Understanding Corrosion —
Galvanic, Pitting, Crevice
and Erosion-Corrosion

Most corrosion failures are preventable. Not by using expensive materials, not by specifying exotic alloys, but by understanding what is actually happening at the metal surface and making design decisions that remove the conditions under which each mechanism operates. The failures that could not have been prevented are rare. The failures that happened because someone put two different metals in contact, or left a crevice where fluid could stagnate, or ignored the velocity limit for the chosen alloy in the specified fluid — these are the bulk of what gets investigated after the fact.

This article covers the four corrosion mechanisms responsible for the majority of preventable engineering failures in process plant, pipework, and fabricated equipment. For each: the mechanism, the conditions required for it to operate, the material and design combinations that are most vulnerable, and the practical design decisions that prevent it.

Uniform General Corrosion — The Baseline

Before the four specific mechanisms, it is worth establishing general corrosion as the baseline case. General (or uniform) corrosion is the even, predictable removal of metal from an exposed surface by chemical or electrochemical attack. Rusting of carbon steel in water is the most familiar example. The rate is relatively predictable for a given metal-environment combination, measured in millimetres per year (mm/yr) or mils per year (mpy), and the design response is straightforward: add a corrosion allowance to the calculated wall thickness sufficient to last the intended design life.

General corrosion is the least dangerous mode because it is predictable and gives visible warning. The four mechanisms covered below are more dangerous precisely because they are localised, accelerated, and often occur with minimal or no prior warning on the external surface of the component.

Galvanic Corrosion

The mechanism

When two dissimilar metals are in electrical contact in the presence of an electrolyte (any conductive liquid — water, process fluid, moisture on a surface), they form an electrochemical cell. The less noble metal (the anode) corrodes preferentially — it oxidises and gives up metal ions to the electrolyte — while the more noble metal (the cathode) is protected. The driving force is the potential difference between the two metals on the galvanic series. The greater the potential difference, the more aggressive the galvanic attack on the anode.

The galvanic series

The galvanic series ranks metals and alloys by their electrochemical potential in seawater. Metals at the active (anodic) end are corroded preferentially; metals at the noble (cathodic) end are protected. From most active to most noble, the approximate order of common engineering materials:

Magnesium → Zinc → Aluminium → Carbon steel → Cast iron → Low-alloy steel → 316 stainless (active) → Lead → Tin → Copper → Brass → Bronze → Monel → Silver → Titanium → Platinum → Gold → 316 stainless (passive)

The separation between two materials on this series determines the severity of galvanic attack. Carbon steel and copper in contact in seawater is a particularly aggressive couple — the steel corrodes rapidly. Aluminium and stainless steel in contact will attack the aluminium. Carbon steel and zinc are a deliberate coupling — zinc is the sacrificial anode in galvanising, protecting the steel.

What makes it worse

The area ratio of cathode to anode is critical. A small anode coupled to a large cathode accelerates corrosion of the anode severely — the large cathodic area drives a high current through a small anodic area, concentrating attack. A steel bolt in a copper alloy plate is far more dangerous than a copper bolt in a steel plate. The classic failure mode: stainless steel fasteners in a carbon steel flange are relatively benign (small noble cathode, large active anode). Carbon steel bolts in a stainless steel flange — the steel bolt is the small anode with a large cathodic stainless flange driving the attack — will corrode aggressively.

Prevention

• Avoid dissimilar metal couples where possible — use the same alloy throughout a system
• Where dissimilar metals must be joined, insulate them electrically — PTFE isolation gaskets, nylon bolt sleeves, and isolation flanges between carbon steel and stainless or copper alloy pipework
• If the couple cannot be avoided, make the anode large relative to the cathode — never the reverse
• Apply coatings to the cathode (not the anode — a break in a coating on the anode concentrates attack dramatically)
• Use sacrificial anodes deliberately — zinc anodes on steel structures in seawater, magnesium anodes on buried pipework
• Reduce or eliminate the electrolyte — dry assemblies are galvanically inert

Pitting Corrosion

The mechanism

Pitting corrosion is localised electrochemical attack that produces small, deep cavities (pits) in an otherwise intact surface. It is the characteristic failure mode of passive alloys — stainless steels, aluminium alloys, nickel alloys — in the presence of specific aggressive ions, most commonly chlorides. The passive oxide film that gives stainless steel its corrosion resistance breaks down locally at microscopic defects or heterogeneities in the film. Once the film breaks down at a point, the bare metal beneath is exposed to the electrolyte and becomes a small, highly active anode. The surrounding intact passive surface acts as the cathode. The highly unfavourable area ratio (tiny anodic pit, large cathodic passive surface) drives intense localised attack — the pit grows rapidly downward while the surrounding surface remains visually unmarked.

