Concrete spall repair and restoration field guide

Spall restoration, and the rule that the cause comes first

Spall restoration starts with one rule: find and fix what caused the damage, or the repair fails. A spall is a piece of concrete that has broken or flaked off the surface, usually because the reinforcing steel underneath corroded and the rust pushed the cover off. Patch the hole without dealing with the corrosion and you have hidden the problem, not fixed it. The steel keeps rusting under your fresh patch and the spall comes back, often within a few years.

The work has a fixed order. Find the cause. Find the full extent of the damage, including the delaminated concrete that has not fallen off yet. Remove all the unsound and contaminated concrete, all the way around and behind the corroded bar. Clean or replace the steel. Stop the corrosion from migrating to the edges of the repair. Then rebuild with a material that moves with the parent concrete, cure it, and protect the surface so the same thing does not happen again.

None of those steps is the hard part on its own. The discipline is doing all of them, in order, on every spall, instead of troweling a cosmetic skim over a problem that is still alive under the surface. That is the difference between a repair that lasts twenty years and one that fails before the next inspection cycle.

Why does spalled concrete happen?

Spalled concrete is concrete that has flaked, chipped, or broken away from the surface, and the most common reason is the reinforcing steel underneath rusting. Steel in good concrete is protected by the high alkalinity of the cement paste, which holds a passive film on the bar. Two things break that protection. Chlorides, from deicing salt, seawater, or a chloride-bearing admixture, reach the steel and depassivate it. Or carbonation, where carbon dioxide from the air reacts into the concrete over years and drops the pH, takes the protection away from the front face inward.

Once the bar starts to corrode, the rust itself does the damage. Rust occupies several times the volume of the steel it came from, so it expands and pushes outward against the cover. That pressure cracks the concrete along the line of the bar, the cracks connect, and a plate of cover lets go. Thin cover spalls first, which is why cover is the variable that decides how long a structure lasts. That whole mechanism is the subject of the rebar placement and cover guide, and it is worth reading alongside this one.

Corrosion is not the only cause. Freeze-thaw breaks down concrete that is saturated and not properly air-entrained, scaling the surface off in layers. Alkali-silica reaction, ASR, is an internal expansion between reactive aggregate and the pore solution that maps and cracks the concrete from inside. Overload and impact spall mechanically, at edges and bearing points. And a leaking joint funnels water and chloride to one spot and spalls it ahead of everything around it. Before you repair anything, you name which of these you are looking at, because the repair is different for each.

Why do most concrete patch repairs fail?

Most concrete patch repairs fail for a reason that has nothing to do with the patch material. They fail because the cause was never fixed. If active corrosion or chloride-contaminated concrete is left in place, the steel keeps corroding the day after the patch cures, and the spall comes back. The material did its job. The diagnosis did not.

Poor surface preparation is the next one. A patch is only as good as its bond to the substrate, and a weak, dusty, or unsound surface gives the new material nothing to grab. Feather edges fail too. Where a patch tapers to nothing at the perimeter, that thin edge is brittle, debonds, and breaks away, which is why the edges get saw-cut square.

Then there is the failure that surprises people, because the patch itself is fine. A ring of fresh corrosion appears in the parent concrete right around the repair, and a new spall forms a few inches outside the one you just fixed. That is the incipient anode effect, also called the ring anode or halo effect, and it is common enough that on chloride-laden structures you plan for it before it happens. The next section is how it works.

What is the ring anode effect?

The ring anode effect is new corrosion that forms in the parent concrete just outside a fresh patch, and it is one of the main reasons spot repairs on chloride-laden concrete fail within a few years. It also goes by incipient anode effect and halo effect. The names all describe the same electrochemistry.

When you cut out the spalled area and replace it with clean, chloride-free repair material, you change the balance of the corrosion cell. The steel in the new repair becomes relatively cathodic, which is to say protected. The steel in the surrounding chloride-contaminated concrete, which was cathodic to the old spall before, is now anodic by comparison, so corrosion shifts to the perimeter. You fixed one spot and started the next one, sometimes within months, often within five years.

