Palm Jumeirah has no concrete foundation — it's 94 million m³ of dredged marine sand behind a rock breakwater. It sinks ~5mm/year, loses sand continuously to wave action, and one replenishment operation moved 3.5 million m³. Dredging also buried the coral reef that used to dissipate the wave energy.
Between 2001 and 2006 Nakheel dredged sand off the Persian Gulf seabed and sprayed it into the shape of a palm tree. 5.72 square kilometres of new land. 94 million cubic metres of marine sand. Around $12 billion. It added 78 kilometres of coastline to Dubai, which was the whole economic case — beachfront commands premiums inland property doesn't, so manufacturing coastline is manufacturing margin.
What it doesn't have is a foundation. There's no concrete base, no seawall at the trunk. It's sand for the fronds and trunk, plus a crescent rock breakwater to shield the fronds from open-ocean wave action — 7 million tonnes of rock, quarried in the Hajar Mountains and barged 100 kilometres.
Which makes it structurally different from everything else people file it alongside. The Falkirk Wheel is steel and concrete with a 120-year design life. The Wuppertal Schwebebahn is structural steel and has been running since 1901. The Palm Jumeirah is built out of the one material water is specifically good at relocating, and its service life is a function of how much money Dubai spends putting sand back versus how fast the Gulf takes it away.
The maintenance is existential, not preventive
NASA satellite data has the island subsiding at roughly 5 mm/year. The fronds lose sand continuously to wave action and tidal currents — one major replenishment operation moved over 3.5 million cubic metres.
Dredged marine sand is a worse structural material than it looks, because it lacks what natural coastal sand has: shell fragment content, organic matter, and the compaction history that gives a natural beach cohesion. You're placing a poorly-graded, uncompacted, cohesionless fill and then asking it to hold a shape against continuous hydraulic loading.
The comparison worth drawing is the Delta Works, which protects land that exists naturally and would flood without intervention. Failure there means inundation of terrain that still exists underneath. Palm Jumeirah protects land that doesn't exist naturally and would dissolve without intervention. Failure means the terrain stops being terrain. Same category of engineering, fundamentally different failure mode.
Self-defeating by construction
The dredging buried roughly 1.2 square kilometres of coral reef under sediment.
Coral reef is a natural breakwater. It dissipates wave energy before it reaches shore — which is exactly the function the 7-million-tonne rock crescent was built to perform. The construction destroyed the free, self-maintaining, self-repairing wave attenuation structure and replaced it with a quarried one that requires monitoring and repair, on the same site, to protect the thing that was built on top of the first one.
The sediment plume also smothered oyster beds and seagrass under about five centimetres of silt across a wide radius, and turbidity cut light penetration enough to kill photosynthetic organisms in the affected zone. Artificial reef structures have since been installed along the breakwater with some documented recovery, but the remediation is cosmetic against the scale of the original loss.
Two consequences nobody designed for
The first is downstream erosion. Dubai's beaches lose 10,000 to 15,000 cubic metres of sand a year naturally. The island disrupts alongshore sediment transport — the mechanism that moves sand from one beach to the next along the coast. Sand that would have travelled east now accumulates on the western side of the island, and the beaches east of it starve. Some sections have retreated at up to 10 metres per year, and Dubai has spent millions on seawalls, groynes, and rock revetments to arrest erosion the island itself caused.
The second is water quality. The crescent breakwater that protects the fronds from storm waves also prevents tidal flushing, so the lagoons don't circulate properly. That produces stagnant zones, algal blooms, and the smell tourists have been reporting since it opened. Engineers deepened channels and installed circulation infrastructure — you're now running engineered water movement to substitute for the tidal exchange the breakwater blocks. Partially effective.
Both are the same class of problem: a structure that solves its primary design objective by interrupting a natural transport process, and then requires secondary infrastructure to replace the process it interrupted.
They're building the bigger one
Palm Jebel Ali — 50 percent larger, shelved after 2008 — relaunched in 2023. Redesigned masterplan with 80 hotels, homes for 35,000 families, six marinas. Jan De Nul took an AED 810 million marine works contract covering dredging, reclamation, beach profiling, and sand placement. First eight fronds targeted site-ready by early 2025, completion revised to 2027.
Meanwhile The World — 300 islands arranged as a world map, four kilometres offshore — sat largely undeveloped for a decade after the crash, visibly dissolving on satellite imagery. Individual island owners carry their own shoreline protection costs, running into millions of dirhams annually per island, which is a reasonable explanation for why most of them remain empty. A Monaco-themed resort opened in 2022 and a Sweden Island development is in progress.
Full write-up on the construction, the sediment transport problem, the reef loss, and Palm Jebel Ali: https://unteachablecourses.com/palm-jumeirah-dubai-artificial-islands-2026/
The engineering question I'd actually like an answer to: is there a design approach for large-scale sand reclamation that gets you to something self-sustaining, or is perpetual replenishment inherent to the method? Geotextile containment, cement or polymer stabilisation of the fill, deliberately engineered reef structures as primary rather than supplementary wave attenuation — do any of these change the maintenance curve meaningfully at 5+ km² scale, or is the honest answer that if you build land out of cohesionless dredged fill in open water, you have signed up for an annual sand budget in perpetuity and the only variable is its size?