On this page
Maintenance · 9 min read
Microcement common issues — and how to spec around them
Most microcement failures are not material failures. They're spec failures and prep failures. Here's the recurring shortlist — what each issue looks like, what actually caused it, and the question to have asked before signing the quote.
Hairline cracks tracking through the finish
What it looks like: a fine line, typically less than 0.3 mm wide, often dead straight, occasionally branching, appearing anywhere from three weeks to twelve months after install. The line tends to follow something underneath — a screed expansion joint in a floor, a plasterboard joint in a wall, a corner, a door reveal, the edge of a pipe chase. Two giveaway patterns: cracks that run perfectly straight (always tracking a substrate joint), and cracks that radiate from a single point (a fixing, a hidden noggin, or a substrate void).
What actually caused it: in 80% of cases, missing fibreglass mesh in the primer layer. Microcement is rigid; the substrate beneath it moves slightly with temperature, humidity and load. Without the embedded mesh distributing those movements across the surface, all the movement concentrates at the line of least resistance — a substrate joint or a structural change in the wall — and cracks the finish along that line. Other causes: substrate moving more than expected (a freshly extended wall that hasn't fully settled, plasterboard with unsecured joints), thermal cycling on a heated floor where mesh weight was specified too light, or applying over a substrate that wasn't properly identified at survey (the installer assumed plasterboard but the wall was actually skimmed render over plasterboard with an air gap).
How frequent is it: by far the most common issue we get called to inspect after the fact, accounting for ~40% of all "something has gone wrong" call-outs. Almost always a missing-mesh issue. Almost always preventable at quote stage.
How to spec around it: any quote should explicitly include fibreglass mesh embedded in the primer, with 10 cm overlap at joints, corner reinforcement at internal corners, and 100 g/m² weight on heated floors or floors with known small movement. Ask the installer where they apply mesh and how — a vague answer or a "we don't really need it on this substrate" is a flag. The full kit is in our microcement toolkit guide; the heated-floor specifics are in underfloor heating compatibility.
The fix when it does happen: rake out the crack to depth, re-prime that line with a flexible primer, embed a strip of mesh across the crack, re-apply matching base + finish coats, re-seal locally. The patch is invisible on a hand-applied finish — usually you can't tell where it was. Repair cost typically £400–£900 for a single cracked line on a residential floor. The repair only works if the underlying cause is fixed too: if the mesh is genuinely missing across the whole floor, the cracks will keep appearing in new places.
Colour blotching across the finish
What it looks like: irregular dark or light patches across an otherwise even surface, most visible in directional light or under low evening lamps. Doesn't track straight lines (which would be cracks); follows the trowel-strokes path of where one batch was applied. Often a circular or roughly oval patch a metre or so across, sometimes with a slight halo at the edge where the next batch overlapped. Worst on smooth-finish surfaces because there's no texture to mask the variation; less visible on stucco or coarse finishes.
What actually caused it: most often, re-tempering — adding extra water to a batch that started to set, to make it usable for longer. Re-tempered material has a different water-cement ratio than fresh material, which changes the cured colour. The patch is the area where the over-watered batch was applied. Less commonly: mixing two slightly different production batches of the same colour without blending them first (each manufacturer's batches have small lot-to-lot variation, around 2–5% colour shift at most, but enough to read as patches when laid side by side); or applying onto a substrate that was uneven in absorbency (e.g. patches of fresh skim mixed with old plaster) without an adequate primer to even out the porosity.
How frequent is it: ~15% of issue call-outs. More common in summer installs (faster set times, more pressure to extend batches) and in cheap-end jobs (where the installer is working alone and rushing batches to keep ahead of the cure clock).
How to spec around it: ask how the installer handles batch sizing and timing, and how many batches per day they typically use on a project the size of yours. The right answer is "small batches, 30–45 minutes pot life, never re-watered, batches blended before use if mixing across production lots." A typical 5 m² wall takes 5–8 batches over the course of a day; a "we mix it all in one go" answer is a flag.
The fix when it does happen: this one is harder than cracks because the issue is in the finish coats themselves, not in a single line. Sometimes a deeper sealer can mask mild blotching; for visible blotching the answer is to sand back the finish coats and re-apply. Cost typically £35–£60/m² for re-finishing only (the base coats stay), so a 20 m² wall might be £700–£1,200 to put right. Worth pushing the original installer for a remediation quote if they're still trading.
