Concrete crack injection, explained properly

First, the confusion that costs people money: are you here about a sunken slab?

The phrase resin injection names two different trades that share a chemical and almost nothing else. One injects into the concrete to stop water. The other injects into the subgrade - the ground underneath a slab or footing - to compact weak soil, fill voids and lift the structure back toward level. Both use polyurethane. Only one of them is waterproofing.

Ground improvement will not stop a single drop of water and does not seal a crack. Equally, nothing on this page will lift a driveway that has dropped 30 mm. If the symptoms are a stepped driveway, doors and windows binding, a hollow drummy sound underfoot or a level difference across a floor with no water involved, a ground engineering firm is the right call and they do that work well.

One thing worth knowing before ringing either trade: re-levelling does not fix the cause of the settlement. Reactive clay movement, a leaking sewer or stormwater line washing fill out from under the slab, or tree roots drawing moisture out of the subgrade all have to be corrected first, or the slab settles again. And where the work amounts to underpinning a footing, engineer involvement and building approval requirements come into it.

The five resin families, and what each one is actually for

There is no best resin. There is one resin that matches the water state, the movement state and the drying state of a specific crack, and four that do not. Choosing wrongly is the single commonest cause of a failed injection repair in this country, and it is almost always chosen wrongly in the same direction: whatever was already on the truck.

Two figures worth carrying into any conversation with a contractor. Acrylate gel runs at 10 to 50 mPa.s, which is roughly the viscosity of water, and that is why it permeates soil and hairline cracks that no polyurethane will enter. And the accepted international practice for epoxy - ACI 503.7 and ICRI RAP Bulletin 1 - calls for a low-viscosity epoxy of 500 cP or less on cracks of 0.3 mm and finer. Neither is an Australian code requirement; both are the body of practice that a competent injection contractor works to. The same goes for EN 1504-5, the European product conformity standard that nearly every injection resin on the market declares against, and its classification of injection products as force-transmitting (F), ductile filling (D) or water-swelling (S). It is a product specification, not an Australian rule, and a contractor who says otherwise is bluffing.

• DRY, dormant, and the intent is strength: epoxy.
• DAMP - staining, no visible movement of water: hydrophilic PU or acrylate gel.
• WEEPING - beads forming, slow trickle: hydrophilic PU, or hydrophobic if that element dries out seasonally.
• RUNNING - continuous flow: hydrophobic PU on a fast accelerator dose.
• GUSHING - flow under head: hydrophobic PU on a very fast gel, usually after the flow is knocked down with oakum packing or a hydraulic cement plug.
• Then two more questions before committing. Does it move? If yes, no epoxy. Will it dry out permanently? If yes, no hydrophilic.

