Corrosion Engineering: The Complete Engineering Guide — Electrochemistry, the Eight Damage Mechanisms, Materials Selection by PREN, Coatings & Cathodic Protection Design, Inspection & RBI, and What Corrosion Costs India

Corrosion is the largest unmanaged materials bill in industry — NACE's IMPACT study priced it at US$2.5 trillion a year, 3.4 % of global GDP, and a July…

·

Corrosion Engineering: The Complete Engineering Guide

Every kilogram of steel industry installs is on a slow schedule to become iron oxide again — the only questions are how many years it takes, whether it fails gracefully or suddenly, and who pays. That payment is enormous and measurable. NACE International's IMPACT study put the global cost of corrosion at US2.5 trillion per year, 3.4 % of world GDP (2013 basis), and — the part that makes it an engineering problem rather than a law of nature — estimated that 15–35 % of that bill is avoidable with technology that already exists: coatings, cathodic protection, correct alloy choice, and inspection. For India the arithmetic has only sharpened. The IMPACT India cost study used India's 2011–12 numbers as one of its five national datasets: direct cost US26.1 billion (2.4 % of GDP), indirect cost US39.8 billion, avoidable share US9.3 billion. A Nomura report released in July 2026 estimates India now loses ₹14 lakh crore a year — 4.3 % of GDP — to corrosion-induced damage, and notes that effective corrosion protection could lift GDP by up to 1.5 %. Corrosion is not a maintenance footnote. It is one of the largest line items in the national accounts, and it is almost entirely an engineering decision.

This guide is the whole discipline with the arithmetic attached: the electrochemistry that decides which metal dissolves, the eight damage mechanisms and where each one kills equipment, the alloy-selection law (PREN) that separates a 20-year seawater line from a 2-year one, coatings and cathodic protection designed with real numbers, the inspection mathematics that turns a UT reading into a remaining life, and what all of it costs in India in 2026. If you specify, fabricate, inspect, or pay for industrial equipment, this is the maths behind every material decision you make.

1. The Corrosion Cell: Four Requirements, One Equation

Corrosion that matters industrially is electrochemical: it moves electrons through metal and ions through water. Every wet-corrosion failure you will ever investigate needs four things simultaneously:

  1. an anode — where metal dissolves (oxidation): \mathrm{Fe \rightarrow Fe^{2+} + 2e^-}
  2. a cathode — where something accepts the electrons (reduction, most commonly oxygen: \mathrm{O_2 + 2H_2O + 4e^- \rightarrow 4OH^-}, or hydrogen evolution in acids)
  3. an electrolyte — an ionically conducting liquid film (water with any dissolved salt; even a condensation film on steel in humid air)
  4. a metallic path — direct contact between anode and cathode (the pipe wall itself is enough)

Remove any one and corrosion stops. That sentence is the entire engineering playbook in disguise: coatings and insulation kill the electrolyte; galvanic isolation and CP change the electrical conditions; inhibitors poison the electrode kinetics; alloy changes move the thermodynamics. Every mitigation you will ever specify attacks one of the four.

Standard potentials and the Nernst equation

Metals differ in how badly they want to dissolve, measured as electrode potential. The standard reduction potentials (vs the standard hydrogen electrode) set the hierarchy: \mathrm{Au^{3+}/Au} +1.50\ \mathrm{V}, \mathrm{Cu^{2+}/Cu} +0.34\ \mathrm{V}, \mathrm{H^+/H_2}\ \ 0.00, \mathrm{Fe^{2+}/Fe} -0.44\ \mathrm{V}, \mathrm{Zn^{2+}/Zn} -0.76\ \mathrm{V}, \mathrm{Mg^{2+}/Mg} -2.37\ \mathrm{V}. More negative = more willing to dissolve = "less noble." But real environments are not standard conditions, and the tool that corrects for that is the Nernst equation:

Worked example that shows why it matters: iron in a stagnant pocket where \mathrm{[Fe^{2+}] = 10^{-6}\ M} sits at

— nearly 200 mV more negative than the standard value. Low dissolved-oxygen, low-flow pockets (under deposits, under gaskets, in dead legs) both accelerate the anodic side and become the cathode's blind spot (oxygen can't get in, so the adjacent aerated area becomes a large cathode). This is why stagnant water under a deposit corrodes steel several times faster than flowing water — the electrochemical arithmetic in one line.

Pourbaix diagrams: the map

Plotting equilibrium potential against pH gives the Pourbaix diagram, which partitions the iron–water system into three territories:

The water stability window itself is drawn by two lines: the hydrogen line E = -0.0592\,\mathrm{pH} and the oxygen line E = 1.229 - 0.0592\,\mathrm{pH}. At pH 7 they sit at −0.41 V and +0.82 V; keep the metal's potential below the oxygen line and dissolved oxygen cannot even be reduced on its surface.

The chromium trick — the reason stainless exists. Above roughly 10.5–12 % chromium, a chromium-rich passive film (\mathrm{Cr_2O_3}, a few nanometres thick) forms and self-repairs in air. That single material insight is why "stainless" steel stains less. It is also fragile: chloride ions puncture it (Section 3.3), and it fails in any oxygen-starved crevice, because the film needs oxygen to rebuild (Section 3.4).

Kinetics: Tafel and mixed potential theory

Thermodynamics says whether a reaction can occur; kinetics say how fast. The rate of an electrode reaction follows, near its reversible potential, an exponential law — the Tafel equation:

where \eta is the overpotential that must be applied to push current i, and i_0 is the exchange current density at equilibrium (oxygen reduction on steel has i_0 in the \mathrm{10^{-8}}–\mathrm{10^{-11}\ A/cm^2} class — sluggish; steel dissolution is far easier, which is why steel supplies the electrons and oxygen consumes them).

