TANKER SHIP

PHASE 1: Make It Float

You're standing on the dock, looking at a slab of steel 20 mm thick. Density: 7,800 kg/m³. Water: 1,025 kg/m³. Steel is 7.6 times denser than water. Drop that plate in the harbor and it sinks in 2 seconds. Now explain how 45,000 tonnes of that same steel floats while carrying 300,000 tonnes of oil. The answer is Archimedes' principle, and it's 2,200 years old: F_buoyancy = ρ_water × V_displaced × g A body immersed in fluid experiences an upward force equal to the weight of the fluid displaced. The steel doesn't need to be lighter than water. The hull needs to be — because the hull encloses air. Shape the steel into a hollow box, and the box displaces far more water than the steel alone. The air inside pays no weight penalty but claims enormous volume. The Numbers for a VLCC (Very Large Crude Carrier) A typical 300,000 DWT tanker: ├── Length overall (LOA): 330 m ├── Beam (width): 60 m ├── Depth (keel to deck): 30 m ├── Design draft (loaded): 22 m ├── Lightweight (empty ship): ~45,000 tonnes ├── Deadweight (cargo + fuel + stores): 300,000 tonnes └── Total displacement: ~345,000 tonnes For the ship to float at 22 m draft, it must displace 345,000 tonnes of seawater: V_displaced = mass / ρ_seawater V_displaced = 345,000,000 kg / 1,025 kg/m³ V_displaced = 336,585 m³ That's the volume of the hull below the waterline. A box 330 m × 60 m × 22 m = 435,600 m³, but the hull isn't a box — it narrows at bow and stern. The ratio of actual underwater volume to the bounding box is the block coefficient (Cb): Cb = 336,585 / 435,600 = 0.82 Tankers are fat. A Cb of 0.82 means the hull fills 82% of its bounding rectangle. Compare: a container ship Cb ≈ 0.65, a destroyer Cb ≈ 0.50. Tankers are slow, so they don't need slender hulls. They need volume.

Reserve Buoyancy — The Margin of Life The ship floats with 22 m of its 30 m depth underwater. That leaves 8 m of hull above the waterline — called freeboard. This freeboard is reserve buoyancy: if the ship takes on water (flooding, heavy seas), it can sink 8 more meters before the deck goes under. Reserve buoyancy for a VLCC: Volume above waterline ≈ 330 × 60 × 8 × 0.75 = ~118,800 m³ Extra buoyancy available = 118,800 × 1,025 = ~121,770 tonnes That's a 35% margin beyond the loaded displacement. Lose a tank to flooding, take on thousands of tonnes of seawater — the ship keeps floating. This margin is mandated by the International Load Line Convention (the Plimsoll line painted on every ship's hull).

DESIGN SPEC UPDATED: ├── Buoyancy: F = ρ_water × V_displaced × g ├── Total displacement: ~345,000 tonnes at 22m draft ├── Block coefficient: 0.82 (hull fills 82% of bounding box) ├── Freeboard: 8m → reserve buoyancy of ~122,000 tonnes (35% margin) └── Steel floats because the hull encloses air — geometry defeats density

PHASE 2: Make It Strong

Imagine a 330-meter steel beam balanced on a wave. The wave has a length of 165 meters — half the ship. One end of the ship is supported by the wave crest. The other end hangs in mid-air over the wave trough. Gravity pulls it down. The ship bends like a bridge with no middle support. Except this bridge weighs 345,000 tonnes and the load shifts every 10 seconds. A ship is a beam. The longest, most heavily loaded beam humans have ever built. And unlike a bridge, it's loaded from below (buoyancy), from above (cargo weight), and the supports (wave crests) move continuously. Hogging and Sagging — The Two Ways a Ship Bends

