Corrosion Of Steel In Concrete

Why Your Concrete Structure Is Cracking (And What To Do About It)

I spent last Tuesday pulling cores out of a mid-rise parking garage built in 1987. The concrete looked fine on the surface. Underneath, the rebar was reduced to something between a rod and a mesh screen in several locations. The spalling patterns were textbook rust-jacking — the oxide layer took up roughly five to seven times the volume of the original steel, and the concrete literally exploded outward from the pressure. Corrosion Of Steel In Concrete is one of the most expensive deterioration mechanisms in infrastructure. The American Society of Civil Engineers estimates it costs the U.S. over $100 billion annually. That number doesn't include the secondary damage — degraded structural capacity, serviceability issues, premature replacements. The economics are brutal.

Corrosion Of Steel In Concrete: The Actual Mechanism

Portland cement pore water has a pH around 12.8 to 13.2. At that alkalinity, the steel passivates almost instantly. A thin, stable iron oxide layer forms on the rebar surface — the passive film — and corrosion effectively stops. This is why newly poured concrete protects embedded steel for decades under the right conditions. The protection fails when two things happen. Carbonation slowly reduces the pH of the concrete matrix by reacting CO with calcium hydroxide to form calcium carbonate. When carbonation reaches the rebar depth and pH drops below approximately 9, the passive film breaks down. Alternatively, chloride ions penetrate the concrete and locally disrupt the passive layer even at high pH. Both pathways lead to active corrosion. Both are preventable with proper mix design and cover depth. Once corrosion initiates, the rust products are voluminous. Ferrous hydroxide and later hydrated iron oxides occupy roughly 2 to 6 times the volume of the original steel depending on the exact oxidation state. This creates tensile stresses in the surrounding concrete. Concrete has very low tensile strength — typically 8 to 15% of its compressive strength — so cracks form within months of active corrosion starting. Those cracks let in more water, more oxygen, more chlorides. It's a self-accelerating feedback loop.

Prevention: Where Most Projects Go Wrong

I've reviewed more repair specifications than I care to count. The most common failure point is inadequate concrete cover. The ACI 318 minimums exist for a reason. A 2-inch cover for a parking garage deck in a de-icing salt environment isn't being conservative — it's the floor. Go below that and you're gambling with the structure's remaining service life. Clinker type matters more than people realize. Type I/II portland cement will carbonation-slow compared to older Type I mixes. But the real lever is supplementary cementitious materials. Fly ash at 25 to 35% replacement substantially reduces chloride diffusivity by refining the pore structure and consuming calcium hydroxide through pozzolanic reactions. Slag cement at 50% replacement gives even better resistance. The tradeoff is slightly slower early strength gain, which matters if you're working tight schedules. You plan around it. It's worth it. Low water-cement ratio is non-negotiable. I've seen specs call for 0.45 w/c on exposed decks. That's generous. Aim for 0.40 or lower. Every 0.05 reduction in w/c cuts chloride penetration roughly in half based on accelerated testing data from multiple lab studies.

Here's something most spec writers skip: curing. Proper moist curing for at least 7 days — 14 for slag mixes — dramatically reduces surface porosity and carbonation rate. A well-cured 4,000 psi mix outperforms a poorly cured 5,000 psi mix every time. I had a project where the contractor stripped forms too early and we lost about 15% of the designed surface strength. Had to core-test three separate locations before we could confirm the slab was still within tolerance. Nobody wants that phone call.

Diagnosis: Testing Methods That Actually Work

Half-cell potential mapping (ASTM C876) is the standard field tool. You place a copper-copper sulfate reference electrode on the concrete surface and measure the voltage between it and the rebar. More negative readings indicate a higher probability of active corrosion. The thresholds are: above -200 mV means less than 10% probability of corrosion. Between -200 and -350 mV is ambiguous. Below -350 mV means greater than 90% probability. But here's the thing nobody tells you — half-cell potential only tells you about corrosion likelihood, not corrosion rate. A structure could show highly negative potentials across a large area and be corroding very slowly if oxygen availability is limited. Conversely, a small area with moderately negative readings could be actively attacking. Always pair potential mapping with resistivity measurements (ASTM C1760). Low resistivity plus negative potential is a red flag. High resistivity with negative potential might just mean moisture is present without active corrosion. Rapid chloride permeability testing (ASTM C1202) on core samples gives you a quantitative measure of chloride ingress resistance. Values below 1,000 coulombs indicate excellent durability. Between 1,000 and 2,000 is good. Above 4,000 means your concrete is letting chloride through like it's not even there. I once saw a deckPour clock in at 6,200 coulombs. The w/c was right on paper. The contractor had over-vibrated the mix, segregating the aggregate and creating continuous permeability pathways. Paper specs don't catch that. Field observation does.