In a real "mixed potential" system both reactions run on the same surface, and the potential settles where anodic current = cathodic current. That balance point is the corrosion potential — and the magnitude of the balancing current is the corrosion rate. Faraday's law (next section) converts that current into millimetres per year. That chain — potential → current → metal loss per year — is the spine of everything that follows.

2. Faraday's Law: The Only Three Formulas You Need for Rates

Michael Faraday did the book-keeping in 1834. His law says the mass of metal dissolved is proportional to charge passed: m = \dfrac{ItM}{nF}, with M atomic mass, n electrons per atom, F = 96{,}485\ \mathrm{C/mol}. Rearranged and unit-massaged, it gives the three formulas that every corrosion engineer actually carries:

1. From weight loss (coupons, removed specimens):

where W = mass loss in mg, \rho = density in g/cm³, A = exposed area in cm² (or in² for the mpy version), t = exposure time in hours. Both constants are just F, the calendar, and the unit conversions — use them, don't re-derive them.

2. From measured corrosion current density (probes, LPR, electrochemical work):

with EW = M/n the equivalent weight (27.92 for iron, 31.77 for copper, 32.69 for zinc). For steel this collapses to a number worth tattooing somewhere: 1 µA/cm² = 11.6 µm/y = 0.46 mpy. (Zinc: 1 µA/cm² = 15 µm/y.)

3. The charge-to-mass shortcut: 1\ \mathrm{Ah} of anodic current dissolves 1.04 g of iron — so one ampere-year (8,760 Ah) eats 9.1 kg of steel. A stray current of 10 A wandering out of an uninsulated welding transformer through a buried pipe for a year removes 91 kg of iron from wherever it exits. Stray-current failure investigations are usually won or lost on exactly this arithmetic.

Worked example (coupon). Two mild-steel coupons, total exposed area 100 cm², are hung in a cooling-water sump for 90 days (2,160 h). After cleaning per ASTM G1, weight loss = 4.2 g:

Interpretation: double that for the pitting factor on carbon steel and you have 0.4 mm/y local attack — a six-inch pipe (7.1 mm wall) retires in under two decades unless the water is treated.

Worked example (current). A tank floor of 1 m² carries a uniform corrosion current of 10 µA/cm² (a modest number — no acid, no chlorides, just aerated water):

91 kg of steel dissolved by a current smaller than a phone charger's. Corrosion is a current problem dressed up as a materials problem.

Typical rates, by environment (design-table numbers)

Environment (carbon steel unless noted) · Typical rate · Mechanism

Atmosphere C2 (clean inland), ISO 9223 · 1.3–25 µm in year 1 · uniform

Atmosphere C3 (urban/industrial-typical India) · 25–50 µm in year 1 · uniform

Atmosphere C4–C5 (coastal India, C5 up to 200 µm) · 50–200 µm in year 1 · uniform + pitting

Aerated seawater, ambient, unaided · 0.1–0.3 mm/y · uniform + pitting

Quiescent polluted water / under deposits · 0.3–1 mm/y · pitting, MIC

Soil, buried, unaided · 20–100 µm/y average; pitting far worse · pitting, MIC

CUI (under wet insulation, 60–120 °C) · up to ~1 mm/y · pitting/grooving

Flow-accelerated corrosion (100–220 °C water) · 0.1–>1 mm/y local · oxide dissolution

ISO 9223's first-year carbon-steel bands (µm/y): C1 ≤ 1.3, C2 1.3–25, C3 25–50, C4 50–80, C5 80–200, CX 200–700. Design rule of thumb used across process industry: average rate ≤ 0.1 mm/y is "manageable" with allowance; 0.1–0.5 mm/y needs active mitigation or alloy; > 0.5 mm/y is a materials change, not a paint job.

3. The Eight Damage Mechanisms — and Where Each One Kills You

Franklin Fontana's "eight forms" taxonomy survives because failure investigation still runs on pattern recognition. Here they are with the numbers that decide each one.

3.1 Uniform (general) corrosion

The honest one: metal thins more or less evenly. Design response is arithmetic — a corrosion allowance. Typical allowances: CS pipe 1.5–3 mm; pressure vessels 3 mm (plus a minimum for handling); storage tanks per API 650/653 practice. At 0.1 mm/y, a 3 mm allowance buys 30 years of operation; every doubling of the corrosion rate halves the life in a straight line. Uniform corrosion is the only mechanism you want, because it is predictable and inspectable by a single UT grid.

3.2 Galvanic corrosion: the series decides

Two different metals in one electrolyte form a battery — and the less noble one pays. The galvanic series (flowing seawater, potentials vs saturated calomel electrode, per ASTM G82 and ASM Handbook Vol. 13C) ranks the engineering metals:

Material · Potential (V vs SCE) · Role in a couple

Magnesium · −1.60 · extreme anode

Zinc / galvanised steel · −1.05 · sacrifical anode

Aluminium alloys · −0.85 to −0.70 · anode

Carbon steel / cast iron · −0.65 to −0.60 · anode vs Cu/SS

410 SS (active) · −0.55 to −0.45 · anode when active

316 SS (active), 304 active · −0.45 to −0.35 · active-state only

Naval brass / copper alloys · −0.30 to −0.15 · cathode vs steel

316 SS (passive) · −0.05 to +0.05 · cathode

Titanium Gr.2 · −0.05 to +0.10 · cathode (safe)

