In the footsteps of Scott: Richard Griffiths Architects add to the gothic fantasy of London’s St. Pancras Hotel.

In the world of conservation architecture, new-build is a ticklish issue. New-build that sets out to fit in with, rather than act as a foil to, a historic building is still more of a challenge. You run an extreme risk of being accused of pastiche, which is regarded as a bad thing. This has been the case since the days of William Morris, the foundation of the Society for the Protection of Ancient Buildings, and the whole principle of “honest repair”. When adding new to old you should, in other words, be able to see the join. So consider this: a new hotel wing, at George Gilbert Scott’s St. Pancras Station in London, will do nothing less than revisit Scott’s take on Ruskinian Gothic. It is, however, subtly different.

This is a long and complex story – in both planning and structural terms – but it boils down to this. As part of the reinvention of St. Pancras station as a transit hub and terminus for Eurostar trains, the former St. Pancras Chambers hotel (empty for years, and prior to that used as offices for years) is being reborn as a hotel and apartments, developed by the Manhattan Loft Corporation (a London company, with no American connections other than the name). To make such a development economically viable in today’s conditions, the hotel needs more rooms. Accordingly an extra wing is planned as a 190-bed “Renaissance” hotel, Marriot’s top brand. It is to run down the western side of the existing building opposite the British Library – itself an early contextual response to the massing, materials and colours of the station. What to do in this highly charged urban setting?

The first response of MLC’s architects RHWL was to distinguish the new building totally from its forbears by putting two stacks of rooms, slightly staggered like passing trains, in a glass box perched on the side of the Grade 1 listed red-brick complex. It was by no means a bad solution, but English Heritage ruled it out absolutely. The new building had to be in the “Scott manner”, it insisted. MLC then brought forward Richard Griffiths Architects – already working with RHWL on the conservation aspects of the St. Pancras Chambers scheme – to design the building. So began a fascinating saga.

Griffiths is described by MLC’s managing director Angus Boag as “a brilliant conservation architect who is also brilliant at new buildings”. And he cites Griffiths’ group of Millennium buildings at Southwark Cathedral as “unsurpassed” in this context. He’s right: the way Griffiths, in his refectory building at the cathedral, makes a sequence from external shallow stone arch via internal roman-tile arch to a Hopkins-like precast concrete vault – all the while working in the context both of the cathedral and its 1970s additions – tells you plenty about the practice’s capabilities. And at first, at St. Pancras, Griffiths and his colleagues tried – as they did at Southwark – to avoid the apparent cliché of the pointed arch. After all, square-topped windows also feature in the original St. Pancras composition, where the design of the western side differs significantly from the grand Euston Road frontage. But in the end, the pointed arch it had to be.

Did the ghost of Ruskin speak through Scott? Already a ground-level arcade and energy centre building for the station on the western flank have been built by others, slavishly copying some of Scott’s gothic details. That was never Griffiths’ intention. More prosaically, it was a piece of structural engineering that finally dictated the form. The buildings on this side of the station are built above the new loading bay in the undercroft. A massive cross-braced steel transfer structure had to be incorporated into the new building. Square window heads did not fit into the structure. Pointed ones did. Form, in the end, followed function.

That is only part of the structural story. The new building had to utilise foundations originally intended for a different building – which, once the massing and the building line of the Griffiths design settled down, meant cantilevering some rooms off at the side. What with one thing and another, sky-hooks were starting to look increasingly attractive. But eventually, a way was found. “What we’ve achieved is a masonry skin at least a brick or a brick and a half thick, using lime mortar,” says Griffiths. Given the complex structural underpinnings there will be movement joints too, carefully concealed.

In the meantime, the aesthetic debate raged, with everyone from the Victorian Society to CABE weighing in. Opinions were often contradictory, some saying that the design was too Victorian, some not enough so. An early design was deemed to be too over-simplified: rather than encrust it with detail, Griffiths and his associate Richard Hill went back to first principles and started again with what he calls “a more constructive Gothic – more Philip Webb than Scott, perhaps – omitting some of the richness of decoration and bringing out what lies behind that.” Key to the final design solution, however, was the use of a Scott technique of incorporating two floors into a single-storey order – something he did when he turned the corner of the hotel onto Midland Road. In the Griffiths design, the floor level within this order is dealt with by means of a metal balustrade panel, in a pattern relating to the aesthetics of both Scott and Barlow, engineer of the great train shed behind. The metal roof and rear elevation of the hotel, meanwhile, is clearly in Barlow’s realm while certain details of its end gable to the north pick up on Sandy Wilson’s British Library opposite.

But in the end, what cleared the air was none of these details, but a simple acknowledgment that it was OK to borrow from Scott’s composition of the front of St. Pancras, which consists of two layers of pointed windows, then a zone of triple-arched windows, and finally attic rooms in a steep roof. A lot of the early design problems had stemmed from trying precisely to avoid this. Or as Richard Hill puts it: “Rather than seeing this as a subsidiary extension to the Scott building, there was virtue in it being a more literal version of the Scott order. We were compelled in the end to continue that.”

In a final twist, the scheme that won planning permission and listed building consent then had to be modified because it was found that a screen wall of brick arches continuing the first-floor façade treatment to enclose the hotel’s garden forecourt would be too heavy for the finely-balanced structural gymnastics of the complex. The proposed screen wall had to be removed and replaced with a balustrade which opens up the building more visually on this side.

After all the to-ing and fro-ing, the hotel building that arises along Midland Road will be rather grander than Griffiths and Hill originally imagined, and also much closer to Scott than they had envisaged. But there is logic to this, as well, in that the building will be a new expression of the famous hotel, bursting out in a new direction. You can certainly question the original English Heritage directive to build in the “Scott manner”. But if you are constrained to work within this style, why be coy about it? After going back to read Ruskin’s Stones of Venice, noting the Venetian elements that Scott had incorporated into his hotel, and starting all over, Griffiths was led to an inescapable conclusion. The way Scott composed the facades of his original hotel was ingenious, adaptable, and pretty much fit for purpose in the vastly changed circumstances of today.

RIBA Journal: http://www.ribajournal.com
Richard Griffiths Architects: http://www.rgarchitects.com
Manhattan Loft Corporation: http://www.manhattanloft.co.uk
RHWL Architects: http://www.rhwl.co.uk/rhwl_94.aspx

Text © Hugh Pearman, all drawings by Richard Griffiths Architects. First published in the RIBA Journal, December 2006.