Auto-Screw Mooring Unit (ASU) — Feasibility Review & Recommended Design

Tension-leg mooring system for a 44-ft triangular, three-foil seastead | Feasibility-level engineering (not a substitute for a stamped detailed design)

Verdict: Yes — this is engineerable at a reasonable price. The ASU is a marriage of three mature, well-understood technologies: (1) helical screw anchors (150+ years of use, with a proven torque↔capacity correlation that gives you a built-in proof test on every install), (2) Kelly-bar rotary drilling (hex shaft sliding through a drive bushing — exactly your concept), and (3) submersible gearmotor drivetrains (wastewater mixer / borehole-pump technology, IP68, cheap, proven). Nothing here requires invention — only careful integration, corrosion engineering, and testing. The main risks are operational (very soft sand, biofouling on retrieval, thruster scour, snatch loads), and all are manageable with the design features described below.

Screw pair spacing

36 in (0.9 m) c-c

Motor power per ASU

4–6 kW total (2 × 2–3 kW)

Active screw-in time

~2–5 min (pair)

Full evolution per unit

~15–30 min

Cost per ASU @ qty 60

≈ US$6–9k FOB China

Cost per seastead (3 ASUs)

≈ US$22–32k landed

Prototype cost

$6–12k land rig / $25–40k in-water

Engineering package

$30–80k, ~6–12 months

1. System Concept & Operating Sequence

1.1 Components (per ASU)

Base frame — 316L stainless weldment, ~4.5 × 3.5 ft footprint, four legs with sand pads and short skirts. Reacts the differential drive torque, keeps the unit upright, carries the camera/lights.
Two helical screws — one right-hand helix, one left-hand helix, so counter-rotating them drives both downward while their torques largely cancel in the drive carriage. Solid 2205 duplex hex shafts.
Drive carriage — two submersible gearmotors (recommended) or one motor + splitter gearbox, each output ending in a long-engagement hex drive bushing (your "Kelly bushing") that slides along the shaft as the screw advances. Carriage rides a short guide mast on the frame for vertical travel (~5 ft).
Load clamps ("slips") — wedge-type clamps that grip the hex flats anywhere along the shaft, self-energizing under upward pull, releasing when slacked. This solves the real problem that the screws will stop at different, unpredictable heights (dense vs. soft patches), so a fixed-height attachment won't work.
Equalizer beam — the lift line connects to a small rocker beam spanning both clamps, so each screw automatically carries ~half the 3,500-lb tension.
Floats — one closed-cell float (30–50 lb buoyancy) near the top of each screw, as you specified, keeping shafts vertical during deployment and marking them for the camera.
Lift/power umbilical to the surface — an 18 mm polyester double-braid rope (strength member) plus a PUR-jacketed subsea power/signal cable clipped alongside with slack loops. A 10-ft nylon snubber section at the top absorbs dynamic snatch. Using synthetic rope — not wire — is also your galvanic isolator between the duplex ASU and the aluminum seastead.
Corner winch on the seastead — 5,000-lb WLL electric winch with fail-safe brake, load-cell pin, and level-wind, mounted in a weathertight enclosure at each corner; stows horizontally in a rubber-lined cradle as you described.

1.2 Installation sequence

Seastead points into wind/waves, holds station on GPS with thrusters at low power.
Winch lowers the ASU to the sand (~15–50 ft). Operator watches the ASU camera; frame lands, floats keep the screws vertical.
Operator starts each drive motor individually for a few seconds (confirm both screws bite), then runs both together. The carriage walks down the shafts as the screws advance.
The VFD logs motor current → torque. Acceptance criteria: stop at target depth (~8 ft) and final torque ≥ ~500 ft-lb per screw. If torque < ~300 ft-lb at full depth → sand too soft → recover, move, retry. If refusal above ~5 ft → hard layer/rock → move. This torque reading is the proof test — log it per install for insurance.
Winch takes up: the slips self-engage on the hex shafts, the equalizer splits the load, and the winch tensions to 3,500 lb (read on the load pin). Thrusters can now stand down.
Repeat for the other two corners (~15–30 min each; can be run in parallel with three operators/cameras, ~30–40 min total).

