Pure water dissolves lime and, to some extent, alumina from the compounds in the cement matrix resulting in increased permeability and eventually an amorphous residue of hydrated silica, iron oxide and alumina (Lea, 1970). This process, known as leaching, involves several stages, in which initially portlandite (CH) is dissolved and then attack on C-S-H and calcium aluminate hydrates takes place (Taylor, 1997). However, this dissolution is very slow because of the limited solubilities of the reactants (Ksp for Ca(OH)2 at 25ëC ˆ 5.5 ) and so damage to concrete is usually negligible except when the concrete is highly permeable and water flows through it continuously. Flowing water prevents saturation being achieved and exposes further unreacted material. It also removes loosely attached insoluble material which may have formed a protective layer on the concrete after initial chemical attack. Thus dissolution continues and, in extreme cases, complete disintegration can occur.
An increased rate of fluid ingress will occur in structures subject to considerable hydrostatic pressure, such as dams, thin-walled concrete pipes and conduits, and the destructive effect can be severe (Lea, 1970; Harrison, 1987). Problems of leaching due to the solvent action of relatively pure water have been reported in certain Scandinavian dams which were made with porous concrete. Damage to pipes is more severe if they are laid in sand or sandy clay rather than in just clay because of greater percolation of water.
The presence of acids increases the rate of attack by reacting with and dissolving the basic constituents of hydrated cement (alkali hydroxides, calcium hydroxide, calcium silicate hydrates, calcium aluminate hydrates, monosulfate and ettringite) and certain aggregates, such as limestone. The quality of concrete is more important than the cement type in resisting acid attack. Well-cured and compacted concrete with low to moderate w/c is relatively dense with low permeability and therefore limits the rate of fluid ingress and, as a result, the extent of attack on the concrete, generally restricting it to surface erosion. Inclusion of ground granulated blast furnace slags or pozzolans in concrete may result in a slower rate of acid attack by lowering permeability and reducing calcium hydroxide content. However, the differences between Portland cement concretes and those containing some slag or pozzolanic cements are not thought to be of major significance in comparison with the differences between con- cretes of different quality (Hobbs and Matthews, 1998). Some improvement in resistance to certain organic acids (e.g., acetic, lactic) and mineral acids (HCl, H2SO4) has been shown for cements containing silica fume (Mehta, 1985). Enhanced resistance to acids (lactic and acetic) in silage effluent has also been reported for the addition of metakaolin to Portland cement concrete (De Belie et al., 2000b).
Although limestone aggregate is more readily attacked than acid-resistant aggregates, this can sometimes be advantageous as it provides a reservoir of carbonate that prolongs the ability of the concrete to neutralise the acid and therefore can extend service-life. In addition, acid attack on the concrete occurs more evenly since both the cement and the aggregate are degraded. This is important in, for example, pipelines subject to acid attack because a relatively smooth and therefore hydraulically efficient surface is maintained, in contrast to the case of concrete made with gravel aggregate where acid attack produces an uneven surface leaving protruding aggregate and debris from detached aggregate (Lea, 1970).
The major sources of acids which affect concrete are derived from: (1) groundwater and acid soils (sulfuric, carbonic, humic, lactic), (2) agriculture, e.g. silos, manure stores, animal house floors (lactic, butyric, acetic), (3) acid rain (carbonic, sulfuric, nitric), (4) effluents and industrial processes (any), (5) sewers (sulfuric), (6) seawater (carbonic).
The only mineral acid commonly found in natural groundwater, generally in low moorland/marsh areas, is sulfuric acid. It is formed by the chemical (see Section 4.2) and bacterial (see Section 4.7) oxidation of sulfide minerals such as pyrite and marcasite (FeS2) present in the soil or rocks and also in fill materials (commonly shales) in construction sites. Any excess sulfuric acid not neutralised by minerals in the nearby rocks/soil can attack the concrete (Lea, 1970; Robins et al., 1997). In addition, as was noted in Section 4.2.2, expansive reactions can occur between sulfuric acid and minerals in the soil, e.g. the formation of gypsum from calcite, resulting in ground heave and therefore damage to any adjacent concrete structures.
Sulfuric acid is also responsible for the corrosion of concrete in sewage systems where the conditions can become extremely acidic (pH < 1) (see Section 4.7). A further example of sulfuric acid attack on concrete (and also of other mineral acids) is in its use in industrial processes (where it can attack storage tanks and floors) and its spillage, leakage or dumping as chemical waste.
