Substrate flatness and moisture checks before floor coverings

Floor coverings do not fail because the covering was defective. They fail because the substrate underneath was wet, uneven, or both, and nobody checked before the covering went down. By the time the blister, the debond, or the cupped board appears, the covering is in place, the rooms are finished, and the builder is the one answering the question.

The verification work is not complicated, but it has to happen in the right order and be recorded in a way that stands up later. Flatness and moisture are separate criteria. A slab can be dead flat and still too wet for the specified covering. It can be dry and still too uneven for a sheet product. Passing one does not mean passing the other, and the acceptance range depends on which finish is going down.

This article covers the pre-covering verification process: what to check, how to interpret readings across different finishes, and what to record so that if something goes wrong, the evidence is already in the project file.

Why the substrate, not the covering, decides the outcome

Every floor covering is a system. The covering, the adhesive or fixing method, and the substrate are designed to work together, and the covering manufacturer’s installation instructions assume a substrate within a stated moisture and flatness range. If the substrate is outside that range, the manufacturer can decline a warranty claim on the basis that the installation conditions were not met. That is the commercial risk: the covering supplier is not liable for a failure caused by a substrate they did not install and did not test.

On a residential project, the flooring trade is usually a subcontractor engaged by the builder or by the owner. The subcontract may say the subbie is responsible for checking the substrate, but the principal contractor still carries the programme risk and the reputational risk when the floor fails. A failure after practical completion means returning to an occupied house, moving furniture, removing the covering, drying or grinding the substrate, and replacing the covering. That is a variation at best and a defect claim at worst.

The reason moisture causes so much trouble is that it moves. A slab that reads dry on the day of testing can pick up moisture from below, from a wet area on the other side of a wall, or from the environment if the building is not yet sealed and dried. Timber and sheet substrates respond to changes in the surrounding air, swelling and shrinking as humidity changes. That is why the reading at the time of covering matters more than any reading taken months earlier during structural works.

Flatness and moisture are two separate acceptance criteria

Flatness is a geometric property. It is measured against a straightedge, or by a method the specification nominates, and compared with the tolerance for the intended finish. A resilient sheet product needs a tighter tolerance than a carpet because a sheet telegraphs every bump and hollow. A tile needs flatness because lippage between adjacent tiles is visible and can be a trip hazard. The tolerance is not a single number for all finishes; it comes from the covering manufacturer and from the relevant standard or specification for the substrate.

Moisture is a condition of the material. It is measured in different ways depending on what is being measured and why. A surface meter gives an indication of moisture near the surface. An in-situ probe measures moisture deeper in the slab, which is where the moisture that causes failures often sits. A relative humidity test measures the moisture condition in a way that some covering manufacturers require for their warranty. These methods are not interchangeable, and a reading from one method cannot be compared with a limit stated for another.

The practical consequence is that you cannot simply take a reading and compare it with a number you remember. You have to know which method the covering manufacturer nominates, take the reading that way, and compare it with the limit they state for that method. If the specification is silent, that is a question for the designer before the covering is ordered, not after it is laid.

The verification sequence before covering goes down

The sequence matters because each step can be invalidated by doing it out of order. Cleaning after testing can remove material that was part of the reading. Testing before the building is dried in can give a reading that will change. Releasing the substrate before recording the results leaves no evidence if the covering later fails.

The first step is to confirm exactly which finish is specified and where. A project can have several finishes: tile in wet areas, resilient sheet in kitchens and laundries, carpet in bedrooms, timber in living areas. Each has its own substrate requirements, and each needs its own verification. Assuming the whole slab is one condition is how a failure gets missed in the one room that mattered.

The second step is inspection and cleaning. Laitance, curing compounds, dust, and construction debris all affect both flatness measurement and moisture movement. A straightedge laid across a ridge of dried compound will show a hollow that is not in the slab. A moisture reading taken through a layer of dust is a reading of the dust, not the slab. Clean the area first, then test.

The third and fourth steps are the actual measurements. Flatness is measured with the instrument and method the specification requires. Moisture is measured with the method the covering manufacturer requires, at the frequency and locations they state. If they do not state a frequency, a common approach is to test each room or each defined area, plus any area with a history of dampness or a known risk. The point is to have enough readings that a localised problem cannot hide.

