Sand-and-Cement Screed: Mix Ratios, Depths, Drying Times, and When to Choose It

Your builder quotes "sand-and-cement screed" as a provisional line item for your extension floor. It sounds like a sensible default. Six months later, your underfloor heating is sluggish, the bedroom above the kitchen is always cold, and the porcelain tiles you laid are lifting at the grout lines because the screed underneath them was still wet when the tiler started. Every one of those failures is predictable from the product choice. Sand-and-cement screed is cheap, traditional, and still right for some jobs. Picking it for the wrong job is one of the most expensive mistakes in a kitchen extension.

What it is and what it's for

Sand-and-cement screed is a semi-dry mortar mix, laid on a sub-floor as a smooth level bed for your final floor finish (tiles, vinyl, engineered wood, or carpet). The mix is 4 parts sharp sand to 1 part Portland cement, wetted just enough to bind, then hand-laid and compacted between timber or aluminium rails with a wooden float.

It is governed in the UK by BS 8204-1:2003+A1:2009 (Screeds, bases and in-situ floorings), which sets out minimum depths, surface regularity classes, and material specifications. Any competent screeder or flooring contractor will specify against this standard. When a product datasheet or contractor quote references "BS 8204-1 compliant", this is what it means.

The consistency is important to understand. Sand-and-cement screed is not fluid. It arrives in a wheelbarrow looking like damp, earthy soil. A handful squeezed tightly holds shape without water weeping out. This is called semi-dry consistency and the test has a name: the squeeze test. If water runs out of the squeezed handful, the mix is too wet and the cured screed will be weak and prone to shrinkage cracking.

The finished material achieves 25-30 N/mm² compressive strength at a 4:1 mix (the domestic standard), or 30-40 N/mm² at a 3:1 mix (used for high-traffic commercial and garage floors). This is the product's strength, not its flatness. Flatness is a separate specification measured under a 2m straightedge, covered later.

Click to enlarge

Mix ratios, sand, and why building sand wrecks a floor

There are two mix ratios you need to know: 4:1 for domestic, 3:1 for heavy-duty. The difference matters more than most homeowners realise.

The most common DIY mistake is using the wrong sand. Screed must use sharp sand (also called grit sand or concreting sand), conforming to BS EN 13139. Sharp sand has angular, coarse particles that interlock under compaction, producing a dense, strong floor.

Building sand (soft sand, bricklaying sand) is fundamentally different. Its particles are rounded and fine, which is perfect for a workable mortar that presses neatly into brickwork joints. But those same rounded particles slide against each other in a screed mix, producing a weak, dusty floor that cracks within months. Merchants stock the two side by side in suspiciously similar bulk bags. Specifying the wrong one is depressingly common.

Depth requirements: the BS 8204-1 minimums

BS 8204-1 sets minimum screed thicknesses by application type. Get the depth wrong and the screed either debonds (too thin) or wastes material and takes forever to dry (too thick). These are the numbers your contractor should be working to:

The floating floor with UFH case is the one that causes the most confusion. For a floating screed on rigid insulation the minimum is 65 mm, and with UFH pipes underneath you need at least 65mm of screed above the insulation. That is the screed depth, not the total floor build-up. Under that screed you still have rigid insulation (typically 100-150mm of PIR), a DPM, and possibly a perimeter expansion strip. Your finished floor height above the sub-slab will be more like 175-250mm once everything is stacked.

Curing vs drying: the distinction that wrecks flooring programmes

This is the single most misunderstood point about sand-and-cement screed, and it's where most tiling failures originate. Curing and drying are different processes.

Curing is the chemical reaction. Cement reacts with water over 28 days to form the crystalline structure that gives the screed its strength. Covering the screed with polythene does not harm curing, in fact it helps by preventing rapid surface moisture loss. Temperature must stay above 5°C throughout (hydration stalls below that). After that period, chemical curing is effectively complete and the screed has reached full strength.

Drying is the evaporation of excess mixing water from the screed mass. This is the process that governs when you can put a floor covering down. Drying happens at an uncomfortably slow rate:

• First 40-50mm of screed depth: approximately 1mm/day
• Above 40mm, drying slows significantly to approximately 0.5mm/day (roughly 2 days per mm), per BS 8203 and Homebuilding & Renovating guidance. The moisture path to the surface lengthens as depth increases, which is why thicker pours take disproportionately longer to reach a flooring-ready moisture level.

