Corrosion and materials of construction

In short

Corrosion is the electrochemical (or chemical) destruction of a material by its environment, and materials of construction (MOC) is the engineering decision of which material will survive a given process fluid for the design life at acceptable cost. In a chemical plant it drives wall thickness, inspection intervals, leak and fire risk, and a large share of capital cost — a 316L reactor can cost five to ten times its carbon-steel equivalent, while the wrong cheap choice causes a shutdown. It is examined in GATE (CH) under Process Equipment Design / Plant Design, and in university papers on Corrosion Engineering and Materials Technology. Employers who test it hardest are EIL, L&T Hydrocarbon, Toyo, Technip, Reliance, IOCL/BPCL/HPCL refineries, UPL, Deepak Nitrite, Aarti Industries, Thermax and Tata Chemicals, for roles in process design, static equipment, inspection and reliability.

Explained

Electrochemical basis

• Wet corrosion is a short-circuited galvanic cell on one metal surface: anode (oxidation, metal dissolves) Fe → Fe²⁺ + 2e⁻; cathode (reduction) consumes the electrons.
• Cathodic reactions: oxygen reduction O₂ + 2H₂O + 4e⁻ → 4OH⁻ (neutral aerated water, cooling water); hydrogen evolution 2H⁺ + 2e⁻ → H₂ (acids); metal-ion or ion-reduction reactions in process liquors.
• Four essentials: anode, cathode, metallic path, continuous electrolyte. Remove any one and corrosion stops — the logic behind coatings (electrolyte), insulating gaskets (metallic path) and cathodic protection (polarity).
• Galvanic series (seawater, active → noble): Mg, Zn, Al, mild steel, cast iron, lead, tin, brasses, copper, Monel, passive 304/316, titanium, graphite. The more active metal corrodes; the area ratio matters — a small anode with a large cathode corrodes catastrophically fast.
• Electrode potential is given by the Nernst equation; Pourbaix (E–pH) diagrams map zones of immunity, corrosion and passivity for a metal in water, and are read, not derived, in exams.
• Polarisation: at the corrosion potential E_corr the anodic and cathodic currents are equal, giving i_corr. On a Tafel plot (E vs log i) the two straight branches extrapolated back to E_corr give i_corr, which converts to a penetration rate by Faraday's law. Linear polarisation resistance (Stern–Geary) is the on-line version used by plant corrosion probes.
• Passivity: Cr, Al, Ti, Ni and stainless steels form a thin, adherent, self-healing oxide that cuts the rate by orders of magnitude. Oxidising conditions maintain it; chlorides, reducing acids and oxygen starvation break it locally.

