SD1:2005 Part B - Draft 1 v5 8/12/04 1

Part B: Chemical attack on concrete

B1. General
This Part deals first with sulfate attack and acid attack,
these being the principal types of chemical attack that are
of concern for concretes placed in the ground in the UK.
The aggressive chemical agents responsible commonly
occur in both natural ground and land contaminated by
activities of man.
Additionally, this Part identifies the more rarely occurring
forms of chemical attack, caused by high levels of
chemical species such as ammonium and chromium, and
organics such as phenols. Generally these agents are
found in troublesome concentrations only in
contaminated land. With some exception, specific
guidance is not given in this Special Digest in respect of
protecting concrete from the action of these less
commonly destructive agents. Generally the protective
principles applied in Part D will be beneficial, for example,
specifying a well-compacted concrete, with a low
water/cement ratio or providing an appropriate protective
coating. Specialist advice should be sought when
appropriate.

Finally, this Part explains the action of aggressive carbon
dioxide in respect of concrete in contact with flowing
water. While not taken into account in a standard ground
investigation or general concrete specification, the
possible damaging effect of high levels of aggressive
carbon dioxide are catered for in Part F in the design of
specific precast concrete products such as pipeline
systems. The potential for this form of attack should also
be taken into account when designing cast-in place
structures that carry flowing water, eg culverts.
B2. Principal types of chemical attack on
concrete
B2.1 Sulfate attack
The essential agents for sulfate attack are sulfate anions
(SO
4
2-
). These are transported to the concrete in various
concentrations in water, together with cations, the more
common of which are calcium, magnesium and sodium.
Where porous concrete is in contact with saturated
ground the water phase is continuous across the
ground/concrete interface and sulfate ions will be readily
carried into the body of the concrete. Well compacted,
dense, low water/cement ratio concrete in such an
environment will, however, initially restrict access of the
ions to the surface layer.
Migration of sulfate ions from unsaturated ground into the
concrete can take place by diffusion provided there is
sufficient water to coat the particles of soil, but the rate will
be slow and dependent on the sulfate concentration.
The reactions that take place when sulfates enter the
concrete matrix are complex and contentious. There is
extensive research literature on the topic, including some
recent collaborative books and conferences
[1- 4]
. A simple
guide is given here in order to understand the basic
chemistry and resultant effects. The reactions have been
demonstrated to depend on the type of cement, on the
availability of reactive carbonate in, for example, the
aggregate and groundwater, and on the temperature. Two
separate forms of sulfate attack on Portland cement
concretes are described here:
a well-known type (commonly called the conventional
form of sulfate attack) leading to the formation of
ettringite and gypsum;
a more recently identified type producing thaumasite.
In practice, both can operate together to some extent in
field conditions in buried concrete.
Sulfate attack can only be diagnosed when the concrete in
question is showing physical signs of degradation such as
expansion (with or without notable cracking), surface
erosion or softening of the cement paste matrix. The
identification of levels of sulfate significantly greater than
4% by weight of cement within the surface of a visually
sound concrete does not automatically imply that sulfate
attack has taken place; it may only be a warning of
potential attack in the future.
B2.1.1 Conventional form of sulfate attack
For sulfate attack to occur leading to the formation of
ettringite and gypsum in susceptible concrete the following
must be present:
a source of sulfates, generally from sulfates or sulfides
in the ground;
the presence of mobile groundwater;
calcium hydroxide and calcium aluminate hydrate in
the cement matrix.
SD1:2005 Part B - Draft 1 v5 8/12/04 2
In the highly alkaline pore solution (pH>10) provided by
the sodium, potassium and calcium hydroxides liberated
during the cement hydration reactions, sulfate ions that
have penetrated the hardened concrete react with
calcium aluminate hydrate to form calcium sulfo-
aluminate hydrate (ettringite, CaO.Al
2
O
3
.3CaSO
4
.31H
2
O).
The formation of this mineral can be destructively
expansive since it has a solid volume greater than the
original constituents and it grows as myriad acicular
(needle-shaped) crystals that can collectively generate
high internal stresses in the concrete.

