All processes
PRC.41 · Construction Process

Wood Shingle and Stone Slab Roofing

These are the oldest roof coverings there are, and they work on a principle no modern material has improved on: no element is waterproof, no joint is sealed, and yet the water does not get in. It gets in and finds its way out. Watertightness comes not from closure but from overlap, which is why a shingle or stone slab roof is not designed in millimetres of membrane: it is designed in degrees of pitch and centimetres of cover.

Watertightness principle
overlap, not sealing
Typical head lap
one third of the length or more
Weight of a stone slab roof
several times that of clay tiles
Ventilation below the covering
required on both faces

Overview

TAV. 00

Timber shingles and stone slabs belong to worlds far apart in weight and workmanship, but they share the same constructional logic, which is why it is worth treating them together. In both cases the covering is made of small, loosely laid elements overlapping in successive courses: water that gets past the first element meets the second, and water passing between two side-by-side elements finds the element of the course below beneath it. Three practical consequences follow. First, pitch is not an aesthetic matter but the parameter governing watertightness, and when it drops the head lap must grow, meaning more material to cover the same area. Second, the covering must be able to breathe and dry from both faces, because it is made of materials that get wet for a living: a discontinuous covering laid on a closed substrate rots, and not because of the rain. Third, weight: a stone slab roof can weigh several times a clay tile covering, and the structure must be checked beforehand, not after seeing the quotation for the stone.

The site score

TAV. H
NOT TO SCALE1 of 5 layer in placeTimber load-bearing structure

The detail of the craftthe weight that changes everything

COPPILOSEstessa superficie, carico molto diverso:la verifica si rifà, non si eredita
01Days 1-3

Checking the structure: the weight comes before the covering

The existing structure is surveyed, or the new one sized, on the actual weight of the chosen covering, taking the site's snow load into account as well. Bearings, member sizes and the condition of the timber are checked, and it is decided whether the build-up will be ventilated in one cavity or two.

Why it is done this wayA stone slab roof can weigh several times a clay one, and that changes the nature of the design: you are not choosing a finish, you are loading a structure. The commonest mistake in refurbishment is replacing a light covering with stone because stone is what was originally there, without redoing the check: the roof stands, but deformations grow for years and the timbers work beyond what was assumed. Then there is a factor that in the mountains weighs as much as the stone: snow, which does not merely add load but accumulates asymmetrically, more on one slope than the other, and which on discontinuous coverings tends to stay rather than slide off. Self-weight is known by calculation; the rest is known by looking at the site and asking the people who live there.
peso reale · not that of the covering being replacedneve · load to assess as an asymmetric drift
In handresistograph for the existing timbergeometric survey of the structure
at altitude you work in the dry season, not right before winter
The mistake that costs dearlyRebuilding a stone slab roof as it originally was without redoing the structural check: the structure of that time had different sizes and quality, and it has been working for a century since.
Site supervision checkCheck the calculation against the covering's real weight and the local snow load; inspect the beam ends bearing on the masonry, which is where old timber fails first.
«You do not choose the stone by looking at the roof: you choose it by looking at the beams underneath.»site notebook — editorial synthesis
The beats — with site daysbeat 1 of 5
1-2 seasonsnewly laid shingles settle and change colour unevenly: the covering should be judged after the first wet-and-dry cycle, not at handover

click a beat · arrow keys ← → walk the site · dashed pauses are the chemistry at work

The principle that holds it all up

No element is waterproof, and yet the roof holds

TAV. G1

A discontinuous covering does not stop water: it escorts it. Every element is laid loose, without glue or sealant, and between one element and its neighbour a gap always remains. Watertightness arises because beneath that gap there is not a void but the element of the course below, deliberately staggered. Water entering through a gap meets a sloping surface that carries it back out, and to reach the structure it would have to cross two or three aligned elements, which head lap and stagger make impossible. It is a system that accepts being permeable at every single point in order to be reliable as a whole, and it is why these roofs are repaired a piece at a time instead of being replaced.

FILA SUPERIOREFILA INFERIOREL'ACQUA ENTRA DALLA FESSURAe l'elemento sotto la riporta fuoriRICOPRIMENTOpermeabile in ogni punto, affidabile nell'insieme

A qualitative diagram of the principle; head lap values depend on pitch, wind exposure and element type, and should be taken from local established practice, which on these coverings remains the most reliable reference.

The arithmetic that surprises

Less pitch does not mean less material: it means more

TAV. G2

On a continuous covering the pitch has little effect on quantity: a membrane covers the same area however it is inclined. On a discontinuous covering the relationship reverses, counter-intuitively. Because watertightness depends on head lap, and the lap must grow as the roof flattens, lowering the pitch means bringing the courses closer together: the useful part of each element, the part actually exposed, shrinks, and covering the same area needs more courses and therefore more elements. This is why a choice made for reasons of appearance or eaves height later shows up as a surprise in the quantities, and why on these coverings pitch should be discussed at the outset, while it is still a design decision and not a budget problem.

