Coiled-Tubing Heave Compensation:
Holding the Tension Window
Riserless coiled tubing is the most fragile thing anyone hangs off an intervention vessel. A 2⅜-inch string 400 m long yields at 53 tonnes and buckles if it ever goes slack — so it has to be held inside a tension window at both ends while the ship above it moves with every wave.
The string is the governing fragility
400 m of 2⅜″ × 0.190″ CT-90-class tubing has an axial stiffness of EA ≈ 174 MN. Over that length that is 435 kN of tension per metre of stretch — so ±1.5 m of uncompensated relative motion asks the string for ±650 kN, past its 522 kN yield in a single wave.
What happens with no compensation
Hang the injector platform rigid in Hs 3.0 m / Tp 9.0 s and the string takes the full motion of the ship. The sea asks it for 1,308 kN against a 522 kN yield, and it parts about fifteen seconds into the job. Across twelve wave realisations of that sea state the rigid hang-off parts the string in every one, between 8.7 s and 32.5 s.
The worked example as a film — the tension window, the stuck-pipe overpull and the parted-string arrest, rendered frame-for-frame from the CONSTELLATION simulation.
What the compensator has to do
Through one ANTARES 125 t / 5.5 m on its softest charge, the same waves move the tension 93–146 kN about a 118 kN setpoint — inside ±3 tonnes — with zero slack events, while the unit absorbs 4.5 m of relative motion.
The mechanism is stiffness, not force: a gas spring soft enough that a metre of ship motion barely changes the pressure, in series with a string that is 435 kN/m stiff. The soft element takes the motion; the string keeps its tension.
Stuck pipe: the same unit, a different working point
When the tubing sticks, the recovery pull has to rise without the window opening. An adaptive passive unit retunes its gas working point in place: the string then carries a 42.6 t recovery pull — the 80 % working line — with wave peaks reaching 89 % of yield, window still held, no slack. No re-rigging, no second unit.
The parted string, and why the end of the stroke matters
The accident the industry actually fears is the string parting at maximum overpull. Forty-two tonnes of tension vanish in a heartbeat and the platform is thrown upward — 3.2 m in about a second on the realisation shown.
| Parted-string arrest | Shown | Worst of 12 | Limit |
|---|---|---|---|
| Contact with anything hard | none | none | ≤ 0.5 m/s |
| Peak deceleration | 2.1 g | 3.7 g | 5.0 g |
| Main-fall peak | 0.61 × SWL | 0.90 × SWL | 1.00 × |
The compensator must damp the end of its own retraction. With that function the platform is arrested without reaching anything hard. Without it, the same platform on the same wave reaches the end of its travel at speed — a critical failure rather than a controlled stop.
The main fall is sized by the parting, not by the lift. It goes momentarily slack as the platform is thrown up, then catches the load again. A 36 t platform on a routine day would need a fraction of that rating.
Where the envelope closes
Stroke demand runs at roughly 1.6 × Hs for this arrangement, so a 5.5 m stroke puts the operability ceiling near Hs 3.1–3.4 m — which is why Hs 3.0 m is the honest “workable but demanding” design point on the NCS, and why stroke, not static load, is usually what sizes the unit.
Frequently asked
Send us your case
String size and grade, water depth, vessel and sea state. We return a worked example like this one — the window through the hold, the overpull case, the parted-string arrest, and the ensemble behind every number.
Talk to an engineer