Two Surveys, Not One

A bespoke kitchen is cut to a room that does not yet exist, which is why we measure once to draw it and again before anything goes on the saw.

Every kitchen we make is measured at least twice, and the two surveys are not the same job done twice over. They happen months apart, they answer different questions, and only one of them is a set of dimensions we will cut timber to. A maker who takes a single survey is either building a kitchen loose enough to absorb the error, or gambling that the room will not move. Rooms move.

The room you measure is not the room you fit into

At the point a client first asks us to look at a space, the room is usually one of two things: still in use, with the old kitchen in it, or a shell with the ceilings out and the floor down to joists. Neither is the room the kitchen goes into.

Between that visit and installation the space will typically gain two coats of plaster or a skim on board, a floor build-up that may include insulation, screed and underfloor heating, a tiled or stone finish, and quite often a structural alteration that was still being priced when we first walked in. Each of those takes something off the opening. Plaster alone can lose you a good deal of width across a run between two walls, and a screed and stone floor will take real height out of a ceiling that was already tight for a tall run.

None of this is a problem if you know it is coming. It is a serious problem if you have already cut a 4.2 metre bank of in-frame cabinets to the shell dimension.

The first survey: measuring in order to draw

The first survey is a design survey. We are there to find out what the room will let us do.

We record overall dimensions, but the useful work is everything else: where the services actually run, what is structural and what is not, ceiling heights at several points rather than one, window reveal depths and cill heights, door swings, the direction and depth of the joists if we are hanging anything heavy or dropping a waste through a floor, and the route an extractor can realistically take to an outside wall. On older houses we are also looking at what the building is doing — whether a wall is bowed, whether a floor falls away towards an outside wall, whether a chimney breast is genuinely square to anything.

Everything drawn from this survey is provisional, and we say so on the drawing. Its real output is not the layout; it is the setting-out information the builder needs. Waste centres, hot and cold positions, socket and switch heights, appliance supplies, the extractor duct, floor boxes to an island, and the finished floor level agreed in writing. That drawing goes to the builder or the project manager, and first fix follows it. If it does not, the second survey is where we find out.

The second survey: measuring in order to make

The second survey happens when the room is plastered, the services are first-fixed and capped, the windows are in, any steel is in and cased, and the finished floor level is either laid or confirmed as a build-up we can rely on. This is the survey the workshop works to. Nothing is cut before it.

The method matters more than the tape. We set a datum — a single laser line struck at a metre above finished floor level and marked on every wall in the room — and every vertical dimension in the kitchen is taken from that line, never from the floor. Floors are not flat and are rarely level; the datum is. Worktop heights, tall unit heights, cornice lines and window cill relationships all hang off it, which is why a run of cabinets set out this way reads level across the room even where the floor falls 12mm from one end to the other.

Then we measure the room’s errors, deliberately, and write them on the survey sheet. Diagonals across the plan to establish how far out of square the room is. Plumb on each wall, taken top and bottom. A straightedge or string line along each wall to find the bow and where it sits. Ceiling height at every metre along a tall run. Reveal depths at head, middle and cill. In a new-build you might find a stud wall bowed 5mm and a ceiling falling 15mm over six metres. In a Suffolk farmhouse we have measured walls out of plumb by 30mm over a single cabinet height, and that is not unusual.

Why in-frame punishes a bad survey

A face-frame kitchen presents a continuous line of timber across the whole run, and a continuous line is a straightedge held up against the building. A 2mm variation in a gap around a door is visible to anyone standing in the room, whether or not they could tell you what they were looking at. On the handle-less OAK range, where the shadow gap is the only detail, a deviation of the same size changes the gap by a proportion you can see from the doorway [CHECK — confirm our standard shadow gap dimension].

So the carcasses have to go in plumb, level and square to each other regardless of what the walls are doing, and the wall’s error has to be taken up somewhere else. We design that somewhere in: scribe allowance on every end panel, plinth and cornice return, and on badly out-of-plumb walls a loose scribe piece that is cut on site so the cabinet itself never has to twist. Our standard allowance is generous enough to swallow the error and small enough not to read as a filler [CHECK — confirm the scribe allowance we specify as standard].

You cannot design a scribe you have not measured. That is the whole argument for the second survey in one line.

What it costs to skip it

A single survey does not usually produce a disaster. It produces a bad week.

The tall run needs packing 20mm at one end, so the cornice no longer runs parallel to the ceiling. The island is 15mm narrower than drawn because the plaster came on thicker than anyone allowed, so the stone template is wrong and the fabricator returns. An integrated appliance aperture is tight, the door catches the adjacent frame, and someone eases a hand-painted edge with a block plane on site. The fit takes ten days instead of five, and the finishing that should have happened under lights in the workshop happens on a dust sheet in someone’s house.

The economics are simple. A survey visit is a day of somebody’s time. A re-made in-frame door is timber, machining, a full set of joints, four coats of paint and a week in the finishing room, and it holds up the whole installation while it dries. We would rather drive to Gloucestershire twice.

In truth, there are three

The second survey is the one timber is cut to, but it is not quite the last time the room is measured. Stone is templated off the cabinets after they are fitted, not off our drawings, because a worktop that follows a wall’s real line is the only worktop that will sit tight to it. Some fabricators will template from a CAD file. We do not let them.

The rod pinned above the bench is drawn from the second survey — full size, in pencil, the mortices and tenons and the door gaps at their true dimensions. It is only as honest as the dimensions that made it. Measuring twice is not carefulness for its own sake; it is what makes the rod worth drawing.