The eight forms of corrosion

1. Uniform (general) attack — even thinning, e.g. steel in dilute acid or atmosphere. Least dangerous because it is predictable; handled by a corrosion allowance.
2. Galvanic — dissimilar metals coupled in an electrolyte; classic failures are a brass valve in a steel line, or carbon-steel bolts on a stainless flange. Control: same metal, insulating kits, coat the cathode, keep the anode large.
3. Crevice corrosion — oxygen-depleted, acidified, chloride-rich stagnant gaps under gaskets, washers, deposits, lap joints and tube-to-tubesheet crevices. Design out crevices; use full-penetration butt welds; raise Mo content.
4. Pitting — local breakdown of passivity, usually by Cl⁻; deep, narrow, autocatalytic pits penetrate a wall with negligible weight loss, so weight-loss coupons miss it. Measured by critical pitting temperature (CPT) and ranked by PREN.
5. Intergranular corrosion — attack at grain boundaries. In austenitic SS, holding 425–815 °C (welding) precipitates Cr₂₃C₆ and depletes adjacent Cr (sensitisation, "weld decay"). Remedies: low-carbon grades (304L/316L), stabilised grades (321 with Ti, 347 with Nb), solution annealing. Stabilised grades can still fail by knife-line attack next to the weld. Caustic/exfoliation variants occur in Al alloys.
6. Selective leaching (dealloying) — one element is removed: dezincification of brass (use inhibited admiralty brass, aluminium brass), graphitic corrosion of grey cast iron, denickelification of cupronickel.
7. Erosion–corrosion — flow strips the protective film: pump impellers, elbows, control-valve outlets, heat-exchanger tube inlets, slurry lines. Includes cavitation and impingement. Control by velocity limits (copper alloys ≈ 2 m/s in water; 1.5–3 m/s typical for CS cooling water), larger-radius bends, impingement baffles, ferrules, harder alloys.
8. Stress corrosion cracking (SCC) — tensile stress (often residual weld stress) + specific environment + susceptible alloy = brittle cracking at stresses far below yield.
   • Chloride SCC of austenitic SS, typically above about 60 °C — the reason 304/316 are avoided for hot chloride duty and for insulated hot lines near chloride-bearing insulation.
   • Caustic embrittlement of carbon steel in hot NaOH (read against the NACE/standard caustic service curves; stress-relieve above the curve).
   • Wet H₂S services: sulphide stress cracking (SSC), hydrogen-induced cracking (HIC), SOHIC — controlled by hardness limits (≤ 22 HRC for most steels), low-sulphur HIC-tested plate, PWHT, per NACE MR0175/ISO 15156 (oilfield, sour) and NACE MR0103/ISO 17945 (refinery).
   • Amine SCC (MEA/DEA units), ammonia SCC of brass and of carbon steel in anhydrous ammonia, polythionic acid SCC on shutdown of hydrotreaters.
9. Also examined: hydrogen damage (blistering, embrittlement, high-temperature hydrogen attack — Nelson curves), microbiologically influenced corrosion (MIC) by sulphate-reducing bacteria in stagnant water, cooling towers and buried pipe, corrosion under insulation (CUI) in roughly 50–175 °C intermittently wet zones, and fretting corrosion.

Environment and high-temperature attack

• Governing variables: pH, chloride and dissolved-oxygen content, temperature (rate roughly doubles per 10–30 °C until a film or solubility limit intervenes), velocity, concentration, impurities, and continuous vs intermittent wetting.
• Concentration effects are not monotonic: carbon steel is used for ≥ 70% H₂SO₄ at ambient but destroyed by dilute acid; 316 handles nitric acid well (oxidising, film-forming) but fails in hydrochloric acid at any strength.
• Dry (high-temperature) oxidation: scale grows by ion diffusion; protection depends on the Pilling–Bedworth ratio — PBR < 1 (Mg) gives a porous cracked scale, 1 < PBR < 2 (Al₂O₃, Cr₂O₃) gives protective adherent scale, PBR > 2 (W, Fe at high T) spalls.
• Other hot mechanisms: sulphidation (modified McConomy curves), carburisation and metal dusting in reformers, fuel-ash/vanadium attack, creep (above roughly 0.4 T_melt the allowable stress is time-dependent, not yield-based), graphitisation of CS above ~425 °C, 475 °C embrittlement and sigma-phase formation that cap duplex SS at about 300 °C.
• API 571 is the industry catalogue of damage mechanisms used in risk-based inspection; API 581 quantifies risk.

