Overboarding or Taking Down a Lath and Plaster Ceiling

Everything the Screw Has to Cross

This is the arithmetic that quietly kills a lot of overboarding quotes, and it belongs on the back of an envelope before anyone agrees a price. A 12.5 mm sheet over 20 mm of old plaster over 8 mm of lath is a little over 40 mm of material before the point has touched timber. On top of that sits the bite into the joist the fastener requires — ASTM C840 and the Gypsum Association's GA-216 state a minimum penetration into wood framing for single-layer work, commonly given as 5/8 in (16 mm), while the British and European systems publish their own figures for their own screws. That modest example already totals 56 mm, and a 15 mm board over 30 mm of plaster passes 70 mm — either of which is where a collated screwgun stops being able to help you.

Measure the old ceiling rather than assuming it. Take the cover off the rose and measure at the cable hole, or pull a downlight and measure at the cut edge, and do it in more than one place: hand-floated work is thicker at the edges of a room than in the middle, and thicker again over a patch.

Head profile is not a detail here. A board screw has a bugle head that sinks into the paper face and stops; a countersunk timber screw keeps going and punches through, at which point the fastener holds nothing but gypsum core. Long board screws exist for multi-layer work and are what this job wants — coarse thread, in the length the stack-up demands.

Then the layer in the middle, which is why overboarded ceilings loosen. The screw is not clamping the board to the joist; it is clamping it to a brittle crust of century-old plaster, and that crust crushes. Drive hard and you excavate a crater around the head, the board never pulls tight, and the fixing goes slack within a season. Set the depth stop light and drive to just-tight.

All of which changes how you count. On a bare joist ceiling you can separate a perimeter spacing from a field spacing, because you know where the sheet edges land. Overboarding, every screw has to find a hand-set joist through the twenty-eight-odd millimetres of opaque old ceiling in front of it, and the edges land where the joists allow, so the practical answer is a uniform grid at the closest centres the board will accept. Count it the way you are going to fix it.

Board Over a Surface You Cannot See Behind

Finding the joists is the first hour and it defeats anyone expecting a detector to do it. A magnetic finder lights up on the lath nails, and there are hundreds of those at thirty-odd millimetre intervals across the whole ceiling. Establish one joist properly — a fine pilot hole probed until it goes solid, or a patch of skim scraped back until the nail line reads — then step off from it, confirming every line rather than laying them out with a tape. These joists were set by hand and are genuinely irregular.

Snap chalk lines on those centres, then carry each mark down onto the wall a few centimetres below the ceiling line. The moment the first sheet goes up every line you drew is invisible, and the marks on the wall are what keep you hitting timber for the rest of the day.

Run the sheets across the joists and cut them so the ends land on timber. That is where the waste allowance goes: on a new build you work in whole sheets, and here you cut nearly every one to a measured length to chase joists set out by a man with a folding rule in 1897. Fifteen per cent is realistic rather than generous.

Before ordering, count what has to come down by the thickness of the board plus the skim: every downlight, the rose plate, the smoke alarm base, the loft hatch lining. Each is a hole cut in a new sheet and a return visit by whoever owns that trade.

1. Prove dead every circuit feeding a fitting in that ceiling before a screw is driven.
2. Establish one joist by probe or scrape, then confirm each of the others individually.
3. Snap the joist lines and transfer every one onto the wall below the ceiling line.
4. Set the sheets across the joists, cut ends onto timber and staggered between rows.
5. Fix from the middle of each sheet outward, and stop the gun the moment it is tight.
6. Cut back to the cornice line rather than running board over it.

Bringing It Down Without Losing the House to Dust

If the decision goes the other way, the first day is containment rather than demolition. Clear the room, sheet or lift the floor, seal the doorway with a zipped screen rather than a sheet and a nail, tape the loft hatch, and put an extractor in the window. On the top floor the void you are about to open connects to the roof space and therefore to every other ceiling on that floor.