Hull Girder Bending Moment The wave bending moment for a ship on a wave equal to its length: M = ρ_water × g × B × L² × h / C Where: ├── ρ_water = 1,025 kg/m³ ├── g = 9.81 m/s² ├── B = beam = 60 m ├── L = length = 330 m ├── h = wave height = 12 m (design wave) └── C = constant (~8 for sagging, ~11.5 for hogging) For sagging (worst case): M = 1,025 × 9.81 × 60 × 330² × 12 / 8 M = 1,025 × 9.81 × 60 × 108,900 × 12 / 8 M = 9.96 × 10⁹ N·m ≈ 10 GN·m Ten billion Newton-meters. For perspective: the bending moment on a large highway bridge is about 50 MN·m. This ship experiences 200 times more bending than a bridge — and the load reverses every few seconds. Hull Girder Stress The stress at the deck or keel: σ = M / Z Where Z is the section modulus — a measure of the hull cross-section's resistance to bending. For a VLCC, Z ≈ 50-70 m³ (yes, cubic meters — massive). σ = 10 × 10⁹ / 60 σ = 167 MPa Yield stress of hull steel: 235-355 MPa (depending on grade). The safety factor is only 1.4 to 2.1. Not much margin. This is why classification societies (Lloyd's, DNV, ABS) are fanatical about steel quality and welding standards.

DESIGN SPEC UPDATED: ├── Bending: hogging (wave amidships) and sagging (waves at ends), alternating every ~10s ├── Wave bending moment: ~10 GN·m (200× a highway bridge) ├── Hull stress: σ = M/Z → ~167 MPa, safety factor 1.4-2.1 against yield ├── Hull steel grades: 235-355 MPa yield, controlled by classification societies └── Cross-section: I-beam principle — thick deck/bottom plates + longitudinal stiffeners

PHASE 3: Fill It Without Sinking

You've built a steel box that floats. Now fill it with 300,000 tonnes of liquid. Liquid that moves. Liquid that sloshes from side to side every time a wave hits. Each slosh shifts the center of gravity. Shift it far enough, and the ship rolls over. You haven't changed the weight — you've changed where the weight IS. And that's enough to kill everyone on board. This is the free surface effect, and it has sunk more ships than storms. How Free Surface Kills Imagine a tank half-full of oil, running the full width of the ship (60 m). The ship rolls 5° to starboard. The oil flows to starboard. Now the center of gravity is to starboard — which makes the ship roll MORE to starboard — which makes more oil flow — which makes it roll MORE... It's a positive feedback loop. Self-amplifying. The ship doesn't need to be overloaded. It just needs one tank with a free liquid surface. Free surface correction: GG' = (ρ_cargo × i) / Δ Where: ├── GG' = virtual rise in center of gravity (m) ├── ρ_cargo = density of the liquid (crude oil ≈ 850 kg/m³) ├── i = second moment of area of the free surface (m⁴) ├── Δ = displacement of the ship (kg) The critical term is i. For a rectangular tank: i = (l × b³) / 12 Where l = tank length, b = tank breadth. Note: b is CUBED. Breadth dominates.

Why Narrow Tanks Save Ships Compare one wide tank vs two narrow tanks:

For our VLCC with 4 longitudinal subdivisions per tank group: GG' = (850 × 26,042) / 345,000,000 GG' = 0.064 m Versus the single-tank case: GG' = (850 × 416,667) / 345,000,000 GG' = 1.026 m A 1-meter virtual rise in the center of gravity is catastrophic — it can reduce the metacentric height to zero or negative, causing capsize. The subdivided arrangement keeps it to a manageable 6.4 cm.

Tank Arrangement on a VLCC A modern VLCC has approximately: ├── 15-17 cargo tanks (3 rows across × 5-6 groups along the length) ├── Center tanks: widest, ~24 m breadth ├── Wing tanks: narrower, ~16 m breadth each ├── Slop tanks: 2 (for tank washing residues) ├── Segregated ballast tanks: in the double hull spaces └── Total cargo capacity: ~320,000 m³ Loading sequence matters. You can't just pump oil into random tanks. Load plan: ├── Start with center tanks (keeps ship upright) ├── Alternate port and starboard wing tanks ├── Monitor trim (bow-stern angle) throughout ├── Monitor free surface effect at every stage └── Final condition: all cargo tanks 95-98% full (minimize free surface)

