Elevated Station Design for the South Pole Redevelopment Project at Amundsen-Scott South Pole Station

Click a section to display its information.



Historically, the facilities of the Amundsen-Scott South Pole Station have been designed as below-surface structures.  Within several seasons of their initial construction on the surface, they drift over with snow and except for pedestrian and vehicular access points which are manually cleared, remain buried.  Despite the potential advantages of facilities designed to remain elevated above the snow surface, such construction has until now, been limited to smaller ancillary buildings.

To remain permanently above-surface, structures at the pole must be designed to overcome both the localized snow drifting issues that they cause, and the annual accumulation of snow at the site.  The current redevelopment of the Amundsen-Scott South Pole Station represents the most ambitious attempt by any nation to establish a significant above-surface facility in an environment such as the South Pole.

This paper will discuss the rationale for the development of permanent above-surface facilities at the South Pole Station, examine previous examples of such structures that have been, or are currently operational in Antarctica, and review the features of the new Amundsen-Scott Station that qualify it as being state-of-the-art in South Pole elevated station design.



The first International Geophysical Year, conducted in 1957-58, triggered the construction of numerous research stations in Antarctica.  By 1980, there were 34 year-round stations maintained by 12 different countries:  Argentina (8), USSR (6), United Kingdom (4), United States (4), Chile (3), Australia (3), Japan (2), South Africa (1), New Zealand (1), Poland (1), and France (1).  The majority of these stations are located along Antarctica’s coastline and constructed on grade using traditional cold regions techniques.  Some Stations, however, have been established on the permanent ice shelfs or further inland on the interior plateau, most notably the U.S. Amundsen-Scott Station, which is located at the geographic south pole.  For these non-coastal stations, Antarctica’s environmental conditions pose significant challenges, and traditional cold regions construction techniques are seldom adequate to cope with them.  The primary challenges for stations located on the ice shelfs or further inland is annual snow deposition.  With no frost cycle, snow accumulates year after year, ultimately burying structures built on the surface.  Strong prevailing winds often hasten the burial process by contributing drifting snow on and around structures.  To overcome this problem several countries have experimented with innovative above-surface “elevated” stations, initially constructed 3 to 5 meters above the snow surface, and often incorporating a means of periodically raising the buildings to keep ahead of the ever accumulating snow.

The redevelopment of the U.S. Amundsen-Scott Station, currently in the beginning phases of construction, will include the largest and most ambitious example of an elevated station in Antarctica.  Scheduled to be fully operational in 2005, with a winter-over population of 50, the new station will set the standard for many years to come.  The next several sections of this paper gives a brief history of elevated stations constructed in Antarctica to date, followed by a discussion of the Amundsen-Scott station’s design.


Old Casey Station

Although a number of Antarctic coastal stations have employed the idea of a pier-foundation to allow for the seasonal scouring and control of drifting snow, the first emergence of a truly “elevated” station was Australia’s Casey Station.  Located on the shore of the Bailey Peninsula in Vincennes Bay, Casey Station was constructed to replace Wilkes Station, approximately two kilometers away, and became operational in 1969.  Originally a U.S. station, Wilkes had been constructed in a topographic hollow.  Over the years, snow that didn’t melt built up around the buildings and eventually buried them.  Accordingly, one of the objectives of the Casey Station design team was to permanently control the problem of drifting snow, and therefore, not repeat the experience at nearby Wilkes.

Figure 1 - Old Casey StationThe design of the new 20 person station by the Australian Antarctic Division staff involved a long row of thirteen inexpensive modular buildings elevated 3 meters above the surface on scaffold piping.  As a result of wind tunnel testing during the design, the row of buildings was oriented at right angles to the prevailing wind and connected together by a single walkway covered by semi-circular corrugated galvanized steel siding on the windward face.  The visual effect was that of an elevated tunnel, rounded on the windward side and squared off on the leeward side.  Winds were channeled below the elevated structure and effectively scoured the snow away, eliminating drifting problems as hoped.

