33/1

Corrosion

33.1 Corrosion basics
33.1.1 Definitions of corrosion
Corrosion is generally taken to be the wastage of a
metal by the action of corrosive agents. However, a
wider definition is the degradation of a material through
contact with its environment. Thus corrosion can include
non-metallic materials such as concrete and plastics and
mechanisms such as cracking in addition to wastage (i.e.
loss of material). This chapter is primarily concerned with
metallic corrosion, through a variety of mechanisms.
In essence, the corrosion of metals is an electron
transfer reaction. An uncharged metal atom loses one or
more electrons and becomes a charged metal ion:
M ~ M + + electron
In an ionizing solvent the metal ion initially goes into
solution but may then undergo a secondary reaction,
combining with other ions present in the environment
to form an insoluble molecular species such as rust or
aluminium oxide. In high-temperature oxidation the metal
ion becomes part of the lattice of the oxide formed.
33.1.2 Electrochemical corrosion
The most important mechanism involved in the corrosion
of metal is electrochemical dissolution. This is the basis
of general metal loss, pitting corrosion, microbiologically
induced corrosion and some aspects of stress corrosion
cracking. Corrosion in aqueous systems and other cir-
cumstances where an electrolyte is present is generally
electrochemical in nature. Other mechanisms operate in
the absence of electrolyte, and some are discussed in
Section 33.1.4.
Figure 33.1 depicts a metal such as iron, steel or zinc
immersed in electrolyte such as sodium chloride solution.
M+
Aqueous environment
Metal \ //c
M ~
Electrons
Anode
(a)
M-I-
H2
Aqueous environment
\ Cat,o
M ~ ~
Electrons
Anode
(b)
Figure 33.1
Corrosion basics 33/3
The fundamental driving force of the corrosion reaction
is the difference in the potential energies of the metal
atom in the solid state and the product which is formed
during corrosion. Thus corrosion may be considered to be
the reverse of extractive metallurgy. Metals are obtained
by the expenditure of energy on their ores. The greater
the energy that is required, the more thermodynamically
unstable is the metal and the greater its tendency to revert
to the ore or any other oxidized form. Thus gold is found
native (i.e. as the metal) and it resists corrosion. Iron,
aluminium and most metals exist naturally combined as
oxides or sulphides and, once reduced to the metallic state,
they attempt to revert to the combined state by corrosion.
In the systems illustrated in Figure 33.1 the anodic
reaction has to be electrically balanced by the cathodic
reaction, since electrical charge cannot build up at any
location. A continuous electrical circuit is required through
the metal (for electron conduction) and the environment
(for ionic conduction).
There are four main aspects to the corrosion chemistry
of this solution:
1. The presence of dissolved ions (an electrolyte) facili-
tates the passage of an ionic current through the solu-
tion.
2. Certain anions, especially chloride, penetrate the pro-
tective films which are naturally present on some met-
als (e.g. aluminium and its alloys and stainless steels).
This process is an initiator for corrosion, especially for
localized corrosion.
3. The pH of the solution is a measure of the availability
of one of the two most important cathodic agents, H +
(the other being dissolved oxygen, see below). It is
also largely responsible for determining whether the
metal surface will be active or passive (see below).
4. Dissolved oxygen (and even water) can passivate cer-
tain metals such as titanium, aluminium and stainless
02
e
j H+
e
33/4 Corrosion
steels, thereby rendering them, to varying degrees, cor-
rosion resistant. Because oxygen is a strong cathodic
agent it stimulates corrosion of many metals, especially
those such as iron and zinc, which tend not to form
stable, adherent oxides.
Thus in the system depicted in Figure 33.1, for any
metal which is thermodynamically unstable with respect
to its dissolved ions in the solution (and this includes most
metals of industrial importance) corrosion will occur. The
rate of corrosion is determined by several factors:
1. The conductivity of the environment; low conductivity
hinders the ionic current flow hence distilled water is
less corrosive than a solution of sodium chloride with
the same pH and dissolved oxygen content.
2. The presence of cathodic agents; dissolved oxygen and
H + promote corrosion of most metals in an electrolyte
solution (see Figure 33.2).
3. The supply of cathodic agents to the metal surface:
stirring a solution or increasing the flow rate in a
pipe will increase corrosion rates in systems which
are under diffusion control. As soon as corrosion
initiates in the system depicted in Figure 33.1 there
is a local depletion of cathodic agent and a build-up
of metal ions. In stagnant conditions this soon reduces
the corrosion rate, and the diffusion of these species
governs the overall corrosion rate. The system is then
under diffusion control.
4. The build-up of scales or other deposits on the metal
surface can stifle corrosion. The reaction of anodi-
cally generated M + with cathodically generated OH-
can, depending on M +, be an insoluble product. It
may or may not adhere to the metal surface. If it
does adhere strongly it can stifle corrosion. Other
deposits can form, in addition to corrosion products.
The locally high pH at the cathode can precipitate
insoluble carbonates. With complete coverage of the
cathodic surface corrosion is generally stifled, since
the cathodic reaction cannot usually occur through the
deposit. However, some cathodic scales (e.g. magnetite
(Fe304)) and some forms of iron sulphides can support
the cathodic reaction, and, since anodic area is limited,
this can result in intense local attack leading to pitting
corrosion at bare areas.
5. The effect of temperature is generally to increase
corrosion rate as temperature rises. This is a chemical-
kinetic effect and can sometimes be overridden by
\ 33.1.2.1 Pitting corrosion
\ Aluminium
\\ Zinc
\
\ Steel
\
\ attacked. Rates of pitting penetration can be very high;
\
\ type 316 stainless steel in warm sea water can suffer pit
\\
2 4 6 8 10
F i g u r e 3 3 . 2 V a r i a t i o n in c o r r o s i o n rate w i t h pH
physical effects such as deposition of scales. The
solubility of oxygen in a saline solution decreases
from approximately 8 - 1 0 ppm at 5°C to virtually zero
near the boiling point. Thus in a system in which
dissolved oxygen is the only cathodic agent corrosion
rates would pass through a maximum and decrease to
very low levels as temperature is increased and the
availability of cathodic agent is diminished.
In Figure 33.1 the region from which metal is lost is
the anode. The atoms of the metal are ionized and release
electrons:
M > M n+ + ne-
where M denotes metal, M n+ its ion and e- is an electron.
This reaction can proceed to a significant extent only
if the electrons which are released are consumed by a
cathodic reaction. The commonest cathodic reactions are
essentially:
02 + 2H20-t- 2e- > 4OH-
as shown in Figure 33.1(a), in which oxygen is con-
sumed, and
H + + e- > H atoms > H2 gas
in which hydrogen is liberated (Figure 33.1(b)).
Which cathodic reaction is preferred for any combi-
nation of metal and environment depends on the relative
amounts of 02 and H + available, and kinetics of 02 reduc-
tion and H + reduction on the metal surface.
The thermodynamic driving force behind the corrosion
process can be related to the corrosion potential adopted
by the metal while it is corroding. The corrosion potential
is measured against a standard reference electrode. For sea
water the corrosion potentials of a number of constructional
materials are shown in Table 33.1. The listing ranks metals
in their thermodynamic ability to corrode. Corrosion rates
are governed by additional factors as described above.
Stainless steels each appear twice in the list. The more
active potentials are those which the metal adopts when
corroding as in a pit. The more cathodic potential is
that adopted by the bare surface around the pit. The
potential difference constitutes a significant driving force,
analogous to the situation where the coupling of dissimilar
metals such as copper and iron promotes the corrosion of
the more anodic of the two (see below).
Pitting is a form of localized corrosion in which part
of a metal surface (perhaps 1% of the exposed area) is
penetration rates of 10mm per year. This is a natural
// consequence of the low ratio of anodic area to cathodic
°
/ area, coupled with the self-acidification processes that
occur in some metals.
v
Figure 33.3 illustrates a pit in a stainless steel such
12 pH
as type 304 or 316 austenitic alloy. Pitting starts at a
heterogeneity in the steel surface, such as an outcropping
 Corrosion basics 33/5

Table 33.1 Galvanic series in sea water a The second phase of pitting corrosion is propagation.
Because of the geometry of the small pit that has been
Metal Corrosion created, oxygen diffusion to the corroding region is hin-
potential b (V) dered. This creates a potential difference between the pit
and the surrounding region. The solution in the pit acidi-
Magnesium and alloys -1.6 fies through hydrolysis of the Fe z+ and Cr 2+ ions present.
Zinc - 1.0 The pH in the pit falls to very low values (0.6 in the
Aluminum alloys -0.8/- 1.0
Steel (incl. low alloys) - 0 . 5 5 / - 0.75 case of 316 stainless steel), irrespective of the bulk solu-
Stainless steels (304, 321,410, 430) -0.5 c tion conditions, which may be neutral or even alkaline.
Stainless steels (316, 317) -0.4 c This enhances corrosion by dissolving the pit walls and
Copper -0.35/-0.2 ensures that the pit walls cannot be repassivated. The sur-
Bronzes (Si, Sn) -0.35/-0.25 rounding metal surface is passive (i.e. is covered with an
Stainless steels (410, 430) -0.3/-0.2 d adherent oxide film of mixed Fe203. Cr203 oxide). This
Lead -0.25 e is dissolved in strongly acidic conditions such as those
Copper nickel alloys -0.25 prevailing in the pit. Because the conditions in the pit pro-
Bronze (Ni-A1) -0.15 mote self-acidification, pit penetration rates often increase
Nickel -0.15
with time.
Stainless steels (304, 321,316, 317) -0.1 - 0 d
Titanium 0/-0.1 A third phase is sometimes identified in pitting cor-
Platinum, graphite +0.2/0.3 rosion, i.e. termination. Pits can become stifled by the
build-up of insoluble corrosion products at their mouths.
aThese values are roughly constant across a range of electrolyte Removal of these mounds of corrosion products, either
environments except where noted but the variations between mechanically or through some change in the environmen-
alloys, heat treatment conditions, etc. creates a range for each tal chemistry, can allow the pits to restart growth.
metal. For some metals such as iron and steel the range is low
(+100 mV), but for lead, nickel, stainless steels a range is given.
bThe corrosion potential is reported with respect to the saturated 33.1.2.2 Crevice corrosion
calomel reference electrode.
cWhen active, as in a pit or a crevice or when depassivated by Crevice corrosion of stainless steels (Figure 33.4(a)) has
mechanical damage of oxide film or chemical removal in non- a similar mechanism to pitting corrosion. The initiation
oxidizing acid.
dWhen passive, on a bare surface around a pit or crevice in normal phase is assisted by the creation of a crevice of suitable
conditions. geometry. Crevices are formed by certain fabricational
e-0.5V in environments low in chloride ion content. processes including riveted seams, incompletely fused
welds, interference fits, O-rings, gasketed joints and even
sulphide inclusion, the shielded region beneath a deposit paint markings on components. The solution in the crevice
or even a discontinuity in the naturally present oxide film becomes acidified and locally the metal surface anodic
caused by a scratch or embedded particle of abrasive grit. compared with the surrounding bare metal, as described
This initiation phase of pitting corrosion may take sec- above. Crevice initiation is largely dependent on the
onds or weeks to develop into a feature large enough to geometry of the crevice that is present: shorter initiation
be recognized as a pit. Initiation is facilitated by oxidizing times are brought about by tighter crevices, since the
conditions. Thus hypochlorite ions are particularly aggres- primary effect is the limitation on inward diffusion of
sive towards stainless steels since they raise the potential oxygen to maintain passivity. The lower limit of crevice
of the steel to such a high level that the protective oxide width that will permit crevice corrosion is the point at
layer breaks down, allowing localized attack. which the crevice is watertight.