This is why removing the chloride-contaminated concrete and cleaning behind the bar matters so much, and why on salt-exposed structures the repair includes corrosion protection at the perimeter rather than just material in the hole. Galvanic anodes set around the edge of the patch are the common answer, covered further down. Ignore the ring anode on a parking deck or a marine structure and you have signed up for a repeat visit you could have designed out.

How do you assess spalled concrete before repair?

You map the full extent of the deterioration, not just the part that fell off, and you find the cause, before you remove a thing. The damage you can see is always smaller than the damage that is there.

Start by sounding the surface. Drag a chain across a deck or tap with a hammer and listen. Sound concrete rings sharp; delaminated concrete sounds hollow or drummy because the plate of cover has separated from the steel but has not dropped yet. Chain drag for decks follows ASTM D4580. Mark every hollow area, because all of it has to come out.

Sounding finds what has already delaminated. To find corrosion that is active but has not cracked the concrete yet, run a half-cell potential survey to ASTM C876, reading the bar's electrical potential against a copper/copper sulfate reference cell on a grid. As a rule of thumb the standard treats potentials more negative than about minus 350 mV as a high probability of active corrosion and more positive than about minus 200 mV as a low probability, with the band between uncertain. The numbers are probabilities, not a pass or fail, so read them with the other data.

Fill in the cause with chloride sampling at bar depth, a carbonation test with a phenolphthalein indicator that stains concrete pink where it is still alkaline and stays clear where it has carbonated, and a cover survey to see how deep the steel sits. The cover survey ties straight back to the rebar placement and cover guide. The point of all of it is the same: define the real boundary and the real cause so the removal goes far enough the first time.

Do you remove concrete behind the rebar?

Yes, when the bar is corroded you remove the concrete from behind it, not just the cover in front. This is the step that gets shortcut most, and skipping it is why patches fail at the steel.

Start the perimeter with a saw cut. Cut a square or rectangular boundary around the repair area to a shallow depth, commonly no deeper than the cover over the steel so you do not nick the bar, and remove to that line. The square cut gives the patch a real edge to key into and kills the feather edge, the thin tapered margin that breaks off and starts a debond. ICRI guidance is explicit about square, near-vertical edges and no feathering for this reason.

Remove all the unsound concrete inside the cut, down to sound, solid material, usually with light chipping hammers so you do not bruise or microcrack the substrate you are keeping. Where the reinforcing is corroded, keep going past the bar. Undercut it and clear the concrete behind the steel so the full circumference is exposed, commonly about 3/4 in of clearance behind the bar, enough to get a blast nozzle or your hand all the way around. There are two reasons. You cannot clean the back of a bar you cannot reach, and the concrete pressed against the back of a corroding bar is the most chloride-contaminated, most aggressive concrete in the whole repair. Leave it and you have buried the cause inside the fix.

Then prepare the substrate to take the new material. ICRI ties this to a concrete surface profile, CSP, with higher numbers being rougher, and most repair mortars want a roughened, sound profile rather than a smooth saw-cut face. Abrasive blast, hydroblast, or scarify the bonding surface, then get the dust off. The substrate the manufacturer's data sheet assumes is clean, sound, and roughened. The one you start with is none of those until you make it so.

Cleaning and assessing the reinforcing steel

Once the steel is exposed all the way around, clean it to bright, bare metal. Abrasive blasting is the standard way, taking the bar back to roughly the near-white metal condition so no rust scale, no flaking, and no chloride film is left on the surface. Needle guns and wire brushing can work on small areas but rarely get the back of the bar as clean. Rust left on the steel is both a bond problem and a corrosion problem you are sealing in.

While the bar is clean, assess section loss. A lightly rusted bar that has lost no real diameter goes back to service as is. A bar that has lost meaningful cross-section, pitted or necked down, no longer carries the load it was designed for, and that is an engineer's call, not the crew's. The thresholds for supplementing or replacing steel come from the engineer of record and the project specification. When steel is added, it is lapped and tied to sound bar per the development and lap rules, which is the same territory as the rebar placement guide.