PENDENZA FORTEricoprimento minorefile rade, meno elementiPENDENZA DEBOLEricoprimento maggiorefile fitte, più elementistessa superficie coperta: la pendenza si paga in numero di file

A qualitative comparison for the same covered area; the size of the effect depends on the element format and the lap rule adopted, but the direction of the relationship never changes.

How it ages (and what betrays it)

TAV. P
MARCESCENZA DAL BASSOShingles fail on their underside while the exposed face still looks sound: the ventilation cavity is missing or the covering was laid on closed boarding (phase 02). Decay is already extensive by the time it shows from outside.
VIA D'ACQUA ALLINEATAWater penetration recurring at the same point after every heavy rain: vertical joints aligned between successive courses (phase 03). It is not a material defect but a laying one, and is corrected only by relaying the courses involved.
FISSAGGI CONSUMATIThe covering lifts in sheets in strong wind after decades of apparent calm: nails incompatible with the timber's tannins, corroded down to the head (phase 04). The covering is still good, but it is no longer attached to anything.
SPACCATURA LUNGO LA FIBRAElements split at the nails after only a few seasonal cycles: fixings driven home hard, preventing the timber from moving with humidity (phase 04). Every split becomes a water path down to the course below.
CEDIMENTO PER SOVRACCARICOProgressive deformation of the structure under a stone covering: the structure was checked against the previous covering and not the one laid (phase 01), often made worse by asymmetric snow drift. It shows as slow sagging, not as a break.

Questions from the site

TAV. Q
How long does a shingle roof really last?

The honest answer is that it depends on three things, and none of them is the quality of the shingle. The first is exposure: a north-facing slope, which dries slowly and stays damp, has a noticeably shorter life than a south-facing one, on the same roof and in the same material, and it is normal to find historic roofs where the two slopes were renewed at different times. The second is ventilation: a covering that dries from both faces lasts far longer than one laid on a closed substrate, and this is the difference the designer can actually govern. The third is the timber species and how the shingle was produced, because a shingle split along the grain behaves much better than a sawn one, where the cut severs the fibres and opens a path for water. Put these conditions together and you commonly speak of decades, with maintenance replacing individual elements along the way. But a well-made covering on the wrong roof will not reach half that time.

Can shingles be treated or painted to make them last?

Generally no, and the reason is worth understanding because it applies to all exposed timber. Any film-forming treatment, meaning one that leaves a surface skin, makes things worse rather than better on a discontinuous covering: the film slows evaporation without really stopping water entering at the edges and joints, so moisture stays inside the element longer. The film also ages and flakes unevenly, and renewing it would mean treating a whole roof element by element, which nobody ever does. The correct strategy is the opposite and simpler: choose a naturally durable species, guarantee ventilation on both faces, and accept that the covering will go grey. That grey is not decay, it is the surface layer oxidised by ultraviolet light, and it is the natural protection of the timber beneath. If the client does not want the grey, the problem is not the treatment: it is that they probably do not want a timber roof.

Can a stone slab roof go on any roof?

No, and the reason is almost always structural before it is technical. Stone weighs many times more than a clay covering and still more than a metal one, so the first step is not choosing the stone but verifying that the structure can carry it, including the snow load, which in the places where stone slabs are traditional is never negligible. In refurbishment there is a further fact often taken for granted: that the roof originally had stone slabs does not prove it can carry them today, because the structure has worked for decades, the beam ends at their bearings are the most degraded point, and the sizes of that time were not calculated as we calculate now. Then there is a geometric condition: stone slabs need generous pitches, because their head lap cannot grow indefinitely and their irregular surface slows the run-off. On a flat roof plane a stone covering does not behave as it should, and no laying technique recovers the pitch that is missing.

What maintenance does it need, in practice?

Very little, but not none, and the difference between the two is the entire service life of the roof. The useful inspection is annual, preferably after winter, and consists of three simple things. Look at the covering from outside for split, slipped or missing elements, which should be replaced individually and at once, because while they are few the repair is trivial. Clear gutters and valleys of leaves and needles, which accumulate on these roofs more than elsewhere because of the irregular surface. And above all go up into the roof space with a torch on a sunny day and look upwards: points of daylight visible from inside and dark stains on the structure tell you in minutes what cannot be seen from outside. The client should also be left a stock of elements from the same batch, because years later they can no longer be matched for colour and format, and a repair made with different material shows forever.

Materials involved

TAV. M