Why Do Slate Roofs Lose Individual Slates?

Roofs Do Not Lose Slates at Random

If nail corrosion were the only thing happening, slates would come off more or less evenly across an elevation. They do not. They come off in places, and the places are consistent from roof to roof.

The verges at each gable. The top two or three courses under the ridge. The corners. The bottom course above the gutter. Around any penetration - a chimney, a vent, a roof window.

That distribution is the clue, and around Macclesfield, where properties on the higher ground take considerably more weather than the town centre, it is more pronounced than most places.

Why the Edges

Wind does not push a roof. It pulls it. Air flowing over a roof accelerates as it passes the edge and the ridge, and accelerated air is at lower pressure, so the roof experiences suction from above.

That suction is not evenly distributed. It is strongest along the perimeter zones - the verges, the eaves, the corners and the ridge line - and comparatively mild over the middle of a slope. On the same roof in the same gale, a slate at a corner is carrying several times the load of one in the middle.

So a slate near an edge is being lifted and dropped by the wind thousands of times a year, and its nails are working every time. Fatigue and corrosion together get there far sooner than corrosion alone. That is why edges go first.

It also explains something people find counter-intuitive: a slate can be lost in a gale on a roof where nothing is wrong with the covering at all. The wind found the one place where the fixing had reached its limit.

Reading the Pattern

Where losses cluster in one specific place, look for a local cause. A run along one verge points at that verge - a failed mortar bed, a missing undercloak, an exposed edge with nothing holding it down. Losses around a chimney point at the flashing or the soakers disturbing the slates around them. A line above a gutter points at the eaves detail.

Where losses are scattered singly across the whole elevation with no cluster, the answer is the fixings as a population, and the pattern is telling you the roof is ageing rather than that something specific has failed.

Where losses are on one elevation only, exposure is doing it. On the hills east of Macclesfield that is usually the west or south-west face.

Stone Slate Is a Different Animal

Some older property around Macclesfield and the hill villages carries graded stone slate rather than Welsh slate - heavy, laid in diminishing courses from large at the eaves to small at the ridge, and traditionally hung on oak pegs over riven battens rather than nailed.

Those roofs lose stones for a different reason: the peg fails, or the batten it hooks over does. And because each stone is sized for its course, a lost stone cannot simply be replaced with any stone - it has to be the right size for that course or the coursing goes wrong and stays wrong.

They are repairable and they are worth repairing. They are not a job to hand to somebody who has only worked with concrete tile, because the individual stones are valuable and getting the diminishing courses wrong is visible from the road for the next hundred years.

What It Means for the Repair

On an exposed roof, putting a slate back exactly as it was puts it back into the same conditions that took it off. Where we refix in a perimeter zone here, the fixing specification reflects the exposure rather than simply matching what was there.

It is also worth looking at the whole edge rather than the one slate. If the verge that lost a slate has lost its mortar bed, the next one along is next, and dealing with the run costs far less than three separate visits.