Conserving Decorative Plaster Richard Ireland This article first appeared in The Building Conservation Directory 2005. From plain walls and ceilings, through elaborate architectural panelling and complex cornices to the most exquisite figurative modelling, plaster has adorned and decorated buildings for centuries. Work in this most basic of materials spans the full breadth of achievement, from coarse craft to highest art. Conservation can be required where decay or failure threatens its survival and often entails preservation, repair and frequently reinstatement within the context of existing historic fabric. To achieve an appropriate methodology and specification of materials it is vital that all variables are given due consideration. Though ideologically sound, employment of traditional materials and techniques on a like-for-like basis may have an unwanted impact. Plaster may itself be fragile, be dependent upon a fragile substrate, or even provide the ground for delicate decorative coatings and wall paintings, all of which may be archaeologically and historically important. This article is not intended as a prescriptive ‘How To’ manual, but rather as a means of raising awareness to the intricacies and complexities that can surround such an apparently narrow subject. Techniques should be tailored to each particular situation and adjusted accordingly. Too often, for a variety of reasons, techniques are implemented without due thought and consideration. Success depends on a thorough understanding of materials and techniques, both old and new, so as to minimise intervention, maximise preservation and, where applicable, reinstate aesthetic integrity. A thorough primer on lime and gypsum plasters and their use can be found in many articles, pamphlets and textbooks detailing the subject in the technical and conservation press. A basic guide is outlined below.
Plaster Plastering materials and methods remained largely unchanged over several centuries in Europe until the older methods were gradually ousted during the 19th century by faster setting alternatives including sand and cement based renders and fibrous plaster. These new technologies were coupled with the use of cast gypsum plaster enrichment which was ushered in during the late 18th century. Two distinctive material types are generically termed ‘plaster’: lime and gypsum. Broadly, till the latter part of the 18th century, it was lime plaster that predominated. This was typically produced by the calcination of limestone rock (calcium carbonate) at temperatures in excess of 1,000 C forming quick lime (calcium oxide). Slaking with water produced a non-hydraulic lime putty – that is, a form of lime (calcium hydroxide) which does not set on contact with water. The lime slowly reverts to a chemically identical material as its parent rock by slow absorption of carbon dioxide during carbonation. The addition of set-enhancing pozzolans such as crushed brick to non-hydraulic lime plasters reduces setting time at the expense of malleability, as does the use of naturally hydraulic lime. Ultimately artificial cements made pure limes all but redundant in new 1
building construction, and gypsum plaster emerged as the predominant finish for interior walls and ceilings. Lime’s unique properties, when intimately combined with aggregates like sand, include plasticity and controlled setting. These selfsame traits restrict maximum coat thickness to some 18mm and necessitate several days between applications to allow for shrinkage and development of adequate strength. The inclusion of differing grades of aggregate and of organic ingredients like cattle hair modify and adjust performance to suit the work in hand. The resulting mixture may be used to render walls and ceilings, run mouldings, press ornament and model in situ. Internal flatwork on walls and ceilings is traditionally comprised of three layers: a render or pricking coat, floating coat and setting coat applied to solid or lathed backgrounds. Coarse sharp sands reduce the effects of shrinkage whilst inclusion of cattle hair increases tensile strength. A physical key for each successive coat is formed by scratching the partially set surface of the preceding one, and the final coat is given a finer finish applied around 3mm thick. External rendering is usually carried out without hair. 19th century experimentation led to the common use of very dense hard finishes often over softer coarse textured base coats. Run mouldings are built up layer by layer in the same manner as flatwork, the shapes formed with pre-cut metal profiles incorporated on a timber frame run onto the flatwork. By the 20th century, casting of large flat and curved plain faced sections became increasingly economical and heralded today’s extensive use of the techniques from domestic interiors through to the largest shopping malls. Modelling of lime plaster is an additive process by which material is gradually built up from the surface by the craftsman. This individual hand working of each element enabled the deep undercut and layering which enriches so many buildings of the late 17th and 18th century and distinguishes it from the later mechanical repetition of cast plaster produced from moulds. Large projections such as limbs, foliage or instruments required the use of an armature till adequately carbonated. These can be ferrous or organic, such as wood and bone – indeed anything capable of providing suitable support for the carbonating plaster was used and often became a significant element in later deterioration. While lime is inherently weather resistant and could be used inside or outside, the same is not true of gypsum, otherwise known as plaster of Paris or casting plaster (calcium sulphate). Calcined at no more than 120 C to produce a white powder, it typically sets rapidly and rigidly within 15 minutes of mixing with water. This is a material for internal use only. It was often used from the latter half of the 18th century as an admixture combined with lime plaster to achieve an early setting and to counteract shrinkage. Mixed on its own with water to a creamy consistency, gypsum is particularly suited to pouring into low relief moulds. Exploitation of these attributes in the late 18th century, together with the rise in popularity of the neo-classical style, enabled large quantities of repetitive low relief ornament to be churned out in a fraction of the time taken to model lime in situ. This led to the rapid decline of the lime plaster modeller. By the mid 19th century, flexible gelatine moulding materials allowed a degree of undercut to be achieved in a single cast. However, cast enrichment cannot capture the organic vivacity of the hand modelled lime it superseded.