What actually happens on site, in the order it happens

1. Find where the water gets in, before anything gets drilledThe crack you can see inside is very often not where water enters outside. Water tracks laterally between a membrane and the concrete, and between a structural slab and its topping, then exits at the weakest point - frequently metres from the entry. Before anything is quoted we rule out the non-injection causes: blocked or absent ag lines (the perforated subsoil drain behind a wall) and weep holes, missing or reversed falls, a downpipe discharging at the footing, a failed sealant joint, a service that can be pressure-tested, and condensation. We inspect after rain, and during rain where access allows. If we cannot explain how the water gets in, we cannot warrant that injection will stop it.
2. Classify the crack, and fit tell-tales if it is ambiguousWidth measured with a comparator card or feeler gauge, plus length, orientation, whether it passes through the full element, and whether it is active or dormant. AS 3600 treats cracking as a serviceability condition with limits broadly in the 0.1 to 0.4 mm range depending on element type and exposure classification, so a fine crack is not automatically a defect. Where movement is uncertain, tell-tales (simple gauges fixed across the crack with a scale on them) get fitted and read over 4 to 12 weeks across a temperature and moisture cycle rather than injecting on the first visit. It feels slow. It is much cheaper than injecting a moving crack twice.
3. Rate the flow, then choose the resinUsing the dry / damp / weeping / running / gushing rating above, with the movement and drying questions applied after it. Cause matters as much as flow: a plastic or drying-shrinkage crack, a thermal crack and a flexural crack look similar on a wall and want completely different responses. A structural or corrosion-driven crack is not a waterproofing job at all.
4. Scan, then set out and drillGPR or a good cover meter on any structural element. On a post-tensioned slab, scanning plus written engineer approval before a bit touches the concrete - cutting a tendon is not a recoverable mistake. Holes go in at 45 degrees to the face, alternating left and right of the crack in a stitch pattern so they cross the crack plane from both sides. Offset from the crack line by roughly half the element thickness so the 45-degree hole meets the crack at mid-depth: on a 250 mm wall, about 125 mm off the line. Drill to two-thirds or three-quarters of the element thickness and never break through the far face - break through and all pressure is lost into the ground or into somebody's storeroom. A 13 mm bit paired to a 13 mm mechanical packer is the workhorse. Drilling concrete releases respirable crystalline silica, workplace exposure standard 0.05 mg/m3 over 8 hours, so on-tool extraction or wet drilling is not optional in an enclosed basement or lift pit.
5. Set packer spacing to the element, not to habitWorking range is 150 to 400 mm centres. Tight, 150 to 200 mm, for hairline cracks and for diffuse paths such as core-filled besser block. Wider, 300 to 500 mm, for wide open cracks and cold joints. A widely used rule is that spacing should not exceed the thickness of the element, and roughly half the element thickness is a safe default on critical work. Whichever number is chosen, the test is empirical rather than theoretical: if resin does not travel from one packer to the next, the spacing is too wide or the crack is not connected.
6. Flush with water and prove the crack is connectedClean water pumped through each packer before any resin, working bottom upward on a vertical crack. It does three jobs at once. It proves the crack is hydraulically connected between packers - if water does not surface at the next packer, the set-out is wrong, and drilling more holes is far cheaper than wasting resin finding out. It clears laitance, drilling dust, efflorescence (the white salt bloom left behind when water evaporates out of concrete) and calcite that would otherwise block penetration. And it pre-wets a dry crack so a moisture-cured polyurethane has water to react with. Do not water-flush ahead of an epoxy injection unless the crack can then be dried with oil-free compressed air. Never flush solvent into the ground.
7. Inject bottom to top, and move on only at refusalOn a vertical crack, always start at the lowest packer and work upward. Resin then displaces water up and out of the crack instead of trapping it, the lower section fills first so the seal builds off a sealed base, and resin appearing at the packer above is direct confirmation of travel. On a horizontal crack or joint, work consistently from one end to the other. A single gusher gets killed first with a fast-set so it stops washing resin out of the neighbouring section, then the sequence resumes. Move to the next packer when resin shows at that packer, or at refusal - pressure climbs, holds, and nothing more goes in. Cap each packer on leaving it.
8. Keep the pressure down and watch the element, not the gaugeStart at the bottom of the band for that element and escalate only if the crack genuinely refuses to take. The governing limit is the structure, not the pump: a single-component PU pump rated to 200 bar will happily destroy a 150 mm precast panel. Bands by element type and the four signs of over-pressure are set out below.
9. Twenty-four hours, then cut back and make goodPackers stay in a minimum of 24 hours, longer in cold weather or where a slow-cure product was used. Pulling them the same day to finish in one visit lets water back in before the seal has developed. Shafts are snapped or cut 10 to 20 mm below the finished surface, the recess filled with a non-shrink hydraulic plug or polymer-modified repair mortar, then ground or trowelled flush. Cured polyurethane is very difficult to remove from a face, so surfaces get masked and protected beforehand rather than cleaned afterwards. Reinstating render, paint or a wall lining is a separate scope item and should be priced as one.
10. Verify, then hand the record overObservation for 24 to 48 hours minimum and, where the program allows, across a rain event. Moisture meter readings at the same marked points before and after. Flood testing wherever the geometry permits. Then the paperwork: photographs before and after, taken from the same positions, and a written description of what we found and what we did. That record is what settles an argument in two minutes if water reappears in eighteen months.

Pressure bands by element, and the four signs you have gone too far

These are indicative field ranges, not code requirements, and they are ranges for a reason. The correct starting pressure is the bottom of the band. Pressure gets escalated only when a crack will not take resin at the low end, and on slender, prestressed or post-tensioned elements an engineer should nominate the maximum in writing before the pump is switched on.

