Deciding Whether a Cavity Wall Should Be Filled

What the Borescope Is Being Pointed At

The gap is not spare space. In traditional cavity construction the two leaves are deliberately discontinuous so that water crossing the outer leaf — and it does cross, in every driving-rain event, on every elevation — arrives at a void it cannot bridge, runs down the cavity face, and leaves at the base. The ties are the exception the design tolerates: a small number of stiff metal links that carry wind load across the gap and are shaped, or should be shaped, to shed water before it reaches the inner leaf. Everything a survey looks for is a way that discontinuity has already been broken.

That is also the distinction the sales conversation usually skips. A wall built from the late 1970s onward with a partial-fill board in place, a clear residual cavity, and trays over every opening was designed around insulation. A 1930s or 1950s wall with a clear gap, no trays, and mortar squeezed off the back of the outer leaf onto every tie was not. Both are cavity walls. Only one of them was drawn with a fill in mind, and the survey's job is to work out which is standing in front of it.

So the borescope goes in more than once and on more than one elevation, and someone writes down what it shows. Holes go into a perpend rather than a brick face, they go in high and low, and at least one goes into the elevation that takes the weather. What comes back matters more than the fact that the drill went in.

1. Open a hole in a perpend on the exposed elevation first, high on the wall, where wind-driven rain and mortar droppings both concentrate.
2. Measure the residual clear width between the leaves — not the nominal width the era of the house suggests, and not the width at one hole taken as the width everywhere.
3. Look down the cavity for mortar extrusions off the back of the outer leaf, droppings resting on ties, and debris heaped at the base against the damp-proof course.
4. Find a tie, and look at it: type, condition, whether it slopes toward the outer leaf as it should, and whether it is carrying a bridge of hardened mortar.
5. Check whether the cavity is already filled, or partly filled, or holds a partial-fill board with a residual gap that a second product would now be blown into.
6. Record whether the inner leaf face is wet, stained, or dark, and photograph every hole with the depth gauge in shot before it is pointed up.

Whether the Outer Leaf Can Be Trusted to Stay on the Outside

Rain load is not the same on two houses in the same street, let alone two counties. BS 8104, Code of practice for assessing exposure of walls to wind-driven rain, is the document that turns local wind and rainfall into an exposure category for a given wall, and BS EN ISO 15927-3 covers the calculation of a driving rain index for a vertical surface from hourly climate data. This matters more than any other single input, because the exposure category is what most Agrément certificates use to limit where their system may be used: the same bead, in the same cavity width, is acceptable on a sheltered mid-terrace and refused on an exposed gable with no roof overhang. A surveyor who has not established which category the elevation sits in has not started the assessment.

Then the leaf itself. Cracked or eroded perpends, spalled and frost-damaged faces, a hard cement render crazed over soft brick, brickwork that has been painted with something impermeable, a chimney breast with open flaunching, a single-skin outrigger returning into the cavity wall — every one of these lets more water into the gap than the gap was designed to shed, and none of them is repaired by putting insulation behind it. Fill is not a waterproofing measure and no certificate claims it is. The order of work runs the other way: make the leaf good, let it dry through a season, then reassess.

There is also a consequence of filling that is easy to leave out of the sales conversation, and it works against the outer leaf. Once the cavity is insulated, the outer leaf loses most of the heat that used to leak out through it. It therefore runs colder and stays wet for longer after each event, which raises the number of freeze-thaw cycles it experiences while saturated. On a leaf built from bricks with a freeze-thaw designation appropriate to a sheltered wall — the durability designations sit in BS EN 771-1, Specification for masonry units — Clay masonry units — that change of regime is exactly how a wall that had been stable for decades starts shedding faces after a fill.

Absorption is where the leaf's character shows up as a number. The initial rate of absorption test in ASTM C67 and its European counterpart in BS EN 772-11 measure how fast a dry unit pulls water in through the bed face in the first minute. Be honest about what that tells you: IRA is a bricklaying number, used to decide whether a unit needs docking before it is laid so it does not rob the mortar of its water. It is not a rain-resistance rating and no one should present it as one. It is useful here for a narrower reason — a very thirsty leaf on an exposed elevation, laid dry into a strong mortar decades ago, is a leaf whose bed joints were probably poorly bonded from day one, and open bed joints are how bulk water reaches ties.

