A paper presented at the workshop of the 'Contribution of Science and Technology in the Conservation of Cultural Heritage in the Mediterranean Basin', Tunis, June 1997


My research at the Institute of Archaeology, University College London, addresses the issue of conserving and presenting archaeological sites under protective structures when these are considered to be the best in situ measures for preventive conservation. Its aim is to develop methodologies and approaches related to this issue. The research further aims at establishing an information network to share experiences related to this issue and to make available all adverse information relating to intervention projects, particularly in the Mediterranean countries. A similar initiative has recently been undertaken by the Instituto Centrale del Restauro in Rome, Italy. However, its aim is confined to collecting and examining information on shelters built on Italian archaeological sites.

The Problem

The deterioration of excavated sites would be inevitable if measures for appropriate protection were not taken in exposed environmental conditions. Archaeological remains that have reached a state of equilibrium in the buried environment become vulnerable to weathering factors when they are uncovered. Thus, various approaches and solutions have emerged to protect excavated archaeological sites. Among these solutions, which range between reburial and reconstruction, is the erection of permanent shelters or enclosures on excavated areas. Sheltering or enclosing a site within a new structure is, also, often considered as an advantageous measure in presenting the site and making it accessible to the public. However, the range of problems in conserving sites under protective structures varies widely.

Developing a methodology for shelter design has received little attention. Indeed, there is a need for a methodology which primarily specifies protection measures and requirements for the conservation of the various types of in situ material remains, their setting, and the values held therein. In fact, the design of a new structure on an archaeological site should be thought of as a research process that aims at responding to the physical material conservation needs as well as to the presentation requirements for visitors to the site. Therefore, design should consider the enhancement of the physical and environmental context, use and visitor understanding and, above all, improved and stable environmental conditions for the archaeological materials and structures themselves. Collectively, satisfaction of these conditions would ensure the prevention or minimising of the decay of sites so as to prolong their life and present them accessibly to the public. However, in practice this policy has not proved to be an easy task to achieve.

The following selected examples will illustrate the growing need for a more cautious approach that responds to all the requirements of site sheltering. In fact, learning from previous experiences in this area is imperative in addressing issues that have often been overlooked in the specification of the requirements and design of individual protective buildings on archaeological sites. These protective buildings may be open-air shelters or totally closed buildings (enclosures).


Shelters are preventive conservation measures of immediate effect. They are usually concerned with keeping water away from the site materials, primarily in an attempt to prevent moisture from causing physical damage problems.

Simple shelters are usually metal structures, timber constructions, or tensile structures. Corrugated steel/fibreglass sheets or ceramic tiles are common roofing materials for these structures. Such shelters are usually built simply to protect the site from rain and sunlight. Many of them were built for temporary purposes but have turned out to be permanent buildings on site. In addition, maintenance is required for these types of shelters because of their vulnerability to natural damage. Damage caused due to lack of maintenance can be illustrated by examples of shelters in Crete: Niro Khani (Figure 1) and the recently dismantled shelter at Villa Dyouisos in Knossos (Figure 2). The latter, which caused damage to the gypsum and mosaics underneath it, due to water leakage in the roof, is to be replaced by a newly designed shelter.

site photo

Figure 1: A shelter at the site of Niro Khani, Crete.

site photo

Figure 2: The site of Villa Dyouisous in Knossos.

A newly built shelter on the site of the Byzantine church in Petra, in Jordan, is composed of an advantageous light-weight space-truss construction. However, this new shelter can be recognised in the cultural landscape from a great distance and was therefore criticised for its intrusive impact on the surrounding environment (Figure 3).

site photo

Figure 3: The shelter at the Byzantine church in Petra, Jordan.

At the 94 million years old palaeonotological site of Lark Quarry in Australia, a flat pentagonal steel roof with central skylight was erected in 1979 with the aim, again, of keeping the site dry. The shelter did not prevent wind-driven rain from flooding the site, and, being open on the side, it did not prevent dust from exacerbating the cracking of fossil remains. Additionally, the shelter did not prevent thermal shocks during summer rain storms. Further damage to the fossil bed resulted from the 12-support columns set in a concrete foundation. Also, prior to the later fencing of the site, animals such as kangaroos affected the surface by scratching and leaving droppings and corrosive urine. In addition, prior to the erection of a walkway under the roof, visitors had walked on the surface itself. In every aspect of the design of the shelter, as in many shelters of this type, there were deficiencies which could have been anticipated in the design phase. Solutions to problems which should have been foreseen were formulated, too late, at a later stage after damage had occurred.

