Cracked slabs that leak water

Is your slab leaking, or has it sunk?

The word injection covers two unrelated trades and that is the single biggest source of wasted money in this category. Leak sealing injects resin into the concrete, through 13 mm holes drilled at 45 degrees so they intersect the crack at mid-depth, to fill the water path right through the element. Ground re-levelling injects a high-density structural polyurethane or geopolymer into the subgrade beneath the slab, through small 12 to 16 mm tubes, so expansion pressure compacts the soil and lifts the concrete back up.

The materials, the rigs, the skill sets and the outcomes are different. Ground re-levelling will not seal a crack. Leak sealing will not lift a slab by a millimetre. If a contractor offers you both from the same drum, that is a warning.

• Signs it is a leak: water appears during or after rain, a wet line you can point at, white powdery salt bloom (efflorescence) beside the crack, damp carpet or delaminating floor coverings, a drip mark on the ceiling below.
• Signs it has sunk: a step or lip you can feel with your foot, a hollow drummy sound when the slab is tapped, doors and windows binding, a level difference across a crack rather than water through it.
• Signs it is both: a settled slab that has cracked, and the crack now leaks. Order matters here. The cause of the settlement is corrected first, then the slab is re-levelled if required, and only then is the crack sealed. Seal it first and the next movement reopens it.

Crack width, and when water changes the answer

Almost every enquiry opens with a version of the same question: is this crack normal? AS 3600 answers it. The standard treats fine cracking in reinforced concrete as expected behaviour and sets serviceability crack width limits broadly across a 0.1 mm to 0.4 mm band, with the tighter figures applying as the exposure classification becomes more aggressive. The classifications run from A1, a sheltered internal element, through to C1 and C2 for the harshest marine and tidal conditions.

Two practical notes before you measure anything. First, confirm the specific limit for your element and exposure class against the edition of AS 3600 currently called up in Queensland before anyone puts a number in a report. Second, width is only half the question. A dormant 0.5 mm crack in a dry internal slab and a live 0.15 mm crack passing water into a car park below are not the same problem, and the water is what decides.

The four ways water gets through a concrete slab

A slab that leaks is leaking through one of four things, and they need different treatment. Getting this wrong is why repairs come back.

• A discrete crack through the slab. The classic case, and the one injection is built for. Shrinkage and thermal cracks are usually dormant and injectable; flexural cracks caused by load are not a waterproofing job.
• The kicker joint, where the floor slab meets the base of a wall. This is a construction joint, not a crack, and in a basement or undercroft it is usually the real path even when a crack higher up the wall is what you can see. Joints take more resin than cracks and need packers at 200 to 400 mm centres.
• The gap between a structural slab and a topping slab. Water enters at one point on the deck above and travels between the two layers before exiting metres away. On a suspended deck this is a mapping problem before it is an injection problem.
• The body of the slab itself, across a broad area. Porous concrete under hydrostatic pressure passes water over a whole surface with no line to inject. There is nothing for resin to grip, and the answer is a negative-side tanking system, a drained cavity floor or getting the groundwater down with drainage and a sump.

Water coming up through a slab on ground

In South East Queensland this is a garage, an undercroft or a ground-floor storeroom that goes wet during the wet season and dries out by winter. The first thing worth checking costs nothing: where does the surface water go? A driveway falling toward the building, a downpipe discharging at the footing, a blocked or missing ag line (the perforated subsoil pipe in gravel that carries groundwater away) and blocked weep holes are behind a very large share of the enquiries we get. Fixing those is cheaper and more permanent than sealing anything, and it is the first thing we look at.

Where there genuinely is a crack passing water, a slab on ground is thin, so pressure discipline matters. Drilling to two-thirds of a 100 mm slab leaves a 65 mm hole, and the packer has to seal inside that. Injecting under hydrostatic uplift can lift and crack a thin slab, so pressure starts low and only escalates as the crack demands it.

The other slab-on-ground trap is the void underneath. If there is a drainage layer, a washout or poorly compacted fill below, resin will disappear into the ground without ever sealing the face. We set a hard volume stop of 1 to 2 litres per packer. If the packer takes that and nothing appears at the next one along, we stop, and we have a different conversation with you rather than pumping a drum into the subgrade at your expense.

Suspended slabs leaking to the level below

Water showing on a ceiling, a car park soffit or a garage roof is a suspended slab leak, and the visible drip is very rarely above the entry point. Water tracks laterally between the structural slab and the topping, or along the underside of a membrane, and exits at the weakest point, which can be several metres away. Mapping that path is more valuable than any amount of injecting, and it is the step cheap quotes skip.

Stalactites and hard calcite deposits on the soffit are worth reading properly. They mean water has been moving through that path continuously for a long time, not occasionally, and they physically block the crack the resin needs to travel through, so they have to be flushed out first.

