Why Grout Shrinks: Plastic, Drying and Autogenous Shrinkage
On this page
- Why a solid joint changes volume at all
- The main kinds of shrinkage
- Plastic shrinkage: losing water before the grout sets
- Chemical and autogenous shrinkage: the reaction itself
- Drying shrinkage: the big one
- How sand, polymers and additives limit shrinkage
- How standards treat shrinkage
- Shrinkage in epoxy, urethane and premixed grouts
- What shrinkage looks like in the finished floor
- Frequently asked questions
- Sources and standards
Cement grout shrinks because it loses water. Some evaporates from the fresh surface (plastic shrinkage), some leaves the hardened pores over weeks (drying shrinkage), and some is consumed by the cement reaction itself (autogenous shrinkage). Sand and polymers restrain these movements, but when shrinkage exceeds the grout's stretch or bond, it cracks or pulls away from tile edges.
Key takeaways
- Cement paste is the part that shrinks. Sand grains do not, so more sand means less shrinkage, which is the main reason unsanded grout is limited to narrow joints.
- Drying shrinkage is the largest component in tile grout. It comes from capillary tension in the pores and water leaving the calcium silicate hydrate gel.
- More mixing water means more evaporable water, more pores and more shrinkage.
- Polymers mostly reduce cracking rather than shrinkage itself: they let grout stretch further and grip tile edges harder.
- Shrinkage cracks appear early, are fine and random, and often run along tile edges. Cracks that line up with substrate joints or keep growing point to movement instead.
Why a solid joint changes volume at all
Freshly mixed cement grout is a suspension of cement, sand, pigment and additives in water. Over the following days and weeks, that water does one of three things: it reacts with the cement, it stays in very fine pores inside the hydration products, or it evaporates. Each route changes volume. The cement reaction produces solids that take up less space than the cement and water that formed them, and evaporation from fine pores pulls the solid skeleton inward. The result is a joint that ends up slightly smaller than the space it was packed into.
The numbers are small in percentage terms, but a grout joint is bonded to rigid tile on both sides and to the setting bed below. It cannot simply shrink freely. Restrained shrinkage turns into tensile stress, and cement-based materials are weak in tension. That is the link between this invisible chemistry and the hairline cracks and edge gaps people see. For the hydration chemistry itself, see how cement grout cures.
The main kinds of shrinkage
| Type | When | Cause | Typical sign in tile grout | Main controls |
|---|---|---|---|---|
| Plastic shrinkage | First hours, before set | Surface evaporation faster than water can rise from below | Fine surface crazing, shrink-back of the joint face, edge gaps forming as grout firms | Water retention agents, no drafts or heat, damp porous tile and substrate first |
| Settlement | First hours | Solids settle and water bleeds up; grout slumps into voids | Low or dished joints, especially wide or deep ones | Stiffer mix, full compaction, properly cleaned joint depth |
| Chemical and autogenous shrinkage | From mixing through the first weeks | Hydration products occupy less volume than reactants; internal drying as water is consumed | Usually minor at normal grout water contents; matters in low-water, high-cement and rapid-setting products | Formulation: cement content, expansive components |
| Drying shrinkage | Days to months after set | Water leaving capillary and gel pores in hardened grout | Hairline cracks across joints, separation along tile edges, cracks in wide unsanded joints | Sand content, low mixing water, polymers, shrinkage-reducing additives |
| Carbonation shrinkage | Months to years | Calcium hydroxide reacting with carbon dioxide from air | Rarely visible on its own; adds to surface microcracking in porous grout | Dense, well-cured grout |
| Thermal contraction | Hours after mixing, then with seasons | Cooling after hydration heat; temperature swings in service | Minor in thin joints; relevant on sun-exposed or heated floors | Movement joints per TCNA EJ171 |
Plastic shrinkage: losing water before the grout sets
For the first few hours, grout is a soft paste. Water at the surface evaporates, and if it leaves faster than bleed water can rise to replace it, tiny curved water surfaces (menisci) form between particles at the face. Surface tension in those menisci pulls the particles together, and the top layer contracts while the grout beneath does not. The paste has almost no strength at this stage, so it tears.
In concrete practice, evaporation rates above about 0.2 lb per sq ft per hour (1 kg per sq m per hour) are treated as high risk for plastic cracking. Grout joints have it worse in one respect: the surface-to-volume ratio of a joint is high, and porous tile edges and dry substrates draw water out from the sides and bottom at the same time. Warm air, low humidity, fans, heaters and direct sun all push evaporation up. This is why cellulose ethers (water retention agents) are in almost every grout formula, and why data sheets warn against grouting in drafts. Site measures belong to grouting in hot and cold weather.
