Waterproofing paint vs membrane, by the numbers
| Coating (paint, unreinforced) | Reinforced liquid membrane | Torch-on sheet | |
|---|---|---|---|
| Finished thickness | roughly 0.15 – 0.3 mm | roughly 1 – 1.5 mm | 3 – 4 mm per layer |
| Reinforcement | None | Polyester cloth between coats | Built-in polyester or glass carrier |
| Bridges a static hairline crack | Marginally | Yes | Yes |
| Bridges a moving joint | No | Within limits | Yes, with a detail |
| Survives standing water | No | Poorly | Yes |
| Indicative service life | 3 – 5 years | 8 – 12 years | 15 – 20 years, two-layer |
| Indicative installed cost | R95 – R180 / m² | R120 – R240 / m² | R220 – R380 / m² |
| Fails by | Chalking, peeling, splitting at cracks | Cloth showing through, edge lifting | Lap failure, blistering, UV at end of life |
Over a twenty-year horizon the cheapest option in that table is the most expensive one, because it gets bought four or five times. That is the whole argument in one line — but it only holds where a membrane is the right answer, and it is not always.
What waterproofing paint genuinely does well
It is not a bad product. It is a product with a narrow correct use, sold as though it had a wide one:
- Shedding water off a vertical surface. Walls do not hold water against a coating; gravity helps. A good exterior coating on a sound wall is doing a real job.
- Detail work. Liquid gets into upstands, around penetrations and into awkward corners where sheet has to be cut and lapped. Even on a torch-on roof, the fiddly parts are often liquid.
- Re-coating a sound existing system. Restoring the UV layer on a membrane that is otherwise intact is the cheapest maintenance in waterproofing, and it genuinely extends the life of the roof below.
- Small areas and complex shapes where a sheet system would be more joint than field.
- Heat. A white coating measurably reduces the surface temperature of a roof, which slows the ageing of whatever is underneath it.
What it cannot do, no matter how many coats
- Hold back standing water. A roof that ponds needs thickness and laps, not a film. This is the most common failure, and it is a design problem — the falls are wrong — that no coating can solve.
- Bridge a moving joint. A crack that opens and closes with temperature will split any unreinforced film within a season or two. It needs a bond-breaker detail and a reinforced system.
- Resist water under pressure. Below ground, water arrives with force behind it and pushes any coating off the wall from the back. That needs tanking or a cavity drain membrane.
- Work under tiles. Waterproofing paint under a tile bed is not a wet-area system. Showers need a bonded sheet — see shower waterproofing.
- Fix a wall that is wet from the inside. Coating the face of a wall carrying rising damp or condensation traps the moisture and lifts the coating, taking plaster with it.
- Stick to a damp or contaminated substrate. Adhesion is decided before the first coat goes on. Most coating failures are preparation failures.
The four-question test
On the roof, in about ten minutes:
- Does it pond? Wet the roof and watch. Water standing anywhere more than twenty minutes after the flow stops means the falls are wrong. A coating will not survive it — fix the falls or lay a sheet system.
- Is the existing surface sound? Press a thumbnail into it and try to lift an edge. If it is chalking, blistered or lifting, a new coat over it fails with it, and you are buying a strip.
- Is there a crack, and does it move? Mark across a crack with a pencil line in mid-winter and look again in mid-summer. If the line has broken, the crack moves and you need reinforcement across it.
- How long do you need it for? Selling in three years, and a re-coat is honest maintenance. Staying twenty and the deck is sound, and a sheet system is cheaper per year by a distance.
If the answer to the first question is yes, the rest do not matter. Ponding is the one condition that defeats every liquid system on the market, regardless of price or brand.
Reading a South African product label
Three things on the data sheet decide whether you have bought a system or a coating, and none of them are on the front of the tin:
- Does it specify a membrane cloth? If yes, the cloth is part of the product. Skipping it is not a cost saving, it is a different and much weaker system.
- Does it specify a primer? Concrete, fibre cement and previously coated surfaces usually need one, and adhesion failure is the most common way these systems fail early.
- What is the wet film thickness per coat, and how many coats? Multiply it out. If the specified system lands under a millimetre finished, you are buying a coating whatever the tin calls itself.
Which product sits in which family, including the torch-on, cementitious and repellent ranges, is set out in waterproofing products in South Africa.
The Cape Town complication
Two local conditions push this decision further toward membrane than it would be elsewhere. The first is the rainfall pattern: our rain comes in concentrated winter spells, arrives close to horizontal, and lands on roofs that have just spent a summer being baked. Thermal cycling opens cracks in summer that driven rain then finds in winter.
The second is the weather window. Liquid systems need a dry substrate and a dry cure — typically several consecutive dry days. Between May and August Cape Town rarely offers one, so coatings get applied to substrates that are not properly dry and then fail for reasons that have nothing to do with the product. If the work has to happen in winter, that is an argument for a sheet system, which does not care nearly as much.
Where a roof is sound and you are re-coating, or where the shape makes sheet impractical, a properly reinforced liquid system is a sensible, defensible choice. Where the roof ponds, moves, or is already at end of life, it is not. The systems and what each costs installed are on waterproofing systems; roof-specific advice is on roof waterproofing in Cape Town.
The edge of a flat roof is where paint-on systems fail first, because the upstand moves differently from the deck; the detail is on parapet wall leaks.
Related reading
Start with Waterproofing Products in South Africa, Honestly Compared for the overview.
Common questions
Is waterproofing paint as good as a membrane?
Not as a primary roof system. An unreinforced coating is roughly a tenth to a twentieth of the thickness of a torch-on sheet and has no reinforcement, so it lasts three to five years rather than fifteen to twenty. Applied with its membrane cloth as specified it becomes a reinforced liquid membrane, which is a genuine system at around eight to twelve years.
Does waterproofing paint work on a flat roof?
On a sound flat roof with correct falls and no ponding, a reinforced liquid system works and is a reasonable choice. On a roof that holds water anywhere, no liquid system survives — that is a falls problem and needs screeding or a sheet membrane.
Can I paint waterproofing over a crack?
Only if the crack is static and hairline. A crack that moves with temperature will split any unreinforced film. It needs routing, a flexible sealant and reinforcement taken across it, or the coating tears in the same place within a season.
How many coats of waterproofing paint do I need?
Whatever the data sheet specifies, applied at the stated wet film thickness, with the membrane cloth if one is specified. Two thin coats without the cloth is the commonest shortcut and the commonest cause of early failure.
Which is cheaper over ten years, paint or membrane?
Indicatively, a reinforced liquid system at R120–R240 per m² lasting 8–12 years and a two-layer torch-on at R220–R380 lasting 15–20 land in a similar place per year. An unreinforced coating at R95–R180 every three to five years is the most expensive of the three.
Can waterproofing paint stop rising damp?
No. Rising damp arrives from inside the wall, and sealing the face traps it — the coating lifts and usually takes plaster with it. Rising damp needs a chemical damp-proof course and salt-retardant replastering.