Graphite · +0.25 to +0.30 · worst cathode in industry

Two rules govern practice. Rule 1 — voltage gap: keep the couple difference under ~200–250 mV where possible; above 300 mV, assume attack unless designed otherwise. Rule 2 — area ratio: the corrosion current is set by the cathode, and it is discharged over the anode's area, so

A large cathode feeding a small anode is disastrous. Worked case: a stainless steel or CuNi accessory plate (500 cm²) sustaining a modest 5 µA/cm² of oxygen reduction in aerated seawater drives a carbon-steel fastener of 5 cm²:

That is why a steel bolt in a copper-alloy or large stainless component at sea perforates in a season or two, while the reverse (a small stainless bolt in a large steel structure) is comparatively benign — the anode is big, so the current density is diluted. Fastener-materials practice follows this single sentence more than any other.

Practical corollaries from the plant floor: never install copper or brass upstream of galvanised pipe (copper ions redeposit downstream and set up pitting cells; and galvanised–copper couples sit ~700 mV apart on the series); use dielectric isolation kits (isolating gasket + sleeves + washers, continuity-tested after assembly) at every buried steel-to-stainless or steel-to-copper flange; in marine atmospheres never couple aluminium to steel or stainless (the aluminium becomes the anode and its passive film is chloride-fragile); and remember stainless behaves as two different metals depending on whether it is passive (noble) or active (barely better than steel) — a distinction that turns on oxygen availability and is the hidden variable in half the "unexplainable" marine failures.

3.3 Pitting: chlorides versus the passive film

Pitting is the signature failure of stainless steels and aluminium: a chloride ion punctures the passive film at a weak point, the pit becomes its own tiny anode with a large passive cathode around it, and the local acid chemistry (metal chlorides hydrolyse; pit pH can fall to 1–3 while the bulk stays neutral) keeps it eating inwards. Wall loss at the surface looks like nothing; the pit goes through the wall.

Quantify the alloy's resistance with the Pitting Resistance Equivalent Number:

Grade · UNS · Typical PREN · CPT in 6 % FeCl₃ (ASTM G48-class) · Seawater verdict

304L · S30403 · ≈ 19 · ~10–15 °C · no

316L · S31603 · ≈ 25 · ~15–25 °C · not standalone

904L · N08904 · ≈ 35 · ~40 °C · with care

Duplex 2205 · S32205 · ≈ 35 · ~35–50 °C · with CP limits (crevices)

254SMO (6Mo) · S31254 · ≈ 43 · ~60–80 °C · yes

Super duplex 2507 · S32750 · ≈ 42 · ~60–80 °C · yes (≤ ~50–60 °C, chlorination-limited)

Alloy 625 / C-276 · N06625/N10276 · ≈ 51 / 68 · > 85 °C · yes, hot everything

The decisive number on the world's coastlines: the critical pitting temperature (the minimum temperature for attack, ASTM G48, 6 % FeCl₃). The industry acceptance law for seawater — NORSOK M-001 and ISO 21457 — is blunt: PREN ≥ 40 is treated as "seawater resistant"; below ~38 a stainless cannot be used in aerated seawater without cathodic protection. That is why the "marine grade" myth fails in practice: 316L (PREN ≈ 25) is routinely installed on seawater and routinely fails by pitting, especially chlorinated, warm, or stagnant. Temperature and chlorination shift everything: NORSOK caps super duplex and 6Mo with tight crevices (screw couplings) at about 20 °C when seawater is chlorinated to 1.5 mg/L; chlorination measurably cuts the CPT. Seawater itself is 19,000–24,000 mg/kg of chloride — no chlorine-free option exists, only better alloy, lower temperature, fewer crevices, or cathodic protection.

3.4 Crevice corrosion: the geometry is the killer

Under a gasket, a bolted washer, a weld spatter bead, a scaling deposit, or between a pipe and its support, a stagnant pocket runs out of oxygen first. The pocket becomes the anode (it is the oxygen-starved zone), the open surface around it becomes the cathode, and inside the gap the same self-acidifying chemistry as a pit develops. The pattern of failures is almost embarrassingly geometric: flanged joints, under-deposit zones, threaded connections, and tank-bottom sludge piles.

Design answers, in order of effectiveness: eliminate crevices entirely (full-penetration welds; crevice-free valve seats; no lap joints in immersion); if you can't, select for it (crevice temperature limits are 10–20 °C below pitting limits — always design to CCT, not CPT); gasket the joint properly (spiral-wound with PTFE or graphite per service, correct seating stress); and never let 316L do one of these jobs in seawater. Complete drainage, no dead legs (rule: no dead leg longer than ~2–3× pipe diameter), and sludge-free tank bottoms are corrosion engineering, not housekeeping.

3.5 Intergranular corrosion and sensitization

Heat 304/316 between roughly 450 and 850 °C — precisely the band the heat-affected zone experiences during welding — and carbon diffuses to the grain boundaries, combining with chromium into \mathrm{M_{23}C_6} carbides. The chromium is pulled out of the adjacent zone, which drops below the ~12 % passivity threshold. The result is a microscopic sheet of chromium-depleted, non-passive alloy along every grain boundary. In a corrosive service, the material dissolves like a bag of sand along its own grain structure — weld decay, with the classic band parallel to the weld bead 3–15 mm into the HAZ.

Control is metallurgical and well codified:

The same logic bites in shut-down corrosion: sensitised stainless in refinery service, contacted by moist air and sulfur compounds during a shutdown, cracks by polythionic acid SCC — which is why shutdown procedures specify soda-ash neutralisation and dry-air purges. Corrosion engineering is procedural too.

