ACT · Act now — correcting soonHydrocyclones →Cost of fault ≈ $100k+ / incident
Why it matters
A roping apex isn’t separating — it’s a rope of solids and fluid with no air core. Roping or plugged cones send drilled solids straight back to the active system: exactly the fraction the cones exist to remove.
Likely causes
Low feed head — manifold pressure too low; below ~75 ft the cones can’t form the air core.
Worn feed-pump impeller / volute — head quietly lost even though the pump “runs.”
Apex too small for the load — overloaded, it ropes instead of sprays.
Cone plugging — large particles or trash lodged at the apex.
Too many cones for pump capacity — head split too thin.
Suction air leak — pump losing prime/pressure.
How to diagnose it
Calculate feed head: h = 19.2 × p ÷ ρ (p in psi, ρ in ppg). Below ~75 ft is your problem.
Read the apex: umbrella spray = healthy; rope = overload or low head; dry/intermittent = plugging.
Check manifold pressure against design and trend it for drift.
Inspect the feed pump — impeller/volute wear is the usual hidden cause.
Count cones running vs pump capacity.
The fix — step by step
Restore feed head above ~75 ft first — most roping is a head problem. Repair/replace the feed-pump impeller.
Open or replace the apex to suit the load (umbrella, not rope).
Clear plugged cones; fit a trash screen/desander upstream if trash recurs.
Match cone count to pump capacity — don’t run more cones than the pump can feed.
Confirm it's fixed
✓ Verify: Every cone showing a steady umbrella spray with a visible air core, manifold pressure giving ≥75 ft feed head, and a discard that’s solids-laden — not whole mud. Re-check head after any pump or cone change.
Field note. Running is not cutting. The single most expensive solids-control fault I see is a cone bank “running” at low head — everyone sees the cones spinning, nobody calculates the head, and fines quietly recirculate for a whole section. Two minutes with h = 19.2p/ρ settles it.