The hard stop on suspended slabs is post-tensioning. A post-tensioned slab contains steel tendons still under enormous tension, and they are common in car park decks and apartment slabs. We do not drill one without a ground-penetrating radar scan mapping the tendons and written approval from the structural engineer, and neither should anyone else. Cutting a tendon is not a recoverable mistake.

How a leaking slab crack is actually sealed

1. Map the water path before anything is drilledEstablish where water enters, not just where it appears. Inspect after rain and during it if we can. Rule out the cheap causes first: blocked ag lines and weep holes, a reversed fall, a failed sealant joint, a leaking service that should be pressure-tested, and condensation. If we cannot explain how water gets in, we cannot warrant that sealing will stop it.
2. Classify the crack and decide whether to waitWidth with a comparator card or feeler gauge, length, orientation, and whether it passes through the full slab. Where activity is ambiguous, we fit tell-tales and read them over 4 to 12 weeks across a temperature and moisture cycle. That feels slow. It is far cheaper than injecting a moving crack twice.
3. Scan, then set out at 45 degreesGPR or a cover meter on any structural element. Holes are drilled at 45 degrees, alternating either side of the crack so they cross the crack plane from both directions, offset from the crack line by roughly half the slab thickness. On a 200 mm suspended slab that is about a 100 mm offset and a drilled depth of 130 to 150 mm, which is two thirds to three quarters of the thickness. Never through the far face, or you lose all pressure into the space below.
4. Mechanical packers, not surface portsA 13 mm mechanical packer is drilled in and locked by tightening a nut, which squashes a rubber sleeve against the sides of the hole. It seals inside the slab and takes high pressure. Surface ports glued to the face with a bead of filler over the crack are low pressure only and will not adhere to wet concrete, which is why they blow off a leaking slab. Spacing runs 150 to 400 mm centres, tighter for hairlines, wider for open cracks and joints.
5. Water-flush to prove the crack is connectedClean water is pumped through each packer before any resin. It proves the crack is hydraulically connected from one packer to the next, clears laitance, dust and salt deposits that would block the resin, and pre-wets a dry crack so a water-reactive polyurethane has something to react with. If water does not appear at the next packer, the set-out is wrong and we fix it before injecting.
6. Inject from the low point, in sequenceOn a vertical face, always from the lowest packer upward, so resin displaces water out of the crack ahead of it rather than trapping it. On a horizontal slab crack, consistently from one end to the other. If a single point is running hard, that point is killed first with a fast gel, a hydraulic plug or oakum packing, then we return to the sequence. Mechanical-packer polyurethane work commonly runs 35 to 140 bar, always starting low.
7. Cut back at 24 hours and make goodPackers stay in the slab for a minimum of 24 hours, longer in cold weather. They are then cut off 10 to 20 mm below the finished surface, the recess is filled with a non-shrink hydraulic plug or polymer-modified mortar and ground flush. That means a short second visit, not a same-day finish. Re-screeding, tiling or coating is a separate line item and we quote it as one.

When injecting a leaking slab is the wrong answer

These are the cases where we will tell you not to spend the money with us, and what the right scope is instead.

• The slab is still moving. Ongoing settlement or flexural movement means the crack reopens. Cause first, engineer if the cracking is diagonal, widening upward, or accompanied by a lean or a level change. Seal afterwards.
• Water is coming through the body of the slab across a broad area rather than along a line. There is no crack to grip. That is negative-side tanking, a drained cavity floor, or a drainage and sump scope.
• The crack follows the line of the reinforcement, with rust staining, spalling or drummy hollow-sounding patches. That is corrosion driving the crack open, and sealing it hides an expanding problem. The correct scope is concrete repair: break out to sound concrete behind the bar, treat or replace the steel, reinstate with a compatible mortar.
• It is a designed movement joint or control joint. Filling one with resin defeats its purpose and transfers the movement into the concrete beside it. That is a backing rod and a movement-capable sealant.
• A failed membrane over a whole deck or podium. Injection seals a defined path through concrete. It cannot restore a membrane that has debonded or split across an area. Replacement to AS 4654.1 and AS 4654.2 is the honest answer.
• Damp with no correlation to rain, on a cold surface in a still, poorly ventilated space. That is very often condensation, and injection achieves nothing except the cost of it. Run a 48-hour plastic sheet test before anyone quotes.
• A leaking pipe or service under or beside the slab. Pressure-test it. Quick to rule out, expensive to get wrong.

How we prove the water has stopped

Verification is where a documented repair separates from a hopeful one. During injection, the proof that resin is travelling is that it appears at the next packer along. Afterwards we observe for a minimum of 24 to 48 hours and, wherever the program allows, across a rain event, because a slab that is dry on a sunny Thursday has proved nothing.