Another example of a simple shelter in the Middle East is at the presargonic palace site in Mari, Syria (Figure 4), where a covering of modular elements of plastic on metal elements was erected with the base mounted on parts of the original structure using concrete blocks. Problems there included the rapid drying of the interior walls of the palace which in many parts turned into sand; drainage; and the passage of children and animals on the roof. Therefore, very regular maintenance was much required.

site photo

Figure 4: The site of Mari, Syria.

It could be said that most of the shelters described above were, conventionally, erected from a felt need to 'provide a roof' over a site. There has been a more recent attempt in shelter design which has adopted a more rational, reasoned approach to design. The Getty Conservation Institute aimed to develop a lightweight modular shelter which was easy to erect and could be temporary if desired - the 'Hexa shelter.' The Hexa shelter was developed in Paphos, Cyprus, and has also been further tested at Fort Selden in New Mexico (Figure 5). The framework is of aluminium tubes and the fabric of tri-laminated PVC with open-knit aero-textile panels for the sides. The evaluation of its effectiveness was undertaken in 1991 on the Fort Selden Site. There, protective indices were determined by structuring meteorological parameters and temperature variability on two identical adobe test walls inside and outside the shelter. Although the design of the shelter proved to be aesthetically low profile and neutral in colour, it did not provide complete environmental protection. Despite the fact that footings on the surface do not require subgrade excavation and can be cast in situ, structural movement at the Fort Selden site occurred when the soil underneath the footings was wet and slippery. The Hexa shelter could also be affected by lifting with high wind which necessitated additional reinforcing bars in the structure. Moreover, this type of shelter cannot stand vertical loading resulting from snow in cold climates.

site photo

Figure 5: The Hexa shelter in Paphos, Cyprus.

In summary, in many of these open-air shelters, it is likely that their benefits are more psychological than actual.


Enclosures, in contrast to shelters, have taken several forms. Among temporary structures on ongoing excavations are pneumatic forms which, in spite of their relative complexity, prove efficient where short term protection is required. A pre-formed continuous membrane of plastic sealed around its base and with air locked entrances can be kept inflated by small gasoline or electrically powered fans with balanced ventilation (Weaver 1973). Simpler methods of temporary enclosures made of polythene sheets and wooden frameworks can be achieved in such cases too.

Several Madaba Churches in Jordan were enclosed in order to present mosaic remains to the public. Protective enclosures were built in such a way as to give the impression of traditional church architecture (Figure 6). The building over the Apostles church site, which cannot now be distinguished from the original remains, consists of a concrete structure with traditional stone cladding. It is easy to comment critically on this practice since it does not conform with the concept of distinguishibility set by known conservation charters and ethics.

site photo

Figure 6: The shelter (under construction) over the Madaba church in Jordan.

At the temple of Apollo Epikourious at Bass, Greece (Theolakis 1993), thermal fluctuations affected the argillios veins of the limestone causing cracking and damage. Therefore, in 1987, a canopy with anti-seismic scaffolding to protect the site from weather conditions and earthquakes was erected and the measurement and assessment of environmental environmental factors proceeded to control the microclimate. Recording of the temperature and relative humidity data were taken. The site has been monitored since then, and improvements to the conservation conditions have been achieved.

Preventive measures in enclosed structures are usually concerned with controlling the relative humidity and temperature around the archaeological materials. Unlike artefacts within museums or in the interior of existing buildings, controlling the environment of exposed archaeological remains is complex. One may note the proposal of Eurocare Eurobuild for a protective building to cover the Hammar Cathedral in Norway (Apeland 1993). At Hammar, among other climatological factors, the major problem was caused by frost damage to the historic stone material. The aim, there, was to develop a method for simplifying the design methodology of protective enclosures. Projects of this type are important to the development of preventive conservation measures in this area, but their study results are often not disseminated. Another project with similar scope was the erection of a protective air-conditioned shell over the column of Marcus Aurilius in Rome (Bruno 1987) due to the polluted environment. There, the control of relative humidity and temperature was also essential. In these examples, glass panes technology was an essential element of the research programme. The research further considered cleaning equipment and technical solutions for the anti-reflective glass panels. Again, information on results of this study are not made available to practitioners.