DESIGN SPEC UPDATED: ├── Free surface effect: GG' = ρ_cargo × i / Δ (virtual rise in G) ├── i = l × b³/12 — breadth cubed means narrow tanks are exponentially safer ├── Halving tank width → 4× reduction in free surface effect ├── VLCC: 15-17 cargo tanks, 3 across × 5-6 groups longitudinally └── Tanks loaded to 95-98% full to minimize sloshing

PHASE 4: Push 300,000 Tonnes

The ship is loaded. 345,000 tonnes of steel, oil, and water. You need to move it at 15 knots (28 km/h) across the Indian Ocean. The resistance is enormous — you're pushing a building through treacle. And you'll do it with a single propeller, 10 meters in diameter, turning at 80 RPM. One propeller for the weight of 3,000 blue whales. Resistance — What Fights the Ship Total hull resistance at 15 knots for a VLCC: R_total = R_friction + R_wave + R_appendage + R_air ├── Friction drag: water sticking to the hull. ~75% of total │ R_f = ½ × ρ × V² × S × Cf │ Wetted surface S ≈ 35,000 m² (the area of 5 football fields) │ At 15 knots: R_f ≈ 1,050 kN │ ├── Wave-making drag: energy lost creating bow and stern waves. ~15% │ R_w ≈ 210 kN │ ├── Appendage drag: rudder, propeller hub, bilge keels. ~5% │ R_a ≈ 70 kN │ └── Air drag: superstructure above waterline. ~5% R_air ≈ 70 kN R_total ≈ 1,400 kN Power required: P = R × V = 1,400,000 × 7.72 = 10,800 kW Add propulsive efficiency losses (~50%): Brake power ≈ 21,600 kW ≈ 25,000 kW installed

The Propeller — Momentum Theory A propeller is a rotating wing that accelerates water backward. Newton's third law: push water back, the water pushes you forward. Thrust: T = ṁ × (V_jet - V_ship) Where: ├── ṁ = mass flow rate of water through the propeller disc (kg/s) ├── V_jet = velocity of water leaving the propeller ├── V_ship = ship speed (approach velocity of water) For a 10 m diameter propeller at 80 RPM: Disc area: A = π × 5² = 78.5 m² Mass flow: ṁ = ρ × A × V_avg ≈ 1,025 × 78.5 × 9.5 = 764,000 kg/s That's 764 tonnes of water flowing through the propeller every second.

The Engine — A Cathedral of Combustion A VLCC engine is a two-stroke, low-speed marine diesel. The numbers are staggering: ├── Type: MAN B&W 7S80ME-C (typical) ├── Cylinders: 7 ├── Bore: 800 mm (a piston you could sit inside) ├── Stroke: 3,450 mm (3.45 meters!) ├── RPM: 72-80 (so slow you can count individual firings) ├── Power: 25,000 kW (34,000 hp) ├── Weight: ~1,350 tonnes ├── Height: ~13.5 m (4 stories tall) ├── Fuel consumption: ~170 g/kWh └── Thermal efficiency: ~50-52% (best of ANY heat engine) Daily fuel consumption at sea: ~80-100 tonnes/day of heavy fuel oil. 15,000 km voyage at 15 knots: ~35 days → ~3,000 tonnes of fuel.

DESIGN SPEC UPDATED: ├── Total resistance at 15 knots: ~1,400 kN (friction 75%, waves 15%) ├── Propeller: single, 10m diameter, 80 RPM, momentum theory T = ṁΔV ├── Disc loading: one big prop > multiple small (less wasted kinetic energy) ├── Engine: 2-stroke diesel, 25,000 kW, 50-52% thermal efficiency └── Fuel consumption: ~90 tonnes/day at sea, ~3,000 tonnes per voyage

