Electrolytic sea structures

Seawater electrolysis coats a steel cathode with a thin mineral skin and stops the steel rusting while current flows. It does not grow a rubble-mound’s worth of rock. Numbers below are tagged as measured, published claim, or input.

The process is Hilbertz mineral accretion (1979; later “Biorock” / Seacrete). Cathode: water is reduced to H2 + OH. If surface pH stays moderate, CaCO3 (aragonite) deposits. If pH is higher, Mg(OH)2 (brucite) deposits instead — softer, two electrons per formula unit. Lab plates: compact deposit at ≤150 µA/cm² (1.5 A/m²); brucite dominates as current rises (Electrochimica Acta 2023).

Independent Danish work grew ~0.4 cm/yr and ~7.2 kg/yr on 50 kg of rebar using 215 kWh/yr (EUDP final report). Goreau’s “<1–2 cm/year” and “several times harder than Portland cement” are proponent claims (Goreau & Prong 2017). Independent puncture tests on slow-grown material were in the same class as C20/25 concrete, not 3× Portland (Frontiers 2021).

The Pulau Gangga installation used a 2 m³ river-stone gabion under each mesh sheet. Goreau writes that gabions give faster initial wave slowing, and that electrolysis is what stops the gabion wire rusting. No measured wave-transmission coefficient (Kt) for those modules was published. Photographs show waves interacting with the mesh (their Fig. 8).

Power-off: steel is protected only while current flows (Hilbertz 1979; Electrochimica: <55 µA/cm² allowed corrosion). Goreau 2012 says biological growth benefits drop when power is off. A 2020 lab study found calcareous/sediment agglomerate can dissolve after current stops (Johra et al., CBM). A “hotels switched it off and the structures collapsed” story is not in the papers opened for this page, so it is not used here.


Cost comparison — published figures only

These are not the same structure and not the same wave climate. A modern rubble-mound armor layer sized for a 3 m significant wave is a different job from a shallow gabion + mesh on an Indonesian reef flat. Showing both prevents fake equality.

Modern rubble-mound (armor layer) Electrolytic mesh / Biorock
What was priced Primary armor layer (not core + filter + armor as a full mound) No itemized project invoice published for Pulau Gangga
Design sea Hs = 3.0 m, Tp = 10 s Shallow seagrass / intertidal reef flat (Gangga 2016). Wave height not given as a design Hs
Unit price used 2,500,000 IDR/m³ marine concrete Not stated for steel, stone, cable, or labor
Published cost / m shoreline 85–215 million IDR/m (KD 24 → KD 2) Claimed $20–1,290/m vs “other methods” $60–155,000/m
Evidence peer-reviewed Suwondo et al. 2025, Indonesian market prices. Armor-layer total from Hudson sizing, not a built bid. claim Goreau & Prong 2017 cite “[16]” for those dollars. Their [16] is Ferrario et al. 2014 Nat. Commun. on coral-reef wave dissipation — not a cost study. Treat the dollar range as unsourced in that paper.
Grown mineral vs concrete, per kg Concrete ~€0.04/kg (80–100 €/m³) Deposit €1–12/kg at 7.2 kg/yr from 215 kWh/yr
Evidence independent report EUDP Table 5 (Hanstholm / Aalborg). Authors say this is not a breakwater-per-metre comparison.
Electricity None after placement Gangga: “about one air conditioner” for the whole project — no kWh. EUDP: 215 kWh/yr, 0.2 A, 1.2–2.5 V, 50 kg rebar (~280 m)
Stone / steel actually used (Gangga) Each module: gabion 1×1×2 m of 20–50 cm river stone; welded mesh ~2.1×5.4 m, 15 cm grid; 48 modules; 1 supply per 4 units, ~100 m cable to shore
Independent per-metre audit of a Biorock shore job None found

IDR figures left in rupiah as published. Converting them to dollars would add an extra assumption that is not in Suwondo et al.

Gangga bill of materials × your local prices

Quantities are from Goreau & Prong 2017. Prices are yours. Empty fields stay out of the total. This is not a bid; placement, boats, anodes, and controllers are unpublished so they are omitted unless you add them.

Block sizer — Texas solar and battery

Enter the cage you want (L × W × H). Bars are an orthogonal mesh on all six faces (shared edges counted once); if height is below one pitch it is a single sheet. Default 2 × 1 × 1 m is the Gangga gabion core, not the 2.1 × 5.4 m sheet. Bar Ø is not in Goreau 2017 — 5.4 mm is implied by EUDP 50 kg / ~280 m. Current I = j × steel area does not depend on voltage; power, PV, and battery do. PV is PVGIS 5.2 for Port Aransas (Hilbertz 1979 field site), not all of Texas. 14% system loss is already in those Ed numbers.

2.5 V

Voltage vs yield and December PV

Top: kg CaCO3 per kWh (100% Faraday, n = 1 mol e/mol as in Goreau’s 0.62 kg/kWh at 6 V). Bottom: December PV (Wp) at Port Aransas. Night battery Wh = P × 13.76 h, so it is the same shape as the amber line. Mass/year at fixed j is flat in V; energy and hardware grow with V.

Monthly PVGIS Ed and array to match this load

Port Aransas 27.833°N, 97.061°W; 25° tilt, south; PVGIS-NSRDB 2005–2015. December night 13.76 h is geometric solstice at this latitude (no refraction).

Month Ed (kWh/kWp·d) Wp to cover 24 h at this V

Sources