Pitting Resistance Equivalent Number (PREN)

The susceptibility of stainless steels and nickel alloys to pitting corrosion in chloride environments is characterised by the Pitting Resistance Equivalent Number:

PREN = %Cr + 3.3×%Mo + 16×%N

Higher PREN indicates better pitting resistance. As a practical guide:

• 304L (PREN ~18) — suitable for low-chloride environments (<100 ppm Cl⁻ typically)
• 316L (PREN ~25) — improved resistance, suitable for moderate chlorides. The standard upgrade from 304L for chloride-containing process streams.
• Duplex 2205 (PREN ~35) — significantly better, suitable for seawater injection and moderately aggressive chloride service
• Super Duplex 2507 (PREN ~43) — seawater service, produced water, aggressive chloride environments
• Alloy 625 (PREN ~51) — highly aggressive chloride environments, marine, sour gas

Temperature significantly affects pitting susceptibility — the critical pitting temperature (CPT) is the temperature above which pitting initiates in a standard test solution. 316L pits readily in warm seawater above ~25°C; 2205 is reliable to ~50°C; 2507 and higher-PREN alloys extend this to 80°C and above.

Prevention

• Select an alloy with PREN appropriate to the chloride concentration and temperature — not the cheapest alloy that looks like stainless
• Maintain fluid velocity above the critical deposition velocity to prevent chloride concentration under stagnant deposits
• Avoid surface contamination during fabrication — iron contamination from carbon steel tooling initiates pitting in stainless. Dedicated tooling, passivation after fabrication, and freedom from embedded iron particles are process control requirements for stainless fabrication
• Maintain the passive film — avoid mechanical damage, weld discolouration (heat tint), and crevices where the passive film cannot be maintained

Crevice Corrosion

The mechanism

Crevice corrosion occurs in confined spaces — the gap under a gasket, the contact area between two overlapping plates, the space under a bolt head, the annulus between a tube and tubesheet — where the electrolyte is present but fluid exchange with the bulk environment is restricted. The mechanism involves progressive depletion of oxygen in the crevice (the oxygen is consumed by the corrosion reaction and cannot be replenished from the stagnant fluid in the crevice) and accumulation of aggressive ions (particularly chlorides, which migrate into the crevice to maintain electrical neutrality as metal ions are produced). The result is an acidic, oxygen-depleted, chloride-enriched environment within the crevice — conditions that break down the passive film and produce intense localised attack at the crevice walls.

Crevice corrosion is particularly insidious because it occurs in locations that are not visually inspectable in service. By the time it is detected — usually during a maintenance shutdown, when a gasket is removed or a fitting is disassembled — the damage may already be through-wall.

Critical crevice gap

Crevice corrosion requires a gap narrow enough to restrict fluid exchange (typically <0.1–0.5mm for most alloy-electrolyte combinations) but wide enough for the electrolyte to enter. A perfectly tight metal-to-metal contact with no gap is not a crevice. A very wide gap with free fluid circulation is not a crevice. The dangerous range is the narrow gap that traps stagnant electrolyte — exactly the geometry produced by a standard soft-seated gasket, a fillet weld attachment, or an overlapping joint.

Prevention

• Eliminate crevices by design — butt welds rather than fillet welds, full penetration welds at tube-to-tubesheet joints, flush-mounted fittings, elimination of dead legs and stagnant zones
• Use non-metallic gasket materials that conform tightly to the flange face and eliminate the gap at the gasket inner diameter — spiral wound gaskets with compression stops are better than flat cut gaskets
• Where crevices cannot be eliminated, seal them — weld the gap closed, fill with a compatible sealant, or maintain a biocide in the fluid to suppress biological growth that would otherwise accelerate attack
• Use higher-PREN alloys in crevice-prone geometries — the critical crevice temperature for a given alloy is always lower than the critical pitting temperature for the same alloy, so an alloy that resists pitting in open surfaces may still suffer crevice attack at the same temperature
• Avoid painting over surfaces that contain crevices — a paint break at a crevice creates a small anode in a large cathodic painted surface, concentrating attack at exactly the most vulnerable geometry

Erosion-Corrosion

The mechanism

Erosion-corrosion is the combined and synergistic action of mechanical erosion (physical removal of metal by fluid flow, entrained particles, or cavitation) and electrochemical corrosion. The two mechanisms accelerate each other: erosion removes the protective corrosion product or passive film from the metal surface, exposing fresh metal to corrosive attack. Corrosion then attacks the freshly exposed surface, producing a softened or pitted surface that is more susceptible to the next cycle of erosion. The combined attack rate is typically greater than the sum of the two individual rates.