Work through what this means for a 75mm floating screed over UFH. The first 40mm takes 40 days. The remaining 35mm takes another 70 days. Total: roughly 110 days (about 16 weeks) to reach the moisture threshold for most floor coverings. For engineered timber or LVT, which require below 65% RH, it's closer to four months.

Your builder will almost certainly not have planned for this. Quotes routinely assume "the tiler can start in a few weeks". Push back and plan the programme around the drying time, not against it.

Protecting the pour in the first 72 hours

The first three days after the pour are when surface damage is most likely. Direct sun or a warm draught dries the surface faster than the core, causing the top to shrink against a wetter base. The result is curling at edges and hairline cracking across the slab.

Protection rules for the first 72 hours:

• Close windows and doors. No draughts.
• Cover the screed with polythene sheeting if there's any risk of rapid drying.
• No gas heaters anywhere near the screed. Gas combustion produces water vapour which stalls drying, and the direct heat causes surface cracking.
• Electric heaters are acceptable but shouldn't be pointed at the floor.
• Dehumidifiers are effective, but only after 14 days. Using them too early pulls moisture from the surface faster than the core can release it, and you get curling.

Why it's the wrong choice for UFH (and when it's still fine)

Liquid anhydrite screed is pumped from a mobile mixer as a genuinely fluid product. It flows around and under the UFH pipework, encapsulating 100% of the pipe surface in direct thermal contact. Sand-and-cement screed cannot do this. It's hand-compacted as a semi-dry mix, and no matter how carefully the screeder tamps it, air voids form beneath and around the pipes.

The numbers matter. With sand-and-cement screed over UFH, typical pipe encapsulation is around 60%. The remaining 40% of pipe surface area is in contact with air voids instead of screed. Air is an insulator. The consequences stack up:

• Cold spots on the finished floor directly above pipe voids. Heat transfers unevenly across the room.
• Higher flow temperatures needed to reach target floor surface temperature, increasing gas or electricity consumption by 10-20% over the life of the system.
• Slower heat-up times from cold start. Sand-and-cement has a thermal conductivity of around 1.4 W/mK, compared to anhydrite at 2.0 W/mK. Anhydrite warms a room 15-20% faster for the same pipe spacing and flow temperature.
• Deeper screed required: 65-75mm minimum for sand-and-cement versus 40-50mm achievable with flowing screeds. More thermal mass means slower UFH response.

Click to enlarge

That said, sand-and-cement is still the correct choice in these specific cases:

1. No UFH installed. The encapsulation and conductivity advantages of anhydrite don't apply. Anhydrite's premium price buys you nothing.
2. Area under 40m². Liquid anhydrite pours are pumped from a dedicated mixer truck. The fixed mobilisation cost (typically several hundred pounds for the pump and truck setup) makes small pours economically pointless. Hand-mixing a 20m² utility floor is straightforward.
3. Wet room with falls. Liquid screed self-levels by definition. It cannot be laid to falls for a shower drain. Sand-and-cement is the only option when you need a sloped finish to a linear drain or gully.
4. Budget is primary. Sand-cement is materially cheaper than liquid anhydrite on a supply basis, see the Cost section below.
5. Future repair is likely. Sand-and-cement is easy to grind and patch with standard cementitious repair products. Anhydrite is incompatible with Portland cement-based products, which makes repairs specialist work.
6. Floor finish is tile or carpet. These finishes tolerate SR2 surface regularity (±5mm under 2m). LVT and engineered timber demand SR1 (±3mm) which usually requires a self-levelling overlay on top of sand-and-cement.

See liquid anhydrite screed for the alternative's full specification and cost profile.

Surface regularity: SR classes and why they matter

BS 8204-1 defines three surface regularity classes, measured under a 2m straightedge:

• SR1: maximum deviation ±3mm. Highest flatness. Required for LVT, engineered timber, large-format tiles (600mm+).
• SR2: maximum deviation ±5mm. Standard for most ceramic tiles and carpet. What sand-and-cement typically achieves with a competent screeder.
• SR3: maximum deviation ±10mm. Acceptable only where a self-levelling compound or thick adhesive bed will correct the surface before flooring.

Sand-and-cement readily achieves SR2. Reaching SR1 is very difficult with a hand-laid screed; professional screeders manage it only with experienced teams, careful rail setup, and skilled long-board finishing. For most homeowners the realistic plan is SR2 followed by a 3-5mm self-levelling compound overlay wherever SR1 is needed for the final floor finish.

