A concept primer built around a reference design: a 44‑foot equilateral‑triangle living platform on three streamlined buoyancy legs, shipped in a single 45‑ft High‑Cube container.
All numerical examples are illustrative back‑of‑envelope estimates, intended to teach method — not to substitute for a professional design review.
Naval architecture is the engineering discipline that answers six questions about anything that floats:
Our reference design is a three‑column semi‑submersible: a 44‑ft triangular living deck, 7‑ft walls, carried on three vertical foil‑section legs (~21.5‑ft long, NACA‑0035 section, ~3‑ft max thickness), each half‑submerged with ~7.25‑ft draft, rated at 27,500‑lb buoyancy per leg at the design waterline. That geometry appears in every section below, because it sits at an interesting point in the design space: part SWATH, part trimaran, part mini tension‑leg platform.
Archimedes' principle: a floating body displaces a weight of water equal to its own weight. In salt water (≈64‑lb/ft³):
Displacement Δ (lb) = 64 × submerged volume ∇ (ft³)Three legs × 27,500‑lb rated buoyancy = 82,500‑lb total displacement (~37.4 metric tons, ~1,290‑ft³ of seawater). The shipping container caps the shipped structure at 62,000‑lb. The design allocates ~25% of displacement (~20,600‑lb) to LiFePO₄ batteries low in the legs.
Note the arithmetic is tight: 62,000 + 20,600 ≈ 82,600 ≈ the full displacement. Whether or not the batteries travel inside the container, there is little slack for people, water, food, and stores. A running weight spreadsheet — every plate, bolt, panel, and tank, with its weight and vertical position — is the single most important design document.
A useful derived number is the load sensitivity: how much the draft changes per unit of added weight. This equals the weight of a one‑inch (or one‑foot) slice of the waterplane. The designers state that a 1‑ft change in draft corresponds to about 1/7 of total buoyancy:
1/7 × 82,500‑lb ≈ 11,800‑lb per foot of draft, i.e. a total waterplane area of about 11,800 ÷ 64 ≈ 185‑ft² (~61‑ft² per leg). Per inch: ≈ 980‑lb/inch. So 2,000‑lb of stores and crew sinks the seastead about two inches — tangible and manageable, but a number to design the ballast and trim plan around.
Reserve buoyancy is watertight volume above the waterline that can still be immersed. Here the top half of each leg (plus the enclosed deck) is reserve — roughly 100% of the displaced volume. Subdivide it: multiple airtight compartments per leg mean one puncture is an inconvenience, not a foundering. Check the damaged stability case: can the craft stay upright with one corner compartment flooded?
Three points govern initial stability. G is the center of gravity (where the weight acts — pulled down by heavy things, especially those 20,000‑lb of batteries low in the legs). B is the center of buoyancy (centroid of the displaced volume). When the craft heels, B shifts sideways toward the immersed side; the vertical through the new B intersects the centerline at M, the metacenter. The distance GM sets the initial righting arm:
GM = KB + BM − KG, where BM = I ⁄ ∇Here I is the second moment (inertia) of the waterplane about the heel axis. This is where multi‑leg geometry shines: I is dominated not by the size of each leg's footprint but by their separation — each leg contributes roughly A·d², its waterplane area times the square of its distance from the center.
Legs at the corners of a 44‑ft equilateral triangle sit at radius R = 44/√3 ≈ 25.4‑ft from the centroid. With ~61‑ft² waterplane per leg:
I ≈ 61 × (3/2) × 25.4² ≈ 59,000‑ft⁴ → BM ≈ 59,000 ÷ 1,290 ≈ 46‑ft
With KB ≈ 3.6‑ft (half the draft) and an assumed KG ≈ 9–11‑ft (structure above water, batteries low), GM ≈ 38–42‑ft. For comparison, a 44‑ft cruising monohull might have GM of 2–4‑ft. This craft is extremely stiff: safe against capsize, quick in its motions. The wide stance converts a small waterplane into enormous initial stability — the central trick of every multi-hull and semi-submersible.
Partly full tanks let liquid slosh to the low side, effectively raising G and reducing GM. With GM this large the effect is unlikely to be fatal, but subdivide and press-full the water and fuel tanks anyway. And measure the real GM after assembly with an inclining experiment: shift a known weight a known distance athwart, measure the heel angle θ, then GM = w·d ⁄ (Δ·tan θ).
