operation and the core is well wetted, ready for the grouting process. The entire wall or walls are tested in this manner. Any large areas of masonry not showing signs of leaking during the process can be deep drilled through joints on a grid system of 0.5 m vertically, 1.0m horizontally, and staggered in order to connect with long voids which will act as both grout injection and proving holes during the works. All marked water escape points are fitted with plastic tubes set into the wall that are of a diameter to accept the grout injection nozzles


operation and the core is well wetted, ready for the grouting process. The entire wall or walls are tested in this manner. Any large areas of masonry not showing signs of leaking during the process can be deep drilled through joints on a grid system of 0.5 m vertically, 1.0m horizontally, and staggered in order to connect with long voids which will act as both grout injection and proving holes during the works. All marked water escape points are fitted with plastic tubes set into the wall that are of a diameter to accept the grout injection nozzles. Repointing of the wall or walls can now take place and in doing so the plastic tubes are fixed into the masonry. Grout is always introduced from the base of a wall, not from the top. This important practice is to avoid blocking through air locking and to prevent fine debris working down and closing up the voids. For this reason, it is sensible to plan repointing at open joints from the bottom upwards, so that the mortar can gain enough strength to contain the grout without leakage. In some circumstances the open joints can be plugged with foam backer rod or some alternative packing and repointed after grouting. Introduction of grout is ideally carried out using a simple diaphragm pump which works by pushing and pulling the operating lever (Figures4.48 – 4.49). A pressure gauge is fitted to the pump and this, during pumping, should hardly register, as introduction of grout into the wall needs to allow for the grout to spread slowly and horizontally along the wall, rather than creating a vertical head of grout locally at the injection point and then allowing it to spread along the voids within the wall. The lift heights for grouting will vary on buildings. Large granite ashlars with a fine joint system could be raised 1.5 –2.0 m in one day. Alternatively, small stones in a weak mortar should not be raised above 0.5 m without the risk of hydrostatic pressure moving faces away from the core. Decisions on height of grout per day can only be taken on site and must be based on extensive previous experience. If possible, each lift of grout introduced to a wall should be left for one day before resuming grouting. This is in order to give the grout time to dewater and stabilise before more grout is added to the wall. Grout continues in this manner until the head of the wall is reached. At this stage it is advisable to apply a head of grout to the upper proving holes, to compensate for the reduction in volume of grout due to de-watering within the voids. Quantities of grout taken up at each injection point should be recorded to provide an indication of the volumes filled. On completion of grouting and after a period of three to four days, the plastic sleeves can be removed and the injection points can be deep tamped and pointed.

 

Stone replacement

 

Within the philosophical context already described it is unusual to find any substantial replacements of new stone. All stones, in whatever condition and especially any that have to be replaced, must be recorded and evaluated. All types need to be petrographically identified and provenanced as closely as possible, and archaeological evidence such as lifting tenons, quarry-working marks, tooling, identification marks and mortices must be recorded. When different types of stone have been used in construction, identification may indicate important changes in the history of the building or, for instance, the use of salvage material from other sites.6 Where replacement is necessary, to support or protect historic material at risk, the rule of ‘like with like’ is generally thought to be appropriate. In many cases this can amount to no more than geological compatibility, because

