POTENTIAL DEGRADATION MECHANISMS LIBRARY

TABLE OF CONTENTS

1 AQUEOUS (ELECTROLYTIC) CORROSION .............................................................................. 2

1.1 Corrosion by Acids ................................................................................................................. 2
1.2 Corrosion by Acid Salts ......................................................................................................... 4
1.3 Corrosion by Carbon Dioxide .............................................................................................. 5
1.4 Corrosion by Caustics ........................................................................................................... 6
1.5 Corrosion by Chlorides .......................................................................................................... 7
1.6 Corrosion by Oxidants ........................................................................................................... 9
1.7 Corrosion by Water .............................................................................................................. 10
2 HIGH-TEMPERATURE EFFECTS ............................................................................................... 13
2.1 Mechanical Effects.............................................................................................................. 13
2.2 Chemical Effects .................................................................................................................. 15
3 Corrosion Under Insulation .................................................................................................. 23
3.1 Introduction........................................................................................................................... 23
3.2 Location for CUI ................................................................................................................... 24
3.3 Causes ................................................................................................................................... 25
3.4 Mechanism ........................................................................................................................... 26
3.5 Types of Corrosion Under Insulation ................................................................................. 26
3.6 Prevention ............................................................................................................................. 28
3.7 Monitoring and Inspection ................................................................................................. 29
3.8 Specification ......................................................................................................................... 31
3.9 Conclusions ........................................................................................................................... 31
4 Erosion-Corrosion ................................................................................................................. 32
4.1 Introduction........................................................................................................................... 32
4.2 Mechanism ........................................................................................................................... 33
4.3 Erosion Assessment .............................................................................................................. 33
5 Microbiologically-Induced Corrosion ............................................................................... 34
5.1 Introduction........................................................................................................................... 34
5.2 Mechanisms of MIC ............................................................................................................. 35
5.3 Microbiologically-Induced Corrosion Assessment......................................................... 36
6 Corrosion in Boiler System ................................................................................................... 41
6.1 Introduction........................................................................................................................... 41
6.2 Normal Boiler Conditions .................................................................................................... 41
6.3 Damage Mechanisms in Boiler System ............................................................................ 42
6.4 External Corrosion and Erosion .......................................................................................... 47
6.5 Corrosion control.................................................................................................................. 48
6.6 Monitoring and inspection ................................................................................................. 48
7 Corrosion in Cooling Water System ................................................................................... 49
7.1 Introduction........................................................................................................................... 49
7.2 Common Degradation Mechanism to Cooling Water System .................................. 49
7.3 Corrosion Control ................................................................................................................. 51
7.4 Carbon Steel ......................................................................................................................... 52
7.5 Carbon Steel in Cooling Tower ......................................................................................... 53
7.6 Stainless Steel Corrosion...................................................................................................... 53
7.7 Monitoring ............................................................................................................................. 54
 1
Chapter

1 AQUEOUS (ELECTROLYTIC) CORROSION

1.1 Corrosion by Acids

1.1.1 General
Acids are often classified as oxidizing or reducing. Some acids can show more than
one behavior, depending on concentration and/or temperature. Materials
selection for acids and their derivative compounds depends in part on whether
they are reducing or oxidizing. In addition, the corrosivity of the solution often
depends on the presence of strong oxidizing salts such as ferric chloride (FeCl3) or
cupric chloride (CuCl2). Both salts are also strong pitting agents. Such oxidizers are
sometimes present as contaminants. Particularly in reducing acids, corrosivity can
be dominated by aeration and/or the presence of oxidizing contaminants.

In general, oxide-stabilized corrosion resistant materials perform well in the

presence of oxidizing acids. For alloys, it is a general rule that the higher the alloying
content, particularly for chromium, the higher the concentration and temperature
limits for which the alloy will be suitable.

Increasing the Ni content increases resistance to chloride pitting and to chloride

stress corrosion cracking. Increasing the Mo content reduces susceptibility to
localized corrosion phenomena such as pitting and crevice corrosion.

Alloys designed to operate well in reducing acids may perform poorly in such acids
if they are aerated or contain oxidizing contaminants. Accordingly, it is important
to determine if reducing acids will contain oxidizing contaminants or will be
aerated.

When austenitic stainless steels are selected for acid service, it is conventional to
specify the low carbon grade, that is, the "L" grade. Many acids will attack the
sensitized band in the weld heat affected zones of the conventional grades.

In surveying materials for a specific application, keep in mind the following

alternatives:

 Non-metallic materials such as fiber-reinforced plastics are available for both

piping and for most equipment such as vessels, tanks and pumps.
 Liners such as rubber, polymer or glass are frequently cost effective. Plastic-
lined piping is a common choice in acid systems.
  Cladding and/or weld overlays, using carbon or low-alloy steel for pressure
containment and a corrosion resistant alloy for corrosion resistance, are
sometimes used.