PHASE 5: Steer a City Block

You're approaching the Strait of Malacca. Shipping lane: 2 nautical miles wide. Your ship is 330 meters long and weighs 345,000 tonnes. At 12 knots, you need over 2 km to stop. Your turning circle is over 1 km in diameter. You're steering a city block through a hallway, and the brakes barely work. The Momentum Problem At 15 knots (7.72 m/s), the ship's momentum: p = m × v p = 345,000,000 × 7.72 p = 2,663,400,000 kg·m/s ≈ 2.66 GN·s 2.66 billion Newton-seconds. To stop, you need to remove all of that momentum. The only tool: reverse the propeller. But the engine takes 10-15 minutes to stop, reverse direction, and reach full astern power. Full astern thrust: ~60% of ahead thrust ≈ 840 kN Time to stop from 15 knots: t = p / F = 2,663,400,000 / 840,000 = 3,170 seconds ≈ 53 minutes But that's the theoretical minimum using constant full astern force. Real stopping distance accounts for the time to reverse the engine: Crash stop distance ≈ 15-20 ship lengths ≈ 5-6.5 km

Rudder Force and Turning The rudder is a flat plate, approximately 9 m tall × 6 m wide, hinged at the stern. Rudder lift force: F_rudder = ½ × ρ × A × V² × C_L At 15 knots, rudder area 54 m², C_L ≈ 0.8 at 35° deflection: F_rudder = ½ × 1,025 × 54 × (7.72)² × 0.8 F_rudder = ½ × 1,025 × 54 × 59.6 × 0.8 F_rudder = 1,321 kN That sounds like a lot. But the ship's lateral resistance is enormous (you're trying to push 345,000 tonnes sideways through water). The result: ├── Turning circle diameter: ~1,200 m (3.6 ship lengths) ├── Time for 90° turn: ~4 minutes ├── Time for 360°: ~15 minutes ├── Maximum rudder angle: 35° └── Helm response time: 28 seconds (hard-over to hard-over) At 6 knots in a harbor, rudder effectiveness drops to 16% of open-sea values (force ∝ V²). This is why you need tugs.

Tugs — External Steering Muscles In port, a VLCC typically requires: ├── 3-4 tugs, each producing 60-80 tonnes of bollard pull ├── Total tug force: ~250 tonnes ├── Positioned: 2 forward, 1-2 aft ├── Can push the ship sideways at ~0.5 knots └── Docking speed: 0.1-0.3 knots (5-15 cm/s) At 0.2 knots, the ship's kinetic energy is still: KE = ½ × 345,000,000 × (0.103)² = 1,830 kJ Enough to crush any dock structure. A VLCC touching a berth at 0.5 knots instead of 0.2 applies 6× the energy. Docking fenders are rated to absorb specific energy levels. Exceed them and the dock crumbles.

DESIGN SPEC UPDATED: ├── Momentum at 15 knots: 2.66 GN·s — takes 5+ km and 15-20 min to stop ├── Rudder force: F = ½ρAV²C_L → ~1,321 kN at 15 knots ├── Turning circle: ~1,200 m diameter (3.6 ship lengths) ├── Harbor maneuvering: 3-4 tugs, 60-80 tonnes pull each └── Docking speed: 0.1-0.3 knots — even at creeping speed, KE is enormous

PHASE 6: Don't Explode

The cargo is crude oil. Crude oil vapor mixed with air between 1% and 6% concentration will explode on contact with any ignition source — a spark, a hot surface, static electricity. Here's the terrifying part: the FULL tanks are relatively safe. The EMPTY tanks are bombs. When you pump oil out, the space above fills with hydrocarbon vapor at exactly the right concentration to detonate. The Flammability Triangle Three things needed for a fire or explosion:

Inert Gas System — Removing the Third Side You can't remove the fuel (it's your cargo). You can't always remove the heat (static, lightning, mechanical sparks). So you remove the oxygen. The inert gas system takes exhaust gas from the ship's boilers or a dedicated gas generator, scrubs it of soot and sulfur, cools it, and pumps it into the cargo tanks. Composition of inert gas: ├── N₂: ~83% ├── CO₂: ~12-14% ├── O₂: ~2-4% (critical — must be below 5%) ├── H₂O: ~trace └── SOx: scrubbed out Fire requires O₂ above ~11%. Inert gas at 2-4% O₂ makes combustion physically impossible. The tank atmosphere cannot burn no matter how much fuel vapor is present. When Inert Gas Fails ├── 1969: SS Marpessa, Mactra, Kong Haakon VII — three tanker explosions in one month │ All during tank washing. No inert gas. Hydrocarbon vapor + O₂ + static spark. │ Combined death toll: dozens. │ ├── Result: IMO mandated inert gas systems on all tankers >20,000 DWT (SOLAS 1974) │ └── Post-regulation explosion rate dropped by >90% Before mandatory inert gas: ~30 tanker explosions per decade. After: ~2-3 per decade (usually due to equipment failure or human error).