Every floating body is a pendulum. Disturb it and it oscillates at its natural roll period:
Troll = 2π·k ⁄ √(g·GM) ≈ 0.44·B ⁄ √GM (ft, seconds)where k is the roll radius of gyration (how far the mass sits from the roll axis) and B the beam. Resonance occurs when waves arrive at the natural period: each wave adds energy to the last, and motion grows until damping (viscous drag of appendages, plates, and the hull) caps it. Low damping = tall, narrow resonance peak; high damping = low, broad peak.
| Wave type (Caribbean) | Typical period |
|---|---|
| Local chop, protected anchorages | 1.5 – 3 s |
| Trade-wind seas | 3 – 6 s |
| Swell wrapping into anchorages | 6 – 10 s |
| Long ocean swell | 10 – 15 s |
What matters is the encounter period — the period you actually feel, shifted by your speed and heading. Heading into waves at speed shortens it: at 5‑kn into 5‑s seas, the encounter period drops to about 3.8‑s. Speed and heading are therefore free "detuning knobs" while underway.
With GM ≈ 40‑ft and k ≈ 13–15‑ft: Troll ≈ 2.3–3‑s — squarely in the local-chop band. Two mitigations keep this comfortable:
① Small excitation. Roll moment from waves scales with waterplane area and is partly self-cancelling when the wavelength is comparable to the 44‑ft leg spacing. Small waterplane ⇒ small push. ② Damping. The bolt-on heave plates on the lower legs drag quadratically through the water — exactly where you want energy extracted. Expect behavior like a big catamaran: very safe, somewhat "snappy" if excited, with damping doing the comfort work.
Heave tells the same story: undamped, Theave ≈ 2π√(m ⁄ ρgAwp) ≈ 3‑s; the plates' added mass (water they entrain) stretches that toward ~4–5‑s while their drag suppresses the resonant peak.
Detune (move natural period away from local wave periods — hard here since geometry fixes GM), damp (heave plates — the chosen lever), and avoid (moor in protected water; the stated Caribbean plan). Verify at the dock with a roll decay test: heel and release, record the period and the decay rate of successive swings.
The waterplane is the shape the hull slices out of the sea surface. Its area controls three different things, and shrinking it trades between them:
At 1‑ft ≈ 1/7 of displacement, this design deliberately sits between SWATH and conventional hulls: soft enough to be gentle, stiff enough to absorb provisioning, a crowd of visitors on one walkway, or battery swaps without drama. The price is that heave resonance (~3–5‑s) lives closer to the wave spectrum than a SWATH's does — acceptable in protected Caribbean anchorages, questionable in open-ocean swell. The waterplane number is arguably the single most defining parameter of the whole concept.
Total hydrodynamic resistance splits into:
Wave-making is governed by the Froude number Fn = V ⁄ √(g·L). The classic "hull speed" rule, V ≈ 1.34·√LWL (knots, feet), marks where wave-making starts to dominate: for 21.5‑ft legs, ≈ 6.2‑kn. Friction uses the Reynolds number and a friction line such as Cf = 0.075 ⁄ (log₁₀Re − 2)².
At 5‑kn (8.4‑ft/s), dynamic pressure q = ½ρV² ≈ 71‑lb/ft². With ~1,300‑ft² of wetted surface (three half-submerged legs plus appendages) and Cf ≈ 0.0028 at Re ≈ 1.5×10⁷, friction is ~250–300‑lbf; form, wave-making (Fn ≈ 0.32), and appendages bring a realistic total to ~350–450‑lbf. Power into the water ≈ 4–5‑kW; at ~50% overall propulsive efficiency, roughly 8–10‑kW of electrical power. Power scales ~V³ at these speeds, so 6‑kn needs roughly 1.7× as much — speed is an energy decision, not just a throttle setting.
Barnacles in warm Caribbean water can double skin friction within a season: budget antifouling or haul/wipe access from the walkways. And remember the heave plates, so valuable at anchor, are pure drag underway — the bolt-on detail should allow removal or fairing for long passages.
Same physics, different fluid: F = ½ρairV²·A·Cd. Air is ~830× less dense than water, but wind is fast and the house is big — a 7‑ft wall around a 44‑ft triangle presents ≈ 308‑ft² of projected area per side.
qair at 30‑kn ≈ 3.1‑lb/ft². With Cd ≈ 1.1 (bluff flat wall with railing and walkway clutter): F ≈ 1,000–1,100‑lbf — roughly three times the entire hydrodynamic drag at 5‑kn. At 40‑kn it grows to ~1,800‑lbf (V² law).
