|Home | FAQ | Finishing | Sump Pumps | Foundations|
Water-retaining structures such as reservoirs, swimming pools, sewage tanks, etc., are normally designed and constructed in accordance with the British Standard Code of Practice, BS 5337.
Leakage generally arises from three main causes: cracks; porous and /or honeycombed concrete; defective joints. In most cases this type of leakage is repaired with the structure empty, and the repair is effected on the inside (water face).
However, with structures holding industrial liquids, the emptying of the structure and the execution of the repair on the inside would cause great dislocation to the operation of the plant. In such cases the owner wishes the repair to be effected with the structure in operation. This means that the repair has to be effected against the pressure of the contained liquid. The author’s experience is that it's seldom that such a repair results in achievement of even reasonable watertightness, unless, after all individual areas of seepage have been effectively sealed, the whole of the outside is given a coat of structural gunite of adequate thickness.
If the work is carried out by an experienced contractor (who is unlikely to be the lowest tenderer if there is competitive tendering!), then considerable improvement can be reasonably expected, but the author doubts whether such a contractor would give a guarantee for complete elimination of seepage.
From time to time materials are offered on the market which are claimed to seal leaks against the pressure of the contained liquid by the process of the growth of crystals into the concrete. It is claimed that the crystal growth forces back the water in the pore structure of the concrete and the growth takes place after the proprietary mortar has been applied to the outside of the structure and has hardened. The author has not seen any evidence to support the claim of crystal growth inwards from an externally applied mortar or slurry. In hydrated cement paste in hardened concrete there are two distinct classes of pores, namely capillary pores and gel pores. The size of capillary pores is about 0.1um (0.0013 mm) (a human hair is about 0.1 mm). Gel pores are much smaller, being of the order of 15 - 20 Angstroms (one Angstrom is 10^-7 mm). In reasonable quality concrete capillary pores become discontinuous after some months. It seems to the author that the chances of crystal growth displacing the water in capillary pores is about zero.
9.4.1 Repair of Cracks
It has been stated in Section 5 that to repair a crack successfully it's first necessary to know why the crack has formed, and then to decide whether further movement is likely to occur which would cause the crack to reopen if it were repaired with a rigid material.
In new structures, cracking is usually more prevalent in the walls than in the floor and is generally caused by thermal contraction stresses in the early life of the concrete. This type of cracking has been discussed in Section 3, and a typical thermal crack in a cantilever slab is shown in Fig. 3.4. Figure 9.2 shows such a crack in a reservoir wall.
FIG. 9.2. Typical thermal contraction crack in reservoir wall.
For roofed structures which are either in the ground or are embanked, the chances of significant movement, once the tank is in operation, are quite small, unless foundation movement occurs. Therefore cracks in this type of structure can be safely sealed with epoxide resins and similar materials which are rigid. It is advisable for the sealing to be carried out as late in the construction process as possible. When the structure is filled with water the cracks will tend to close due to moisture movement as the concrete becomes progressively wetter. In this way, the sealant in the crack, whether it's inserted or injected, will be put into compression. Unless the structure remains empty for a considerable time and the concrete dries out, the cracks should remain closed.
These thermal contraction cracks penetrate right through the wall or floor and therefore constitute a plane of weakness. If they are very fine and have practically sealed themselves while under test by the deposition of calcium carbonate (sometimes known as autogenous healing), they should be repaired as described below. While epoxide resins are normally rigid when cured, specially formulated resins can now be obtained which possess some degree of flexibility. Polyurethanes can be formulated so as to be very flexible. However, for drinking water reservoirs, the sealant should be non-toxic, non-tainting and should not support bacterial growth. Bitumen compounds, if used in large quantities, can, under some circumstances, impart a phenol taste to the water.
A decision must be taken as to whether the cracks will be repaired by crack injection methods, as described in Section 5, or by a crack filling and sealing technique. A brief description of the latter is as follows:
(a) Carefully tap down the line of the crack with a chisel so as to remove any weak edges caused by honeycombing behind. It should be noted that this is not the same as cutting out the crack. For cracks of this type, the author does not consider that cutting out is either necessary or desirable.
