Winch Selection for Tidal Tension-Leg Mooring

Slow line take-up / pay-out at 2,000–3,000 lb working tension, quiet enough to live above, sized for 3 ft tides and beyond.

Short answer: yes — a self-locking worm-gear winch is close to ideal.

The “boat-lift hoist” class of electric worm-gear winch is almost purpose-built for this duty: it holds a multi-thousand-pound load for months with zero power and no brake (the worm cannot be back-driven), it is nearly silent at creep speed, it is cheap, marine-rated, and available in 12/24 V DC. Pair it with a high total reduction (roughly 3,000–5,000:1), a small brushless DC motor, an inline snubber/compliance element, and deadband control so it only moves a few inches every few minutes — the tide is so slow that the winch is essentially always parked and self-locked.

Best alternatives: a self-locking Acme-screw linear actuator (if you ever cap tidal range at ~2 ft of line travel, or use a 2:1 bridle), a harmonic-drive winch (premium, whisper-quiet, custom), or a passive hydro-pneumatic tensioner (silent and power-free, but bigger and costlier).

1. Why the worm-gear winch fits this exact job

Your requirement	How the worm winch answers it
Hold 2,000–3,000 lb for weeks/months	Self-locking gear set: load cannot back-drive the motor. No holding brake, no holding current, no heat, no noise while parked.
Move extremely slowly (tides)	Tide-following needs only 0.05–0.6 in/min. With ~4,000:1 total reduction the motor turns a few tens of RPM — or you jog it in small steps.
Quiet (bedrooms above)	At these speeds a greased worm is effectively silent. No ratchet/pawl clicking (choose a pawl-less unit), no brake engage/disengage clunks.
Safe on power failure	Fail-safe by physics: a dead winch simply freezes in place. Add a snubber so a frozen winch during a rising tide is an hours-scale problem, not a minutes-scale one.
Marine environment	Boat-lift hoists live outdoors over salt water by design; sealed gearcases, 12/24 V motors, hand-crank backup options.
Fits redundancy philosophy	Small and cheap enough for one per corner plus a spare in the container; each can run off that leg’s independent inverter/battery bus.

Nuance: worm gears are self-locking when efficiency < ~50% (ratios ≥ ~40:1 with a small lead angle). Specify the worm stage for static self-locking; if you worry about vibration slowly “walking” it, add a small spring-applied holding brake as belt-and-suspenders.

2. How slow is “slow”? (the numbers that drive the design)

Maximum tidal rate of rise for a semidiurnal tide ≈ (range/2) × 2π/12.42 h. With a 6 in diameter grooved drum (1.57 ft of line per revolution) and ~4,000:1 total reduction:

Tide range	Max rate of rise	Drum speed	Motor speed @ 4,000:1	Comment
1.5 ft (Anguilla)	~0.08 in/min	1 rev / ~4 h	~17 RPM	Jog a few seconds every several minutes
3 ft (design point)	~0.15 in/min	1 rev / ~2 h	~32 RPM	Creep or jog — either is inaudible inside a leg
6 ft	~0.30 in/min	~0.016 RPM	~64 RPM	Still trivial
12 ft (“interesting places”)	~0.61 in/min	~0.032 RPM	~130 RPM	Continuous creep is fine
30 ft (Brittany / Cook Inlet class)	~1.5 in/min	~0.08 RPM	~320 RPM	Still slow for machinery

Key insight: the required speeds are so low that you should not try to gear for continuous motion at full motor RPM. Instead, gear 3,000–5,000:1 and either (a) run the motor slowly, or (b) jog: move 1–2 in of line, stop, self-lock, wait. Total line travel for a 3 ft semidiurnal tide is only ~12 ft/day — any winch on Earth has the duty cycle for this.

Deployment speed: the same gearbox at full motor speed (~3,000 RPM) gives ~1–1.5 ft/min, so hauling 3 ft of pretension takes ~2–3 minutes, and winding a 100 ft scope takes ~1.5 h. Since deployment is rare, that is acceptable; otherwise preset the scope by hand and let the winch manage only a tensioning pendant (see §7).

