Every waterproofing system on a building has one place where it is guaranteed to be tested to destruction, and it is almost never the open field.
It is the joint.
A membrane over a slab is asked to resist water. A membrane over a joint is asked to resist water while being pulled apart twice a day, every day, for as long as the building stands. Those are different jobs, and only one of them can be solved by choosing a better coating.
This is the rule the rest of the article exists to justify: you cannot coat across a joint that moves. Not with a superior elastomeric, not with three coats instead of two, not with a reinforcing mesh. If the gap moves and your system is simply painted over it, the only open question is which monsoon it fails in.
A crack and a joint are not the same thing
They get treated as the same item on site constantly, and the consequence is expensive.
Where the building moved without permission
Damage. Shrinkage, settlement, thermal movement, overload, corrosion of reinforcement. Nobody designed it. It may be dormant — the movement is finished — or live, still opening and closing with temperature and load.
Where you gave the building permission to move
Designed. An expansion joint, a construction joint, an isolation joint between blocks, a movement joint in a podium. Put there deliberately, and it will move for the entire life of the structure. That is its job.
| Crack | Joint | |
|---|---|---|
| Origin | Unplanned movement | Designed movement |
| Will it move again? | Maybe — depends whether it is live or dormant | Yes. Always. Forever |
| Correct response | Repair: fill, bond, restore continuity | Detail: bridge with something that can move |
| Can you coat over it? | A dormant crack, yes, after treatment | Never |
And the rule that connects them: a live crack must be treated as a joint. It is not a repair problem any more. The building has told you where it intends to keep moving, and the only sensible response is to formalise that.
Telling a live crack from a dormant one
You do not need instrumentation. You need patience and a date.
- Mark and date it. Draw a line across the crack in two places, mark the ends, write the date. Check in a month, and again across a season change.
- Watch the seasons. Thermal cracks are widest on a cold January morning and narrowest on a hot May afternoon. If the width changes with the weather, it is live.
- Use a tell-tale. A rigid patch of plaster or a glass strip bridged across the crack: if it fractures, the crack is moving.
- Read the pattern. Cracks following a structural line — over a beam, at a column junction, at a change of material — are usually movement. Fine map-pattern cracking in plaster is usually shrinkage of the finish.
Treating a live crack as a dormant one is the most common way a competent repair fails within a year.
Why a coating tears at a joint, even a very elastic one
This is the part that surprises people, because the tin says 300% elongation and the joint only opens two millimetres.
A coating applied over a joint is bonded to the substrate on both sides. When the joint opens, all of that movement has to be absorbed by the tiny width of film spanning the gap — a strip perhaps a millimetre or two wide. The film is not stretching over its whole area. It is stretching over almost nothing.
Two millimetres of movement across two millimetres of unbonded film is 100% strain, concentrated at the exact line where the film is thinnest and least supported. Do that daily, add UV and heat, and the film tears along the joint like a perforation — then holds water in the gap rather than out of it.
The answer is never a better coating. It is a debonded span — something that bridges the gap while being deliberately not stuck across the moving zone.
The backing rule
Joint tapes are usually discussed by their membrane. They should be specified by their carrier — the backing — because the carrier decides how the tape is installed, whether it can be covered, and what it can be bonded with.
This is the rule worth writing on a drawing.
| Carrier | What it means on site | Where it belongs |
|---|---|---|
| AAluminium | Stays exposed. Never coated over, never buried under a finish | Exposed expansion joints, visible movement joints, anything that must stay inspectable |
| BPolypropylene fleece | Gets coated over. The fleece is embedded into the liquid system so membrane and tape become one | Joints inside a coated waterproofing system — terraces, wet areas, tanks |
| CThin felt | Flexible enough to form an L-shaped corner detail | Wall-to-floor upstands, corners under tiles, sunken slab perimeters |
| DHole-punched TPE, thick | For real structural movement. Epoxy-bonded rather than embedded in a membrane | Structural expansion joints, podium-to-building separations, high movement and pressure |
Two consequences fall straight out of that table, and both are violated on Indian sites weekly.
An aluminium-carrier tape that has been coated over has been installed wrong. The carrier exists precisely so the tape stays a separate, exposed, replaceable element. Burying it defeats the design.
A fleece tape left exposed has also been installed wrong. The fleece is not the waterproofing — it is the anchor for the membrane that goes over it.
The tape is not a product decision. It is a detail decision, and the carrier is the part of the specification that tells the applicator which detail he is building.
Two classes of movement
Once the carrier is settled, the second question is how much movement and how much water pressure the joint has to take. The range separates into two classes.
Embedded into the liquid membrane. Joints and details inside a coated system.
Epoxy-bonded. Structural movement joints, submerged and high-pressure locations.
A joint in a terrace parapet upstand and a structural separation between two blocks are not the same specification — and the gap between 139% and 580% is exactly why.