How to Calculate Galvanic Corrosion Life: Area Ratio, Potential Difference and Faraday's Law

A 304 stainless steel panel fixed with carbon steel bolts rusts through in three months on a coastal site. Turn the assembly around, use a carbon steel plate with a handful of 304 bolts, and the same two metals in the same salt air last twenty years. The materials did not change. What changed is which one has the larger area.

Galvanic corrosion is usually discussed at the level of a warning: these two metals must not touch. But service life is decided by three quantities you can actually compute. The potential difference tells you whether there is a driving force. The area ratio tells you how tightly that current is squeezed onto the anode. Faraday's law converts the current into millimetres of metal lost per year. This article works through all three, gives six examples you can reuse directly, and states the single rule that is both most violated and most damaging on site: coating belongs on the cathode.

Step 1: Potential Difference, the Driving Force

When two metals share an electrolyte (seawater, rainwater, a condensation film), the more negative one becomes the anode and dissolves, while the more positive one becomes the cathode and is protected. The driving force is the difference between their corrosion potentials. Values below are typical measurements in quiescent seawater against a saturated calomel electrode (SCE).

A three-band judgement on the potential difference is standard practice:

• Below 0.05 V — the driving force is negligible and direct contact needs no special treatment. 304 against 316L sits in this band.
• 0.05 to 0.25 V — acceptable indoors in dry conditions; outdoors, in rain, condensation or coastal exposure, isolation is required.
• Above 0.25 V — isolation is mandatory regardless of indoor or outdoor exposure. This is the threshold long used in aerospace and defence standards.

On this table, passive 304 sits 0.55 V from carbon steel and 0.70 V from aluminium, both far beyond the 0.25 V threshold. That means whenever stainless steel is electrically connected to carbon steel or aluminium in a wet environment, isolation has to be considered.

Basis of calculation: Potentials are typical measured corrosion potentials in quiescent aerated seawater (SCE reference); published values vary by roughly ±50 mV with temperature, flow velocity and surface condition, so site measurement or closer literature data should govern in practice. The oxygen reduction limiting current density in the worked examples is taken as 100 µA/cm², within the usual 50 to 200 µA/cm² range for a passive stainless surface in quiescent aerated seawater, and is a deliberately conservative figure. Rates derived from this model are therefore upper bounds; electrolyte resistance, polarisation and calcareous deposits reduce them, and measured values typically fall between 10% and 50% of the estimate. Cross-checked against seawater exposure data, measured values typically run at 16% to 27% of the theoretical figure (about 20% on average); see the section on cross-checking below. All examples assume uniform attack spread over the whole exposed anode area and do not account for localised morphology.

Step 2: Area Ratio, the Quantity That Decides Failure

The potential difference answers whether corrosion happens. The area ratio answers how soon. The reason is simple: the total oxygen reduction reaction on the cathode scales with cathode area, and all of that current is collected onto the anode, where it is divided by the anode area.

where i_lim is the limiting diffusion current density for oxygen reduction on the cathode surface (µA/cm²), A_c is the cathode area, A_a is the anode area, and i_gal is the galvanic current density actually carried by the anode.

This expression contains the most important rule in galvanic corrosion: a small anode coupled to a large cathode is the dangerous configuration. Shrinking the anode does not reduce the total amount of corrosion. It concentrates the same total attack onto a smaller area, multiplying penetration depth. Bolts, welds, rivets and washers occupy that dangerous position by their very nature.

The reverse, a large anode with a small cathode (a big carbon steel plate with a few stainless bolts), spreads the cathodic current over an enormous anode area and adds almost nothing to the corrosion rate.

Step 3: Faraday's Law, Converting Current into Millimetres per Year

With a current density in hand, Faraday's law converts it to a penetration rate:

EW is the equivalent weight (atomic mass divided by valence; for alloys the reciprocal is weighted by mass fraction) and ρ is density in g/cm³. Values used here:

The equivalent weight of 304 comes from the mass-fraction weighted reciprocal:

Carbon steel at 1 µA/cm² gives 11.62 µm/y, which matches the textbook rule of thumb that 1 µA/cm² is about 11.6 micrometres per year for steel, and serves as a cross-check on the constant.

Example 1: 304 Panel with Carbon Steel Bolts, Failure in Three Months

A 1.0 m² 304 stainless panel is fixed with 20 carbon steel bolts, each exposing roughly 2.5 cm² of head and shank.

At 232.3 mm/y, an M20 bolt (10 mm radius) would in theory lose half its radius in under a month. Even discounting to 10 to 50 percent of the theoretical figure, giving 23 to 116 mm/y, the life is still measured in months. This is why stainless equipment on a coastal site loses its carbon steel fasteners within a season.

Example 2: Swapping the Areas Cuts the Rate by 784 Times

Now reverse the geometry: a 1.0 m² carbon steel plate as the main body, with 50 cm² of 304 stainless bolts as fasteners. Identical pair of metals, only the roles and the areas exchanged.