In sulfate-resisting Portland cement (SRPC), the
tricalcium aluminate (C
3
A in cement notation) level is kept
to a minimum so reducing the extent of this reaction.
Incoming sulfate ions may also react with calcium
hydroxide Ca(OH)
2
to form gypsum (calcium sulfate
dihydrate, CaSO
4
.2H
2
O). This reaction product also has a
greater solid volume than the original constituents and in
some cases can contribute to degradation of the
concrete. If magnesium ions accompany the sulfates,
they may also react with calcium hydroxide, producing
brucite (magnesium hydroxide, Mg(OH)
2
) which because
of its low solubility precipitates out of solution, also
leading to increase in solid volume. Magnesium ions may
also attack calcium silicate hydrates, the principal
bonding material in set concrete.
Laboratory tests show that the first effect of the
conventional form of sulfate attack is to increase the
strength and density of the concrete as the reaction
products fill the pore space. When it is filled, further
ettringite formation induces expansive internal stresses in
the concrete which, if greater than the tensile strength of
the concrete, will expansively disrupt the affected region.
This cracking together with white crystalline
accumulations are the characteristic signs of the
conventional form of sulfate attack.
B2.1.2 Thaumasite form of sulfate attack (TSA)
A comprehensive account of this form of sulfate attack
has been given in the Report of the Thaumasite Expert
Group
[1]
and

in

the proceedings of a special conference
[4]
and so only the essentials are mentioned here. Several
factors must generally be coincident for TSA to occur in
susceptible concrete:
a source of sulfates, generally from sulfates or sulfides
in the ground;
the presence of mobile groundwater;
a source of calcium silicate hydrate, mostly derived
from cementitious calcium silicate phases present in
Portland cements;
the presence of carbonate, generally in coarse and/or
fine concrete aggregates, as bicarbonate in
groundwater or as a constituent of the cement;
low temperatures (since thaumasite formation is most
active below 15C).
The availability of carbonate ions (CO
3
2-
) changes the
reaction products when sulfates enter the concrete. Below
about 15C in the presence of water, the reactions
between the calcium silicate hydrate, the carbonate and
the sulfate ions produces thaumasite (CaSiO
3
.CaCO
3
.
CaSO
4
.15H
2
O). The calcium silicate hydrates provide the
main binding agent in Portland cement, so this form of
attack weakens the concrete as well as causing some
expansion and, in advanced cases, the cement paste
matrix is eventually reduced to a mushy, incohesive mass
(as in Figure B1).
Since TSA does not depend on the level of calcium
aluminate hydrates, SRPC concretes can be vulnerable to
this form of attack. Concretes containing granulated
ground blastfurnace slag (ggbs) as part of the cement
have good resistance to TSA. Concretes made with other
cement types must rely on achieving very low permeability
for resistance.
Figure B1 Sample of formerly high quality concrete from
highway bridge foundation that has been severely affected by
TSA. The outer 50 mm of concrete has been reduced to a mushy
SD1:2005 Part B - Draft 1 v5 8/12/04 3
reaction product rich in thaumasite. White haloes of pure
thaumasite can be seen around dolomite aggregate particles.
The effect of temperature variation on the severity of
TSA has been explored in accelerated laboratory tests.
Concrete specimens, which showed no sulfate attack
when immersed in sulfate solutions at a normal
laboratory temperatures of around 20C were
progressively more severely affected by TSA when the
temperature was lowered below 15C.
In the field the significance of temperature on the
reactions is not so well understood. However, it is likely to
be significant since there is a variation within the critical
temperature band of 5 to 15C in near-surface ground. In
central and southern England, Meteorological Office data
[5]
indicate that the seasonal ground temperature variation
progressively decreases with depth, converging to a
value of about 10 - 12C. At shallow foundation depths
down to 1.2 m below ground level, the typical
temperature range is from a minimum of 4C in March to
maximum of 17C in September. At a depth of 3.0 m, the
temperature range is from a minimum of 8C in April to a
maximum of 12C in October. As in the laboratory, it is
likely that the extent of TSA will be increased at the
cooler temperatures if the chemical conditions are
satisfied. It is relevant to note, however, that minimum
ground temperatures below some types of construction
will be raised significantly above natural levels due to
heat loss from the building, and these sub-structures
may, therefore, be less prone to TSA.
B2.2 Acid attack
The acids most commonly encountered by concrete (all
found in some natural groundwaters) are carbonic acid,
humic acid and sulfuric acid. The first two are only
moderately aggressive and will not produce a pH below
about 3.5. Sulfuric acid is a highly ionised mineral acid
and may result in a pH lower than 2. Other similarly
aggressive mineral acids may occasionally be found in
ground contaminated by industrial processes.

