Estimating Coatings for Block Walls and Timber Fences

Two Walls, One Tin, and a Factor of Four

The north elevation had been coated twice before. Same product, same sleeve, same afternoon, and it gave back very nearly the three hundred and fifty square feet to the gallon printed on the pail. The south elevation was bare lightweight block, identical dimensions off the same drawing, and it took the first coat at something closer to ninety. Nobody had done anything wrong and nothing had been thinned. The two walls were simply not the same surface, and only one of them resembled the flat panel the number on the pail was measured against.

That gap does not close by buying better material. A spread rate is a statement about a coating, not about your wall: it is derived rather than observed, and the derivation assumes every drop that leaves the pail ends up as dry film on a plane. Open concrete block and rough-sawn softwood both break that assumption, and they break it by multiples rather than by the ten or fifteen per cent that a waste allowance is built to absorb. Getting it wrong is rarely a disaster on a single elevation. It is a disaster across a run of six, when the shortfall shows up on the last one and the merchant has shut.

So the skill worth having is not hunting for a better published figure. It is knowing the handful of places material goes when it does not become film, carrying each of them as its own line, and settling the argument on a measured panel wherever the job is large enough that being wrong costs a day. The two substrates below punish this hardest for opposite reasons: block loses coating into itself, and sawn timber loses it along the grain.

Where the Printed Rate Comes From, and What It Assumes

The number is arithmetic, and it is worth being able to reproduce. Coverage in square metres per litre is ten times the volume solids percentage divided by the dry film thickness in microns. In the other system it is 1,604 times the solids fraction divided by the film thickness in mils. A coating at 38 per cent volume solids specified to 50 microns dry works out at 7.6 square metres per litre, which is a shade under 310 square feet to the gallon, and both routes agree because they are the same sum in different clothes. Nothing in it is a fudge, and nothing in it is marketing.

Each term traces back to a test. Volume solids is measured to ASTM D2697 Standard Test Method for Volume Nonvolatile Matter in Clear or Pigmented Coatings, which is what makes the figure comparable between two manufacturers rather than a claim. Film thickness has its own instruments and its own practices: ASTM D4414 Standard Practice for Measurement of Wet Film Thickness by Notch Gages for the comb you keep in the van, ASTM D7091 for gauge readings over ferrous and non-ferrous metal, and ASTM D6132 for ultrasonic readings over substrates a magnetic gauge cannot see. What none of those documents will tell you is what your wall does, because they were all designed around a surface that behaves itself.

The protective coatings world names the gap outright, which the decorative end of the trade largely does not. A specifier working to ISO 12944-5 Paints and varnishes — Corrosion protection of steel structures by protective paint systems — Part 5: Protective paint systems is dealing in a theoretical spread rate and a practical one as two distinct quantities, and the second is always the smaller. ISO 19840, which covers measurement and acceptance of dry film thickness on rough surfaces, goes further and publishes correction values keyed to the profile grades of ISO 8503-1: a gauge sitting on the peaks of a blast-cleaned surface reads high, because coating has gone into the valleys where it does no protective work. Those corrections are written for steel and do not transfer numerically to block or timber. The principle behind them transfers exactly.

Break the loss into destinations and it stops being mysterious. Some of the pail becomes film on the plane you measured, which is the only part the printed rate describes. Some follows a profile the tape measure never saw, because the developed area of a textured face is genuinely larger than its measured area. Some is absorbed below the surface into pores, where it is spent without contributing anything to hiding or to film build. Some never leaves the equipment. And some goes past the work altogether. Only the first of those five is in the published number, and on a bare block wall it can be the minority share.

Rough-Sawn Timber Drinks Along the Grain

Wood does not absorb evenly in every direction, and that single fact explains most of what surprises people about staining a fence. The vessels and tracheids that carried sap run along the length of the board, so a cut end is a bundle of open tubes and a face is a wall of closed ones. Stain applied to end grain travels axially until the liquid runs out, and it goes far past anything that could be called a film. Every picket top on the run, every rail cut at a post, every post top and every rip down a board to close out a bay is a piece of end grain, and collectively they are why the first litre disappears faster than the second.

The reference for the underlying behaviour is the USDA Forest Products Laboratory's Wood Handbook — Wood as an Engineering Material, published as General Technical Report FPL-GTR-190, which covers how wood weathers and how finishes behave on it in far more depth than a data sheet can. Two of its themes matter at the counter. A sawn face carries torn and lifted fibre that both raises the developed area and opens capillaries a planed face keeps closed, which is why the same product on machined capping and on rough-sawn boards is two different consumption figures on one fence. And weathering does the same thing over time: ultraviolet breaks down surface lignin, fibres lift and grey, and a fence left bare through a summer will drink noticeably harder than the same fence coated on the week it was built.

New pressure-treated stock is the opposite failure, and it costs material rather than saving it. Timber that has come straight from the treatment cylinder still has cell cavities full of treatment solution, so stain sits on the surface, runs, and cures badly if it cures at all. AWPA U1 Use Category System: User Specification for Treated Wood is what specifies the treatment for a ground-contact post or an above-ground rail, and AWPA M4 Standard for the Care of Preservative-Treated Wood Products covers handling and field-cut treatment on site. Neither publishes a waiting period before finishing, and neither should, because it depends on the treatment, the species, the section, the stock rotation at the yard and the weather since. Sprinkle water on a face: beads mean not yet, a dark patch that soaks away in seconds means ready. That test takes ten seconds and it is worth more than any number somebody quotes you.