Now referred to as “Old Casey Station”, it was replaced in 1989, after 20 years of effective service and scouring, by the current New Casey Station.  Interestingly, New Casey Station is constructed on-grade, in the traditional cold regions manner.  Snow drifting at the new station has been controlled by careful placement of the buildings in non-drift high areas and periodic grooming.  The unique design of Old Casey was not repeated, but not because it hadn’t performed as planned.  It had actually performed so well that it was never necessary for personnel to expose themselves to the elements.  This luxury was later perceived as a possible cause of lower productivity.


Filchner Station

As Old Casey was approaching the last several years of its occupancy, Germany was implementing a new approach to overcoming snow drifting and deposition on the opposite side of the continent.  In 1982, they completed construction of a small above surface summer station for up to twelve personnel on the Filchner-Ronne Ice Shelf in the southern portion of the Weddell Sea.  The ice shelf site of the Filchner Station was characterized by an annual snow accumulation of approximately half a meter, strong winter winds, and seaward drift of the ice shelf on the order of 1000 meters per year.  The German designers’ solution to overcoming the annual snow deposition was to place their modular accommodations atop a jackable structural platform on steel columns.  The platform was initially elevated 3 to 4 meters above the snow surface.  Every 2 to 3 years, the platform was lowered to the surface using a system of winches and cables, the columns were extended by approximately 1 meter, and then the platform was rehoisted to its new height.  The entire process took 3 to 4 days, and continued to work effectively until February of 1999 when a several thousand square meter portion of the ice shelf calved and took Filchner with it (unmanned at the time).  The station has since been removed from the iceberg and is in storage.


Halley V

The success of the Filchner Station’s jackable platform concept did not go unnoticed.  In 1982, the same year that Filchner became operational, the British Antarctic Survey (BAS) had constructed and opened the third replacement of Halley Station (Halley IV) about 1000 kilometers from Filchner, on the Brunt Ice Shelf off of Coats Land.  The site conditions on the Brunt Ice Shelf were severe.  Annual snow deposition was on the order of 1.5 meters, gale force winds were common 180 days out of the year, and the annual seaward movement of the ice shelf was approximately 850 meters.  From the station’s inception in 1957 with Halley I, structures were designed to withstand being buried and station life was essentially subterranean.  The life expectancy of each new replacement station was only 8 to 10 years.  Weary of living below the surface and hopeful of reducing the ever-increasing costs of rebuilding an entirely new station every decade, the BAS determined to change tactics and design Halley V as an elevated station based upon the jackable platform concept at Filchner.

Figure 3 - Halley VRetaining Christiani and Nielson of Hamburg for the design, Halley V was completed and became operational in 1992.  The 1,255 square meter station for 30 personnel was the most ambitious elevated facility in Antarctica at the time.  It consists of three separate buildings on jackable platforms, set 300 meters apart from one another at the three points of an equilateral triangle site plan.  Each platform is set initially 4 to 5 meters above the surface.  The working facilities on the platforms are created from an interconnected series of prefabricated building modules.  The largest facility (The Accommodation Building) is approximately 930 square meters and contains the living, working, and technical support spaces.  The smaller two buildings are roughly 140 and 185 square meters, and contain laboratories.  As a result of wind tunnel testing at the Cold Regions Research Engineering Laboratory (CRREL), the long axis of each building is oriented parallel to the prevailing winds in an effort to minimize any platform level drifting which would impact exterior pedestrian activity and access.  Unfortunately, this orientation has had the adverse effect of aggravating drifting below the platform at the leeward end of each complex.

Halley V contends with a much greater annual snow deposition and drifting problem than either Filchner or Old Casey.  As a result, jacking must be performed annually.  Similar to Filchner, the platforms are lowered to the surface, the columns are extended (in this case by 2 meters), and the platforms are then raised to their restored heights and the cycle begins again.  The entire jacking process takes approximately one week, and originally required one bottle jack and two operators at every column, working in unison (now the process is performed with electric jack motors).  During this period the station is kept fully operational.