The primary effect of any type of acid attack on concrete
is the dissolution of the cement paste matrix. This
weakens the affected concrete, but unlike sulfate attack,
the degradation does not involve significant expansion.
Neither ettringite nor thaumasite are stable in acid
solution so that the reaction product from sulfuric acid
attack will be primarily gypsum.
In concrete with siliceous gravel, granite or basalt
aggregate, the surface attack will produce an exposed
aggregate appearance. However, in concrete with
limestone (calcium carbonate) aggregates, the aggregate
may dissolve at a rate similar to that of the cement paste
and leave a smoother surface.
The rate of attack depends more on the rate of water
movement over the surface and on the quality of the
concrete, than on the type of cement or aggregate:
Acidic groundwaters that are not mobile appear to
have little effect on buried concrete.
Mildly acidic (pH above 5.5 ) mobile water will attack
concrete significantly, but the rate of attack will be
generally slow, particularly if the acids are primarily
organic in origin.
Flowing acidic water may cause rapid deterioration of
concrete, therefore high quality concrete is needed.
In the case of humic acid, reaction products formed on the
surface of concrete are mainly insoluble and tend to
impede further attack.
Several cases of acid attack on concrete in the UK are
described by Eglinton (1975)
[6]
. Occurrence of acidic
ground conditions is dealt with in Section C2.2 and
assessment of the ground conditions in relation to acidity
and mobility of water is included in Section C5.
B3. Other types of chemical attack on concrete
A large number of chemicals have been reported as
attacking concrete, albeit most in the longer term and /or
at high concentrations. For instance a USA document
[7]
lists more than 100 potentially destructive inorganic and
organic substances. The chance of encountering the vast
majority of these in the ground is remote, however, and
only the more likely ones are described in this Section or
referred to in the Sections on site investigation (Part C).
B3.1 Magnesium ions
Magnesium is a common element in soil and groundwater
but is generally only hazardous to concrete when the Mg
2+
cation is present in high concentrations in association with
certain other chemical agents the key one being sulfate
anions. Laboratory studies have found concretes made
with some cements are attacked to a greater degree by
high concentrations of magnesium sulfate (MgSO
4
) than
SD1:2005 Part B - Draft 1 v5 8/12/04 4
by equivalent concentrations of sodium sulfate. Because
of this effect, recommendations for concrete specification
in this and previous Digests have differentiated between
low and high magnesium levels when combined with high
sulfate concentrations. In practice the high Mg levels will
be found in the UK only in ground having industrial
residues. Other than the above, magnesium chloride
(MgCl
2
) is reported
[8]
to be especially aggressive.
The action of magnesium ions in concrete is complex, but
a key mechanism is the replacement of Ca in calcium-
silicate-hydrates that form much of the cement paste.
This leads to a loss of the binding properties. Formation
of brucite ( Mg(OH)
2