3. Winch types compared

Type	Self-locking?	Noise at creep speed	Efficiency	Cost	Verdict
Worm-gear drum winch (boat-lift class)	Yes — no brake needed	Near-silent	40–70%	$	Recommended
Planetary winch + spring-applied brake	No (brake holds)	Silent parked; gear whine moving; brake clunk	85–95%	$$	Viable — brake is an extra failure point
Harmonic-drive winch (custom)	Usually (not guaranteed)	Near-silent, zero backlash	70–85%	$$$$	Premium option
Cycloidal-drive winch	Partially	Very quiet	80–90%	$$$	Viable
Acme-screw linear actuator (tensions a bridle)	Yes (Acme thread)	Silent	30–50%	$$	Good if stroke is enough — e.g., 3–4 ft actuator on a 2:1 bridle = 6–8 ft of line travel
Capstan + separate line storage	No	Quiet	—	$$	Unnecessary complexity
Hydraulic winch	No (valve/brake)	Pump noise	—	$$	Avoid — no HPU aboard, leaks
Offshore chain jack	Yes (grippers)	Clunks loudly	—	$$$$	Overkill — the industry big brother, wrong scale
Passive hydro-pneumatic tensioner (riser-style)	n/a — no motor at all	Truly silent	n/a	$$$	Interesting alternative — follows tides and waves with zero power; see §8

Search terms for off-the-shelf hardware: “electric worm gear winch 12/24V”, “boat lift hoist worm drive”, “worm gear power winch grooved drum”. Representative manufacturers to survey (not endorsements): Thern, Jeamar, Bloom, Dutton-Lainson, Fulton, plus boat-lift hoist brands. Many sell the gearbox/drum without a motor so you can bolt on your own BLDC.

4. Recommended per-corner unit (specification)

Mechanical

Type: electric worm-gear winch, self-locking, no ratchet pawl
Rated line pull: ≥ 9,000 lb on first layer (3× working tension)
Drum: 6 in dia, grooved, single layer, ≥ 20 ft of 1/2 in line (~7 in wide) — plus 3–4 dead wraps
Reduction: ~4,000:1 total (e.g., 40:1 planetary or quiet belt primary → 100:1 worm)
Drum torque at working load: ~1,500 ft-lb (6,000 lb bridle case)
Line: 1/2 in double-braid polyester, MBL ~11,000 lb (see §6 on why not HMPE)
Compliance element: inline rubber snubber stack or disc-spring stack, ~±6 in travel over ~1,000–6,000 lb (≈400–500 lb/in)
Manual backup: hand crank on worm input shaft

Electrical & control

Motor: 24 V BLDC, ~1/2 HP (only ~1 N·m needed after reduction), IP67, FOC drive for silent low-RPM torque
Sensors: load pin at fairlead sheave (or inline tension link), multi-turn absolute encoder, travel limits, motor temp
Average power: < 10 W while tide-following; 0 W while holding
Controller: per-leg node on that leg’s power bus (matches your triple-redundant power architecture); networked to central computer
Mode set: Deploy / Retrieve / Set-pretension / Tide-follow / Storm / Manual
Marinization: sealed gearcase (synthetic grease), 316 shaft & fasteners, epoxy coat, galvanic isolation from aluminum structure, desiccant or 5 W heater in the compartment

5. Control strategy: deadband, not constant-tension

Industrial “constant tension / render-recover” winches follow every wave — exactly what you do not want (power draw, wear, and motor noise day and night). With your numbers, 1 ft of waterline change ≈ 1/7 of total buoyancy ≈ 3,900 lb, so waves swing line tension by thousands of pounds at wave frequency. Let the snubber + polyester stretch absorb waves passively, and let the winch chase only the tide:

every 60 s:
    T = moving_average(tension, last 3–5 min)   // filters out waves
    if   T > T_HIGH:  pay_out(STEP)                // e.g., 1–2 in of line
    elif T < T_LOW:   haul_in(STEP)
    else:              sleep                        // worm self-locks, ~0 W, silent

Example setpoints: target 3,000 lb; T_LOW = 2,400 lb; T_HIGH = 3,600 lb; snubber mid-stroke at target.
STEP small (1–2 in) with a rate limit → motion is smooth and inaudible; adjustments happen a handful of times per hour at most.
Optional feed-forward: you know the tide prediction — pre-position along the predicted curve and let the deadband trim the error. Fewer, earlier, smaller moves.
Storm mode: raise pretension so troughs never unload the tethers (pull-down should exceed expected wave height; 2 ft seas ≈ ±7,800 lb total swing by your 1/7-per-foot figure).
Alarms: sustained high/low tension, encoder–load-cell disagreement, snubber near end of travel → notify crew phones.