(b) Remove the laitance on the surface of the concrete for a distance of 300 mm on each side of the crack. This can be done by power operated wire brushes, light grit blasting, high velocity water jets, or light bush hammering. All grit and dust must be removed from the prepared surface.
(c) Brush into the crack and onto the prepared area, a minimum of three coats of low viscosity epoxide resin. The resin must be formulated to bond to damp concrete.
Figure 9.3 shows the above work in diagram form.
FIG. 9.3. Method of sealing crack in r.c. wall or suspended slab.
The author is not in favor of sealing cracks by means of a strip of preformed material such as polyethylene, butyl rubber or polyisobutylene, which is fixed to the concrete by an adhesive. The reason for this is that the adhesive is liable to fail within a year or so because both edges of the strip are exposed to continuous immersion.
Questions are sometimes raised on the subject of ‘autogenous healing’ of cracks in concrete. The author has not seen an authoritative definition of this term, but it's usually taken to mean that the crack seals itself. This can only occur with very fine cracks, probably less than 010mm wide, provided there is no further movement at the crack. The author’s experience is that this ‘healing’ either occurs within the first week or so or the crack does not really seal itself.
There are cases, however, where there is reason to believe that the cracks may open at a later date during the anticipated operating cycle of the structure. This can occur when there is likely to be an appreciable drop in temperature in a structure where provision has not been made for this in the form of movement or partial movement joints or additional reinforcement. The thermal contraction stresses which develop may cause repaired cracks in the walls or floor to open as they are of course planes of weakness in the structure.
An example of this type of operating cycle occurs in a swimming pool. The water is maintained at a temperature of about 27°C, and the air temperature in the hall at about 29°C, for a period of three to six or more years. Then the pool is emptied for complete cleaning, maintenance and repair, usually in the middle of winter. This can result in a drop in temperature of 20—25°C. Pools which are elevated are particularly vulnerable to this, as the area below and around the pool shell is often used for plant rooms and stores, etc. When the pool is emptied, the opportunity is taken to service the plant and equipment as well.
The problem is how to repair cracks so as to allow a certain amount of flexibility and at the same time to ensure watertightness, and this can present considerable difficulty. If the movement across the crack is expected to be very small, then the use of an epoxide resin specially formulated to provide a degree of flexibility is probably as good as anything at present available. This would be suitable when there are a number of cracks parallel to each other and it could be assumed that the total movement would be distributed over say two or three cracks instead of concentrated across one. In such a case, i.e. where it's decided to use a flexible epoxide resin, the crack should be cut out to a width of 20 mm and to a depth of 20 mm; this groove should then be carefully filled in as directed by the supplier of the resin.
Crack injection as described in Section 5, using a resin of low viscosity and low ‘E’, value may also provide an acceptable solution. It must be remembered that the sealing of cracks in the structure itself with a semi-flexible material, does not in any way solve the problem of possible cracking in rigid materials, such as tiling or protective chemically resistant rendering, which may be applied later.
When it's considered that the possible movement across the crack will be greater than can be accommodated by the flexible epoxide resin, then another method should be adopted. The detailing of such repairs requires careful thought and the following is one solution, but this should not be considered as necessarily applicable to all cases.
(a) A channel should be cut in the concrete for the full height of the wall, so that the crack is within the cut-out section as shown in Fig. 9.4. The depth of this channel need not exceed 5 mm. The sides of the channel should be cut with a saw, and the surface of the channel should be smoothed with a carborundum wheel.
Following this preparatory work all grit and dust must be removed.
(b) A low viscosity epoxide resin should be injected into the crack. The resin should bond to damp concrete and possess some degree of flexibility. Special care is needed for this type of injection. If the anticipated movement across the crack is likely to exceed the safe extension of the resin it's better not to use resin injection at all. The reason for this is that the resin may bond so strongly to the concrete that when tension develops across the crack, the concrete may fracture nearby, usually parallel to the original crack.
(c) Line the channel with Hypalon, or chlorinated polyethylene, and carry this along the face of the concrete to a minimum distance of
75 mm each side of the channel, all as shown in Fig. 9.4. The sheet should not be less than 1 mm thick and should be bonded to the concrete with an adhesive which is not moisture-sensitive, over the 75 mm distance.