Fe2+

OH- 02
Aqueous environments
containing chloride ions
and dissolved oxygen

Stainless steel / "~ t Surface of stainless steel

~ ]e is normally passive, i.e
[ _ \ / - covered with natural oxide
Fe i = Fe2+ Cl- FeZ+ ~,-~/Fe film

\ Fe 2+ /
} ~ Walls of pit are depassivated, i.e
bare of protective oxide film
Hemispherical shape of pit
is often distorted by
microstructural effects

Figure 33.3 Pitting corrosion of stainless steel

33/6 Corrosion
Aqueous environment
Metals such as Cathode
stainless steel 0 2--~OH-
Aqueous environment
Copper alloy Cathode
2e-+Cu2+ --~ Cu
Figure 33.4 Crevice corrosion driven by (a) a differential aeration cell and (b) a differential metal ion concentration cell
Crevice corrosion of copper alloys is similar in princi-
ple to that of stainless steels, but a differential metal ion
concentration cell (Figure 33.4(b)) is set up in place of the
differential oxygen concentration cell. The copper in the
crevice is corroded, forming Cu + ions. These diffuse out
of the crevice, to maintain overall electrical neutrality, and
are oxidized to Cu 2+ ions. These are strongly oxidizing
and constitute the cathodic agent, being reduced to Cu +
ions at the cathodic site outside the crevice. Acidification
of the crevice solution does not occur in this system.
33.1.2.3 Galvanic corrosion
Galvanic corrosion is the enhanced corrosion of one metal
by contact with a more noble metal (or a conductive
non-metal such as graphite). The two metals require only
to be in electrical contact with each other and exposed
to the same electrolyte bnvironment. By virtue of the
potential difference that exists between the two metals,
the potential of the less noble (more reactive metal) is
raised. This stimulates corrosion of the more reactive
metal, simultaneously reducing the corrosion rate of the
less reactive metal. This principle is exploited in the
cathodic protection (Section 3317.2) of steel structures by
the sacrificial loss of aluminium or zinc anodes.
The metal that corrodes in any couple is that which
has the most negative corrosion potential in the galvanic
series in that environment. As a guidance, Table 33.1 is
of general but not universal applicability. Thus in the case
of copper and aluminium the aluminium corrodes unless
corrosion inhibitors are employed. Copper and aluminium
are therefore not generally used in close contact with
each other, but copper ions in solution from the corrosion
of copper pipework can plate out onto aluminium items
(including saucepans) downstream in the system. Rapid
corrosion of the aluminium occurs in the region of the
copper deposits. The conductivity of the environment
and the potential difference between the metals involved
Metal or non-metal
Width of crevice
~ Anode typically 0.01-0.5mm
Fe2++H20-'- H+ Fe--~ Fe2++2e-
Cr ----Cr2++2e-
(a)
Metal or non-metal
Width of crevice
~ A n o d e typically
Cu--~Cu++e - 0.01-0.5 mm
(b)
determine whether the corrosion is localized to a small
region at the interface between the metals, or spread
across a larger region. The exposed areas of the two metals
together with the length of the conductance path in the
electrolyte determine the intensity of the attack. Other
conditions, such as the flow regime and the polarization
behaviour of the two metals, influence galvanic corrosion.
Potential reversal is a complicating factor in galvanic
corrosion. Zinc is usually anodic to steel and hence it
is used to protect steel sacrificially. Galvanized steel is
steel coated with zinc. The zinc protects any steel that
is exposed (such as cut ends of sheet) and scratches
through the zinc layer. At temperatures above 80°C,
in low-conductivity water (e.g. condensation) the metals
undergo a reversal of polarity. The steel corrodes in
preference to the zinc, and this causes rapid loss of the
zinc, which can fall off in sheets as corrosion proceeds
along the steel surface undermining the zinc. Tin is
usually more corrosion resistant than steel, hence its use in
tinplate (tin-plated steel). In some organic acids, including
some of those encountered in fruit juices, the relative
potentials of tin and steel are reversed and the tin corrodes,
contaminating foodstuffs with tin corrosion products.
The corrosion potentials of metals and alloys are tem-
perature dependent. The potential of a steel distillation
column, heat exchanger or other item of plant varies in
the temperature zones in which it is operating. Under
some circumstances this creates galvanic corrosion cells
between different areas of one item fabricated from one
material, leading to localized galvanic corrosion problems.
To avoid galvanic problems, different materials of con-
struction may have to be electrically isolated or at least
the electrical resistance between them increased to a level
sufficient to reduce the corrosion current to an acceptable
value. In some instances it is more practical to paint or
otherwise coat the more cathodic of the two parts of the
couple. The anodic material should not be coated, since
 Corrosion basics 33/7

even more rapid penetration would occur at any breaks in (SCC) if the conditions are sufficiently severe. The tensile
the coating. Guidance on the galvanic hazards posed by stress can be the residual stress in a fabricated structure,
the contact of different materials of construction is given the hoop stress in a pipe containing fluid at pressures
in the British Standard publication PD 6484. above ambient or in a vessel by virtue of the internal
hydraulic pressure created by the weight of its contents.
33.1.2.4 Microbially induced corrosion Stresses result from thermal expansion effects, the tor-
sional stresses on a pump or agitator shaft and many more
Various bacteria are involved in corrosion processes. Of causes.
these the most important are some Thiobacillus types The corrosive environments which cause SCC in any
(these generate sulphuric acid which dissolves concrete) material are fairly specific, and the more common com-
and sulphate-reducing bacteria (SRB). SRB are; ubiqui- binations are listed in Table 33.2. In the case of chloride
tous but only create significant corrosion problems in stress corrosion cracking of the 330 series austenitic stain-
conditions suitable for rapid growth. They tIu:ive best less steels it is generally considered that the risk is min-
in anaerobic conditions, with a supply of sulphate ions imal below approximately 60°C. Above this threshold, if
and a carbon source. Anaerobic conditions are created in the chloride ion content of the environment is low a higher
stagnant areas of chemical plant, under slime Jn indus- tensile stress can be tolerated without SCC developing,
trial cooling systems and soils and seabed muds. Sul- and vice versa. It is difficult to design for the control of
phate ions are widely available, especially in the sea
and some chemical processes. Suitable carbon sources
include organic materials such as acetic acid. SRB pro- Table 33.2 Stress corrosion cracking
duce sulphide ions from sulphate ions. In conditions such
as beneath a colony of these bacteria corrosion processes Metal Agent
as illustrated in Figure 33.5 occur. Many metals are sus-
ceptible to this form ofattack, including steel, stainless Steels Nitrate ion
steels, copper alloys and aluminium alloys. Localized pit- Strong alkali
Carbonate/bicarbonate
ting rates in excess of 10mrn/yr are possible on steel Liquid ammoniaa
and low-alloyed stainless steels such as types 304 and Hydrogen sulphide (aqueous)b
316 austenitic and 22% Cr duplex. Good resistance has Cyanide ion
been demonstrated for 25% Cr superduplex and some of Austenitic stainless Chloride (and other halide) ion
the higher grades of superaustenitic stainless steels. Some steels Polythionic acids
Cladosporium fungi cause pitting attack on aluminium High-temperature water
alloys. Gallionella bacteria thrive in solutions containing Hydrogen sulphide (aqueous)
Fe 2+ and produce deposits under which anaerobic pro- Copper alloys Ammonia/amines
cesses can occur. Nitric acid/nitrate fumes
Nickel alloys Strong alkali
Polythionic acids
33.1.3 Cracking mechanisms Chloride ion
Titanium alloys Alcoholsa,b
Cracking mechanisms in which corrosion :is impli- Nitric acid plus nitrogen oxides
cated include stress corrosion cracking, corrosion fatigue, Magnesium alloys Chloride ionc
hydrogen-induced cracking and liquid metal embrittle- Aluminium alloys Chloride ionc
ment. Purely mechanical forms of cracking such as brittle Organics
failure and fatigue are not considered here. High-purity hot water

Environments are aqueous unless specified otherwise.

33.1.3.1 Stress corrosion cracking alnhibited by 1000 ppm of water.
bSometimes considered to be a form of hydrogen-induced
The conjoint action of a tensile stress and a specific cor- cracking.
rodent on a material results in stress corrosion cracking Cln oxidizing conditions.

HS- H2S

Anaerobic environment
containing H2S

Iron sulphide

Steel L~ ~ - - - ~ e ~ Cathode

Electrons
Anode

Figure 33.5
33/8 Corrosion
SCC since it is the local stress and corrodent concentration
rather than average or bulk values that are relevant.
The essential feature of most SCC mechanisms is the
rupture of the protective film on the metal (passive oxide
layer in the case of stainless steels and aluminium alloys)
which allows the corrodent to reach the bare surface of the
metal. Cracking is the result of intense corrosion, concen-
trated by the tensile stress. In austenitic stainless steels,
chloride SCC is transgranular in nature and the cracks
are typically high branched. In ferritic materials cracking
is largely due to a hydrogen embrittlement mechanism
brought about by hydrogen created during corrosion.
33.1.3.2 Corrosion fatigue
Many metals suffer fatigue, which is the result of the
application of an alternating stress. Corrosion fatigue
results from the exposure of a metal, which is sub-
jected to alternating stress, in a corrosive environment.
Unlike stress corrosion cracking, corrosion fatigue does
not require a specific corrosive environment for any metal.
Moist air is adequate in many cases, thus the majority of
fatigue failures are actually corrosion fatigue in nature.
The alternating stress pattern requires only a small tensile
component. This tensile stress may be small relative to the
metal's ultimate tensile strength and millions of cycles of
the stress pattern may be required to cause failure.
33.1.3.3 Hydrogen-induced cracking
Hydrogen-induced cracking (HIC) is most commonly
encountered in steels but other metals are susceptible, as
shown in Table 33.3. The presence of hydrogen atoms in
a metal degrades some of its mechanical properties, espe-
cially its ductility, leading in some cases to embrittlement.
Additionally, hydrogen atoms diffuse through metals and
coalesce to form hydrogen molecules at certain preferred
locations such as inclusions.
Hydrogen atoms are soluble in ferrite (which is the
major phase of most steels). At discrete inclusions in the
Table 33.3 Hydrogen-induced cracking
Metal Agent
Steel Hydrogen gas
Atomic hydrogena
Hydrogen sulphide-containing
aqueous environments b
Waterc
Martensitic Hydrogen sulphide-containing
stainless steels aqueous environments
Copper alloys Sea waterd
Nickel alloys Hydrogen sulphide-containing
aqueous environments
Titanium alloys Anhydrous alcohols
Zirconium alloys High-temperature water
aCreated by overprotection during cathodic protection,
evolved during electroplating or pickling processes and
during welding with wet welding consumables.
bSteels of strength in excess of 550 MPa.
cSteels of strength in excess of 900 MPa.
dHigh-strength alloys, cathodically protected.
steel (e.g. manganese sulphide, which is present in many
steels) the hydrogen atoms combine to form hydrogen
molecules. Hydrogen molecules are insoluble in the steel
lattice and these molecules associate, producing high ten-
sile stresses which can initiate hydrogen cracking. When
laminations are present in the steel hydrogen molecules
form at the interface between them, creating blisters. At
high temperatures hydrogen atoms can react with carbides
in the steel producing methane gas. Methane is insoluble
in the steel structure and causes blistering or cracking. In
clean steels containing none of these features the hydrogen
atoms pass straight through the steel, causing no damage.
Hydrogen can originate from several sources:
1. Electroplating (e.g. zinc, cadmium or hard chromium
plating);
2. Acid pickling (e.g. prior to galvanizing);
3. Cathodic protection, especially if overprotected,
through inadequate potential control;
4. Corrosion, where H + reduction is a cathodic reaction
(Section 33.1.2);
5. The process inside the plant.
The amount of the hydrogen that is liberated on or near a
metal surface which then enters the metal varies according
to the environment and condition of the metal. The main
factor that promotes the entry of hydrogen into a metal
is the presence on the metal of a surface poison such
as sulphide or other species which inhibit the hydrogen
recombination reaction.
33.1.3.4 Liquid metal embrittlement
This is a form of stress corrosion cracking in which the
corrodent is a liquid metal. Mercury at ambient temper-
ature and metals including zinc (from galvanized steel-
work) and copper (from electric cables) when melted
during welding or in a fire cause rapid failure of certain
metals.
Steels and austenitic stainless steels are susceptible to
molten zinc, copper, lead and other metals. Aluminium
and copper alloys are attacked by molten mercury, zinc
and lead. Nickel alloys are attacked by mercury, zinc,
silver and others. Other low-melting point metals which
can attack common constructional materials include tin,
cadmium, lithium, indium, sodium and gallium.
33.1.4 Non-electrochemical corrosion
Although all the types of corrosion discussed in this
section result in the oxidation of metal and some involve
direct electron transfer, they can be described without
reference to electrochemistry.
33.1.4.1 Oxidation
In oxidizing atmospheres (i.e. in the presence of oxygen
or a source from which oxygen can be derived, such as
water vapour) most metals are unstable with respect to
their oxides. Oxidation to create the metal oxide depends
on several factors, the most important being temperature