1.1.2 Inorganic Acids

Hydrochloric, hydrofluoric, nitric, sulfuric and phosphoric acids account for most of
the strong inorganic acids encountered in hydrocarbon and chemical process
plants. Hydrochloric, hydrofluoric and phosphoric acids are reducing. Nitric acid is
a strong oxidizer. Sulfuric acid can be either oxidizing or non-oxidizing, depending
on temperature and concentration. Sulfuric acid becomes increasingly oxidizing
at concentrations of about 25 wt. percent and higher. Below 25 percent, the
uncontaminated acid is regarded as reducing.

Because of their relatively severe corrosiveness, many of the inorganic acids are
difficult to handle with alloys. In addition, use of highly alloyed materials
substantially increases both capital and installed costs. For these reasons, it is
common for materials selection to include plastics (including fiber-reinforced
plastics), elastomers, linings and coatings. Carbon and graphite also find use in
some severe applications.

Fluoropolymers are usually very resistant to most inorganic acids, but may be
permeable. Often, less expensive plastics are suitable. Equipment constructed of
lined carbon steel is often selected. Candidate linings include rubber, plastic,
resistant paint coatings (if backed up with cathodic protection) and glass. Plastic
lined piping is regarded as the normal choice for many industrial applications.

1.1.2.1 Sulfuric Acid

Carbon steel is normally used for storage tanks and sometimes for piping for sulfuric
acid at concentrations of 70 wt. % and above, at temperatures up to 40°C. Typical
industrial concentrations are 93 and 97 %. The selection of carbon steel depends
on controlling velocities to less than about 3 ft/sec (0.9 m/s). The velocity limitation
is critical, since successful use of carbon steel depends on not disturbing the
protective, but non-adherent, soft, insoluble iron sulfate scale layer. Linings and
anodic protection are also specified, sometimes from concerns over product
purity. Note that some design and construction details can be important. Examples
include avoiding accidental entry of water and proper pre-commissioning
cleaning. Refer to NACE RP0391 "Materials for Handling and Storage of
Concentrated (90 to 100 %) Sulfuric Acid at Ambient Temperatures" for further
details.

Hydrogen damage can occur in carbon steels in sulfuric acid services.

 Hydrogen induced cracking (HIC) damage, including blistering, has been

observed in tanks and other components built of plate. Both of these
phenomena can be minimized by using clean carbon steels of the type
recommended for HIC resistance.
  Hydrogen grooving has been observed in carbon steel tanks in sulfuric acid
service. This phenomenon is caused by bubbles of H2 eroding the protective
scale covering from the steel, permitting continued attack of the substrate.
This problem is usually associated with manholes or nozzles and can be
minimized by use of corrosion-resistant linings or alloys.

Type 3I6L SS, rather than carbon steel, is becoming the standard metallic
material of construction for piping concentrated sulfuric acid for
temperatures up to about 27°C.

Lower concentrations of acid or higher operating temperatures require the use of

more resistant materials. Type 3I6L SS can be used, in conjunction with anodic
protection. However, applications are limited because of the possible
consequences of failure or disruption of the anodic protection system.

In common with most mineral acids, sulfuric acid in various concentrations and
temperatures can be handled by fiber-reinforced plastics, liners such as rubber
(e.g., neoprene), polymers such as polypropylene and glass and plastic-lined pipe.

1.2 Corrosion by Acid Salts

The corrosion mechanism by a presence of acid salts is similar to corrosion by
respective acid. Acid salts is the product of reaction between weak base and a
strong acid; the most common are NH4Cl and FeCl3. These salts will produce an
acidic solution when dissolved in water and with a suitable exposed temperature
can become concentrated. Depending on conditions and environmental
parameters such as temperature, acid salt deposits can form and absorb water
that may lead to under-deposit corrosion such as pitting and may be stress
corrosion cracking.

If the acid salts contain the anionic species such as sulfide or other cathodic
poison, cracking mechanisms such as hydrogen stress cracking or hydrogen
induced cracking can be significantly accelerated. Acidic chloride salts such as
NH4Cl can cause not only under-deposit pitting, but also cause chloride stress
corrosion cracking of austenitic stainless steels under conditions that would
normally be considered benign. A wet chloride acid salt deposit can reduce the
under-deposit pH to very acidic values, with a very high chloride concentration. In
addition, the deposit acts as a diffusion barrier, leading to oxygen depletion
beneath the deposit. Galvanic effects from the active-passive cell can accelerate
corrosion mechanisms. This situation can cause stress corrosion cracking at
temperatures well below the 60°C threshold characteristic of the austenitic stainless
steels in neutral saline waters.

Acidic salt solutions will act as weak mineral acids, causing accelerated general
pitting corrosion. These solutions also tend to destabilize otherwise protective scale
formations. In non-turbulent regions, this phenomenon usually causes localized
accelerated pitting, with the production of large quantities of loose, soft scale. In
turbulent areas, erosion corrosion is usually the destructive mechanism.