If the walls are to be rendered or given some other applied finish such as ceramic tiles, then the detail will be somewhat different, and a suggestion for this is shown in Fig. 9.5. It should be appreciated that the crack in the wall will not be directly below the joint in the rendering and tiling and therefore complete absence of subsequent cracking in the rendering, etc., can't be guaranteed. Figures 9.4 and 9.5 show resin coating across the crack on the outside face of the wall. Obviously this can only be applied if the outer face is accessible. The author is in favor of the use of a rigid (hard) resin for this external coating because if this fractures, it would indicate that movement across the crack was taking place and alert supervisory staff.
FIG. 9.4. Suggestion for sealing crack in wall.
FIG. 9.5. Suggested method for sealing crack in tiled r.c. wall.
9.4.2 Repair of Defective Joints
The author’s experience is that defective joints are the principal cause of leakage in water-retaining and water-excluding structures. Joints of any kind in this type of structure can be termed a necessary evil. It is impossible to build a liquid-retaining structure without joints; in addition, the joints must be correctly located and correctly detailed. It is unfortunate, to say the least, that some designing authorities leave matters relating to joints to the contractor and then blame him when leaks occur. This attitude is certainly contrary to the intention of the relevant clauses in Code of Practice BS 5337. The Code makes it clear that the location and detailing of joints is an essential part of the design and is the responsibility of the engineer. Joints detailed and designed to allow thermal movement to take place must be provided unless the engineer is satisfied that other satisfactory provision has been made to control thermal contraction cracking. There are numerous papers on this important subject and a reader who wishes information should refer to the Bibliography at the end of this section.
It is normal practice, although some engineers don't agree with this, to provide water bars in joints. In walls and suspended slabs, these are usually the centrally located dumb-bell type. In addition, the water side of the joint is usually provided with a sealing groove which is filled with a sealant. However, for what are known as monolithic (construction or day-work) joints, some engineers rely entirely on the bond between the old and new concrete for watertightness.
Figures 9.6 and 9.7 show defective joints in a reservoir before and during repair.
FIG. 9.6. Deteriorated sealant in reservoir wall.
FIG. 9.7. Joints in Fig. 9.6 after resealing with ‘Neoferma’ (EPDM) gaskets.
In spite of these rather elaborate and expensive precautions, joints often leak, In the case of walls and suspended slabs, the location of the leak is clearly visible, but with floor slabs supported on the ground, the tracing of the leak is very difficult indeed. Most structures of this type are underdrained, and if the layout of the drainage system has been designed with this problem in mind (i.e. each section of the underdrainage deals with a particular area of floor), then the locating of the leak is much facilitated. Even so, there is likely to be an appreciable area of floor which is under suspicion. Once it's established that the leakage is through the floor, then in most cases it's fairly certain that it's the joints which are responsible. The flow through the under-drains must be carefully monitored as each length of joint is ponded. This is a time- consuming and difficult job and requires patience and care. Sometimes reference to the daily reports from the R.E. or Clerk of Works will be useful in revealing some forgotten hitch in the work which may have resulted in under-compaction of concrete around a water bar, or displacement of a water bar during concreting. It is not possible in this guide to discuss all the possible trouble which can occur at a joint and so a typical example will be given to illustrate what the author considers are important principles.
FIG. 9.8. Suggested method of repair of defective joint in ground slab.
Figure 9.8 shows the situation as finally ascertained after lengthy investigations following the failure of a sewage tank to pass the prescribed water test. The displaced water bar was located by careful drilling alongside the joint. There were welded connections at C and D where the displaced water bar CD joined water bars EF and CK, and DJ and GH. T and T are the positions of the two drill holes which located the displaced water bar.
The problems involved, and the line of argument developed, may be stated as follows:
1. As the water was passing through the joint, it was obvious that neither the sealant nor the water bar were satisfactory. It was relatively easy to remove the sealant (which was a rubber bitumen) and replace it. This may have cured the leak.