Table 33.4 Oxidation in air
Steels 480°C
Low-alloy steels 560°C
Ferritic stainless steel (type 410) 650°C
Austenitic stainless steel (type 316) 900°C
Nickel-based superalloys 1 IO0°C
Superaustenitic stainless steel (type 310) 1150°C
Cobalt-based superalloys > 1150°C
1. Temperatures the those above which oxidation is generally too
rapid to permit the metal's safe or economic use.
2. Atmospheres contaminated with acid or halogen gases reduce
these values.
3. The mechanical properties including strength, ductility and
creep resistance can be affected below these temperature.
(and hence the thermodynamics of the process), availabil-
ity of oxygen and the protective action of any oxide that
is created. Oxygen can migrate through the oxide, where
it reacts with the metal. In other metal-oxide systems,
metal ions migrate outwards through the oxide and form
oxides on the outer surface. The structure and the stability
of the oxide, together with its semiconducting properties,
control the mechanism of oxidation. Oxidation rates can
be linear, parabolic (tending to a steady state oxide thick-
ness) or exponential (termed catastrophic, because the rate
increases with time). Table 33.4 gives an indication of the
temperature limits for some metals through oxidation in
a clean atmosphere.
Sulphidation is analogous, but catastrophic sulphida-
tion is common because of the generally lower melting
points of sulphides than the corresponding oxide. This
is especially true in the case of nickel alloys, when a
nickel/nickel sulphide eutectic is formed.
33.1.4.2 Fretting corrosion
Fretting corrosion occurs as a consequence of oscillating
relative motion between touching surfaces. As little as
3 x 10-9m lateral movement is required. The amount of
damage increases as the normal force between the surfaces
is increased. In dry conditions the corrosion product is
usually the oxide. In the case of steel this is wustite.
This is otherwise the high-temperature form of the oxide,
which implies that locally high temperatures are created
on the fretting surfaces. The surfaces weld together in the
high stress conditions at points of contact and are torn
apart by the relative motion of fretting surfaces.
Materials with hard oxides, including stainless steels
and aluminium and titanium alloys, are particularly sus-
ceptible to this form of attack. In steel it is also known
as 'false Brinelling' because of the high surface hardness
that can be created in work-hardening grades.
33.1.4.3 Molten salt corrosion
Any salt that is present on a hot metal surface can cause
corrosion of that surface. This mechanism is often rapid
and is due to straightforward dissolution of the metal and
any oxide which may be present on the surface. The
mechanism is similar to that of aqueous corrosion. The
high rate of attack is a consequence of the high activities
The implications of corrosion 33/9
of the ions present in the molten salt. The problem is
greatest for salts of low melting point, since these are
present over a wide temperature range and can be very
fluid. Low melting-point salts include many chlorides,
sulphides and sulphates.
33.2 The implications of corrosion
33.2.1 Economics
The principal economic implications of corrosion of a
plant are the initial cost of construction, the cost of
maintenance and replacement, and the loss of production
through unplanned shutdowns. The initial cost of the
plant is influenced by material selection, and a choice
of material which is more corrosion resistant than is
necessary for the safe operation of the plant over its design
life is a very expensive error. This cost involves initial
outlay of money, and plants have been built which could
never be profitable because of the inappropriate materials
selection.
Maintenance costs arising from corrosion can also
prove to be unacceptably high. For plant which requires
regular maintenance for other reasons (e.g. desludging or
batch operation) it is often accepted that certain items are
expendable and annual or even monthly repair or replace-
ment with low-cost materials is preferable to the use of
expensive ones. Plant handling hot, strong hydrochloric
acid is often regarded in this way, since the only possible
construction materials which have acceptably low corro-
sion rates are too expensive to use. Plant which requires
to be run continuously or maintenance-free for long peri-
ods necessitates high-integrity design and hence the use
of more reliable corrosion-resistant materials. The most
appropriate approach to studying the economic implica-
tions for several alternative materials of construction is
considered to be that of life cycle costing. This requires
an estimation of interest rates over the design life of the
plant.
Significant savings can be achieved by optimum mate-
rial selection (guidance is given in Section 33.3), by
considering corrosion at the design stage (Section 33.4),
by employing corrosion-control techniques (Section 33.7)
and corrosion monitoring (Section 33.8). In the industrial-
ized countries of the Western world, corrosion is estimated
to cost some 4% of each country's GNP. In the UK this
represents £12 billion per annum, of which a significant
proportion could be saved by the use of existing knowl-
edge. It cannot all be saved, since the gradual deterioration
of a plant over its operational life is one of the costs
incurred in the process.
33.2.2 Safety
The failure of plant by corrosion can be gradual or catas-
trophic. Gradual failure has few implications for safety
providing it is monitored. Direct corrosion-monitoring
techniques are described in Section 33.8. Indirectly, the
correct interpretation of records relating to metal con-
tamination of products or the loss of efficiency of heat
exchangers, etc. can provide useful information.
33/10 Corrosion
Sudden or catastrophic failure of plant through corro-
sion can result in the loss of product at high velocity
from a failed reactor vessel, high-energy steam from a
steam line, toxic or inflammable materials from storage
vessels. Incidents of this nature from plants around the
world have led to the death and injury of plant opera-
tives and nearby householders. The majority of serious
incidents in the chemical process industry, however, have
been attributed to mechanical failure of plant.
33.2.3 Contamination of product
The metal lost from the inside of pumps, reaction vessels,
pipework, etc. usually contaminates the product. The
implications of this depend upon the product. White
plastics can be discoloured by ppb levels of iron, though at
this level the effect is purely cosmetic. The taste of beer is
affected by ppm levels of iron and other metals. Products
sold to compositional requirements (such as reagent-grade
acids) can be spoiled by metal pick-up. Pharmaceutical
products for human use are often white tablets or powders
and are easily discoloured by slight contamination by
corrosion products.
33.3 Materials selection
The materials selection procedure for new or replacement
plant is crucial to the safe and economic operation of that
plant. There is no one correct way to select the appropriate
construction material since for small plant handling toxic
or inflammable products the integrity of the plant is high
in priority, whereas large plant producing bulk material in
a competitive market is more likely to be made of cheaper
constructional materials, and the occasional leak or failure
may have fewer safety implications.
Wherever possible, items should be made from mate-
rials which have proved themselves in similar service.
Essential items of plant for which long delivery times are
expected should be considered most carefully. The pos-
sible interactions between items fabricated from different
materials should be considered. Section 33.1.2 discusses
galvanic corrosion.
33.3.1 Sources of information
The first source of information for the behaviour of a
material in the proposed service environment is the poten-
tial supplier of the item of plant. Except for new (or
significantly modified) processes, specialist suppliers or
fabricators have relevant information and service experi-
ence. The supplier should be provided with all process
or environmental details that are of possible relevant to
corrosion. The most important are listed below:
1. Environmental chemistry: including complete typical
and worst-case compositional analyses, pH, redox
potential, dissolved gas content;
2. Physical conditions: temperature ranges in bulk and at
surface, heat transfer, mass flow, fluid velocities, the
presence of entrained particles and gases;
3. Operating cycles: commissioning procedures, heating,
cooling and pressure cycling, stagnant periods, clean-
ing procedures;
4. External environment: corrosive gases, humidity, tem-
perature;
5. Mechanical aspects: the presence of tensile or cyclic
stresses.
If the potential suppliers cannot provide documented ser-
vice performance for similar plant on similar duties, mate-
rials suppliers should be consulted.
For all materials other than basic constructional steels
and cast irons, reputable suppliers have information bases
and applications laboratories from which information
can be obtained. Trade organizations representing
categories of materials suppliers are excellent sources
of information; some are listed at the end of this
chapter. The materials suppliers should be consulted
in conjunction with equipment suppliers in order to
ensure that the information generated is fully applicable
to the end use to which the material is to be put.
Fabricational techniques should be agreed between the
two types of suppliers, since some materials cannot be
cast or welded and some items cannot be made by
forging.
For some new or modified processes it will not be
possible to obtain sufficient good information to permit
material selection to be made. Advise is available from
professional corrosion consultants and corrosion-testing
facilities are available at a number of laboratories in the
UK. The Institute of Corrosion publishes a directory of
laboratories and consultants.
The following guidance is intended as a checklist for
general use, since there is no universal, definitive guide
to the selection of materials. It is based on corrosion and
not on mechanical property requirements such as strength,
toughness, hardness or fatigue resistance.
33.3.2 Aqueous systems
33.3.2.1 Water supply
For the purpose of corrosion, water is scaling or non-
scaling. Scaling water tends to deposit generally protec-
tive hardness scales. Soft water does not scale and hence
is potentially more corrosive, especially when it contains
dissolved gases such as oxygen, carbon dioxide, ammonia
and sulphur oxides.
For drinking water, copper has now replaced the use
of lead, except in areas of cuprosolvency. The local
water authority provides information including analytical
data. Stainless steels are being introduced especially for
domestic systems. Experience on the Continent is mixed,
with some corrosion problems such as pitting and stress
corrosion cracking reported.
For process water, steel pipes are used unless iron
pick-up is to be minimized. Plastic pipes (polyethylene
and polyvinylchloride) are used but they sometimes need
external protection from solvents present in industrial
atmospheres, ultraviolet radiation (including sunlight),
freezing and mechanical damage.

For firefighting water, steel pipes are used but corro-
sion products can block sprinklers. Cement asbestos pipes
are utilized but pressure limitations restrict their use. For
critical applications, including offshore oil installations,
duplex stainless steels have now largely replaced cop-
per-nickel alloys. Fire-retardant grades of fibre-reinforced
plastics are now available.
Borehole waters are generally very hard and cast iron
pipes are still used because of the low internal corrosion
rates permitted by the scaling which occurs naturally.
Acidic waters cause graphitic attack on cast irons.
33.3.2.2 Boiler water, steam and condensate
The water supply for boilers is usually treated. Treatment
depends on the quality of the water supply, the pressure
of the boiler, the heat flux through the tube walls and the
steam quality required. Most waters require dealkaliza-
tion. The water produced in this process is non-scaling
and potentially corrosive (see above).
The quality of water thus produced is generally ade-
quate for low-pressure boilers, but deoxygenation is usu-
ally achieved through the addition of sodium sulphite in a
feedwater tank. This should be designed to prevent further
ingress of oxygen, employing a nitrogen blanket if nec-
essary. Pipework from the feedwater tank (or 'hot well'
if condensate is returned to it) to the boiler is steel, since
the water is low in corrosivity.
For higher-pressure boilers demineralization is neces-
sary to minimize total dissolved solids in the boiler. This
water is normally carried in steel pipework, but if con-
densate is returned and the condensate has become con-
taminated (for example, with carbon dioxide or copper
ions) more corrosion-resistant materials such as copper
are required. Downstream of the boiler, steam pipework
is usually steel with steel or stainless steel expansion
bellows.
The presence of corrosive species in steam, however,
creates the following corrosion problems:
1. Oxygen renders the steam corrosive towards steel. If
the feed is not deoxygenated a corrosion allowance
should be added to the steam piping wall thickness,
an oxygen scavenger such as hydrazine can be used
or a volatile corrosion inhibitor could be employed.
Candidate inhibitors include filming amines such as
ethoxylated soya amines. Some amine-based inhibitors
can break down and cause corrosion of copper in the
steam or condensate system. Some products such as
hydrazine are toxic or carcinogenic and cannot be
used in food-processing plant, breweries and steam
sterilizing equipment for hospitals.
2. Carbon dioxide, from the decomposition in the boiler
of temporary hardness salts present in some waters,
causes corrosion of steel steam pipework and cast iron
valves and traps. Corrosion inhibitors may be used, but
the choice of inhibitor must take into account the other
materials in the system. Neutralizing amines such as
morpholine or cyclohexylamine are commonly used.
3. Condensate return lines are often copper. Copper has
good corrosion resistance to oxygen and carbon diox-
ide individually. When both gases are present in the
Materials selection 33/11
condensate, copper is susceptible to corrosion. Copper
picked up in the condensate system and returned to the
boiler causes serious corrosion problems in the boiler
and any steel feedwater and steam pipework. Boiler
tubes should last for 25 years but can fail within one
year in a mismanaged or ill-designed boiler system
suffering from these faults.
. Boiler salts can contain chloride ions. When carried
over into the steam (e.g. during priming) this can
result in chloride stress corrosion cracking of austenitic
stainless steel expansion bellows. In steam systems
where freedom from chloride cannot be guaranteed,
bellows can be supplied in 9% Cr steel or austenitic
alloys containing 40% nickel.
Reputable water treatment specialist companies are able
to provide complete chemical treatment and corrosion
monitoring services.
33.3.2.3 Aqueous processes
The principal corrosive species as described in
Section 33.1.2, combined with the reducing or oxidizing
nature of the environment, can be used to select candidate
materials for any aqueous system. The process variables
discussed in Section 33.3.5 must also be considered.
Steel is the most common constructional material, and
is used wherever corrosion rates are acceptable and prod-
uct contamination by iron pick-up is not important. For
processes at low or high pH, where iron pick-up must
be avoided or where corrosive species such as dissolved
gases are present, stainless steels are often employed.
Stainless steels suffer various forms of corrosion, as
described in Section 33.5.2. As the corrosivity of the envi-
ronment increases, the more alloyed grades of stainless
steel can be selected. At temperatures in excess of 60°C,
in the presence of chloride ion, stress corrosion crack-
ing presents the most serious threat to austenitic stainless
steels. Duplex stainless steels, ferritic stainless steels and
nickel alloys are very resistant to this form of attack.
For more corrosive environments, titanium and ultimately
nickel-molybdenum alloys are used.
The following brief summary of materials for acid
duty is very simplified, and further guidance should be
obtained from the sources of information outlined in
Section 33.3.1:
1. Sulphuric acid: Plain carbon steel is used at strengths
in the range 70-100% at temperatures up to 80°C.
In flowing conditions the corrosion rate increases,
thereby rendering steels unsuitable for pumps. Storage
tanks for sulphuric acid require care in design to
prevent the possible ingress of water vapour. This
creates a layer of dilute acid on top of the bulk acid,
creating rapid attack at the fill line. At strengths below
70°C chemical lead is the preferred material for tanks
but at temperatures above 120°C corrosion becomes
significant. For castings such as pumps, valves, fittings
and some heat exchangers cast iron containing 15%
silicon is used at strengths up to 100%, at temperatures
up to the boiling point. Where iron pick-up must be
avoided, conventional stainless steels can be used in
33/12 Corrosion
certain ranges of strength and temperature. Anodically
protected titanium is used at 70% (60°C) and 40%
(up to 90°C). For higher temperatures and strengths
above 100%, nickel-based alloys were molybdenum
are used. Glass, gold and platinum are required for
specific combinations of acid strength and temperature.
2. Hydrochloric acid: The selection of materials for han-
dling hydrochloric is exceptionally difficult on account
of the sensitivity of corrosion to minor impurities. Oxi-
dizing hydrochloric acid is highly corrosive; dissolved
oxygen or trace levels of metal ions in high oxidation
states (especially ferric or cupric) ions are the com-
mon oxidizing species. Dilute hydrochloric acid (2%
at up to 70°C or 10% at 20°C) is handled by copper,
nickel and some of their alloys. Moderate-strength acid
(up to 10% at up to 70°C, up to 30% and 30°C and
up to 40% at 15°C) requires the use of more resistant
materials such as cast iron alloyed with silicon and
molybdenum or, in the absence of oxidizing agents,
silicon bronze. Strong hydrochloric acid at moderate
temperatures (below 70°C) can be handled by some of
the nickel-molybdenum alloys or non-metallic linings
such as rubber. At higher temperatures silver, platinum,
zirconium and tantalum are suitable across the whole
range of acid strength.
3. Nitric acid: Stainless steels are the most commonly
used materials for nitric acid duty. Because molybde-
num is widely considered to be detrimental to corro-
sion behaviour in nitric acid, type 304L is generally
preferred to type 316L. It is used at temperatures up
to 100°C and acid strengths up to 70%. At 100%
strength the temperature limitation is 30°C. Higher
alloys, based on 20% Cr, 25% Ni, extend the tem-
perature range to nitric acid's boiling point at con-
centrations up to 50%. An austenitic stainless steel
containing 18% chromium, 10% nickel and 2 - 3 % sil-
icon was developed specifically for this service. Cast
iron with 15% silicon is used for valves, pumps, etc.
in acid strengths of 0-40% (up to 70°C) and 40-80%
(up to 110°C). Some aluminium alloys may be used for
nitric acid above 80% strength at temperatures up to
35%. Titanium, gold, platinum and tantalum are used
to extend the temperatures range where required.
For alkaline duty, steel can sometimes be used up to
approximately pH 11. Zinc, aluminium and similar metals
and their alloys have limited use in alkaline conditions
because they dissolve, giving complex anions. Iron and
steel react in this way above about pH 12. Approximate
limits of use for zinc are pH 6-12 and aluminium alloy
pH 4-8. Stainless steels, including the lower grades, can
be used even in the presence of chloride ions at pH levels
of approximately 8. In the absence of halides they can be
used up to about pH 13.
Materials selection cannot be based on any simple com-
bination of common corrosive species. There are many
complicating factors, including the harmful or beneficial
effects of contaminants at the p.p.m, level, the relative
proportion in which certain combinations of species are
present (H + and C1- are often synergistic in their effect,
whereas SO42- and C1- often counter each other) and the
presence of naturally occurring corrosion inhibitors.
33.3.3 Non-aqueous processes
33.3.3.1 Solvent systems
Most organic solvents, except for protic solvents such as
alcohols and amides/amines, have reasonably low ionic
conductivities and hence do not support electrochemical
corrosion to any significant extent. Steel is commonly
used except in systems in which water can separate and
where the conductivity is sufficient to permit the flow of
ionic current.
33.3.3.2 Dry gases
Standard materials for handling dry gases include: for
chlorine, UNS N10276 (steel is used for liquid chlorine);
for bromine, UNS N10276 (below 60°C); for fluorine,
copper; for hydrogen chloride, UNS N10276; for hydro-
gen sulphide (sulphur dioxide and trioxide), type 316
stainless steel. The presence of water in even trace quan-
tities changes this significantly.
33.3.4 High-temperature environments
33.3.4.1 Chimneys, flues and ducts
The main corrosion process that occur in these items arise
from condensing liquids on the internal surface. Although
often lagged, heat loss frequently causes internal skin
temperatures to fall below the dewpoint of one or more
components of the gas stream, albeit locally, such as at
support points. Even at temperatures above its dewpoint
a gas can dissolve in condensed water. Rapid corrosion
can then occur in this thin film of corrosive liquid.
Condensing species of relevance to corrosion include
water and all acid gases. The dewpoint of water is
obtained from standard tables, requiring only the water
content (i.e. relative humidity) of the gas stream. Above
the water dewpoint, corrosion problems include condens-
ing acids (Section 33.3.2), dry acid gases (Section 33.3.3)
and erosion. Below the water dewpoint acid gases dissolve
in the water film to create an acidic solution, including:
1. Carbon dioxide produces a solution of carbonic acid
(as in boiler condensate, see Section 33.3.2). Carbon
steel is often employed but corrosion rates of up
to 1 mm/yr can be encountered. Coatings and non-
metallic materials may be employed up to their temper-
ature limits (Section 33.5.6). Basic austenitic stainless
steels (type 304) are suitable up to their scaling tem-
peratures.
2. Sulphur oxides produce sulphurous/sulphuric acid.
Because oxygen and halides are often also present,
high-nickel alloys are required. Current practice for
some flue gas desulphurizer ducts operating in the
temperature range 60-120°C involves lining the steel
structure with 2mm thick UNS N10276 or UNS
N06022. Non-metallic linings, including glass flake
epoxy materials, have limited application in the lower
part of the temperature range.
Other acid gases such as hydrogen chloride and oxides
of nitrogen produce similar corrosion problems. The