Most acid salts are hygroscopic, that is, they can absorb water vapor. This results in
two conditions under which salt deposits may cause corrosion problems in
nominally "dry" systems.
 In a water-saturated vapor system, salt deposits may absorb enough water vapor
to produce a wet spot under the deposit, leading to severe underdeposit pitting.
This phenomenon is a cause of pitting failures in carbon and low-alloy steels, but is
particularly common with oxide stabilized alloys such as stainless steels. In austenitic
stainless steels, this can also cause accelerated chloride stress corrosion cracking,
as described above.

Particularly in heat exchangers, the metal temperature under the deposit may be
below the water dew point of the otherwise "dry" system. A hygroscopic salt deposit
may form an aggressive corrosion cell under such conditions.

Neutralization, via injection of a neutralizer such as caustic soda or a neutralizing

amine, is sometimes used to control the problems caused by acid salt solutions.
However, water washing is probably the most common mitigation method used to
dilute and dispose of harmful acid salt solutions and deposits. Note that the wash
water must be free of dissolved oxygen, or destructive corrosion will occur. A
concentration of 50 ppbw dissolved oxygen is usually an acceptable limit. In a
refinery, stripped sour water is often used as wash water because of its very low
oxygen content.

For carbon steels, the recommended concentration threshold for safe operation is
usually on the order of 2 to 3 wt. percent. Some users accommodate up to 8 wt.
percent in carbon steel. Concentrations exceeding 10 percent are considered
destructive to carbon steel even at low velocities.

Corrosion susceptibility for common materials is as follows:

 Carbon steel : Most susceptible

 Aluminum : S usceptible
 Stainless steel (300-series): velocity sensitive; can be stress corrosion cracked
by chloride excursions

1.3 Corrosion by Carbon Dioxide

Wet carbon dioxide (carbonic acid, a weak inorganic acid) can cause severe
pitting and/or grooving in carbon and low-alloy steels. A rule of thumb is that
carbon steel is usually acceptable for wet CO2 if the CO2 partial pressure is less
than about 4 psia (27 kPa). A corrosion allowance of up to ¼" (6.4 mm) is usually
specified. However, it is easy to estimate the corrosion rate of carbon steel, using
the deWaard-Milliams nomograph. If the estimated corrosion rate is unacceptable,
consider the use of corrosion inhibitors, increase the corrosion allowance or use
high alloyed steels.

The deWaard-Milliams nomograph is based on corrosion rates measured in

carbonic acid, for clean steel surfaces. Surfaces protected by scales such as mill
scale (Fe3O4) or other surface deposits are usually at least partially protective. In
addition, these rates are valid only for non-turbulent systems. Thus, the rates
predicted by the nomograph can be influenced by several factors.
  Surface scales produced by carbonic acid corrosion such as FeCO3 can
result in significantly reduced corrosion rates. Protection by such scales is
influenced by several factors, including temperature, pH and velocity.

 Experimental and field data indicate that nomograph rates are unreliable
for systems in which the carbonic acid is condensing, that is, in systems
involving the formation of dew point water. The nomograph rates are
therefore needs to be derated by a multiplier of one-tenth.

 At higher temperatures, the wet CO2 corrosion rate begins to decrease due
to the formation of a protective corrosion scale. Hence, the rates estimated
for design conditions may actually be less than the rates estimated for
operating conditions. Thus, the user should check the rates under operating
conditions before determining the basis for materials selection.

 Carbon steel protected by mill scale or other surface deposits may corrode
at rates substantially less than those predicted by the de Waard-Milliams
nomograph. However, such scales (particularly mill scale, Fe3O4) may be
susceptible to slow dissolution by carbonic acid, eventually resulting in
accelerated corrosion at rates in accordance with the de Waard-Milliams
diagram. During the dissolution of mill scale by carbonic acid, the carbon
steel surface usually develops a characteristic appearance, leading to the
descriptive term "mesa" corrosion.

 In process streams containing oxygen or other cathodic depolarizers,

carbonic acid pitting rates may be much higher than predicted by the de
Waard-Milliams nomograph.

 In turbulent systems, the rate may be greater than 1 in. per year (25 mm/yr).

1.4 Corrosion by Caustics

Corrosion-induced metal loss by caustics is uncommon. However, a localized form
of pitting, called caustic gouging, can occur with carbon steel, particularly in high-
temperature services such as boiler systems. Such systems typically have some form
of caustic-based water treatment program. Debris or deposits permit the formation
of under-deposit solutions of hot concentrated sodium hydroxide, which attack the
substrate steel. The key to preventing this form of pitting is to chemically clean the
caustic-exposed system before commissioning and then keep the system clean
during operation. Modern phosphate-based treatments avoid this problem.

Metals and alloys that can form amphoteric hydroxides are subject to accelerated
corrosion in alkaline environments. (Amphoteric hydroxides are soluble in alkaline
solutions.) Aluminum and zinc are the most common metals exhibiting this
behavior. Their use in alkaline processes is usually not recommended. Likewise, their
use in buried or submerged alkaline environments can lead to rapid metal loss.