Consequences: ① The six fixed RIM thrusters must together exceed wind + water drag to hold station or make headway to windward — size bollard pull against a chosen "design wind," not against calm-water drag. ② Moored, the tension-leg screws and tendons carry this load continuously. ③ Orientation matters: presenting an apex rather than a flat side to the wind trims the load somewhat; the tension mooring fixes the heading, so choose it deliberately.
Stabilization options divide by whether they need forward speed and whether they need power:
| Device | Type | Works at zero speed? | Cost |
|---|---|---|---|
| Low ballast (batteries in legs) | Passive | Yes | Free (already planned) |
| Heave plates | Passive damping | Yes | Drag underway |
| Anti-roll tanks | Passive/semi-active | Yes | Free-surface penalty, volume |
| Fin stabilizers | Active | No — need flow | Power, through-hulls |
| Gyro stabilizers | Active | Yes | Heavy, kW-hungry |
| Thruster-based motion damping / DP | Active | Yes | Power, control software |
| Tension-leg mooring | Mechanical | Yes | Site-bound |
The essential seasteading fact: a seastead lives at zero speed, where fins do nothing. The sensible stack here is exactly the one chosen — passive plates for wave-frequency damping, and the mooring as the ultimate stabilizer when parked. The six fixed-orientation thrusters give surge and yaw authority (plus differential-thrust maneuvering and turn-in-place), but essentially no lateral (sway) control, so they can assist station-keeping and damp slow drift on the mooring rather than replace it. Two seasteads linked by a walkway become a two-body coordinated-control problem: both computers sharing IMU/GPS data and gently matching low-frequency motion is realistic; cancelling wave-frequency relative motion is not — board only in benign conditions.
Design passive stability to be adequate alone; treat active systems as comfort and capability layers. The triple-redundant power architecture (per-leg batteries, charge controllers, inverters, and thruster pairs) is exactly the right fault-containment pattern for this.
The design belongs to a well-proven family. In the early 1960s drillers noticed that pontoons ballasted below the surface with only slender columns piercing it barely responded to waves — the column-stabilized semi-submersible was born, and it now dominates deep-water drilling and production. Relatives include the SWATH (Small Waterplane Area Twin Hull — fully submerged torpedoes on thin struts, e.g. SSP Kaimalino, 1973), ballast-down heavy-lift ships (Blue Marlin), and tension-leg platforms (first: Hutton TLP, 1984), where excess buoyancy pulls taut vertical tendons to the seabed and heave is nearly eliminated. In seasteading proper, DeltaSync's ClubStead study reached the same conclusion this design embodies: for a stationary floating community, semi-sub geometry is the sweet spot.
Air gap: with 7.25‑ft of leg above water, the deck underside clears the surface by ~7‑ft. Wave slamming on the deck is impulsive and violent; this gap is comfortable for 2–4‑ft anchorage seas but argues for storm avoidance or a very sheltered berth in anything like 6–8‑ft swell.
Tension legs: pulling down 3‑ft adds 3 × 11,800 ≈ 35,000‑lb of pretension (~12,000‑lb per corner group). Sanity check against slack lines: a 2‑ft wave amplitude changes buoyancy by ±61‑ft² × 2‑ft × 64 ≈ ±7,800‑lb per leg — pretension wins, so lines stay taut in moderate conditions, confirming the "3‑ft is enough" intuition for protected sites. In ~6‑ft waves the margin vanishes: pull deeper or leave. Helical screws in sand must be rated with a healthy factor of safety above the cyclic loads, and the tidal window (small in the Caribbean) must fit inside the pull-down allowance.
The drag coefficient Cd packages everything about shape into one number in F = ½ρV²·A·Cd (mind the reference area A — frontal for bluff bodies, planform for foils):
| Shape (flow normal to it) | Approx. Cd |
|---|---|
| Flat plate | 1.17 – 1.3 |
| Square/rectangular box (the house wall) | ~1.0 – 1.3 |
| Circular cylinder (subcritical) | 0.8 – 1.2 |
| Sphere | ~0.47 |
| Streamlined strut / fairing (3D) | 0.05 – 0.10 |
| NACA foil section (2D, on planform area) | 0.006 – 0.01 |
| Streamlined body of revolution | ~0.04 |
Per foot of height at 5‑kn (q ≈ 71‑lb/ft²): a 3‑ft-diameter cylindrical column sees ~71 × 3 × 1.0 ≈ 213‑lbf. A 3‑ft-thick, 8.5‑ft-chord foil sees ~71 × 8.5 × 0.008 ≈ 5‑lbf — roughly 40× less drag. Streamlining is the difference between a platform that can genuinely cruise and one that can only drift.
