On this page15 sections
- 01Joint width, joint depth and the ribbon of mortar between tiles
- 02Concepts to hold first
- 03Estimating the negative space
- 04Small tiles drink, large tiles sip
- 05Flush from bed to surface
- 06The tub is the cross-check
- 07How the method works
- 08Try it, verified
- 09What each input represents
- 10Worked example
- 11Reading the result
- 12Common mistakes
- 13Questions readers arrive with
- 14When this calculation is used
- 15Assumptions and guards
Joint width, joint depth and the ribbon of mortar between tiles
The formula is the geometry of a grid of channels. Each square metre of tiling carries a network of joints whose total length depends on the tile’s footprint — its perimeter relative to its face area — and each metre of joint is a slot of the spacer’s width and the tile’s depth. Multiply the slot’s cross-section by the network’s length and by the grout’s density factor, and the kilograms per square metre fall out.
Tile format is the strongest lever, and it works in reverse of intuition about size. A mosaic sheet of tiny tesserae is nearly all edges, so it drinks grout; a large-format slab floor of the same area is nearly all face, so it sips. Halving the tile dimension roughly doubles the joint network, which is why a grout order must never be carried from one format to another by habit.
The spacer decides the width, and fashion has been moving it. Rectified tiles laid at a two-millimetre joint take a fraction of the grout that a rustic floor at a wide joint takes; the joint depth, meanwhile, is simply the tile thickness, because a properly grouted channel is filled flush from bed to surface. Both dimensions enter in millimetres and both multiply the answer directly.
The density factor translates channel volume into bag weight, and it belongs to the product. The default is the conventional figure for a sanded cement grout; the manufacturer’s own coverage formula may state another, and epoxy systems play by their own rules entirely. It is the one input that comes from the tub rather than the tape measure.
The answer is reported twice: as a consumption per square metre — the number printed on grout packaging, useful for sanity-checking against the tub — and as the total mass for the area with wastage applied. Wastage here is the grout that cures in the bucket, gets sponged off during clean-down, or is lost to over-filled joints; it is a modest allowance, because grout is mixed in small batches as the work proceeds.
Concepts to hold first
The grid of channels a tiled surface carries between its tiles. Every square metre of a given format contains a knowable length of it, and grout quantity is that length multiplied out — which is why grout is estimated from geometry where adhesive was estimated from a datasheet rate.
How much edge a tile has relative to its face. It is the property of the format that drives the joint network’s length: small tiles are nearly all edge, large ones nearly all face, and halving the tile dimension roughly doubles the network in the same floor.
The cross-section of one channel. The spacers set the width — narrow for rectified tile, wider for handmade and rustic formats — and the tile’s thickness sets the depth, because a sound joint is filled flush from the adhesive bed to the surface. Both multiply the answer directly.
The constant from the grout’s own coverage formula that converts channel volume into bag weight. It is the one input that comes from the tub rather than the tape measure: cement grouts cluster near the conventional figure, and epoxy systems play by their own rules entirely.
Estimating the negative space
Every other take-off on this job measured something you can stand on. Grout measures the gaps — and the gaps have a geometry as definite as the tiles that make them. Fix the format, and each square metre of floor contains a fixed length of joint; fix the spacer and the tile thickness, and each metre of that joint is a slot of known cross-section. Volume follows, and the density factor turns volume into the kilograms a bag must supply. Nothing is estimated by habit; every term traces to a dimension you chose or a constant the tub states.
That traceability is the practical point of doing it this way. When a grout order looks odd, each input can be interrogated separately — did the spacer change, did the tile get thicker, is the density factor the product’s own — instead of arguing about a single mysterious allowance. The formula is transparent because the geometry is.
Grout reasons from the tile format, through the joint network it creates and the channel cross-section the spacers and thickness set, to the mass to buy
The reasoning runs format, network, cross-section, mass. The floor’s area enters only at the end — the format decides how much joint each square metre carries.
Small tiles drink, large tiles sip
Size intuition runs backwards here. A mosaic sheet of tiny tesserae is nearly all edges — its perimeter is enormous relative to its face — so a mosaic floor is dense with channel and drinks grout. A large-format slab floor of exactly the same area is nearly all face: few joints, little channel, a fraction of the demand. Halving the tile dimension roughly doubles the joint network, which is a stronger lever than any plausible change of spacer.