1.3 Removal

Slack tension (slips release), reverse the motors. Screws back out cleanly — this reversibility is the big advantage of helical screws over deadweight anchors. Breakout torque spikes briefly, then unscrewing is easy.
Carriage rises as the screws extract; hoist the unit, rinse, stow in cradle.
Contingency: if a screw is fouled/stuck, torque + steady winch pull together almost always free it. Worst case, unclamp and abandon one screw (~$700–900 part) — cheap insurance to carry 2 spare screws per seastead.

2. Recommended Sizes (feasibility-level, to be verified by detailed design)

Item	Recommendation	Notes
Working tension per ASU	3,500 lb (1.6 t)	Your figure; verify freeboard: 3 × 3,500 lb ≈ +2.7 ft draft on your waterplane (~3,930 lb/ft)
Design ultimate per ASU	≥ 10,000 lb	FS ≈ 3 on working load, covering uneven sharing & dynamics
Screw shaft	2 in across-flats hex, 2205 duplex (UNS S32205), ~12 ft overall, in two 6-ft pinned couplings for container shipping & handling	~150–170 lb per screw. 2" hex has large torsional margin (shear ≈ 7 ksi at 1,000 ft-lb vs ~30+ ksi allowable) plus wear/bending margin. 1.5" hex would barely work; 2" is the right call for repeated use.
Helices	Two 8 in (200 mm) Ø plates, 3/8 in (10 mm) thick, 3 in pitch, 24 in apart, chamfered leading edges, pilot point at tip	Twin-helix gives better capacity and more stable torque correlation in layered sand than a single plate.
Embedment target	6–8 ft into sand	~4 ft of shaft remains above sand for clamps/float access
Screw pair spacing	36 in (0.9 m) center-to-center (acceptable 30–42 in)	= 4.5 helix diameters → negligible group-capacity reduction (≥3–5× Ø is the geotechnical rule of thumb); wide enough for frame stability and clamp access; compact enough to stow. This is the answer to your spacing question.
Drive per screw	2.2–3 kW submersible gearmotor, 25–30 RPM output, 600–900 ft-lb	Total 4–6 kW per ASU. See §4.
Hex drive bushings ("Kelly bushings")	Nickel-aluminum-bronze (C95800) or 2205 sleeve, hex broached/wire-EDM, 6–8 in engagement, ~0.5–1 mm clearance, grease grooves + Delrin/bronze scraper wipers top and bottom	NAB is the classic seawater bearing bronze — resists galling against duplex. Wipers keep sand and biofouling from jamming the sliding fit.
Load clamps	Wedge/slip type on the hex flats (flat dies in tapered pockets), self-energizing in tension, release on slack	Prototype alternative: drive to a fixed stop and use a bolted split collar (see §8).
Frame	316L weldment, ~4.5 × 3.5 ft, four legs, 12-in Ø pads with 2-in skirts, guide mast with ~5 ft carriage travel	Frame ~200–250 lb; whole ASU ~450–600 lb dry, ~350–500 lb in water
Floats	2 × 30–50 lb buoyancy, closed-cell foam or fender type	Enough righting moment to keep a ~140-lb (in water) shaft vertical while its tip is on the sand
Winch (per corner)	5,000 lb WLL, 1.5–2 kW, spring-applied brake, load pin, ~100 ft line	Must both hoist the ASU (~500 lb wet) and hold 3,500 lb tension indefinitely
Lift line	18 mm polyester double-braid (MBL ≈ 9–11 t) + 10-ft nylon snubber	Avoid HMPE/Dyneema for the permanent tension leg — it creeps under sustained load over weeks. Polyester has minimal creep and its stretch is a free shock absorber.
Electrical	Deck VFD per ASU (soft start, torque control, reverse, current→torque readout) + 4-core 2.5 mm² PUR subsea cable, glanded into housings (no underwater connectors)	Wet-mate connectors are expensive and unnecessary — gland the cable permanently into each housing; connect/disconnect only on deck.
Instrumentation	1080p subsea camera + 2 × 10 W LED, motor current/RPM (from VFD), winch load pin, line-counter for depth	Camera is essential, as you already planned — also lets the operator confirm clamp engagement.
Anodes	2 × ~2 kg zinc on frame (replace annually)	Protects crevices and any coating damage on the duplex/316L