corrosion effects produced by acid condensates are ampli-
fied by the motion of the gas stream (typically 20-33 m/s)
and erosion effects due to entrained solids and impinge-
ment at bends, damper plates, reheaters, etc.
In the complete absence of acid gases, steel or gal-
vanized ducting can be suitable (but 80°C is often the
operational temperature limit for galvanized steel).
33.3.4.2 Furnaces
Table 33.4 gives scaling data for classes of heat-resisting
alloys in air. Mechanical properties, particularly creep
strength, may be reduced significantly at temperatures
well below scaling temperatures.
In addition to scaling and loss of mechanical properties
the third major factor in material performance in fumaces
is the detailed chemical environment. Specific problems
are encountered with the following:
1. Oxygen or steam in reformer tubes (typically up to
1000°C), steam superheater tubes (700-900°C), regen-
erators, etc. promote oxidation. Catastrophic oxidation
occurs in some materials including molybdenum-
bearing stainless steels because of the volatility of
MOO3. Alloys containing 25% Cr and 20% Ni are used
up to 1100°C but it is sometimes necessary to employ
35% Cr, 25% Ni or even 45% Cr, 25% Ni types. Type
316 austenitic stainless steel can suffer catastrophic
oxidation at temperatures as low as 760°C, although
it is often used up to nearly 900°C provided that a free
flow of gas is achieved over the entire surface.
2. Sulphur compounds, whether organic or inorganic in
nature, cause sulphidation in susceptible materials. The
sulphide film which forms on the surface of many
construction materials at low temperatures becomes
friable and melts at higher temperatures. The presence
of molten sulphides (especially nickel sulphide) on a
metal surface promotes the rapid conversion to metal
sulphides at temperatures where these sulphides are
thermodynamically stable. High-alloy materials such
as 25% Cr, 20% Ni alloys are widely used, but these
represent a compromise between sulphidation resis-
tance and mechanical properties. Aluminium and sim-
ilar diffusion coatings can be of use.
3. Carbon sources permit carburization. In cracker
furnaces, ethylene pyrolysis tubes, etc. carburization
results in rapid metal loss in the temperature
range 850-1100°C. A lower-temperature form of
this mechanism (sometimes termed 'metal dusting')
occurs in the range 400-800°C in the carbon
monoxide/carbon dioxide/hydrogen/hydrocarbon gas
streams. Preferred materials include HK40 (25% Cr,
20% Ni, 0.4% C) or higher alloys (35% Ni, 25% Cr
or 30% Cr, 30% Ni), depending on the severity of the
environment.
4. Nitrogen-containing compounds, including ammonia
and amines, cause nitriding in susceptible materials.
Ammonia synthesis (500-600°C, 250-400 bar pres-
sure) requires the use of austenitic stainless steels
or nickel alloys. The rate of penetration of nitriding,
which causes severe embrittlement, can be 0.2 mm/yr
in basic austenitic grades (e.g. type 304), falling by
Materials selection 33/13
more than an order of magnitude for 40% Ni base
alloys.
5. Molten salts promote rapid corrosion of many con-
structional materials at relatively low temperatures.
Low melting-point salts include sodium salts from
saline atmospheres, fireside ash, silicate insulation,
contaminants in the feed, etc. Corrosion rates of sev-
eral mm/yr can be observed at temperatures as low as
520°C. High chromium- and nickel-containing alloys
up to 50% Cr/50% Ni are employed.
Other detrimental factors which require to be taken into
account in the materials selection process include temper-
ature cycling and the presence of halide gases. Specialist
alloys containing rare earth element additions such as
cerium, lanthanum and yttrium have been developed for
use in certain environments up to 1300°C.
33.3.5 Influence of process variables on
material selection
33.3.5.1 Flow
The main effects of flow on corrosion of most materi-
als is deleterious. Increasing flow through the laminar
region serves to create a thinner hydrodynamic bound-
ary layer on the metal surface. This increases the rates of
diffusion of cathodic species and anodic products, gen-
erally increasing corrosion rates. Protective deposits and
scales are removed or prevented from forming. Damage
is caused at locations such as those shown in Figure 33.6.
The most important exception to this is the behaviour
of stainless steels in sea water. Grades which are suscep-
tible to pitting or crevice corrosion, including types 304
and 316, are liable to greatest pitting attack in stagnant
conditions where debris settles onto upward-facing hori-
zontal surfaces. In conditions of higher flow (greater than
1 m/s) the debris is less likely to settle and thus pitting
(Section 33.5.2) is less likely.
If flow becomes turbulent the corrosion rate increases
even more rapidly. In practice, most engineering materials
have a 'critical velocity' above which the corrosion rate
is unacceptably high. This does not correspond with the
laminar-to-turbulence transition. Surface roughness is an
important consideration.
Impingement attack (sometimes termed erosion cor-
rosion) is a result of the combined effect of flow and
FLOW
J
Figure 33.6 Areas affected by erosion corrosion
33/14 Corrosion
corrosion on a metal surface and it occurs when metal
is removed from the surface under conditions where pas-
sivation is insufficiently rapid. It is a function of flow,
corrosion and passivation.
At higher flow rates cavitation is a serious degradation
mechanism, where vapour bubbles created by pressure
fluctuations brought about by the flow of liquid past the
surface collapse on the metal surface with tremendous
force. This damages any protective oxide which may
be present, leading to pitting corrosion. It also causes
mechanical damage to the metal.
The corrosive and mechanical effects of flow are
observed in pipes, especially at bends and downstream of
flow disturbances, tube and shell heat exchangers, valves
and pumps. More corrosion resistant and/or harder materi-
als are used in such areas. Austenitic stainless steels work
harden and hence are superior in flowing conditions to
ferritic stainless steels of otherwise similar corrosion resis-
tance. Hard facings, such as hard chromium electroplate "
or cobalt based alloys, are used for local improvement of
flow-assisted corrosion resistance where it is not required,
practical or economic to fabricate the complete item of
high resistant material. Hard facing by cobalt-based alloys
can be achieved by weld overlay and plasma spray tech-
niques. Correct choice of alloy is important to optimize
corrosion resistance, though the resistance of most of the
alloys in use is superior to the more common grades
of stainless steel. Hard chromium plating contains micro
cracks; this allows corrodents to gain access to the sub-
strate. Layers of copper and/or nickel are often required
to prevent the corrosion of underlying steel.
The alternative option for counteracting cavitation dam-
age is the use of a resilient material such as rubber. The
mechanical forces attendant on collapse of the bubbles are
absorbed by elastic deformation of the resilient material.
33.3.5.2 Intermittent operation
This is generally harmful to plant, since deposits can form
in periods of stagnation. The common grades of stainless
steel, such as types 304 and 316 and other materials
(including aluminium alloys and some copper alloys),
are susceptible to under-deposit attack and are at risk in
circumstances where the process liquid is left to stagnate
in between operation cycles. Evaporation can also take
place, creating more aggressive conditions than occur
during operation. Condensation can form on vessel walls,
roofs and support points as heat is lost. This condensation
can absorb corrosive gases, creating localized corrosion
effects of greater severity than the bulk environment that
is normally present.
In rare circumstances, however, intermittent operation
is beneficial. Plant which would pit if operated contin-
uously can be operated successfully if the operational
cycle is shorter than the initiation period for pitting
attack and it is cleaned between batches. This is mainly
the case for stainless steel plant operating with high
enough chloride levels (or a low pH or high temperature)
to initiate pitting if operated continuously. The clean-
ing regime between batches promotes repassivation; hot
water, steam, chloride-free caustic or sodium orthophos-
phate solution may be required to achieve the required rate
of passivation. Plant is rarely designed on this principle,
but a significant number of cooking vessels, especially in
the food and drink industries, operate this way.
33.3.5.3 Commissioning
The first use of new plant, or start-up after a shutdown,
poses corrosion hazards additional to those encountered in
normal operation. New plant such as boilers requires spe-
cial water treatment, involving boil-out, passivation and
possible chemical cleaning. Actual requirements depend
on the boiler type, the proposed service, the quality
of water available during commissioning and the inter-
nal condition of the boiler. The condition of the boiler
depends on for how long and in what conditions it has
been stored. The presence of any salts, dirt or rust is
harmful. Boiler internals are covered by an adherent, pro-
tective layer of magnetite in normal operation. To create
an intact layer of magnetite of the correct thickness, crys-
tal type and morphology all debris must be removed from
the boiler, manually or by means of a vacuum cleaner.
If rusting is more than superficial, or if millscale is
present on the steel of the boiler tubes or shell, chem-
ical cleaning is necessary. BS 2486 contains details of
approved chemical cleaning techniques. These are based
on inhibited acids (e.g. citric acid). The most appropri-
ate technique is determined by a number of factors, and
the most important include the nature of deposits to be
removed (i.e. rust, hardness scales, millscale and metallic
copper).
Degreasing may be necessary, especially on new plant,
and alkaline cleaning techniques are described in BS 2486.
For optimum boiler life, passivation treatment is recom-
mended. This is described in BS 2486.
All boiler treatment and cleaning should be carried out
by reliable contractors, preferably within the guidelines
contained in BS 2486. Boilers are fired pressure ves-
sels and are subject to mandatory insurance inspections.
Significant benefits in safe and economic operation, par-
ticularly by reducing unnecessary chemical treatment, can
be achieved by monitoring the condition of boilers.
Items of plant fabricated from stainless steels should
be inspected before first use and after any maintenance
work or unplanned shutdown. All materials which rely
for their corrosion resistance on the presence of an oxide
or similar passive layer are susceptible to localized attack
where that layer is absent or damaged. Damage is most
commonly caused by scratching, metallic contamination
(nearby grinding or touching with ferrous tools), embed-
ding of grit and weld spatter.
Serious damage of this nature should be rectified before
the plant is used. Scratches, embedded grit and weld splat-
ter can be removed by careful grinding. All grinding and
abrading on stainless steel must be carried out using alu-
mina or a similar abrasive, rather than silicon carbide
consumables. Wire brushes must be stainless steel, rather
than carbon steel. Local dressing should be fine-finish,
since coarse scratches provide sites for the initiation of
pitting attack. In critical applications, if stainless steel is
to be used near its limit (in terms of corrosion), and for
cases such as welds, where a good finish cannot be oth-
erwise achieved, additional passivation is required. Nitric