) and Mg-silicate hydrates is an
indication of attack.
The determination of magnesium ion content is a routine
part of site investigation for brownfield sites and is further
discussed in Section C5.1.2.
B3.2 Ammonium ions
Ammonium ions (NH
4
+
) will only be a problem to concrete
in ground having chemical residues left by man (including
in this case agriculture). Ammonium salts are reported
[8]
to act as cation-exchange compounds, transforming the
insoluble calcium in the hardened cement paste into
readily soluble calcium salts that are subsequently
leached away. During the reaction, ammonia is liberated
and escapes as a gas. The removal of both reaction
products results in an increase in the porosity of the
concrete, leaving it vulnerable to further attack.
Ammonium salts are also reported to act as weak acids
[8]
and neutralise the alkaline hardened cement paste; the
removal of the hydroxide ions results in softening and
gradual decrease in strength of the concrete.
In addition to the corrosive action of the ammonium ion,
some further deterioration may be caused by the action
of the associated anions. Ammonium sulfate, (NH
4
)
2
SO
4
,
is one of the most aggressive salts to concrete; cases of
attack have occurred resulting from spillage of the
material around fertiliser stores.
UK guidance is not available on the concentration of
ammonium ions that can be tolerated by different types of
concrete. However, EN 206-1: 2000 (Table 2) does
indicate that a level of NH
4
of 15 to 30 mg/l should be
regarded as slightly aggressive, 30 to 60 mg/l as
moderately aggressive, and greater than 60 mg/l highly
aggressive.
Because of the rarity of chemical attack attributed to
ammonium ions, assessment of ammonium concentration
is not specifically included in the scheme presented in this
document for assessment of ground aggressive to
concrete, or for guidance given for specification of
chemically resistant concrete. Specialist advice should be
sought if the presence of ammonium ions is suspected.
B3.3 Chloride ions
Chloride (Cl
-
) is a common anion in soil and groundwater,
in most cases being associated with sodium (NaCl is
common salt). However, the levels of chloride found in the
ground are generally chemically innocuous; indeed, they
may be beneficial since there is considerable evidence,
from seawater studies, that the presence of chloride
generally reduces sulfate attack. This is taken into account
for brackish water (12000 - 17000 mg/l chloride) in Note
(e) to Table C2.
No recommendations are given here for concrete exposed
to seawater (~18000 mg/l chloride). Reference should be
made to BS 6349: Part 1 for maritime structures and BS
8500-1.
While not generally causing chemical attack on concrete,
chlorides originating in the ground can lead to degradation
of concrete through a physical mechanism involving
crystallisation of chloride salts near to the surface of the
concrete; this is sometimes called salt weathering see
Section B6.
The risk of corrosion of embedded metals in buried
concrete in non-aggressive soil is generally lower than in
externally exposed concrete. However, high chloride
concentrations in the ground will increase the risk of
corrosion, since chloride ions may migrate into the
concrete and lead to a reduction in passivity at the metal
surface. The recommendations for the protection of steel
reinforcement given in BS 8500-1 should be followed.
On brownfield sites that have industrial residues, the
presence of chloride ions, together with a pH below 5.5,
could indicate the existence of hydrochloric acid that may
cause acid attack. It will, therefore, be important to
determine the amount of chloride in the soil and / or
groundwater during site investigation, as described in
Section C4. The procedure for taking account of the
SD1:2005 Part B - Draft 1 v5 8/12/04 5
measured chloride content in this particular circumstance
is given in Section C5.1.2.
Apart from this and the need to identify brackish and sea
waters, no account is taken of chloride concentration in
the procedure for concrete specification given in Section
D.
Specialist advice should be sought if exceptionally high
chloride levels are encountered, for example, related to
past industrial use of land. Such high concentrations
have been reported
[8]
as chemically affecting hardened
concrete. Detrimental mechanisms include the reaction of
calcium and magnesium chlorides with calcium aluminate
hydrates to form chloroaluminates which may result in
low-medium expansion of concrete.
B3.4 Chromium
A potential for attack on concrete from chromium has
been reported in the literature
[8]
. However, in practice
this would appear to be limited to concrete in contact with
chromite ore processing residues.
B3.5 Organic compounds
Phenols are the most commonly encountered
troublesome organic group. These are contaminants
typically generated as by-products during the
manufacture of town gas, tar and coke. The
concentrations present are rarely sufficient to attack
hardened concrete. However, their presence may well
affect the setting of concrete through an inhibition or
modification of the hydration of the cement
[9]
. Where in-
situ concrete is placed directly against ground suspected
of substantial contamination by phenols, consideration
should be given to the use of a barrier, such as
polyethythene sheeting, as protection during the setting
and hardening period.
It has been reported
[8]
that where phenol is present in
exceptionally high concentrations (eg several thousand
mg/l), it has the potential to attack hardened concrete.
The phenol is said to react with calcium hydroxide in the
cement paste to form calcium phenolate. This crystallises
in the pores of the concrete causing deterioration as a
result of physical expansion.
Some organic compounds are mildly acidic and affect
concrete as described in Section B2.2. In addition to
naturally occurring humic acid derived from decay of
organic matter, acids may also occasionally be produced
by activities of man, for example lactic acid, acetic acid
and butyric acids.
B4. Attack from aggressive carbon dioxide
Aggressive carbon dioxide is relevant only to certain
situations where water is continually flowing over (or
seeping through) the concrete. Diversion pipes or culverts
around dams retaining moorland waters containing high
concentrations of aggressive carbon dioxide can be
subject to erosion, as can poorly compacted concrete, or
permeable concrete products (for example some
aggregate concrete blocks) used in foundations.
Calcium carbonate ( CaCO
3
) is practically insoluble in
water but calcium bicarbonate (Ca(HCO
3
)
2
) is soluble.
Where carbon dioxide dissolves in water, carbonic acid
(H
2
CO
3
) is formed and this will react with any
carbonated cement (or limestone aggregate) to form
calcium bicarbonate that goes into solution. This is how
swallow holes and caves are formed in limestone. With
bicarbonate ions (HCO
3

