Recovering lost features

"It's always nice to prove an architect wrong," says a grinning Brian Phipps. Halfway through cleaning and restoring the West Front of St Paul's, not only has the Wates project manager set the architect straight, but all the other experts on Christopher Wren's masterpiece as well.

When the restoration project started last May, nobody knew for sure what kind of stone the grand flights of steps at the West Front were, including the only original one, the 'Wren step' at the very top. Drawing the wrong conclusions from vague documentary evidence, the architects suggested an Irish quarry on the basis that Wren's stone agent was Irish. The real place of origin, though, lay somewhere else, as Phipps discovered when colleague Alan West, project manager for masonry subcontractor Stonewest, investigated the matter further.

Because of the huge numbers of visitors and inadequate foundations, the West Front steps have tilted and lifted over the years, and rain is now entering the vaults below. In the 19th century the same problem led to the replacement of the originals with different stone. Just the 22 stones in the Wren step were in good enough condition to stay. Now they're knackered. But what was the stone precisely?

Architects tend to rely on archival records, but years of experience told West. He was convinced that the Wren step wasn't an Irish stone. Not that that helped very much. Nor did the suggestions of the other architects and stone companies who came to take a look and pointed every which way: Belgium, South Africa, Italy. All proved duff when samples came back and failed the laboratory analyses.

So West sent off a sample of the Wren step to geology consultant David Jefferson. Black limestone from Pooil Vaaish quarry on the Isle of Man was his verdict. And when a sample came from the quarry, which also produced documents recording consignments that tally with Wren's stone delivery notes, it convinced the architects.

Blemish-free beds
Pooil Vaaish's stone beds proved to be deep enough to quarry the stones, but length was more of a problem. Now, the Wren step is 30m long. It's not a single piece of stone, either. It's made up of 22 individual stones. West had to recreate the step exactly. That meant finding runs of stone between 600mm and 2900mm without too much calcite veining or fool's gold. It took the quarry almost three months to haul out the longest stone required.

For the rest of the step repairs, West took plugs of rock from the underside of the existing stones to patch the small holes left after removing the bronze handrails.

But getting the stone right is only part of the story. Wren modelled the West Front's flights of steps on those of the Parthenon, incorporating an entasis or bulge in the middle to counteract the optical illusion which makes a long horizontal look as though it is sagging in the middle. So before taking up the stones – 565 individual slabs, between 1m and 3m long, and weighing up to 350kg – West took laser-level readings at the four corners of each. Averaging the results allowed him to plot a curve for the steps that reproduced the original bulge.

Also at work on the West Front are masons carving new blocks of stone to replace the most disastrously eroded parts of the facade. A gentle touch is a necessity. "The masons can't be overzealous in chipping out decayed mortar on the carvings," says Phipps. "Don't dig out that last little bit – the Surveyor of the Fabric will go ape." The Surveyor, Martin Stancliffe, tours the building every week to decree what shall and, more importantly, what shall not be touched.

Stone love
Such is Stancliffe's reverence for the stone that the restoration almost didn't get off the ground. He was adamant that Phipps could not fix any access scaffolding for the stonemasons and cleaners to the facade. "He only backed down when I told him the job couldn't be done without some," says Phipps. "Even so, scaffolding was a nightmare. It took five months to put up."

So Phipps halved the number of fixings that penetrated the fabric, keeping the scaffold stable by vastly increasing the amount of cross bracing. The result is 54 miles' worth of scaffold that earns your hard hat repeated and heartfelt gratitude.

The 50-odd ties that do penetrate the walls are fixed in recesses or areas that will subsequently be covered by asphalt. The ties also use neoprene sleeves rather than the expanding type. When the scaffold goes down, unscrewing the ties from the wall will make the sleeves go flat; they can then be pulled out of the wall rather than left in.

Previous attempts to remove the carbon deposits on St Paul's Portland stone blasted off several millimetres of surface, but Phipps is using the gentler Hodge Clemco system. This allows stone cleaners to regulate streams of air, sand and water independently. Cleaning is a craft skill, not a drudge's post, and clinker that has fused with stone is left well alone.

"People are proud of what they're doing," says Phipps. "There are never any problems with subcontractors messing about. Quality control here is easy."

And the pride has always been there. The original masons chiselled their names in copperplate, far above the visitors' gaze, where Wren himself saved a few pennies by using a cheaper grade of Portland, which bristles with scallop shells.