Mechanical packers or surface ports - the difference that decides the job

A packer is the fitting installed so a hose can be connected and resin pumped in. It carries a one-way valve so resin goes in and nothing comes back out. There are two families of them and the difference is not a preference, it is what separates a repair from a wasted day.

A mechanical packer has a steel or plastic body with a rubber sleeve that is compressed against the sides of the drilled hole by tightening a nut. It seals inside the wall body, not on the face. It takes high pressure and it delivers resin into the crack at mid-depth rather than onto it. This is the correct choice for effectively all remedial waterproofing injection.

A surface port is glued or paste-fixed to the face of the wall over the crack, with a stiff epoxy paste capping the crack between ports. It is a low-pressure device. It has genuine uses: thin sections that cannot safely be drilled, structural epoxy work at 10 to 50 psi, and cracks too fine or too shallow to drill without doing more harm than good.

Here is the blunt version, because it is the single most useful thing a homeowner or a committee can know before reading quotes. Surface ports on a wet wall will not adhere. The epoxy paste capping the crack needs a clean, dry substrate to bond to; on a wall with water coming out of it, the seal never keys, and the first few bar of pressure lift it straight off the face. The resin ends up on the wall rather than through it. That is what a cheap quote on an actively leaking basement crack usually is, and it is why that repair typically fails inside one wet season.

Verification: the proof to demand before you pay

Injection is concealed work. Once the packers are cut off and the holes are patched, a good repair and a bad repair look identical, and the difference only surfaces in the next storm. That asymmetry is exactly why this trade has a reputation problem, and the fix is documentation rather than assurances.

• Resin arriving at the next packer along, seen and photographed during injection. That is the only real-time evidence the crack is filled between two points.
• Photographs before and after, taken from the same positions, together with a written description of what was found and what was done.
• Moisture meter readings at the same marked points before and after, so there is a number rather than an opinion.
• Twenty-four to 48 hours of observation as a minimum, and where the program allows, a rain event before sign-off.
• Where the geometry permits, a water test: a lift pit dewatered and watched, a tank filled and held and tested to the leakage criteria in AS 3735, a planter box flooded.
• For structural epoxy where penetration has to be proven, coring across the repaired crack is the only genuine verification. It is destructive and expensive and should only happen where an engineer specifies it.

What it costs in South East Queensland, and what moves the number

The published prices in this market are incoherent - some contractors quote per crack, some per linear metre, some per square metre and some per job, which makes three quotes genuinely impossible to compare without normalising them first. The bands below are indicative South East Queensland ranges excluding GST. They are a guide for budgeting and for reading someone else's quote, not a quote.

The single figure most operators will not put in writing is the first one. Site attendance covers travel, set-up, the pump and plant on the truck and a minimum block of labour, and almost no legitimate injection contractor can attend, set up, inject and make good below it. A short single crack therefore prices as a minimum job rather than at a linear metre rate, and long repetitive runs on one site drop toward the bottom of the band because there is one mobilisation, one set-up and one wall to learn.

Fifteen times we will tell you not to inject

This is not a disclaimer. It is the most commercially useful page section on the site, because every one of these situations gets injected by somebody every week in South East Queensland, and every one of them comes back.