The coating question comes up every time. Coating the cleaned bar with a cementitious or epoxy corrosion-inhibiting coat is common, and on some details required, but there is a real argument against a continuous film-forming coat on the bar inside a patch. If it holidays, or the coating ends partway along the bar, you can set up a small anode at the gap and concentrate corrosion right there. Many specs now favor a cementitious, alkaline-rich bonding and protection coat over a film-forming epoxy for exactly that reason, or rely on galvanic protection instead. Follow the spec and the manufacturer; do not freelance the coating.

The substrate: saturated surface dry and the bonding-agent question

The substrate has to be sound, clean, roughened, and at the right moisture before the repair material goes on. For cementitious repair materials, that moisture target is saturated surface dry, SSD: the pores of the concrete are full of water but the surface has no standing film. Saturate the substrate ahead of placement, often for hours, then let the free water disappear before you place. A dry substrate pulls the mix water out of the repair material and starves the cure right at the bond line, which is where you can least afford a weak, porous layer.

The bonding-agent question is where field practice and old habit part ways. The traditional move was to brush a bonding agent onto the substrate every time. The modern view is that with a properly prepared, sound, roughened, SSD substrate, many polymer-modified repair mortars bond well without a separate bonding agent, and a neat portland cement slurry scrubbed into the surface immediately ahead of the mortar works as well as a proprietary product for most cementitious repairs.

The real risk with a bonding agent is timing. Apply it, let it skin over and dry before the repair material goes on, and it stops being glue and becomes a bond breaker, the exact opposite of what you wanted. So the rule is simple. Follow the repair material manufacturer's data sheet, because the bond system is part of the tested system. If the data sheet calls for a bonding coat, use it wet on wet. If it calls for SSD and no bonding agent, do not add one because it feels safer.

Choosing the repair material

Pick the repair material to match the parent concrete, not just to hit a strength number. The property that wrecks more patches than low strength is dimensional incompatibility. A repair material that shrinks more than the concrete around it, or that has a much higher elastic modulus, builds up stress at the bond line and cracks or debonds even though the material itself tested strong. You want the patch to move with the slab, not fight it.

Polymer-modified cementitious mortars are the common choice for that reason. The polymer lowers permeability and improves bond, and these mortars can be formulated with low shrinkage and an elastic modulus close to the substrate, so they share load and movement instead of concentrating it. Match the thermal expansion too, because a repair that expands at a different rate than the parent shears the bond every time the temperature swings.

Match the material to the placement, because the same hole can want a different product depending on where it is. A large horizontal area can take a form-and-pour or a flowable micro-concrete. A vertical or overhead patch needs a material that holds its shape and resists sag, built up in lifts. A big area on a wall or a column can be shotcrete, sprayed and built up fast. Strength matters, but on a durability repair the bigger questions are low shrinkage, low permeability, and a modulus near the parent. Read the data sheet for the placement method it was tested for, and do not use a horizontal patch material overhead because it was on the truck.

Placement methods: form-and-pour, hand-applied, shotcrete, and dry-pack

The placement method follows the geometry and the size of the repair. Each method has a material range and a technique that goes with it.

Form-and-pour, or form-and-pump, suits larger repairs and full-depth work. You build a form, then pour or pump a flowable repair concrete or micro-concrete into it, with the form detailed so the material fills the cavity and vents the air. It gives a dense, well-consolidated repair and is the go-to for columns, beam soffits with enough depth, and large vertical areas. Dry-pack, a very stiff, low-water mortar rammed into place in thin layers, is the opposite end. It suits small, deep, confined holes like form-tie pockets and tight cavities where consolidation is by hand pressure.

Hand-applied, trowel-placed mortar covers most spot repairs on walls, slabs, and overhead. The skill on vertical and overhead is the material and the lifts. Use a repair mortar formulated to hang without sagging, place it in lifts within the thickness the product allows per pass, and scratch each lift so the next one keys to it. Overhead is unforgiving. Too thick a lift sags and debonds under its own weight before it sets. Shotcrete, wet-mix or dry-mix sprayed at the surface, builds up large areas quickly and consolidates on impact, which is why it dominates on big facade and parking-structure repairs, but it is a trade of its own and the nozzle technique decides whether it comes out dense or full of voids.