3.6 Selective leaching (dealloying)

One alloying element dissolves preferentially, leaving a porous skeleton of the rest — dezincification of brass (\mathrm{Cu-Zn} with > 15 % Zn: zinc leaves, copper-red porous metal remains; the part looks fine and holds nothing), graphitic corrosion of grey cast iron (iron leaves, graphite network remains, pipe looks like pipe until it fails in your hands), and aluminum loss from aluminium bronzes. Fixes are metallurgical: DZR (arsenical) brass, naval brass, bronze or CuNi instead of plain brass for wet valves; ductile iron or coated CI instead of grey iron for buried mains; and awareness that dezincification accelerates in stagnant, slightly acidic, high-chloride water — another reason dead legs and stagnant valves fail first.

3.7 Erosion-corrosion and flow-accelerated corrosion (FAC)

When flowing fluid strips the protective film or oxide layer faster than it can rebuild, the metal beneath corrodes at an accelerating rate. Two species matter most:

Velocity-limited erosion-corrosion (film stripping) has hard industrial ceilings per material in seawater:

Material · Practical velocity limit (seawater)

Copper (plain) · ~1–1.5 m/s

Aluminium brass · 0.8–2.5 m/s (larger bores; less on small tubes)

90/10 CuNi · 3.2–3.5 m/s max; keep design above ~0.8 m/s to avoid fouling/stagnation

70/30 CuNi · up to ~4.5 m/s

Carbon steel · ~1.5–2.5 m/s in water with solids; watch elbows at higher

Titanium, super duplex · high (limited by other mechanisms, not erosion)

Also velocity-limited is cavitation (vapour bubbles collapsing on pump impellers and valve bodies — attack scales roughly as velocity⁵–velocity⁶, so trim impellers rather than throttle 20 % open) and droplet-impingement in wet-steam piping.

FAC is the subtler species, and it has killed more power-plant and industrial boiler feedwater personnel than any single corrosion mode: deoxygenated, alkaline water dissolves the protective magnetite (\mathrm{Fe_3O_4}) layer on carbon steel, the flowing water sweeps the dissolved iron away, and the oxide thins to near-zero. Key numbers from the EPRI/industry corpus: FAC is active in the ~100–280 °C window with the rate peak near 150 °C in single-phase flow; it is worst at pH ≈ 8.5 and collapses above pH ~9.2 for single-phase flow; oxygenated treatment (30–150 ppb DO) essentially eliminates it by building a dense \mathrm{FeOOH} overlayer, which is why modern all-ferrous cycles specify OT/AVT(O) with feedwater pH 9.2–9.6. When you cannot change chemistry, change material — 1 % chromium low-alloy steel (e.g., 1.25Cr) reduces FAC rates by roughly 75 % versus carbon steel, and that single percentage point of Cr is why power-plant feedwater piping is often specified in alloy. Turbulence concentrates the damage: elbows, tees, reducers, and anything downstream of a control valve or orifice runs 3–5× the straight-pipe rate, which is exactly where a FAC program puts its CMLs (Section 7).

3.8 Environment-assisted cracking: SCC, SSC, caustic, ammonia

Not thinning — cracking. All env-cracking needs three things at once: a susceptible material, a specific environment, and tensile stress (residual stress from welding is usually the culprit that tips a part over). It fails suddenly at nominal pressures, which is why it owns so many of industry's worst accidents.

Chloride stress corrosion cracking (Cl-SCC) of austenitic stainless steels. 304 and 316 under tensile stress crack in chloride-containing water above roughly 50–60 °C — and the attack is branched and fast, often through-wall with no measurable wall loss. The classic industrial scene is a hot stainless line: oxidised by insulation or process, chloride-rich from insulation leachate or coolant, warm, and stressed by a weld. Mitigations: keep stainless cool and dry (or insulated per ASTM C795 low-chloride specs with good water shedding); move to duplex (much more Cl-SCC resistant) or high-Ni alloys; thermal-spray aluminium under insulation; strictly control how the stainless welds are stress-relieved (post-weld anneal — not a mild stress relief, which does nothing).

Sulfide stress cracking (SSC) and the sour-service law. In wet \mathrm{H_2S}, atomic hydrogen enters high-hardness steel and cracks it. NACE MR0175/ISO 15156 — the standard every oil-and-gas material specification cites — sets the entry threshold and the controls:

The rest of the cracking family the working engineer should recognise on sight: caustic SCC (concentrated NaOH above ~50 %, hot, stressed — caused famous tank-wagon catastrophes; fix = full stress relief of welds + steam/nitrogen purge procedures); ammonia SCC of copper alloys in condenser service (hence CuNi or naval brass, and vent the ammonia); HIC/SOHIC (hydrogen blisters and stepwise cracking in wet-H₂S carbon steels — killed by clean steel: low-S, HIC-tested plates per NACE TM0284); liquid-metal and carbonate cracking (stay out of the carbon-steel H₂S-free "caustic concentrations × temperature" boxes — these are in API 571, the damage-mechanism encyclopedia worth its price for any plant engineer); and hydrogen embrittlement of high-strength fasteners — which is why 12.9-grade bolts are banned in hydrogen and cathodically protected service while the whole industry drills hydrogen embrittlement relief bake into plating specs.