• Ongoing structural movement. Monitor with tell-tales, get an engineer to determine the cause, stabilise or accommodate the movement, then seal. Epoxy will hold and the concrete will crack alongside it; foam will be torn open.
• Active settlement. A footing still moving on reactive clay, poorly compacted fill, a leaking service washing out subgrade or tree root moisture extraction. Geotechnical and structural problem first.
• A membrane at end of life. Injection cannot restore a membrane that has debonded, blistered, split or UV-degraded across a whole deck, balcony or planter box. That is removal and replacement to AS 4654.1 and AS 4654.2 externally, or AS 3740:2021 for internal wet areas. Injecting here buys 6 to 18 months.
• Drainage. Blocked or absent ag lines, blocked weep holes, no drainage cell, no fall, a downpipe discharging at the footing. This sits behind a very large share of injection enquiries, and sealing the structure without restoring drainage raises the head against it.
• Rising damp. A tide mark to roughly a metre with salt bloom and perished plaster is capillary moisture through porous masonry, treated with a silane or siloxane chemical damp-proof course. Resin crack injection will not touch it.
• Corrosion-driven cracking, commonly called concrete cancer. Cracking along the line of the reinforcement with rust staining, spalling or drummy areas. Injecting seals the problem in while the steel keeps expanding. That is a concrete repair scope.
• Designed movement joints. Filling one defeats its purpose and transfers the movement into the concrete beside it. Backing rod and a movement-capable sealant.
• Post-tensioned slabs without GPR scanning and written engineer approval. Never drill.
• Condensation misdiagnosed as a leak. Damp on cold surfaces in still, humid, poorly ventilated spaces with no correlation to rain. Ventilation, dehumidification, insulation.
• A leaking pipe or service. Pressure-test before injecting. Quick to rule out and expensive to get wrong.
• General porosity with no defined line. Water through the body of a wall or slab across a broad area gives the resin nothing to grip. That is negative-side tanking, a drained cavity, curtain injection behind the wall, or excavation.
• Sealing water into somewhere it cannot escape. A planter box with a blocked outlet, a saturated cavity, an under-slab void. Always ask where the water will go instead; if the answer is nowhere, that is a design problem.
• Heritage and soft lime-mortar masonry. High-pressure injection destroys the mortar bed and blows joints irreversibly.
• Where injection is being asked to do a strengthening job. Epoxy restores monolithic action across a crack; it does not add capacity. An under-designed, overloaded or section-lost member needs a strengthening design.
• Where the source can simply be redirected. A downpipe extension, a re-graded path, a new fall or a diverter is cheaper, more reliable and more permanent than any resin.

The situations this method actually gets used on

Each of these has its own page with its own numbers, because the method is the same but the constraints, the productivity rates and the compliance obligations are not.

• Basements and below-ground walls. The kicker joint - where the wall meets the floor slab - is usually the real path even when a crack higher up is what you can see. Roughly 8 to 15 linear metres of joint per crew per day.
• Garage and undercroft walls on SEQ hillside blocks. Check the driveway fall and the downpipe discharge before anyone quotes injection.
• Retaining walls, core-filled block and reinforced concrete. Packers at 150 to 250 mm on blockwork because the water paths through mortar beds are diffuse rather than linear. Drainage first. Any lean, bulge or step cracking stops the job and gets an engineer.
• Basement car park slabs and soffits. Roughly 15 to 30 linear metres of joint per crew per day in accessible conditions; far less at height or with traffic management. Stalactites and calcite deposits mean chronic flow, not a recent event.
• Lift pits. A confined space under AS 2865 with a legal permit, atmospheric testing, standby person and rescue plan attached to it, and buffer or hydraulic oil that has to be degreased or the repair simply fails.
• Planter boxes and podium garden beds. Hydrophilic or acrylate because the substrate never dries out, and usually an interim measure while a membrane replacement is planned honestly.
• Leaking cold and construction joints. Packers at 250 to 400 mm centres, and a joint takes considerably more resin than a crack of the same length.
• Form tie and tie-bolt holes. A straight pipe through the wall and the most direct water path in the building. 20 to 50 holes per crew per day - count them at quoting stage, because a basement wall can carry hundreds.
• Curtain injection behind the wall, where there is no single injectable crack. Acrylate gel on a 250 to 500 mm grid at low pressure, priced from a trial panel per point or per kilogram, never as a lump sum.
• Cracked slabs that leak water, as distinct from slabs that have sunk.
• Concrete water tanks, fire tanks and pools. AS/NZS 4020 certification for every product touching potable water, and AS 3735 to establish whether the loss is even outside the permissible rate.
• Active leaks during a storm event. Fast-gel hydrophobic plus a hydraulic plug or oakum to kill flow, then a proper repair when the weather allows.

What it costs

Indicative only, and excluding GST. Build a line-by-line estimate for your own job with the price book, or read the working behind the rates in the 2026 cost guide.