The Ties Are the Thing You Never See Again

Wall ties carry every unit of wind load across the cavity, and a fill buries them for the remaining life of the building. Any question about their condition has to be settled before the hose goes in, because afterwards the only way to inspect one is to take bricks out. In pre-1980s housing the tie is usually a galvanised steel strip or a twisted wire, and the galvanising has a finite life in a damp cavity, especially where black ash mortar was used. Corroding steel expands as it goes, and expanding steel in a bed joint jacks the outer leaf upward.

That produces the diagnostic every surveyor should know before they ever look at insulation: horizontal cracking in the bed joints at a regular vertical interval, matching the tie rows rather than any pattern of settlement, along with joints that look thicker than the ones above and below, a bulged panel, and a lifted gable or verge. Where those appear, the wall is a tie replacement job, and it is a structural one. Nothing about filling the cavity is the next step. Fill placed against corroding ties simply guarantees that the eventual repair is done blind, through a cavity full of wet material.

What the ties are supposed to be is written down. BS EN 845-1, Specification for ancillary components for masonry — wall ties, tension straps, hangers and brackets, covers the component; PD 6697, Recommendations for the design of masonry structures to BS EN 1996-1-1 and BS EN 1996-2, is where UK practice for their spacing and density lives, and it is the document a remedial tie scheme will be designed against. Do not take the count from the era of the house: on any wall built before the current recommendations, a metal detector survey and a few opened joints tell you what is actually there, which in a surprising number of houses is fewer ties than anybody assumed and none at all around the openings.

The Base of the Cavity Is a Drain, and the Fill Lands on It

Water in a cavity has to leave somewhere, and it leaves at the bottom: down the inner face of the outer leaf, onto a damp-proof course or a cavity tray, out through weep holes in the course above. That drainage path is the reason an unfilled cavity tolerates a leaky outer leaf for years without anything showing indoors. It is also the first thing a survey should look at, because if the path is already compromised, the fill converts an intermittent nuisance into a permanent one — bulk water that used to run down a clear face and out now has a wet material to soak into and hold against the inner leaf.

Two failures dominate. The first is debris: mortar droppings heaped on the damp-proof course until they form a ramp from the outer leaf to the inner one, and snots squeezed off the back of the outer leaf that hang on ties like shelves. The second is missing trays. Cavity trays over openings, at abutments, above air bricks and wherever the cavity is interrupted were not standard in older construction, and where they are absent the head of every window is an unprotected bridge that the fill will now be pressed against. Neither failure is visible from outside, and neither is fixed by the person holding the hose.

Weeps are the checkable half of it. Count what the elevation ought to have above each tray and at the base course, then count what it actually has, and count again after any fill: an open perpend that has been rendered over, painted shut, or blown full of bead is a drainage point that no longer exists. Guidance on spacing differs by jurisdiction, so use the one that governs the wall — NHBC Standards for external masonry walls in UK housing, or the adopted building code elsewhere — and treat any run with no weeps at all as a finding that goes in writing.

Everything That Still Has to Breathe on Monday Morning

A surprising amount of an ordinary house depends on the cavity staying open in specific places. Sleeved vents cross it to serve the void under a suspended timber floor; air bricks feed appliance ventilation; balanced flue terminals and extract ducts pass through it; a cavity barrier closes it at a party wall or at the eaves. Every one of these is a place the fill has to be kept out of, locally and deliberately, and keeping it out is not the same as stopping the drill pattern short — beads flow, and a broken or corroded sleeve fills from the inside. Blocking an underfloor void is how a fill turns into a joist-end problem three years later, and blocking combustion air is a great deal more serious than that.

So the survey marks them up on an elevation drawing, the installer works to that drawing, and someone checks the sleeves and terminals afterwards from both ends. Where the wall carries a cavity barrier, the fill also has to respect the compartmentation the barrier exists for, which is a question for whoever is responsible for fire separation in that building rather than for the insulation crew.