The most common corrosion problem involving caustics is alkaline stress corrosion
cracking. The threshold concentrations and temperatures for caustic soda service
can be obtained from Figure 1.1.4
 Figure 1.1.4: Graph for Caustic Soda Service. (Corrosion Data Survey, Metals Section, NACE
International, Houston, 1985)

1.5 Corrosion by Chlorides

Aqueous chlorides provide an excellent electrolytic environment for corrosion.
However, ambient temperature neutral chloride solutions are not particularly
aggressive to carbon and low-alloy steels. Acid aqueous chlorides, below a pH of
about 4.5, can be very aggressive to such steels.
Carbon steel, unprotected by coatings or cathodic protection, usually provides a
useful life of at least several years in water with chlorides at concentrations of 3.5
to 5.5 wt. %. Pitting rates for this materials is relatively low, usually on the order of 3-
5 mpy (0.08-0.1 mm/yr).

In environment with pH is below about 4.5, general corrosion in the form of small,
closely spaced pitting is the normal form of attack. The mechanism is essentially
one of mild acid attack. The presence of aeration (at least 1 ppmw dissolved O2)
accelerates corrosion due to the action of dissolved oxygen as a cathodic
depolarizer. Under-deposit corrosion, which is a form of concentration cell
corrosion, can be very severe in aerated waters with low pH and chlorides. Velocity
can also accelerate the rate of metal loss due to erosion corrosion.

Chlorides in water can disrupt scales that would otherwise protect the substrate
steel. Pitting corrosion is the normal form of attack. If the scale is mill scale (Fe3O4)
pitting rates can be quite severe, since mill scale is both cathodic with respect to
the substrate steel and is a relatively good conductor. (Unlike most scales, mill scale
is not dielectric.) The threshold concentration for disruption is somewhere between
about 50 and 500 ppmw chloride. The threshold is affected by stream velocity, with
15 ft/sec (4.6 m/s) being considered safe for even concentrated brines.

Aeration greatly accelerates pitting rates. Aeration in chloride-bearing waters

poses a high risk of under-deposit pitting corrosion. Consequently, a minimum flow
velocity of at least 2.5 ft/sec (1.5 m/s) should be maintained to keep entrained
debris from settling to form a deposit. The minimum flow velocity may have to be
increased in streams having settled debris. Designs should avoid deadlegs.

Stainless steels, both the 12 Cr and austenitic grades, can be susceptible to pitting,
under-deposit and crevice corrosion in aerated chloride-containing waters. Free-
flowing, clean chloride solutions such as seawater permit the use of Type 316 SS
(Type 316L SS if welded) if the temperature is not allowed to exceed 60°C. The latter
temperature is the threshold temperature for chloride stress corrosion cracking of
Type 316 SS in neutral chloride-containing waters. See Figure below:
 Figure 1.1.5: Chloride stress corrosion cracking of Type 304 and Type 316 stainless steels as
a function of chloride concentration and temperature. (D.R. McIntyre, Experience Survey,
Stress Corrosion Cracking of Austenitic Stainless Steels in Water, MTI Publication No.27,
Materials Technology Institute, St Louis, 1987.

Austenitic stainless steels are much more susceptible to chloride damage. Because
of the potential galvanic couple that can exist between active and passive
stainless steel surfaces, chloride-induced damage can occur rapidly in stainless
steel, for example, crevice corrosion. Microbiologically induced corrosion is also a
threat in improperly layed-up stainless steel piping and equipment. Chloride stress
corrosion cracking is usually not a concern, since the exposure temperature is less
than 60°C.

The 50 ppmw chloride concentration threshold, as with many rules of thumb, can
be misleading. Low-chloride water («50 ppmw) can cause serious corrosion if such
pocket of chloride-containing water can concentrate chlorides by evaporation. In
the presence of air, which acts as a cathodic depolarizer, such pockets can cause
corrosion problems.

1.6 Corrosion by Oxidants

Strong oxidants such as chlorine and oxygen are usually handled in carbon steel
at ambient temperatures, if dry. They require special alloys for higher temperature
use. Carbon steel is permitted at temperatures up to 150°C for chlorine and for
temperatures up to 65°C for oxygen. However, greases, oils and other
hydrocarbons can spontaneously ignite in such services.
1.7 Corrosion by Water
Pure water is non-corrosive to ordinary engineering materials. However, most
waters contain dissolved salts and may contain dissolved oxygen and/or acid
gases such as CO2 or H2S. Such waters are often quite corrosive.

Anaerobic water may become corrosive due to the presence of anaerobic

microbes such as sulfate-reducing bacteria (SRB) or similar bacteria. Sulfate-
reducing bacteria metabolize dissolved sulfates and generate H2S as a metabolic
product. SRB activity is most commonly observed in anaerobic seawater. Refer to
the section below for more detail '"Microbiologically Induced Corrosion".

Dissolved oxygen plays a key role in determining the corrosivity of a water. Oxygen
acts as a cathodic depolarizer, not as a corrodent. Experience has shown that
water that contains less than about 10 ppbw dissolved oxygen is essentially
deaerated. Waters containing less than about 200 ppbw are usually not corrosive
unless they contain acid gases or have pHs less than about 5. Neutral waters
containing from 200 ppbw to about 1 ppmw are mildly corrosive to carbon steel.

