We propose unfinished concrete to roughly half the residential clients we work with. We end up building it for roughly a third of those — about one project in six. The clients who agree to it before construction tend to be nervous; the clients who live in it afterwards almost never want to take it out. This essay is a working note on the gap between those two states, and on the decisions, mostly small, that determine which side of the gap a wall ends up on.
The case is not that concrete is fashionable. The case is that concrete is, in a residential interior, very nearly the only contemporary material that gets better with time, with children, with weather coming in off a balcony door, with the scuff of a chair leg. Plaster has to be patched. Painted drywall has to be repainted. Engineered timber gets a wear-strip before the homeowner notices. Concrete, properly poured and modestly sealed, simply darkens, slowly, in the lower foot of every wall, the way the bottom of a stone column darkens in a cathedral. The patina is the room learning that you live there.
§ I.The two fears.
When a client refuses concrete, they usually refuse it for one of two reasons, both of which are misinterpretations of what the material actually is.
The first fear is cold. Concrete will be cold, the client says, and they are correct in the same trivial sense that a stone hearth is cold. The objection assumes the wall is going to behave like a refrigerator panel. In fact, a board-formed concrete wall in a heated room sits within two or three degrees of the ambient air after the first twenty-four hours. It is thermally massive, which means it is slow to change — but the steady state it settles to is not cold; it is still. The wall feels less aggressive than a sheet of cold-rolled steel, less neutral than a painted drywall surface. It feels like the room is sitting down.
The second fear is institutional. The client has, somewhere in their associative memory, a parking garage, a 1970s university lecture hall, the back wall of a dry-cleaner. This is a more legitimate fear and it has to be answered with evidence. The evidence is the board-form. A wall poured against a wall of timber boards — rough-sawn fir, ideally re-used from an earlier pour — comes out with the timber's grain pressed into its surface, half a millimetre deep, in a pattern that no photograph really conveys. It is not a parking-garage wall. It is a record of a piece of timber, in negative.
§ II.Notes on the pour.
The studio has, over the last six years, settled into a rough specification for residential board-form. We are writing this down here partly because we wish other people would, and we keep being asked.
The boards should be rough-sawn — never planed — Douglas fir or larch, ideally three to five inches wide. We prefer boards that have already been used for one previous pour, because the first pour leaves a release pattern on the timber that makes the second pour's grain considerably richer. The first pour, in our experience, is the cost of the second.
The pigment, if any, is added at the batch plant, not on site. We have used a soft warm-grey pigment — roughly the colour of a wood ash — on three projects, and a completely unpigmented mix on the rest. Pigment, in our hands, works best when you can barely tell it is there. Anything more emphatic and the wall begins to look like a colour decision rather than a structure.
The aggregate matters more than most clients will believe. A 3/8″ rounded river-gravel aggregate produces a denser, slightly creamier surface than a crushed stone of the same nominal size. We specify river gravel almost always, even though it costs about 15% more. The crushed stone reads, faintly but unmistakeably, as industrial; the river gravel reads as building.
The tie-rod pattern is the most argued-over decision in the studio. We default to a 24″ horizontal and 30″ vertical grid, with the rod holes left open and chamfered slightly with a release-cone. We do not fill them. Filling tie-rod holes is, almost always, an attempt to make concrete pretend to be a different material. The honesty of the open hole is the entire reason to choose board-form over a skim-coat plaster system. Hide it and you have lost the case.
§ III.On sealing.
Sealing is where most residential concrete walls go wrong, and the wrong direction is almost always too much rather than too little. A heavy acrylic sealer will leave the wall looking wet — permanently wet — in raking light. The client will not be able to articulate why the wall feels off, but they will feel it. The wall will look like a photograph of itself.
We use a single coat of a reactive lithium-silicate densifier, applied roughly six weeks after the pour, well after the material has finished its slow chemical conversation with the air. The densifier sits inside the surface, not on top of it. It does not change the colour. It does not change the sheen. It simply makes the concrete five to ten percent harder and considerably less dust-prone. A wall sealed this way still accepts a small spill and lets it darken and fade over a few days. It is, in the right sense, a porous wall.
The patina is the room learning that you live there. A sealed, embalmed wall cannot learn anything.
We have only ever specified a heavier sealer for one residential wall, and that was at a kitchen splashback, where the client was going to be cooking professionally and the wall was going to be hit with tomato and turmeric and red wine. For everywhere else, the densifier is enough. The wall is, after all, the building. Buildings are allowed to age.
§ IV.Where it does not belong.
I want to be honest about the rooms in which board-form concrete is, in our view, a mistake. We do not specify it for bedrooms. We do not specify it for libraries or studies. We do not specify it for any room intended for sustained quiet, because the material — even sealed, even pigmented — has an acoustic signature that is, frankly, hostile to a long evening of reading. It bounces sound. It bounces, in particular, the consonants. A conversation in a concrete-walled bedroom feels, we have found, slightly louder than the actual decibel level warrants.
We use it almost exclusively for the rooms in a house that are already loud — kitchens, hearth rooms, entry galleries, sometimes a bathroom that wants to feel more like a small bathhouse than a domestic en-suite. In all of those rooms the material does the acoustic work it should be doing: it gives back exactly enough liveliness for a room of five people having a meal to feel like a room of five people having a meal.
§ V.A coda about restraint.
The reason this material continues to interest the studio, eight years in, is that it is a kind of test of architectural restraint. Concrete will reward, almost endlessly, a small gesture: a slight chamfer at a corner, a deliberate tie-rod rhythm, a single horizontal joint where two pours meet. It will punish, more or less immediately, a large gesture: a decorative texture, a coloured pigment too saturated to be believed, a sealer that makes it shine.
We end up, on every concrete project, making fewer decisions than we would on the equivalent painted wall. The wall arrives with most of its decisions already made — by the timber of the formwork, by the river gravel of the aggregate, by the slow chemistry of the cement curing for six weeks before we touch it. Our job, by the time the formwork comes off, is mostly to leave it alone. That is a discipline worth practising. Most of what the studio is trying to learn, on most projects, is when to stop.