The question that started one of our own brands wasn’t a product question. It was a material question: the ceramics factories around Foshan discard millions of tonnes of tile offcuts and kiln waste every year. Could that material — ground, sorted, and re-bound — produce something worth putting in a room?
The answer turned out to be yes, but not immediately. The interesting part wasn’t the idea; it was the conversion. Getting the aggregate gradation right so the surface read as intentional rather than accidental. Getting the binder strong enough to survive shipping but open enough to take a polish. Getting the colour stable batch to batch when the source aggregate varies. None of that came off a spec sheet. It came from working the process with the workshop until it was repeatable — and then repeatable again, with a different batch of waste.
The material, in section
The three levers
Aggregate. Particle size, colour distribution, density and material mix (ceramic, glass, stone, shell) all shift the surface character. Coarse aggregate reads bold and graphic from a distance; fine aggregate reads as texture up close. Mixing sources — ceramic with glass, for instance — adds reflectivity the way solid stone never does. The aggregate is also where the recycled story lives or dies: a controlled waste stream gives you a repeatable palette, an uncontrolled one gives you surprises.
Binder. Cement and resin produce meaningfully different materials, not just different looks. Cement reads industrial and matte, takes a honed finish well, and is more UV-stable; it’s also heavier and more brittle in thin sections. Resin can be made nearly transparent — letting the aggregate dominate — and tolerates thinner sections, but scratches more visibly and can yellow outdoors. Cost and lead time also differ: resin systems usually cure faster but cost more per kilo.
Finish. A fine grind, a coarse grind and a bush-hammered face are three different materials to the eye and the hand. Sealing choice (penetrating sealer vs. surface coating) changes both the colour depth and the maintenance story. Decide the finish with the binder, not after it.
What it’s good for
Flat and gently curved surfaces: table tops, panels, lighting bases, objects. Compressive strength is high; tensile is not — design for compression. Colour range is wide and the speckle pattern has a depth that solid stone and synthetic surfaces don’t. Good for pieces where weight reads as intention.
Where it fights you
Sharp undercuts and thin sections are difficult — the material needs mass and draft to demould cleanly. Batch-to-batch variation is real: the aggregate source matters, and a different run of ceramic waste will shift the palette. Plan for that variation, or plan to control your input material. Weight is a constraint for larger pieces and for anything that ships internationally: a 600 × 600 × 20 mm cement-bound top is roughly 17 kg before packaging.
Where it can be pushed
Aggregate composition is the most under-used variable — most specifications stop at “terrazzo, white base”. Oversized aggregate (40 mm+) turns the material graphic. Layered pours can put different mixes on different faces of the same piece. Resin binders pushed toward transparency produce something closer to cast composite than stone. Pigmenting the binder and the aggregate separately gives two-axis colour control.
Cost drivers
- Mould complexity — flat moulds are cheap; curves, returns and integral details are not
- Finishing labour — grinding and polishing time scales with surface area and finish grade
- Weight — freight cost is a function of the design, decide it at the drawing stage
- Batch setup — small runs carry the mould and mix-development cost on few units
- Sealing system — penetrating sealers are cheap; high-build coatings add a process step
What to put on the drawing
- Aggregate spec: material mix, size range, density of distribution — with a physical sample reference
- Binder: cement or resin, pigmented or natural
- Finish grade: grind grit, honed or polished, sealed with what
- Minimum section thickness and draft on any vertical face
- Acceptable batch variation — define it with two boundary samples, not adjectives
- Edge profile: a 2–3 mm chamfer survives shipping; a sharp arris doesn’t