Halley VI Antarctic Research Station by Hugh Broughton Architects, Brunt Ice Shelf, Antarctica

In designing the new Halley Research Station for environmental science organisation the British Antarctic Survey (BAS), getting the cladding right might seem like the least of a daunting set of problems. Conditions at the BAS base on the Brunt Ice Shelf in east Antarctica are extreme.

Temperatures regularly plummet to below -50˚C, winds can reach over 100mph and nearly a third of the year is spent in total darkness. Each year the ice shelf moves 700m and over a metre of snowfall accumulates on its surface, gradually entombing and crushing built structures.

In fact, the cladding has proved to be one of the greatest challenges. Architect Hugh Broughton has pioneered the use of glass-reinforced plastic (GRP) on the building, which will open formally in January 2013. Although the design of the panels was carefully trialled, when the first ones went to Antarctica they failed, with small cracks opening up in the surfaces.

The designers and manufacturers have now solved those problems and all the units are clad. Fortunately the BAS proved understanding. ‘If we had had these kinds of problems on a building in Cambridge [the BAS headquarters] they might have asked questions,’ says Broughton. ‘But this is the first time that this kind of cladding system has been put together, in response to very challenging conditions.’

This is the sixth Halley Research Station. The first four were buried and crushed by snow buildup. Halley V, which is still in use, overcomes this problem by being jacked up on stilts. This allows it to be raised every year, though it requires the collective effort of 40 people over several days. And it does not solve the problem of the relentless movement of the Brunt Ice Shelf towards the Weddell Sea, so that the area on which the station sits will eventually break off.

To overcome these problems, in 2004 BAS launched a competition (the first of its kind for a Halley base) won by Broughton and engineer AECOM (then Faber Maunsell). Their design consists of a series of linked modules on stilts that can be raised with far less effort than Halley V, allowing more of the team brought to the Antarctic to carry out scientific rather than maintenance work.

The modules are on skis and can easily be pulled to a new location. Broughton’s design also improves living conditions - especially important for the overwintering crew, who are there for eight months, including 105 days of total darkness.

Modules are brightly coloured - strong blue for the science and sleeping quarters, and a vivid red for the largest, central module, which forms the base’s social hub. When he won the competition, Broughton originally intended to clad the modules with structural insulated panels (SIPs) similar to those used on the new US station at the South Pole.

But a manufacturer suggested the possibility of using GRP so, once Broughton had appointed Billings Design Associates (BDA) as cladding consultant, the design team considered the pros and cons of SIPs, GRP and cold-store panels.

SIPs are relatively small (limited by the size of plywood sheets) and require a lot of handling on site. With a short summer season of only three months and each construction worker costly to maintain, this was a real concern. Jointing could also be difficult, and there was a further worry that high winds could suck the timber face off the insulation layer below.

On Halley V this problem was overcome by using a timber batten to join the front and back faces. But in the extreme cold, even timber will act as a cold bridge. Cold store panels are simple to construct and are obviously adapted to low temperatures. But they degrade rapidly, especially when exposed to ultraviolet light.

Design life would only have been five to 10 years. The advantages of GRP were obvious. It forms large panels and is light, making it easy to handle and install. It is used in cryogenic applications, so can evidently withstand low temperatures. But this project pushed GRP technology - more commonly used in aircraft or train construction - to its limits.

The contract was awarded to South African company MMS Technologies, partly because it was one of the few manufacturers capable of creating both steel frame and GRP cladding as a complete package, and partly because of the technology it used to make the GRP.

There are two components in GRP: a mat of fine glass fibres and the resin that infuses them. The mat is threaded through the insulation, in this case in the form of trapezoidal blocks. Unusually, MMS used a vacuum method of infusing the resin.

‘They had massive truck bodies that they were making in one piece,’ says Sean Billings of BDA. ‘They put them in a big plastic bag and poured the resin in through little tubes and just sucked it through.’ This approach allowed the design team to develop large panels and create a semi-monocoque structure, with panels fixed to rubber mountings.

Early test castings were encouraging, but there were problems caused by the demanding requirements for fire resistance. To meet these, the design team added a ‘filler’, aluminium trihydrate, to the resin. This has the effect of giving off water vapour in a fire, and so improves performance. But it also makes the resin more viscous and so more difficult to infuse under vacuum suction.

With the largest panels measuring 10.4 x 3.3m, this was a problem. It slowed down production, and meant that instead of having all the panels for the blue modules ready for shipping for the summer season of 2007/2008, the panels for only one module were ready.

This proved a blessing in disguise because once the panels reached the Antarctic, small surface cracks rapidly started to appear. Many of these were around the complex moulding in the joint between panels, but some were also in the panel centres.

The contractor continued with the erection, but also appointed David Kendall, a structural engineer specialising in composite materials, to investigate the problem and help devise a solution. Kendall’s investigation showed that none of the cracks were structural and that in fact, the structural performance of even the worst affected panels was very good.

The source of the problem was ‘resin-rich’ areas where the resin had pooled, without adequate fibre, because of the difficulty that the filled resin had in passing through the moulding. Working with the architects and structural engineer, he came up with a solution that allowed many of the panels, which had already been fabricated, to be remediated rather than having to be entirely remade.

The joints between the panels were redesigned to be much less sharp, and the original gaskets replaced with aluminium cover plates. The joints were ground down to create the new shape, and a similar joint was designed for the red panels, which were yet to be fabricated.

The outer faces were also ‘re-skinned’ using resin without filler, since only the interior face of the panels are vulnerable to fire. The new panels underwent extensive fire and thermal testing and all were ready for shipping for the 2009/2010 season.

Erection of the panels went without a hitch - indeed it took less time than anticipated, which was fortunate, as the ship carrying the construction crew was delayed by 10 days, eating into the nine-week building timetable.

Broughton is delighted with the result. The new joints are crisper than the original gaskets, which had proved to run into difficulties where they had to turn corners. Valuable knowledge has been accumulated on using materials in extreme conditions. Broughton is now so confident of GRP’s properties that a new station he has designed for the Spanish in the Antarctic will be entirely of GRP construction, with no supporting steel frame.