Questions we actually get asked

How far apart should the drill holes be, and does the spacing genuinely matter?

It matters more than the choice of resin on a lot of jobs. The working range is 150 to 400 mm centres: tight at 150 to 200 mm on hairline cracks and on diffuse paths like core-filled block, wider at 300 to 500 mm on open cracks and cold joints. A widely used rule is that spacing should never exceed the thickness of the element, and about half the thickness is a safe default on anything critical. The real test is not arithmetic though - it is whether resin travels from one packer to the next during injection. If it does not, the spacing is too wide or the crack is not connected at that point, and the honest response is to drill more holes rather than pump more resin.

Someone already injected this crack and it is leaking again. Can it be re-injected?

Usually yes, and the first job is working out why it failed rather than repeating it. Three causes cover nearly all of them: the diagnosis was wrong and the water is not coming through that crack at all; surface ports were glued to a wet wall so the resin went on the concrete rather than through it; or the resin chosen did not suit the water and movement state of that crack. Ask the original contractor which resin was used and to see their before and after photographs. If there is nothing to show, that is informative in itself. Re-injection usually means drilling a fresh set of holes on a different offset so the new packers intersect the crack in undisturbed concrete rather than into the old, partly filled path.

Two quotes for the same crack are three times apart. What is the difference actually buying?

Nearly always, they are quotes for different work. A cheap quote is commonly surface ports glued to the face with a bead of filler over the crack, injected at 10 to 50 psi, with no flushing and no making good - which on an actively leaking wall puts resin on the concrete rather than through it. The dearer one is angled 13 mm drilling to intersect the crack at mid-depth, mechanical packers that seal inside the wall, a water-flush connection test, injection at 35 bar and up, a return visit at 24 hours to cut back and patch, and a written record. Both are legitimately described as crack injection. Only one of them survives a wet season on a wall under head.

What is the difference between a crack and a construction joint, and why does the joint cost more per metre?

A crack is a fracture through concrete that was meant to be continuous. A construction joint, or cold joint, is the line where one pour of concrete met the next and the two never fully bonded - it is a designed interface that has become a leak path. Joints cost more per linear metre, typically $200 to $500 against $150 to $450 for a crack, for three reasons: they are wider and more open so they take considerably more resin, they are hydraulically connected across their whole length so resin travels further and needs to be chased, and they generally need closer packer spacing at 250 to 400 mm. In a basement or a car park, the joint is the most likely leak path in the building, and the kicker joint at the base of the wall is the single most common of all.

Does the render, paint or tiling have to come off first?

Yes, wherever it hides the crack. You cannot inject what you cannot see, and painted or rendered walls conceal the true extent of a crack, which is how a contractor discovers on day one that a 2 metre crack is actually a 6 metre one. Exposure should be quoted as a separate line rather than assumed, and so should reinstatement - re-rendering and repainting is a different trade to injection and pretending it is included is one of the reasons quotes end up so far apart. On bare concrete in a car park or plant room, the finished job is a line of patched 13 mm holes at 150 to 400 mm centres running at an angle either side of the crack, and that is usually the finish.

Our slab is post-tensioned. Can it still be injected?

Yes, but nothing gets drilled until two things have happened. The slab is scanned with ground-penetrating radar to map the tendons and the reinforcement, and an engineer gives written approval for the drilling positions and nominates a maximum injection pressure. Cutting a post-tensioning tendon releases enormous stored energy and is not a recoverable mistake, which is why this is one of the very few absolute rules in the trade. If a contractor is prepared to drill a car park deck or an apartment slab without asking whether it is post-tensioned, that answers a different and more important question about them.

Is the 24-hour wait before cutting the packers off real, or a way to charge for a second visit?

It is real, and the second visit is a consequence rather than a purpose. The resin gels within minutes but the seal keeps developing for hours afterwards, and pulling a packer out on the same day opens a 13 mm hole straight into the crack you just filled, at the exact moment the plug is least able to hold. Twenty-four hours is the minimum, longer in cold weather or with a slow-cure product. On a remote job the second visit effectively doubles mobilisation, so the program should be designed around it - batching several jobs into one area, or sequencing a larger site so the crew is on site anyway.