When a sealed joint leaks, the sealant gets blamed. Occasionally that’s fair. Far more often the product performed exactly to specification and something about the joint, its geometry, its preparation, or what was underneath made failure inevitable from the day it was installed.
The useful thing about sealant failure is that it’s legible. A failed joint tells you what went wrong, if you know how to read it.
Reading a Failed Joint
Two failure modes, distinguishable at a glance.
Adhesive failure the sealant separates cleanly from one face, leaving that surface bare and the bead intact. The bond never developed properly. Causes: contaminated or dusty substrate, moisture on the surface at application, no primer where the substrate needed one, or a cure chemistry incompatible with the material.
Cohesive failure the sealant tears down its own middle, with material still bonded to both faces. The bond was fine; the sealant was asked to stretch further than it could. Causes: joint too narrow for the movement, three-sided adhesion, or insufficient movement capability for the application.
Which one you’re looking at determines whether the fix is better preparation or a different joint design. Repairing a cohesive failure with more of the same sealant in the same geometry guarantees a repeat.
The Five Things That Actually Cause Failures
- Three-sided adhesion. The most common and least understood. A sealant bonded to both sides and the back of a joint cannot stretch; the movement is concentrated in a thin section instead of distributed across the bead. It tears.
Backer rod exists solely to prevent this. It’s a closed-cell foam rope pushed into the joint before sealing, giving the bead a defined depth and a surface it won’t bond to. It costs almost nothing and it’s skipped constantly. Sizing matters the rod should be roughly 25% wider than the joint so it compresses and stays put. Rod is commonly stocked in 6, 8, 10, 13, 15, and 20 mm diameters, covering most joint widths on a building.
- Wrong joint geometry. For a movement joint, the working rule is a width-to-depth ratio of roughly 2:1, with a practical minimum joint width around 6 mm. A bead as deep as it is wide is far stiffer than intended and fatigues quickly.
The related error is designing joint width from what looks tidy rather than from calculated movement. If a joint will move ±5 mm and the sealant is rated ±25%, that joint needs to be at least 20 mm wide. Building it at 10 mm because it looked better means asking for ±50% performance the product doesn’t have.
- Contaminated substrate. Dust, form release agent, curing compound, old sealant residue, oil, and surface moisture all prevent adhesion. Concrete is the worst offender because release agents and curing compounds are invisible once dry and are specifically designed to stop things bonding.
Proper preparation is mechanical cleaning back to sound material, solvent wiping with the correct solvent, and letting it flash off completely. Where an old joint is being redone, all previous sealant has to come out; new silicone does not bond reliably to cured old silicone.
- No primer where one was needed. Some substrate and sealant combinations develop adequate adhesion unprimed; many don’t, particularly porous materials and certain coated metals. The manufacturer’s data states where primer is required, and it’s the first step to disappear when the programme is tight.
- Wrong product for the substrate. Acetic-cure silicone against concrete, marble, or coated metal will fail at the bond line because the acid attacks the substrate. This is a specification error presenting as an installation failure six months later, and matching cure chemistry to substrate is the highest-value decision in the whole sequence.
Conditions on the Day
Temperature at installation sets the joint’s baseline. A joint sealed on a hot afternoon is at its narrowest the sealant then has to accommodate the full movement in extension when everything contracts overnight. Sealing at a moderate temperature centres the joint in its movement range. On a large exterior run this is worth planning around.
Surface moisture. Most silicones need a dry substrate. Some hybrid and polyurethane products are formulated to be applied through surface moisture, which is genuinely useful in Thai conditions but that’s a specified property of particular products, not a general property of sealant.
Tooling. Tooling the bead within its skin-over time presses the sealant into contact with both faces and produces the slightly concave profile that distributes stress properly. An untooled bead has poor edge wetting and a shape that concentrates stress at the corners.
Cure time before service. Silicone skins in minutes and reaches full cure in days, depending on depth, temperature, and humidity. Loading, washing, or moving a joint before full cure compromises it permanently.
Where This Actually Pays Off
The economics are lopsided. Backer rod, primer, and thirty extra minutes of substrate preparation cost very little at installation. Diagnosing a leak through a finished façade, arranging access, cutting out the failed joint, and resealing costs a large multiple of that — and on a high-rise, the access equipment alone can exceed the entire original sealant package.
The sealant is rarely the expensive part of a joint. The joint is.
Practical Takeaways
- Diagnose before repairing a clean peel means adhesion failure, a torn middle means the joint was too narrow
- Always use backer rod in a movement joint, sized about 25% wider than the gap
- Target a width-to-depth ratio around 2:1, minimum joint width about 6 mm
- Size joint width from calculated movement, not from what looks tidy
- Strip old sealant completely new silicone won’t bond to cured silicone
- Seal at moderate temperature so the joint sits mid-range in its movement cycle
Where to Start
Getting the product right is the first of the five decisions above, and a Silicone sealant (EN head term) range organised by cure type makes that decision faster First Sealant Technology’s is split between neutral-cure options for mixed and porous substrates and acetic-cure for non-porous glazing work, with an anti-fungal variant for wet areas.
For primer requirements, movement capability data, or backer rod sizing, their team can be reached at 096-818-9168, 097-959-5111, or 097-959-1222, Monday through Saturday, 08:30–17:30.
Technical specifications referenced above are drawn from the manufacturer’s published product information. Confirm substrate compatibility, primer requirements, and movement capability for your specific application directly with the supplier.