acid (10-15% by volume) is the best passivator.It also
dissolves iron contamination. In circumstances where the
use of nitric acid is not possible for safety or physical rea-
sons (such as the underside of vessel roofs) passivation
paste is appropriate. Both materials are used at ambi-
ent temperature and require a contact time of approxi-
mately 30 min. They must be removed by thorough rinsing
with low chloride-content water. Guidance is available in
ASTM A380.
New plant can be contaminated with corrosive species
such as sea salt if it has been transported by sea as deck
cargo. Corrosive species present on building sites can
gain ingress into items of plant. Many metallic materials,
particularly aluminium, zinc alloys and galvanized steel,
are susceptible to attack by alkaline cement dust and
acid floor cleaners. Significant corrosion can occur by
any of these agents before or during commissioning.
Titanium and, to a lesser extent, tantalum are at risk of
hydriding attack when contaminated by iron smears (see
Section 33.5.5).
The materials selection procedure should take into
account the commissioning requirements of all parts of
the plant in order not to include any materials that cannot
safely be degreased or chemically cleaned.
33.3.5.4 Heat transfer
During the materials selection procedure isothermal
corrosion testing may indicate the suitability of a
material for handling a corrosive process fluid. In many
cases where heat transfer is involved the metal wall
temperature experienced in service is higher than the
bulk process fluid temperature. This, and the actual
heat transfer through the material, must be taken into
account since both factors can increase corrosion rates
significantly.
33.3.5.5 Liquid line
The corrosion conditions can be different at the fluid
line from the bulk condition. Aqueous liquids have a
concave meniscus which creates a thin film of liquid on
the vessel wall immediately above the liquid line. Some
corrosion processes, particularly the diffusion of dissolved
gases, are more rapid in these conditions. Additionally,
the concentration of dissolved gases is highest near the
liquid surface, especially when agitation is poor. Locally
high corrosion rates can therefore occur at the liquid
line, leading to thinning in a line around the vessel. This
effect is reduced if the liquid level in the vessel varies
with time. Any corrosion tests undertaken as part of the
materials selection procedure should take this effect into
account.
33.3.6 Influence of external environment
The external environment experienced by plant can
be more corrosive than the internal process stream.
Any construction material must be chosen to withstand
or be able to be protected from external corrosive
agents.
Materials selection 33/15
33.3.6.1 Corrosive atmospheres
Corrosive species in the atmosphere include water, salts
and gases. Clean atmospheres contain little other than
oxygen, nitrogen, water vapour and a small quantity of
carbon dioxide. These species are virtually non-corrosive
to any of the common constructional materials for plant at
normal temperatures. Steel is susceptible to corrosion in
even fairly clean air where water can exist as liquid. For
plant operating at temperatures up to approximately 100°C
coatings are employed to protect steel if required. In clean
air corrosion rates are low, and corrosion is primarily
a cosmetic problem, although it may be necessary to
prevent rust staining of nearby materials. Organic coatings
(i.e. anti-corrosion paints) are used. Above 100°C the
steel will be dry provided it is sheltered from rain. It
is also likely to be lagged and therefore not required to
be further protected from corrosion. Few organic coatings
are suitable for use in this temperature range, but epoxy-
based coatings are used up to 120°C and vinylesters up
to 160°C. Silicones and inorganic and metallic coatings
can be used at higher temperatures. At temperatures above
480°C, oxide scaling is a serious problem for steel, and
altemative materials such as stainless steels or nickel-base
alloys are required.
Contaminated atmospheres create additional corrosion
problems. In marine environments (e.g. coastal power sta-
tions, offshore oil platforms and above the splash zone on
ships) the chief corrosive agents is sea salt. The chloride
content of sea salt precludes the use of stainless steel,
bare steel or aluminium in these areas. Steel requires
to be coated as described in the previous section. Any
coating scheme used for the protection of steel must be
properly selected and applied. BS 5493 contains guide-
lines for coating systems for a variety of environments.
The most common reasons for coating failure are incor-
rect application and inadequate surface preparation. The
use of an independent coating inspector, as discussed in
Section 33.7, is recommended. Aluminium alloys can be
coated with plastics, such as U-PVC, epoxy finishes or
conventional paints, with the correct etch primer where
applicable to ensure adhesion. Some alloys are suitable
for anodizing, but the corrosion protection afforded by
anodizing is variable and requires tight specification and
control to ensure correct anodized film thickness and seal-
ing efficiency. The lower grades of stainless steels are not
suitable for external exposure in marine atmospheric con-
ditions unless they are cleaned regularly with fresh water.
Higher grades of stainless steel, nickel alloys and most
copper alloys do not normally benefit from additional pro-
tection. The development of the coloured patina on copper
alloys can, however, be avoided by coating with inhibited
lacquer.
Industrial environments can be contaminated with sul-
phur oxides, other acid gases, ammonia, hydrogen sul-
phide, nitrogen oxides, chlorine, etc. Most of these are
corrosive towards steel, which is normally protected as
outlined above. Copper alloys are corroded by hydro-
gen sulphide at levels above 60 p.p.b, although generally
the corrosion product is stable and adherent and fur-
ther protection is not usually required. Corroded in this
way, copper is unsightly and the functioning of electronic
33/16 Corrosion
equipment is affected. Ammonia causes corrosion and
stress corrosion cracking of some copper alloys and if
it is present in significant quantities, most alloys require
protection by coating. Chlorine and most acid gases are
sufficiently corrosive towards aluminium alloys and the
lower grades of stainless steel that protection is required.
Coatings are the only practical solution if these gases are
present at high enough levels.
33.3.6.2 Corrosion under lagging
Thermal insulation of vessels and pipework usually em-
ploys glass fibre or foamed polyurethane products. In
their pure forms these pose little corrosion risk. Gener-
ally, they are contaminated with leachable acids and/or
chloride ions. Chloride-free lagging can be specified and
this should be used for contact with metals which are
susceptible to chloride pitting or chloride-induced stress
corrosion cracking. The lower-grade austenitic stainless
steels, including types 304, 321,347, 316, 317 and their
low-carbon variants, are at greatest risk. If the lagging
remains dry there is no corrosion problem. Lagging can
become wet through rain ingress at damage sites, spillage
around entry ports or leakage from steam coils or the
lagged vessel wall. Chloride leached from the lagging or
introduced with the water tends to concentrate on the pipe
or the vessel, and even trace quantities create serious cor-
rosion problems. If chloride-free lagging is not used, or
the risk of water ingress cannot be removed, the vessel
can be coated prior to lagging. Any barrier coating can
be used but if it is damaged it will not prevent corro-
sion. Aluminium, in the form of foil, is used to prevent
chloride-induced stress corrosion cracking of austenitic
stainless steels. Applied to the vessel or pipe prior to lag-
ging, it acts as a barrier to water and possibly by slight
cathodic protection.
33.3.6.3 Corrosion in concrete
Intact, good-quality, void-free concrete creates an alka-
line environment and protects embedded steelwork such
as reinforcing steel. Penetration by chloride ions, car-
bon dioxide, oxygen and other corrosive species reduces
this protection. Voids in the concrete created by inade-
quate compaction at the time of application accelerate
this penetration. Carbon dioxide reacts chemically with
the concrete, reducing its alkalinity. Chloride and oxygen
reach the reinforcing steel and initiate corrosion. The cor-
rosion products occupy a greater volume than the steel
from which they were formed. This expansion of some
three to fourteen times creates a pressure greater than
the tensile strength of the concrete. Spalling of the con-
crete allows further corrosion to take place. Footings for
structural steelwork and vessel supports disintegrate and
blocks of detached concrete fall off the roofs and walls
of buildings. For reinforced concrete bases for vessels,
especially where acidic or chloride- or sulphate-containing
spillages are anticipated, the following options should be
considered at the materials selection process: acid- and
sulphate-resisting concrete are available; specialist coat-
ings can be applied to concrete to prevent the ingress
of species that would attack the concrete or corrode the
reinforcing steel or the process vessel that is encased in
the concrete; reinforcing material can be galvanized steel,
stainless steel (BS 6743) or plastic coated; reinforced steel
can be cathodically protected.
33.3.6.4 Corrosion in soil
Pipework is the item of plant that is most usually buried
in soil. In addition to the internal environment, the soil
can present a significant corrosion hazard.
Light, sandy, well-drained soil of high electrical resis-
tivity is low in corrosivity and coated steel or bare stain-
less steels can be employed. It is unlikely that the whole
pipe run would be in the same type of soil. In heavier
or damp soils, or where the quality of backfilling cannot
be guaranteed, there are two major corrosion risks. Steel,
copper alloys and most stainless steels are susceptible to
sulphide attack brought about by the action of sulphate-
reducing bacteria in the soil. SRB are ubiquitous but thrive
particularly well in the anaerobic conditions which per-
sist in compacted soil, especially clay. The mechanism
of corrosion where SRB are involved is described in
Section 33.1.2. Cathodic protection alone, described in
Section 33.7.2, is not completely effective against this
type of corrosion. Where SRB activity is likely or the
resistivity of the soil is less than 4000 f2 cm the pipeline
should be coated. Anti-corrosion paint systems at a min-
imum thickness of 0.2 mm and rubber, plastic or bitu-
minous coatings are preferred protection systems. Close
control of backfill quality is necessary to ensure that
sharp stones do not damage the coating, and full on-site
inspection is required to ensure defect-free application of
site-applied coatings and field joints. It is usually the case
that a combination of coating and cathodic protection is
the most economic alternative. For critical application or
the upgrading of old pipes, cathodic protection is gener-
ally suitable. Whether or not cathodic protection is used,
care is required to avoid stray current corrosion prob-
lems through interaction with nearby cathodic protection
or other electrical systems such as electrified railway sys-
tems.
33.3.6.5 Corrosion in timber
Metals in contact with or in the proximity of timber can
suffer enhanced corrosion attack. Some species of timber,
especially oak and Douglas fir, contain high levels of
acetic acid. These are volatile and cause corrosion of
nearby metals, especially iron, steel and lead alloys.
Metals in contact with timber can be corroded by the
acetic acid of the timber and by treatment chemicals
present in it. Treatment chemicals include ammonium sul-
phate and ammonium phosphate flame retardants. These
are particularly corrosive towards steel, aluminium and
copper alloys. Preservative treatments include copper salts
which, at high timber moisture contents, are corrosive
towards steel, aluminium alloys and zinc-coated items.
33.4 Design and corrosion
The design of a plant has significant implications for its
subsequent corrosion behaviour. Good design minimizes