Sealer failure — chalking, lifting, or yellowing
What it looks like: three different visible failure modes, all sealer-related but with different causes. Chalking: white powdery residue when wiping the surface, which keeps coming back after each clean. The sealer is degrading and shedding tiny particles. Most visible on darker colours where the white shows. Lifting: patches where the sealer has visibly delaminated from the microcement underneath — sometimes you can lift a flap of clear film with a fingernail. Yellowing: the surface darkening unevenly over months, especially on lighter colours and especially in zones with sustained UV exposure or thermal cycling.
What actually caused it: usually the wrong sealer for the conditions. Standard interior PU sealers don't hold up in wet zones — they need the wet-zone variant with chlorine-tolerant chemistry. Standard sealers in heated rooms can yellow under repeated thermal cycling. Aromatic-based PU sealers (cheaper) yellow under UV; aliphatic-based PUs (more expensive) stay colour-stable. Sealer applied over a still-curing base coat can lift later because the base finishes curing and pushes the sealer off. Sealer applied too thickly in a single coat — instead of two thinner coats with intermediate sanding for adhesion — can also lift.
How frequent is it: ~12% of issue call-outs, but the proportion is rising as more bathroom microcement gets installed by contractors who haven't worked in wet zones before.
How to spec around it: ask which sealer system is being used and whether it's specifically rated for the conditions of your project (wet zone, heated floor, kitchen worktop, full-immersion pool). The same sealer for every job is a flag. For UV-heavy rooms (south-facing kitchens, sunrooms, pool surrounds), specifically ask whether the sealer is aliphatic. The wet-zone-specific spec is covered in microcement bathrooms; the heated-floor variant in underfloor heating compatibility.
The fix when it does happen: most sealer failures need the entire sealer stripped and re-applied. Stripping uses a chemical stripper and sanding (the underlying microcement is fine, only the top coat goes); re-application is the same two-coat sequence as a new install. Typical cost £25–£45/m². For a localised lifting (a small patch in one corner), spot-strip and re-seal is feasible at lower cost. Yellowing on aromatic sealers cannot be fully reversed — only stripping and replacing with aliphatic sealer fixes it.
Slow or sticky cure
What it looks like: a finish that should be hard and walk-on-able after 24 hours but is still tacky to the touch a week later. Footprints leaving permanent impressions in floors. Shoe scuff marks that you can't wipe off because the surface is sticky. Furniture legs adhering to the surface they sit on. The visible surface looks fine; the problem is in the cure.
What actually caused it: usually one of two things. Sealer mixing-ratio error: the two-part polyurethane is mixed off-ratio (e.g. 3:1 instead of the specified 4:1) — the components don't fully crosslink and the surface stays partially uncured forever. Wrong temperature/humidity for cure: PU sealers cure poorly below 10°C or above 80% relative humidity. Winter installs in unheated rooms, summer installs during a humid weather window, or installs where someone left a window open and the room temperature dropped overnight. Less commonly: contamination of the sealer from cleaning solvents on the substrate, or sealer applied over a still-uncured base coat.
How frequent is it: ~8% of issue call-outs, with a strong winter skew. We see almost no slow-cure issues in May–August installs; ~25% of January–February installs have some level of slow-cure if temperature isn't actively managed.
How to spec around it: ask if the installer uses calibrated dispensers for the sealer parts (eyeballed ratios are a leading cause). For winter installs, ask whether the room will be kept above 15°C through the cure window — that's portable heating, sealed door openings, and a temperature/humidity log on site. For summer installs in humid weather, ask about dehumidification. For any install ask about the temperature target during sealer application and the 14-day cure window after. A vague answer here is a flag.
The fix when it does happen: sometimes you can recover by getting the room to ideal conditions (heat to 20°C, dehumidify to 50% RH) and waiting another 1–2 weeks; sometimes the sealer never cures and has to be stripped and re-applied. The diagnostic is whether the surface has hardened at all in the first 7 days — if there's any progress, full cure usually happens given the right conditions; if it's exactly the same after 7 days, it's probably a mix-ratio error and stripping is the only fix.