original sources of stone are often no longer available. Matching colour, grain size and texture are thought to be important in the repair of any masonry buildings, but in the conservation of old, weathered masonry the overriding objective must be to so modify the situation in the wall that its life expectancy is significantly increased. Some general points on the selection of appropriate material may be made. The stone should be of the same type as the original, and of the same overall size and character, achieved through cutting and dressing techniques. Other characteristics are more problematical. Even when a replacement stone is near to identical with the original stone it will be different by reason of its unweathered state; its surface, at least, is going to be more resistant to weathering and decay agencies. Bedded and grouted into the old masonry, even in appropriate mortars, the new work is commonly much sounder than the old and may initially cause additional local stress to the fabric it is intended to assist. Admittedly, this is only likely to be the case when the old masonry is in very poor condition, but such situations are not unusualto be appropriate. In many cases this can amount to no more than geological compatibility, because original sources of stone are often no longer available. Matching colour, grain size and texture are thought to be important in the repair of any masonry buildings, but in the conservation of old, weathered masonry the overriding objective must be to so modify the situation in the wall that its life expectancy is significantly increased. Some general points on the selection of appropriate material may be made. The stone should be of the same type as the original, and of the same overall size and character, achieved through cutting and dressing techniques. Other characteristics are more problematical. Even when a replacement stone is near to identical with the original stone it will be different by reason of its unweathered state; its surface, at least, is going to be more resistant to weathering and decay agencies. Bedded and grouted into the old masonry, even in appropriate mortars, the new work is commonly much sounder than the old and may initially cause additional local stress to the fabric it is intended to assist. Admittedly, this is only likely to be the case when the old masonry is in very poor condition, but such situations are not unusual in ruin conservation. With these concerns in mind, alternatives to stone replacement are sometimes specified, and these are described below. The requirement for new stone is usually identified during the condition inspection and becomes incorporated in a schedule of repairs. Regardless of irregular arrises or damages, new stone should be scheduled giving large enough dimensions (length on face _ depth on bed _ height – L _ B _ H) to allow for dressing back to the final sizes required. In many countries new stone is now supplied four or six sides sawn as standard (Figure 4.51); such blocks would have been very expensive to produce in centuries past, when it is usual to find only the face and short-return dimensions worked square, with the back or tail of the stone irregular as quarried. Replacements should normally be 100 or 150mm on bed, and sometimes longer, depending on their functions, but the face dimensions (L and H) should match those of the lost stone as closely as possible and, most importantly, should be set to the original face line on the building (Figures 4.52 and 4.53). In weathered and decayed masonry, this means that the new stone will project, often forming a ledge on the top bed and creating a shadow line under its bottom bed. In spite of this, there should be no deviation from this rule or the evidence of the original face line and its profiles will be lost, and the authors of future replacements denied essential information. Projecting stones of this kind do not need conspicuous mortar frames, and especially not mortar fillets applied to the top bed as a ‘weathering’ (Figure 4.54). Mortar bedding and pointing should follow the weathered profile of the masonry, so that joints never increase in thickness over the original width and are not enlarged to fill out local damages. On the top bed, the stone may be given a very slight weathering and a pencil-rounded front arris to assist the discharge of water from the face (Figure 4.53). Stones should be set into well washed, cleaned cavities on a full bed of mortar (see ‘Mortars’). The top bed joint should be filled with ‘dry-pack’ mortar (the bedding mortar with a low water ratio). This crumbly material can be firmly and tightly packed using a tamping iron, to within 20 –30 mm of the face, and later pointed. Unless dry packing is used there will be a tendency for the top mortar bed to shrink, leaving the stone above unsupported. Replacement stone should always be identifiable, and there is a tradition of incising the date of placement on new stone. In some situations stone replacement within ruins may be structurally necessary and a stylistic difference may be made to differentiate new from old. Examples of this approach are shown at Battle Abbey (Figure 4.55) and Furness Abbey (Figure 4.56). In both cases a structural solution has been provided with the ‘correct’ stone, but stone which has not been moulded to imitate the original, deliberately staying short of any accusation of restoration. It can be argued, in the case of the Furness illustration, that some ghosting of the vertical roll and hollow moulding would have been an aesthetic gain; certainly, the course lines of the blocks should have been carried through. In cases where severely decayed masonry is in need of local replacement, an alternative to whole block indenting may be the use of tiles. The Society for the Protection of Ancient Buildings in England has for decades recommended the use of coursed clay tiles for philosophical reasons, in both hydraulic lime and cement and lime. The argument for this approach is that structural support can be introduced in a way that does not compromise the truth, even if the tile repairs were rendered, as they often were. An example of such a repair is shown in Figure 4.57, at Guildford Castle, carried out in 1914. A more elaborate example, including full arch rings, is shown used at Wolvesey Castle, Winchester (Figure 4.58). There are, however, some situations where the use of tiles, clay or stone, may be the most suitable and successful replacement option, quite apart from the philosophical issues of honesty in repair. Tile slips may, for instance, be introduced with less loss of original material, in the repair of decayed vault ribs, performing perfectly well in compression within the arch thrust of which they will become a part. Or within a very weak, cavernously decayed context such as the poorly lithified limestone of Herod’s Northern Palace at Masada, they may be almost the only option, as described in Chapter 10.2. The importance of this technique is primarily in its compatibility with the ancient fabric. Porous, permeable tiles of clay or weathered stone set in porous, permeable mortars are receptive to water vapour movement and the passage of salts in solution; they do not create new crises for the beleaguered walls in which they are set and can, in later years, be replaced. Unlike new stone or brick they can be expected to perform in a truly sacrificial role for the benefit of the old work they are trying to sustain. In setting tiles into a cavity created by the removal of decay, bedding mortar is placed on the upper face of the top tile and the joint below is dry-packed with a tamping iron to ensure that the soft mortar is pushed up into the irregular roof of the cavity.

 

 

 

Stone repairs

 

The retention, rather than the replacement, of the stones or bricks in a ruin should be an obvious choice. Old techniques of repairing stones included the ‘tinning’ of open fractures by running in liquid cement, joining detached fragments with shellac or clasping elements such as mullions and tracery with copper tape. Minimum replacement is frequently coupled with techniques of filling small lacunae with mortar (‘dental repairs’), or with small pieces of stone or tile within the boundaries of an old stone (‘piecing-in’), or of drilling and stitching fractured stones. The design of appropriate mortar fills must follow the general principles set out under ‘mortar specification for joints’, but in the repair context it is even more crucial to have good water vapour permeability and moduli of elasticity in addition to good texture and good, stable colour in both wet and dry conditions. Dental repairs, like stone repairs, should always be sacrificial and records of their composition must be retained so that future repair material can be repeated or modified. The technique of repair, preceded by photographic recording of the conditions, is to cut away the decay with small points or chisels, or even a spatula, scraping back to a sound surface and forming slight undercuts around the perimeter of the