How to interpret moisture readings across different finishes

The same reading can be acceptable for one finish and unacceptable for another. This is the part that catches builders who treat moisture as a single pass or fail number.

A resilient sheet product, such as vinyl, traps moisture vapour between the covering and the slab. If the slab is still emitting moisture, that vapour has nowhere to go. It builds up under the sheet, and the result is blistering, adhesive failure, or a visible bubble that appears weeks after handover. For these products, the acceptable moisture range is narrow, and the manufacturer will usually nominate a specific test method. A reading that would be fine for a carpet can be a rejection for a sheet vinyl.

Ceramic and porcelain tiles are bedded in adhesive and grouted. The tile itself is not damaged by moisture, but the adhesive bond can be. Moisture in the slab can weaken the adhesive, and if the slab is also not flat, the combination produces debonded tiles and hollow spots. Tile also has a flatness requirement because lippage between tiles is a visible defect and a trip risk. On a tiled floor, both criteria have to pass.

Timber and engineered boards are the most sensitive to moisture because timber moves. The boards and the substrate need to be in equilibrium with the service environment. If the substrate is drier or wetter than the boards, or if the building is not at service conditions, the boards will take up or lose moisture after installation. That produces cupping, crowning, gaps, or squeaks. The reading has to be taken at the time of covering, and the ambient conditions need to be representative of the occupied building. WoodSolutions publishes guidance on timber and moisture that is a sensible reference for how timber behaves in service, and it is worth reading before you accept a timber floor substrate.

Epoxy and resin coatings are also affected by moisture. Moisture vapour pressure under a coating can cause blistering and delamination, and the surface preparation requirements are strict. For these systems, the reading method and the surface preparation standard are both specified by the coating manufacturer, and both need to be followed.

Timber and sheet substrates: the same principles, different mechanisms

Concrete slabs are not the only substrate. Particleboard, plywood, and other sheet flooring substrates are common on upper floors and on timber-framed construction. These materials also have moisture content, and they also have flatness and fixing requirements. A sheet substrate that has swelled from moisture will not return to its original thickness, and a covering laid over a swelled joint will show a ridge.

The mechanism is different from a slab. A slab dries from the surface inward over time, and the moisture that matters is often deep in the section. A sheet substrate takes up and releases moisture from the air, so its moisture content tracks the humidity around it. That means the timing of the reading is different. A sheet substrate can read acceptable in a dry, well-ventilated shell and unacceptable after the building is closed up and the wet trades have finished. The reading has to be taken when the building is in the condition it will be in when occupied, or as close to it as the programme allows.

WoodSolutions guidance is useful here because it explains how timber and wood-based products respond to moisture in service. The practical takeaway is that you cannot treat a timber substrate like a slab and expect the same reading to mean the same thing. You need to know what the covering manufacturer requires for that substrate type, and you need to measure it that way.

What to record, and why the record matters

The record is the difference between a defensible position and an argument. If the covering fails after handover and the cause is disputed, the question will be what the substrate condition was when the covering was laid. Without a record, that question cannot be answered, and the builder is left relying on memory and on the covering supplier’s assessment of a floor they did not see before it was covered.

A useful record includes the method used for each test, the locations and number of readings, the ambient conditions at the time, the acceptance range for the specified finish, the name of the person who took the readings, and the date. If any area was excluded or requires a retest, note that too. The record does not need to be elaborate, but it does need to be specific enough that someone reading it later can tell what was tested and what it was compared against.

It also helps to notify the relevant parties before covering. If the substrate is marginal, or if the finish is one with a narrow tolerance, putting the reading in writing to the covering installer and the designer gives them the chance to raise an objection or to nominate a different product. That notification is not an admission of a problem; it is the normal flow of information on a project where several trades depend on each other’s work.

Worked example: a builder in western Sydney

A builder in western Sydney is completing a single-storey brick veneer house on a waffle raft slab. The slab was poured months earlier, the frame and roof are complete, and the building is at lock-up but not yet fully dried out. The contract specifies tile in the bathrooms and laundry, resilient sheet in the kitchen, and carpet in the bedrooms.

The flooring subcontractor arrives to start the tile. The builder has not arranged any moisture testing because the slab “looked dry”. The tiler starts in the laundry, and within a few weeks the tiles in that area sound hollow. The resilient sheet goes down in the kitchen later, and within a month it blisters near the sink. The carpet in the bedrooms is fine, because carpet is more tolerant of the conditions that caused the other two failures.