Corrosion rates are usually on the order of 3-5 mpy (0.08-0.1 mm/yr). Waters
containing oxygen in excess of one ppmw should be considered corrosive and
subject to chemical treatment or some other mitigation measure.

In many facilities, the source water for the fire water system is aerated. However, if
the fire water system is stagnant, the dissolved oxygen is soon scavenged by the
carbon steel piping, making the water non-corrosive. Again, if this water contains
dissolved sulfates (such as seawater), SRB can become a major problem.

In aerated waters, dissolved salts and/or velocity control corrosivity. If the dissolved
salts promote the formation of a dense scale, the water is usually noncorrosive. (The
scale layer acts as a barrier to oxygen diffusion and promotes rapid polarization of
the anode surface by anion saturation.)

While scale formation may favorably reduce corrosivity, it can cause other serious
problems such as plugging and "hot spots" in heat transfer equipment. If a water is
determined to be "scaling," the process should be analyzed to determine if the side
effects are acceptable.

The velocity of an aerated water affects its corrosivity. Some alloys such as
admiralty brass are limited in their use because of their sensitivity to velocity in
corrosive waters. In some systems, low-velocity components such as piping may be
made of carbon steel. Higher velocity components such as pumps and control
valves may require alloys such as 12 Cr SS.

Waters with low total dissolved solids (TDS) have a wide range of corrosivity. Very
pure waters, such as distilled and totally deionized waters, are non-corrosive if
uncontaminated. They can be very corrosive if they absorb an acid gas such as
carbon dioxide, which can rapidly reduce the pH of the water. This problem can
be severe in boiler systems if the steam condensate is allowed to absorb CO2 in
the condensate handling and storage system. (The source of CO2 is usually
carbonate salts in the boiler feed water, which decompose in the steam system
and carry over into the condensate system). Partially deionized waters such as
waters that have anions but no cations (other than hydrogen) can be very
corrosive. Such waters are uncommon. Waters containing high concentrations of
dissolved salts also become more corrosive if they absorb acid gases. However,
the presence of the dissolved salts buffers the water, preventing a rapid reduction
in the pH value. Thus, high-TDS waters containing dissolved CO2 are much less
corrosive than low-TDS waters that have absorbed CO2.

Waters that have absorbed both oxygen and an acid gas such as CO2 or H2S can
be very aggressive, whether the water is hard or soft. Such waters must be
chemically treated if they are to be rendered non-corrosive to carbon steel. In
some cases, other measures such as alloys, paints, coatings, linings or cathodic
protection are employed.

Galvanic couples or crevices can aggravate the local corrosivity of water. In such
cases, the key issue is the presence or absence of dissolved oxygen. If there is
insufficient oxygen to act as a cathodic depolarizer, the couple or crevice will
polarize and will not be a serious problem. Galvanic couples in aggressive water
should be examined to ensure that unfavorable anode/cathode area problems
are mitigated. Galvanic couples that are also crevices can be particularly
vulnerable to accelerated corrosion. For example, titanium tubes have caused
severe crevice corrosion in alloy tubesheets in aerated seawater.

Materials selection for waters can be a real challenge. In most plants, it is safe to
assume the cooling, utility and fire waters to be non-corrosive, because of either
chemical treatment or because of design. An example of the latter is a fire water
system that is stagnant and not subject to SRB. However, prudence requires the
determination the corrosivity of the various waters to be handled. In the event that
corrosivity is a concern, mitigation measures include chemical treatment, cathodic
protection, alloys, linings such as cement-lined pipe, and paints or coatings. Some
general rules for materials selection follow.

For non-corrosive, mildly corrosive or chemically treated water, carbon steel is the
normal material of construction. Galvanized carbon steel is often recommended
for moderately corrosive waters.

For corrosive, untreated waters, the common alternatives are:

 Carbon steel with corrosion protection such as the use of coatings, linings
and/or cathodic protection.

 Corrosion-resistant alloys.
o The austenitic stainless steels are usually not employed in corrosive
waters because off the dangers of chloride-induced pitting and
stress corrosion cracking. However, Type 316 SS (Type 3l6L if welded:
CF-3M for castings) has found extensive successful use in freely
flowing, clean aerated seawater for temperatures below 60°C.
o Cu-Ni alloys are often used in corrosive waters or relatively low
velocity applications such as heat exchanger tubes.

 Corrosion-resistant non-metallic such as polyethylene, PVC, or fiberglass.

o Fiberglass is commonly used for piping and vessel shells in corrosive
water service. Note that none of the non-metallic materials are
suitable for services in which they are likely to be mechanically
 abused. Examples include pipe movement, water hammer and
hydraulic surges.
o PVC, CPVC and polyethylene piping are commonly used in
corrosive waters.

There are other measures that can be used to prevent corrosion:

 Cement linings are usually successful. Select the cement with due regard for
the pH expected of the water. Note that cement linings are usually not
recommended for waters with pH < 6.