Whatever the method, the cavity has to be prepared and the steel handled the same way first. The placement method does not change the diagnosis or the prep. It only changes how the material gets into the hole.

Corrosion protection: inhibitors, galvanic anodes, and cathodic protection

On chloride-laden concrete, the repair includes something to stop corrosion from migrating to the patch perimeter, because clean material in the hole does not address the steel just outside it. There are three tools, and they are not mutually exclusive.

Galvanic, or sacrificial, anodes are the common answer to the ring anode effect. These are small zinc elements in an activating mortar, tied to the reinforcing steel and set around the perimeter of the repair. Zinc is more electrochemically active than steel, so it corrodes preferentially and feeds a small protective current to the bar, holding the surrounding steel cathodic so a new anode does not form at the patch edge. Discrete embedded zinc anodes are placed at the perimeter for exactly this, and ACI and ICRI both publish guidance on installing them.

Migrating corrosion inhibitors are surface-applied or admixed chemicals that move to the steel and slow the corrosion reaction. They are used on broader contaminated areas, often together with patching, where you cannot remove all the chloride. Cathodic protection is the heavy version: an engineered system, either galvanic over a whole element or impressed-current driven by a power supply, that protects steel across a large structure rather than one patch. It is a designed system with monitoring, not a jobsite add-on.

The choice scales with the exposure. A dry interior slab spall often needs none of this. A parking deck, a bridge, or a marine structure usually needs at least perimeter anodes, and a badly chloride-loaded structure may need area cathodic protection designed by a specialist. The engineer and the corrosion specialist size it; the crew installs it where the detail says.

Curing the repair

Cure the repair, and start sooner than you would on a big pour. A patch is small, has a large surface area relative to its volume, and is often on a vertical or overhead face exposed to wind and sun, so it loses moisture fast and can dry out before the cement has hydrated. A repair that dries early gains less strength, shrinks more, and is more likely to debond at the edge. The mechanics are the same as ordinary concrete curing, just less forgiving because of the geometry.

Follow the repair material's curing instructions, which for cementitious mortars usually means keeping the surface wet or covered for the period on the data sheet, with wet burlap, a curing compound compatible with any later coating, or plastic. Polymer-modified mortars sometimes call for a short wet cure followed by air drying so the polymer can form its film, which is the opposite of a long wet cure, so read the sheet rather than assuming. The general curing principles carry over from the mix design side of the work. The difference here is that nobody is watching a small patch, and a small patch left to dry in the sun is the one that fails quietly.

Protective coatings and surface restoration

Once the repair has cured, the last step on a durability job is protecting the surface so the cause does not return. A patch fixes the spot that failed. A protective coating or sealer addresses the whole face that is still taking on chloride or carbonating, including the concrete you did not repair.

The two common families do different jobs. A penetrating sealer, often a silane or siloxane, lines the pores and makes the surface water-repellent so chloride-laden water does not soak in, while still letting the concrete breathe. A film-forming coating, such as an elastomeric or an anti-carbonation coating, lays a barrier on the surface that slows carbon dioxide and water ingress and can bridge fine cracks. On a carbonating facade an anti-carbonation coating buys years; on a salt-exposed deck a penetrating sealer or a traffic-bearing membrane keeps the chloride out.

There is an aesthetic side too, and it is not just vanity. A patch rarely matches the color and texture of aged concrete, and on a visible facade a coating or a finish unifies the appearance so the repair does not read as a scar. Match the coating to the exposure and to the repair material, confirm the surface is dry enough and cured enough to accept it, and check that the coating is compatible with any corrosion protection system already in place.

Do you inject a crack or route and seal it?

Whether you inject a crack or route and seal it comes down to two questions: is the crack structural, and is it still moving. The two repairs do opposite things, and using the wrong one wastes the work.

Epoxy injection bonds the concrete back together. You inject a low-viscosity structural epoxy into a dormant crack under pressure, it fills the full depth, and once cured the section behaves monolithically again, often stronger at the crack than the surrounding concrete. That is the right repair for a structural crack that is not moving, where you want to restore the load path and keep water out. It is the wrong repair for a crack that is still working, because the rigid epoxy just cracks again next to the old line.