2. However, it was known that sealants are not long lived materials, particularly rubber bitumens and polysulfides. Although neoprene has a much longer life, it's also likely to deteriorate in the course of time. As soon as deterioration starts and /or bond with the sides of the sealing groove begins to fail, leakage will recommence because the water bar is in some way defective. The water bar was specified as an additional safeguard, but due to some failure by the contractor, it no longer fulfilled its purpose.
3. The contract was already behind schedule and the tank was urgently needed. The contractor accepted responsibility for the remedial work.
The real question then was, what was a practical and reasonably satisfactory method of repair?
One suggestion was to cut out a strip of concrete about 100 in wide parallel to the joint CD. Another idea was to completely remove bays A and B, but this immediately raised the question of how to deal with the water bars in the joints EG, F’H, EF and GH. A further suggestion was to try to pressure grout the defective concrete around the water bar in bay A. This was rejected as it would probably result in the grouting-in of part of the underfloor drainage.
The author’s suggestions are given below; these were based on the principle that as little concrete as possible should be cut outs
(a) The existing sealant should be cut out and the sealing groove deepened by about 15 mm.
(b) Both sides of the sealing groove should be reformed by first cutting the concrete with a saw to widen the joint, and then truing-up the sides with an epoxide mortar.
(c) After the reformed groove had been carefully cleaned out, a cold-cured epoxide resin should be inserted to a depth of about 15 mm. This resin should possess low viscosity, flexibility and should bond to the damp concrete sides of the groove. The reason for providing this under layer of resin was that the final (top) sealant described in (e) below, would not bond to the existing sealants in the adjacent joints.
(d) After completion of (c), a cellular neoprene jointing strip should be inserted in the groove above the epoxide resin under layer. The strip would be bonded to the concrete sides of the groove by ‘a suitable primer. It would not be bonded to the epoxy under layer.
(e) For a distance of 600 mm each side of the joint and for its full length, the surface of the concrete should be prepared by mechanical scabbler, grit blasting or high velocity water jets, so as to lightly expose the coarse aggregate to a depth of about 5 mm.
(f) An epoxide resin mortar should be laid on the prepared surface of the concrete to a minimum thickness of 5 mm, the resin to be specially formulated to bond to damp concrete.
The work was carried out generally as described, and no further leakage was observed. It was completed in five working days, compared with an estimated minimum of three weeks for any of the alternative schemes. The formation of additional joints, with corresponding hazards of leakage, was thus avoided.
9.4.3 General Repairs and Lining of Water-retaining Structures
While the principles of chemical attack on concrete are dealt with in Section 2, there are many borderline cases of mild chemical attack on concrete in service reservoirs due to certain aggressive characteristics of the raw or partially treated water. This occurs particularly where the water is derived from an upland gathering ground.
Such waters are characterized by:
Occasionally, the acids include sulfuric derived from the breakdown of sulfur compounds in peat and marshy ground by bacteria. The presence of sulfuric or other mineral acids will greatly increase the severity of attack. It is often found that the pH varies over a range from about 4 to 6 in some parts of the country there is a significant drop in pH after heavy rainfall.
While it's customary in some cases to treat the raw water before it enters the reservoir by filtration through limestone chippings and /or addition of lime, with sudden variations in the acid content of the water the pH correction may not always be as successful as one would desire.
These facts are mentioned because they have an obvious bearing on the assessment of any damage done and on the method of repair.
The pH value itself has no direct relationship to the amount of acid present; its potentional for attack can only be determined when the type of acid causing the low pH has been established by chemical analysis.
Water with a pH in the range of 7.0 – 8.0 is alkaline but this does not necessarily mean that it will not attack Portland cement concrete. Under certain conditions such water can attack concrete due to its potential in dissolving out the calcium hydroxide (lime). The lime dissolving power of such waters depends on the amount of dissolved carbon dioxide, the hardness expressed as calcium carbonate, the alkalinity measured by methyl orange and expressed as calcium carbonate, the temperature, and the original pH of the water, and total dissolved solids.
The Langelier Index, or calcium carbonate saturation index, can be used to determine whether the water is lime dissolving (denoted by a negative value to the Index) or lime depositing (denoted by a positive value to the Index).