Debonding — microcement coming away from the substrate
What it looks like: the microcement lifts from the wall or floor in sheets, sometimes with a clean separation (the underside of the microcement is smooth and the substrate is clean), sometimes with bits of substrate attached (a layer of paint or a flake of plaster came up with it). On floors the sheets often come up at edges first — near doorways, around fixtures — then progress inwards. Tap-test along the surface will sound hollow over the failing area before any visible lifting starts. Sometimes there's an audible crack as you walk over a debonded patch on a floor.
What actually caused it: wrong primer for the substrate, or no primer at all. Different substrates (tile, plasterboard, screed, ply, MDF, concrete, marble, terrazzo) need different primer chemistry — broadly, absorbent substrates need a consolidating primer that soaks in and bonds to the porosity, non-absorbent substrates need a chemical-grip primer that creates a bondable surface where nothing would otherwise stick. Wrong primer = no mechanical or chemical bond, and the microcement is essentially held on by surface tension alone, peeling off the first time the substrate moves. Less common causes: applying over a contaminated surface (release agent, oils, dust between primer and base coat), or applying at the wrong primer-to-base-coat timing (the primer needs to be tacky-but-not-fully-dry when the base coat goes on).
How frequent is it: ~10% of issue call-outs. Almost always traces back to a single mistake at survey: substrate not properly identified, or one-primer-for-everything used.
How to spec around it: ask which primer is specified for your substrate type, and confirm the answer matches what's actually under the proposed install. There should be a different answer for tile (a non-absorbent primer like Primer Plus or equivalent), plaster (a consolidating primer), screed (an absorbent-screed primer), and any plywood or MDF substrate (a wood-specific primer). One-primer-fits-all is a flag. Ask whether the installer has worked on your specific substrate type before.
The fix when it does happen: this is the most expensive issue to fix because the entire microcement system has to come off the affected area. Mechanically remove the debonded sheets, prep the substrate properly with the correct primer, re-apply the full system from scratch (primer + mesh + base coats + finish coats + sealer). Cost is roughly the same as the original install for the affected area, plus the demolition cost — typically £150–£250/m² total for a re-do. Sometimes the entire wall or floor has to come off because debonding tends to spread once it starts.
Efflorescence — white salt deposits
What it looks like: white powdery deposits appearing on the surface, especially around the perimeter of a room (near skirtings or wall-to-floor junctions), on walls that share a face with outside, or in any area with a known damp history. Wipes off with a damp cloth but comes back within days. Sometimes accompanied by faint surface blistering or a slight tonal shift in the microcement near the affected zone.
What actually caused it: rising damp, penetrating damp, residual moisture in a fresh screed or plaster, or trapped moisture from a hidden leak. The dissolved salts in the wet substrate dissolve, migrate up through the microcement to the sealed surface, and crystallise as the moisture evaporates through the sealer. The microcement is a victim, not the cause — but applying microcement over damp makes the damp problem visible at the surface much faster than paint or wallpaper would, because the sealed surface concentrates evaporation at the perimeter where the salts gather.
How frequent is it: ~8% of issue call-outs, almost always in older buildings (pre-1940s solid masonry, no DPC, or DPC compromised), or in new-build projects where the screed was applied but not given enough time to cure before microcement went on top.
How to spec around it: damp survey before specifying microcement on any wall that touches the ground, any wall that has previously had peeling paint or salt deposits, or any below-grade space (basement, sunken family room). For new-build screeds, the moisture-content rule (below 3% before microcement application, verified with a calibrated meter) is non-negotiable. Fix the damp first; then microcement. The worst thing to do is press on regardless and seal the damp under a waterproof finish — that just hides it for a while and makes the eventual repair much worse.
The fix when it does happen: the damp itself has to be diagnosed and remediated first — DPC injection, exterior render repair, leak isolation, screed dry-out time. Until that's done, any patch on the microcement will just push the efflorescence to a new zone. Once the underlying damp is fixed and the substrate has dried, the microcement can usually be repaired locally with a wash, primer refresh and re-seal. If salts have crystallised within the microcement layers themselves (the layers feel powdery to the touch), that affected area has to come off and be re-applied.