3.9 Corrosion under insulation (CUI): hidden, expensive, predictable

The most expensive square metre of corrosion in any plant is the one nobody can see: the underside of a pipe under its rock wool. Water enters at the jacketing seams and penetrates at every penetration, breather, and damage point; the insulation holds it against the hot steel; and the mixture of heat, water, oxygen, and concentrated chloride leachate drives localised external corrosion at rates up to ~1 mm/y — worst between 60 and 120 °C, within the overall susceptible band of roughly −12 to 175 °C (API RP 583, with NACE SP0198 covering the coating/insulation systems side and ASTM C795 covering chloride limits for stainless insulation). Two mechanisms hide there:

Detection is its own discipline — insulation is not removed everywhere, so the industry screens: profile/real-time radiography (wall thickness through the jacketing), pulsed eddy current (averaged wall loss on ferrous piping without stripping), guided-wave UT (long-range screening from a single access point), infrared and neutron backscatter (wet insulation detection). Then it strips and inspects "at the money": the CUI-risk-ranked locations, per API RP 583's likelihood×consequence ranking. Expect a well-run CUI program to be 80 % ranking and 20 % inspection — and to find its worst corrosion at the bottom of vertical runs and under removable blankets, where the databases always say "not inspected."

3.10 MIC: when the bugs do the chemistry

Microbiologically influenced corrosion is not a different electrochemistry; it is bacteria (and archaea) changing the local chemistry to set up aggressive cells. The headline actor is sulfate-reducing bacteria (SRB), which reduce sulfate to sulfide in low-oxygen niches: the sulfide reacts to black iron sulfide, depolarises the cathode, and can accelerate steel corrosion to mm/y localised rates at pit sites. MIC favours: stagnant water (firewater mains with dead sections, hydrotest water left in lines, cooling systems under low flow), hydrocarbon/water interfaces, and any regime where oxygen is low and organics are present. 316L is as vulnerable as carbon steel — MIC doesn't care about your PREN — and its damage looks like random pitting with black biofilm in the pit and a rotten-egg smell when opened. Program against it, because you cannot select against it: continuous or shock biocides (chlorination, glutaraldehyde-type programs), pigging/flow management so nothing sits stagnant for months, clean dry commissioning water (treat that first fill as a corrosion job, because it will become one), coupon + microbial monitoring per NACE TM0212, and — the cheapest fix of all — design out the dead legs so the system cannot hold a stagnant pocket.

4. Materials Selection: The Grade Ladder (PREN Is Your First Pass, Never Your Last)

The grade ladder for aqueous and chloride service, with Indian per-kg market context from Section 8 overlaid:

Grade · PREN · SMYS (MPa) · India ₹/kg band (Sep 2026) · Use it when

Carbon steel + protection stack · — · 240 · 45–65 · the default; geometry allows coating + CP; big weight

304L · ~19 · 205 · 185–260 · mildly corrosive, non-chloride, < 60 °C, L for welds

316L · ~25 · 205 · 295–390 · chemicals, moderate chlorides non-aerated; not hot seawater

Duplex 2205 · ~35 · 450 · 450–520 · chloride service, moderate temp, high strength; CP in seawater

Super duplex 2507 · ~42 · 550 · ~550–750 · seawater to ~50–60 °C, chlorinated water, high strength

904L / 254SMO (6Mo) · ~35–43 · 220 · 500–800 · acid + chloride; where duplex welding is not available

Alloy 625 / C-276 · ~51 / ~68 · 415+ · premium (nickel-driven) · sour + hot + chloride; last ladder rung before titanium

Four selection laws to carry:

  1. PREN ≥ 40 for seawater (NORSOK M-001/ISO 21457 convention); PREN < 38 requires CP in aerated seawater. Cavities, crevices, and chlorination each knock the effective threshold up — design to CCT in any bolted or gasketed joint, and remember the NORSOK ~20 °C crevice cap on chlorinated seawater.
  2. Temperature is the multiplier. Chlorides + oxygen + each ten degrees make the attack statistically worse; the Cl-SCC anxiety threshold (≈ 60 °C for austenitics) and the CUI band (60–120 °C) are the same physics in different clothing.
  3. Strength pays for alloy. Duplex's 450–550 MPa SMYS against 240 for carbon steel means roughly half the wall for the same pressure duty. At 20 bar in DN150: CS Schedule 40 wall (7.1 mm) versus a duplex wall that calculates to ~0.6 mm calc + minimum — duplex buys its extra ₹/kg back in wall thickness, weight, supports, and welding hours. On long seawater lines that arithmetic is frequently decisive.
  4. Weldability and availability trump the datasheet. Duplex demands qualified procedures (heat input window, nitrogen-bearing shielding, ferrite count 35–65 % in weld and HAZ per ASTM E562, corrosion-tested per ASTM A923). If your fabricator cannot get duplex WPS/PQR in place, 6Mo or a protected carbon steel is a more honest choice than an unqualified duplex weld.

Protection beats alloy whenever geometry allows. The classic economic result: for a long, simple, inspectable line, carbon steel + coating + CP is 3–10× cheaper over 25 years than an CRA; for a short, complex, uninspectable, immovable component (seawater pump internals, small-bore instrument fittings, heat-exchanger tubing, buried valves), the CRA is cheaper because the protection and inspection are what cost money there, not the metal. Choose by life-cycle cost of the system, never of the kilogram.

5. Design and Fabrication Rules That Decide the Life

Most corrosion failures in India are designed in at the drawing stage. The checklist that matters:

Drainage and geometry. Every line drains at a low point; every tank bottom slopes to a sump; no water sits anywhere. No dead legs over 2–3 pipe diameters. Vents and drains on jacket/insulation systems. Slope steelwork so water leaves; avoid "cups" between structural members — the standard outdoor detail of a horizontal member pair 100 mm apart is a corrosion cell that holds water for six months a year.

