Veridella

Notes on traditional copper, lead and zinc roofing, from the first survey to the last patina.

What a 1930s copper roof tells you before you strip it

An old copper roof has already written most of its own replacement specification in bay widths, opened welts and torn clips. Here is the order to read it in before anyone lifts a seam.

By Ruth Albescu, , nine-minute read, in Survey

Before a single clip is lifted, an old copper roof has already given you most of the specification you need to replace it. The trouble is that it gives it in a language of tarnish, sag and staining, while the job sheet usually asks only for ‘strip and re-cover like for like’. What follows is the order I work in on an interwar copper roof, written up after a run of three on civic buildings in the West Riding this summer.

Start with the bay widths, and measure every bay, not one. Interwar copper was usually supplied as 1.8 m × 0.6 m sheets, which, once you take off the turn-ups for a double-lock seam, leaves a finished bay of about 525 mm. If you find bays of 450 mm or less, either someone was working with narrower sheet or they were nursing a structure that moved, and you want to know which before you specify 600 mm strip. On the second of our three roofs, two bays by the parapet ran at 410 mm, and the boarding under both had been patched twice.

Next, the cross-welts. Short-sheet roofs of this age have a transverse joint every 1.8 m or so, staggered between bays. Count the welts that have opened and note which side has lifted. A welt that has lifted on its downslope edge has been working against clips that should have been able to slide. One that has split along the fold usually means the sheet was about 0.45 mm (roughly 20 oz to the square foot) rather than the 0.6 mm you will be fitting now, and that the fold was dressed too hard on a cold morning.

Then the clips. Open one seam at the ridge and one at the eaves with a seam opener and look at the cleats. Original work typically has copper cleats at 300–375 mm centres, fixed with copper clout nails into softwood boarding. If the cleats have pulled their nails, the boarding has gone soft and you are now pricing a deck, not a covering. If the cleats are sound but the sheet has torn round them, the roof never had sliding clips. Those sheets have spent ninety years fighting their own expansion, which comes to about 1 mm per metre for every 60 °C swing between a winter night and a July afternoon.

Look at the patina last, and don’t read it as a sign of health. Uneven green across slopes is normal, and a roof that has gone fully green has not necessarily stayed watertight. Bright green streaking below a single seam usually marks a split or an opened welt just above it. A black, almost lacquered patch that has never turned is generally a dip where water stands after rain. Both belong on the drawing, because both tell you where the new sheet will be asked to work hardest.

Then draw it all. Make a 1:50 roof plan with every bay width written in and every opened welt and torn clip marked, plus a 1:5 section at the worst abutment, usually where the copper turns up against a parapet or a stair turret. The conservation officer will want this for the consent. The next roofer, in another ninety years, will want it more.

Finally, like for like rarely means identical. On these three roofs we kept the bay rhythm and the double-lock profile, because both read clearly from the street. We changed the sheet to 0.6 mm, used long-strip lengths of up to 6 m where the falls allowed, put a fixed zone of clips in the middle third of each length with sliding clips beyond it, and laid a structured underlay over new boarding. Nobody looking up from the pavement will see the difference until the new copper starts to turn.

Longer notes

Single lock or double lock on a 9° pitch: the case for the second fold

At low pitches, the extra fold in a standing seam is cheap insurance against wind-driven water, and a seaming machine has made it cheaper still.

By Tomasz Wierzbicki, , six-minute read, in Seams & Rolls

On a 9° slope a single-lock seam is a bet on the weather. Rain driven up the roof by wind sits against the seam far longer than it would at 30°, and a single 180° fold leaves a short, straight path for water to be drawn in by capillary action. The second fold of a double lock adds another turn of metal that the water has to climb, and that turn is what the guidance we work to is paying for when it moves to double lock on shallow pitches.

The cost is mostly time. On a 525 mm bay, closing a double lock entirely by hand takes our two-person team noticeably longer per metre than a single lock. With a seaming machine, the gap almost disappears: we set the first metre by hand, clip the machine on, and let it run the full length of the strip. The machine also gives a more even fold than tired hands manage on the fortieth seam of the day.

We still use a single lock at low pitch in one place: on short, sheltered runs such as a bay window or a dormer cheek, where the seam length is under 2 m and the wind cannot get a run at it. Even there, if the building faces the prevailing weather, we fold the second lock anyway.

Brochantite, antlerite, atacamite: reading the colour of a patina

The green on a copper roof is not one mineral, and which one forms depends on what the air carried while it weathered.

By Hannah Oduya, , seven-minute read, in Patina

Copper does not go straight to green. It first darkens to a brown-black layer that is mostly cuprite, a copper oxide, and that layer is what most new roofs look like for their first several years. Green appears later and patchily, as copper sulfates and chlorides build up over the oxide. In inland towns the stable green is mostly brochantite, a basic copper sulfate. Where sulfur pollution was heavy for decades, as on many older industrial-town roofs, more acidic conditions favour antlerite. Near the sea, chlorides give atacamite, which tends to look slightly bluer and more powdery.

Timing varies more than clients expect. On an inland British roof, the green typically arrives somewhere between ten and twenty years in, and it comes faster on slopes that take the prevailing rain and slower on sheltered ones. A copper roof that is still brown after a decade is usually not a fault, just a dry, clean site.

Factory pre-patinated sheet has its place, mainly for repairs that need to sit beside an old green roof without looking like a patch. Tell the client that it will not stay exactly that colour. The factory layer will carry on weathering in the local air, and in a few years the repair will drift toward whatever the roof around it is doing.