 Coatings (e.g., coal tar) are often used, in conjunction with internal cathodic
protection, for tanks and vessels in corrosive water service.

 In deaerated water storage tanks, inert gas blanketing can be used

 2
Chapter

2 HIGH-TEMPERATURE EFFECTS

2.1 Mechanical Effects

2.1.1 Introduction
High-temperature tends to force the selection of expensive materials of
construction. Whenever possible, the materials selection engineer should review
the design data or design basis to see if there is opportunity to justify reducing the
maximum design temperature.

2.1.2 Creep
Most metals and alloys exhibit a temperature above which the grain boundaries
become weaker than the grains themselves. This temperature is the threshold
temperature above which the material is susceptible to creep.

For metals and alloys at temperatures less than their creep thresholds, strain is not
time-dependent for constant stress. However, at temperatures above the creep
threshold, creep can occur. Creep is defined as time-dependent strain at constant
stress, or, stated another way, the strain rate is greater than 0 for a stress rate of 0.
In the creep range, stresses above the creep threshold cause the nucleation and
propagation of grain boundary voids. Figure 2.3A shows an idealized
representation of the three stages of creep.

Figure 2.3A: The three stages of creep.

 In primary creep, the material plastically deforms while undergoing rapid work
hardening. No significant damage is generated. During secondary creep, grain
boundary yielding produces local work hardening and the nucleation of grain
boundary voids. During tertiary creep, the grain boundary voids link up and gross
necking or thinning develops. Tertiary creep ends in an exponentially accelerating
strain rate, rapidly leading to fracture. Fracture by this phenomenon is called stress
rupture.

The lower (threshold) end of the creep temperature range for carbon steel is about
400°C. The Cr-Mo steels have creep threshold temperatures of about 480°C and
higher. The conventional austenitic stainless steels have creep threshold
temperatures of 565 to 595°C. A safe estimate for the creep threshold temperature
of a material is the upper temperature limit permitted by ASME Section VIII, Div. 2
[1]. Some engineering codes such as ASME B31.3 [2] for piping and ASME Section
VIII, Div. 1 [1] for vessels contain provisions for creep design.

If creep is a concern, coarse-grained materials are favored. Carbon steels killed

with silicon are usually recommended for temperatures where creep is a concern.
Examples include ASTM A106 for pipe and ASTM A515 for plate.

2.1.3 Stress Rupture

If sustained maximum operating temperatures will create significant creep strain,
the component is at risk of failure by stress rupture, typically associated with tertiary
creep. Stress rupture design includes "safety" factors intended to define inspection
intervals and/or schedule replacement before the onset of tertiary creep. In typical
stress rupture designs, the design life of the component is the expected period of
secondary creep. During secondary creep, component dimensions such as tube
diameter and length slowly increase. In wrought materials, secondary creep strains
of 10 to 20 percent are not uncommon.

When operating in the creep range, care must be taken to avoid temperature or
stress excursions; of the two, temperature excursions are by far the more damaging.
A 28°C excursion can easily reduce the operating life by 50 percent or more.
Furnaces and heaters are about the only equipment having both temperatures
and stresses high enough to require creep to be taken into account during plant
design.

Thermal fatigue produces fractures that are virtually identical to creep failures.
Maximum code-allowable stresses are high enough to permit thermal fatigue.
Accordingly, if thermal cycles are a feature of equipment design, thermal fatigue
analysis is usually recommended.

2.1.4 Spheroidization and Graphitization of Carbon Steels

Spheroidization and graphitization can occur in carbon and C-½Mo steels at high-
temperature. Low-alloy steels with chromium contents of about 0.7 wt. percent or
more are considered immune to these effects. Spheroidization and graphitization
occur at temperatures above 455°C. Since most users avoid the selection of
carbon steels for use above 425°C, the effects of these two mechanisms are often
not considered. However, many plants use carbon steel lined with refractory for
high-temperature services. Refractory failures occasionally expose these carbon
 steels to temperatures substantially in excess of 425°C. Selecting carbon steel for
such services does carry a risk, so understanding what could happen is important.

Iron carbide (cementite: Fe3C), one of the primary components of carbon steel, is
thermodynamically unstable. However, its rate of decomposition in carbon and C-
½Mo steels is negligible at temperatures less than 455°C. The rate of decomposition
begins to accelerate at temperatures exceeding 455°C. Like most high-
temperature, diffusion-controlled phenomena, the rate is exponentially related to
temperature. At 480°C, 50 percent conversion to graphite occurs in about 10,000
hours. At 595°C, the conversion time is only about 1000 hours.

The process of iron carbide decomposing to form iron and graphite is called
graphitization. Decomposition is accompanied by a moderate reduction in the
strength of the steel. This reduction will accelerate the formation of creep damage
if the applied stress is large enough to cause creep. Decomposition of the iron
carbide can "embrittle" the steel if the graphite that develops forms a continuous
(or closely spaced discontinuous) band. Ruptures have occurred from this cause,
most frequently in C-½Mo steels. This alloy is no longer being recommended as a
material of construction for high-temperature services.