Two methods, and they produce different waste. Working the plaster off first with a wide scraper, from a corner and along the run of the laths, leaves the laths standing to be levered off afterwards joist by joist — two separate streams, mineral arisings and clean timber, which is worth having when it comes to the container. Taking plaster and lath down together in sections is faster and leaves one mixed heap nobody will recycle. On a small room the mixed method is defensible; on a whole floor it is not.

Then the quantity, which almost everybody gets wrong in the same direction. In place the ceiling is a thin slab: sixteen square metres at a combined twenty-eight millimetres is just under half a cubic metre, which sounds like five or six barrow loads. Broken up it is nothing of the sort — fractured plaster carries voids and laths tangle rather than stack.

One honest caveat about the arithmetic. The swell factors published in the demolition estimating literature, and therefore the ones the calculator offers, are for concrete, block, brick and stone, because those are the materials the literature covers. There is no consensus swell factor for lath and plaster and this site will not invent one. Run it at the highest listed factor, treat the answer as a floor rather than a forecast, and correct it against the first container you actually fill.

1. Settle any textured coating question before the first tool touches the ceiling.
2. Empty the room, protect or lift the floor, seal the doorway with a zipped screen.
3. Prove the lighting circuit dead at the rose, not at the consumer unit — the drops are stapled to the joists.
4. Strip the plaster first, corner outward along the lath run, keeping the mineral waste separate.
5. Lever the laths off joist by joist, drawing or punching the old nails as you go.
6. Vacuum the joist tops and the wall plate before anything new is offered up.

The Container Fills by Volume and Fails by Weight

There are two ways to get the container wrong and they pull in opposite directions. Order for loose volume alone and a mixed heap of plaster arisings can reach the weight limit at half full — a container you have paid for and cannot legally have collected. Order for weight alone and the laths, which weigh almost nothing and stack almost not at all, fill the box before the tonnage is anywhere near. Check the volume and the weight limit as two separate constraints and take whichever binds first.

Access decides more than size in a terrace. No driveway means the container stands on the highway, which in the United Kingdom needs a permit from the highway authority together with lighting and markings, and in many streets means no container at all. The fallback is bulk bags handed through the house, and that changes the arithmetic completely: a bag rated for a tonne fills with plaster arisings long before it reaches one, and an overfilled bag is one the crane will refuse. Settle this before the ceiling comes down — a heap on a bedroom floor is a concentrated load on a floor designed for furniture.

The New Offcuts Are a Different Stream Entirely

Whatever route the ceiling took, boarding it produces a second waste stream that must not be tipped in with the first. Gypsum in anaerobic landfill can generate hydrogen sulphide, which is why board waste is separated from biodegradable waste — in England and Wales the regime derives from the Landfill (England and Wales) Regulations 2002, and comparable restrictions apply elsewhere. Clean offcuts are also genuinely recyclable back into new board.

Board waste is light and extremely bulky, so the conversion between volume and tonnage is what tells you whether the second container is a container at all or two bags in the front garden. Be clear about what that density describes: broken plasterboard offcuts, not the material that came off the ceiling. Old plaster bonded to timber lath is contaminated as far as any recycler is concerned, and quoting both streams as one number misjudges the disposal twice over.

The Joists You Can Finally See

The first hour after a ceiling comes down is the most informative of the job. You find the joist notched half its depth for a 1970s heating flow and return, the one cut clean through for a soil pipe and never trimmed, the split running back from a bearing, and occasionally beetle damage where joist ends sit in damp external masonry. None of it was visible an hour earlier and all of it was there.

The second thing you find is that they are not in a plane. Joists set out by hand and loaded for a century sit at different levels, and boarding straight onto them reproduces the undulation you just paid to demolish. The fix is a grid of battens run across the joists and packed to a laser line: it straightens the ceiling, gives a service void above the board, and lets you fix at whatever spacing the board wants. That grid is arithmetically identical to the joists themselves — members at a chosen spacing along the room, each spanning its width — so new joists, sistering members and levelling battens are all one count and one total length of timber.