Repair: The Practical Reality

When you're dealing with active corrosion, you have three main approaches. Patch repair with polymer-modified cementitious materials is the most common. Remove all delaminated and contaminant-affected concrete back to sound material — which usually means 25 to 50 mm beyond the visible deterioration boundary. Clean the rebar to White Metal blast profile (SSPC-SP 5), apply an inhibiting primer if the spec calls for it, then place the repair material. The key detail is the bond. Surface preparation determines whether your repair lasts 5 years or 25. I've seen contractors use needle scalers and call it done. It wasn't done. The old concrete surface was already saturated with chlorides and micro-cracks. Those defects transfer straight through to the new material. Cathodic protection is the more aggressive option. Impressed current systems can halt active corrosion in place without removing concrete. The investment is steeper — roughly $50 to $75 per square foot installed — but the system can protect the structure for 30-plus years after installation. I deployed this on a marine pier where access for patch repairs was logistically impossible due to tidal windows. The cathodic system was installed in sections over three months. Current densities were maintained at 4 to 6 mA/m² of rebar surface area. Post-installation monitoring showed potentials shifting positive by 150 to 300 mV within 90 days. Corrosion effectively stopped. Electrochemical chloride extraction is another option, particularly for structures where chloride contamination is the primary issue rather than carbonation. The process uses an applied current to drive chloride ions out of the concrete toward an anode placed on the surface. Treatment typically takes 6 to 12 months for a full deck. Post-treatment, you need to neutralize the remaining alkalinity and restore the passive environment. This is niche but effective for the right application.

A Problem I Ran Into With Epoxy-Coated Rebar

I was overseeing a bridge deck replacement where the spec called for epoxy-coated reinforcement throughout. During inspection, I noticed the stirrup ties at several supports were standard black steel, not epoxy-coated. The detail drawing had specified stainless steel ties, but the fabricator had substituted. Here's what happens when you couple epoxy-coated rebar with uncoated carbon steel ties: the coating creates a large anode-to-cathode area ratio. The exposed tie at any damage point in the coating becomes a tiny cathode relative to the vast coated anode surface, and the corrosion rate at that breach accelerates dramatically. It's the opposite of what most people assume — the coated bar corrodes faster at any scratch or cut end, not slower, because the coated surface acts as an efficient cathode driving localized attack at the defect. The workaround was straightforward but costly. We pulled the non-compliant ties and replaced them with Type 316 stainless steel wire. Then we inspected every bar for coating damage during placement — a handheld spark tester at 9,000 volts caught about 12% of the bars having holiday defects. Most were minor nicks at the bend points. We touch-coated everything with an approved epoxy repair compound and retested. Took an extra afternoon. Far cheaper than replacing the deck in five years.

What These Methods Don't Fix

Cathodic protection stops corrosion but doesn't restore lost cross-sectional area. If your rebar has lost more than 20% of its original diameter, you need structural strengthening regardless of whether the remaining steel is still corroding. Carbon fiber wrapping or external steel plate bonding brings the capacity back. I've seen engineers skip this step because the cathodic system "worked" and the potentials normalized. The structure was safer from further degradation but still carried a reduced load capacity. That's a liability you carry forever. Resistivity-based durability predictions assume stable environmental conditions. Freeze-thaw cycles, wet-dry cycling, and temperature fluctuations all affect resistivity independently of chloride content. A resistivity reading of 20,000 ohm-cm in summer doesn't mean the same thing in winter when the concrete is partially frozen and resistivity spikes artificially. Don't use a single seasonal reading to make durability claims. Take multiple readings across seasons if you can. And here's a counterintuitive point about carbonation: carbonation itself doesn't directly cause corrosion. It removes the alkaline protection that keeps corrosion suppressed. In a dry environment where carbonation reaches the rebar but moisture is absent, you won't see active corrosion. Water and oxygen are the other two ingredients the electrochemical reaction requires. This is why some old carbonated buildings in arid climates still have sound reinforcement after 60 years, while structures in humid coastal environments with identical cover depths show severe deterioration in 20.

The bottom line is that Corrosion Of Steel In Concrete is predictable if you respect the chemistry and measure what's actually happening rather than what the spec assumes. The tools exist. The methods are standardized. The failures usually come from cutting corners on cover depth, curing, and surface prep — the unglamorous details that determine whether a structure makes it to 50 years or needs a major intervention at 20.