3. Load & Capacity Engineering (the important part)

Torque ↔ capacity correlation. For small-shaft helical anchors the industry-standard empirical relation is ultimate capacity ≈ K_t × installation torque, with K_t ≈ 10 per ft (33 per m). So 500 ft-lb of verified installation torque ⇒ ~5,000 lb ultimate per screw — FS ≈ 2.9 at 1,750 lb/screw working load. This is why the drive must be able to reach at least ~500 ft-lb per screw in firm sand: the installation torque is your acceptance test, logged every time.
What the sand gives you. A single 8-in helix at ~5 ft embedment in medium-dense submerged sand develops roughly 3,000–4,500 lb ultimate by bearing theory; twin helices at 6–8 ft comfortably exceed 5,000 lb. In loose sand, capacity can be half that — which is exactly why the torque gate and the "move and retry" rule matter more than any calculation.
Cyclic/wave loading. Apply a ~0.7–0.8 cyclic reduction factor in sand for repeated tension cycling; your FS ≈ 3 absorbs this. Keep the system taut (your 3-ft pull-down margin is the right idea) — slack-then-snatch is what kills moorings.
Uneven sharing. Assume 60/40 worst-case sharing between the two screws despite the equalizer (one clamp bites early). The design FS covers it.
Snatch loads. With the seastead pulled down ~3 ft in a protected site, residual heave works against a stiff system. Mitigations: polyester stretch + nylon snubber, load-cell monitoring, and — a nice side benefit of your architecture — the winches can re-tension remotely at any time (e.g., ahead of a squall, pull to 4,000 lb; in calm, relax to 3,000 lb).
Freeboard check. 10,500 lb total pull-down ≈ 2.7 ft additional draft. Confirm the walkway and wall bottoms still have adequate clearance above wave crests in your chosen protected anchorages, and that the foils' structure takes the up-load at the clamp points (it does if the winches are bracketed to the wall/frame nodes, not skin).

One stall scenario: if both screws are geared rigidly together and one hits a dense layer, the stalled one keeps rotating in place and can auger out a loose pocket — destroying its own capacity. This is the main reason §4 recommends two independent drives rather than one motor with a splitter.

4. Drive Train Options & Motor Sizing

	Option A — your concept: 1 motor + splitter gearbox	Option B — recommended: 2 independent submersible gearmotors
Configuration	One 4–5.5 kW submersible motor → planetary reducer → custom dual-output gear case (counter-rotating hex bushings)	Two 2.2–3 kW submersible gearmotors, each with its own hex bushing, mounted on the shared carriage; one runs electrically "in reverse"
Torque reaction	Cancels in the gear case; difference goes to frame	Cancels in the shared carriage; difference goes to frame pads — same principle you described
Stall behavior	Bad (both shafts forced to same RPM)	Good — each screw advances at its own rate; small height differences absorbed by the sliding bushings
Custom machining	Significant (dual-output gearbox)	Minimal — mostly off-the-shelf drives + adapter plates
Redundancy	One motor failure = dead unit	Can finish a hole on one motor at half speed
Cost	Roughly equal (~$1.5–2.5k per ASU drivetrain in China)