Carbon steel self-corrodes at about 25 µA/cm² in seawater, so the total is 25.5 µA/cm²:

Bare carbon steel alone gives 3.27e-3 × 25 × 3.552 = 0.2904 mm/y, so the increase is only:

Putting the two examples side by side:

The same two metals, the same contact interface, and the rate differs by a factor of 784 purely because of which side is large. This is why the sentence "stainless must not touch carbon steel" is both true and useless: the real question is which component is the small one.

Example 3: Aluminium Grating with 304 Bolts, Safe

2.0 m² of aluminium grating with 30 stainless bolts, each exposing about 1 cm².

Aluminium self-corrodes at roughly 3 µA/cm² in seawater, giving 3.15 µA/cm² in total:

Bare aluminium gives 3.27e-3 × 3 × 3.330 = 0.0327 mm/y, an increase of 3.15 / 3 - 1 = 5.0%. This direction is safe; in practice a nylon washer under the bolt head is enough to stop foot traffic wearing through the oxide film.

Example 4: Aluminium Trim Against 304 Posts, Perforated in 40 Days

Reversed again: five 304 posts totalling 0.5 m², with 200 cm² of aluminium trim. This is a very common facade and balustrade detail.

Time to perforate 3 mm aluminium trim:

Example 5: 95% Isolation Extends Life 20 Times

Add nylon washers and sleeves to Example 4, isolating 95% of the conducting contact area so only 5% of effective cathode remains.

The reduction, 1 - 1.361 / 27.2 = 95.0%, equals the isolation ratio, which is the internal consistency check on this model. Life rises from 1.32 months to:

Example 6: The Rule — Coating Goes on the Cathode

When Example 4 fails, the most common reaction is to paint the aluminium trim. It is also the most damaging mistake in galvanic protection. No coating is perfect; pinholes, scratches and thin edge coverage all leave defects. At a 1% defect rate the effective anode area collapses from 200 cm² to 2 cm²:

Time to perforate 3 mm aluminium:

Bare metal survives 40 days; painted, it perforates in under half a day, one hundred times faster. The coating compressed the anode into pinhole-sized points, and the entire galvanic current discharges through them.

The correct action is to coat the cathode, the five 304 posts. Covering 95% of the cathode leaves 250 cm² effective, which gives exactly the 1.361 mm/y of Example 5. Reducing cathode area by coating helps; reducing anode area is catastrophic. That is the one rule worth memorising.

Six Examples Compared

The number worth remembering is not any single value but the relationship between cases 4 and 6: one well-intentioned coat of paint compresses 40 days of life into half a day.

Cross-Check Against Measured Data: What Discount to Apply

Every example above is a theoretical upper bound derived from the limiting diffusion current density for oxygen reduction. Polarisation, electrolyte resistance and calcareous deposits all suppress it in service, so the real number is lower. Three external datasets show by how much.

The first row is a pure Faraday conversion with no environmental term to discount, and the two routes agree to 99.9%, which validates the constant and the equivalent weights used here. The second is an independent theoretical model: an i_L of 0.1 A/m² (10 µA/cm²) multiplied by the area ratio of 100 gives 1000 µA/cm², and substituting it yields 11.62 mm/y, within 100.7% of NovaSolver's 11.7 - model against model needs no discount. The third is real seawater exposure, where the discount is largest: 26.7%, 16.1% and 18.1% across the three ratios, averaging about 20.3%.

In practice, multiply the figures above by roughly 0.2. Even then, Example 1 still gives 46.5 mm/y from 232.3, so an M20 bolt remains a matter of months and the conclusion is unchanged.

Two Corroborations from Standards Bodies

The Australian Stainless Steel Development Association (ASSDA) publishes two judgements that agree with this article. First, the potential difference between passive 304 and passive 316 is small, so galvanic corrosion between them is not expected even at large area ratios. That is direct support for the claim that upgrading to 316L does not solve a galvanic problem. Second, the rule of thumb is that galvanic effects are not serious once the wetted area of the corroding metal reaches ten times that of the noble metal. Examples 2 and 3 here sit at area ratios of 0.005 and 0.0015, far below the 0.1 threshold, and both are judged safe accordingly.

The same source also marks the limit of Example 3: a stainless screw in aluminium is frequently used, yet the aluminium immediately around the stainless can corrode, because the wetted area ratio shifts from 1:50 towards 1:1 as the joint dries after rain. That is why Example 3 still recommends a nylon washer even though the geometry is safe.

Sizing a Sacrificial Anode: W = I × t × EW / F

Where isolation is impossible and sacrificial protection is the only option, the same law gives the consumption:

For zinc, EW = 32.69. With a protection current I = 0.1 A over one year, t = 31536000 s:

So every 0.1 A of protection current consumes 1.07 kg of zinc per year. Bare carbon steel in seawater needs about 120 mA/m², so 10 m² requires 1.2 A, consuming 12.8 kg per year and 128 kg over ten years; at a realistic 85% utilisation that is 151 kg to install. This is why sacrificial anodes are mostly confined to the immersion zone, with coating and isolation preferred for atmospheric exposure.