Materials of construction

• Carbon steel (IS 2062 plate, IS 1239/IS 3589 pipe, ASTM A106 Gr B / A516 Gr 70): default choice — cheap, weldable, strong; used wherever a modest corrosion allowance buys the design life. Brittle below about −29 °C unless impact-tested (LTCS).
• Low-alloy steels: Cr–Mo (1¼Cr–½Mo, 2¼Cr–1Mo) for hydrogen and high-temperature strength; 3.5% Ni and austenitic grades for cryogenic duty.
• Cast iron: cheap, good in concentrated H₂SO₄ and as pump casings; brittle. High-silicon iron (14.5% Si) is excellent in sulphuric acid but almost unmachinable.
• Stainless steels — Cr ≥ 10.5% gives passivity:
  • Ferritic (409, 430) — magnetic, cheap, chloride-SCC resistant, poor weldability.
  • Martensitic (410, 420) — hardenable; trim, shafts, blades.
  • Austenitic (304/304L, 316/316L, 321, 347, 310, 904L) — the workhorses; Mo in 316 resists pitting; weak point is chloride SCC and chloride pitting.
  • Duplex (2205) and super duplex (2507) — ~50/50 austenite + ferrite: about twice the yield strength of 316L, far better chloride SCC and pitting resistance, lower Ni cost; limited to ≈ 300 °C and needs tight weld-procedure control of ferrite balance.
• Nickel alloys: Monel 400 (Ni–Cu) for HF, seawater, caustic; Inconel 600/625 for high temperature and oxidising-chloride duty; Incoloy 800/825; Hastelloy B-2 for hot reducing HCl, Hastelloy C-276 for oxidising + reducing + chloride mixtures; Alloy 20 for sulphuric acid.
• Titanium — outstanding in wet chlorine, seawater, hypochlorite and oxidising chlorides (passive TiO₂); fails in dry chlorine, HF and reducing acids. Zirconium for hot HCl and sulphuric; tantalum for the most severe acid duty (cost is the limit).
• Copper alloys: copper and cupro-nickel for condenser tubes and potable water; brass and bronze for fittings; attacked by ammonia, oxidising acids and sulphides. Aluminium for cryogenic exchangers and some organics; lead historically for sulphuric acid.
• Non-metals and linings: rubber-lined steel (HCl, brine, slurries, to ~80–90 °C), glass-lined (glass-flaked/enamelled) steel for pharma and agrochem reactors, PTFE/PFA/PP/PVDF linings, FRP and dual-laminate vessels and ducting, PVC/CPVC/HDPE piping, impervious graphite heat exchangers for HCl, acid-proof brick and chemical-resistant cement, carbon brick, ceramics and glass, concrete with sulphate-resisting cement.
• Selection logic (how to answer a design question): fluid composition and impurities → temperature and pressure → mechanical and fabrication needs → corrosion data/experience/coupon tests → compare clad or lined steel against solid alloy → then life-cycle cost. Clad plate or weld overlay (e.g. CS + 3 mm 316L) gives alloy surface at steel prices for thick vessels.
• Corrosion allowance: 1.5 mm (mild/clean service), 3 mm (common carbon-steel default), up to 6 mm (severe) — added to the calculated pressure thickness. Alloy and lined equipment is normally given zero allowance; if a material corrodes fast enough to need a large allowance, it is the wrong material.

Corrosion control

• Design: avoid crevices, dead legs and pockets; provide full drainage and venting; equal-area or no dissimilar-metal couples; smooth flow transitions; adequate thickness at known wear points; access for inspection; weld (not rivet) and grind welds; stress-relieve where SCC is credible.
• Coatings: metallic (hot-dip galvanising, zinc/aluminium thermal spray, electroplating, cladding), inorganic (phosphating, anodising, enamel), organic (epoxy, zinc-rich epoxy primer, polyurethane top coat, coal-tar epoxy, 3-layer PE/FBE for buried pipe). Surface preparation by blasting is the single biggest factor in coating life.
• Inhibitors (dosed, mg/L): anodic (chromate, nitrite, molybdate, phosphate — dangerous if underdosed, causing pitting), cathodic (zinc, polyphosphate), mixed/organic film-formers (amines, benzotriazole for copper), vapour-phase inhibitors for storage; plus oxygen scavengers (hydrazine, now largely replaced; carbohydrazide, sulphite) and biocides for MIC.
• Cathodic protection: make the whole structure a cathode.
  • Sacrificial anode (galvanic): Mg, Zn or Al anodes; no power needed, self-regulating, limited driving voltage — used for tank bottoms, jetty piles, small pipelines, ship hulls, water heaters.
  • Impressed current (ICCP): rectifier plus inert anodes (MMO/Ti, graphite, Si-iron); high output, adjustable, needs power and monitoring — used for long cross-country pipelines and large tank farms. Criterion commonly used: potential of −850 mV vs Cu/CuSO₄ (polarised, "on-off" measured).
  • Anodic protection raises the potential into the passive range — only for passivatable metals in specific electrolytes (e.g. SS in sulphuric acid).
• Monitoring: weight-loss coupons, LPR and ER probes, hydrogen probes, ultrasonic thickness surveys, radiography, dye penetrant and MPI for cracks, IR for CUI, intelligent pigging of pipelines, process-side iron counts.

