degree sweeps. Two identical cables in that ratio is the worst case.
• A second cable also doubles the friction of the pull and means one pull can damage an already-installed, already-tested cable.
• The first cable already showed damage under tension in this same route.

Recommendation: install a second innerduct and pull one cable per innerduct, or upsize the innerduct so the jam ratio moves out of the critical band. Do not simply stuff a second cable into the existing 1.25 in innerduct on this route. If the customer insists, the pull must be re-engineered with full lubrication, tension monitoring, and OTDR re-acceptance of the existing cable afterward.