Static Electricity — The Invisible Igniter When crude oil flows through pipes, it generates static charge. When it splashes into a tank, the spray creates a charged mist. Voltage can build to tens of thousands of volts. Safety measures: ├── Loading rate limited to 1 m/s until inlet is submerged (no splashing) ├── After submerged: increase to 3-4 m/s ├── All metal structures bonded and grounded ├── No sampling, gauging, or tank entry until 30 minutes after cargo operations │ (allows charge to dissipate — relaxation time) └── Tank washing with inert gas atmosphere ONLY

DESIGN SPEC UPDATED: ├── Explosive range: 1-6% hydrocarbon vapor in air (empty tanks are the danger) ├── Inert gas system: fills tanks with 2-4% O₂ atmosphere (fire needs >11%) ├── Post-inert-gas regulation: tanker explosions dropped >90% ├── Static control: loading rate limited, 30-min relaxation time after ops └── Flammability triangle: remove oxygen side → explosion impossible

PHASE 7: Survive the Storm

North Atlantic, winter. Wave height: 15 meters — a 5-story building of water hitting the ship every 12 seconds. The ship rolls 25° to port. Then 25° to starboard. The deck tilts like a seesaw. 300,000 tonnes of oil shift inside the tanks. The question isn't whether the ship will rock. The question is: will it come back upright? The answer depends on one number: metacentric height (GM). Stability — The Righting Arm When a ship rolls, three points matter:

For a loaded VLCC: ├── KG (keel to center of gravity): ~13.5 m ├── KM (keel to metacenter): ~16.0 m ├── GM = KM - KG = 2.5 m └── Minimum required GM: 0.15 m (IMO standard)

The GZ Curve — Maximum Righting Arm At larger heel angles, the simple GM × sin(θ) approximation breaks down. You need the full GZ curve — computed by integrating the hull shape at each angle:

Roll Period — The Comfort vs Safety Tradeoff The natural roll period: T = 2π × k / √(g × GM) Where k = radius of gyration (≈ 0.35 × beam for a VLCC = 0.35 × 60 = 21 m) T = 2π × 21 / √(9.81 × 2.5) T = 131.9 / 4.95 T = 26.6 seconds A 27-second roll period means a full cycle (port → starboard → port) takes 27 seconds. This is comfortable — a slow, gentle motion. But if GM increases (ship more stable): ├── GM = 5.0 m → T = 18.8 s (still okay) ├── GM = 8.0 m → T = 14.9 s (getting uncomfortable) ├── GM = 15.0 m → T = 10.8 s (snappy roll — cargo shifts, crew can't work) Too much stability is dangerous too. A "stiff" ship snaps back so violently that: ├── Lashing on deck cargo fails ├── Crew suffer injuries ├── Tank baffles take impact loads └── Structural fatigue accelerates The sweet spot for a VLCC: GM = 2.0-4.0 m, giving T = 20-30 seconds.

DESIGN SPEC UPDATED: ├── Metacentric height GM: 2.0-4.0 m for VLCC (M must be above G) ├── GZ curve: max righting arm ~2.3 m at 40°, vanishing stability ~75° ├── Roll period: T = 2πk/√(gGM) → ~27 seconds (comfortable, safe) ├── Too much GM → snappy roll (damages cargo, injures crew) └── IMO minimum GM: 0.15 m (below this → capsize risk)

PHASE 8: Don't Break in Half

January 1943. The SS Schenectady, a T2 tanker, is sitting calmly at its fitting-out dock in Portland, Oregon. No storm. No collision. The air temperature is -3°C. Without warning, the hull cracks from deck to keel with a sound like a gunshot. The ship breaks nearly in two, sagging into the water. It was brand new. It had never gone to sea. This wasn't an isolated incident. During WWII, over 1,500 Liberty Ships and T2 tankers suffered brittle fractures. About 12 broke completely in half. The cause: a fundamental misunderstanding of how steel fails in the cold. Brittle Fracture vs Ductile Fracture

