Corrosion, through-hulls and bottom paint — the practical course for the haul-out crew. What eats boat metal, how to measure and service a through-hull right, and how to read what the paint is telling you.
Before you start
The practical course for what lives under the waterline — corrosion, through-hulls and bottom paint. Enough ABYC to know the difference between galvanic and stray-current corrosion and how to stop it, plus the hands-on skills: measure a through-hull right, remove and replace it, service the seacocks, and read the paint. Each chapter ends in a quick check; an 80% final unlocks your certificate.
Size by bore, match the thread standard (NPS/NPT/BSP), and never mix them.
Galvanic vs DC/AC stray current; isolators, isolation transformers, and bonding.
Flowering paint, pink bronze and bright anodes all tell you where to look.
Chapter 1
The haul-out is where you catch the thing that puts boats on the bottom.
Before any theory, the reason this course exists. The data on sinkings is blunt: for every boat that sinks at sea, about four sink right at the dock — quietly, while nobody is aboard. And when you look at why dockside boats sink, the water almost always comes in below the waterline through a fitting.
| Cause of a dockside sinking | Share |
|---|---|
| Leak at an underwater fitting / hose (stuffing box, through-hull, seacock, hose, raw-water plumbing) | ~50% |
| Rain / snow / sleet accumulation (cockpit drains, covers) | ~32% |
| Fitting near the waterline that submerged (heeled, loaded, low transom) | ~9% |
| Mooring / dock arrangement (hung up, holed) | ~8% |
| Other | ~4% |
This is the practical ESI manual for the metal and plastic that lives under the waterline: corrosion, through-hulls, and bottom-paint inspection. It covers enough ABYC corrosion theory to explain galvanic corrosion, DC and AC stray current, galvanic isolators and isolation transformers, and the US-vs-European bonding difference — but it is not the full ABYC corrosion-specialist course. When a job goes past what is here, you will know to escalate (Chapter 13).
Sources: BoatUS Marine Insurance / Seaworthy claim studies (≈4 boats sink at the dock for every 1 at sea; ~50% of dockside sinkings begin at an underwater fitting, stuffing boxes leading); ABYC framing.
Chapter 2
Which metal eats which — and why stainless is a trap below the waterline.
Put two different metals in seawater and connect them, and you have built a battery. The galvanic series ranks metals from active (anodic / least noble) to noble (cathodic). The rule is simple: the more active metal corrodes to protect the more noble one, and the further apart they sit on the series, the faster it happens.
| Most active (corrodes / sacrifices) | → Noble (protected) |
|---|---|
| Magnesium · Zinc · Aluminium | the anode metals — they give themselves up |
| Mild & galvanised steel · cast iron | hulls, keels, fasteners |
| Stainless (active, in a low-oxygen crevice) | the danger state — see below |
| Lead · tin · brasses · manganese bronze | brass is high-zinc → dezincifies |
| Silicon / aluminium bronze · copper | good below-waterline metal |
| 316 stainless (passive) · titanium · graphite | noblest — protected at others’ expense |
Stainless steel is confusing because it lives in two places on the series. In clean, oxygen-rich, moving water it is passive (a thin oxide film) and very noble. But starve it of oxygen — buried in a crevice, packed in a stuffing box, under marine growth, or inside a through-hull — and the film breaks down. The metal goes active and suffers crevice corrosion and pitting, often hidden, often without warning. That is why stainless is the wrong choice for below-waterline through-hulls and seacocks.
| Metal | Where / how it behaves |
|---|---|
| Silicon / aluminium bronze | The proper below-waterline metal — through-hulls, seacocks, struts. Noble, durable. Must be real bronze, not high-zinc brass (Ch 8). |
| Marelon (composite) | Not a metal at all — glass-reinforced polymer. Immune to corrosion and electrolysis. ESI-preferred (Ch 8 & 12). |
| Stainless (304/316) | Fine above the waterline; poor below it (crevice corrosion). Shafts are a managed exception. |
| Aluminium | Outdrives, saildrive legs, some hulls. Very active — must be protected by anodes and kept away from copper paint. |
| Zinc / aluminium / magnesium | The sacrificial anodes — deliberately the most active metal aboard (Ch 3). |
Sources: standard marine galvanic series (ABYC / Steve D’Antonio / Practical Sailor); stainless active-vs-passive crevice-corrosion references.
Chapter 3
The slow, normal corrosion every boat fights — and the zincs that win the fight.
