Independent author of Built by Chemistry. Every edition sourced from EN / ANSI / BIS / ASTM standards or peer-reviewed literature. About →
Last week we opened the bucket and looked at what makes a cementitious slurry flexible. This week we do the unglamorous but decisive part: we read the rulebooks. A flexible cementitious waterproofing membrane is the same grey material whether it is sold in Madrid, Miami, Melbourne, Mumbai or Guangzhou — but the certificate on the bag answers to a different standard in each, and those standards do not measure the same thing. This is a straight, side-by-side comparison of how five regions decide whether that membrane is fit to keep water out, with no product names and no marketing — just what each standard actually requires, and where the numbers stop being comparable.
Europe — EN 14891: adhesion plus a bridged crack
EN 14891 is the European reference for liquid-applied water-impermeable products used under ceramic tiles. It sorts them into three families by chemistry — CM (cementitious), DM (polymer dispersion) and RM (reactive resin) — and sets a floor a product must clear to be sold: tensile adhesion of at least 0,5 N/mm² measured initial, after water immersion, after heat ageing, after freeze–thaw and after lime-water contact, plus watertightness under head, plus the ability to bridge a crack of at least 0,75 mm. As Edition #18 flagged, that crack-bridging test is run at 23 °C; performing it in the cold is an optional class — O1 to −5 °C, O2 to −20 °C — and a separate letter P covers chlorinated pool water. Europe, in other words, bundles waterproofness, bond and flexibility into one certificate and expresses “flexible” as a deflection in millimetres. (For façade coatings rather than under-tile membranes, Europe uses a different crack-bridging scale entirely, EN 1062-7, graded A1 to A5 by crack opening — a reminder that even within one region the yardstick changes with the application.)
The United States — ANSI A118.10 + A118.12: two products, not one
The American system makes a distinction Europe folds together. ANSI A118.10 governs the load-bearing, bonded waterproof membrane for tile: it must pass a hydrostatic head test and hold a shear bond to cement above 50 psi — about 0,34 MPa — wet and dry, and survive the Robinson floor test that loads the tiled assembly. What A118.10 does not certify is movement. That job belongs to a second standard, ANSI A118.12, for crack-isolation membranes, and it rates flexibility not as a deflection or a stretch but as the width of substrate crack the membrane can absorb without telegraphing it into the tile: a standard grade to roughly 1,6 mm, a high-performance grade to about 3,2 mm. So in the US a single European “flexible waterproof” product corresponds to two different specifications, and its flexibility is written as a crack width.
Australia and New Zealand — AS/NZS 4858: rated by how far it stretches
AS/NZS 4858 takes yet another route. It classifies wet-area membranes by a pure material property — elongation at break — into three classes: Class I at 10–59 %, Class II at 60–299 %, and Class III at 300 % and above, on top of durability tests for immersion, heat ageing, water-vapour transmission, water absorption and cyclic movement. The elegant part is that the class then drives the detailing: the less extensible the membrane, the wider the bond-breaker it needs at a joint — roughly 100 mm for Class I, 35 mm for Class II, 12 mm for Class III. Here “flexible” is a percentage, and it is deliberately tied to how the installer must treat movement.
China — GB/T 23445: a materials specification
China has a dedicated product standard for exactly this material: GB/T 23445, the polymer-modified cement waterproofing membrane, first issued in 2009 and refreshed in 2025. It reads like a materials spec rather than a system test — a Type I membrane must reach a tensile strength of at least 1,2 MPa and an elongation at break of no less than 200 %, with a higher-strength Type II for structural use. Where Europe asks “does it bridge a crack?” and Australia asks “how far does it stretch?”, China sets both a strength floor and a stretch floor and expects the product to meet them in the lab.