Specify the SR class in your flooring brief. If the contractor cannot guarantee SR1 and your final finish needs it, budget for an SLC overlay up front. Expect to add a small extra per-m² in materials on top of the base screed cost, enough to matter on a whole-house footprint but not on a single utility room.

Additives that actually matter

Three additives are worth knowing about. The rest are marketing.

SBR latex (Styrene-Butadiene Rubber). Dual purpose: a bonding primer for bonded screeds, and a mix additive that improves flexibility and water resistance. Applied as a slurry onto hardened concrete before the screed is placed, then the screed goes down while the primer is still tacky. Typical dosage as a mix additive is 5 litres of SBR per 25kg of cement, used 1:1 with water as the gauging liquid. Essential for bonded screeds, thin-section pours, and wet rooms. Named products: Bostik SBR, BAL Bond SBR, Flowcrete Isopol SBR.

Polypropylene fibres. Short (6-18mm) synthetic fibres added to the mixer at around 0.6-0.9kg per m³ of screed. They distribute tensile stress as the screed dries, dramatically reducing plastic shrinkage cracking. They replace A142 mesh in most domestic applications (saving labour and reducing the risk of mesh voids). Worth specifying on any UFH screed because of thermal movement, and on any unbonded floating screed where curling risk is high.

General-purpose plasticiser. Improves workability without adding excess water. This is the additive that lets a screeder produce a consistent semi-dry mix without fudging the water content. Typical dosage is 100-200ml per 25kg of cement, per the manufacturer's datasheet. A good contractor will use one by default.

What's not worth it: "rapid-dry" cement replacements that claim to get the floor tiled in 7 days. They exist (products like Cemfloor rapid and Ardex A38) but they're specialist cementitious screeds, not additives to standard sand-and-cement, and cost approaches liquid screed. If you need to tile quickly, use liquid anhydrite or a proprietary rapid-drying system, not a sand-and-cement mix with additives.

How much do you need

A 1:4 mix at 75mm depth over 1m² works out to roughly 0.075 m³ of compacted screed. Before compaction, allow 10% for waste and shrinkage.

The 25kg bag route is only sensible for small rooms. For 30m² or more, order sharp sand by the bulk bag (roughly 850kg each) and cement by the 25kg sack. Sharp sand bulk bags are typically 40-50% cheaper per tonne than 25kg bags. For 50m² pours, consider a loose tipper load of sand delivered to a clean hard standing (around 1 tonne is enough).

DIY mix materials for the 30m² at 75mm example:

• 34 x 25kg cement bags at £6.50–8.50
• 3 x 850kg sharp sand bulk bags at £50 – £80
• Or 101 x 25kg sharp sand bags at £2.80 – £4.50
• Fibres, SBR, polythene sheeting, rail timber: modest allow-for budget
• Total DIY materials: mid to high three figures depending on sand format chosen

Cost: material only vs installed

The graph price for sand-and-cement screed is £11–18/m². This is the material-only figure, useful if you're hand-mixing yourself or comparing merchant quotes for supply. The installed figure is materially higher once you add screeder labour.

for raw material. Professional supply and install adds a meaningful increment on top of the material-only rate once you factor in screeder day rates for a two-person team; published 2026 quotes from Property Workshop, MyJobQuote, and PriceYourJob broadly converge on an installed rate somewhat above the material figure.

For typical extension floor sizes, total installed costs look like:

• 30m² floor: £330–540 material only; a few hundred pounds more on a supply-and-install basis once labour and ancillaries are added.
• 50m² floor: £800–£1,500 material only; expect a supply-and-install total into the low four figures for a competent two-person team on a straightforward pour.

Compare this against liquid anhydrite at £18–34/m² installed. The premium for going liquid over sand-and-cement is typically 40-80% more for the screed stage, but you save on UFH running costs over the life of the heating system, get a faster dry-out (floors tileable in 28-40 days rather than 100+), and achieve SR2 flatness reliably.

Where to buy the materials

Sharp sand and cement are stocked by every UK builders' merchant. For bulk bag sand, Travis Perkins, Jewson, Wickes, and MKM Building Supplies are the mainstream choices. Independent local merchants are usually 10-20% cheaper on bulk bags. Cement in 25kg bags is priced broadly the same across Screwfix, Wickes, Toolstation, and the builders' merchants; trade accounts save 5-15%.

Bulk sand tipper deliveries (1-10 tonnes loose, tipped on your driveway) come from aggregate suppliers like CEMEX, Tarmac, and regional quarries. Cheaper per tonne than bulk bags but you need hard standing to tip onto and a wheelbarrow route to the screed area.