4.1 How many watts? (direct answer)

Install 4–6 kW total per ASU (2 × 2–3 kW). Physics check: 700 ft-lb (950 N·m) per screw × 2 screws at 25 RPM needs ~5 kW at the shafts; at 15 RPM needs ~3 kW. You will rarely pull max torque and full speed simultaneously — peak torque only occurs in dense sand, where you should slow down anyway (the VFD does this).
Speed target: 20–30 RPM at the screw. With a 3-in pitch that is 5–7.5 ft/min of penetration — i.e., active screw-in of ~2–5 minutes for 8 ft, and ~3–6 minutes out (brief breakout-torque spike, then easy). Add lowering, clamping and tensioning: 15–30 min per unit end-to-end.
The off-the-shelf shortcut (important): submersible wastewater mixers (Flygt/Xylem, Sulzer ABS, Wilo, Landia — and many Chinese clones at $600–1,200 each) are IP68, oil-filled, moisture-sensored, slow-RPM/high-torque gearmotors rated for continuous submersion at 20 m. Two clones on a carriage is 80% of your drivetrain, off the shelf. A 3 kW unit at ~30 RPM delivers ~700 ft-lb — right on target. Confirm a 20 m depth rating for your ≤50 ft requirement.
Alternative off-the-shelf stack: 4-in water-filled submersible borehole-pump motors (~$200–400 each, any depth you like) + planetary gearboxes in oil-filled, pressure-compensated housings. More custom work, but depth-proof and cheap.

5. Materials & Corrosion Notes

Screws: 2205 duplex — correct choice, and better than 316L for your case. 2205 has ~2× the yield strength of 316L (wear and torsion margin for repeated screw-in/out) and much better chloride pitting/crevice resistance (PREN ≈ 35 vs ≈ 24) — which matters in 28–30 °C Caribbean water, where 316L is genuinely marginal in crevices. If you ever upgrade, super-duplex 2507 (PREN ≈ 42) is the next step, at ~1.3–1.5× material cost. Add zinc anodes and inspect annually; avoid tight unsealed crevices in the design.
Welding duplex requires 2209 filler, heat-input control, and pickling/passivation after — use a Chinese fabricator with documented duplex experience (common in Jiangsu/Zhejiang shipbuilding and pressure-vessel shops) and require weld procedure qualifications + dye-penetrant testing on helix welds.
Galvanic isolation seastead↔ASU: the polyester lift line is your insulator — do not add a wire-rope section. Keep the power cable jacket intact (no exposed armor). The rubber-lined storage cradle you planned handles the on-deck case. Note the shore-power/VFD ground: bond the ASU frame to the seastead's ground through the cable's protective earth so your corrosion-protection and electrical-safety grounds are one controlled path — have the electrical designer confirm this against your anode strategy.
Bushings: nickel-aluminum bronze (C95800) — the standard seawater bearing material, galvanically comfortable near duplex, gall-resistant. Grease with a waterproof (e.g., calcium-sulfonate) grease.
Biofouling on the exposed shaft section (weeks submerged: slime/algae; months: barnacles) can jam the bushing on retrieval — the scraper wipers handle slime; for long stays, plan to "exercise" each unit (drive the carriage up/down a few inches) monthly via remote command.

6. Cost Estimates (indicative 2025 USD, FOB China, qty 60 ASUs / 120 screws)

6.1 Bill of materials per ASU

Item	Est. cost (USD)	Notes
2 × duplex screws (2" hex × 12 ft, twin 8" helices)	$1,400 – 2,000	~75 kg 2205 each @ $6–9/kg + forming/welding/NDT/passivation. Largest cost driver — tracks nickel/molybdenum prices.
Drivetrain (2 × submersible gearmotors or mixer clones, adapters)	$1,500 – 2,400	Chinese mixer clones $600–1,200 ea; or pump-motor + planetary + housings
2 × hex bushings (NAB, wire-EDM/broach, wipers)	$250 – 400
2 × slip/wedge clamps + equalizer beam	$350 – 600
316L frame weldment (~150–180 kg) with pads, mast, camera bracket	$1,000 – 1,400
Floats, fasteners (duplex), anodes, misc hardware	$250 – 400
Camera + lights + sensors	$250 – 450
PUR subsea cable (~100 ft) + glands	$300 – 500
Assembly, pressure/torque test, QA, packing	$400 – 700
Per ASU, FOB China	≈ $5,700 – 8,900 (target ~$6–7k)
Corner winch + load pin + rope/snubber (per ASU)	$900 – 1,500	Marine-grade electric, brake, level-wind