This is why a grout order must never be carried from one job to the next by habit. The same room, retiled from a mid-size ceramic to mosaic, multiplies its grout demand while its area stands perfectly still — and a fixer who bought on last job’s memory discovers it at the worst moment, sponge in hand. The format is an input because the format is the answer’s biggest driver.
Fashion moves the width lever too. Rectified tiles laid at the narrow joints modern spacers allow take a fraction of what a rustic floor at a wide joint takes, in the same format. Width multiplies the answer directly, so a change of spacer is never cosmetic to the order.
Flush from bed to surface
The depth input has a physical meaning worth pausing on: it is the tile’s thickness, because a properly grouted channel is filled flush from the adhesive bed at the bottom to the tile surface at the top. The joint is not a bead sitting in a groove; it is a full-depth ribbon of mortar standing between the tiles, keyed to the bed the last take-off priced. Thicker tile means deeper channel means more grout, entirely independent of format and spacer.
Joints deliberately finished shy of the surface use less, but specifying that is the product literature’s business, not this model’s — the formula prices the sound, flush joint that tiling standards describe. And a wide gap is not a wide joint: past the guard’s limit, a gap moves with the building and must be sealed as a movement joint, which is silicone’s job. The engine refuses to price it as grout rather than dress a structural detail as a filling.
The build-up in cross-section: structural floor, adhesive bed, then the tile with its grout joint filled flush to the surface
The joint in its build-up: the channel runs the tile’s full thickness, from the cured adhesive bed below to the finished surface above. Depth is not a choice; it is the tile.
The tub is the cross-check
The answer arrives twice, and the order of reading matters. First comes consumption per square metre — the same kind of number printed in the coverage table on every grout tub. Check it against the tub you intend to buy: for your format, joint and thickness, the two should agree closely, and a disagreement means an input is wrong — usually a dimension in the wrong unit or a density factor from the wrong product. Only then read the total mass, which is that consumption carried across the area with wastage applied.
Wastage here is modest by nature. Grout is mixed in small batches as the work proceeds, so the losses are the batch that stiffens, the sponge-off during clean-down, and the little that over-fills. It is the smallest allowance on the job — and if it wants to be large, the joints are probably being raked deeper or cut wider than the inputs admit, which is a measurement problem, not a mixing one.
How the method works
The tile’s plan dimensions fix how much joint a square metre of the format contains — its perimeter relative to its face area — and the spacer width and tile thickness fix each channel’s cross-section.
Channel length times cross-section times the grout’s density factor gives the consumption per square metre, the figure to check against the tub’s own coverage table.
That consumption is carried across the tiled area and grossed up by the wastage percentage to give the total mass to buy.
The certified engine performs this calculation. This page explains what it does; it does not reproduce it, because a second implementation of a specified method is a second answer waiting to disagree with the first.
Try the worked scenario
The engine below is the calculator page’s own certified engine, verified before it opens and pre-filled with the pack’s declared vector: the journey’s floor in a common square format on a narrow joint, at the conventional sanded-cement density factor. Swap the tile dimensions down toward mosaic sizes and watch the consumption multiply with the area untouched; then widen the joint one spacer size and see the width lever move the order on its own.
Read the per-square-metre consumption first and check it against the coverage table on the tub you intend to buy — agreement is the sign the inputs describe your job. Every figure is computed live by the verified engine; nothing on this page stores an answer.
What each input represents
The area whose joints are being filled, in square metres — the same surface the layout and adhesive take-offs describe, because every tile that is bedded is eventually grouted.
The longer plan dimension of one tile, in millimetres. Together with the width it fixes how much joint a square metre of this format contains.
The shorter plan dimension of one tile, in millimetres. Square formats enter the same figure twice; plank formats are where the two dimensions earn their separate fields.
The gap the spacers hold between tiles, in millimetres — narrow for rectified tile, wider for handmade and rustic formats. The answer scales directly with it.
How deep each channel runs, in millimetres — the tile’s thickness, since a sound joint is filled flush from the adhesive bed to the tile surface.
The bulk-density constant from the grout’s own coverage formula, converting joint volume to mass. The default is the conventional sanded-cement figure; take the product’s own value where the manufacturer states one.
The percentage added for mixing losses, sponge-off during clean-down and the batch that stiffens before it is used.
Worked example
The scenario
The pack’s declared vector: thirty square metres of three-hundred-millimetre square tile on a three-millimetre joint, eight millimetres deep, at the conventional sanded-cement density factor and the default tenth of wastage.