-
) now in solution, a certain
amount of the dissolved carbon dioxide will be needed to
stabilise it, so there will be less available to attack further
carbonated concrete. It is this remaining available portion
of the dissolved carbon dioxide that is referred to as
aggressive. Measures to take account of aggressive
carbon dioxide for some uses of specific precast concrete
products are incorporated into guidance for specific
precast concrete products in Part F. These measures are
also relevant to cast-in-situ structures that are in contact
with flowing water containing aggressive carbon dioxide
(see Note d in Table C1).
B5. Attack from pure water
Pure (soft) water, which contains low concentrations of
dissolved ions is aggressive when it flows in quantity over
a concrete surface. Concrete surfaces that are carbonated
are less prone to this form of attack.
B6. Damage to concrete from crystallisation of
salts
In addition to causing chemical attack on concrete, soluble
compounds originating in the ground can lead to
degradation of concrete through a physical mechanism
involving crystallisation of salts, usually sulfates or
SD1:2005 Part B - Draft 1 v5 8/12/04 6
chlorides, near to the concrete surface. A classic
scenario for this is where concrete of high permeability is
partly buried in wet sulfate or chloride-bearing ground
and partly exposed to air. Sulfates or chlorides in solution
may be drawn through the concrete by capillary suction
to evaporate at or close to the free surface.
Crystallisation of salts in pores close to the surface of the
concrete may generate expansive stresses that disrupt
the concrete, while surface salt deposits form a
characteristic efflorescence. The process may be
aggravated by repeated wetting and drying of the
exposed concrete surface; this leads to cyclical salt
precipitation and dissolution and fatigue stressing of the
concrete fabric.
Additionally, where crystallisation initially occurs at a
relatively high temperature producing an anhydrous salt,
subsequent wetting may lead to conversion to a hydrous
crystalline form of substantially greater volume. A salt
particularly implicated in this latter mechanism is sodium
sulfate which, subjected to alternate wetting and drying,
may alternate between anhydrous thenardite (Na
2
SO
4
)
and hydrous mirabilite (Na
2
SO
4
.10H
2
O), with a change in
crystalline volume of some 300% and a potentially large
cyclical stress change. A comprehensive discussion of
the topic is included in Sulfate attack on concrete
[3].
In the UK, degradation of partly buried concrete due to
crystallisation of salts originating from the ground is rarely
a problem. . For most ground conditions, the measures
recommended here to mitigate chemical attack on
concrete (and in particular specified free water/cement
ratios of 0.5 or less) should also be effective against
physical degradation due to crystallisation of likely salts.
Further guidance for extreme ground conditions in arid
areas is given CIRIA Report C577
[10].

B7. Microbial contribution to chemical attack on
concrete
Activity of micro-organisms in the ground can result in
changes to the chemical environment that can contribute
indirectly to concrete attack. The most widely recognised
damage of bacterial origin is the deterioration of concrete
in sewers caused by sulfate-reducing bacteria which feed
on sulfate in the effluent. Sulfuric acid is produced which
attacks the concrete. In contrast, sulfate-oxidising bacteria
such as thiobacillus ferroxidans aid in oxidation of pyrite
(FeS
2
) in the ground, producing both sulfuric acid and
sulfates that subsequently lead to sulfate attack of
concrete. The need to take pyrite oxidation into account
where pyritic soils will be disturbed by construction is
discussed in Section C5.1.3.



SD1:2005 Part B - Draft 1 v5 8/12/04 7
References Part B
[1] Department of Environment, Transport and the Regions. The thaumasite form
of sulfate attack: Risks, diagnosis, remedial works and guidance on new construction.
Report of the Thaumasite Expert Group. DETR, January 1999.
[2] Marchand J and Skalny J P. Materials Science of Concrete - Special Volume: Sulfate Attack Mechanisms.,
American Ceramic Society. Westerville, Ohio, USA. 1999.
[3] Skalny J P and Marchand J. Sulfate Attack on Concrete. Spon Press. London. 2002.
[4] Building Research Establishment. Proceedings of First International Conference on Thaumasite in Cementitious
Materials. BRE, Garston June 2002. CRC, London, 2002.
[5] Meteorological Office. Averages of earth temperature at depths of 30 cm and 122 cm for the United Kingdom
1931-1960.Report MO 794. HMSO, 1968.
[6] Eglinton M S. Review of concrete behaviour in acidic soils and groundwaters. Technical Note 69. CIRIA, London.
1975.
[7] Portland Cement Association. Effects of substances on concrete and guide to protective treatments. Concrete
Information IS001. PCA. Skokie, Illinois. 2001. (can be web-downloaded from www.cement.org).
[8] Environment Agency. Risks of contaminated land to buildings, building materials and services: A literature review,
R&D Technical Report P331, Environment Agency, Swindon, 2000.
[9] Paul, V. Performance of building materials in contaminated land. BRE Report BR 255. Garston, BRE, 1994.
[10] Walker M. Guide to the construction of reinforced concrete in the Arabian Peninsular. Report C577. CIRIA,
London, 2002.
British Standards Institution
BS EN 206-1: 2000 Concrete: Part 1: Specification, performance, production and conformity.
BS 8500-1: Concrete Complementary British Standard to BS EN 206-1 Part 1: Method of specifying and guidance
for the specifier
BS 6349-1: Code of practice for maritime structures: Part 1:2000 General criteria.