6.2 Program totals (your 20-seastead order)

Scope	Est. cost
60 ASUs + 60 winch sets, FOB China	$400k – 620k (target ≈ $460–500k)
Spares package (6 spare screws, 2 spare drivetrains, bushings/seals/wipers kits)	$25 – 40k
Freight + duty (≈ 2 × 40-ft HC containers, disassembled: nested frames, 6-ft shaft sections)	$20 – 35k
Per seastead (3 ASUs + 3 winches, landed, with spares share)	≈ $22,000 – 32,000

Cost-down paths if needed: galvanized carbon-steel frame + anodes (save ~$400/ASU, acceptable since the frame is retrievable and inspectable); single-motor Option A (marginal); larger qty. Do not cost-reduce the screws — they are the safety-critical part.

7. Off-the-Shelf Availability (direct answers)

Part	Off the shelf?	Sources / notes
Hex-shaft helical mooring screws (galvanized steel)	Yes	Boat-mooring "helix anchors" (US/EU suppliers) and utility screw anchors. Fine for the prototype where coating wear doesn't matter. Note: many utility anchors (Hubbell Chance SS series) use square shafts (1.5–2 in) — the Kelly-bushing concept works identically with square; hex just wears more kindly.
Duplex / 316L helical mooring screws	No (effectively)	316 yacht mooring screws exist (Italy/Turkey, small sizes only). Duplex helical screws are a custom fabrication — but an easy one: 2205 hex bar + formed plate + qualified weld. Any competent Chinese duplex fabricator can make them from your drawings.
"Kelly bushings" / hex drive sleeves in ~2" size	Partially	Oilfield Kelly bushings exist but are sized for 3–5¼-in Kellys and weigh hundreds of pounds — far too big. Small hex-bore hubs (PTO, robotics, ½–1½ in) are too small. The useful off-the-shelf items are anchor-drive "drive tools" / hex output hubs in 2 in and 2-9/16 in (Digga, Pengo, Eskridge, Belltec ecosystems) — these can serve as the core of your bushing. A proper long-engagement sliding hex sleeve is a quick custom part: wire-EDM the hex through-hole in NAB (~$80–200 each in China at prototype qty, less in production).
Machines that drive hex screw anchors	Yes — and yes, you can bolt two together	Hydraulic helical-pile drive heads (2,000–12,000 ft-lb, 2-in hex outputs) are standard skid-steer/excavator attachments. Two drive heads on a welded frame = your land-based prototype ASU. For in-water use, hydraulics means hoses/HPU — fine on a beach test, not on the seabed at 50 ft; that's where the submersible-mixer drives come in.
Submersible gearmotors	Yes	Submersible wastewater mixers (Flygt, ABS, Wilo, Landia + Chinese clones) — see §4.1. Also 4-in borehole-pump motors (any depth) + planetary gearboxes.
Winches with holding brake; load pins	Yes / Yes	Standard marine/industrial; cheap Chinese load pins are fine here (non-safety-critical readout — the rope is the safety path).
Subsea cable & connectors	Cable yes; wet-mate connectors avoid	PUR/TPU-jacketed multi-core cable is commodity. Skip underwater-mateable connectors ($$$) — gland the cable into the housings, connect on deck only.
Slip/wedge clamps for hex shafts	No — custom	Concept borrowed from drilling "slips"/pipe elevators; must be designed for hex flats. Simple machining, one of your few genuinely custom mechanisms (see §8 for a prototype shortcut that avoids them initially).
VFDs, cameras, floats, polyester rope, NAB bar, 2205 hex bar	Yes	2205 hex bar is stocked by Chinese mills/stockists (~$6–9/kg).