We take moisture meter readings at marked points before and after, so there is a number rather than an opinion. Where the geometry allows a water test, we run one. And we hand over photographs taken before and after from the same positions, along with a written description of what we found and what we did.

That is not paperwork for its own sake. If water reappears in six months, it tells us whether it is the path we sealed or a new one, which is the difference between a warranty return and a fresh scope.

What slab leak sealing costs

Slab crack sealing is priced per linear metre of crack, and the honest band for accessible, straightforward work is $150 to $450 per linear metre plus GST. What moves you within it, and outside it:

• Slab thickness, because it drives packer length, drilling time and resin volume.
• Flow rate. A damp weep and a running crack are different jobs with different gel times and different volumes.
• Whether floor coverings, screed or a topping have to be removed to expose the crack, and reinstated afterwards. Quoted separately, never assumed.
• Access. At height off a scissor lift, in an occupied space needing protection, or in an operating car park, rates push toward $450 to $900 per linear metre.
• Repetition. Long continuous runs on one visit trend down toward $80 to $200 per linear metre. A single short crack does not price as a rate at all, it prices as a minimum job.

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

Water is coming up through the middle of my garage floor, not out of a crack. Can that be injected?

Usually not, and it is worth knowing why before you pay anyone. If the water follows a defined line, most often the joint where the floor meets the base of the wall, that joint can be injected at 200 to 400 mm packer centres. If water is pushing up through the body of the slab across a broad area, there is no path for resin to fill and nothing for it to key into. That is hydrostatic pressure through porous concrete, and the answers are a negative-side tanking system, a drained cavity floor, or lowering the groundwater with drainage and a sump. We will tell you which of those you are looking at before quoting.

My slab crack leaks every wet season and dries out completely by winter. Does that change the resin?

It changes it more than anything else about the crack. A hydrophilic resin, which seeks water out and holds it inside the cured material, bonds beautifully to damp concrete but can shrink if the crack dries out permanently. That is called desiccation shrinkage and it is the failure mode that matters. A seasonally drying slab crack in South East Queensland is hydrophobic territory: it uses water to cure but does not hold any, so it stays dimensionally stable through wet and dry cycling. Anyone who tells you one chemistry is always correct is repeating a manufacturer's marketing position, because both camps have one.

How do I measure the crack before I call anyone?

You need three things and none of them cost money. Width: hold the edge of a business card against it, which is about 0.3 mm, and a credit card edge, which is about 0.8 mm. If the business card edge will not enter, you are under 0.3 mm. Extent: mark both ends of the crack with a pencil and write the date beside them, so in a month you know whether it has grown. Behaviour: note whether the water appears only during and after rain, all the time, or only in the wet season. Photograph all of it while it is wet. Those three answers get you a far more useful conversation than a crack width alone.

Can you seal a suspended slab from underneath, on the ceiling of the level below?

Often yes, and frequently that is the better face to work from because it is where you can see the water and where the leak actually exits. The catch is that the drip on the soffit is regularly metres from where the water gets into the slab above, so the first job is mapping the path, not drilling under the wet patch. It also means overhead drilling, resin running down the hose and onto whatever is beneath, and access equipment. Protection of vehicles and fit-out below is a real line item, because one resin drip on a car bonnet costs more than the repair.

Will drilling and injecting lift or crack a thin slab?

It can, and that is exactly why pressure discipline is part of the method rather than an afterthought. Injecting a thin slab that has hydrostatic pressure pushing up underneath it can heave and crack it. We start low and escalate only as the crack demands, we watch for the over-pressure signs, which are the crack visibly widening, resin appearing somewhere unexpected, spalling around a packer, or a sudden pressure drop with no resin return, and on any slender or prestressed element the maximum injection pressure should be nominated in writing by the engineer, not chosen by the operator.

One corner of the slab has dropped and the crack there leaks. What order do we do things in?

Cause, then level, then seal, and skipping straight to the seal is money spent twice. The settlement has a cause: reactive clay, poorly compacted fill, a leaking sewer or stormwater line washing out subgrade, or tree roots drawing moisture out. That has to be found and corrected first. If re-levelling is warranted, that is a ground improvement contractor with a two-component rig and live level monitoring, not us. Once the movement has genuinely stopped, which tell-tales read over 4 to 12 weeks will confirm, the crack can be sealed and it will stay sealed.

The builder says it is only shrinkage cracking. Do we still need to do anything?

If it is dry, very possibly nothing. Concrete shrinks as it cures and AS 3600 works to serviceability crack width limits in the range of roughly 0.1 to 0.4 mm depending on the element and its exposure classification, so fine cracking is designed for rather than a defect. The question is not whether the crack exists, it is whether water is coming through it and whether it is still growing. A dormant hairline in a dry internal slab genuinely needs nothing. The same crack passing water into a car park below needs sealing, because water sitting against reinforcement is what turns a $1,500 job into a concrete repair scope in a few years.