Substrates The plaster surface frequently forms the last built layer in a complex interrelationship of building materials from stone, timber and brick through a variety of plasters and renders. Internally a timber substructure behind ceilings and built out elements like arches is constructed to reflect the topography of the design and minimise excessive thickness and weight of features such as running mouldings, ribs and cornices. Externally, substrates are 2
often brick or stone. Well into the 20th century elements such as cornices could combine both solid running and planted cast elements, but eventually fibrous plaster lengths of decorative cornice became the norm. Each material has intrinsic qualities which make it best suited to a specific role in a building. However, no material is without some drawbacks to its use, and where these are not fully appreciated at the time of use, they may well cause problems later. Inadequate structural timber, poor loadings and hard inflexible plasters are typical problems associated with historic work. In all cases they will have an impact upon the means of conservation.
Inspection and investigation Whilst the splendid enriched and decorated architectural surfaces of house interiors are much admired, the conservator should also bear in mind the artifice frequently employed to represent one material by another. Papier mâché, paint and plaster frequently masquerade as materials as diverse as wood, metal and stone. Successful conservation (of plaster and of the structural skeleton behind it) depends on thorough investigation. Quick fixes are rarely satisfactory in the long term. Programme and cost constraints as well as availability of skills are very real factors which must be carefully appraised and accounted for in developing solutions. Ceiling backs and salvaged debris may provide evidence both of failure as well as previous repair campaigns which may only detectable from above. Changes in plaster colour, integrity of plaster key, changes of lath and evidence of intervention in supporting structures all suggest potential problems impinging on plaster stability. Mortar analysis breaks down the components of plaster and identifies the proportions of binder (lime and/or gypsum), the type and quality of aggregate and the presence of any organic components. Care is needed in interpreting results as often crushed limestone or old mortar can be included in the mix which can skew the apparent binder to aggregate ratios. The analysis can also shed light on the durability or failure of the existing plaster. Whilst not a dating tool in itself, mortar analysis can provide useful comparisons with both the plasters used in different areas and on suspected remodelled sections of buildings, as well as between different schemes created in the same geographical area in the same period. It is especially helpful where colour and texture need to be accurately matched for bare plaster or render surfaces. Where salts appear as a white efflorescence or otherwise causing a problem, analysis using a laboratory flame photometer will indicate the type of salts and help trace the cause which in turn is vital to treatment. Sulphate salts are a product of air borne pollution but are also present in cements and hydraulic lime and a component of gypsum from which they may be leeched to cause decay. Nitrates tend to be associated with decaying organic matter whilst sodium chloride can come from a variety of sources, including road salt and, in the case of a garden temple, the source has been traced to the urine of livestock. X-Ray and other remote non-interventionist techniques such as metal detectors and infra red thermography are also invaluable in the right circumstances for the detection of structures such as ferrous armatures that may be rusting and splitting plaster sculpture and modelling. Plasterwork is rarely signed by the modeller, which can make dating and provenance difficult. Stylistic and structural context may be the only guides in the absence of bills and other archival sources. Dendrochronology of related structural timbers and even nail chronology can be useful in assisting dating.