Nuances: ① The 35%-thick section is fat — chosen for buoyant volume and container packing, not minimum drag (a NACA‑0012 would be sleeker but hold far less battery). ② Truncating the last 0.5‑ft of chord leaves a ~5-inch blunt trailing edge ("flat-back" foil) plus the wire conduit: base drag rises, perhaps doubling section Cd — still an order of magnitude better than a cylinder, and the packing benefit is worth it. ③ Foils only work when pointed into the flow: fixed legs plus cross-current means operating at a yaw (crab) angle, which grows drag and side-force fast. ④ Surface-piercing struts add spray and interference drag at the waterline and at the leg–deck junction — fair those intersections.
The triangle's roof area is (√3/4)·44² ≈ 838‑ft²; at realistic packing and modern panel density expect roughly 12–15‑kW peak solar, and a tropical-day average well below that. Against the ~8–10‑kW draw at 5‑kn from §6: 4–5‑kn solar-electric cruising in good sun is plausible, with 6‑kn available as a battery-assisted sprint. The ~20,000‑lb LiFePO₄ budget is on the order of 1,000‑kWh at pack level — hundreds of nautical miles of low-speed autonomy, and the legs double as ballast for that mass. RIM drives of 1.5‑ft diameter sitting ~5‑ft below the waterline are deep enough to avoid ventilation in small chop; their combined bollard pull should be checked against the §7 wind loads, since wind — not water — sets the worst-case thrust demand.
| Concept | Question to ask of this design |
|---|---|
| Weight budget | Is there a line-item weight + center-of-gravity ledger showing structure + batteries + crew + stores ≤ 82,500‑lb with margin? What is the tracked KG? |
| Stability | Has GM been confirmed by inclining after assembly? Is the one-compartment-flooded case upright and afloat? |
| Roll period | Roll-decay test done? Is the measured ~2.5–3‑s period acceptable given site wave climates and the heave-plate damping? |
| Waterplane | Is there a ballast/trim procedure for the ~980‑lb/inch load sensitivity (crowds on one walkway, water-tank transfers)? |
| Water drag | Fouling plan? Heave plates removable for passages? Leg–deck junctions faired? |
| Wind drag | Does combined thruster bollard pull exceed the ~1,100‑lbf 30‑kn wind load with margin? Are tendons/screws sized for 40‑kn+? |
| Active systems | Is passive stability alone adequate? Do the two-vessel walkway controls degrade gracefully on comms loss? |
| Semi-sub specifics | Is the ~7‑ft air gap respected by the storm plan? Does pretension exceed wave-induced load swings at the intended sites? Tide range within pull-down? |
| Shape / Cd | Trailing-edge and conduit drag accepted? Crab-angle drag in cross-currents considered in the energy budget? |
| Structure (beyond scope here) | Global bending at the leg–deck connections in beam seas; walkway diagonal braces; dinghy suspension loads in a seaway. |
Added mass — water effectively entrained by an accelerating body; raises inertia and lengthens natural periods. Air gap — clearance between waterline and the underside of the deck. Bollard pull — static thrust at zero speed. Cd, Cf — drag and skin-friction coefficients. Damping — energy extraction that caps resonant motion. Displacement (Δ) — weight of water displaced = craft weight. Encounter period — wave period as felt by a moving craft. Froude number — speed vs. wave-making scaling, V/√(gL). Free surface effect — stability loss from slack tanks. GM — metacentric height; initial stability. Heave plate — horizontal plate adding vertical damping and added mass. Hull speed — ≈1.34√LWL knots, where wave-making climbs steeply. Inclining experiment — measured heel from a shifted known weight; yields true GM. KG, KB, BM — heights of gravity, buoyancy, and metacenter contributions. Reserve buoyancy — sealed volume above the waterline. Resonance — amplification when excitation matches a natural period. Reynolds number — viscous scaling, VL/ν. Semi-submersible — buoyancy deep, small columns at the surface. SWATH — small-waterplane twin-hull ship. Tension-leg platform — buoyant hull pulled down against taut vertical tendons. TPI — tons per inch immersion; load sensitivity. Waterplane — the hull's footprint at the surface. Wetted surface — submerged area that generates friction.