The Charpy Impact Test How do you measure DBTT? Drop a pendulum hammer onto a notched steel sample at different temperatures: Charpy V-notch test: ├── Cut a 10mm × 10mm × 55mm bar of steel ├── Machine a 2mm deep V-notch in the center ├── Cool/heat to test temperature ├── Swing a 300J pendulum hammer into the notch ├── Measure energy absorbed by the break

Fatigue — Billions of Stress Cycles Even with the right steel, the ship is being bent back and forth by every wave: Typical wave period: 10 seconds Stress cycles per minute: 6 Cycles per day: 6 × 60 × 24 = 8,640 Cycles per year (assuming 300 days at sea): 2,592,000 Cycles over 25-year life: ~65,000,000 But it's worse. Superimposed on the wave-frequency bending are: ├── Springing: hull vibration at natural frequency (~2-3 Hz) ├── Whipping: transient vibration from slamming ├── Thermal cycling: deck temperature varies 50°C over 24 hours Effective fatigue cycles over 25 years: ~10⁸ to 10⁹ Critical fatigue locations: ├── Deck longitudinal connections at transverse bulkheads ├── Bracket toes (stress concentration factor: 3-5×) ├── Welded joints (fatigue strength = 50-70% of parent metal) └── Hatch corners (if any — tankers have fewer than bulk carriers) Classification societies require fatigue life assessment showing minimum 25-year fatigue life with a safety factor of 2.0 — meaning the design must survive 50 years of calculated fatigue loading.

DESIGN SPEC UPDATED: ├── Brittle fracture: steel shatters below DBTT (Liberty Ships: DBTT = +10°C) ├── Charpy test: 27J minimum at service temperature (modern: -40°C for Grade E) ├── Fatigue: 65 million wave cycles over 25 years + springing/whipping ├── Welded joints: fatigue strength 50-70% of parent metal └── Classification rules: 25-year fatigue life with 2.0× safety factor

PHASE 9: Don't Poison the Ocean

March 24, 1989. The Exxon Valdez runs aground on Bligh Reef, Alaska. 37,000 tonnes of crude oil pour into Prince William Sound. 2,100 km of coastline blackened. 250,000 seabirds dead. 2,800 sea otters dead. 300 harbor seals dead. Cleanup cost: $7 billion. Ecological recovery: still incomplete 35 years later. The ship had a single hull. The Exxon Valdez changed everything. The result: the Oil Pollution Act of 1990 (OPA 90) in the US, and MARPOL Regulation 13G internationally, mandating double hulls on all new tankers. Single Hull vs Double Hull

Segregated Ballast — Keeping Oil and Water Apart When a tanker is empty, it rides too high in the water — the propeller comes out of the water, the rudder loses effectiveness, the ship is dangerously unstable in wind. You need to add ballast water (seawater) to push the ship down. Old method (pre-MARPOL): pump seawater directly into the cargo tanks. Problem: when you discharge the ballast before loading, oily water goes into the ocean. Modern method: Segregated Ballast Tanks (SBT) ├── Dedicated ballast tanks in the double hull space ├── Ballast water NEVER contacts cargo tanks ├── Separate piping, separate pumps ├── Ballast water management: must be treated before discharge │ (to prevent invasive species transfer — IMO BWM Convention) └── Ballast capacity: ~40% of deadweight for a VLCC A loaded VLCC carries zero ballast. An empty VLCC carries ~100,000-120,000 tonnes of ballast water to maintain safe draft and stability.

DESIGN SPEC UPDATED: ├── Double hull: inner + outer hull with ≥2m gap (MARPOL 13G) ├── Exxon Valdez (1989): 37,000 tonnes spilled, $7B cleanup → OPA 90 ├── Major spills dropped 99.7% from 1970s to 2020s ├── Segregated ballast: ballast water never contacts cargo tanks └── Ballast capacity: ~100,000-120,000 tonnes when empty

PHASE 10: Navigate the World