Settlement happens in the same window. Heavy particles sink and water rises, and in deep or wide joints the grout can slump into voids left by incomplete packing or by thinset that was not cleaned out to a consistent depth. That kind of loss shows up as low or recessed grout joints rather than cracks. Good packing is covered in grout compaction and joint filling.
Chemical and autogenous shrinkage: the reaction itself
When portland cement hydrates, the solid products (mainly calcium silicate hydrate, calcium hydroxide and ettringite) occupy less volume than the cement and water that went into them. This chemical shrinkage amounts to roughly 0.06 to 0.07 milliliters per gram of cement at complete hydration. Before the grout sets, that volume loss simply lets the paste settle a little. After set, the rigid skeleton resists, and the missing volume appears as tiny empty spaces inside the pores.
As those pores empty, the internal humidity drops even though no water has left the material. This self-desiccation creates the same capillary tension that drying does, and the bulk material shrinks slightly. That external change is called autogenous shrinkage. It is significant when the water-cement ratio is low (below roughly 0.40) and cement content is high. Most tile grouts are mixed wetter than that, so autogenous shrinkage is usually a small share of the total. It can matter more in high-strength, low-water and rapid-setting formulas, which is one reason some of them include expansive components that form extra ettringite to offset the early contraction.
Drying shrinkage: the big one
Once grout has set, the water left in its pores gradually evaporates toward equilibrium with the room. Three mechanisms pull the solid inward as this happens:
- Capillary tension. As water retreats from capillary pores, menisci form in progressively smaller pores. The finer the pore, the stronger the pull on its walls.
- Loss of adsorbed water. Water layers on the enormous internal surface of calcium silicate hydrate prop the gel sheets apart. As the air dries, those layers thin and the sheets draw closer.
- Interlayer water loss. At low humidity, water held between the gel layers leaves, causing further and partly irreversible contraction.
Some drying shrinkage reverses when the grout gets wet again, which is one reason grout in showers and on exterior surfaces experiences repeated small movements. A portion of the first drying is permanent.
The amount depends mostly on how much paste there is and how much evaporable water it holds. Typical orders of magnitude from cement and concrete research show the effect of diluting paste with aggregate:
| Material | Paste content | Typical drying shrinkage | Grout equivalent |
|---|---|---|---|
| Neat cement paste | All paste | About 0.2-0.4% or more | Closest to fine, filler-light unsanded grout |
| Cement-sand mortar | Moderate | About 0.1-0.2% | Sanded grout |
| Concrete | Low (coarse aggregate dominates) | About 0.04-0.08% | No direct grout equivalent |
To put that in tile terms: an unrestrained 10-foot (3 m) run of grout shrinking 0.2 percent would shorten by about 1/4 inch (6 mm). Bonded between tiles it cannot shorten, so the strain is absorbed as stress, as countless invisible microcracks, or as a few visible cracks where the grout is weakest.
How sand, polymers and additives limit shrinkage
Sand and fillers
Silica sand does not shrink. In sanded grout, graded sand forms a stiff skeleton that the shrinking paste has to pull against, and it replaces paste volume that would otherwise shrink. The finer and more uniform the filler, the more paste is needed to coat it, so unsanded grout has a much higher paste fraction. That is the mechanism behind the common rule of using unsanded grout for joints under 1/8 inch and sanded for 1/8 inch and wider: in a wide joint, a paste-rich grout has too much shrinking volume and cracks. Modern high-performance grouts relax this rule with fine, dense filler packing and polymer modification, which is why some fine-textured products are rated for wider joints; the data sheet sets the limit. See sanded grout and unsanded grout.
Water content
Every extra bit of mixing water is evaporable water that will eventually leave and pull the paste inward. It also makes coarser pores, which shrink more when they empty. Holding mixing water to the low end of the data sheet range is the cheapest shrinkage control there is. The broader consequences are covered in why excess water weakens grout.
Polymers
Redispersible polymer powders and liquid latex additives do not dramatically cut the amount of shrinkage. What they change is the grout's response to it. A polymer film threaded through the cement matrix raises tensile strength and strain capacity, lowers the stiffness of the composite and improves adhesion to tile edges, including dense porcelain and glass. The grout can therefore absorb more shrinkage strain before it cracks or debonds. Polymers and cellulose ethers also slow early water loss, which reduces plastic shrinkage. More detail is in polymer-modified grout.
Shrinkage-reducing and expansive additives
Some formulations include shrinkage-reducing admixtures, typically glycol-based compounds that lower the surface tension of pore water and so weaken capillary tension. Others use calcium sulfoaluminate or similar components that form ettringite early, producing a slight expansion that offsets later contraction. Which approach a product uses is rarely disclosed in detail, but the result is reflected in its standards compliance.