The saturation pH (pH is calculated from the total dissolved solids, the temperature (°C), calcium hardness as CaCO3 and the methyl orange alkalinity as CaCO3 The saturation pH is then subtracted from the original pH and the result is the Langelier Index of the water. For further information on the Langelier Index reference should be made to the book, Water Treatment for Industrial and Other Uses, by E. Nordell, and a paper by W. F. Langelier in the Journal of the American Water Works Association in 1936; both are listed in the Bibliography at the end of this section. Unfortunately there are wide differences of opinion among chemists on the aggressive effects of waters with different negative values of the Langelier Index.
The only specific information the author has seen on this important matter is a draft prepared by the International Organization for Standardization, Technical Committee ISO/TC77, Reaction of Asbestos- cement Pipes to Aggressive Waters. This draft recommends that a water should be considered as highly aggressive to asbestos—cement pipes if the Langelier Index is more negative than —2. It would be moderately aggressive with a Langelier Index in the range of 0 to — 20.
The author considers it's prudent to have water checked for the Langelier Index in those cases where the normal chemical analysis suggests that the Index may be negative.
The Repair of Etched Surfaces
Attack on the concrete by soft acidic waters, and waters with a negative Langelier Index, normally show as etching of the concrete surface (see Figs. 9.9 and 9.10). The etching will vary in depth depending on the degree of attack which in turn will depend on the aggressiveness of the water and the chemical resistance of the concrete.
FIG. 9.9. Concrete deeply etched by soft moorland water.
FIG. 9.10. Deteriorated concrete channels carrying soft moorland water.
Provided the deterioration is confined to the surface of the concrete and there has been no significant corrosion of the reinforcement, the following measures can be adopted; the one actually selected will depend on the circumstances of each case:
1. The surface of the concrete must be prepared by wire brushing or water jetting so as to provide a clean strong base on which the new coating will be applied. It is essential to do this preparatory work carefully so that the best possible bond is obtained.
2. On the prepared surface of the concrete, apply 2 thick brush coats of grout made with 2 parts ordinary Portland cement and 1 part styrene—butadiene latex by weight. The first coat must be well brushed in and the second coat should be at right angles to the first.
3. Apply to the prepared surface of the concrete two coats of epoxide or polyurethane resin, to a minimum thickness of 03 mm. For drinking water reservoirs it's likely that polyurethane would not be acceptable as the coating must be non-toxic, non-tainting and should not support bacterial growth.
4. Where the attack has been severe and there has been some penetration of water to the reinforcement resulting in limited corrosion, the application of a 50 mm thick layer of gunite, reinforced with a light mesh, would be justified. Detailed information on gunite is given in a later section.
Bituminous compounds have in the past been used for this type of protection, but generally have not proved particularly durable. They don't bond well td damp concrete, and some may impart a phenol taste to the water.
The Complete Lining of Deteriorated Structures
Cases sometimes arise, particularly with water towers and similar elevated structures, where the provision of a complete lining is required in order to ensure a satisfactory standard of watertightness. For a complete lining, there are two basic alternatives: insitu material or preformed sheeting.
Insitu materials. These include cement grout composed of OPC and SBR latex, epoxide and polyurethane resins, and gunite. As previously mentioned, for structures holding drinking water, the coating must be non-toxic and non-tainting.
For the coatings, the correct preparation of the base concrete is essential. The author considers that it's most advisable for the supplier of the material to be responsible for all aspects of the work, i.e. preparation of the concrete, and supply and application of the coating, as this avoids divided responsibility.
Prior to the commercial use of organic polymers, such as epoxides and polyurethanes, mastic asphalt and bituminous emulsions were used for lining water-retaining structures. These latter two materials are now very seldom used; they have been replaced by Portland cement and styrene— butadiene slurry coats for mildly aggressive conditions and by polymer resins (certified as being suitable for use with drinking water when used in potable water reservoirs).
Latex grout: polymer emulsions and Portland cement. The treatment of the floor and walls of a water-retaining structure with a grout composed of two parts Portland cement and one part styrene—butadiene resin latex emulsion (SBR) by weight can be effective in sealing the surface of slightly porous concrete subjected to a low head of water. There are no figures available for maximum safe water pressure, but the author would not feel confident in the use of this material under a pressure exceeding 150m head.