Dissimilar-metal discipline. Keep couples under ~200–250 mV; fix the area ratio (big anode, small cathode); use isolation kits at mixed flanges (gasket + sleeves + washers, verified with a continuity meter after bolt-up); never aluminium-to-steel in marine; never copper upstream of galvanised; insulate CP-protected structures from new "grounds" (a new earthing connection can eat the CP budget).

Welding. For stainless: control heat input and interpass temperature; brush and pickle/passivate every weld per ASTM A380/A967 — the rainbow heat tint left on a 316 weld is a chromium-depleted starter coupon for pitting, and free iron from carbon-steel tooling (wire brushes, grinding wheels used on other jobs) rusts out of the stainless surface within weeks. For duplex: qualified WPS, nitrogen in the shielding, ferrite checks, A923 corrosion acceptance. For sour CS service: hardness ≤ 22 HRC, PWHT ≥ 621 °C where required, hardness surveys on each weld. And do not casually stress-relieve austenitic stainless — you will sensitise it faster than you fix it; use L grades or stabilised grades and let them be.

Insulation and CUI details. Closed-cell or water-shedding insulation in the CUI band; low-chloride insulation (ASTM C795) for stainless; metallic jacketing properly lapped and sealed, with no unsealed penetrations; coatings under insulation specified per NACE SP0198 — and design so that any water that does get in drains out rather than pooling at the bottom of the run. For hot stainless lines, specify whether the answer is coating (TSA) or material (duplex) — not both accidentally neither.

Corrosion allowance and inspection access. Put the allowance where the rate is (an elbow allowance is not a straight-pipe allowance). Provide access for UT at every CML — 200 mm clear straights at each end of every run, test points reachable by scaffold-free means. An uninspectable design is an unmanaged design; RBI (Section 7) is only as good as the accessibility you drew in.

6. The Protection Stack: Coatings, Galvanizing, and Cathodic Protection

Coatings: the barrier, in microns

A protective coating is the cheapest way to kill the electrolyte of the corrosion cell, and its life is decided in the first millimetre of process — surface preparation. The specification shorthand: blast to ISO 8501-1 Sa 2½ (near-white, anchor profile 40–75 µm), apply the specified system, verify DFT and holiday-test buried pipe (ASTM G62). Systems by ISO 12944 corrosivity category:

System (typical) · Where · Thickness

Alkyd / 2-coat paint · C2 indoor · 100–150 µm

Epoxy primer + epoxy intermediate + PU topcoat · C3–C4 industrial · 240–320 µm

Zinc-rich primer + epoxy + PU · C5 coastal India · 320–500 µm

FBE (fusion-bonded epoxy) · buried pipe · 350–500 µm

3-layer PE (epoxy + adhesive + HDPE) · buried high-spec pipe · 2.5–4 mm total

Thermal-spray aluminium (TSA) · CUI service, stainless, marine · 200–300 µm

Every one of those systems is competee against replacement on a life-cycle basis, not first cost — a 320 µm C5 system that lasts 25 years beats three rounds of repainting at 8-year intervals by a factor of 2–3 once scaffolding is counted.

Galvanizing: the sacrificial shortcut

Hot-dip galvanizing doesn't just barrier — the zinc is sacrificial, and it protects steel at scratches and cut edges. Coating per ISO 1461/IS 4759: typically 70–85 µm (roughly 610–865 g/m²) on structural steel. Life = zinc thickness ÷ zinc loss rate (ISO 9223 zinc bands: C3: 0.7–2.1 µm/y; C4: 2.1–4.2; C5: 4.2–8.4), so an 85 µm coating computes to roughly 20–50 years in C3, 10–20 in C4, 5–10 in C5 — with the caveat that zinc rates decline as the patina forms, and that chlorides in coastal India push real sites toward the C5 end. In India, job-work galvanizing runs ₹19–24/kg (₹/kg of steel, per TradeIndia market listings) — against a painted re-coating cycle at ₹150–300/m² of surface, which is why virtually every outdoor structural job in this country is galvanized rather than painted. The duplex system (galvanize + paint) is the C5/CX answer, and the one detail to never miss: cut ends and drilled holes of galvanized steel are protected by the zinc itself; cut ends of galvanized then painted are not — field touch-up rules exist for a reason.

Cathodic protection: making the structure the cathode

CP stops corrosion thermodynamically: push the steel below its dissolution potential (Section 1) and it simply stops dissolving. The criteria every technician should know (NACE SP0169 family):

Design, worked end to end. The current demand comes from surface area × current density × coating condition. DNV-RP-B401 defaults: bare steel buried in sediments 0.020 A/m²; bare steel in seawater roughly an order higher; aluminium components 0.010 A/m²; heated surfaces add 0.001 A/m² per °C above 25 °C; and the mean (maintenance) design current is about 50 % of the initial/final value, because the calcareous scale and fouling layer that forms once you polarise the steel cuts the demand.

Take a tank farm with 2,000 m² of wetted steel in soil:

That pair of numbers is the entire economics of CP: a good coating reduces CP current demand 100×, and the coating is what you are actually buying when you "design cathodic protection." Nobody clever solves bare-steel CP on big areas; the coating breakdown factor is the design's main lever.