What to do when one of these does happen
Localised issues — single crack, single blotchy area, sealer chip in one spot — can almost always be repaired in place without a full re-install. The repair sequence is the same regardless of which issue it is:
- Identify the cause. The single most important step. Patching a crack without fixing the missing mesh underneath just moves the crack to a new spot. Patching a blotch without fixing the batching discipline issues just produces more blotches in the next finish coat. The diagnostic is what determines whether the fix is local or whole-surface.
- Mechanically remove the failing area down to a clean substrate. For cracks this is raking out the crack; for blotching it's sanding back the affected finish coats; for debonding it's lifting the debonded sheets; for sealer failure it's stripping the sealer. The area removed extends 50–100 mm beyond the visible failure to ensure clean edges.
- Re-prime, re-mesh if needed, re-apply matching base + finish coats. The base coats are usually re-applied even if only the finish coats are failing, because the bond between old finish coats and new finish coats is much weaker than two coats applied within the original system's working window.
- Re-seal the affected area with the correct sealer for the conditions. Edges are feathered to blend with surrounding sealer where possible.
Hand-applied finishes hide patches well — most local repairs are invisible after the sealer cures, especially on stucco and coarse finishes. Smooth finishes are slightly less forgiving because there's no texture to vary the patch into the surrounding surface; sometimes a smooth-finish repair has a faint shadow visible only under directional light. Local repair costs typically £400–£1,200 depending on the issue and area.
The exception is whole-surface issues: a wrong-sealer-everywhere job, a no-mesh-anywhere floor, a wrong-primer-on-tile install. These have to be addressed at the whole-surface level — usually stripping and re-applying the sealer or the entire system. Whole-room re-installs cost roughly the same as the original install, sometimes more because of demolition costs.
One thing worth saying clearly: most issues we see were preventable at the spec stage by a competent installer asking the right questions and refusing the wrong substrate. If you're unhappy with an existing microcement install, the first conversation is with the original installer — most reputable contractors will return to remediate genuine workmanship issues within their warranty period at no cost. Second-opinion remediation by a different installer is a last resort and typically more expensive than the original work.
The spec that prevents all of it
Almost all of the above is preventable at the spec stage. The vast majority of issues we get called to inspect would have been avoided by a single quote-stage conversation that confirmed the right materials and the right time. A microcement quote that's worth signing should include all of:
- Substrate type identified at survey, and primer specified accordingly. Different chemistry for tile vs plaster vs screed vs ply.
- Fibreglass mesh embedded in the primer layer, with overlap details (10 cm minimum), corner reinforcement strips at internal corners, and weight specified for the conditions (60 g/m² for unheated walls, 100 g/m² for heated floors and any high-movement substrate).
- Number of base coats and finish coats explicit — typically 2 base + 2 finish, with cure times between each.
- Sealer system identified by name and grade — wet-zone PU for bathrooms, heated-floor PU for UFH rooms, kitchen-grade for worktops, pool-grade for full-immersion. Same sealer for everything is a flag.
- Cure time between coats explicit — typically 24 hours minimum per coat, longer for sealer between coats. The total project timeline reflects the cure stages, not just the application time.
- Total project timeline of 7–14 working days for a typical residential project. Anything under 7 days is skipping cure stages.
- Substrate moisture testing with a calibrated meter, for screeds, plaster and timber substrates. 3% maximum moisture content before priming.
- Application conditions — room temperature 15°C+ during application and through the 14-day cure window after sealer; humidity below 70% RH; no opening windows mid-cure.
- Warranty covering at least the workmanship (5 years minimum), ideally with the materials manufacturer's warranty pass-through (10–15 years on the materials themselves).
- Damp survey if any wall touches the ground, has had previous damp issues, or sits below grade.
A quote that's missing any of those isn't a complete quote. The questions to ask are in the FAQ; what realistic pricing looks like (and what suspiciously low quotes have skipped) is in the cost guide; the kit and skill that goes into a proper install is in the microcement toolkit.
The single biggest predictor of whether a microcement install will hold for fifteen years or fail in eighteen months: whether the installer is willing to refuse a job that isn't ready. The right answer to "the screed isn't dry yet but we need the room ready in two weeks" is "we'll come back when it's dry," not "we'll go ahead and hope for the best." A contractor who pushes ahead on a substrate they shouldn't is the one whose work will appear in this article a year from now.