Product class swings the figure more on timber than it does on masonry, so decide it before you measure anything. BS EN 927-1 Paints and varnishes — Coating materials and coating systems for exterior wood — Part 1: Classification and selection sorts exterior wood coatings by the film they build and by how much dimensional movement they are meant to tolerate, and the classes are not interchangeable quantities. A penetrating semi-transparent oil is bought to soak in, so the thirstier the timber the more of it goes; a film-forming opaque stain is bought to bridge, so it behaves more like paint and less like a treatment. Running the two through the same coverage assumption is a mistake that only shows up when the pails run out.

Loss That Never Reaches the Substrate At All

Some of the pail is spent before the wall sees any of it, and on small jobs this is a bigger share than anyone admits. A block filler is a high-build material with the consistency of thin plaster, and a deep-nap sleeve loaded with it holds a startling amount that is thrown away with the sleeve at the end of the day. A hopper gun and its hose hold a measurable volume before the first pass leaves the tip, and that volume is charged to the first wall of the job whether or not it lands there. So does the film left round the sides of a pail that has been rolled out of rather than boxed and scraped.

Spray adds the loss that is genuinely hard to bound. Transfer efficiency — the fraction of what leaves the gun that ends up on the work — depends on the gun, the tip, the pressure, the distance, the operator and, outdoors, on the wind, and there is no published figure that describes your setup on your fence line. Airless onto a spaced picket fence is the worst case anyone routinely meets, because a large share of the fan passes between the pickets and lands on the grass, and back-brushing afterwards is a labour line rather than a material saving. Where the regulators care about transfer efficiency they measure it for a specific process rather than publishing a constant, and that is the honest position to take on site too: set the allowance from what your own jobs have actually used, and write down the used figure each time so next year's allowance is evidence rather than habit.

Cut-in and detail scale with length, not with area, which is why they behave so oddly on small jobs. A single-bay gate return costs almost the same in edge work as a ten-bay run and carries none of the area to absorb it. Keep it as its own line on anything with a lot of arrises, mouldings, capping or trim, and stop trying to bury it in a percentage of the face area, where it is invisible until it is not.

Settle It on a Trial Panel, Not in the Van

Everything above narrows the range. Only a panel closes it. The test is one square metre of material and about twenty minutes, and it replaces every multiplier on this page with a measured rate for the actual wall, the actual product and the actual application method — which is the only combination that matters and the only one no document can know about in advance. On anything above a few hundred square metres of open block, it is the cheapest insurance on the job by a wide margin.

Do it on the worst face rather than the most convenient one. Coating estimates fail at their extremes, not at their averages, and a trial run on the sheltered dense elevation tells you comfortable things about a wall you were never going to run short on. Where a delivery has visibly changed partway along a building, or where one elevation has weathered and another has not, treat them as two walls and prove both.

1. Mask a measured square metre on the worst face on the job, not the handiest one.
2. Gauge or weigh the pail, coat the panel to the finish you have actually sold, and gauge or weigh it again.
3. Divide the panel area by the volume that left the pail: that quotient is the practical spread rate for this wall, this product and this method.
4. Repeat wherever the substrate visibly changes — a second block delivery, a weathered elevation, a stretch that was coated before.
5. Divide the net measured area by the practical rate, then add the application allowance as its own line rather than folding it into the rate.
6. Order up to the next whole pail and record the rate you measured against the substrate you measured it on.
7. Leave the trial panel where the main coat will run straight over it, so the test costs a square metre of material and nothing else.

What Governs This, and What Nobody Publishes

It is worth being clear about which parts of this have a document behind them and which do not. The spread rate arithmetic, volume solids, film thickness measurement and the theoretical-against-practical distinction all do, in ASTM D2697, ASTM D4414, ASTM D7091, ASTM D6132, ISO 12944-5 and ISO 19840. Where dry film thickness has to be demonstrated rather than assumed, SSPC-PA 2 Procedure for Determining Conformance to Dry Coating Thickness Requirements, now published by AMPP, sets out how many readings, where, and what constitutes conformance. Substrate preparation and condition have ASTM D4258, ASTM D4259 and ASTM D4263, and the masonry unit itself has ASTM C90 and ASTM C140. All of those are real, and citing them is not decoration: they are what an argument with a specifier is settled against.

What does not exist is the thing everyone actually wants — a published table converting a texture class into a coating multiplier, or a species and sawing factor for fencing timber. No standards body publishes one, because the quantity depends on the coating as much as on the substrate and would have to be republished for every product. The multipliers behind the block filler tool on this site are described in its own sources as a rule of thumb rather than dressed up as a table, and that is the correct status for them. They are the right thing to use on the way to the merchant and the wrong thing to keep using after you have a measured panel.

The document that does know your product is the data sheet, and it is under-read. Block filler sheets publish coverage as a span rather than a single figure precisely because the manufacturer cannot see your wall — the Sherwin-Williams PrepRite Block Filler sheet is the widely stocked example of the form — and where a sheet offers a coarse-block or rough-sawn figure of its own, that measured number beats any rule of thumb, including the ones here. Use it, and do not then apply a multiplier on top of a rate that has already been derated. Finally, keep the record. A survey sheet that says lightweight block, open under a raking lamp, 2.1 square metres per litre measured on a panel, 46 litres used across 90 square metres is worth more on the next job than every published figure in this article combined, because it is the only one that was measured on a wall you have actually stood in front of.