The rectification involves removing the affected tiles and the sheet vinyl, drying or grinding the slab where necessary, and replacing the coverings. The cost is carried by the builder, because the covering manufacturers decline warranty on the basis that the substrate conditions were not verified. The builder also loses the programme time for the rework and the goodwill of the owner, who has already moved in.

What should have happened is straightforward. Before any covering was ordered or laid, the builder should have confirmed the moisture test method each manufacturer required, tested the slab in each area at the appropriate time, compared the readings with the acceptance range for each finish, and recorded the results. If the slab was still too wet for the resilient sheet, the builder could have delayed that installation, specified a different product, or used a surface treatment the manufacturer accepted. Any of those options would have been cheaper than removing a finished floor.

The example is not about a bad tiler or a bad sheet product. It is about a verification step that did not happen, and about the fact that the tolerance differs by finish. The carpet passed because carpet tolerates more. The tile and the sheet failed because their substrates did not meet the conditions those products required.

Where this fits in the project programme

Pre-covering verification is a hold point, not a background task. It sits after the substrate is complete and clean, after the building is sufficiently dried in for the readings to be meaningful, and before any covering is laid. On a project with several finishes, it may need to be repeated at different times for different areas, because the building conditions change as the wet trades finish and the shell dries.

The programme implication is that the verification takes time and can delay the covering if the readings are not acceptable. That is a reason to plan it early, not a reason to skip it. A delay of a few days to allow a slab to dry is a small cost compared with the cost of removing and replacing a finished floor. It is also a reason to raise the requirement with the client early, so that the expectations about handover dates are realistic.

For builders working in NSW, the verification record is also part of the evidence that the work was done in accordance with the plans and specifications. Keeping it with the project documentation is good practice regardless of whether anyone asks for it. If you want to understand how this kind of evidence fits into the broader CPD and compliance picture, the NSW builder CPD guide explains how the requirements work, and the PointCPD course catalogue has courses that cover building materials and quality processes. The PointCPD home page is the starting point for NSW builder CPD courses.

Common mistakes and how to avoid them

The most common mistake is testing too early. A slab that is tested before the building is dried in can read dry and then gain moisture later, or read wet and be unnecessarily delayed. The reading needs to be taken when the building is in, or close to, the condition it will be in when occupied. If the covering is going into a space that will be air-conditioned, the reading should reflect that environment as closely as the programme allows.

The second mistake is using the wrong method. A surface meter reading is not comparable with an in-situ probe limit, and neither is comparable with a relative humidity limit. If the covering manufacturer nominates a method, use that method. If they do not, ask them in writing and keep the response.

The third mistake is assuming one acceptance range applies to the whole project. Different finishes have different tolerances, and a single reading can pass for one and fail for another. Test per area, per finish, and record which reading belongs to which area.

The fourth mistake is not recording the result. A verbal “it looked fine” is not evidence. A written record with method, location, reading, and acceptance range is.

The fifth mistake is covering a marginal substrate without notifying anyone. If the reading is close to the limit, or the finish is one with a narrow tolerance, put it in writing. That protects you and gives the other parties the chance to make a decision they can own.

Where to go for the actual limits

This article deliberately does not state numbers, because the limits come from the covering manufacturer, the relevant Australian Standard, and the project specification. Those are the documents to check. If the specification is silent, ask the designer. If the covering manufacturer’s instructions are not on site, get them before the covering is ordered. The article on slab edge thickness and tolerances covers a related set of slab checks, and the waffle raft slab defects checklist is useful for the structural side of the same slab. For timber substrates, the WoodSolutions guidance on timber and moisture is a practical reference for how timber behaves in service.

The principle to carry onto every job is that the substrate has to be verified against the finish that is going onto it, using the method that finish requires, at a time when the reading means something, and the result has to be recorded. Do that, and the covering has a chance of surviving. Skip it, and the failure will surface after practical completion, when it is most expensive to fix.

Questions builders ask

• Moisture
• Quality
• Slabs

Published 14 September 2026. General information for builders, not legal, financial or engineering advice for a specific job. Check current requirements with the relevant regulator or standard before relying on them.