The Fishbourne site near Chichester in England (Figure 7) is a classic example where important Roman mosaics were presented in an enclosed modern display that formed part of the restored subterranean foundations, lined out at ground level in the archaeological park. Elevated walkways with carefully located footings were designed for visitors, and the displays are highly effective. Problems developed later and included rising damp and biological growth due to ground water. The lowering of the ground water table by dry wells or other means was not desirable because of the deferential settlement potential of the foundations. This situation illustrates that even when sites are presented in a carefully controlled environment, they are not exempt from further deterioration problems that arise later. It further illustrates that constant monitoring is important.

site photo

Figure 7: Fishbourne site, England.

At the Peterborough rock art site in Ontario, Canada (Bahn et al. 1995), an enclosure built in 1984 has received several criticisms in recent years. Major issues included the uncertainties in the planning phase both about the function of the enclosure and in identifying potential causes of deterioration. Damage to the rock art surface has resulted from carbon dioxide from precipitation and human visitation. This has caused the dissolution of the calcitic surface. In the beginning, however, conservators thought that acidity of rain water alone was the cause of this deterioration. Other problems were related to the materials used in the new construction and the effects of condensation. This case is notable for it includes not only the technical aspects of presentation and environmental control but also ethical problems concerned with enclosing rock-art sites in a manner that relegates them to the status of museum objects and, so, might alter their cultural meaning indelibly.

Piazza Armerina in Sicily (Minissi 1961) is an example of a structure, the first of its kind, which has been currently reviewed. In the 1950s, a modern 'abstract' structure enclosed rooms of the mosaic villa around a courtyard. Elevated metal walkways were installed over the ancient walls. Translucent panels of plastic, attached to metal framing, were built, with louvres in wall areas for ventilation. In some areas, suspended panels of plastic created flat ceilings to reduce heat transformation and glare. These, however, proved to be insufficient to control solar radiation and heat. Discoloration of the panels and material failure of the exterior sheathing occurred due to sunlight. Additionally, summer visitors complained about high temperatures despite the ventilation design.

At the Montreal Museum of Archaeology and History in Quebec, Canada (Figure 8), remains of an excavated crypt area were left in-situ and form part of the display at Pointe-à-Calliere. A complete structure has been built on all sides to ensure its long-term preservation. This structure is itself enclosed in the museum building. Although temperature remained stable, there was a fluctuation in relative humidity which enhanced the formation of salts, which, upon crystallisation, were disintegrating the surface of the crypt. Relative humidity is well below that at which sodium chloride, the major existing salt, crystallises from solution, so a major problem was presented as a never-ending migration of salts. Options included stopping the water migrating to the surface and keeping the required relative humidity above 75%, which might be too high for visitor comfort and could also introduce biodeterioration. This last option, however, was implemented in a similar situation at the Archaeological Museum at Vergina in Greece in order to protect mural paintings (Dimacopoulos 1995).

site photo

Figure 8: The archaeological remains in the crypt of the Montreal Museum for History and Archaeology.

Another example of a pre-designed controlled environment is the Roman Bath Museum in Heerlen, Holland (Boekwijt 1985). Condensation and, mainly, a salt problem of 'harmless' calcium sulphate (gypsum) occurred later and was merely an ethical problem that had to be removed chemically using EDTA paste. Fortunately, there was an easy solution to this salt problem. Nevertheless, this case shows that environment control measures did not take conservation needs into account and that, as a result, condensation and salt formation could not be prevented at this site.

A more high-tech example of an enclosure is the intervention at the Cathedral of Atri in Italy (Scichilone 1985). It consists of four oversized showcases offering a strictly controlled environment on and around excavated remains that are permanently displayed to the public. Laminated glass panes use a thermostatically controlled system that can heat the glass electrically to prevent condensation. PVC pipes provide for ventilation through filters, so preventing contamination. A programmed timer in the panes was installed to operate anti condensation. Lighting of low voltage fluorescent lamps was provided for the visitor display. This case has proved to be successful so far.

And, as a final example, the Church of Our Lady in Bruges, Belgium (De Witte 1985). Here environmental parameters were controlled and mirrors to help visitors to have a complete view of the paintings were arranged with spot lighting of standard chromatic index and UV range. This is another successful example which considered responses to conservation and interpretation needs.

