What a Change of Roof Covering Reaches Back Into

The covering record puts one question to professionals in plain terms: is the existing structure being asked to carry a covering of different weight from the one it was designed for. That is a structural question rather than a roofing one, and it is asked before the covering is chosen rather than after the old one has come off.

It runs in both directions. A heavier covering asks more of rafters, purlins and the walls beneath them. A lighter covering changes the balance between the weight holding the roof down and the wind trying to lift it, which is why the restraint arrangement is reconsidered rather than assumed to carry over. What either means for a specific roof is established by the engineer responsible for it.

The record describes fixings as tying the covering to the wind rather than to the roof, and notes that uplift concentrates at eaves, verges, ridges and around anything that interrupts the surface, so the restraint pattern is not uniform even though the covering looks uniform. Where clips are left out at those edges, units lift and the underlay is exposed to a duty it was never the primary defence against.

Batten position and covering overlap are described as a single decision rather than two. Moving the battens further apart increases what is left exposed on each course and shortens the head lap, so the covering becomes more open to wind-driven rain at exactly the moment the underlay is being asked to do more. A new covering with a different unit size therefore arrives with a new batten setting-out, and the old battens are not a starting point.

Once the covering is off, the layer beneath it is visible for the only time in the roof's life. The record notes that covering, underlay and fixings do not weather at the same rate, and that an underlay degraded by ultraviolet exposure at the eaves, or fixings corroding against an incompatible metal, can end a roof whose visible units still look entirely sound.

There is also a coordination question. The record asks whether the specified underlay assumes ventilation elsewhere in the roof, and whether that ventilation is actually present. A sheet chosen for a roof that dries upward behaves differently in a roof that does not, so an underlay substitution made on the scaffold is a change to the build-up rather than a change of product.

Where a re-cover happens alongside insulation added above the rafters, the covering plane rises, and the rafter-line record describes the consequence bluntly: the eaves has to come further out, the verge has to be rebuilt in a new position, the abutment flashing sits higher and any rooflight kerb grows. A decision taken about a layer in the middle of the roof arrives as a redesign of the perimeter.

Even without added insulation, a change of covering can move the plane, because unit thickness, batten depth and whether a counter-batten is introduced all sit between the structure and the visible surface. Anything that moves that plane moves the relationship between the covering and every termination formed around it.

Where the roof meets a wall, the abutment record describes weathering as deliberately layered rather than sealed: the roof turns its own covering up the face, the vertical element oversails that upstand with a flashing retained in a chase or a bed joint, and in a cavity construction a tray inside the wall collects water descending within it. A re-cover reaches the first two of those and cannot reach the third, because a tray sits inside the wall and cannot be added later without opening the construction up.

The verge is the other edge worth naming. The junction record describes it as the position most exposed to wind uplift, where the covering has to shed water sideways without delivering it onto the gable wall, and where the last course has support on one side only. Securing, weathering and support all meet on one line, which is why a verge is rebuilt with a new covering rather than reused.

A change of covering is frequently a change of metal somewhere. The covering record asks which metals will sit in contact at fixings, flashings and rainwater goods, and whether they are compatible, and notes that fixings corroding against an incompatible metal can end a roof whose units still look sound. The standing seam record extends the question to run-off, asking which metals the water leaving the roof passes over before it leaves the building.

That is a question about the whole path rather than about a component, and the records treat it as part of the design rather than a decision made on the roof. It is also one of the few things about a re-cover that can be settled on paper before anything is ordered.

Where the roof abuts a wall belonging to an adjoining building, the abutment record notes that working on a wall belonging to another owner may carry legal obligations, and asks who is entitled to carry out work on it. That question sits outside this reference entirely, and it is worth raising before a programme depends on the answer.

The same record also points out that provision for a concealed tray has to be made while the wall is being built, so where a re-cover arrives at an abutment with no tray behind it, the remedy involves taking part of the wall apart. Establishing what is there, rather than assuming, is what turns that from a surprise into a decision.