Sizing is circuit arithmetic. Anode output follows the resistance: for a slender stand-off anode, Dwight's equation gives

For \rho = 0.25\ \Omega\cdot\mathrm{m} (warm seawater), L = 1.0\ \mathrm{m}, d = 0.15\ \mathrm{m}: R_a \approx 0.118\ \Omega, and with a driving voltage of 0.25 V (anode −1.05 V open, design protection at −0.80 V) each anode delivers ≈ 2.1 A initially, decaying as it consumes. Current demand sets the initial anode count; charge consumption sets the mass, \text{mass} = I_{\text{mean}} \times \text{life} \times 8{,}760 / (\text{capacity} \times u) — zinc anodes (≈ 780 Ah/kg) need ~2.6× the mass of aluminium for the same charge, which is why Zn survives only in low-temperature/polluted niches (pipelines in anaerobic soil, ballast tanks) and Al–Zn–In rules everywhere else. After installation: commission the CP — baseline structure-to-electrolyte survey, then annual ON/OFF+DC current trend. A CP system that was never surveyed is indistinguishable from no CP, except in the accounts.

7. Inspection, Monitoring and RBI: From a UT Reading to a Decision

Corrosion management closes the loop only when measurements turn into intervals. The toolkit, by mechanism:

Method · Measures · Best for

Weight-loss coupons (ASTM G1/G4) · average rate over months · baseline corrosivity, inhibitor program verification

ER probes · metal loss (electrical resistance), continuous · trending, online systems

LPR probes · instantaneous rate via polarisation resistance · cooling water, inhibitor response (needs conductive water)

UT thickness grid (CMLs) · wall thickness · the workhorse: pipe, vessel, tank

Guided-wave / PEC / profile RT · screening without access · CUI, buried, hard-to-reach

Half-cell potential (ASTM C876) · rebar corrosion probability in concrete · infrastructure, RC structures

Holiday testing (ASTM G62), DFT · coating integrity · new buried pipe acceptance

LPR gives instantaneous rate from the Stern–Geary relation i_{\text{corr}} = B/R_p with B \approx 26\ \mathrm{mV} for steel: a probe reading R_p = 5{,}000\ \Omega\cdot\mathrm{cm^2} corresponds to i \approx 5\ \mu A/\mathrm{cm^2} \approx 0.06\ \mathrm{mm/y} (healthy-ish); R_p collapsing to 1{,}000\ \Omega\cdot\mathrm{cm^2} means 0.3 mm/y — alarm territory — usually traceable to an inhibitor feed failure within the same day.

The four numbers that turn survey data into a plan (API 570/510/653 practice):

Worked case — a tank bottom plate: original 6.0 mm, UT now 4.6 mm, 8 years in service, required retirement thickness 3.0 mm.

Cross-check the long-term rate: if the previous UT was 5.1 mm two years ago, the "short-term rate" (5.1-4.6)/2 = 0.25 mm/y is faster than the long-term 0.175 — the corrosion is accelerating (chemistry changed, coating failed, MIC started), and the conservative answer is to inspect on the short-term number. Long-term rate for inventory, short-term rate for urgency — that is the whole philosophy of API 570 in one line.

Above this sits risk-based inspection (API 580/581): rank every circuit by Probability × Consequence, where probability folds in damage mechanism, rate, remaining life, and inspection effectiveness, and consequence folds in safety, environment, and production loss. RBI is how a plant gets from "inspect everything every four years" to "spend the 20 % of inspection budget that buys 80 % of the risk reduction" — and it is, in the end, exactly FabFlow's kind of arithmetic: schedule, consequence, and money per millimetre of metal.

8. What Corrosion and Its Prevention Cost in India

The macro number (Section 0, restated because it belongs here): Nomura's July 2026 estimate — ₹14 lakh crore per year, 4.3 % of GDP, with up to 1.5 % of GDP retrievable through effective corrosion protection; the IMPACT global study's avoidable fraction — 15–35 % — is the same claim in different units. This is not a rounding error; it is the budget of the Ministry of Defence.

The material ladder in rupees (indicative stockist/mill rates, Sep 2026). Structural/MS steel ₹45–65/kg; SS 304 ₹185–260/kg (Jindal mill list of June 2026: ₹218–258/kg ex-Mumbai, 304L +₹2); SS 316L ₹295–390/kg; duplex 2205 ₹450–520/kg; 904L ₹500–800/kg; super duplex 2507 above 2205 (grade- and form-dependent); nickel metal ₹1,500–2,000/kg — and nickel is what drives every stainless surcharge. Hot-dip galvanizing job work: ₹19–24/kg delivered-to-bath. Every number above is indicative of published Indian market listings, not a quotation; alloy surcharges move monthly.

The two comparisons that decide real projects:

(1) Corrosion allowance versus alloy upgrade. DN150 pipe, 20-bar service, 20-year life. Extra 2 mm of carbon-steel wall: weight goes 28.26 → 35.76 kg/m (+7.5 kg/m) ≈ +₹450/m at ₹60/kg steel. Upgrading the same wall to 316L instead: +₹7,600/m at the prices above — 17× the cost of the allowance. This is why the first ₹ of corrosion budget goes to steel + coating + CP + inspection, and alloy is reserved for the components where protection genuinely cannot work.

(2) 2,000 m² of buried steel, bare versus coated (from Section 6). Bare: ~2.2 tonnes of aluminium anode metal over a 10-year design. Coated (fc = 0.01): ~22 kg. The coating costs on the order of ₹150–300/m² at C5 spec — about ₹3–6 lakh for the area — and saves roughly two tonnes of anode metal plus its replacement labor. Coatings are the highest-return rupees in the entire corrosion budget, which is why the cheapest quote on a buried pipe is usually the most expensive by year ten.