The steel columns supporting the jackable platforms of Halley V are stabilized by tensioned cable “X” bracing.  Because of strain induced in the frame from eccentric snow creep and heave resulting from the leeward snow drifting, unanticipated problems have been encountered in maintaining column verticality.  Off-vertical column inclinations of up to 200mm have had be corrected, and guying systems are now routinely utilized to maintain the columns in a vertical position, particularly prior to the annual jacking process.

After seven years of expensive and problematic station maintenance, the BAS is seriously researching the possibility of using on-grade non-jackable sled-based relocatable buildings to serve as Halley V’s eventual replacement.

The experience at Halley illustrates that an elevated station design which can work effectively in one location (such as the case for Filchner) may be less successful in another, and it reinforces the idea that successful architectural and engineering solutions must be carefully tailored to address site specific conditions.


A New Vision for Amundsen-Scott Station

Less than a year before Halley V became operational, the National Science Foundation’s Office of Polar Programs was already creating a dramatic new vision for redeveloping the aging U.S. Amundsen-Scott station.

Similar to the British experience at Haley, the United States had continuously maintained and operated its Amundsen-Scott station at the geographic South Pole since 1957.  With a relatively milder climate (less wind and snow) to contend with than the British, the original buildings of wood and canvas initially sufficed, but the relentless annual snow deposition of .2 meters and prevailing winter winds eventually led to corrugated steel arches being placed over the living and working facilities, both for protection from the elements, and to contend with the inevitable burial beneath the rising snow plain.

The original station was abandoned after 18 years of operation and replaced with an entirely new station in 1975.  By that time, the old station had become buried by 8 meters of snow.  The 1975 station design perpetuated the concept of placing facility buildings within steel arches which could withstand being buried, and also placed facilities within a dramatic 51 meter diameter geodesic dome.  The landmark dome housed laboratories, offices, dining, and berthing space while the arches enclosed a garage/shop, fuel storage, the power plant, and other support areas.  The dome resisted drifting over better than the arches did, but over time, the unequal snow loading imposed on the dome’s base ring began to compromise its structural integrity.

As Amundsen-Scott approached the 1990’s, the combined pressures of outdated and fatigued facilities, a population which far exceeded original design capacity, and the general disadvantages of a mostly  below-surface existence led NSF to seek for a better and longer term solution.

In early 1991 NSF orchestrated a high-level design retreat in Enfield, New Hampshire.  That June they published the results of their charette in a document entitled the Enfield Concept Design.  Most notable in the visionary design concept was the idea that the pending redevelopment of the station would be substantially above grade in elevated, jackable structures.

Figure 4 - Enfield Concept Design

The concept rendering produced by the Enfield retreat depicts a series of identical separated two-story elevated structures connected by enclosed pedestrian bridges.  This line of elevated buildings was itself connected by an enclosed bridge back to the geodesic dome, which had been retained as the signature keystone of the station.

In 1992, Ferraro Choi And Associates a Honolulu based architectural firm with 10 years of antarctic experience, was retained to develop the concepts contained in the vision initiated by the Enfield retreat to a stage of final design and construction documents. Over the next 8 years studies, research, and design gradually led to current elevated, jackable design for the new Amundsen-Scott station buildings.  The design involves two separate elevated C-shaped two-story buildings connected by a two-story pedestrian link. The entire complex has been arranged in a linear configuration with the long axis perpendicular to the prevailing winter winds, and is elevated approximately 3 meters above the snow surface.  The windward face is chamfered like an airplane wing for improved aerodynamic performance.  Each of the C-shaped structures have been designed as “Pods”.  Each Pod is designed to account for differential settlement and to facilitate periodic minor leveling in addition to an ability to be raised a full floor height twice during the life of the station.  Day to day station operations (science, medical, administration, food service, berthing, communications) occur primarily on the second floor while support spaces (mechanical, electrical, emergency power generation) and gymnasium occur on the first floor.  The entire elevated complex (exclusive of the remote science facilities) is connected to a below surface main power plant, cargo facility, garage/shop, and fuel storage by means of a vertical circulation tower, which includes a staircase, cargo lift, and all plumbing and utility risers.  Above surface elevated facilities total approximately 6,040 square meters, easily the largest station of its type in Antarctica.  The landmark dome (shown on the site plan) is scheduled to be removed for cost and maintenance reasons.