Galvanic corrosion is the everyday, low-voltage corrosion every boat lives with. It needs three things, and removing any one stops it:
Close that loop and the active metal gives up millivolts of potential and slowly dissolves. It is normal. We do not eliminate it — we redirect it onto a metal we are happy to lose.
A sacrificial anode is a lump of a more active metal bonded to the metals we want to keep. Because it is the most active thing in the circuit, it corrodes first and the protected metals do not. Everyone calls them “zincs,” but the metal depends on the water:
| Anode metal | Use it in | Notes |
|---|---|---|
| Zinc | Salt water | The traditional choice. Passivates (crusts over and stops working) in brackish/fresh. |
| Aluminium | Salt and brackish | ESI default for most boats: lasts longer, more capacity, keeps working in brackish, lighter, more eco-friendly. Safe in salt too. |
| Magnesium | Fresh water only | Very active — perfect for low-conductivity fresh water; consumed almost instantly in salt. Never use in salt. |
Sources: BoatZincs / Performance Metals anode-by-water guidance; ABYC E-2-2025 cathodic protection; standard marine anode practice (50% rule, never paint, clean contact).
Chapter 4
The fast killer: not millivolts over years, but volts over days.
Galvanic corrosion is slow and measured in millivolts. Stray-current corrosion is the abnormal one — driven by stray volts from the electrical system or the dock — and it is brutally fast. A stray-current fault can eat a through-hull or a prop in days to weeks, not years. When you see rapid, localized metal loss, think stray current.
Somewhere on board, DC current is escaping its wire and taking a shortcut through the water. A classic source: a bilge-pump or accessory wire with chafed insulation sitting in bilge water, or a corroded connection, leaking current into the bonding system. The current flows out through one underwater metal, through the water, and back — and the metal where the current leaves into the water erodes ferociously.
The moment you plug into shore power, the green safety ground wire connects your boat’s underwater metals to the dock and to every other boat on it. Two things ride that shared wire:
Sources: ABYC E-11-2025 (AC/DC), ABYC corrosion guidance; Steve D’Antonio / BoatUS on stray-current and ESD; marine electrical references.
Chapter 5
Two ways to break the dockside loop. One cheap, one complete.
From Chapter 4: the green shore ground links your metals to the whole marina, carrying galvanic current and stray current onto your boat. You cannot just cut the safety ground — it has to stay connected for AC fault protection. So we interrupt the corrosion path while keeping the safety path. Two devices do this.
A galvanic isolator sits in the green ground wire. It uses diodes that will not conduct below about 1.2–1.4 V — so the small galvanic voltages (millivolts to ~1 V) that drive dockside galvanic current are blocked, while a real AC fault (which easily exceeds that threshold) still passes through to keep the safety ground working.
An isolation transformer takes shore AC into one coil and induces it into a second, separate coil that feeds the boat. There is no metallic connection between dock wiring and boat wiring — the energy crosses magnetically. That single break does it all:
Sources: ABYC A-28-2024 (galvanic isolators) & E-11 (AC systems / isolation transformers); Victron / Blue Sea / ProMariner technical notes on isolators vs transformers.
Chapter 6
Two philosophies under the floorboards. Know which boat you are on.
A bonding system is a heavy green wire that ties the underwater metals — through-hulls, seacocks, shaft, rudder, struts, engine — together, and to the sacrificial anodes (and, per the standard, to DC negative/ground). It exists so all those metals share cathodic protection from the anodes and so stray current and lightning have a defined path. Here is the part that surprises techs: the US and Europe disagree on whether to bond at all.
US boats built to ABYC generally bond the underwater metals to a common bonding system tied to anodes.
Many European builders (think Beneteau, Jeanneau, and others, to ISO 13297 practice) leave the underwater through-hulls UNBONDED — each bronze fitting stands electrically alone, relying on good-quality bronze’s own corrosion resistance, while the engine/shaft/saildrive get their own anode.
Sources: ABYC E-2-2025/E-11 bonding; ISO 13297; Steve D’Antonio and Practical Sailor on the “to bond or not to bond” US-vs-Europe difference.
Chapter 7
Stop guessing whether a boat is protected. Read it in millivolts.