Codes, standards and Indian practice

• Pressure equipment: ASME Boiler and Pressure Vessel Code Section VIII (Div 1/2) with IS 2825 (unfired pressure vessels) still cited in Indian specifications; piping to ASME B31.3 (process) and B31.4/B31.8 (pipelines). Fired boilers fall under the Boilers Act and Indian Boiler Regulations (IBR), administered by state Boiler Directorates and the Central Boilers Board.
• Materials: IS 2062 (structural/plate steel grades E250 onwards), IS 1239 and IS 3589 (steel tubes and pipes), IS 1161 (structural tubes), IS 6911 (stainless-steel sheet/plate), with ASTM/ASME SA specifications used for most process equipment plates, pipes and fittings.
• Protection: IS 2629 (recommended practice for hot-dip galvanising), IS 4759 (zinc coatings on structural steel), IS 1477 (painting of ferrous metals in buildings), plus Indian Standards on phosphating and on protection of structures against atmospheric corrosion — quote the subject rather than the number if you are unsure, since several older IS numbers have been withdrawn or revised.
• Corrosion-specific industry practice: NACE/AMPP (now AMPP) standards MR0175/ISO 15156 for sour oilfield service, MR0103/ISO 17945 for refineries, SP0169 for buried pipeline CP, SP0198 for CUI; API 571 (damage mechanisms), API 510/570/653 (inspection of vessels, piping, tanks); ISO 9223 for atmospheric corrosivity categories (C1–CX), which is how coastal Indian sites (Mumbai, Chennai, Jamnagar, Paradip) get specified for heavier coating systems than inland ones.
• Statutory links: the Occupational Safety, Health and Working Conditions Code, 2020 (in force 21 November 2025, replacing the Factories Act, 1948) governs plant safety duties including the integrity of equipment; PESO/CCOE licenses pressure vessels and petroleum/gas installations; the MSIHC Rules cover hazardous chemicals; Pollution Control Board consents cover effluent from pickling and plating. Corrosion failures become statutory incidents, which is why MOC decisions are documented and auditable.

Key points

• Corrosion - electrochemical destruction of a metal by its environment; needs anode, cathode, metal path and electrolyte.
• Anode corrodes - oxidation happens at the anode; the cathode is protected. All of CP follows from this.
• Area ratio rule - small anode + large cathode = very high local rate; never coat only the anode.
• Passivity - a thin self-healing oxide (Cr, Al, Ti, Ni); chlorides and reducing conditions destroy it locally.
• Pitting - localised, autocatalytic, chloride-driven; dangerous because weight loss is tiny but penetration is fast.
• Crevice corrosion - oxygen-depleted acidified stagnant gaps; a design problem more than a material problem.
• Sensitisation - Cr-carbide precipitation at 425-815 °C depleting grain boundaries; cured by L, 321 or 347 grades.
• SCC - stress + specific environment + susceptible alloy; chloride SCC of 304/316 above about 60 °C.
• Sour service - wet H₂S cracking controlled by ≤ 22 HRC hardness, HIC-tested steel and PWHT per NACE MR0175/ISO 15156.
• PREN - %Cr + 3.3(%Mo) + 16(%N); ranks pitting resistance: 316L ≈ 26, 2205 ≈ 35, 2507 ≈ 42.
• Duplex 2205 - double the strength of 316L, better chloride resistance, less Ni; limited to about 300 °C.
• Titanium - excellent in wet chlorine and seawater, useless in dry chlorine and HF.
• Corrosion allowance - 1.5/3/6 mm added to the pressure thickness for carbon steel; normally zero for alloys and linings.
• Clad or lined steel - alloy surface on a steel pressure shell; the economic answer for thick vessels.
• Cathodic protection - sacrificial anodes (simple, self-limiting) or impressed current (powerful, needs rectifier and monitoring).
• Inhibitors - anodic types must never be underdosed; organic film-formers and biocides handle cooling water and MIC.
• Acceptable rates - below ~0.1 mm/y is good for process equipment; above ~0.5 mm/y means change the material.
• Non-metals - rubber, glass lining, PTFE, FRP, impervious graphite and acid brick handle duties no economical metal survives.

Formulas and examples

Example 1 — Coupon test to remaining life. A carbon-steel coupon (ρ = 7.87 g/cm³, area 20 cm²) loses 45 mg in 30 days of cooling-water exposure.

