Splashbacks and upstands in New Addington

Yes. New Addington is on the patch, and the postcodes we work here are CR0. It is worked on our south London run, along with the rest of Croydon. A skip, a delivery lorry or a run of scaffolding takes the frontage the van needed, so the street is worth a look the day before rather than on the morning. Panels come off the van in one trip where possible, since a tall piece handled repeatedly is a piece asking to be chipped. A narrow gap between parked cars decides how a panel turns. Nothing needs visiting and nothing needs collecting: the template is in your kitchen and so is the fit.

The window sits in the middle of the run over the sink, and its board and cill set the limit on how tall an upstand can be. Post-war terraces are typically blockwork or early lightweight block with a hard render coat, and the kitchen has usually been reconfigured at least once, leaving patched areas of newer plaster next to original render that sit at different depths. A hundred millimetres covers more than it looks. A run can measure dead right at both ends and still belly out in the middle, which only appears once the template plots the intermediate points. It is usually what pushes a job towards a 100mm upstand instead of a full-height panel, because an upstand rides what a panel exposes. The reveal decides where the stone dies.

Where the stairwell is narrow enough that the stone travels on edge the polished face runs within an inch of the plaster, so the walls get boarded as well as the steps. It is usually the full-height panels rather than the tops that will not make the turn, and that only shows up when they arrive. A polished face close to an edge does not survive a knock. Where parking runs down both sides the road is left as a single lane, and a van standing in it has to be somewhere it does not stop a bin lorry. Splashbacks are normally the last thing off, so a short parking window is exactly when they get rushed. These travel flat and get carried on edge, and both want room.

It is on the short list of three for this patch, with worktop replacement first. Same template, same workshop, same fitters whichever it is. A converted flat is measured off the original house, so no two units in the same building have quite the same kitchen run. Splashbacks and upstands are cut from the same slab run at the same time, which is the only way the batch actually matches. Cut with the tops while the batch is open. Kitchens go in near the end of a conversion, behind the plastering and the flooring, by which point the programme has already absorbed whatever slipped earlier. Ordered later from a different batch, the match is not reliable exactly where the pieces meet. Cut as a set, so one can replace another.

Where the sink and the hob face each other across the aisle, everything carried between them crosses the walkway, and that is the one movement the shape cannot design out. The same shape decides where splashback pieces break, and a joint in a corner reads as a detail while one in a clear run reads as a mistake. Return ends are polished, not left sawn. Material sits between £110 and £550 per square metre supply only, and templating, fabrication and fitting are quoted line by line.

Where a corner is a couple of degrees off, a mitre cut to a theoretical ninety degrees opens on the wall side and there is no closing it afterwards. A tall panel that fits at the worktop will stand off the wall at the top unless the wall is dubbed out or the panel is scribed. A bowed wall shows on a panel and hides under an upstand. Boiler position, flue and any pipework drops must be settled and the wall skimmed level before templating the splashback run. An owner paying for this directly tends to want the reasoning as much as the number, so a quote that explains why a join sits where it does is worth more than a cheaper one that does not. A 100mm upstand is the low-maintenance answer, and full-height stone is the one that has to be committed to early. What repeats across flats is recorded once and repeated.

A join cannot cross a cut-out, so a sink mid-run takes away the position you would have used. A join is glued and clamped flush on the day, so it is fixed rather than something to be undone. Slabs come in finite sizes, so a long run needs a join whatever the room looks like. A run that stops in mid-air at the open end needs a finished end rather than a cut one, and that end takes the same edge work as the whole front. Long runs almost never take a single panel, so the joint positions want agreeing before anything is templated. Panel widths answer to the slab before the wall.

The alternative stands on the surface, so the choice is losing worktop permanently or only while you are draining. Machining is permanent. If a later kitchen moves the sink, the grooves stay where they were cut. Blockwork with a hard render that has been patched over decades looks modern and behaves nothing like it. What decides between a slim upstand and full-height stone is nearly always the wall rather than the taste. Panel joints land over a unit line, never mid-span.

Doorways, the newel post and the handrail usually take the room away, rather than the stairs themselves. Splitting a run costs a join, and that is the price of getting stone into the room at all. If a piece cannot be got in at all, we say so at the template rather than on the day. A window sill or a radiator on the half-landing takes the corner out of the turn at precisely the point the piece needs that corner. Where the route is tight we break a long splashback at a natural vertical joint instead of forcing one panel through. Panel size is checked against the tightest point rather than the average one.

Ordered later from a different batch, the match is not reliable where the two pieces meet. The joint between splashback and worktop is siliconed, never grouted, so it can move a little without cracking. Stone up a wall is heavier than tile and bonded rather than hung, so the wall has to be sound. Base units from the seventies and eighties are often only 500mm deep, so a 600mm stone top overhangs at the front and the doors want checking against it. That step is exactly what an upstand covers and what a full-height panel exposes, so unless the wall is skimmed level end to end, a 100mm band is the reliable choice. The plaster behind it is the real question.