3
Decay The agents of decay wreak havoc behind the scenes. Ferrous armatures rust and expand, salts effloresce, insects and fungi take hold of the timber substructure. In the British Isles water is the single most destructive agent. Much damage is due to poor water management and neglect. Leaks, penetrating damp, rising damp, overgrowth of flora, frost, poor maintenance and erosion are just some of the ways it innocuously permeates buildings. While lime plaster itself is generally not directly affected by water, many of the materials it is applied over are more vulnerable. To make things more challenging, these materials are often buried within or beneath the plaster. Persistent exposure to water will cause gypsum, alone or combined in a lime plaster matrix, to gradually decay and soften until physical failure occurs. Some loss of strength is of little consequence, but for later 19th century plasters which may contain high proportions of gypsum, this can be catastrophic on ceilings. Heavy and repeated applications of decorative coatings to hide, freshen and ‘hold back’ signs of damp on plaster have the reverse effect. Moisture is withheld in the plaster and may be spread over a much wider area, driven by capillary action and steep moisture gradients which actively draw moisture to drier air. Typical methods employed include the use of impermeable oil based decorative coatings, thick lead carbonate paste primers, varnishes and even bitumen. Persistent problem areas may be exacerbated by the introduction of harder and ‘better’ modern plasters or cements applied in the belief that their durability will provide both repair and barrier. Sadly this generally leads to accelerated decay of original fabric at boundaries where moisture and salts become more concentrated.
First Aid Once a problem has been identified, whether catastrophic physical failure like a ceiling collapse or noting a spreading leak, the subsequent action has a large influence on the success of any conservation and repair. Ceilings are frequently propped in a mad panic with heavy timbers and large sheets of plywood. Whilst effectively restraining the remaining ceiling, it considerably compounds the difficulty of assessment, and may cause more breakage, particularly around perimeter areas that have been forced back against falling debris from the ceiling back. In addition, it can be difficult to remove the props without causing still further damage to fragile plaster. Provision of evenly spaced free standing props with manageable platforms of no more than a metre square, topped with blanket or foam held just shy of the surface enables easy access and avoids wholesale removal of supports prior to undertaking work. Furthermore, any fragments which subsequently fall off will be safely collected and their source pinpointed, aiding their reinstatement and indicating the areas still at greatest risk. Where decay has rendered plaster particularly friable or where armatures are splitting modelling, the plasterwork can be faced using acid free tissue applied with reversible conservation grade adhesives to form a temporary protective skin pending consolidation and repair. Where sections of plaster need to be removed, a rigid ‘case’ of applied plaster of Paris or fibreglass is required to provide sufficient support for the object once removed from its backing substrate. When using this technique, additional protection can be achieved by adding layers of canvas or hessian. Temporary props may be needed to allow for the added burden on a ceiling section and strengthening battens and handles incorporated in the protective case to support the combined weight of object and protective case and assist carrying. 4
Techniques Good craftsmen acquire and refine the knowledge and skills particular to their job. The conservator is no different. Wherever possible it benefits the conservator to develop as simple technique as possible for any given task, knowing that it is the breadth and depth of their knowledge and its application with skill and experience that distinguish their work, not the complexity of their solution. Frequently, over-complication in an effort to justify ‘specialist’ work can have the reverse effect and, worse, may be later replicated by operatives with less knowledge and skill, to the detriment of the object. Once a deteriorating object has been made physically stable, a strategy must be formulated for the next step. Though factors vary in importance for any object and situation, they will include historical value, contextual importance, causes of decay, physical integrity and state of repair, accessibility, practicality and need of conservation and repair, cost and programme. Other important factors will relate to impact of techniques on the object itself, as well as adjacent existing fabric. Sometimes all that is needed is environmental change or control to inhibit or arrest decay. A non-interventionist approach is generally preferable, though the need for conservation often arises as rescue rather than prevention. Many strategies are used to stabilise collapsing ceilings. Some are logical, some prescriptive, and others potentially disastrous. Attempts to straighten sagging ceilings is often asking for trouble. Unless painstakingly cleared, debris caught beneath laths and joists can only increase stress in those areas as attempts are made to lift or push areas back. Likewise, the levelling of plasterwork that has sagged over a long period will only move stress and strain onto other hitherto stable areas providing the potential for later disaster. The layered nature of the typical ceiling construction can result in intercoat separation, though generally collapse results from deflection of the timber substructure from which the plaster depends, coupled with failing key and laths. The method I have developed and favour for securing relatively coherent sections of existing or barely intact ceiling plaster provides a simple, dry, lightweight, flexible, cost effective and readily reversible means of restraint. Shallow inset washers are inserted into the ceiling face and tied off on perforated straps secured between the joists behind. Treatment can be localised and specific to major cracks and delaminating areas. The use of thin wire ties suspended from flexible band introduces a degree of suspension and flexibility absent from the