Route and seal is for non-structural cracks and for cracks that move. You widen the crack into a clean groove, often a V or a U, and fill it with a flexible sealant that can stretch as the crack opens and closes with temperature and load. It keeps water and chloride out without pretending to restore structural continuity. For active cracks you sometimes build a true movement joint instead of fighting the movement.

The line between structural and non-structural is the engineer's to draw, not the crew's. If a crack might affect capacity, or if you cannot tell whether it is still moving, that is a call for the engineer of record before anyone picks a method.

Structural repairs and the engineer of record

Some spall repairs are cosmetic and some are structural, and the difference decides who is in charge. The moment the deterioration touches load-carrying capacity, section loss in the reinforcing, deep removal in a beam or a column, anything that reduces the structure's ability to carry its loads, it stops being a crew decision and becomes the engineer of record's.

The engineer sets the things a crew cannot judge from the ground. How much steel section loss is acceptable before bars must be supplemented or replaced. Whether the member needs shoring before concrete is removed, because removing a chunk of a working beam without shoring can overload what is left and bring the consequence forward. How deep the removal can go and in what sequence. And the repair detail itself: the material, the reinforcement, and how load transfers back into the repair.

ACI 562 is the code written for this, the requirements for assessment, repair, and rehabilitation of existing concrete structures that an engineer works to when a repair carries load. ACI 546 is the companion repair guide with the methods. Neither replaces the engineer of record's judgment on the specific structure. The blunt version: if you are not sure whether a repair is structural, treat it as structural and get the engineer involved before you start removing concrete. Removing load-bearing concrete on a hunch is how a repair becomes a collapse.

Testing the finished repair

Test the finished repair the same way you found the damage, plus a quantitative check where the spec calls for one. Sound the cured patch with a hammer, listening for the hollow, drummy note that means it did not bond or has a void behind it. A patch that sounds hollow has failed regardless of how it looks, and it is cheaper to find that now than after the coating is on.

For a measured acceptance, the common test is a direct-tension pull-off to ASTM C1583. A core is drilled partway through the repair into the substrate, a steel disc is glued to the top, and a tester pulls it straight up until something breaks. The result is the tensile stress at failure, and where the break happens tells you as much as the number. A break down in the substrate means the bond beat the parent concrete; a clean break at the bond line means the bond was the weak point. The project specification sets the acceptance value and the number of tests, because the right figure depends on the material and the application.

On a larger program, half-cell surveys and chloride sampling get repeated over time to confirm the corrosion protection is working and a new ring of activity has not started around the repairs. Testing is not a formality at the end. It is how you know the work you cannot see did what it was supposed to.

Where this work lives: garages, bridges, facades, and industrial slabs

This work concentrates where chloride and water concentrate. Parking structures are the classic case. Cars track in deicing salt all winter, it dissolves, runs to the low spots and the joints, and soaks into the deck, so parking garages spall on a schedule. The repairs there are almost always corrosion-driven, almost always need perimeter anodes or broader cathodic protection, and live or die on keeping water and chloride off the deck afterward with sealers or membranes.

Bridges are the same physics at a larger scale, with deicing salt on the deck and, on coastal structures, salt spray and tidal exposure on the substructure. Bridge deck repair is where chain drag, half-cell surveys, and chloride mapping were largely worked out, and where the ring anode effect got learned the hard way on early patch programs. Facades and balconies carbonate and spall from rain and carbonation rather than deicing salt, and there the falling-concrete hazard makes the assessment urgent, because a spall off a tenth-floor balcony is a life-safety problem before it is a durability one.

Industrial and data center slabs are a different animal. The concern is less chloride and more flatness, joint integrity, and surface durability under heavy wheel loads and tight equipment tolerances. A spalled joint arris under a loaded pallet jack or a server-row cart is both a tripping and an equipment problem, and the repair has to take traffic and meet a flatness the operation actually runs on. The diagnosis and the prep are the same. What changes is the exposure you are protecting against and the acceptance the owner cares about.

What to document

Document each repair so the next person can see what was wrong, how far you went, and what you put back. On a repair program of any size, the record is what proves the corrosion was addressed and not just covered, and it is what a later survey gets checked against.

