Floating Islands That Work: TVNI’s Revised Vetiver Floating Island Specification v2.1
A Vetiver Floating Island is a simple idea: a buoyant raft planted with vetiver grass, anchored on a polluted pond or lagoon, with the plant’s dense vertical roots hanging one to two meters into the water. The root curtain takes up nitrogen, phosphorus, and ammonia, hosts the microbial biofilms that do much of the treatment work, traps sediment, shades out algae, and shelters fish. TVNI has just issued a fully revised design specification for these systems, and the revision changes some long-standing assumptions.
Start the plants before you need them. The single biggest determinant of success in hostile water — sewage, saline water, industrial effluent — is not the island design but the plant. Vetiver pre-grown for about three months in small pots, deployed with its root ball intact, adapts rapidly where bare-root slips struggle or die. The new specification makes pre-grown stock a requirement, not a preference, for any hostile water body.
Placement beats coverage. The old rule of thumb — cover 3–5% of a pond and expect results — is not supported by the research. Paired-pond studies show that around 10% coverage produces no measurable whole-pond improvement, while 20% does. But the most useful recent finding points the other way: islands concentrated at a pond’s inlet, where the polluted water arrives at full strength, achieved major nutrient-removal gains at just 1–5% total coverage. Placement substitutes for area. For large lakes and reservoirs the lesson is decisive: forget whole-lake ratios entirely and station the islands across the inlet flows and polluted tributary mouths.
China has already built the numbers. Dozens of village-scale sewage systems in China use vetiver islands as the polishing stage after a small biochemical reactor, with years of monitoring behind them (Feng, ICV-7, 2023). They supply design values the field has lacked: vetiver uptake is strongest at COD concentrations of 700–1,600 mg/L, and the grass dies near 2,000 — so high-strength effluent must be pretreated before it reaches the islands. Notably, the grass itself was less than 0.2% of project cost; the raft hardware is what you pay for.
Harvest is the treatment. Nutrients captured in plant tissue only leave the water when the shoots are cut and carried away; biomass left to decay simply returns its load to the pond. Where the water carries nutrients alone, that harvest is a resource — young regrowth is good livestock forage, and the leaves make excellent mulch. Where the water carries metals, hydrocarbons, or leachate, it is not: test the tissue or dispose of it safely. The rule of thumb in the new specification: if you wouldn’t drink the water after treatment, don’t feed the grass without testing it.
Every figure in the revised specification carries an evidence tag separating what has been demonstrated in the field from what is inferred and still needs testing — and the open questions are listed, because they are invitations. The full document is available on TVNI’s website, specifically: https://drive.google.com/file/d/17XHDTV5HcoiwbjYd5H80eSFAOQtLN73v/view
Also Paul Truong’s photo essay of vetiver pontoons, their construction and placement is very illuminating and can be found at: https://vetiver.org/AUS_Pontoons.pdf