You can argue about anodes all day, or you can measure. A silver/silver-chloride (Ag/AgCl) reference electrode hung in the water, with a multimeter on millivolts connected to the bonding system, tells you the boat’s hull potential — exactly how well-protected the underwater metals are.
| Vessel / metal | Protected hull potential (mV, more negative = more protected) |
|---|---|
| Aluminium (outdrives, ali hulls) | −950 to −1100 mV |
| Fibreglass boat, common metals (bronze/steel, no aluminium) | −750 to −1100 mV |
| The protection rule | A metal must be driven at least 200 mV more negative than its own resting potential to count as protected. |
Under-protected (not negative enough): the metals are still corroding — add/parallel anodes or fix the bond. Over-protected (too negative, past about −1100 mV, especially on aluminium and wood): you start stripping paint, generating hydrogen, and causing alkali damage. Over-zincing is a real failure mode — more anodes is not always better.
Sources: BoatZincs Corrosion Reference Electrode guide; Steve D’Antonio reference-cell testing; ABYC E-2-2025 protected-potential ranges (Ag/AgCl: ali −950/−1100 mV, fibreglass −750/−1100 mV; 200 mV shift; unplug shore cord to test).
Chapter 8
Real bronze, composite, and the dezincification test that saves boats.
Proper silicon or aluminium bronze is the classic below-waterline metal. The danger is the look-alike: high-zinc yellow brass plumbing valves and cheap import fittings. In seawater the zinc leaches out of brass — dezincification — leaving a porous, weak, copper-coloured shell that crumbles and can fail without warning.
| Sign of dezincification | What you see / do |
|---|---|
| Colour shift | Healthy bronze is golden; dezincified brass goes pink / coppery. |
| White residue | Powdery white zinc-oxide deposits around the fitting. |
| Scratch test | Scrape with a knife/screwdriver: bright gold = good bronze; pink/copper under the surface = dezincified → condemn. |
| Tap test | A sharp tap: solid bronze rings; dezincified metal sounds dull and may crumble or flake. |
Marelon (Forespar) is a glass-reinforced polymer composite, not a metal. Forespar describe it as offering complete freedom from corrosion and the ravages of electrolysis, above and below the waterline. For ESI that means a lot:
| Type | Verdict |
|---|---|
| Flanged seacock (bronze or Marelon), through-bolted to a backing block, straight (NPS) threads | Best. Supported, serviceable, correct threads. ABYC intent. |
| Ball valve threaded onto a through-hull mushroom | Common but not ideal — thread mismatch risk (NPT vs NPS, Ch 9), no flange support. Acceptable only with matched straight threads and proper backing. |
| Gate valve | Banned below the waterline. Corrodes, jams, the stem fails, and you cannot tell open from shut. Replace on sight. |
Sources: Forespar Marelon material data; Steve D’Antonio / Practical Sailor on dezincification and seacock vs ball-valve vs gate-valve; ABYC H-27-2021 seacock practice.
Chapter 9
If you remember one chapter, make it this one. Measure the bore and the thread — not the outside.
More wrong parts get ordered — and more seacocks leak or shear — from bad measuring than from anything else under the waterline. Through-hull sizing is genuinely counter-intuitive, so slow down and do it right.
A through-hull or seacock is named by its nominal bore / pipe size — the hole the water flows through — not by the outside diameter of the threads. A “1-inch” through-hull has roughly a 1" bore / 1" pipe thread, but its thread OD measures larger (around 1.3"). Measure the OD with calipers, call it “1¼-inch,” order that, and the part is wrong.
| Thread | What it is | Where |
|---|---|---|
| NPS (National Pipe Straight) | Parallel/straight threads — through-hulls & proper seacocks use these so they mate fully along the whole thread. | US marine through-hulls & seacocks |
| NPT (National Pipe Taper) | Tapered threads — a hardware-store ball valve. Threaded onto a straight through-hull it grabs only a few threads. | Plumbing / wrong on a straight through-hull |
| BSP (British/parallel, metric world) | Different angle & pitch from NPS/NPT — common on European boats. | Euro / imported fittings |
Never mix NPS and NPT. A tapered NPT valve on a straight NPS through-hull engages only a turn or two of thread — it looks tight, then leaks, strips, or shears off below the waterline. European boats are frequently BSP/metric; measure and match the standard, do not force a US fitting on.
Sources: Groco / Forespar / Marelon sizing data and ABYC H-27-2021; Compass Marine and Steve D’Antonio on through-hull sizing by bore and NPS-vs-NPT thread matching.
Chapter 10
Heat the stuck ones out, prep the hole for a real bond, and bed it in the right Sikaflex.
| Arrangement | Notes |
|---|---|
| Mushroom through-hull + flanged seacock | The good install — back-bolted to a backing block. |
| Through-hull + bronze or Marelon ball valve | Common; check thread match & support. |
| Old tapered-plug seacock (Wilcox-Crittenden / Groco) | Serviceable bronze classic — see lubrication, Ch 12. |
| Gate valve | Remove on sight (Ch 8). |
A through-hull is held by its bedding, and the fastest way to release bedding is to soften it with heat. ESI’s preferred method on a stuck fitting is to heat it — not to fight it cold or jump straight to cutting.