It can be seen that the design, performance, and maintenance of enclosures are very site specific. However, learning from previous experiences, such as those mentioned above, is very essential. The examples further illustrate the necessity, in the early design phase of these structures, to allocate a considerable part of the conservation budget to continuous maintenance and to necessary modifications.


Despite the fact that conservation is case-specific, these examples illustrate that for the benefit of future projects, it is important that the experiences gained be documented and assessed. They further illustrate the need for a thorough scientific methodology that will serve architectural approaches to the design of shelters and enclosures on archaeological heritage sites. Planning for a long-term commitment to mitigate the effects of time is essential as is the commitment at the very early planning stages to maintain and monitor the site and structure. Additionally, we have seen that the impact of protective structures on archaeological sites is not always beneficial. This, too, requires documentation.

This research programme further intends to develop guidelines and planning procedures which consider the unique environmental qualities of archaeological sites. It is hoped that it will help practitioners in assessing requirements and site management options in the different climatic regions. It will create an empirical base for future strategies in archaeological site management which can deal with protection as a preservation measure and a presentation issue at the same time.


The study is supported by the Conservation and Restoration Centre in Jordan, a centre which was under establishment in Petra by the Jordanian Government and the German Technical Co-operation (GTZ), and which has lately been patronised by HRH Crown Prince Hassan.

I am grateful to Dr. Clifford Price of the Institute of Archaeology for his continuous support and advice and to Dr. Paulette McManus for her help in discussing this paper and advising consistently on the research.

References and Bibliography

ACOR. 1997. "The shelter over the Petra Church." The American Centre of Oriental Research (ACOR) Newsletter Summer 1997. Amman: ACOR.

Agnew, N. et al. 1996. "Performance of a lightweight modular site shelter." Conservation and Management of Archaeological Sites 1(3), 139-150.

Apeland, K. 1993. "EU 446 Eurocare Eurobuild." In: Thiel, M.J. (ed.) Conservation of Stone and Other Materials. London: E&FN Spon, 748-757.

Bahn, P.G. et al. 1995. "The Peterborough Site: Reflections on massive intervention in rock art." Rock Art Research 12(1), 29-41.

Bahn, P.G., et al. 1996. "RAR debates." Rock Art Research 13(1).

Boekwijt, W.O. 1985. "Salt and crack problems in the Roman at Heerlen and their therapy." In: Preventive Measures during Site Excavation and Site Protection Ghent 6-8 XI 1985. Rome: ICCROM.

Bruno, A. 1987. "Protecting and preserving the Column of Marcus Aurelius." Museum 39(1), doi:10.1111/j.1468-0033.1987.tb00657.x.

Caperton, T.J. 1994. "An evaluation of geotextile shelters: Fort Selden, NM." US/ICOMOS Committee on Earthen Architecture Newsletter 10.

De Witte, H. 1985. "The Burges situation and the example of the Church of Our Lady." In: Preventive Measures during Site Excavation and Site Protection Ghent 6-8 XI 1985. Rome: ICCROM.

Dimacopoulos, J. 1995. A shelter in the style of a tumulus. Vergina: an underground archaeological site and museum in the type of a crypt. Athens: Archaeological Receipts Fund.

Jianzheng, C. 1981. "Xian: an archaeological site museum at Banpo." Museum 32(4).

Jerome, P. 1995. "Proposed permanent shelter for building 5 at the Bronze Age site of Palaikastro, Crete." Conservation and Management of Archaeological Sites 1, 35-42.

Lewis, H. 1981. "Experimentation in mudbrick conservation at Tepe Nush-I-Jan." In: Third International Symposium on Mudbrick (Adobe) Preservation, 24 September- 4 October 1980. Ankara: ICOM/ICOMOS.

Minissi, F. 1961. "Protection of the mosaic pavements of the Roman villa at Piazza Armerina (Sicily)." Museum 14.

Schmidt, H. 1988. Schutzbauten. Stuttgart: Theiss.

Theoulakis, P. 1993. "Microclimatic monitoring at the temple of Apollo Epikourios at Bass, Greece." In: Thiel, M.J. (ed.) Conservation of Stone and Other Materials. London: E&FN Spon.

Weaver, M. 1973. "The use of an inflatable 'air-dome' to produce controlled conditions for an archaeological site." Studies in Conservation 18.