Floor flatness gets specified early and understood late. It sits in the documentation as a pair of numbers; it gets priced as a line item, and it becomes urgent at the point a racking installer arrives and tells the builder the aisles are out.
Here is what the numbers mean and why the way they are measured matters more than the values themselves.
FF and FL
FF is flatness, the measure of how bumpy a floor is over a short distance. FL is levelness, the measure of how much it departs from horizontal across the slab. They are independent. A floor can be beautifully flat and sloped, or dead level and rippled. Most specifications set both.
Higher numbers are better. A general commercial floor might specify FF 25 and FL 20. A superflat warehouse floor is in a different world entirely.
The part that catches people out
Standard FF and FL numbers are statistical. They average the good areas in with the bad, which means a floor with a genuine problem in one location can still pass.
Very narrow aisle warehouses are not measured that way. VNA floors are assessed against Fmin numbers, and rather than averaging, they are judged on the single worst measurement in each aisle. It does not matter how much of the floor is perfect. One defect in one aisle sets the result for that aisle.
That is the right way to measure it, because it matches how the floor gets used. A turret truck at fourteen metres of lift does not care about your average. It cares about the spot it is standing on.
Why data centres are pushing this harder
The data centre pipeline is bringing a new set of floor requirements into a market that has spent twenty years optimising for distribution centres. The loads are different, the equipment that moves across the floor is different, and the tolerance for remediation after handover is close to zero, because the fit-out schedule behind it does not have the float.
Grinding a floor after the fact is possible. It is also slow, dusty, expensive and happens inside a building that other trades are trying to work in.
How it gets achieved
Not at the finishing stage. By then the result is largely set.
Flatness is decided by the screed, the mix, the placement rate, the joint layout and the curing regime, and it is measured continuously through placement rather than checked at the end. We run dipstick measurements through the pour, not after it, because a floor that is drifting can still be corrected while the concrete is workable.
At Qube Warehouse 5 at Moorebank Logistics Park, the floor measured FF 109.9 and FL 112.5 across 48,000 m², a combined 222.4 against a previous record of 153 that had stood since 2007.
That result came from planning, not from luck, and it is the same planning we apply to a floor with a far less exotic specification.