India — the product standard that isn’t there
India is the revealing case, because the gap is the finding. There is a code of practice, IS 13182:2020, for waterproofing and damp-proofing of wet areas — but it governs design and installation, not the membrane as a product. There is IS 2645:2003, but it specifies integral waterproofing compounds, the admixtures dosed into concrete, not a coating. What India does not have is a dedicated product standard for a flexible cementitious membrane equivalent to EN 14891 or GB/T 23445. In practice the fast-growing Indian market runs on European classes or on each manufacturer’s own datasheet — which means two “waterproof” bags on the same shelf may have been judged against completely different bars, or against none that is nationally defined.
One sourced, no-brand construction-chemistry breakdown in your inbox every Thursday. Subscribe here.
Why the numbers don’t convert
Line the five up and the real problem appears. Europe measures flexibility as a bridged-crack deflection in millimetres at a fixed temperature. The US measures it as the crack width a separate membrane can isolate, also in millimetres, but obtained a different way. Australia and China measure it as elongation at break, in per cent. India measures it, as a product, not at all. A membrane that is “Class III” in Sydney, “O1” in Frankfurt, “high-performance A118.12” in Chicago and “Type I” in Guangzhou has four certificates whose headline numbers cannot be translated into one another, because they are not the same physical quantity. A deflection is not a crack width; a crack width is not an elongation. This is why a genuinely global specifier cannot simply ask for “a flexible membrane” — the word means a different measured thing at every border, and a product optimised to pass one test is not automatically strong on another.
This matters commercially, not just semantically, because the metrics pull against one another. The chemistry from Edition #18 is the reason: push the polymer up to chase a 300 % elongation and you soften the film — often at the expense of the tensile strength or the bond that another standard rewards; tune instead for a stiff, high-adhesion product and you give up stretch. No single formulation sits at the top of all five scales at once. A membrane engineered to ace Australia’s elongation class is not automatically the one that maximises European adhesion or Chinese tensile strength, and vice versa. So when a datasheet quotes a spectacular figure on one axis, the useful question is what it traded away on the others — and, more to the point, which axis the project in front of you actually loads: a bridged shrinkage crack, a moving structural joint, standing water, or freeze–thaw.
Where the rulebooks are heading
Two shifts are already on the record. In Europe, the revised Construction Products Regulation entered into force in January 2025, and its effect on this product will be less about performance than about paperwork: alongside the CE-marked classes, membranes are heading toward a Digital Product Passport and declared environmental and carbon data, phased in from 2026. In China, GB/T 23445 was refreshed in 2025, holding a hard, testable materials bar. The direction of travel in both is the same — performance classes are becoming table stakes, and the next axis of competition is the environmental declaration behind them.
India is the open question, and here we will offer an opinion rather than a citation. As the country’s tiling and waterproofing volumes scale, the absence of a dedicated membrane product standard becomes harder to sustain: it leaves buyers unable to compare like with like and honest manufacturers unable to prove they are better. The obvious move is a BIS product standard aligned with EN 14891 — close enough to let exporters and multinationals reuse existing test data, adapted for Indian climate and site reality. We would expect that gap to close within a few years, not because a committee has announced it, but because a market this size eventually demands one language. Until it does, the safest specification in India is still an explicit European or equivalent class, written into the contract, rather than the word “waterproof” alone.
The one-line takeaway
There is no world standard for a flexible cementitious waterproofing membrane. There are five regional ones that each measure a different property and call the result “flexible,” plus a large market that has no product standard at all. If you specify, buy or sell across borders, treat the class letter as a local dialect, not a universal number — and always name the standard, the class and the test temperature, never just the promise on the bag. The trilogy closes with Part 3: “what “no standard” means on the ground” — the India market read that turns the missing rulebook into a paradox where the premium is the only defensible place to compete.
Built by Chemistry — construction chemistry, sourced and brand-free. The trilogy is now complete — read Part 3: India’s waterproofing market and the premium paradox. Subscribe here.
- Why tile grout crumbles: the joint that can’t cure
- India’s waterproofing market and the premium paradox
- Flexible cementitious waterproofing: mortar to membrane
- Why tiles fall off: inside a tile-adhesive failure
- Open time, setting time, pot life, adjustment time: four clocks the whole market runs as one — and the one that quietly pulls the tiles off the wall.