corrosion risks whereas bad design promotes or exacer-
bates corrosion.
33.4.1 Shape
The shape of a vessel determines how well it drains
(Figure 33.7). If the outlet is not at the very lowest point
process liquid may be left inside. This will concentrate
by evaporation unless cleaned out, and it will probably
become more corrosive. This also applies to horizontal
pipe runs and steam or cooling coils attached to vessels.
Steam heating coils that do not drain adequately collect
condensate. This is very often contaminated by chloride
ions, which are soon concentrated to high enough levels
(10-100 ppm) to pose serious pitting and stress corrosion
cracking risks for 300-series austenitic stainless steel
vessels and steam coils.
Flat-bottomed storage tanks tend to suffer pitting cor-
rosion beneath deposits or sediments which settle out.
Storage tanks may be emptied infrequently and may not
experience sufficient agitation or flow to remove such
deposits.
Flange face areas experience stagnant conditions. Addi-
tionally, some gasket materials, such as asbestos fibre,
contain leachable chloride ions. This creates crevice and
stress corrosion cracking problems on sealing surfaces.
Where necessary, flange faces which are at risk can
be overlaid with nickel-based alloys. Alternatively, com-
pressed asbestos fibre gaskets shrouded in PTFE may be
used. Graphite gaskets can cause crevice corrosion of
stainless steel flanges.
Bends and tee-pieces in pipework often create locally
turbulent flow. This enhances the corrosivity of the pro-
cess liquid. These effects should be minimized by the use
of flow straighteners, swept tees and gentle bends. Flow-
induced corrosion downstream of control valves, orifice
plates, etc. is sometimes so serious that pipework requires
lining with resistant material for some twelve pipe diam-
eters beyond the valve.
33.4.2 Stress
The presence of tensile stress in a metal surface ren-
ders that surface more susceptible to many kinds of
--i
Liquids can concentrate
and deposits can form
Figure 33.7 Details of design creating corrosion problems
Design and corrosion 33/17
corrosion than the same material in a non-stressed condi-
tion. Similarly, the presence of compressive stress in the
surface layer can be beneficial for corrosion, and espe-
cially stress corrosion cracking, behaviour.
Tensile stresses can be residual, from a forming or
welding operation, or operational from heating-cooling,
filling-emptying or pressurizing-depressurizing cycles.
The presence of a tensile stress from whatever origin places
some materials at risk from stress corrosion cracking, as
described in Section 33.1.3. Some items of plant can be
stress-relieved by suitable heat treatment, but this cannot
prevent operational stress arising.
Cyclic stresses can also give rise to fatigue or corrosion
fatigue problems. Information relating to the fatigue life
of the material in the service environment is required,
together with the anticipated number of stress cycles to
be experienced by the item over its operational life. The
fatigue life (the number of cycles to failure) or the fatigue
strength (the stress level below which it does not exhibit
fatigue problems) is then used in the design.
The presence of stress raisers, including sharp comers
and imperfect welds, produces locally high stress levels.
These should be avoided where possible or taken into
account when designing the materials for use in environ-
ments in which they are susceptible to stress corrosion
cracking or corrosion fatigue.
33.4.3 Fabrication techniques
Most fabricational techniques have implications for corro-
sion performance. Riveted and folded seam construction
creates crevices as shown in Figure 33.8. Those mate-
rials which are susceptible to crevice corrosion should
be fabricated using alternative techniques (e.g. welding).
Care should be taken to avoid lack of penetration or lack
of fusion, since these are sites for crevice corrosion to
initiate.
Welding should be continuous, employing fillets where
possible, since tack welds create locally high stresses and
leave crevice sites. Welding consumables should be cho-
sen to create weld metals of similar corrosion resistance
to the parent material. This often requires the use of a
slightly over-alloyed consumable, to allow for loss of
volatile alloying elements during the welding process and
IV
)ot weld
Heat exchanger tube
expanded into tube
g plate
Figure 33.8 Details of jointing processes creating additional cor-
rosion risks (crevices and stress concentrations)
33/18 Corrosion
to compensate for the inherently poorer corrosion resis-
tance of the weld metal structure. Strongly over-alloyed
weld consumables can create galvanic corrosion problems
if the weld metal is significantly more noble than the par-
ent material. In all welds the heat-affected zone is at risk.
The new structure which forms as a consequence of the
thermal cycle can be of lower corrosion resistance, in
addition to the often poorer mechanical properties, than
parent material. Austenitic steels such as type 304 and
316 are also susceptible to sensitization effects in the heat-
affected zone. In these materials carbide precipitation dur-
ing the welding thermal cycle denudes the parent material
of chromium. This creates areas of significantly dimin-
ished corrosion resistance, resulting in knife-line attack in
many corrosive environments. This is avoided by the use
of the low-carbon equivalents (304L, 316L, etc.) or grades
such as type 321 or 347 which are stabilized against sen-
sitization. With correct welding techniques, however, this
should be necessary only with thick sections (5 mm for
304 and 8 mm for 316).
Some materials, particularly certain aluminium alloys,
duplex stainless steels in certain reducing environments
and most steel plate, are susceptible to end-grain attack.
Penetration along the end grain can be very rapid,
with corrosion exploiting the potential differences that
exist between inclusions and ferrite crystals in steel and
between austenitic and ferrite grains in duplex stainless
steel. Where end-grain attack is significant this should
not be exposed to the corrosive environment. It can be
covered by a fillet 'buttering' weld if necessary.
33.4.4 Design for inspection
Unseen corrosion can be the most damaging type of
attack. Items should be designed to permit periodic
inspection. This involves the provision of sufficiently
large manways, the installation of inspection pits, the plac-
ing of fiat-bottomed vessels on beams instead of directly
onto concrete bases and the facility for removal of thermal
insulation from vessel walls.
33.5 Uses and limitations of
constructional materials
33.5.1 Steels and cast irons
Steel is essentially iron with a small amount of carbon.
Additional elements are present in small quantities. Con-
taminants such as sulphur and phosphorus are tolerated at
varying levels, depending on the use to which the steel
is to be put. Since they are present in the raw mate-
rial from which the steel is made it is not economic to
remove them. Alloying elements such as manganese, sil-
icon, nickel, chromium, molybdenum and vanadium are
present at specified levels to improve physical properties
such as toughness or corrosion resistance.
Steels are used primarily for their strength and other
mechanical properties rather than for corrosion resistance.
The corrosion rates of steels is very low only in a
few environments, i.e. clean or hard water, certain non-
electrolytes, clean atmospheres and properly constituted
and compacted ordinary Portland cement. The economics
of any high-volume chemical processes, including oil
and gas production and transport systems, require the
widespread use of steels. By incorporating a corrosion
allowance into the design of a plant, corrosion rates of
up to 1 mm/yr can be accommodated, particularly for
plant designed for an operational life of only a few
years. Steels generally corrode in a uniform manner. This
is beneficial, since localized attack, as experienced by
stainless steels, is very difficult to allow for in plant
design. The use of corrosion inhibitors (Section 33.7.4)
extends the economic range of use of steels. Steel is also
available coated with zinc (galvanized, see BS 729) or
aluminium-zinc (BS 6830); see also BS 2659, BS 1706
and BS 4921.
Limitations on the use of steels include:
1. Acidic solutions below about pH 5 due to general
corrosion;
2. Alkaline solutions above about pH 11 due to general
dissolution and caustic cracking;
3. Marine and industrial atmospheres (without the use of
protective coatings), because of high corrosion rates;
4. Atmospheric exposure at temperatures in excess of
480°C, because of oxide scaling;
5. Certain environments containing nitrate, cyanide, car-
bonate, amines, ammonia or strong caustic, due to the
risk of stress corrosion cracking. Temperature is an
important factor in assessment of each cracking envi-
ronment;
6. Solutions of hydrogen sulphide, because of hydrogen-
induced cracking. Grades of steel are available for
certain ranges of pH and hydrogen sulphide partial
pressure.
Cast irons are iron with high levels of carbon. Heat
treatments and alloying element additions produce grey
cast iron, malleable iron, ductile iron, spheroidal cast
iron and other grades. The mechanical properties vary
significantly. Nickel-containing cast irons have improved
hardness and corrosion resistance. Copper or molybdenum
additions improve strength.
Chromium, silicon and other alloying elements are used
to create cast irons for corrosion resistance in specific
environments. Silicon-containing cast irons are used for
sulphuric acid duty.
Limitations on the use of cast irons are similar to
those for steel, since in many environments most cast
irons have poor corrosion resistance. Most grades are
also susceptible to graphitization (the loss of iron, leaving
a weak structure of graphite) in acidic environments
below a pH of approximately 5.5. This attack occurs in
soils.
33.5.2 Stainless steels
Stainless steels are iron-based alloys which contain at
least 11% chromium. Hundreds of grades of stainless
steel are available with alloying elements including nickel,
molybdenum, manganese and copper. Stainless steels are
not electrochemically inert, but are protected by a thin

layer of oxide. This passive layer is unstable in oxidiz-
ing and reducing environments. If lost completely, the
underlying steel corrodes rapidly in corrosive (particu-
larly acidic) environments. Where the oxide is lost locally,
pitting, crevice corrosion or stress corrosion cracking
can proceed at rates of several millimetres per year.
Chemical species including the halide ions assist in the
breakdown of the passive film in even mild conditions
such as neutral sodium chloride solution. For satisfac-
tory service, stainless steels should be passivated con-
tinuously (or intermittently, see Section 33.3.5) with a
suitable oxidizing species. The redox potential of an
environment is of prime importance in the selection
of a stainless steel. In reducing conditions the protec-
tive oxide cannot form, and general corrosion takes
place; in strongly oxidizing conditions pitting corrosion
is possible.
Stainless steels are used in a wide variety of appli-
cations and are most often selected because steel or
cast iron would corrode at an unacceptably high rate or
produce high levels of iron contamination in the pro-
posed service environment. The main limitations on their
uses are:
1. The lower grades of austenitic stainless steels suf-
fer chloride-induced stress corrosion cracking at tem-
peratures in excess of 60°C. Note that SCC can
occur at lower temperatures in low-pH, high chlo-
ride environments. Grades containing high levels of
nickel are more resistant. Sensitization (Section 33.4.3)
occurs near welds in certain non-stabilized grades.
The maximum hardness limitations described in NACE
MR-01-75 are necessary to prevent sulphide stress
corrosion cracking in hydrogen sulphide-containing
environments.
2. Ferritic stainless steels have inferior corrosion resis-
tance compared with austenitic grades of equiva-
lent chromium content, because of the absence of
nickel. Stress corrosion cracking can occur in strong
alkali.
3. Martensitic stainless steels are of limited use in chem-
ical environments because of their inferior corrosion
resistance compared with ferritic and austenitic grades.
Their mechanical properties allow them to be used
where high hardness is required, but they are suscep-
tible to hydrogen embrittlement.
4. Duplex stainless steels are mostly composed of alter-
nate austenite and ferrite grains. Their structure im-
proves resistance to chloride-induced stress corrosion
cracking. In certain reducing acids, such as acetic and
formic, preferential attack of the ferrite is a serious
problem.
In all categories the lower grades suffer pitting and
crevice corrosion in even low-chloride environments. As
the chloride content or temperature increases, or the pH
falls, higher grades are required. Reducing acids, includ-
ing acetic and formic, attack most grades, particularly
at elevated temperatures. Free-cutting grades, containing
deliberate additions of sulphur or selenium, have signifi-
cantly poorer corrosion resistance than the corresponding
standard grade.
Uses and limitations of constructional materials 33/19
33.5.3 Nickel alloys
Commercially pure nickel has good corrosion resistance
to a variety of aggressive environments and is specially
used for hot caustic service. It has moderate resistance to
acid attack but cannot be used in oxidizing acidic con-
ditions, i.e. in oxidizing acids or acidic media containing
oxidizing agents. Nickel is widely used for electroplat-
ing steel. Electroless nickel plating is used, especially for
items of complex shape that cannot be successfully elec-
troplated. As with all coatings, the presence of defects
such as pores allow the corrosive environment to reach
the substrate. Since nickel is more noble than steel,
rapid, galvanically assisted pitting of the substrate steel
occurs at defect sites. Nickel is not resistant in the com-
bined presence of ammonium compounds and oxidizing
agents.
Nickel is usually alloyed with elements including cop-
per, chromium, molybdenum and iron for strengthening
and to improve corrosion resistance for specific applica-
tions. Nickel-copper alloys (and copper-nickel alloys;
see Section 33.5.4) are widely used for handling water.
Pumps and valve bodies for fresh water, sea water and
mildly acidic alkaline conditions are made from cast
Ni-30% Cu type alloys. The wrought material is used for
shafts and stems. In sea water contaminated with sulphide,
these alloys are subject to pitting and corrosion fatigue.
Ammonia contamination creates corrosion problems as for
commercially pure nickel.
Nickel-chromium alloys can be used in place of
austenitic stainless steels where additional corrosion resis-
tance is required. These alloys are still austenitic but are
highly resistant to chloride-induced stress corrosion crack-
ing when their nickel content exceeds 40%.
Molybdenum-containing nickel alloys are a family of
alloys which include grades for handling hot hydrochloric,
sulphuric, oxidizing, reducing and organic acids. Some
grades are susceptible to attack in ammonia- or sulphide-
containing environments.
33.5.4 Copper alloys
Copper has excellent resistance to some corrosive
environments, including fresh waters and fluoride-
containing atmospheres. Alloying is necessary to achieve
good strength, but copper lining with steel for strength is
an alternative (BS 5624). Copper and some of its alloys
are susceptible to crevice corrosion, but the mechanism is
different from that which affects stainless steels.
Alloying with zinc produces brasses. Brasses are gen-
erally suitable for fresh and potable waters but dezincifi-
cation is a problem for grades containing more than 12%
zinc, especially in acidic/alkaline conditions. The addi-
tion of small amounts of tin, with arsenic or phosphorus,
prevents dezincification and is particularly necessary for
sea water applications. For handling flowing sea water
or other corrosives, copper-nickel alloys are favoured.
Brasses are susceptible to stress corrosion cracking in
the presence of ammonia or ammonium compounds. This
includes lavatory fittings.
Copper-aluminium alloys (aluminium bronze) have
good general corrosion resistance but de-aluminification
33/20 Corrosion
is possible in some grades in certain environments.
Copper-nickel alloys are used where flow-related corro-
sion is a problem, particularly in sea water heat exchanger
tubes. The grades containing a small iron addition have
the highest critical velocity. These alloys are all suscep-
tible to pitting by sulphides (created in stagnant condi-
tions in sea water systems), especially in the presence
of chlorine and stress corrosion cracking in the presence
of ammonia. To minimize their susceptibility to sulphide
attack in estuarine or other polluted waters it can be
advisable to pre-passivate and, if necessary, inhibit con-
tinuously, with ferrous sulphate.
33.5.5 Miscellaneous metallic materials
33.5.5.1 Aluminium alloys
Commercially pure aluminium has good resistance to
atmospheric corrosion, except where chloride is present in
significant quantities, i.e. within 1 - 2 km of the coast and
in the vicinity of chemical plants. To achieve reasonable
strength, alloying additions of silicon, manganese, zinc
and copper are made. Aluminium-copper alloys have
poor corrosion resistance and should not be used in
corrosive environments. The other aluminium alloys have
good corrosion resistance to most near-neutral media but
are attacked in acidic and alkaline conditions. They are
susceptible to pitting corrosion, especially where deposits
form. Chloride ions also induce pitting of these alloys;
chloride is often generated by the hydrolysis of chloride
containing organic chemicals. Other organic materials can
be handled safely when dry, except alcohols. Metal ions
such as copper, which may be introduced by the corrosion
of other plant, storage vessels or pipework, can plate
out onto the aluminium and create a galvanic cell which
produces intense pitting.
The use of anodic films on aluminium alloys is only
applicable to some mildly corrosive environments, includ-
ing architectural purposes (BS 3987) and where abrasion
resistance is required (BS 5599).
The presence of liquid mercury poses a liquid metal
embrittlement problem for many alloys. This occurs by
the spillage of mercury in aircraft and the condensation
of mercury vapour from mercury pumps in cryogenic
applications.
33.5.5.2 Titanium
Titanium is available unalloyed in several grades of purity,
and alloyed. Each grade or alloy has specific uses. Gener-
ally, titanium has excellent resistance to sea water and
oxidizing conditions, including acids and concentrated
hypochlorite solution. It is susceptible to corrosion in
reducing acids, rapid attack in dry chlorine gas, crack-
ing in alcohols and embrittlement by hydriding. Hydrid-
ing is possible when titanium or a metal with which
it is in contact corrodes, liberating hydrogen. For this
reason, titanium plant should be inspected carefully for
traces of metallic contamination (e.g. scratch marks from
steel-nailed boots). Titanium can be used in the anodized
form; a thick anodic coating greatly reduces hydrogen
absorption. It is used with anodic protection for handling
4 0 - 7 0 % strength sulphuric acid and 36% hydrochloric
acid.
33.5.5.3 Other metals
Other metals are available which have very specific uses.
Their cost is very high but their use is often justified
in certain processes such as fine chemical manufacture.
They are generally used as thin cladding or loose lining
supported by steel or stainless steel.
1. Gold is resistant to many strong acids, but not cyanides
or high levels of fluoride, hydrogen fluoride and chlo-
rine.
2. Platinum is resistant to strong acids and most halogen
gases.
3. Silver has reasonable resistance to inorganic and
organic acids (not strongly oxidizing acids), hot alka-
line conditions and some fluoride-containing environ-
ments. It tarnishes in the presence of sulphides.
4. Tantalum is inert to many environments including
hydrochloric acid at strengths up to 25% at 90°C.
It is attacked by fluorine, hydrofluoric acid, fuming
sulphuric acid containing free sulphur trioxide, and
many alkalines. Above 300°C it reacts with many gases
including air. It is subject to hydrogen embrittlement
in a similar manner to titanium.
33.5.6 Linings and coatings
Many materials with good corrosion resistance have inad-
equate strength or other mechanical properties to enable
their fabrication or economic use. Organic linings and
coatings are used in a range of plant. Organic materi-
als are generally susceptible to organic solvents and have
use only within limited ranges of temperature. All coat-
ings and linings must be applied in a manner to control
defects within them, since the substrate on which they are
applied will be corroded by any contact with the process
fluid.
Inspection techniques are available for specific coat-
ing and lining systems. Proper bonding to substrate is
necessary to avoid blistering and disbonding in case of
a vacuum being created in the vessel. BS 6374 provides
guidance on materials selection, design application and
repair.
33.5.6.1 Resins
Epoxy, polyester, phenolic and other resins are used as
coatings and linings with or without reinforcement (see
BS 6464 and BS 4994). Glass fibre, silica, carbon and
many other materials can be used as filters or reinforce-
ment to produce materials with specific properties of
strength, flexibility, wear resistance and electrical con-
ductivity.
33.5.6.2 Plastics
A large range of man-made polymeric materials is
available, from polyethylene, which is attacked by
most organic chemicals, to fluorinated products such as

