Sealant, collar, wrap, mortar, batt or pillow

Six quotes come back against the same BOQ line. The line reads “supply and install fire stopping to service openings, 120 minutes”. One vendor has priced intumescent sealant by the litre. One has priced coated batt with pipe wraps and sealant. One has priced pillows. One has priced firestop compound. Two have not said what they are pricing at all. Every submittal carries a classification report, and every report says EI 120.

The rating will not help you compare them, because the rating is the same on all six. EI 120 records what one tested assembly did in a furnace: E for integrity, meaning the seal did not let flame or hot gas through, and I for insulation, meaning the unexposed face stayed below a set temperature rise. Six different assemblies can hit both numbers while doing completely different physical jobs.

What decides the product family is mechanical, not commercial. What the service does when it gets hot, how big the opening is, whether the seal has to carry load, and whether anyone will open it again. Get those four wrong and you are holding a valid test report for a building that is not yours.

Start with what the service does when it gets hot

The first question is not opening size. It is whether the service melts, conducts heat, or does both.

A thermoplastic pipe softens and disappears. uPVC, PP, HDPE and PE all leave a void the size of the pipe bore, and something has to close that void faster than the pipe vacates it. Only an intumescent device does that. A mineral seal packed around a plastic pipe, with no intumescent component anywhere in it, is a hole waiting for a fire.

A steel or copper pipe does the opposite. It stays where it is and conducts heat straight through the wall, so the void never forms and integrity is comparatively easy. Insulation is the hard part. The pipe carries temperature to the unexposed face on its own, and no amount of sealant around the annulus changes that. I performance on a metallic service is governed by the pipe and its lagging. This is why a blanket requirement for EI 120 on every penetration produces submittals that cannot exist.

Cables do both. The sheathing melts, the conductor conducts, and the bundle is a different object on every floor. Sort the openings by that behaviour before anyone prices them.

What separates a collar from a wrap

Collars and wraps are usually the same chemistry: graphite-based intumescents that expand under heat to close the bore of a melting pipe. They are priced differently and sold as alternatives to each other, which is where the trouble starts.

The difference is mechanical. A collar carries a steel shell and is anchored to the face of the wall or the soffit of the slab. When the intumescent expands, the shell contains it and drives the expansion inward, and the fixings hold the device against the element. A wrap has no shell. It expands into whatever surrounds it, so the aperture, the mortar or the batt seal hosting it has to provide that restraint. Wraps are therefore tested inside specific host seals, and collars are tested surface-mounted with a stated fixing pattern.

Substitute one for the other to save money and you have changed the load path, with no test report that describes the result.

The tested scope is also narrower than most people assume. One collar test series to EN 1366-3 ran every specimen with a 10 mm annular gap, sealed to a 10 mm depth, with the pipe uncapped on one side and capped on the other. That is what the report covers. A 25 mm gap, an open-ended drainage stack, a collar fixed with the wrong anchor: all outside it.

Sealant is a component of most of the other systems

Intumescent and ablative sealants are qualified over a stated gap width and depth, against a defined backing, in a defined substrate. That is a narrow envelope, and it is meant to be. The product’s job is to close an annulus, a linear gap, or the perimeter of something else.

The point most often missed is that sealant is part of nearly every other system on this list rather than a system on its own. A coated batt seal is a batt plus sealant: sealant over the full thickness of every cut edge, the batt inserted, then a perimeter bead pointed on both faces and worked between individual cables. Leave the sealant out of that build-up and the test report stops applying to it. Same for the annulus around a collared pipe.

The expensive mistake is the large opening. A 300 mm by 400 mm shaft gunned full of mastic because the BOQ said “fire resistant sealant” is an untested seal that burned through cartridges nobody budgeted for, and it comes out again at snagging.

Batt or compound: can the seal be walked on?

Both close large openings, both handle mixed services, and both reach long durations. What separates them is whether the seal has to carry anything.

Coated batt is not a load-bearing seal, and reputable manufacturers say so on the datasheet. It carries its own weight across a moderate span and nothing else. Firestop compound and mortar are the answer where load exists: a shaft opening in a slab that gets walked on for the remaining eighteen months of the programme, a seal that will pond water, a floor opening wide enough that the seal’s own weight is the problem.

Indian sites make this decision for you more often than the drawings do. Slab openings in service shafts are working platforms until handover. A batt seal specified in a floor because the wall detail was copied across is a seal that will be stood on, sagged, and quietly re-formed by someone with a bag of cement.

The second question is re-entry. Cutting a batt and re-sealing it is a short job with a cartridge gun. Breaking out compound is not. If the opening will be opened again, that difference is worth more than the material rate.

Pillows, and the compression figure nobody checks

Pillows are the right answer for one condition: an opening whose contents change. Data centre cable trays that get re-pulled every capacity cycle. Telecom risers in a let commercial floor. Construction-phase seals in openings that are not finished.

A pillow seal is not a stack of bags pushed into a hole. Installed compression is a rated parameter. Manufacturers publish a minimum figure, usually somewhere between a fifth and a third of the uncompressed volume, and some tested systems require steel retaining mesh on both faces. Under-compress and there is no pressure in the seal. Over-compress and the core deforms permanently and never expands the way it did in the test. Both failures look identical from a metre away.

Pillows are also explicitly not for locations subject to traffic or abrasion, which rules out most floor openings and a good number of plant room walls.

The rate sheet compares six scopes as if they were one

NBC 2016, published by BIS as SP 7:2016, required in Part 4 that shafts be sealed at every floor with fire stop material having the same fire resistance as the floor. It named no test standard and no product family. That edition was withdrawn on 30 April 2026 and replaced by SP 7:2026, the National Building Construction Standards, which condenses twelve parts into six and moves the language from “shall” to “should”. It is advisory until a state writes it into its byelaws, so most fire NOC practice is still running on the 2016 text. Neither edition tells a QS how to break up a firestop line item, which is why the line item does not get broken up.

So four decisions collapse into one rate, and the vendor makes all four for you, in the direction of the lowest cost. A second problem follows behind the first. When a batt from one manufacturer is pointed with a sealant from another and hosts a third company’s wrap, the assembly has no test behind it, because those three products were never tested together. ASFP guidance is blunt on this, and it is the easiest thing to spot in a submittal.

Before you accept a firestop rate

• Has the opening been sorted by service behaviour, melting or conducting, before any product was named?
• Does the quoted rate state an opening size band, or is one rate covering 40 mm and 400 mm?
• Is the seal in a floor, and if so, has anyone confirmed it will not be walked on?
• Will this opening be re-entered after handover, and does the quoted system allow it without breaking out?
• Do the batt, the sealant, the collar and the wrap come from one tested system, or has the rate been built from the cheapest of each?
• Does the test report’s annular gap, pipe end configuration and supporting construction match what is on the drawing?
• Is the sealant priced as a component of the seal, or as the whole seal?

There are only four selection inputs and all of them are physical: what the service does when it gets hot, how big the opening is, whether the seal carries load, and whether it will be opened again. Answer those in order and the product family falls out on its own, with the rating following from the tested system rather than choosing it.

Agnimatic supplies and installs across all six families. Most of the work that matters happens before anything is bought.

Agnimatic Technologies LLP, passive fire protection. Standards named for reference: EN 1366-3. Values in the diagrams are illustrative and must not be used for design. This article is general guidance, not a specification for any particular building.