Screed and Underfloor Heating: Why the Drying-Time Numbers Don't Add Up (and What to Trust Instead)
Somewhere between booking the screed pump and commissioning the underfloor heating, most self-builders hit the same wall: nobody can tell them, with any confidence, when the floor will actually be ready. One supplier's page says a week per centimetre. Another gives several contradictory numbers on the same page, ranging from two days to 49 days.
That is not a typo problem. It is a structural one. Drying-time claims for screed over underfloor heating are published as if they were fixed facts, when in practice they depend on thickness, temperature, humidity and airflow — variables no generic rule of thumb can capture. Even trade sources that specialise in screed contradict themselves within a single page.
This piece lays out where those contradictions sit, works through one realistic floor to show how wide the spread really gets, and sets out what to check on site instead of counting days on a calendar.
Key Takeaways
- A 50mm anhydrite screed — a realistic depth for covering underfloor heating pipework — draws estimates from 35 days to roughly 60 days depending on which thickness-based rule is applied, while Barrow Mix Concrete separately publishes 48 hours and 49 days for the product without stating the conditions that separate them.
- Anhydrite screed's gypsum base is moisture-sensitive, so it's generally avoided in bathrooms and wetrooms unless protected with appropriate waterproofing or a specific primer.
- Cement-based liquid screed is self-levelling and UFH-compatible with generally shorter drying windows than traditional sand and cement screed, which conducts heat less efficiently.
- Thickness, temperature, humidity and airflow move the real drying time more than any formula does — which is why both trade sources reviewed here agree on one point: moisture testing, not elapsed time, is what actually confirms a floor is ready for coverings.
- Faster is not automatically better: screed is deliberately dried slowly in many cases, because rapid drying risks cracking.
Screed Types and Underfloor Heating: What Actually Differs
Before any drying-time number can mean anything, it helps to know which screed is being discussed. Three types dominate underfloor heating installations, according to Clockwork Screed's technical guidance: traditional sand and cement screed, anhydrite (gypsum-based) screed, and cement-based liquid screed.
The differences between them are not contested in the same way the drying times are. Anhydrite is widely used under underfloor heating because of its thermal conductivity, which helps heat move efficiently from the pipework to the floor surface — but its gypsum content is susceptible to moisture, so it is generally not recommended for bathrooms or wetrooms unless protected with suitable waterproofing, a moisture-resistant primer, or similar measures. Cement-based liquid screed is also self-levelling and UFH-compatible, with good thermal conductivity, a faster heat response, and generally shorter drying times than traditional sand and cement screed. Traditional sand and cement screed, by contrast, is described as less efficient under underfloor heating because of its lower thermal conductivity.
There is a practical install difference too: cement-based liquid screed typically cures with little surface laitance or dust, though compatibility with a given adhesive — including fast-track systems — still depends on residual moisture, primer choice and the manufacturer's own guidance. Anhydrite screed, on the other hand, often needs surface preparation such as sanding to remove laitance before some floor finishes go down, though the exact requirement depends on the product and finish — a step that sits outside the drying clock entirely but still has to happen before flooring work can start.
| Screed type | UFH thermal behaviour | Moisture tolerance | Typical thickness / finish prep |
|---|---|---|---|
| Traditional sand & cement | Lower thermal conductivity, less efficient with UFH | Reasonable, but not the focus of UFH-specific guidance | 25–75mm; standard finish prep |
| Anhydrite (gypsum-based) | Good thermal conductivity, well suited to UFH | Moisture-sensitive; generally avoided in bathrooms/wetrooms unless protected | 25–40mm; often needs surface prep (e.g. sanding) before some finishes |
| Cement-based liquid screed | Good thermal conductivity, faster heat response | Compatible with fast-track adhesive systems | Self-levelling; typically little laitance or dust after curing, though surface prep may still be needed |
Thickness ranges and finish-prep details draw on Barrow Mix Concrete's and Clockwork Screed's published guidance.
That table is where the agreement ends. Once you ask how long each of these takes to dry, the published numbers start pulling in different directions.
The 50mm Anhydrite Problem: Why the Same Floor Gets Four Different Deadlines
Take a single, realistic case: a 50mm anhydrite screed, thick enough to fully cover underfloor heating pipework. Apply the published formulas to that one floor, and the result is not one answer but four.
Clockwork Screed's stated rule is that anhydrite screeds dry at roughly one week per 10mm of thickness, requiring at least 28 days to cure fully. Applied to 50mm, that works out to five weeks — about 35 days.
Barrow Mix Concrete gives a different mechanism entirely: anhydrite dries at about 1mm per day under optimal conditions (20°C, relative humidity below 65%), but for anything thicker than 40mm, add a further 2 days per extra millimetre. Run that through for 50mm: the first 40mm at 1mm/day costs 40 days, and the remaining 10mm at the 2-day penalty adds another 20 days. Total: roughly 60 days — nearly double Clockwork's figure for an identical thickness.
