Resolving Tolerance Disputes: How Floor Flatness Is Actually Measured

BS 8204 surface regularity classes work by placing a straightedge of a defined length on the floor at specified points and measuring the maximum gap beneath it. Each SR class has an associated maximum permitted deviation, and a floor is compliant if measured deviations fall within that limit across the sampled points.

This is a very different exercise from someone walking across a floor and forming a visual opinion. Light raking across a surface can visually exaggerate very minor, fully compliant deviations, which is why measurement rather than impression should always settle a dispute.

The starting point for any tolerance dispute should be the specification document: what SR class, if any, was called for, and against what method was it to be verified. Without this reference point, both sides are arguing from differing and unverifiable expectations.

Where a dispute cannot be resolved by the parties directly, an independent measured survey using the method set out in BS 8204 provides an objective basis for agreement, and its findings should be recorded and retained alongside other project documentation.

A straightedge survey carried out too early, before the screed has finished shrinking marginally as it dries, can record a deviation that later settles within tolerance on its own.

Timing the measurement against the manufacturer's stated curing period, and noting that date on the survey record, avoids a dispute being raised from a premature reading.

A straightedge survey carried out to settle a flatness disagreement should record the grid points measured, the deviation found at each one and the SR class it was checked against, rather than a single pass or fail statement.

Naming the surveyor, the date and the equipment used turns that survey into a document either party can rely on later, instead of a recollection of who measured what and where.