Aluminum-killed steels are more susceptible to graphitization than are silicon-killed

steels such as ASTM A515. Silicon-killed steels are preferred for high-temperature
services.

Spheroidization refers to the formation of spheroids of iron carbide from the normal
microstructure, pearlite. The mechanism occurs at temperatures above 480°C,
again at rates which are exponential with temperature. However, for sustained
high-temperatures, graphitization becomes the dominant mechanism.

Unless carried to extremes by prolonged exposures at high-temperatures,

spheroidization is often regarded as beneficial, since it greatly improves the impact
toughness of carbon steel with only a minor loss of strength. Normalizing a pearlitic
carbon steel cause partial spheroidization, resulting in improved toughness. Too
much spheroidization will cause an unacceptable loss.

2.2 Chemical Effects

2.2.1 Carburization
Carburization refers to the development of a carbide-rich layer on the surface of
a material exposed to a reducing hydrocarbon environment. This phenomenon is
associated with high-temperature service or, in some cases, to high-temperature
excursions. Carburization of carbon and low-alloy steels is rare since they are not
normally subjected to operating temperatures high enough to induce
carburization. Mild carburization of ordinary 300-series austenitic stainless steels is
sometimes observed since they can be used at temperatures high enough to see
low rates of carburization.

Carburization can cause premature failures or contribute to such failures. Failure is

often caused by cracking due to the large difference in the coefficients of thermal
expansion between the parent alloy and the carburized layer. Such cracking
 causes the carburized layer to disbond, thereby exposing fresh material to
subsequently carburize. Thermal cycling is the normal cause of such failures.

Metal loss is the form of failure in a carburization mechanism known as metal

dusting, which can occur very rapidly. This very limited mechanism involves process
streams with CO/CO2 ratios on the order of 3 to 5, at temperatures in the range
650-845°C, usually involving Fe-Cr and Fe-Cr-Ni alloys having a chromium content
of 25 percent or less. Metal dusting usually causes smooth, circular pits, typically
worst in stagnant areas. In some cases, pitting damage is general, resulting in
overall surface wastage.

When selecting materials for high-temperature hydrocarbon services, the potential

for carburization should be determined. Process licensors usually provide excellent
guidance for materials selection. Alloy manufacturers can also provide excellent
advice. If carburization is anticipated, it is normal practice to provide a nominal
carburization allowance. Do not attempt to mitigate metal dusting with a corrosion
allowance; a change in alloy or in the process is required for this mechanism.

2.2.2 Hydrogen Gas

For the purpose of materials selection, hydrogen service is defined as any service
in which the partial pressure of the hydrogen gas exceeds 100 psia (0.7 MPa).
Hydrogen gas can cause two types of problems: hydrogen embrittlement and
hydrogen attack.

2.2.2.1 Hydrogen Embrittlement

As discussed earlier, hydrogen gas can cause carbon and low-alloy steels to be
embrittled at temperatures ranging from sub-ambient to about 120°C. However,
hydrogen gas itself is not a problem, since it cannot dissolve or diffuse in metal. The
problem is nascent (atomic) hydrogen. At even ambient temperatures, carbon
and low-alloy steels can dissolve nascent hydrogen from hydrogen gas. The
amount of nascent hydrogen capable of dissolving from gaseous hydrogen is
normally quite small, since its concentration in the metal must be in equilibrium with
the concentration of nascent hydrogen in the gas. The latter concentration is very
small except at high-temperatures and high hydrogen partial pressures. As a rule
of thumb, gaseous hydrogen at temperatures less than about 220°C cannot
provide enough nascent hydrogen to embrittle carbon or low-alloy steels.

As is discussed in the following section, carbon steels are not selected for high-
temperature, high-pressure hydrogen services. Accordingly, they are not
susceptible to hydrogen embrittlement by hydrogen gas, unless they are
improperly exposed. As a result, postweld heat treatment is normally not required
for carbon steels in hydrogen service.

Gaseous hydrogen service can cause hydrogen embrittlement in straight

chromium stainless steels and low-alloy steels (including the Cr-Mo steels, which are
favored for high-temperature, high-pressure hydrogen gas service). Hydrogen that
dissolves in the steel at high-temperatures can embrittle the steel upon cool down,
if cooling is too fast to permit the escape of excess hydrogen as the metal cools.
Note that such steels have a very low solubility for hydrogen at temperatures below
about 205°C. Weld repair requires bake out and preheat.
2.2.2.2 Hydrogen Attack
Hydrogen gas can cause surface decarburization as well as internal
decarburization and fissuring. (The latter is called hydrogen attack in carbon and
low-alloy steels.) These types of deterioration involve exposure to high temperature
services having high hydrogen partial pressures. Since surface decarburization
causes only a slight reduction in material strength, it is not normally regarded as a
problem.