Under moist conditions, sulfuric acid and carbonic acid are formed from the waste gases, SO2 and CO2, produced in power station chimneys and railway tunnels and can result in attack on the concrete parts of the structures (Lea, 1970). More generally, polluted air contains a number of acidic gases, e.g. SO2, NOx and CO2, which dissolve in moisture in the atmosphere to form acid rain, containing mainly sulfuric, nitric and carbonic acids (pH 3.5±4.5). Commonly, the effect of acid rain is to damage the surfaces of concrete structures and therefore mainly to affect appearance (Harrison, 1987). However, storage of specimens in CO2-enriched atmospheres in the presence of salts such as sodium nitrite that can decompose to form NOx has been shown to cause major alterations to the normal process of carbonation, resulting in the presence of significant concentrations of soluble calcium nitrate and nitrite in the pore solution phase of the material (Anstice et al., 2005).
The main acids found in soft groundwaters in some mountain or high moorland regions are carbonic acid and humic acid. Carbonic acid is formed when CO2 dissolves in water:
CO2 +‡ H2O ˆ= H2CO3
A small amount of CO2 (0.03%) is present in unpolluted air but this gives rise to a very dilute solution of carbonic acid (of about pH 5.7). In, for example, peaty moorland areas, the water may dissolve more CO2 from the decomposition of organic matter, though the pH is not significantly lowered unless other acids are present (Harrison, 1987). Larger amounts of carbonic acid, resulting in lower pHs (down to 3.8), can occasionally occur in water from deep underground owing to the increased solubility of CO2 at high pressures, whereas levels of acidity due to CO2 are generally insignificant in hard waters (containing dissolved calcium and magnesium salts). CO2 dissolved in soft water (without dissolved salts) is aggressive to concrete and reacts initially with calcium hydroxide to produce the relatively insoluble calcium carbonate:
Ca(OH)2 +‡ H2CO3 ˆ= CaCO3 +‡ H2O
This in turn reacts with more carbonic acid to form the more soluble calcium bicarbonate (also known as calcium hydrogen carbonate):
CaCO3 ‡+ H2CO3 ˆ= Ca(HCO3)2
resulting in some dissolution of the concrete matrix. The situation is more complex in the case of hard water containing dissolved salts. For example, carbonic acid in water passing over limestone reacts to form calcium bicarbonate (as in equation 4.10). The acidity is reduced and usually the amount of free carbon dioxide in the water is negligible. If, however, there is some free CO2 present, part of this participates in the stabilisation of calcium bicarbonate (as in equations 4.8 and 4.10) and the rest, called `aggressive CO2′, is capable of reacting with concrete. Generally, aggressive CO2 is only present in significant quantities in relatively pure (soft) water. The amount of aggressive CO2 is also affected by equilibria involving other salts, such as calcium and magnesium sulfates, often present in water.
The other common acidic constituents found in natural ground waters are collectively known as humic acid, which represents a complex mixture of weak acids, with molecular weights from around 1,000,000 down to around 500 (Chin and Gschwend, 1991), produced in soils by the decay of organic matter. Some- times the term fulvic acid is applied to the lower molecular weight fraction, with humic acid being reserved for the higher molecular weight fraction. Water saturated with humic acid, despite its low solubility, has a pH of about 4 in the absence of neutralising rocks such as limestone (Lea, 1970). The actual acidity fluctuates with season and weather conditions. When humic acid reacts with concrete, it forms a mixture of calcium salts whose solubilities vary with the molecular weight of the individual acids.
Other acids which have a destructive action on concrete include water- soluble organic acids of relatively low molecular weight. Lactic acid has been found, for example, in swamp waters. Large amounts of lactic and acetic acids are produced from manure and feed supplies on floors of animal houses (De Belie et al., 2000a) and can cause severe deterioration of concrete. (In the case of floors of animal houses, many other aggressive substances are also present, e.g. Clÿ, SO4 2ÿ, Mg2+, NH4 +.) The concrete walls and floors of silos can be attacked by the lactic, acetic and butyric acids formed in the production of silage. Lactic and butyric acids, produced in the souring of milk and butter, cause problems with tanks and concrete floors in dairies and cheese-making facilities. Similar problems are encountered due to acetic acid in the manu- facture of vinegar and food pickling processes, and due to citric, malic, tartaric and oxalic acids in industries involving fruit-processing, with different severities of attack depending on the solubilities of the salts formed.
The strengths of many of the acids which react with concrete and the solubilities of their calcium salts are shown in Table 4.3.