Read the per-square-metre consumption first and check it against the coverage table on the tub you intend to buy — the two should agree closely, and a disagreement means an input is wrong. Only then read the total mass, which is that consumption carried across the area with wastage applied.
Every figure in this example is produced by the certified engine when the calculator loads, checked against the signed pack’s declared test vectors; nothing on this page stores an answer. Swap the format down to mosaic dimensions and watch the consumption multiply with the area untouched — the joints, not the tiles, are what grout is bought for.
Reading the result
The per-square-metre consumption is the diagnostic figure: it is what the tub’s own coverage table states, so the two disagreeing means an input is wrong, not the tub.
A demand that jumps when the format shrinks is the geometry speaking — the joints, not the tiles, are what grout is bought for, and small formats carry far more joint per square metre.
If the order still looks high with the inputs verified, look at the joint width and the thickness honestly: wide rustic joints in thick tile are simply grout-hungry, and the model is reporting it rather than flattering it.
Common mistakes
Carrying a grout quantity from a previous job in a different format. Halving the tile dimension roughly doubles the joint network, so habit is the least reliable guide on this take-off in particular.
Entering dimensions in the wrong unit — the tile and joint dimensions are millimetre inputs, and a centimetre slipped in among them moves the answer far more than any judgement call.
Using the conventional density factor for an epoxy or speciality grout. The factor belongs to the product; where the manufacturer states their own coverage constant, theirs wins.
Pricing a wide moving gap as a grout joint. Past the guard’s width limit the gap is a movement joint and wants silicone — the engine refuses it, and the refusal is the advice.
Skipping the tub cross-check. The per-square-metre figure exists to be compared with the packaging before purchase; buying on the total alone throws away the model’s built-in audit.
Questions readers arrive with
Which dimensions do I measure, and in what unit?
The tile’s plan length and width, the spacer-set joint width, and the tile thickness — all in millimetres. Square formats enter the same figure twice; plank formats are where the two plan fields earn their keep.
Where does the density factor come from?
From the grout’s own coverage formula where the manufacturer states one, and from the conventional sanded-cement figure otherwise. It converts channel volume to bag weight, so an implausible value describes no real product — the guard refuses it.
Does the answer cover regrouting an existing floor?
Yes, with one change of posture: the joints are raked out and their real width and depth measured rather than chosen. Raking rarely reaches the full tile thickness, so the measured depth is usually shallower than a new lay’s.
Is the perimeter of the room, or the cut border tiles, treated specially?
No — the formula models the regular network across the area, and the perimeter channel plus the odd irregular cut is noise the wastage allowance absorbs. The perimeter’s own seal, where tile meets wall or bath, is silicone and is excluded.
Why is grout estimated from geometry when adhesive came from a datasheet?
Because the joint is fully described by dimensions you choose — format, spacer, thickness — while an adhesive bed depends on rib collapse and substrate that only the manufacturer’s testing can speak for. Each take-off uses the strongest authority available to it.
When this calculation is used
Buying grout once the tile format, spacer size and tile thickness are all decided.
Sanity-checking the consumption printed on a grout tub against the actual joint geometry of the job.
Seeing what a change of spacer — or a switch from mosaic to large format — does to the order before it is placed.
Estimating regrouting of an existing floor, where the joints are raked out and their dimensions measured rather than chosen.
Assumptions and guards
Joints form a regular grid of the entered width and depth throughout — the formula models the ideal network, so raked, over-deep or irregular joints belong in the wastage judgement.
The joint depth equals the tile thickness, filled flush; tiles grouted shy of the surface use less, but specifying that is the product literature’s business.
The density factor faithfully represents the chosen grout; cement grouts cluster near the default, epoxy and speciality systems must take the manufacturer’s own figure.
The result is joint filler only: silicone at movement joints, perimeter seals and trim beads are excluded, and a wide gap is a movement joint’s job, not grout’s.
A joint wider than fifteen millimetres is beyond what a grouted tile joint means, and the pack ships the refusal as a declared test vector — a gap that size moves and must be sealed as a movement joint, so the engine refuses to price it as grout.
The density factor is held to the physically plausible band for grout products, with its own declared refusal vector: a figure outside it describes no real product and usually means the manufacturer’s coverage constant has been misread.
Tile dimensions, joint dimensions and the area must all be strictly positive — the formula divides by the tile’s face area, so a zero dimension is refused before it becomes a division the order cannot mean.