Figure 5 - Amundsen-Scott Station Site Plan


Amundsen-Scott Station Design Features

If there was an advantage for the development of an elevated station at the south pole site of Amundsen-Scott, (compared to the Halley and Filchner sites), it is the comparatively milder conditions of winter winds and annual snow deposition, and the absence of significant horizontal movement of the ice sheet.  Whereas Halley’s Brunt Ice Shelf site had to overcome gale force winds and annual snow accumulation of 1.5 meters, the Amundsen-Scott site must only deal with 200 mm of annual snow deposition, and moderate winter winds (the strongest recorded wind at the South Pole was 87 kph and the average winter wind which causes drifting is on the order is 24-32 kph).  Compared to the annual horizontal movement of the Filchner site of approximately 850 meters, the site at Amundsen-Scott moves only 10 meters.  These comparatively milder conditions allowed the design team to approach the development of the elevated station from a different point of view.  Instead of being faced with the disruptions and expenses of raising the building every year, they could, through effective snowdrift control, develop schemes that would postpone the need for jacking for more than a decade.

Ultimately, the adopted objective at Amundsen-Scott was to limit the number of times the facility would need to be raised during its design life of 25 years to a maximum of twice.  With this premise in mind, a series of predictive studies at the Canadian based research facility of Rowan Williams Davies & Irwin (RWDI), including wind tunnel testing, Computational Fluid Dynamics, and Finite Area Element computer modeling techniques were initiated.  The studies indicated that the linear complex of C-shaped building forms elevated approximately 3 meters above the surface with the long axis oriented perpendicular to the prevailing winter winds and a tapered windward face would perform well.  The tapered leading edge smoothly channels the wind beneath the station complex. Forced to accelerate, the wind carries the snow well past the buildings where it is deposited in long leeward drifts.  A windward drift also forms just in front of the station.  Over time, the leeward and windward deposits will tend to connect around the ends of the station and gradually build into a rough crater shape.  Eventually, the windward drift will build to a point where it will prevent the wind from channeling beneath the station, and as a result, drifting will begin to fill in beneath the structures.  At that point, the station will be raised approximately 4 meters (one floor level), and the cycle will begin again.  The duration of the initial cycle has been effectively increased by NSF’s decision to initially erect the station on a compacted snow berm which is itself 2 meters higher than the surrounding grade.  Subsequent studies indicate that when all factors are considered, the station’s ability to control drifting could continue until the windward drift approaches the height of the station’s mid section.  This prediction indicates that it may not be necessary to raise the building for 15 years or longer.  The actual time period will ultimately be determined by the growth rate of the windward drift.

Figure 6: Snow Deposition

It was a recognized early on that no matter how well snow drifting could be controlled, at some point the station would need to be raised.  While snow drifting control issues were being studied and tested, therefore, an equal amount of attention was being given to the jacking and structural systems relationships.

Because of the large size of the station (6,040 square meters is roughly 5 times the floor area of Halley V), the duration of the jacking process and the size of the crew which would be needed to do the work were important concerns.  Maintaining full operations during the jacking process was also an objective, which implied the need to develop a way to maintain flexible connections with utilities and plumbing while the station was being lifted to its new height.

The final structural/jacking design involves an integral main building floor “platform”, supported by double trusses that straddle primary 914 mm diameter steel pipe columns located outboard of the building envelope.  The primary columns transfer building loads to welded steel box beams on timber raft footings.  Jacking involves adding a 4 meter column extension to the top of each column, placing hydraulic jacks under spreader beams at the top of each extended column, connecting the spreader beam to the trusses with steel rods, disengaging the trusses from the columns, hoisting the station up a full floor’s height (3m), and then securing the trusses again at the top of the column extension.

