Restoration of the Roofs and Ceiling-Frames of the Oratory of the Partal Palace and of the House of Astasio de Bracamonte

They include the required conservation strategies adapted to the cultural context of the Alhambra.

For your interest

Restoration of the Ceiling Frame

Before beginning the restoration work, we carried out an exhaustive historical study aimed at encapsulating in a single document all the information about the Oratory of the Partal that might be of interest in subsequent restoration work.
This document seeks to identify the different historical strata in the building as it is today, some of which were added in recent restorations of the original building, in order to help select the most appropriate methods for drafting the restoration project and later for doing the work, in particular as regards its division into different phases. In this way this document will help to define the essential criteria on which to base our work in the Project, the points from which to start. The Restoration Project will also take into account the results of various other preliminary tasks including the topographic and photogrammetric surveys of the building, which will help in later analyses of the pathologies and in the decisions as to the best way to treat them.

The first stage of the restoration work involved the renovation of the roofs of the Oratory of the Partal and of the House of Astasio de Bracamante, essentially a construction job.

The second phase involved the restoration of the ceiling frame and of the panelled ceiling, and was divided into three stages:

  •  Structural restoration
  •  Building restoration
  •  Restoration of the paintwork

 

1. Historical background

 
The Oratory is an excellent example of a private chapel and is directly associated with the Partal Palace, of which it stands a few metres to the east. It is situated on the edge of the walled perimeter of the Alhambra and is thought to have been commissioned during the reign of the Sultan Yusuf I (1333-1354), as was the house of Astasio de Bracamonte, to which it is attached on its south-east wall. Its decorative features closely resemble those in the Tower of the Captive and it is thought to have been built at about the same time. The analysis of the party walls between the Oratory and the House of Astasio de Bracamonte suggests that the Oratory was built some time after the House. It has a rectangular floor-plan measuring 4.16m x 3m and houses a small mihrab in its south-east wall, covered by a small mocarabe (honeycombed) vault. The entrance to the oratory is slightly raised with access via a set of stairs in the North West wall. In the area adjacent to the entrance there is an initial, very narrow corridor covered by a horizontal ceiling made of pressed-brick beams, which gives onto the main hall of the oratory through a semi-circular arch. This is the space covered by the lattice-work ceiling frame.

 

2. History of the various restorations

 
The Oratory and the House of Astasio de Bracamonte, Squire to the Count of Tendilla who was granted the house by the King in 1550, were in private hands until the 19th century and in quite a good state of conservation. In 1846 the owner, Mr Francisco Acebal y Arratia restored the building in line with the eclectic fashions of the time completing the decoration as the fancy took him with assorted adornments removed from other places.
In the Alhambra we have extensive documentation on two of the restoration processes performed on the Oratory. The first was the one referred to above by Rafael Contreras in 1846, which includes a plan with a sketch of the polychrome uncovered during the work. The second restoration in 1930 was by the architect Torres Balbás, who described how they cleaned the ceiling by removing a layer of “paint that had been daubed upon it”.


Torres Balbás kept a Works Diary in which he described the work done each day. In addition to the restoration of the decoration on the walls, they also replaced the four-sided roof. The poorly applied coloured paint which he said was staining the ceiling frame and the frieze was cleaned off, returning them to their natural colour. This was the only job that Torres Balbás did on the ceiling frame given that it was in a good state of conservation.

 

3. State of conservation of the ceiling frame

 
Once the roof of the Oratory had been removed we were able to inspect the state of conservation of the ceiling frame from the reverse side of the structure and make an initial appraisal of the causes of the various types of damage. This would later help us resolve these problems and decide exactly what restoration work would be required.

The large deformation visible inside the Oratory on the front side, seems to have been caused above all by the remains of the old roof, which rested on the almizate (central horizontal panel) of the frame and in particular on the crowns of the 4 central rafters (2 rafters from the North side of the roof and 2 from the South). 
This structure was not the only weight bearing down on the ceiling frame, there was also a layer of plaster, which on the almizate was supporting some ceramic bricks, and the weight of the roof of the building itself, made with a double layer of solid brick and tiles. Together they proved an excessive extra load for the ceiling frame, which was supported on the central rafters which in turn rested on the remains of the ridge beam from the old roof.
The layer of plaster applied by Torres Balbás in order to shore up the deformation (already a problem in 1930 when he conducted his restoration work) was applied on top of a previous layer and contains some bolts nailed into the rafters of the frame, on which this frame “hangs” the roof rafters as if they were tie-beams.
The most serious problems we encountered in the frame were as follows:

  • Deformations of the wooden structure due to the fault in the support beam together with the extra weight borne by the almizate, which led to a clearly visible warping of the almizate and the deformation and loosening of the North side of the roof.
  • Generalized surface dirt caused by the accumulation of dust, insect nests and bird excretions located above all on the horizontal surfaces of the almarbate (square wooden panel).
  • Dirt and oxidation of the “protective” layers applied all over the decorated surface of the ceiling frame leading to a general obscuring of the surface.
  • Peeling of coloured paint which has lost its mechanical resistance due to the “protective” layers applied on top of it.
  • Attack by wood-eating insects
  • Repainting
  • Pieces used to rebuild support structures added during previous restoration work.
  • Deformation and warping of planks.