polytetrafluoroethylene and polyethyletherketones, which
have outstanding resistance to virtually all chemicals.
All polymers have their own adhesive, welding and
fabricational limitations which must be taken into account
in the design of the coated item. These materials can also
be used in solid form.
33.5.6.3 Elastomers
Elastomers are used for their flexibility in seals, gaskets
and hoses and to resist abrasion (through absorption of
the kinetic energy of the impinging particles). The range
of materials includes natural and synthetic rubbers and
modem elastomers with chemical resistance.
33.5.6.4 Glass
Glass-lined reactor vessels are widely used in pharma-
ceutical and fine-drug manufacturing processes in which
metallic contamination of the products has to be mini-
mal. Most glasses are susceptible to attack in strong acid
and alkali media. Manufacture and design of glass-lined
vessels is a very specialized area. Such vessels require
periodic inspection to ensure the integrity of the lining,
which is susceptible to mechanical damage and repair is
difficult.
33.6 Specifying materials
Having selected the appropriate material for an item of
plant, it is imperative that this material is used in the
required state of heat treatment, surface finishes, etc. for
its construction.
33.6.1 Compositional aspects
National standards: British Standards, DIN (Deutsch
Industries Normen, from Germany) and ASTM/AISI
(American Society for Testing of Materials/American
Iron and Steel Institute) are those in most common
use for metallic materials. It is always preferable to
select a material for which national or international
specifications exist. Many materials, including steels
and the 300-series austenitic stainless steels, have wide
ranges of chemical composition. If an alloy or a grade
of stainless steel is to be used in a marginal duty
(i.e. where relatively small variations in alloying ele-
ment content, heat treatment or surface condition could
markedly affect its corrosion behaviour) this should
be taken into account when the material is speci-
fied. For stainless steels and some other materials of
construction it is possible to purchase, at a premium
rate, the material with specified minima of alloying
elements. Type 316 stainless steel, when supplied to
AISI specification, can have 10-14% nickel, 16-18%
chromium and 2-3% molybdenum. This wide speci-
fication produces material across a range of corrosion
resistance. It also creates a latitude for mechanical and
physical properties, including strength and magnetic
permeability. For demanding applications, but where
the more resistant type 317 material is not justified,
type 316 with a minimum molybdenum content may
Specifying materials 33/21
be specified. Types 304 and 316 are now often dual-
certified, i.e. 304 conforms also with the requirement
of 304L (the low carbon content grade, to be used
when intergranular corrosion is possible) and 316/316L
similarly.
2. Proprietary alloy designation: Many alloys are the sub-
ject of patents and are the registered trade mark of
the producer. When produced by the holder of the
patent, these alloys are generally of the highest stan-
dard, with tight control of composition, and exhibit
consistent corrosion behaviour. They are mostly orig-
inally developed for a specific corrosive service, but
are subsequently marketed more widely. These mate-
rials can also be manufactured under licence or, when
the patent has expired, by other producers using a
genetic name. Produced in this way, these materials
can be less reliable, with the introduction of unwanted
tramp elements. Where proving tests have indicated
the suitability of a proprietary alloy for specific appli-
cation, the same source of material should be used for
construction. Published compositional data are typical
values; cast analyses should be requested for critical
applications.
3. Government standards: These are developed by gov-
ernment departments, especially the military, and are
primarily for aerospace and marine service materials.
4. In-house specifications: End users requiring larger
amounts of material and fabricators of specialty plant
benefit from preparing their own in-house specifi-
cations. These are often based on existing national
standards modified to include additional requirements.
They include higher compositional ranges of alloying
elements, mechanical properties such as strength and
surface finish, inspection and certification. These spec-
ifications must be written by experts in order to satisfy
the users' real requirements without compromising any
aspect of the materials' performance.
33.6.2 Mechanical properties
Many available standards, particularly national standards,
either do not cover the materials' physical and mechan-
ical properties or leave the user to specify from a range
of options. The aspects of mechanical and physical prop-
erties which have specific implications for a material's
corrosion behaviour include:
1. Strength, particularly where the material will be sub-
jected to fatigue loading or is susceptible to degrading
mechanisms such as hydrogen-induced cracking;
2. Hardness, including surface hardness, especially for
materials for sour service or environments in which
stress corrosion cracking is expected. It is also impor-
tant where erosion corrosion is likely;
3. Surface finish, since good surface finish retards the
onset of certain types of corrosion attack, including
pitting and stress corrosion cracking.
All relevant physical and mechanical property require-
ments should be included in the material specification.
33/22 Corrosion
33.6.3 Certification
All aspects of the material's chemical, mechanical and
physical properties which are included in the specification
should be capable of measurement and certification. For
critical duties all material supplied should be fully tested
and certified by competent approved, independent test lab-
oratories. The nationally recognized laboratory approval
scheme in the UK is NAMAS (National Measurement
Accreditation Service), operated by the Department of
Trade and Industry through the National Physical Lab-
oratory. For less demanding duties material should be
accompanied by test certificates from the supplier. All
items of plant should be purchased with material certifica-
tion. Additional certification is required in cases where the
fabricator, in manufacturing an item of plant, uses tech-
niques such as welding or heat treatment which may affect
the corrosion behaviour of the construction materials.
All bought-in items of plant, especially those supplied
without adequate materials certification, should be subject
to random inspection. Portable instruments are available
for many types of non-destructive examination, chemical
analysis and mechanical testing of fabricated items of
plant.
33.7 Corrosion-control techniques
Economic reasons may dictate the use of an inexpensive
constructional material (i.e. steel protected by one of the
methods discussed below) in place of a more resistant
but more expensive material. Although all these corrosion
control techniques could be used on a variety of materials,
with appropriate design and safeguards, they are only
rarely employed to protect any material other than steel.
33.7.1 Painting
Paints are complex formulations of polymeric binders
with additives including anti-corrosion pigments, colours,
plasticizers, ultraviolet absorbers, flame-retardant chemi-
cals, etc. Almost all binders are organic materials such as
resins based on epoxy, polyurethanes, alkyds, esters, chlo-
rinated rubber and acrylics. The common inorganic binder
is the silicate used in inorganic zinc silicate primer for
steel. Specific formulations are available for application
to aluminium and for galvanized steel substrates.
Because of their generally poor resistance to solvents,
acids, alkalis and other corrosive agents, paints are not
normally used to protect plant internals handling anything
other than waters. Even clean (i.e. potable or better) water
is damaging to most epoxy-based coatings. Limitations on
the use of paint systems include:
1. Erosion resistance: The mechanical properties of paints
prevent their use in conditions where impingement or
erosion by entrained solids is expected.
2. Solvent resistance: This varies from very poor (for
chlorinated rubbers) to good (for polyurethanes). All
paint binders have specific susceptibilities and the
presence of small quantities of the appropriate solvent
in the atmosphere in the region of an item of plant can
cause rapid failure.
3. Saponification: Paints are most commonly used to
protect steel from corrosion by sea water in marine
applications and soil in the case of buffed structures.
Additional protection is often supplied by the appli-
cation of cathodic protection to the steel. Any paint
coating used in conjunction with cathodic protection
must be resistant to the alkali which is produced on the
steel at defect sites in the coating. The amount of alkali
generated depends on the potential to which the steel
is polarized. Some paint binders such as alkyds and
vinyl ester are very susceptible to saponification, and
should not be used on cathodically protected structures.
Cathodic disbondment testing, to BS 3900: Parts F10
and F11, should be undertaken if the relevant informa-
tion is not available.
4. Temperature: Thermal breakdown of the binder lim-
its the service temperature for painted items to 70°C
(e.g. chlorinated rubbers and polyvinylacetates), 120°C
(many epoxies) and 160°C (for vinylesters). The com-
monest reason for coating failure is incorrect applica-
tion. This is a skilled task, involving surface prepa-
ration (usually by grit blasting) and a tight control
over environmental factors (temperature and relative
humidity of the local atmosphere), the cleanliness of
the steel (national standards, including Swedish Stan-
dard SIS 05 59 00, BS 4232 and ISO 8502, cover the
testing of residual contamination and freedom from
millscale, rust, etc.), the paint and the application pro-
cedure. All coating should be supervised by an inde-
pendent, qualified inspector certified to a recognized
standard such as the National Association of Corrosion
Engineers International Coating Inspector Training and
Certification Scheme.
The selection of an anti-corrosion paint coating scheme
should be undertaken by a qualified expert. BS 3900
covers most aspects of paint coating testing.
33.7.2 Cathodic protection
Cathodic protection (CP) is an electrochemical technique
of corrosion control in which the potential of a metal
surface is moved in a cathodic direction to reduce the
thermodynamic tendency for corrosion. CP requires that
the item to be protected is in contact with an electrolyte.
Only those parts of the item that are electrically coupled
to the anode and to which the CP current can flow are
protected. Thus the inside of a buried pipe is not capable
of cathodic protection unless a suitable anode is placed
inside the pipe. The electrolyte through which the CP
current flows is usually sea water or soil. Fresh waters
generally have inadequate conductivity (but the interiors
of galvanized hot water tanks are sometimes protected
by a sacrificial magnesium anode) and the conductivity
of air is far too low to permit the cathodic protection of
buildings or vehicles.
The operating principle of CP is shown in Figure 33.9.
Good CP practice is embodied in BS Code o f Practice
CP 1021. For buried pipelines the power source is most
frequently d.c. voltage derived from a rectified mains
power supply, or in suitable climates such as Australia
and the Middle East, solar cells. Such systems operate