It is simple to apply and relatively cheap compared with more sophisticated coatings. Provided the limitations mentioned above are kept in mind, it can prove a useful solution to minor loss of water from small tanks and channels. There are a number of proprietary latexes on the market, but taking as an example the SBR latex 29 Y 40 manufactured by Revertex Ltd, and sold under proprietary names by various firms, the following is a guide to its use:
(i) Mix: 2 kg ordinary Portland cement to 1 kg latex.
(ii) Coverage in two coat work: 5 m liters of grout.
(iii) The surface of the concrete must be clean and free of weak laitance.
(iv) Each coat must be well brushed in, and allowed to dry before the next one is applied. The second coat must be applied at right angles to the first.
Acrylic resin emulsions can also be used, but they tend to be more expensive.
In addition to the basic emulsions mentioned above there are a large number of proprietary materials on the market; the author feels that the most satisfactory results are likely to be obtained by employing firms who supply and apply the coatings.
Epoxide and polyurethane resin coatings. These materials are basically organic polymers and can be formulated to give the coatings a wide range of characteristics such as chemical resistance, bond strength, modulus of elasticity, viscosity, setting and curing time, ability to cure in very wet conditions including complete submergence, abrasion resistance and colour.
The client (or his professional adviser) should specify as clearly as possible the conditions under which the coating has to operate. If the coating has to possess considerable elasticity, this should be defined as accurately as possible. Such expressions as ‘the coating must be capable of spanning hair cracks’ should not be used as there is no general acceptance of the width of a hair crack.
This type of coating will bond very strongly to clean, good quality concrete, and the bond is an essential factor in the long term serviceability of the coating. When failure occurs it's normally bond failure at the interface with the base concrete. It is essential that all weak laitance and any surface contamination such as oil, etc., be completely removed. It is better if the preparation of the concrete and the formulation, supply and application of the coating are all carried out by one firm as this avoids divided responsibility.
For the complete lining of a liquid retaining structure, a minimum thickness of 0 mm is recommended and this would normally require three coats. The resin can be applied by brush or airless spray. When properly executed it provides a completely jointless lining which is very durable.
As with all applied coatings which require to be bonded to the base concrete, the limiting factor may be the actual quality of the concrete. With a weak porous concrete the effective life of the lining may be disappointing, simply because the concrete is not strong enough to hold the coating in position. In such a case, reinforced gunite may be the best solution.
Reinforced gunite. Gunite is a pneumatically applied material, consisting of cement, aggregate and water. Depending on the maximum size of aggregate used and the grading, it can be considered either as a mortar or a concrete.
The advantage of this material for watertight linings compared with concrete, is that it does not require formwork, is self-compacting, and that except for large structures, or in those cases where the gunite is bonded to the existing structure, all the joints are monolithic. In addition to providing a watertight lining, reinforced gunite can be designed to strengthen an existing structure. This use of gunite has been described in Section 5.
Gunite is used more extensively in the USA, South Africa and Australia than in this country. However, in recent years its use in the UK for new swimming pool shells and the lining of deteriorated pools and other liquid-retaining structures, has increased.
Gunite work is usually offered as a ‘Design and Construct’ package deal type of contract by a few specialist firms. With certain aggregate gradings and the use of specialist ‘gunning’ equipment, equivalent 28-day cube strengths in excess of 40 N/mm can be obtained compared with about 35 N/mm for normal quality water-retaining concrete. The higher compressive strength is accompanied by higher tensile and bond strength as well as impact resistance.
When reinforced gunite is used to line a structure, the reinforcement can be pinned to the old walls and floor and the gunite forms a monolithic lining bonded to the existing structure. An alternative, which is usually adopted when the existing structure is found to be in poor structural condition (low strength concrete, numerous cracks, etc.), is to separate completely the new gunite lining from the old structure. The separation is generally formed by a membrane of 1200 gauge poly ethylene sheeting (03 mm thick). The reinforced gunite lining then forms an independent structure, and would be designed as such.
If a bonded lining is used, then all movement joints in the existing structure should be carried through the gunite. When a liquid-retaining structure has cracked at locations other than along the line of joints, it means that these cracks may act as movement joints. This is the principal reason why an independent unbonded lining is adopted.