The new fitting is only as good as the surface you bond it to. Take the hole back to clean, sound, dry laminate and give the adhesive somewhere to grip.
ESI beds below-waterline through-hulls in Sika Sikaflex polyurethane. Pick the grade to the job:
| Product | What it is | Use it for |
|---|---|---|
| Sikaflex-291 | Multipurpose marine adhesive/sealant — flexible, rated above and below the waterline, the everyday bedding compound. | Standard below-waterline bedding where you want a strong seal that can still be serviced later. |
| Sikaflex-292 | High-modulus structural PU adhesive (3 MPa tensile, ~400% elongation); seawater-resistant and rated by Sika for through-hull fittings below the waterline. | Structural bonds — when the fitting is bonded in, not just bedded. Stronger and more permanent (harder to remove). |
Sources: Sika Sikaflex-291 & -292i product data sheets (291 above/below waterline; 292i structural, seawater-resistant, rated for through-hull fittings below the waterline; clean/dry/grease-free prep, triangular bead, open/skin time); ABYC H-27-2021; Compass Marine / Practical Sailor through-hull replacement (clean to sound glass, bevel for bond area, cored-hull epoxy ring).
Chapter 11
Half of all dockside sinkings start here — at a hose, a clamp, or a stuffing box.
Back to the data from Chapter 1: about half of all dockside sinkings start at an underwater fitting or hose, with stuffing boxes leading. The seacock is only half the connection — the hose and the clamps are the other half, and they are a wear item. This is pure sinking-prevention.
Hose has a service life: it hardens, cracks and lets go. Check every below-waterline hose at haul-out.
Use marine-grade, reinforced hose rated for the application — never thin clear vinyl below the waterline.
A clamp is only as good as its worst part. The ESI standard is the all-316L stainless, non-perforated, worm-drive band clamp — we use PYI (the same solid-band design as AWAB/ABA).
| Use this | Avoid this |
|---|---|
| Solid / non-perforated embossed band — strong, no slots to crevice-corrode or tear | Slotted / perforated bands — the slots trap water, crevice-corrode and break at the slot |
| All 316 stainless, screw and housing included (PYI: one-piece machined 316L screw cage, rolled edges) | “Stainless” clamps with a plated carbon-steel screw that rusts and seizes |
| Rolled-edge band that will not cut soft hose and takes high torque | Sharp-edged cheap bands that bite into and damage the hose |
| Full 316L band AND screw/worm (non-magnetic, marked) | 304 / 18-8 “stainless” — especially an 18-8 worm screw — that pits, streaks and seizes below the waterline |
316 vs 304/18-8 — the screw is the tell. Most hardware-store “stainless” clamps are 304 (18-8), and the worm screw is often the cheapest part of the lot. In salt water 304/18-8 has far less resistance to chloride pitting and crevice corrosion than 316, so the screw rusts, streaks the hull and seizes solid exactly when you need to back it off. Below the waterline ESI uses full 316L — band and screw (PYI/AWAB). Quality 316 marine clamps are non-magnetic and marked; do not trust the word “stainless” on the bag.
Every below-waterline hose connection gets two clamps wherever the barb is long enough to land both fully.
Wet-exhaust hose is large, soft and runs hot. A T-bolt clamp (constant-torque, solid band) delivers high, even clamping force around big soft exhaust hose that a small worm band cannot — so ESI uses T-bolt clamps on exhaust connections.
Sources: PYI Inc. 316L non-perforated solid-band hose clamps; AWAB/ABA all-316 clamp data & Practical Sailor long-term clamp test; Compass Marine double-clamping practice; ABYC H-27-2021 (seacocks/hose) & P-1 (exhaust); SAE J2006 wet-exhaust hose.
Chapter 12
A seacock you cannot turn is no safety device at all.
Every seacock has one job in an emergency: shut. A seized one is useless. So the most important maintenance under the waterline is dead simple — exercise every valve through its full open/close travel, regularly, and lubricate it on a schedule. Forespar’s guidance is a good universal rule: work the handle through the full path every ~30 days, and do a proper lube service twice a year.