Barrow Mix Concrete also publishes a general top-line figure for the same product, separate from either formula: anhydrite screeds, it states, can take about 49 days to dry, against roughly 21 days for Ordinary Portland Cement. That is a third number, and it does not match either of the thickness-based calculations above.
Then there is the fourth figure, from the same source: anhydrite screed becomes walkable in 24 hours and ready for floor finishes within 48 hours under ideal conditions. Forty-eight hours and 49 days are both stated by Barrow Mix Concrete, for what reads as the same product, without the page distinguishing which conditions separate one outcome from the other.
We are not picking a winner here. Thirty-five days, 49 days, roughly 60 days and 48 hours are four outcomes drawn from sources that elsewhere give even shorter windows for the same product — and that spread is itself the finding. None of these figures can be used as a standalone scheduling input.
The pattern repeats on Clockwork Screed's own page, though not as a true contradiction: the 28-day figure quoted above belongs specifically to anhydrite, while the same page's FAQ gives cementitious liquid screed a separate, faster timeline — 24–48 hours for initial setting and 14–21 days for full readiness. Two different screed types, two different numbers, correctly attributed; the risk is a reader conflating them. Combined with Barrow Mix Concrete's genuine 48-hour-versus-49-day spread for the same anhydrite product, the pattern is clear: always check which screed, and which milestone, a figure refers to before trusting it.
What Actually Controls How Fast Screed Dries
The reason fixed-day rules keep breaking down is that drying is not a property of the material alone — it is a property of the material in a specific environment. Temperature, humidity and airflow directly affect how quickly screed dries: warmer temperatures and lower humidity speed the process, and adequate airflow across the surface is essential. Two identical 50mm pours, in two different rooms, can genuinely dry at different rates.
Faster is also not always the goal. As Clockwork Screed puts it, "Screed flooring should be allowed to dry slowly (or cure). This is because the screed needs to retain its moisture for as long as possible to avoid the cracks that can occur when dried rapidly." On some sites, a polythene sheet is used deliberately to slow the process down, not speed it up.
Underfloor heating itself complicates the picture further. Running the system can speed up anhydrite drying, which is why it is popular on tight schedules — but Barrow Mix Concrete is explicit that industrial dehumidifiers and underfloor heating must stay within recommended parameters, or there is a real risk of applying floor finishes too soon. In other words, using heat to accelerate drying only works safely if it is managed, not simply switched on and left.
What to Trust Instead: A Field Checklist
Given how little a generic day-count can be trusted, the practical answer is to stop treating elapsed time as the deliverable. Both Clockwork Screed and Barrow Mix Concrete converge on the same underlying point despite all their other disagreements: moisture testing must confirm readiness for floor coverings, regardless of screed type, rather than relying on elapsed time alone.
A workable sequence looks like this:
- Check the specific product's technical data sheet for the screed actually being poured — not a generic blog rule of thumb that may describe a different thickness or mix.
- Control temperature, humidity and airflow on site rather than assume ideal conditions; the gap between "optimal" and "actual" is exactly where the published formulas fall apart.
- If using underfloor heating to accelerate drying, follow a documented commissioning sequence rather than guesswork — the risk is covering the floor before it is genuinely ready.
- Moisture-test before covering, whatever the calendar says. A hygrometer or carbide reading tells you the actual state of the floor; a day-count does not.
- Budget contingency time. Thickness and ambient conditions move the real number more than any formula can, so a schedule built on a single published figure is a schedule built on sand.
One more hedge worth keeping in mind: these figures come from UK trade sources. Always check the specific product's technical data sheet, and outside the UK, the applicable local standard, before using any of these numbers for scheduling.
Frequently Asked Questions
Can I just use underfloor heating to force-dry the screed faster? You can accelerate anhydrite drying this way, and it is a common approach on tight schedules. The condition is staying within the manufacturer's recommended parameters for the heating and any dehumidifiers used — going beyond them risks covering the floor before it has actually dried out, which can trap moisture under the finish.
Does the one-week-per-10mm drying rule apply to cement-based liquid screed too, or only anhydrite? Only anhydrite. Clockwork Screed's guidance ties that rule specifically to anhydrite screed, including its roughly 28-day minimum cure time. Cement-based liquid screed follows a different curve on the same page: typically 24–48 hours for initial setting and 14–21 days for full readiness, which is part of why it's often chosen when the schedule is tight.
Why do two screed suppliers give different drying times for what sounds like the same product? Usually because they are describing different things: a different thickness, a different ambient condition, or a different stage — walkable versus ready for floor finishes versus fully cured for sensitive floor coverings are three separate milestones, often collapsed into one headline number.
Conclusion
The real gap in this market is not screed chemistry — it is the absence of a shared, independently verified way of disclosing drying times. On the evidence of the two sources reviewed here, a headline "X days" claim functions as an estimate, not a commissioning fact, until it's checked against site conditions. For now, the question worth asking on site is not how many days have passed, but who last checked the moisture reading — and against what.