 

4. Description of the project

 
The first step was to remove the tiled roof so as to enable us to access the extrados of the inner latticework ceiling frame. This would allow us to restore the frame and the horizontal ceiling of the entrance to the Oratory. Once this had been restored we would then rebuild the outer roof and waterproof the sides, before finally retiling it.

 
5. Cleaning systems

 
Due to the fact that the ceiling frame is suffering from a generalized accumulation of dust on the surface and numerous insect nests, we decided to clean it mechanically with a smooth brush and a vacuum cleaner to remove the dirt.
After this dirt had been removed, we then cleaned off the successive coats of varnish, wax and coloured paint applied over the building's long history, which have obscured the original decoration and made its conservation more difficult.
We chemically cleaned the frieze on the South, East and West sides using paint stripper. This caused the repainted layer of brown paint to swell up allowing us to remove it and then clean the frieze with 3AT (mixture of benzyl alcohol, acetone, water and a small percentage of triethanolamine).
After good results were achieved in tests with a laser, the north side of the frieze (replacement of plaster work) was cleaned using this method on the high part of the relief, without touching the painted sunken background, which was cleaned with gentle chemical cleaning.
The roof-sides and the almizate (central panel), which are not coated with brown paint,  are being cleaned with a combination of mechanical and chemical cleaning using cotton swabs dipped in 3AT, which causes the layers of varnish and brown paint to swell up and soften. These layers are then removed with 3AT-impregnated cotton buds before a final rinse-off of the area with ethanol. The areas with no polychrome are covered with coats of wax. This is cleaned mechanically using a scalpel and a laser and by rubbing down with wire wool.

Little remains of the original polychrome and surviving traces are often damaged by repainting and other treatments applied later.  
The polychrome was probably a tempera-based paint applied on glue-coated wood. We cannot be absolutely certain however because it is deeply impregnated with waxes and resins. The analysis shows remains of both oil and tempera painting techniques.
This would suggest that the original paintwork in tempera was painted over with oils, probably pre-19th Century.
The pigments we detected are commonly used in Islamic and Mudejar ceilings, namely plaster, vermillion, white lead, orpiment, lead oxide, realgar, earths and carbon black.
There are also traces of further repainting in the 19th Century or later (dated by the fact that a lead chromate based pigment was discovered). This polychrome was an oil paint with numerous intermediate varnishes of a resinous nature and gilding with pure tin sheet (now degraded to beta-tin, oxides and chlorides), applied on resinous varnish.
On the surface we found large quantities of beeswax and traces of other underlying varnishes, above all shellac. The beeswax has recrystallized and ‘bloomed’, as well as being very dirty with significant amounts of retained dust.
We are completing our initial analysis with other samples taken more recently in order to reach more definitive conclusions.

 

6. Treatments on the inside

 
Fixing and consolidation of the polychromes
Chemical cleaning tests.
Our aim was to find an effective method that respected the original work and that was capable of eliminating the superimposed layers (above all rusted shellac, varnish and waxes) that obscured the original polychromes, so halting any deterioration of the polychrome that these layers may be causing.
We began by testing mixtures in a range from the lowest to the highest polarity so as to find out the minimum polarity necessary to dissolve the material, so minimizing any unfavourable interactions with the artwork.
We then did tests on more polar mixtures of solvents. In most cases the use of this kind of solvent is inevitable, although care must be taken to select the most suitable and in the way it is used, as the volatility and the performance of the solvents can vary greatly. 
Photonic cleaning tests.
Photonic cleaning with a laser is considered very useful when dealing with pieces that are coated with different kinds of materials (brown paint, varnishes and resins) but have no underlying polychromes.
The results of these tests have proved satisfactory in two specific situations:

  • For removing the brown paint on a piece of the frieze on the northern side. This piece is a replacement plaster panel installed in the 19th Century that has no original polychrome.
  • For removing the layers of resin and varnish on the wooden base with no original polychrome.

 

7. Treatments on the outside

 
The first step was to take down the outer roof, which was supported on a wooden superstructure of rafters resting on a central ridge beam. 
Once the roof had been taken down, we began the cleaning work by eliminating the surface dirt with soft brushing and vacuuming.
Once this dirt had been removed, we could make out two quite distinct layers of gypsum plaster on the reverse side. The outer layer was applied during the restoration work by Torres Balbás, which sought merely to prevent further deformation. In some places this layer is on top of another earlier layer of which less survives today.
We removed the most recent layer of plaster applied by Torres Balbás in order to be able to protect and then pull off the older layer. This will be conserved and then put back in its place once work on the reverse side has been completed.
The remains of the old roof have also been dismantled. This was necessary to enable us to push up the warp in the almizate back to its original flat plane. Once this has been done, there will be insufficient space for this old roof, which will therefore not be put back.