How standards treat shrinkage
ANSI A118.7 high-performance cement grouts must meet a maximum linear shrinkage limit, measured on molded test bars over a defined curing schedule, alongside requirements for strength, absorption and color consistency. ISO 13007-3 also includes a shrinkage requirement for cementitious grouts (commonly cited as 3 mm per meter, or 0.3 percent). These are lab tests on specimens cured under controlled conditions. A product that passes can still crack in the field if it is mixed too wet, dries in a draft, or is placed in joints wider than its rating. For how the classifications fit together, see grout standards: ANSI and ISO and high-performance cement grout.
Shrinkage in epoxy, urethane and premixed grouts
Epoxy grouts shrink slightly as the resin crosslinks, but the high filler content and the absence of evaporable water keep total shrinkage low, and epoxy bonds strongly to tile edges. Premixed acrylic and urethane grouts that cure by water evaporation lose volume as they dry, which is why manufacturers specify maximum joint widths and depths and sometimes warn that deep joints may dish slightly. Their flexibility generally keeps that shrinkage from cracking. See premixed grout.
What shrinkage looks like in the finished floor
Because shrinkage happens early and in every joint, its signs are distinctive:
- Edge separation. A hairline gap along one side of a joint, where grout bond to the tile edge was weakest. Dense, smooth porcelain and glass edges give cement grout little to grip.
- Short transverse cracks. Fine cracks across the joint at irregular intervals, usually within the first few weeks.
- Cracks in wide joints. Paste-rich grout in a joint wider than its rating develops cracks down the middle of the joint.
- Surface crazing. A web of very fine cracks on the joint face from plastic shrinkage, often with a weak, dusty surface.
Movement-related cracks look different: they tend to run in long continuous lines, follow substrate joints or framing, cross tiles as well as grout, and grow over time. Telling the two apart is the job of the diagnostic pages why grout cracks and grout cracking after installation. Regrouting a shrinkage problem with the same mix and conditions usually reproduces it; regrouting a movement problem never fixes it, which is the subject of grout keeps cracking.
Bottom line: Shrinkage is built into cement grout. You control how much of it turns into cracks: choose sanded or high-performance grout rated for the joint width, mix with minimum water, keep drafts and heat off fresh joints, and pack joints full. Cracks that follow the structure are not shrinkage and need movement joints, not better grout.
Frequently asked questions
Is it normal for grout to shrink?
All cement grout shrinks slightly as it cures and dries. In a properly mixed product used within its joint-width range, that shrinkage stays invisible or limited to a few hairlines. Visible gaps or many cracks mean something pushed it past its limits.
Why is my grout pulling away from the tile edges?
Drying shrinkage pulls the grout toward the joint center, and it lets go where its bond is weakest, often against dense porcelain or glass edges or dusty, unclean edges. Excess mixing water and fast drying make it worse.
Does unsanded grout shrink more than sanded grout?
Yes. Unsanded grout has a higher proportion of cement paste, which is the part that shrinks, while sand grains stay stable and restrain the paste. That is why basic unsanded grout cracks in wide joints.
Does adding more water to grout make it shrink more?
Yes. Extra water creates more and larger pores that empty as the grout dries, so the grout shrinks more and cracks more easily, besides losing strength.
How long does grout keep shrinking?
Most drying shrinkage happens over the first few weeks, and it continues at a slowing rate for months as the grout approaches equilibrium with room humidity. Shrinkage cracks typically appear within the first month.
Does epoxy grout shrink?
Only slightly. The resin contracts a little as it cures, but heavy filler loading and the lack of evaporating water keep shrinkage low, and strong edge bond resists separation.
Sources and standards
- ANSI A118.7, High Performance Cement Grouts for Tile Installation (shrinkage, absorption and strength requirements)
- ANSI A118.6, Standard Cement Grouts for Tile Installation
- ISO 13007-3, Ceramic tiles: grouts and adhesives, test methods and requirements for grouts
- TCNA Handbook for Ceramic, Glass, and Stone Tile Installation, including EJ171 movement joints
- ACI 305R, Guide to Hot Weather Concreting (evaporation and plastic shrinkage cracking)
- Portland Cement Association, Design and Control of Concrete Mixtures (volume changes of concrete)
- A. M. Neville, Properties of Concrete (drying, autogenous and carbonation shrinkage mechanisms)
- Manufacturer technical data sheets for sanded, unsanded and high-performance grouts (joint width limits and mixing water)
Grout Atlas guidance is checked against ANSI A108/A118 and ISO 13007 requirements, the TCNA Handbook and manufacturer technical data sheets. Product formulations change, so confirm details on the current data sheet for the product you buy. Spotted an error? Report a correction.