Where it occurs

Erosion-corrosion concentrates at flow geometry changes where velocity, turbulence, or impingement is highest:

• The extrados of bends — particularly small-radius bends — where the fluid impinges on the outer wall
• Downstream of orifice plates, control valves, and flow restrictions where turbulence is high
• Pump impellers, particularly in slurry or particle-laden service
• Heat exchanger tube inlets — the first 100–150mm of tube downstream of the inlet tubesheet
• Tee junctions where a branch flow impinges on the main run

Velocity limits

Each material-fluid combination has a critical velocity above which erosion-corrosion becomes significant. Exceeding this velocity produces accelerating attack that no corrosion allowance can address — the rate is not linear, and the corrosion allowance will be consumed far faster than the general corrosion rate predicts. Indicative maximum velocity limits for common material-fluid combinations:

Slurry and particle-laden fluids

Entrained solid particles dramatically lower the threshold velocity for erosion-corrosion. Particle hardness, size, shape, and concentration all affect the attack rate. Angular particles (quartz, alumina) are more damaging than rounded particles of equivalent size. For slurry service, rubber lining, ceramic coating, or high-chrome white iron are the standard materials; stainless steel is generally not hard enough to resist slurry erosion at moderate concentrations.

Prevention

• Keep velocity below the critical limit for the material-fluid combination
• Use long-radius bends rather than short-radius — the impingement angle at the extrados is lower and attack is distributed over a larger area
• Target injections and branches so that the incoming flow does not impinge at 90° on the opposite wall of the main run
• Specify replaceable erosion targets (sacrificial inspection spools or corrosion coupons) at known high-attack locations
• Select harder or more resistant materials at geometry changes — hard-faced or stellite-overlay trim in valve seats and plugs, high-chrome white iron impellers in slurry pumps
• Eliminate or reduce entrained particles — strainers upstream of control valves and pumps, removal of sand and grit from seawater systems

Stress Corrosion Cracking — The Bonus Mechanism

No article on corrosion mechanisms is complete without noting stress corrosion cracking (SCC), even though it falls outside the four headline mechanisms. SCC is the combined action of a susceptible material, a specific corrosive environment, and tensile stress — all three must be present simultaneously. Remove any one of the three and SCC does not occur.

The most important combinations in process engineering:

• Austenitic stainless steel + chlorides + tensile stress — the classic SCC system. Stainless steel vessels and heat exchanger shells crack in the presence of chloride-containing fluids (including humid atmospheres with chloride contamination) under residual or applied tensile stress. Temperature above ~60°C accelerates attack dramatically. Post-weld heat treatment (PWHT) or stress relief reduces residual stress and significantly reduces susceptibility, but cannot eliminate it where the applied stress is high.
• Carbon steel + hydrogen sulphide (H₂S) + stress — sulfide stress cracking (SSC), governed by NACE MR0175 / ISO 15156. High-strength steels are most susceptible. Hardness limits (typically ≤22 HRC for base metal, ≤35 HRC for weld deposits) are imposed in sour service to prevent SSC.
• Copper alloys + ammonia + stress — season cracking. Copper alloy components in ammoniated environments under stress crack. Relevant in refrigeration plant and fertiliser facilities.

Putting It Together — A Design Corrosion Checklist

For any new design involving corrosive service, the following questions identify the most significant risks before a line is drawn:

1. Are dissimilar metals in electrical contact? What is the electrolyte? What is the area ratio? Can they be isolated?
2. Does the fluid contain chlorides? What is the temperature? What PREN is required for the stainless alloy selection?
3. Where are the crevices? Gaskets, fillet welds, flanged connections, threaded fittings — can any be eliminated by redesign?
4. What is the fluid velocity at bends, reducers, valve outlets, and heat exchanger tube inlets? Does it exceed the critical erosion-corrosion velocity for the material?
5. Are there tensile stresses in a susceptible material in a specific corrosive environment? Is SCC a risk?

Summary

Galvanic corrosion requires two different metals, an electrolyte, and electrical contact — remove any one and it stops. Pitting corrosion requires a passive alloy, chlorides, and stagnant conditions — select an appropriate PREN for the chloride concentration and temperature. Crevice corrosion requires a narrow confined gap with stagnant electrolyte — design out the crevice. Erosion-corrosion requires excessive velocity or turbulence at a geometry change — stay below the critical velocity and use appropriate geometry. None of these mechanisms is unpredictable. All of them are, in most cases, preventable through design decisions made before fabrication begins rather than after investigation of the failure that was allowed to happen.