use of more rigid methods using solid studding or thick wire. It also allows for movement of the ceiling/floors whilst limiting point loadings. Drawbacks are the need for intervention at the ceiling face and the need for full access from above. Nonetheless, using wall painting conservation techniques, even decorative paintwork can be removed from restraint points and replaced afterwards with minimum retouching required. Less desirable, due to inflexibility, is the use of inset washers fixed through the ceiling face into the joists behind with screws. However, this may be the only solution where access above is impractical. The use of resin embedded studding into the back of plaster is often used in conjunction with significant resin replacement of failed laths. At the opposite end of the spectrum, the often used prescribed method of pouring plaster of Paris across the ceiling back over wire reinforcement is a potentially disastrous and unsuitable means of repair. This introduces large quantities of wet material setting as rigid monolithic blocks. These in turn accentuate any deflection forces at boundary edges, destabilising otherwise coherent existing plaster. Adhesion to an already limited plaster key, combined with the difficulty of ensuring adequate bonding of material, further compromises this potentially destabilising treatment. It is only partially mitigated when incorporating wire 5
loops through the ceiling face. The incorporation of large quantities of water into a small area, often closed over as soon as complete, can only enhance the possibility of fungal attack and decay. Moreover, removal is virtually impossible without considerable damage. The use of modern resins and adhesives need to be carefully considered. All will gradually harden and lose flexibility as they oxidise and cross link over time. Loss of performance in common commercial products will be greatly exacerbated by temperature fluctuations, humidity and light compared to conservation grade equivalents. Nevertheless, selective use of resins can be extremely beneficial. For example, where patching moisture sensitive existing fabric, such as a decoratively painted surface, an impervious resin can be used to minimise suction and transfer of water into adjacent fabric from the new plaster. Some consolidation techniques employ copious amounts of resin to increase structural integrity, often saturating the object. Careful consideration should be given before using a method that may complicate long term conservation options. Removal of soluble materials from porous objects is rarely easy or practical, and becomes increasingly difficult with age. Grouting of unseen non-structural voids should be carefully tested to determine suitability and success of the method. Much time and effort can be wasted to little or no effect using a technique that cannot be readily measured or checked. Thought is needed when specifying internal crack repairs to ceilings and walls. Whilst likefor-like repairs are often ethically desirable, they may not always be appropriate. Repairs should never be harder than the adjacent existing fabric. Too often cracks are filled without considering the consequences of restricting the natural expansion and contraction of the building fabric – not least on a seasonal basis. Hard fills act as wedges and physically erode adjacent softer material or transfer and increase loads across to other areas which can be further destabilised. The use of lime-based plaster to ensure like-for-like compatibility can also be counter productive. The wider the crack, the more shrinkage may be anticipated together with poor lateral adhesion. Shrinkage cracks may be impossible to avoid without addition of gypsum. Either way, care and tending will be required. Narrow cracks may have to be opened and raked out to provide sufficient purchase for the lime and so cancel the benefits of any friction interference helping to retain structural integrity of plaster sections. The necessary wetting may also be undesirable on sensitive or easily stained surfaces. Some modern gypsum based fillers are particularly hard and are marketed on the strength of their water resistance and durability. These are unsuitable for the repair of soft historic plaster. Others are especially soft and thus eminently suitable. Easily weakened further with the addition of whiting, they should be matched or be weaker than the surrounding fabric. Benefits include good adhesion, rapid drying to a neutral surface, fine texture and soft matrix which makes gentle smooth and controlled sanding an easy task.
Conclusion Frequently the greatest challenge facing the conservator is the treatment of flawed technology of the period. Use of the selfsame materials would perpetuate the problem whilst modern alternatives may not be ethically appropriate. Each and every situation must be assessed on its own merits with a thorough awareness of the consequences of each and any action on that object. Seldom will any material or coating possess every desirable trait without drawbacks. It is the duty of conservation specialists, whatever their discipline, to ensure a responsible and sympathetic treatment of the fabric they are dealing with that encompasses as many favourable conditions as possible and minimises negative attributes. The challenge is 6
identifying the key requirements, minimising compromise, to adopt a sympathetic and effective solution. Strict adherence to the original materials and techniques does not always achieve the long term goals of conservation and preservation, especially where buildings are outside the specialists control. Solutions must be considered both in the conservation of the past for the present as well as their suitability and maintenance for an unknown future. RICHARD IRELAND undertakes conservation on buildings as diverse as castles and farm houses to the Entrance Hall decoration and Reading Room at the British Museum. He operates chiefly around the British Isles and Ireland as both consultant and practitioner and lectures widely.
The Cleaning and Decoration of Decorative Plaster A further article by Richard Ireland, which covers the cleaning of plasterwork and its decoration, will be published in Historic Churches, the annual supplement to The Building Conservation Directory. It will be available from Cathedral Communications at the end of 2005.
7