• t = 30 × 24 = 720 h.
• CR = 87.6 × 45 / (7.87 × 20 × 720) = 3,942 / 113,328 = 0.0348 mm/y (= 0.0348 × 39.37 ≈ 1.37 mpy) — well inside the "good" band.
• A vessel with a 3 mm corrosion allowance would last 3 / 0.0348 ≈ 86 years at this rate — so carbon steel is the right MOC, provided pitting is absent (check the coupon surface, not just its weight).

Example 2 — LPR probe current to penetration rate. An on-line probe in a process stream reads i_corr = 10 µA/cm² = 0.10 A/m² on iron (M = 55.85 g/mol, n = 2, ρ = 7,870 kg/m³).

• r = iM/(nFρ) = (0.10 × 0.05585) / (2 × 96,485 × 7,870) = 5.585×10⁻³ / 1.519×10⁹ = 3.68×10⁻¹² m/s.
• × 3.156×10⁷ s/y = 1.16×10⁻⁴ m/y = 0.116 mm/y. (Useful benchmark: 1 µA/cm² ≈ 0.0116 mm/y for iron.)
• A 3 mm allowance is consumed in about 26 years; if the probe climbs to 100 µA/cm² (1.16 mm/y), the allowance is gone in 2.6 years — ground for inhibitor dosing or an upgrade.

Example 3 — Sacrificial anode sizing for a tank bottom. A 10 m diameter storage-tank bottom sits on sand; design current density 20 mA/m², design life 20 years, magnesium anodes with a practical consumption rate of 7.9 kg per A-year and utilisation factor 0.85.

• Area = (π/4) × 10² = 78.5 m²; current required = 78.5 × 0.020 = 1.57 A.
• Anode mass = 1.57 × 20 × 7.9 / 0.85 = 248 / 0.85 ≈ 292 kg.
• Specify 6 × 50 kg Mg anodes (300 kg) in a ring, and verify the −850 mV (Cu/CuSO₄) criterion at commissioning. Note: 7.9 kg/A·y and 20 mA/m² are typical design figures; actual values come from soil resistivity surveys and the project CP specification.

Example 4 — PREN comparison for a chloride duty.

• For ~5,000 ppm chloride brine at 70 °C, 316L is marginal on both pitting and chloride SCC; 2205 is the usual economical answer, 2507 or titanium for hot seawater-strength chlorides.

Example 5 — Thickness with corrosion allowance. Vertical vessel, inside diameter 2,000 mm, design pressure 1.0 MPa, SA-516 Gr 70, allowable stress S = 118 MPa at design temperature, E = 1.0 (fully radiographed), CA = 3 mm.

• t = PR/(SE − 0.6P) = (1.0 × 1,000)/(118 × 1.0 − 0.6 × 1.0) = 1,000/117.4 = 8.52 mm.
• Add CA: 8.52 + 3 = 11.52 mm → specify the next commercial plate, 12 mm.
• If the service were switched to 316L clad (3 mm cladding, zero allowance on the clad face), the pressure thickness stays about the same but the 3 mm allowance is dropped — roughly a 9 mm base plate plus cladding, and the controlling cost becomes the cladding, not the steel.

Where it is used

• Process design and EPC (EIL, L&T, Toyo, Technip, Jacobs): the MOC column of line lists, equipment data sheets and the MSD (material selection diagram) — a core deliverable graded in design interviews.
• Refineries and petrochemicals (IOCL, BPCL, HPCL, RIL, HMEL): API 571/581 risk-based inspection, sour-service compliance, hydrogen-service Nelson-curve checks, CUI surveys on hot insulated lines.
• Fertiliser, chlor-alkali and agrochemicals (IFFCO, GSFC, GACL, UPL, Aarti, Deepak Nitrite): titanium in wet chlorine, rubber and FRP in brine and HCl, graphite exchangers, glass-lined reactors.
• Power, desalination and offshore (NTPC, ONGC, shipyards, coastal plants): condenser tube materials, seawater intakes in cupronickel/titanium, ICCP on jetties, platforms and cross-country pipelines (GAIL).
• Pharma and food (Sun Pharma, Cipla, Dr Reddy's, dairies): 316L with electropolished finishes and sanitary design to avoid crevices — corrosion control doubling as cleanability and product-purity control.
• Job roles: materials/corrosion engineer, static equipment and piping engineer, inspection and NDT engineer, reliability engineer, painting and coating inspector (NACE/AMPP CIP or BGAS certification is a recognised premium).