8. Prototype Plan (maximize off-the-shelf)

Phase 1 — Land/beach drive-test rig (~$6–12k, 4–8 weeks)

Goal: prove screw geometry, installation torque vs. depth in real sand, pull-out capacity, spacing, and the torque↔capacity correlation — before any underwater hardware exists.

Welded steel or aluminum frame (local fab, $1–2k) holding two hydraulic drive heads (used $1.5–3k each; new $3–5k) at 36-in spacing, powered by a small HPU ($1–2k).
Two galvanized hex mooring screws (or mild-steel hex bar welded to purchased helix sections) — $300–600 each.
Inline torque sensor or HPU pressure correlation, and a load cell ($300) for pull tests — pull with an excavator or come-along against a deadman.
Test matrix: torque vs. depth; pull-out at 5/6.5/8 ft; 30-in vs. 36-in spacing group effects; 50-cycle cyclic test at working load; drive out and re-drive 10× to check hex wear.

Phase 2 — In-water electric prototype (~$25–40k, 2–4 months)

316L frame, two Chinese submersible mixer drives, hex bushings, camera, deck VFDs, PUR cable; deploy from a dock or small barge in 10–15 ft of water; then a full deploy/tension/detension/retrieve cycle test, ideally 20+ cycles including a 2-week soak to check fouling/retrieval.

Custom parts you will need (both phases)

Custom part	Make how	Proto cost (each)
Hex drive bushings (long-engagement sleeves)	Wire-EDM or broach, NAB or 4140 (proto) → NAB/2205 (production). China CNC houses (PCBWay/JLC CNC, FacFox, Xometry) do wire-EDM hex cheaply.	$80 – 250
Carriage/motor adapter plates, mast rails	CNC or laser-cut plate, local or China	$100 – 300
Frame weldment	Local welding shop (steel for land rig; aluminum OK; 316L for in-water)	$1,000 – 2,500
Slip/wedge clamps	Machined wedges + pockets. Prototype shortcut: skip slips in Phase 1–2 — drive screws to a fixed depth stop and clamp with a bolted split hex collar. Add self-engaging slips in the production design once heights/behavior are characterized.	$150 – 400 (wedge set)
Pressure-compensated housings (if not using mixer drives)	Machined aluminum (hard-anodized) or HDPE proto; bladder compensator is off-the-shelf	$500 – 1,200

3D printing: yes and no

Print (PETG/ASA/PA-CF): bushing fit-checks on the hex shaft (do this first — cheap way to nail tolerances), float collars, camera/light brackets, control-box enclosures, clamp mockups, cable clips. $30–300/part.
Do not print: anything in the load path or torque path (bushings, wedges, gears, shaft couplings, lift points) — machine those from metal. For shallow, short-duration tests, printed-then-epoxy-sealed housings can survive a pool test, but machine the real one.
Total custom-parts spend for a Phase 2 prototype: roughly $4–8k in machined parts plus $1–2k in prints — the rest of the $25–40k is the commercial drives, cable, camera, VFDs, frame, and test logistics.

9. Hiring the Detailed-Design Engineer(s)

Who you actually need (3 small roles, not one big firm)

Lead: mechanical/marine engineer with subsea tooling, ROV tooling, or drilling/helical-equipment background. Owns the ASU mechanism, drivetrain, clamps, housings, seals, and the fabrication drawing package for China.
Geotechnical engineer with helical-anchor experience (part-time). Produces the capacity memo: screw sizing vs. Caribbean sand profiles, K_t verification plan, acceptance criteria, cyclic factors. Often available from helical-pile manufacturers' engineering departments or via the Deep Foundations Institute (DFI).
Electrical/controls (small scope): VFD selection, torque-from-current calibration, camera/comms, interlocks, E-stop, grounding/bonding plan.