What the Fill Is Worth Once the Rest of the Wall Is Counted

The fill does not add its resistance to a wall that had none. It replaces a layer that was already doing a little: the tabulated resistance of an unventilated air cavity in BS EN ISO 6946 is modest, of the order of a fifth of a square-metre-kelvin per watt for a still gap with ordinary surfaces, and a cavity that is ventilated to outside is worth even less. The honest sum is therefore the whole assembly built up layer by layer, twice — as it stands, and as it would stand filled — using design values from BS EN ISO 10456, Building materials and products — Hygrothermal properties — Tabulated design values, rather than a headline figure from a leaflet.

Two things routinely get left out of that sum and both flatter the fill. The first is the rest of the wall: the two masonry leaves, the plaster, the render, and the internal and external surface films together carry a real share of the resistance, so the proportional improvement is always smaller than the ratio of the insulation's own R-value to zero. The second is moisture. Tabulated design values assume a material at a design moisture content; a fill that gets wet in service, or a leaf that is saturated for weeks at a time, does not perform at its dry value, which is another reason the exposure question comes before the thermal one.

Watch the units when you compare anything to anything. A metric resistance in square-metre-kelvin per watt and an imperial R-value in hour-square-foot-Fahrenheit per BTU differ by a factor of about 5.678, so an unconverted metric figure dropped into an imperial calculation reads as a wall five times worse than it is, and an unconverted imperial figure reads as a miracle. Get the whole build-up into one system of units before adding anything.

Reading the Number Printed on the Quote

Most quotes state a before-and-after U-value in watts per square metre kelvin. Ask where the pair came from. In scheme work they are usually the standard assumptions a reduced-data energy assessment applies to a wall of that age and description, not a measurement of this wall with its actual leaf thicknesses, its actual clear width and its actual condition. The target the work is being judged against, meanwhile, sits in the energy conservation guidance for the jurisdiction — Approved Document L in England and Wales, Section 6 of the Technical Handbooks in Scotland, Technical Booklet F in Northern Ireland — and is expressed the same way.

Converting that number into the resistance the product is labelled with is the only way to compare the quote with the material. The reciprocal is exact and the conversion is arithmetic; the trap is the unit. For the same wall, the metric transmittance in watts per square metre kelvin is about 5.678 times its imperial counterpart, so a metric U-value has to be divided by 5.678 before it goes anywhere near an imperial calculation. Skip that division and the reciprocal comes back about five times too small: a wall quoted at 0.30 watts per square metre kelvin is R-19 in imperial terms and reads as R-3.3 if the unit is never converted. That error runs against the wall rather than for it, which is exactly why it survives unchallenged — a figure that makes your own house look worse is not one anybody rushes to query, and it has sold a great deal of insulation that was never needed.

The Cold Lines That Survive the Fill

A filled cavity does not warm the whole inside face evenly, and the pattern of what stays cold is the most common complaint after an otherwise sound job. Window and door reveals, where the two leaves close together and the cavity narrows to nothing; lintels, particularly steel ones spanning both leaves; joist ends and the floor line; the junction with a party wall; and the ties themselves, which are point bridges rather than areas. When mould appears in a rectangle around a window that never had it before, the fill did not cause moisture — it lowered the surface temperature everywhere except the bridges, which raised the relative humidity at the coldest lines and let growth start there first.

Sizing that effect properly is a two-tool problem, and the tools are not interchangeable. Linear and point bridges — the reveal, the lintel, the tie — are handled as psi and chi values calculated to BS EN ISO 10211, Thermal bridges in building construction — Heat flows and surface temperatures — Detailed calculations. Whether a given junction will actually grow mould is a surface temperature question, answered against a temperature factor using the method in BS EN ISO 13788, with BRE Information Paper IP 1/06 the usual reference for the criterion applied to dwellings.

What an area-weighted parallel-path calculation gives you is the other half: a strip of the wall that stays uninsulated across a known fraction of its area. Use it where the geometry really is a fraction of area, and a large enough fraction to be worth the arithmetic — a band at the floor line, a section of cavity the fill never reached, an elevation so full of openings that the reveals add up to a tenth of it — and use the standards above where the bridge is a line or a point. Reaching for the area method on a wall tie produces a number, and the number means nothing.