Load deserves its own paragraph, because the ceiling is about to get heavier. Board, battens, insulation between, and anything laid over the joists in the loft all add permanent load, which is governed by BS EN 1991-1-1 in Europe and ASCE/SEI 7 in the United States. A ceiling joist of this age is often also the floor joist of the room above, sized by eye rather than by a table. If anything looks doubtful — a notch through the middle third, a visible sag, a crushed bearing — that is a structural engineer's call, not a boarding decision.

Everything else that wants doing above a ceiling wants doing now: insulation, the cable route for the lighting you actually want rather than the one you inherited, the pipework that has been weeping for a decade. In England, taking the ceiling off a cold roof space also brings the thermal element provisions of Approved Document L into view — read the current edition, because both the trigger and the target performance have moved more than once.

Board Straight Onto Timber

With the old ceiling gone the fixing is ordinary again. The screw crosses the board and a known depth of timber, so its length comes off the board thickness and the required bite rather than off a measured stack of unknowns. Sheet edges can be planned onto framing, which means perimeter and field fastening are two different spacings once more, and the honest way to count them is per sheet and then multiplied by the sheets going up.

Provide timber behind every cut end: sheets across the joists or battens, ends staggered between rows, a noggin at every square end. On a room whose walls are staying, the ceiling perimeter lands on old wall plaster that is neither flat nor square — leave the sheet a little short rather than forcing it. Keep fasteners back from the sheet edges by the distance the board standard states, and drive them to a firm dimple with the paper face intact.

Joints on a Plane That Was Never Flat

Two factory-tapered long edges meeting each other are an easy joint; that is what the recess is for. The cut ends are not, and this job produces more of them than a new build does, because sheets were cut to chase joists rather than run out to their full length. With a repaired plane behind them, the jointing here consumes more per square metre than the same area of new work.

Order against the area at the full multi-coat rate rather than pricing off a count of joints. The published coverage figure covers a whole board area finished to a complete taped system and already assumes a normal density of joints. Where a ceiling like this outruns it is in local filling — the low spot over a batten that needed one more pack, the shallow step where two sheet ends met over a twisted joist.

The junction between the new ceiling and the old wall plaster deserves a decision rather than a default. Scrim and fill it and you have a rigid line between two materials with different movement, which cracks in the first heating season about as often as not. A shadow gap or a flexible sealant line can move without telling anybody, and on a ceiling rebuilt inside walls that have stood since 1897 the flexible answer is the honest one.

Two Millimetres That Cannot Straighten Anything

There are two ways to finish a boarded ceiling and they are regional habits rather than right and wrong answers. British practice skims the whole surface in gypsum finish plaster, giving a monolithic face with no joint anywhere to photograph through the paint. North American practice tapes and coats the joints and paints the paper, which is faster and leaves joints disguised rather than eliminated. The skim route is what the application standards for gypsum veneer plaster cover — ASTM C843 for the plaster and ASTM C844 for the base it goes on — while in Europe the material and its application sit under BS EN 13279-1 and BS 8481.

A skim is two to three millimetres in two passes and it is not a levelling coat. It follows the board with great fidelity: it will not pull a bowed ceiling flat, it will not bridge a step between sheet ends, and over a step it produces a hard edge that catches the light rather than a soft one that does not. Everything flat about a finished ceiling was decided by the battens, the fixing and the sheet layout. The skim only reports on them.

The base has to be right for the plaster to work at all. A board that has been rained on, one fixed with proud screw heads, a surface dusty from cutting — each produces crazing or a blown patch that has to be cut out later. Get the volume before mixing rather than after: it is a small quantity with a short working life, and the way a first skim goes wrong is almost never the arithmetic and almost always a bucket that stiffened while somebody was still working the far corner.