Marelon is corrosion-proof, but the moving parts still need lubrication and exercise. The critical rule is what you lubricate it with:
Sources: Forespar Marelon Valve & Seacock Maintenance tech tip and MareLube product data (non-petroleum PTFE, exercise every 30 days, lubricate twice yearly); Groco tapered-plug service; standard waterproof-grease/LanoCote practice.
Chapter 13
Read the paint and the boat tells you exactly where the trouble is.
Most antifouling is copper-based (cuprous oxide). Beyond keeping growth off, the paint is a free corrosion survey — if you know how to read it. The headline tell is flowering.
Flowering is paint that lifts, bubbles, or blisters in a star/halo radiating out from a through-hull, prop, strut, or fitting. It is the boat flagging an electrochemical reaction at that spot — corrosion or stray current. At a cathode the reaction makes gas and alkali that push the paint off; at an anode the metal itself is going. A flower of lifted paint around a bonded fitting often points to current or over-protection; around an isolated fitting it can mean a developing corrosion cell.
Copper antifouling is noble. Put it in direct contact with a bare active metal — an aluminium outdrive or saildrive, a steel hull — and you have built a galvanic cell that eats the metal fast.
| What you see | What it usually means |
|---|---|
| White/chalky halo, paint peeling near a bonded metal | Possible over-protection (too much / too-negative anode) — confirm with the reference electrode (Ch 7). |
| Pink/copper showing on a “bronze” fitting | Dezincification (Ch 8) — scratch/tap test and likely condemn. |
| Anodes clean and bright (in salt) | Not connected, or wrong metal — no protection happening. |
| Pitting / grooving on prop, shaft, rudder | Stray current or galvanic attack — investigate the source. |
Sources: Practical Sailor / Steve D’Antonio on paint blistering & flowering as corrosion/stray-current signals; antifouling makers (Pettit/Interlux) on copper paint vs aluminium and barrier coats.
Chapter 14
Put it together into a repeatable haul-out survey — and know your limit.
| Pattern | Likely mechanism |
|---|---|
| Slow, general loss near dissimilar metals; anodes wasting | Galvanic — normal; check anodes, bonding, water type |
| Fast, localized destruction; anodes gone in weeks; one fitting hit | Stray current — hunt a DC leak / shared shore ground |
| Hidden pitting/cracking on stainless in a low-oxygen spot | Crevice corrosion — wrong metal below the waterline |
| Pink, porous, dull-sounding “bronze” | Dezincification — high-zinc brass; condemn |
Condemn (replace now, do not “watch”): dezincified/pink bronze; a cracked or weeping seacock; a seized valve; a fitting more than ~50% wasted; stainless below the waterline showing crevice attack.
Escalate to a corrosion specialist / ABYC corrosion-certified tech when: loss is persistent and unexplained; anodes burn through repeatedly; you suspect stray current you cannot isolate; the boat is metal-hulled or has an ICCP system; or the bonding is non-standard/mixed (Ch 6). That is where this practical course hands off to ABYC E-2-2025.
Sources: ABYC E-2-2025 / A-28 / E-11 / H-27; Steve D’Antonio and ABYC corrosion-survey practice; ESI haul-out standard.
Field reference
Field-tested: the symptom, the usual cause, and the fix. Always diagnose before you replace parts.
| Symptom | Likely cause | Fix |
|---|---|---|
| Anodes waste away in weeks | Stray current, over-bonding, or a neighbour’s DC leak on the shared shore ground | Hunt the DC leak, fit a galvanic isolator/transformer, confirm with a reference-electrode reading |
| Anodes stay bright / barely touched | Not actually connected, or the wrong metal (zinc in brackish) | Check the bond stud and contact; switch to aluminium for salt/brackish |
| “Bronze” fitting looks pink & porous | Dezincification of high-zinc brass | Condemn it — replace with proper silicon-bronze or Marelon |
| Bottom paint flowering around a fitting | A corrosion or stray-current cell at that spot | Investigate the cause — do not just sand and repaint over it |
| Seacock seized / won’t turn | No service or lubrication | Exercise and lube on schedule — MareLube (non-petroleum) on Marelon |
| Hull potential reads too negative | Over-protection / over-zincing | Reduce anode mass; protect wood & aluminium from alkali damage |
| Reference-electrode reading looks wrong | Shore cord still plugged in | Unplug the shore cord entirely — breaker-off is not enough |
| Hose weeping at a clamp | Tired hose, a single clamp, or a 304/18-8 screw that rusted | Replace the hose and double-clamp with full-316 PYI clamps |
Final certification
A short check ran at the end of each chapter. Pass this final at 80% to unlock your certificate.