It was mentioned earlier in this section that unless the structure is very large, the gunite-independent lining would have no movement joints but only monolithic day-work type joints. A natural question, therefore, is why should provision have to be made in a concrete lining for thermal contraction (stress relief or contraction joints) while gunite appears to be satisfactory without any movement joints at all. It is not easy to give a completely clear, unambiguous answer to this, but the factors involved are:
(a) A reinforced gunite lining is usually appreciably thinner than a lining in reinforced concrete. The gunite lining has no formwork, resulting in far less build-up of heat from the hydrating cement, and consequently lower thermal stress, than concrete.
(b) Gunite linings are usually reinforced with heavy steel fabric and the amount of horizontal (distribution) steel is likely to be greater per unit thickness of wall or floor than would be provided for concrete.
(c) Good quality, high strength gunite, as placed, has a very low w/c ratio, about 0.33-0.35. Concrete used for lining a liquid-retaining structure would normally have a w/c ratio of about 0.48—0.50.
The above factors add up to the fact that there would be appreciably less thermal contraction in the gunite lining than in an insitu concrete lining, followed by less drying shrinkage, coupled with a heavier weight of reinforcement to control cracking.
However, cracking is not unknown in gunite linings, but is generally caused by inadequate precautions being taken to prevent rapid drying- out of the surface under adverse weather conditions (hot sun, strong winds). This rapid drying-out shows in the form of fine surface cracks. It is advisable for provision to be made for proper covering up of the gunite as the work proceeds.
The bulk density of high strength gunite is about 09 that of normal dense concrete, i.e. 0.9x2380=2100 kg/m^3.
The surface finish of gunite is generally rougher than that of good quality concrete. This may be considered a serious disadvantage when gunite is used to line potable water tanks as some water engineers consider the rough surface would encourage the growth of algal and fungoid growths. This can be partly overcome by finishing the gunite with a wood or steel float.
There are three basic methods of lining water-retaining structures with sheet material: fully bonded lining; partially bonded lining (spot bonding); and unbonded lining.
In the UK for new structures such as reservoirs and water towers a fully bonded lining is normally adopted. In recent years, the technique of using a loose bag’ of PVC sheeting has been introduced for small swimming pools and it could be used for other structures in certain restricted circumstances.
Fully bonded and partially bonded linings. The basic principles for successful application of these two types of lining are very similar and therefore are considered together. A partially bonded lining is used where appreciable further movement in the structure is anticipated. Care is taken to ensure that the lining is left unbonded across all planes of movement and in this way, excessive build-up of stress in the lining material is avoided. Generally about 75 of the area of lining is bonded to the substrate.
The great advantage of a bonded lining compared with an unbonded one is that should the lining be damaged in a limited number of places, it's most unlikely that leakage would occur. The reason for this is that the liquid can't travel behind the lining until the adhesive has become very deteriorated. The adhesives used are water-resistant, but it must be admitted that if water does penetrate through the membrane, in the course of time the adhesive may suffer deterioration. The degradation of the adhesive and consequent loss of bond will show itself by the bulging of the lining and this should be detected in the course of normal maintenance inspections before the water is able to penetrate through the structure.
The disadvantage is the careful preparation which is required of the base concrete; this preparatory work is detailed below. However, on balance, the author is of the opinion that for structures such as water towers, etc., the advantages of a fully or partially bonded lining far outweigh the disadvantages.
As with any material requiring good bond, careful preparation of the base concrete is essential. In many cases, linings of this type are applied to old structures which over the years have been coated with other materials (such as bitumen) to help improve watertightness. All such coatings which are unsound or blistered must be completely removed. It is important that the surface to which the sheeting will be bonded is strong and relatively smooth. A scabbled or grit blasted surface would not be suitable as the irregularities on the surface created by the pieces of coarse aggregate would cause local stress in the sheeting. Therefore if scabbling or similar is required to remove contamination a thin fully bonded coat of rendering must be applied to the base concrete. If the concrete is reasonably clean, but contains rough areas, these must be ground down.