Figure 7: Integral Platform Concept Study with Outboard Building Columns

It was also determined that the most practical solution would be to lift one C-shaped Pod at a time, in an alternating series of 250 mm lifts until the full 3 meter lift was accomplished.  The process requires one 100-ton hydraulic jack and operator at each column during the process.  With 18 columns per C-shaped Pod, a minimum crew of 18 operators and 36 jacks is required (the crew will alternate between Pods for each lift).  The entire process is estimated to take approximately 30 days.



Four Elevated research stations have been constructed in Antarctica since 1969 to both improve the quality of life for station personnel, and to overcome the disruption and expenses to be rebuilt when they become buried and must be abandoned.

Australia’s “Old Casey” station functioned well, but was replaced in 1989 with ground based structures, ostensibly to improve personnel productivity.

Germany’s Filchner station was the first example of a jackable platform concept and was occupied for seventeen summer seasons until it was marooned on a calved section of the Filchner-Ronne ice Shelf, rescued, and placed into storage.

England’s Halley V year-round station was designed on the principals of Filchner’s jackable platform.  It was constructed on the Brunt Ice Shelf in 1992 and continues in operation today.  Unanticipated problems of eccentric snow drifting loads and maintaining the structural integrity of the columns supporting the platforms may lead to surface buildings on towable sled bases for Halley V’s eventual replacement.

The United States’ Amundsen-Scott station at the geographic South Pole is scheduled to be fully operational in 2005.  It will have an indefinite life span because of its ability to remain above the surface. The Amundsen-Scott station represents the state of the art in elevated, jackable station design for its size and environment.

Table 1 lists the comparative features of the elevated stations discussed in this report.

Figure 8 - Amundsen-Scott Station (View From The Geographic South Pole)



Table 1: Comparison Fact Sheet of Elevated Stations in Antarctica

Operational Life 1969-1989 1982-1999 1992-? 2005 – ?
Floor Area (Gross/SM) ±1,210 ±140 1,255 6,040
Population – Winter 20 0 30 50
Population – Summer 25 12 36 110-150
Annual Snow Deposition N/A 50mm 1.5m 200mm
Wind (Moderate/Severe) Severe Severe Severe Moderate
Annual Ice Movement N/A 1000m 850m 9m
Coast/Ice Shelf/Plateau Coast Ice Shelf Ice Shelf Plateau
Height Above Surface 3m 3m 4.5m 3m
Configuration shape Linear Linear Linear Linear/C-Shapes
Aerodynamic Yes No No Yes
Orientation to Wind 90° N/A Parallel 90°
Drifting Mitigation Good * Poor Good
Integral/Platform Integral Platform Platform Integral
C/FA Ration** N/A 1/14 1/42 1/167
Jacking Capability N/A Winch & Cable Hydraulic Jack Hydraulic Jack
Jacking Design Height N/A 1.5m 4m
Jacking Frequency N/A 2-3 years Annual 15+years
Wind Tunnel Yes * Yes/CRREL Yes/RWDI
CFD * * * Yes
FAE * * * Yes
* Information not available
** Column to Floor Area Ration. Floor area given in square meters.



1999, Australian Antarctic Station website, http://www.aad.gov.au/default.asp?casid=405

Blake, D.M.., The Development of Structures For The Brunt Ice Shelf In Antarctica, British Antarctic Survey, Cambridge, England

British Antarctic Survey 1992, ISBN 085655 1532, Halley Research Station

Dahm-Brey, Dr. C.1999, Filchner Station,  AWI website,  http://www.awi-bremerhaven.de/Polar/filchner.html

Ferraro Choi And Associates Ltd. 1996, Jacking Concepts, Prepared for National Science Foundation Office of Polar Programs

Incoll, P.G. 1990, The Influence of Architectural Theory On The Design of Australian Antarctic Stations, Australian Construction Services, Melbourne, Australia

Incoll, Phil G., Architecture in Extremity – Determining Influences On Antarctic Architecture, Australian Construction Services, Melbourne, Australia

Shanklin, J. 1998, The Antarctic Ozone Hole, British Antarctic Survey

Smith, A. 1991, Brief History of Antarctic Stations on Brunt Ice Shelf