And the failure ledger nobody itemises until it happens: a single pinhole leak on a process line is a shutdown, an investigation, five welders, lost production, and (for the right services) an environmental report. The rule of thumb used by plant insurers — failure costs 10–100× the inspection that would have found it — is the strongest argument for the section that follows this one. Corrosion engineering is the discipline of paying ₹1 now instead of ₹100 later, with the discount rate of a fire.

The Numbers to Remember

Frequently Asked Questions

What is the single most common corrosion failure in industrial plants? By count, localised attack — pitting and crevice corrosion — not the uniform wall loss that allowances are sized for. Uniform corrosion is the housekeeping you design for; pitting, crevice, SCC and CUI are the mechanisms that actually fail equipment, and all four concentrate at hidden, stagnant, or geometrically awkward spots: under deposits, at flanged joints, under insulation, and in dead legs.

How do I calculate a corrosion rate from two UT readings? CR = (t_1 − t_2)/\Delta t. Six mm down to 4.6 mm in 8 years = 0.175 mm/y. Then remaining life = (current − retirement)/CR = (4.6 − 3.0)/0.175 = 9.1 years, and the next inspection is due by half of that (≈ 4.5 years) — with the code caps (10 years for piping per API 570, etc.) applied on top. If the short-term rate exceeds the long-term rate, the corrosion is accelerating; inspect on the short-term number.

Is 316L good enough for seawater? No — not on its own. At PREN ≈ 25, 316L sits below the industry's PREN ≈ 38–40 seawater line and fails by pitting and crevice corrosion, especially warm, chlorinated, or stagnant. Seawater-grade answers are super duplex 2507/6Mo (PREN ≥ 40) with temperature and chlorination limits respected, or carbon steel/CuNi with cathodic protection. "Marine grade 316" is a stockist slogan, not an engineering verdict.

What do −850 mV and 100 mV mean in cathodic protection? They are the NACE SP0169 protection criteria: either the structure's polarized potential is at least −850 mV vs a copper/copper-sulfate reference (measured IR-free, i.e., with the CP switched off momentarily), or the CP has produced ≥ 100 mV of potential decay from native. Meet either and steel is protected; measure with an interrupter and correct for IR drop, or you are measuring the voltage drop in the soil, not the protection.

How long does hot-dip galvanizing actually last? Take the coating thickness (70–85 µm typical per ISO 1461) and divide by the local zinc consumption rate from ISO 9223 — roughly 0.7–2.1 µm/y in C3, 2.1–4.2 in C4, 4.2–8.4 in C5. That computes to 20–50 years in C3, 10–20 in C4, 5–10 in C5. Coastal India and chloride-heavy industrial zones land toward the low end; a duplex galvanize+paint system stretches the C5 case further.

Why did our stainless steel crack without any wall loss? Almost certainly environment-assisted cracking — chloride SCC for austenitic grades above ~60 °C or external chloride SCC under insulation (50–60 °C+), or one of the sour-service mechanisms if H₂S was present. Look at the environment's chloride and temperature history, the tensile stress path (weld residual stress counts), and the insulation leachate chemistry — then fix the environment/coat/insulate, because once the environment qualifies, all stressed austenitic surfaces of that system are candidates, not just the cracked one.

How much does corrosion cost India? The most current number: ₹14 lakh crore/year, about 4.3 % of GDP (Nomura, July 2026), against the global benchmark of 3.4 % of GDP from the NACE IMPACT study. The avoidable share — typically 15–35 % — is the annual prize for getting coating selection, CP, alloy choice, and RBI right. It is larger than most companies' annual capital budgets, per country, per year.

Carbon steel with coating + CP, or an upgrade to duplex — which do I choose? Work the geometry: long, large-diameter, inspectable, static equipment favours protected carbon steel (the economics usually win 3–10×); compact, complex, uninspectable, safety-critical components in chlorides and moderate heat favour CRA — and duplex's 450–550 MPa yield lets you halve the wall, which recovers much of its price premium. Then subtract the cost of having no maintenance strategy at all: on critical seawater service, a protected carbon-steel system that is maintained beats an unmaintained alloy that nobody can inspect.

Where FabFlow Fits

Every corrosion decision in this guide terminates in a fabrication order. The duplex spool that replaces a failed 316L seawater line, the new tank bottom in 6 mm CS with the sump and drain that the old one lacked, the hot-dip galvanized structural package (₹19–24/kg and every shop in the country quotes it), the TSA-coated insulated spools for CUI service, the ASTM A923-tested duplex flange set, the isolation kits for mixed-metal flanges, the replacement fasteners in the right galvanic class so the fix doesn't become the next failure — these are drawing-in, quote-out jobs, and they are exactly what FabFlow exists to price.

If you are an owner-engineer planning the fix, post the scope with the service conditions — fluid, chloride level, temperature, pressure, current condition — and the quotes that come back can be compared on the thing that matters (material + coating + weld quality against life-cycle cost), not on the thing that doesn't (the first invoice). If you are a fabricator, the same discipline runs in reverse: the shop that can document its 2205 welding procedure, offer L-grade and stabilised options, hold galvanizing certifications, and pass a third-party A923/PMI check is the shop that stops competing on price. Corrosion is the slowest adversary in manufacturing; the way to beat it is to be the supplier who could read this arithmetic — and quoted accordingly.

A hundred rupees spent at the drawing board is worth ₹10,000 spent at the failure investigation; the whole discipline of corrosion engineering exists to make that exchange rate visible. Start with your own worst line: find its current rate, halve its remaining life, and diarise the date. The calendar, not the chemistry, is what failed the plants that failed.

More FabFlow blog posts