Where to find them

SNAME and MTS (Marine Technology Society) member directories and job boards; IEEE Oceanic Engineering Society.
LinkedIn searches: subsea mechanical engineer, ROV tooling engineer, helical pile engineer, mooring engineer, naval architect small craft. Ask for project photos/references showing mechanisms they designed that went in the ocean.
Application engineers at drive-head makers (Digga, Eskridge, Pengo) and mixer makers (Xylem/Flygt, Sulzer) — free, genuinely useful drivetrain advice.
DFI (Deep Foundations Institute) helical-pile community for the geotech; a university ocean-engineering lab if you want instrumented testing.
Upwork/etc. is fine for CAD drafting support only, not for owning the engineering.
China side: once you have a design package, get DFM feedback from 2–3 candidate factories (free — they do it to win the order), and budget third-party inspections (SGS/TÜV/AsiaInspection, ~$300–500/day) at first-article and pre-shipment.

Fees & timeline (US/EU rates; adjust for region)

Scope	Typical fee	Duration
Senior consultant hourly rate	$125 – 250/hr	—
Phase 0: load cases, concept freeze, vendor RFQs, test plan	$8 – 20k fixed	3 – 6 weeks
Geotechnical capacity memo + acceptance criteria	$3 – 10k	2 – 4 weeks (parallel)
Prototype design package (drawings sufficient for local fab + China CNC)	$10 – 25k	4 – 8 weeks
Test support + design iteration	$5 – 15k	through testing
Production drawing package + BOM + QC plan (China-ready)	$15 – 30k	4 – 8 weeks after tests
Total program	≈ $30 – 80k	≈ 6 – 12 months to delivered production units

Contract in stages with milestone payments, an NDA, and explicit IP assignment. Hire fast (1–2 weeks) with a small paid trial task (e.g., "review this load case and red-line my concept sketch").
If you want stamped calculations for insurance/lender credibility, specify a PE (or equivalent) review up front — it adds cost but the torque-logged install records plus stamped calcs make a strong insurance package.

10. Risks & Watch Items (ranked)

Very loose sand sites — the system honestly can't hold 3,500 lb everywhere. The torque gate + "move and retry" rule is the mitigation; publish a site-selection checklist (sand, 15–40 ft, protected) for captains.
Thruster scour — your RIM drives sit ~2 ft above the leg bottoms, i.e., ~8 ft above the seabed at a 15-ft site, and the ASU lands nearly under the leg. During GPS-hold the jets can fluidize sand around the ASU and tilt it. Mitigations: ASU pads with skirts, minimum-necessary thrust during install, offset the ASU landing point slightly from directly under the jet axes, and confirm with a camera watch during the first deployments.
Biofouling jam on retrieval — wipers, grease, monthly remote "exercise," and the reverse+pull contingency. Test a 2-week soak in Phase 2.
Snatch/dynamic loads — polyester + nylon snubber + load monitoring + re-tension capability. Never let the system go slack (your 3-ft margin handles Caribbean microtides ~1 ft).
Galvanic corrosion — synthetic lift line as isolator, bonded PE ground strategy, anodes, rubber-lined cradle (as you planned). Have the electrical designer and corrosion engineer sign off together.
Cable survivability — weeks submerged, chafe at the waterline: PUR jacket, chafe sleeve, constant tension to minimize working, spare cable per seastead.
Single-screw stall / augering — solved by twin independent drives (Option B).
Freeboard when tensioned (~2.7–3 ft less) — verify walkway/wall-bottom clearance vs. local wave climate at your anchorages.

11. Suggested Next Steps

Freeze the load cases in a one-page document (3,500 lb working, 10,000 lb ultimate per ASU, 15–50 ft depth, sand, ≤1 ft tide, protected sites).
Order two galvanized hex mooring screws + two used hydraulic drive heads and build the land rig (Phase 1) — this can start this month for under ~$12k and answers the biggest unknown (does the sand give you the torque you need?).
In parallel, hire the lead mechanical engineer (small paid trial task first) and get the geotech memo started.
Request quotes/DFM from 2–3 Chinese duplex fabricators using a rough screw drawing — you'll get real prices within 2 weeks and validate the BOM above.
After land tests: commission the in-water prototype (Phase 2), then the production package, targeting first production articles ~9–12 months out.