6. The tether line matters as much as the winch

Use polyester double-braid, not HMPE (Dyneema/Spectra). HMPE creeps under permanent static load — bad for a tether tensioned for months. This is exactly why deepwater taut-leg moorings use polyester.
Polyester’s elasticity is a feature. At 3,000 lb on 1/2 in line (~27% of MBL) it stretches ~4–5%; on a 40 ft tether that is ~1.6–2 ft of compliance — free, silent, zero-maintenance wave absorption that widens your deadband.
Avoid nylon for the permanent section (creep, ~15% wet-strength loss) though a short nylon or rubber snubber section near the top is a fine compliance booster.
Fairlead and sheave diameter ≥ 8× line diameter; drum grooved to line size; at these speeds a level-wind is unnecessary if the fleet angle is reasonable.
Sacrificial chafe sleeve where the line passes the fairlead; inspect on schedule.

7. Rigging the paired screws at each corner

One winch can serve both helical screws with an equalizing bridle (recommended: self-equalizes tension between the two anchors):

seabed:  [screw A]----\
                       bridle == [equalizer] == pendant == fairlead on leg == [snubber] == [WINCH in dry upper leg compartment]
         [screw B]----/

Bridle is 2:1: drum sees 2 × line tension (6,000 lb) but moves 2× the line per revolution — halves your speed requirement, hence the 9,000–10,000 lb drum rating above.
Alternative: twin-groove drum, one line per screw, opposite leads — simpler rigging, slightly uneven tension absorbed by the snubbers.
Placement: put the winch in a dry, vented compartment high in the leg (one of your airtight compartments above the waterline). Water + foam + structure isolate what little sound remains. Route the pendant up the trailing edge alongside the thruster-wire conduit, entering via an above-waterline cable gland — so it is not a below-waterline through-hull.
Scope strategy: in deeper water, set the gross scope once at deployment (shackle the right length at the anchor or a chain locker) and let the winch work only a 20–30 ft tensioning pendant — keeps the drum a compact single-layer unit.

8. Alternatives worth keeping on the table

Passive hydro-pneumatic tensioner (riser-tensioner style): a cylinder + accumulator carries the load and stroking in/out follows tides and waves continuously, silently, with zero power and zero control software. Size stroke = tide range + wave allowance (a 4:1 purchase shrinks cylinder stroke 4×). Cost and nitrogen maintenance are the downsides; it is the most elegant “set and forget” solution and can also be added later as an upgrade in series with the winch.
Ballast-trim tide following: instead of changing line length, pump seawater in/out of the legs. Pumping in ~3,900 lb total (~1,300 lb ≈ 160 gal per leg) per foot of tide rise holds tether tension constant with a fixed-length tether — a slow 0.5 gpm pump is silent and sips power. Cost: freeboard drops foot-for-foot with the tide. Attractive for Anguilla-class microtides (winch may never need to move), and a nice layered backup everywhere else.
Softer tether = smaller winch duty: more polyester scope (or an added snubber) widens the deadband so the winch steps less often. Cheap insurance.

9. Failure modes & redundancy

Failure	Behavior / mitigation
Power loss	Worm self-locks; tension frozen. Snubber + line stretch give hours of margin across a tide cycle; alarm fires; hand-crank backup.
One winch seized during rising tide	Other two corners pay out to shed load; platform heels slightly. Size tethers, bridle and attachments for ≥2× working load. Triple-redundant legs again mirror your power architecture.
Wave spike loads	Self-locking winch cannot render — by design. Spikes go into the snubber/line elasticity, not the gearbox. Confirm structure + anchors for snubber-limited peak (~9–10 kip).
Sensor failure	Encoder vs. load-cell cross-check; fall back to position-scheduled tide-following from predicted tide tables (open-loop is fine for hours).
Line chafe / snubber aging	Sacrificial sleeve, scheduled inspection, monitor snubber travel drift. Whole winch is small enough to carry a full spare — fits your container logistics philosophy.

10. Assumptions & open questions

Assumed 2,000–3,000 lb is per corner; if it is total, everything above gets easier. Pulling down 3 ft ≈ 12 kip total by your 1/7-per-foot figure → ~4 kip per screw pair plus dynamics; verify helical screw capacity for the actual seabed (great in sand/mud, poor on rock).
Intended water depths decide whether the drum stores full scope or just a tensioning pendant (§7).
If you want freeboard to stay perfectly constant for dinghy/dock convenience, favor winch-following; if you want absolute silence and simplicity in microtide areas, favor ballast-trim (§8). The two combine well.
Confirm the worm stage is specified statically self-locking; add a small holding brake only if vibration testing shows creep.

Bottom line: buy (or build around) a 24 V self-locking worm-gear winch in the 9–10 kip class with a grooved single-layer drum, drive it with a small BLDC through ~4,000:1 total, put a snubber in the line, and run deadband control off a load pin with a 3–5 minute average. It will hold station for weeks at zero watts, move a few inches at a time slower than the hour hand of a clock, and be inaudible from the living space.