Interview questions

• Q: Why does the anode corrode and not the cathode? A: Oxidation (metal loss) occurs at the anode where electrons leave the metal; the cathode consumes those electrons in a reduction reaction and is protected.
• Q: A small steel bolt is used on a large stainless flange — what happens and why? A: The steel bolt is the anode against a very large cathode, so the entire corrosion current concentrates on a tiny area and the bolt fails fast; use a bolt of equal or nobler material.
• Q: Why is pitting more dangerous than uniform corrosion? A: It penetrates the wall rapidly with almost no measurable weight loss or thinning, so routine coupons and thickness surveys can miss it until there is a leak.
• Q: Why is 316 not used for hot chloride service even though it resists chlorides better than 304? A: Above about 60 °C austenitic stainless steels are prone to chloride stress corrosion cracking regardless of the Mo content; duplex, high-nickel alloys or titanium are used instead.
• Q: What is sensitisation and how do you prevent it? A: Chromium-carbide precipitation at grain boundaries in the 425-815 °C range during welding, depleting adjacent chromium; prevent with low-carbon (304L/316L) or stabilised (321/347) grades, or solution anneal after welding.
• Q: What does PREN tell you and what are its limits? A: It ranks alloys for chloride pitting resistance from composition, but it is empirical — it says nothing about SCC, crevice behaviour in real geometries, weld quality or cost.
• Q: Carbon steel is used for 98% sulphuric acid but not for 10% — explain. A: Concentrated acid forms a protective ferrous-sulphate film at low velocity, while dilute acid is a reducing acid that dissolves steel with hydrogen evolution.
• Q: When would you choose clad steel over solid alloy? A: For thick-walled vessels where only the process face needs corrosion resistance — the carbon-steel backing carries the pressure at a fraction of the alloy cost.
• Q: Sacrificial anode or impressed current for a 200 km buried pipeline? A: Impressed current, because the current demand and the resistance of coated long pipe need the high, adjustable driving voltage of a rectifier; galvanic anodes suit small or short structures.
• Q: Why is an underdosed anodic inhibitor worse than no inhibitor? A: It passivates most of the surface but leaves small bare areas, creating a small-anode/large-cathode cell and severe local pitting.
• Q: Why do alloy and lined vessels usually carry zero corrosion allowance? A: They are selected for a negligible corrosion rate; if a significant allowance were needed, the material choice itself would be wrong, and a thinning lining fails by perforation, not uniform loss.
• Q: What is corrosion under insulation and where do you look for it? A: External corrosion of steel under wet insulation, typically in the roughly 50-175 °C and intermittently wet range — around nozzles, supports, damaged cladding, below drip points on insulated hot lines.

Common mistakes

• Mixing units in the rate formula — mass loss in mg and time in hours for both constants, but area in cm² with the 87.6 constant (mm/y) and in square inches with the 534 constant (mpy); and 1 mm/y is 39.37 mpy, not 100.
• Treating "stainless steel" as one corrosion-proof material instead of a family with very different chloride, acid and temperature limits.
• Judging a localised-corrosion duty by weight-loss coupons only, and missing pitting, crevice attack or cracking.
• Coating the anode and leaving the cathode bare — a coating holiday then becomes a pinhole anode feeding a huge cathode.
• Assuming the corrosion rate rises monotonically with acid concentration or temperature, and ignoring the strength bands where steel is actually usable.
• Adding a large corrosion allowance to fix a badly chosen material, or forgetting to add the allowance at all to the calculated pressure thickness.
• Specifying a high-strength or as-welded hard steel in wet H₂S service, ignoring the hardness and PWHT limits of NACE MR0175/ISO 15156.

Related topics

Process equipment design · Process plant layout and piping · Process plant safety: HAZOP and PSM · Process design and optimisation · Material science and metallurgy · Electrochemistry and electrochemical engineering