Alternatives

Liquid anhydrite screed is the main alternative and the preferred choice for any floor with underfloor heating over 40m². It's poured as a fluid, achieves 100% pipe encapsulation, dries faster (about 40 days at 50mm to flooring-ready moisture), and reliably hits SR2. Installed cost carries a meaningful premium over sand-and-cement on the screed line item alone; see the Cost section above for the canonical range.

Self-levelling compound is not a screed substitute at structural depths, but is often used as a 3-10mm correction layer on top of an SR2 sand-and-cement screed to achieve SR1 for LVT or engineered timber. Budget £500 – £600 for a full SLC overlay supplied and applied by a screeder; for DIY application, see the materials-cost note in the Surface Regularity section above.

Proprietary rapid-drying screeds (Cemfloor Rapid, Ardex A38, Mapei Topcem) sit between sand-and-cement and anhydrite. They cure chemically like cement but dry much faster (floor coverings in 7-14 days), making them useful when programme is tight. Expect to pay a meaningful premium over standard sand-and-cement on an installed basis.

Where you'll need this

• Screeding - the pour, including substrate prep, expansion joint layout, and protection during curing
• Underfloor heating - the screed directly encapsulates the UFH pipework, so screed choice determines heating performance
• Flooring - moisture testing before any floor covering is applied, and SR class assessment for the final finish
• Tiling - laitance removal, primer selection, and moisture testing before tile adhesive goes down

Screed appears at the first-fix stage of any extension or renovation project where a new floor is being created. The links above point to kitchen extension tasks, but the guidance transfers directly to loft conversions, garage conversions, and single-storey rear extensions.

Common mistakes

Using building sand instead of sharp sand. The single most damaging error. Rounded soft sand particles produce a weak, dusty screed that fails within a year. Always specify washed sharp sand to BS EN 13139 in writing on the quote.

Tiling before moisture testing. Screed feels solid in a week, but holds flooring-critical moisture for 3-4 months on a 75mm floating pour. Fitting tile adhesive onto a damp screed causes adhesive failure, tile lift at the grout lines, and trapped moisture that eventually migrates into the wall base plate. Test with a calibrated hygrometer or CM test and get the result in writing. BS 8203 requires it. The minimum chemical cure time before any tiling is 28 days, and on a thick floating pour you will be waiting considerably longer for drying on top of that.

Too wet a mix. More water makes the screed easier to spread, but dramatically weakens it and causes shrinkage cracking. The squeeze test (handful holds shape without weeping) is the control. Any screeder adding water beyond that point is cutting corners.

Skipping the primer on bonded screeds. Laying screed directly onto dry concrete without an SBR slurry primer causes delamination within weeks. The screed separates cleanly from the substrate in sheets. Prime the substrate and place the screed while the primer is still tacky.

Laying below 5°C without frost protection. Cement hydration stalls below 5°C. Screeds laid in cold weather without heating, insulated blankets, or an enclosed space will be weak or fail to set. In winter, postpone the pour or enclose and heat the area (electric only, no gas).

Firing UFH too early or too aggressively. Heating the floor at full temperature on a fresh screed causes thermal shock cracking. Minimum wait is 21 days after pour, then heat up gradually: 5°C per day rise from cold until operating temperature reached. The UFH commissioning schedule should be written into the heating installer's contract.

Missing expansion joints. Screeds over 40m² continuous, or any screed with UFH, need movement joints at perimeters and around fixed objects (columns, door thresholds). Without joints, thermal and drying movement cracks the screed along its longest dimension. A 10-20mm perimeter expansion strip (compressible foam) is standard; internal joints are marked and cut before tiling.

Ignoring the laitance layer before tiling. A weak, dusty film of cement fines (laitance) forms on every sand-and-cement screed as the surface dries. This layer is too friable for tile adhesive. Before tiling, abrade the surface with a stiff brush or mechanical sander (60-grit), vacuum thoroughly, and prime with an acrylic primer. Skip this and your tile adhesive bonds to dust, not screed.

DIY mixing at scale. A 50m² pour at 75mm needs roughly 4 tonnes of mixed material. In a standard drum mixer that's 80-100 individual batches, and wheelbarrowing it into position is a full day for two people before you start laying. Mix consistency varies batch to batch, producing weak patches. Hire a professional screeder for anything over 20-30m². For small utility or wet rooms, DIY mixing is workable.