Capture the location and the cause, the area and depth of removal, the condition of the reinforcing and whether any was supplemented or replaced, the repair material and the placement method, the corrosion protection installed, and the curing and any coating. If a repair was structural, record the engineer's direction and the detail used. The point is the same as the rest of the job. When the question comes back, an untraceable repair gives you nothing to stand on, while a recorded one shows the corrosion was treated rather than buried.

Common mistakes

• Patching the spall without finding and fixing the corrosion or other cause that produced it.
• Leaving chloride-contaminated concrete behind, especially the concrete pressed against the back of the bar.
• Feather-edging the patch instead of saw-cutting a square, sound perimeter.
• Not removing behind a corroded bar, so the back of the steel never gets cleaned.
• Leaving rust on the reinforcing instead of cleaning it to bright, bare metal.
• Skipping perimeter anodes on chloride-laden structures and getting a new spall ringing the patch.
• Using a repair material that shrinks or stiffens differently than the parent concrete, so it debonds or cracks.
• Letting a bonding agent skin over and dry before placement, turning it into a bond breaker.
• Letting a small patch dry out in sun or wind instead of curing it.
• Treating a structural repair as cosmetic and removing load-bearing concrete without the engineer or shoring.

Field checklist

Standards and references

The standards here come from several bodies, each governing a different part of the work, and the section numbers shift between editions, so confirm them against the version in force on your project.

ICRI, the International Concrete Repair Institute, publishes the technical guidelines for the repair process itself: surface preparation and the concrete surface profile, CSP, for receiving repair material, and the methods bulletins for spall repair and anode installation. ACI covers the code and the engineering side. ACI 562 is the code for assessment, repair, and rehabilitation of existing concrete structures, and ACI 546 is the concrete repair guide that lays out methods. ACI 318 and its durability and cover provisions sit behind the cause, which the rebar placement and cover guide and the mix design and water-cement ratio guide both work through.

ASTM gives the test methods. ASTM C876 is the half-cell potential survey for corrosion. ASTM D4580 covers chain-drag sounding for delamination on decks. ASTM C1583 is the direct-tension pull-off test for bond and surface tensile strength. Material standards and the steel-cleanliness references for abrasive blasting round out the set. Above all of it sits the engineer of record on any structural repair and the repair material manufacturer's data sheet on the material, the bond system, and the cure, because the tested system is the one that carries the warranty.

Units, terms, and conversions

Concrete repair borrows vocabulary from corrosion science, materials, and structural work, so the same idea shows up under different names across a condition report, a spec, and a product data sheet.

Cover and removal depths are in inches in US practice and millimeters in metric. Steel cleanliness is given as a surface grade against the abrasive-blasting standards. Half-cell potentials are in millivolts against a copper/copper sulfate reference cell. Pull-off bond strength is in psi or MPa. Chloride content is reported as a percentage by weight of cement or of concrete, or in pounds per cubic yard, which is the figure a corrosion threshold is usually quoted against.

Spall

A fragment of concrete that has broken or flaked off the surface, usually over corroding steel

Delamination

A separation within the concrete, often at the reinforcing, that sounds hollow before it falls

Incipient / ring / halo anode

New corrosion forming in the parent concrete around a patch as the clean repair turns the surroundings anodic

Carbonation

Loss of concrete alkalinity as carbon dioxide reacts in over time, which depassivates the steel

SSD

Saturated surface dry, the substrate moisture state with pores full of water but no surface film

CSP

Concrete surface profile, the ICRI roughness scale for a prepared bonding surface, higher numbers rougher

Half-cell potential

The reinforcing steel's electrical potential against a reference cell, read to estimate corrosion activity

Galvanic anode

A sacrificial zinc element tied to the steel that corrodes preferentially to protect the bar

EOR

Engineer of record, who controls any repair that affects structural capacity

Related tools

Calculators and readiness checks for this work

Compare your options

FAQ

People also ask

• What is the minimum concrete cover for rebar?
• What is a concrete mix design?
• Why does parking garage concrete spall?
• Why does rebar corrode in concrete?
• What is a concrete overlay?