Why retrofit first should not be retrofit only

We walked the M&S Marble Arch site with the inspector at the start of the planning inquiry last month. The visit helped explain why we are proposing to replace rather than refurbish the buildings on the site.

The Inspector saw an existing retail environment that is heavily constrained: chaotic and confusing for customers, arranged over five poorly connected floors in three separate buildings. None of these buildings were originally designed for their present function.

Servicing this space, with its nine cores and ten separate storage spaces, is hugely challenging. The Marble Arch store is not performing and unless the site can be redeveloped, M&S has been very clear that they will close it.

Retail has got tougher with online competition and high streets need to enhance their offer to compete. Yet decline is not inevitable – new investment at the eastern end of Oxford Street has seen its fortunes rise.

Our plans at Marble Arch will halt the decline at the western end of the street. Selfridges and the other retailing neighbours agree and wrote in strong support of the proposals.

The quality, character and inherent structural constraints of the disparate collection of buildings on the site severely impacts their refurbishment potential. They date from 1930, 1970 and 1985.

The 1930s Orchard House occupies only a third of the site and has been rejected for listing and excluded from any of the three conservation areas that surround the site. The neighbouring buildings, Neale House and 23 Orchard Street, are both of exceptionally poor quality and detract from the setting of the Grade II* listed Selfridges and surrounding Conservation Areas.

A site visit here really brings home the poor quality of the environment created by the existing buildings for pedestrians. A low undercroft on Orchard Street was described by a SAVE witness as ‘unpleasant’ and the dire service-dominated rear on Granville Place makes no positive contribution to the public realm.

Our new build proposals would transform this unacceptable condition by consolidating servicing discretely and safely to the north of the site and demolishing the poor quality 1970’s bridge over Portman Mews. This allows us to create a landscaped square space in Granville Place and to restore historic permeability across the site to Orchard Street through a new top-lit galleria.

On Orchard Street, the undercroft is swept away, and a broad pavement created by setting the new building line back from that of the existing buildings, with new trees and space to enjoy open views to Selfridges. On any view, these are significant benefits to the quality of place.

These important enhancements to the public realm cannot be delivered through a refurbishment of the existing buildings. You can’t relocate the servicing as the existing ground floor level is too low and, even if the whole first floor was removed, you would still also need to take out three cores and nine structural columns to form the service yard.

Nor could you restore the east-west link as the route is blocked by the existing central core (and the quality of the space with the low ground floor would be more of a tunnel than a galleria). Nor, indeed, could you remove the undercroft on Orchard Street because the existing façade line leaves an inadequate pavement width.

The site visit took us on to view David Chipperfield’s modernisation recently completed at the eastern end of Selfridges. Here we were able to experience beautiful five-meter-high spaces, regular open columns grids and great daylight. These are all the characteristics delivered by our new store proposals for M&S and all entirely precluded by the low ceilings, dense column grids and other constraints of the existing buildings on the site today.

We finished our inquiry site visit at the Kensington Building, a retrofit completed by our practice, Pilbrow & Partners earlier this year. The Kensington site has some interesting parallels with Marble Arch – both sites are former department stores located on important retail thoroughfares. Both are next to a Grade II* listed neoclassical neighbour. At Kensington; Derry and Toms, at Oxford Street; Selfridges.

Aspects of our architectural response resonate between the two sites, for example the proposed use of a white Roman Brick, and the enhancement of the public realm through the new permeability and active frontages. I think the scheme shows the quality of detail and material we would bring to the transformation of M&S at Marble Arch.

In evidence, I explained that as a practice, we refurbish in preference to replace where appropriate, that we are signatories to Architect’s Declare and endorse retrofit first where feasible. The nature of the requirements for the site and of the existing buildings must, however, inform the correct approach.

At Kensington, we started with a single building with generous floor heights and a regular open structural grid. The buildings were ugly and hostile at street level, but remodelling and recladding could successfully address these shortcomings. Their structural capacity allowed the site’s full development potential to be realised.

At Marble Arch the situation is very different. Here, we have three separate structures, each with their individual compromises in layout and quality - problems which are amplified in aggregate: dense clashing column grids, low misaligned floors and a floorplate broken up by numerous cores.

At the inquiry, SAVE’s own proposals for refurbishment made for salutary reading. They suggested the removal of all the perimeter cores and the construction of a new central core. More than a quarter of the floorplate was to be demolished and rebuilt.

This might sound better than a new building (aren’t you keeping at least three quarters of the fabric?) until you realise that the patch and repair of existing structures is extremely carbon intensive. From our experience at Kensington, where we only had to demolish 11% of the original structure, the levels of embodied carbon for the overall building were comparable to those anticipated for the Marble Arch new build.

Moreover, the capacity of the existing buildings at Marble Arch means that their refurbishment can only deliver three quarters of the space provided by a new build. This matters because if we are to move to zero carbon, we must make full use of the sites with the best public transport capacity – the alternative is urban sprawl, increased pressure on the greenbelt and more private car usage.

The operational carbon delivered by the new build will also be significantly better than a refurbishment. The building, as assessed by Arup, will be in the top 1% of best performing offices in the UK and is designed to achieve BREEAM outstanding, WELL platinum and Wired Platinum certificates.

Very often, refurbishment will be the right choice. In other cases, as here, it will not. Refurbishment would retain buildings that perform very poorly, underdeliver on the site’s potential and compromise the quality of the public realm.

M&S has been very clear that without the ability to make the transformational investment that planning policies demand for this area, it will have no option but to leave its unsuitable premises, hastening the decline of this end of Oxford Street.

>> Henrietta Billings: Why the M&S public inquiry matters