Throw Distance: Why Atmospheric Galvanic Attack Stays in the Joint

Galvanic current must travel through the electrolyte, and the longer and thinner that path, the higher the resistance:

Immersion in seawater, ρ_e = 0.20 ohm·m, electrode spacing L = 0.1 m, current path cross-section A = 0.01 m²:

In an atmospheric condensation film, take the same saline ρ_e = 0.20 ohm·m, but a film thickness of 50 µm over 0.1 m width gives a cross-section of only 0.000005 m², with a path length of 0.01 m:

Resistance is two hundred times higher, current is suppressed by two orders of magnitude, and the affected zone therefore shrinks to a few millimetres around the joint. This explains the familiar pattern: in seawater, galvanic attack can reach components tens of centimetres away, while in atmospheric exposure it is confined to the bolt hole, the lap seam and the washer edge, decaying rapidly once you move away from the contact line.

Stainless Steel Is Two-Faced: Passive Cathode, Active Anode

Stainless steel has a peculiarity in the galvanic series that causes a great deal of misjudgement. Well aerated, it is passive at -0.05 V and behaves as a strong cathode. Confined in an oxygen-starved crevice (under a bolt head, under a washer, inside a lap joint), the passive film reduces and dissolves, the potential falls to -0.40 V, and it becomes the anode instead.

The same component then forms a couple with itself: the large external passive area acts as cathode and the small active area inside the crevice acts as anode, which is precisely the dangerous small-anode geometry. That is the mechanism of crevice corrosion, and the reason stainless flanges and bolt hole edges fail so often.

The controlling dimension is the gap width. Below about 0.1 mm the solution stagnates and oxygen cannot be replenished; risk rises sharply, and standard crevice corrosion test methods use a gap of this order. In practice, avoid leaving capillary gaps in lap joints, or seal the mouth of the joint so that electrolyte cannot enter.

Will Upgrading to 316L Fix It? No

This is the most common selection error. 316L resists pitting far better than 304 thanks to molybdenum, but its corrosion potential is -0.08 V against -0.05 V for 304, a difference of only 0.03 V. In the galvanic series they occupy the same band. Swapping 304 for 316L barely changes the component's ability to drive corrosion of aluminium or carbon steel.

If the failure mode is galvanic, upgrading the grade spends money in the wrong place; the fix is isolation (washers, sleeves, sealant) or changing the area ratio. If the failure mode is chloride pitting on a single plate, then 316L is the right answer. The two mechanisms differ, the remedies differ, and the mode should be identified before specifying.

Assessment of Common ASPER Joints

The manhole frame and hinge case in detail: 304 hinge effective area 100 cm², cast iron frame 5000 cm².

Cast iron self-corrodes at 30 µA/cm², giving 32 µA/cm² total, with EW/ρ = 27.92 / 7.20 = 3.878:

Cast iron alone gives 3.27e-3 × 30 × 3.878 = 0.3804 mm/y, an increase of 32 / 30 - 1 = 6.7%, which is acceptable.

Site Checklist

1. List every metal in the joint, including filler metal, washers, coatings and fasteners. Treat a coating and its substrate as separate entries.
2. Check the potential difference. Below 0.05 V, accept. Above 0.25 V, continue.
3. Compute the area ratio and identify which part is the small one. If the small part is the anode, change the design.
4. Estimate the rate. Multiply i_lim by the area ratio for current density, convert to annual penetration with Faraday's law, compare against section thickness for a life figure.
5. Choose the measure. Insulation first (washers, sleeves, sealant); then change the area relationship; sacrificial anodes as a supplement in immersion zones.
6. Check coating direction. If coating is unavoidable, coat the cathode, never the anode alone. Painting only the anode is worse than doing nothing unless the coating is truly defect-free.

Frequently Asked Questions

Q1: Must stainless steel and carbon steel never touch?

Not necessarily; it depends on the area relationship. A large carbon steel plate with a few stainless bolts gains only 2.0% (Example 2) and survives twenty years. A large stainless panel with carbon steel bolts runs at a theoretical 232.3 mm/y and fails in months. The same two metals, with only the area ratio reversed, differ by a factor of 784. The criterion is not which metals, but which component is small.

Q2: It is already installed. What is the simplest remedy?

In order of effectiveness: first, insert an insulating layer — nylon or PTFE washers and sleeves separating the shank from the hole wall and the contact faces; 95% isolation extends life about 20 times (Example 5). Second, seal the joint mouth with silicone or polyurethane sealant to break the electrolyte path, which is equivalent to driving the effective area close to zero. Third, coating, and only on the cathode. Never solve the problem by painting the anode (the carbon steel or aluminium part) alone: at a 1% defect rate, perforation time drops from 40 days to 9.7 hours, a hundred times faster than bare metal.

Q3: On a coastal project, will changing 304 to 316L stop the bolts failing?

If the bolts are corroding on their own account (chloride pitting), 316L helps. If they are being driven as small anodes by a large stainless panel (galvanic corrosion), 316L makes almost no difference, because its potential of -0.08 V is only 0.03 V from 304's -0.05 V and the cathodic driving power is essentially unchanged. The correct fix is to use fasteners of the same alloy as the panel, bringing the difference below 0.03 V, or to isolate the joint completely. Identify the morphology first: attack concentrated at the contact interface and running into the crevice indicates galvanic action, while scattered pits indicate pitting.