The sheeting should be well bonded to the concrete with a special adhesive which is not water-sensitive and all seams must be lapped and solvent (cold) welded. An alternative is to use a special tape which is inserted between the sheets and then the joint is fused.
Figure 9.11 shows a previously leaking reservoir after lining with fully bonded sheeting.
It is important that the sheeting be continuous over the floor, and carried up columns and walls to at least 300 mm above top water level. If the sheeting is finished below water level, then it's likely that continuous immersion of the unprotected edges over a long period of time will result in deterioration of the adhesive, causing loss of bond and allowing water to penetrate behind the lining. Even so, the top exposed edge of the sheeting needs careful finishing to avoid ingress of condensation as this is always present in tanks holding liquid at ambient temperature. Reservoirs and water towers contain a number of pipes, both inlet and outlet, and the sheeting must be carefully cut and fitted around these and all joints lapped and solvent welded. Any joints in the structure need special attention and detailing so as to prevent build-up of stress in the lining.
The lining of tanks with flexible sheeting is highly specialized and should only be entrusted to an experienced firm who should be required to give a guarantee of satisfactory performance for not less than 10 years.
Preformed sheeting is fairly easily damaged by small metal tools, hobnail boots, etc., and therefore special precautions have to be taken when carrying out inspections, maintenance and cleaning work.
Unbonded lining. The use of an unbonded lining of flexible sheet material for the waterproofing of water-retaining structures appears to have started with small private swimming pools. The material generally used is polyvinylchloride (PVC) (also known as vinyl), and the sheets vary in thickness from 0 to 1 mm. It is unlikely that PVC would be accepted for lining potable water tanks due to possible leaching of chemicals from the PVC. Chlorinated polyethylene is also used.
The advantages claimed for the use of unbonded lining are that movement of the structure does not induce stress in the lining, and that provided the surface on which it's laid is reasonably smooth, no special preparation is required. Also, with open structures (not roofed), the sheeting can be laid in almost any weather. The author has seen this method used on roofs and for lining small swimming pools.
While the horizontal area which can be covered by unbonded sheeting is virtually unlimited, the vertical height must clearly be limited, probably to a maximum of about 1 m, although the author has not seen any figure quoted. With any form of complete lining, it's important that it should not be subjected to back-pressure from ground water when the structure is empty as there would be a danger that the sheeting would be forced away from the substrate. In the case of unbonded linings it's generally considered essential that all necessary precautions be taken to prevent any back-pressure developing, however slight. The weight (pres sure) of the contained liquid moulds the sheeting to the exact shape of the structure, and if at a later date the lining is forced out of position by pressure from behind, it's unlikely ever to return to its previous position and shape.
FIG. 9.11. Inside of water reservoir lined with fully bonded sheeting.
FIG. 9.13. View of tank in Fig. 9.12 during repair with reinforced gunite
FIG. 9.12. Deteriorated concrete tank holding industrial effluent
FIG. 9.14. View of tank in Figs. 9.12 and 9.13 after completion of repair
The External Use of Reinforced Gunite to Strengthen Liquid-retaining Structures
A previous section has dealt with lining liquid-retaining structures with reinforced gunite. This type of lining is usually applied to below-ground structures. For structures which are above ground, such as many of the tanks at sewage treatment works, sludge and settling tanks for industrial effluent, and water towers, it's sometimes necessary to strengthen the tanks from outside. Reinforced gunite is very suitable for this work.
The gunite coating, reinforced as required to provide the necessary additional strength, is fully bonded to the base concrete. The existing concrete must be carefully prepared to receive the gunite. It is good practice to cut away all spalled and defective concrete, remove rust from reinforcement, cut out and repair all major cracks, and thoroughly clean the surface of the concrete by grit blasting or high velocity water jetting. The new reinforcement is then securely pinned to the old concrete and gunning is commenced.
Figure 9.12 shows a badly deteriorated concrete settling tank used for industrial effluent. Figure 9.13 shows the same tank being prepared for guniting. Figure 9.14 shows the same tank after completion of external structural gunite.
|Top of Page | Home | Prev: Tracing Leaks||Next: Repairs Where Leakage is Inwards Into the Structure||All Articles|