Power supply
Electrolyte(soil or water)
~~,,~ An!de~
Sacrificiaanode
l system Impressed currentsystem
(reactive metal anode) (inert metal anode)
Figure 33.9 Cathodic protection
by 'impressed current' and suitable anodes are platinized
titanium or niobium wire, cast iron, and graphite. They
are not consumed, but can be damaged by the chlorine
generated in sea water CP systems. Reinforcing steel
in concrete jetties, bridge and carpark decks and some
offshore structures are generally protected by means of
impressed current systems.
Sacrificial anode systems operate without external power
source. The anodes are reactive metals such as magnesium
and zinc or aluminium alloys. The energy for the pro-
cess is derived from the anode material. Careful design is
required to match the output and lifetime of the anodes
with the polarization and life-expectancy requirements of
the plant. Sacrificial anode CP is used for offshore plat-
forms, subsea pipelines and the inside of ballast tanks on
tanker ships.
Correct design is necessary to achieve full protection
without overprotection and to minimize wastage of power
or anode materials. Overprotection is undesirable because
alkali is generated at the cathode (this degrades many
plant systems) and hydrogen can be evolved. Hydrogen
is deleterious to some mechanical properties of steels
and pipelines have failed through this mechanism. Steel
is also susceptible to cracking in carbonate-containing
environments in certain potential ranges. A reduction in
the power requirements of a CP system can usually be
achieved by coating the protected structure.
Stainless steel pipes (buried in the ground) and the
interiors of stainless steel heat exchangers have been suc-
cessfully cathodically protected, but CP is rarely used for
materials other than steel. The protection potential usu-
ally adopted for steel is - 8 5 0 mV to the saturated calomel
reference electrode. This varies with temperature and the
presence of other aggressive species in the environment.
33.7.3 Anodic protection
Anodic protection is possible only for material-
environment combinations that exhibit fairly wide passive
regions. Examples include type 304 stainless steel in
phosphoric acid and titanium in sulphuric acid, steel and
stainless steels in 98% sulphuric acid. The effect of the
anodic polarization is to shift the steel into a region of
passivity. This promotes the formation of a protective
film on the steel surface, preventing general corrosion.
Corrosion-control techniques33/23
Inadequate or over-polarization creates corrosion and
hydrogen evolution problems. As with CP, only wetted
surfaces can be protected; this excludes condensed films
or droplets on vessel roofs and walls above the liquid line.
Such areas are subject to rapid corrosion attack.
33.7.4 Corrosion inhibitors
The use of cheap constructional materials such as steel can
be tolerated in certain instances where corrosion inhibition
is possible:
1. Boilers and steam systems: Steel steam lines can be
inhibited by the use of a volatile amine-based inhibitor
such as ammonia, morpholine or cyclohexylamine
introduced with the feedwater. It passes through the
boiler and into the steam system, where it neutral-
izes the acidic conditions in pipework. The inhibitor
is chemically consumed and lost by physical means.
Film-forming inhibitors such as heterocyclic amines
and alkyl sulphonates must be present at levels suffi-
cient to cover the entire steel surface, otherwise local-
ized corrosion will occur on the bare steel. Inhibitor
selection must take into account the presence of other
materials in the system. Some amine products cause
corrosion of copper. If copper is present and at risk
of corrosion it can be inhibited by the addition of
benzotriazole or tolutriazole at a level appropriate to
the system. All aspects of boiler water treatment are
covered in BS 2486 (see also Section 33.3.2).
2. Cooling waters: Once-through cooling systems can-
not usually be inhibited because of the expense of
the chemicals required and the problems of the dis-
posal of the treated water. Open recirculating systems
lose water by evaporation from the tower and pick
up atmospheric dust and other detritus. Inhibition of
such a water is part of the overall water treatment
in which scaling tendency and other factors are also
controlled. Phosphates and phosphonates together with
zinc ions are commonly used. Biocide additions are
also required and algal growth control in the case of
systems into which sunlight penetrates. Closed recir-
culating systems, including those in motor vehicles,
can be treated with the most efficient inhibitors such
as chromates. Disposal problems have restricted the
use of these chemicals, but borates, benzoates and
nitrites are used. If there is copper in the system,
this must be inhibited also. If aluminium is in the
system, nitrite poses a pitting problem in certain pH
ranges. Nitrite is also aggressive towards lead-tin sol-
dered joints and is ineffective as an inhibitor for zinc
and its alloys. Aluminium is only properly protected
by benzoate in acidic solutions; in alkaline conditions
it is not fully effective (borates are preferred). The
compatibility of all materials, including stainless steel
and galvanized steel, must be ascertained before any
corrosion-inhibition system is employed.
3. Heating systems: Where they are virtually sealed
except for a make-up tank (as in the case of domestic
central heating) systems can be treated as for closed
recirculating cooling systems, and benzoates, nitrites
33/24 Corrosion
and borates are used. In a correctly designed system
such that oxygen ingress is minimal and little make-up
is required the oxygen content diminishes as corrosion
proceeds and the system can require no inhibitor treat-
ment. The use of certain toxic corrosion inhibitors is
not permitted in hospitals, food-manufacturing plant
and breweries, where a leak could result in contami-
nation of food or exposure of people to the chemical.
4. Vapour phase inhibitors: These are used for the tem-
porary protection of new plant in transit or prior to
commissioning. Volatile corrosion inhibitors such as
cyclohexylamine derivatives are used. The plant must
be sealed or contained to prevent rapid loss of the
inhibitor. Sachets of these materials are placed in pack-
ing cases. Papers impregnated with them are available
for wrapping steel items. These inhibitors are used pri-
marily to protect steel. Further guidance is available
in BS 1133(6). An alternative is the use of desiccants
where adequate sealing can be achieved.
33.8 Corrosion monitoring
Once a plant is in operation it is important to monitor the
progress of any corrosion which might be taking place.
The four approaches described below vary in sophistica-
tion and cost. The most appropriate for any plant is deter-
mined by a number of factors, including the mechanisms
of corrosion which are anticipated and the implications of
catastrophic or unexpected failure. Key areas of the plant
require closer monitoring than readily replaceable items.
The measures described below do not replace the manda-
tory inspections of pressure vessels, etc. for insurance
purposes. The overall philosophy of corrosion monitor-
ing is to improve the economics of the plant's operation
by allowing the use of cheaper materials and generally
reducing the over-design that goes into plant to combat
corrosion.
33.8.1 Physical examination
Full records of all constructional materials that are used
in the plant should be maintained and updated when
repairs are undertaken. The exteriors of all parts of the
plant should be subjected to frequent visual examination
and the results reported and stored for future reference.
This maximizes the warning time before corrosion failures
occur, since the majority of failure mechanisms cause
leaks before bursting. Key items of plant, those in which
some degree of corrosion is anticipated and those which
might suffer catastrophic failure should be examined in
greater detail. Internal visual inspection during shutdowns
is sufficient to identify most corrosion effects. Cracking
can usually be seen with the naked eye but where cracking
is considered to be a possible mechanism an appropriate
non-destructive test method should be employed. In items
of plant which are shut down only infrequently (relative
to the timescale of possible corrosion or cracking failure)
external non-destructive testing is often possible.
Candidate non-destructive test methods include:
1. Ultrasonic techniques: Wall thickness can be mea-
sured to monitor the progress of general corrosion,
cracks can be detected and hydrogen blisters identified.
Certain construction materials such as cast iron cannot
be examined by ultrasound. Skilled operators and spe-
cialist equipment is required. Plant can be examined in
situ except when it is above 80°C.
2. Magnetic particle inspection: Surface emergent and
some sub-surface cracking can be detected in ferro-
magnetic materials. The technique must be used on
the side of the material in contact with the corrodent.
3. Dye penetration inspection: This is a simple technique,
requiring a minimum of operator training. In the hands
of a skilled operator it is capable of detecting fine
cracks such as chloride stress corrosion cracks in
austenitic stainless steels and fatigue cracks.
33.8.2 Exposure coupons and electrical
resistance probes
If changes have been made to the process (e.g. if incoming
water quality cannot be maintained or other uncertainties
arise) concerning the corrosion behaviour of the construc-
tion materials it is possible to incorporate coupons or
probes of the material into the plant and monitor their
corrosion behaviour. This approach may be used to assist
in the materials selection process for a replacement plant.
Small coupons (typically, 25 × 50mm) of any material
may be suspended in the process stream and removed at
intervals for weight loss determination and visual inspec-
tion for localized corrosion. Electrical resistance probes
comprise short strands of the appropriate material electri-
cally isolated from the item of plant. An electrical con-
nection from each end of the probe is fed out of the plant
to a control box. Instrumentation in the box senses the
electrical resistance of the probe. The probe's resistance
rises as its cross-sectional area is lost through corrosion.
The materials should be in the appropriate form, i.e.
cast/wrought/welded, heat treatment and surface condi-
tion. Metal coupons should be electrically isolated from
any other metallic material in the system. They should
be securely attached to prevent their being dislodged and
causing damage downstream. Simple coupons and probes
cannot replicate the corrosion effects due to heat transfer
but otherwise provide very useful information. It should
be noted that any corrosion they have suffered represents
the integrated corrosion rate over the exposure time. Cor-
rosion rates often diminish with time as scaling or filming
takes place, thus short-term exposures can give values
higher than the true corrosion rate.
33.8.3 Electrochemical corrosion monitoring
A number of corrosion-monitoring techniques, based on
electrochemical principles, are available. These give an
indication of the instantaneous corrosion rate, which is
of use when changing process conditions create a variety
of corrosion effects at different times in a plant. Some
techniques monitor continuously, others take a finite time
to make a measurement.
1. Polarization resistance: The current-potential behav-
iour of a metal, externally polarized around its cor-
rosion potential, provides a good indication of its
 Further reading 33/25

corrosion rate. The technique has the advantage of Corrosion of Stainless Steels, A. J. Sedriks, 2nd Ed, John Wiley &
being well established and hence reliable when used Sons, 1996.
Handbook of Corrosion Data, ASM International.
within certain limitations. This technique can only be Titanium Information Group www.titanium infogroup.co.uk
used for certain metals, to give general corrosion rate Materials Technology Institute of the Chemical Process Industries,
date in electrolytes. It cannot be employed to monitor Inc. 1570 Fishinger Road, Columbus, Ohio 43221, USA.
localized corrosion such as pitting, crevice corrosion or Lees, F.P., Loss Prevention in the Process Industries, Butter-
worths, London (1983).
stress corrosion cracking, nor used in low-conductivity Shreir, L.L. (ed.), Corrosion, 3nd edn Butterworth-Heinemann,
environments such as concrete, timber, soil and poor Oxford (1994).
electrolytes (e.g. clean water and non-ionic solvents). Corrosion Handbook, Dechema (1993).
Equipment is available commercially but professional Plant Corrosion: Prediction of Materials Performance, J. E. Strutt
and J. R. Nicholls, Ellis Horwood (1987).
advice should be sought for system design and location Corrosion Resistance Tables, P. A. Schweitzer, 4th edn, Marcel
of probes. Dekker (1995).
2. Impedance spectroscopy: This technique is essentially Corrosion of Stainless Steels, A. J. Sedriks, 2nd edn, John Wiley
the extension of polarization resistance measurements (1996).
Handbooks of Corrosion Data, ASM International.
into low-conductivity environments, including those
listed above. The technique can also be used to moni-
tor atmospheric corrosion, corrosion under thin films of Journals
condensed liquid and the breakdown of protective paint
coatings. Additionally, the method provides mechanis- British Corrosion Journal is published by the Institute of Materials
tic data concerning the corrosion processes which are Industrial Corrosion is the journal of the Institution of Corrosion
Science and Technology, PO Box 253, Leighton Buzzard, LU7
taking place. 7WB.
3. Electrochemical noise: A variety of related techniques Corrosion is the scientific journal of the National Association of
are now available to monitor localized corrosion. No Corrosion Engineers, PO Box 21 840, Houston, Texas, USA.
external polarization of the corroding metal is required, Materials Performance is the engineering journal of the National
Association of Corrosion Engineers.
but the electrical noise on the corrosion potential
of the metal is monitored and analysed. Signatures
characteristic of pit initiation, crevice corrosion and
Metal Producers' Trade Associations
some forms of stress corrosion cracking are obtained.
Nickel Development Institute
33.8.4 Thin-layer activation The Holloway
Alvechurch
This technique is based upon the detection of corrosion Birmingham
B48 7QB
products, in the form of dissolved metal ions, in the www.nidi.org
process stream. A thin layer of radioactive material is
created on the process side of an item of plant. As Copper Development Association
corrosion occurs, radioactive isotopes of the elements Orchard House
in the construction material of the plant pass into the Mutton Lane
Potters Bar
process stream and are detected. The rate of metal loss is EN6 3AP
quantified and local rates of corrosion are inferred. This
monitoring technique is not yet in widespread use but it Aluminium Federation Ltd
has been proven in several industries. Broadway House
Calthorpe Road
Birmingham
Further reading
Zinc Development Association
Dillon, C. P., Corrosion Control in the Chemical Process Industries, 42 Weymouth St
McGraw-Hill, New York, 2nd edition, 1994. London WIN 3LQ
Fontana, M. G. and Greene, N. D., Corrosion Engineering, 2nd
edn, McGraw-Hill, New York (1978) Titanium Information Group
The Forms of Corrosion Recognition and Prevention, Corrosion www.titaniuminfogroup.co.uk
Handbook No. 1, NACE, Houston, Texas (1982).
Corrosion Handbook, Dechema, 1993.
Plant Corrosion: Prediction of Materials Performance, J. E. Strutt Materials Technology Institute of the Chemical
& J. R. Nicholls, Ellis Horwood, 1987. Process Industries, Inc.
Corrosion Resistance Tables, P.A. Schweitzer, 4th Ed, 1995, 1570 Fishinger Rd,
Marcel Dekker. Columbia, OH 43221, USA.