Hydrogen attack can occur at temperatures above about 220°C. Dissolved

hydrogen can attack iron carbide (Fe3C-cementite), generating methane gas
(CH4), which is trapped in the metal because the methane molecule is too large
for diffusion. Attack is usually at grain boundaries. As the concentration of methane
gas increases, increasing pressure begins to tear the grain boundary apart, causing
first fissures, then cracks, then networks of cracks. Simultaneously, the loss of
carbides lowers the strength of the material. Combined, the two effects can
substantially reduce the expected life of a component. Both chromium and
molybdenum form more stable carbides, leading to a preference for Cr-Mo alloy
steels in hydrogen service. Unless it severely cold worked, austenitic stainless steels
are unaffected by hydrogen attack.

The Nelson curves are used to select materials that will be immune to hydrogen
attack in gaseous hydrogen service. Refer to API Publication No. 941 "Steels for
Hydrogen Service at Elevated Temperatures and Pressures in Petroleum Refineries
and Petrochemical Plants". Because of the "scatter" in the Nelson curve data, it is
common to use the maximum operating temperature plus a 14°C margin when
selecting materials using the Nelson curves. Some users and process licensors
specify the use of a 28°C margin. Make sure that the maximum design temperature
is large enough to include the maximum operating temperature plus the selected
margin. (Note that if the maximum design temperature is used for materials
selection, the use of an additional operating temperature margin should be
unnecessary.)

For the vertical portion of the curves, it is customary to use a 25 or 50 psia (170 to
345 kPa) margin on the maximum operating hydrogen partial pressure. Make sure
that the maximum design pressure is large enough to include the maximum
operating pressure plus the selected margin.

Hydrogen attack is accelerated by inclusions and slag-type defects. Therefore,

killed steels are selected. Inclusion-free welds are often specified. To further protect
materials exposed to hydrogen service, it is common industry practice to impose
weld metal hardness controls. Postweld heat treatment is recommended for all Cr-
Mo alloy steels. Components cold worked more than 5 percent should be stress
relieved. Welded attachments such as reinforcing pads should be vented.

Except at high-temperatures and high hydrogen partial pressures, there is a

significant incubation time before hydrogen damage becomes detectable. Thus,
in situations where the metal is exposed to infrequent and short-term transient
combinations of high-temperature and moderate hydrogen partial pressure, there
may be a significant incubation time before the effects of such attack become
detectable. Investigation of incubation times can often justify the choice of a
 lower-cost material of construction. Refer to API Publication No. 941 for details on
incubation times.

To summarize, when selecting materials for hydrogen service:

 The Nelson curves utilizing the maximum operating temperature plus 14°C
should be used.
 Carbon steels should be fully killed or otherwise deoxidized.
 Low-alloy steels such as the Cr-Mo steels should be postweld heat treated.
 Cold-worked materials should be stress relieved.
 Seamless tubing and pipe are preferred, as they avoid potential problems
associated with longitudinal welds.
 Hardness controls should be employed:
o NACE RP0472. The maximum weld metal hardness permitted for
carbon steel is 200 BHN. Weld procedure qualification testing is
done to ensure that heat-affected zone hardness do not exceed
248 VHN.
o NACE MR0175. It is industry practice to limit the weld metal hardness
of Cr-Mo low-alloy steels (225 BHN for Cr < 3 and 241 BHN for 3 < Cr
< 9). NACE MR0175, which limits the hardness of parent metals and
heat affected zones, should be required.

Heat affected zone hardness should be shown to be satisfactory in the Welding

Procedure Qualification Record via a hardness traverse across the weld metal-
parent metal interface. Since proper use of the Nelson curves will prevent
hydrogen attack on materials of construction, hydrogen gas is regarded as a
crack-inducing agent rather than as a corrodent. As discussed before, hydrogen
gas service involving the Cr-Mo steels is a major concern because of:

 The hydrogen embrittlement that could occur at operating temperatures less

than about 120°C. This concern is usually addressed by not pressurizing at
operating temperatures colder than 120°C.
 The potential for delayed hydrogen cracking that could occur in weld repairs
subsequent to service. This concern is addressed by an appropriate bake out
prior to repair welding.

2.2.3 Oxidation
Virtually all, metals and alloys have threshold temperatures above which they
become susceptible to rapid scale formation and spalling when heated in air or
steam. Table 1.2.2.3 shows the oxidation/scaling threshold temperatures for
commonly used materials.

Materials Scaling
Temperature, °C
 Carbon Steel 540
1¼Cr-½Mo 565
2¼Cr-1Mo 580
3Cr-1Mo 595
5Cr-½Mo 620
9Cr-1Mo 650
12Cr 815
3½Ni 540
9Ni 540
18Cr-8Ni 900
Types 309 & 1095
310 SS

Table 1.2.2.3: Oxidation/Scaling Threshold Temperatures for Commonly Used Materials. (C.P.
Dillon, Corrosion Control in the Chemical Process Industries, 2nd Edition, MTI Publication
No.45)
Note: Nickel content increases spalling resistance. Thus, the higher Ni grades (such as Type
310 SS instead of Type 309 SS) are favored, particularly in cyclical services, where thermal
stressing will encourage spalling.

Materials in applications subject to thickness losses due to oxidation are usually

provided with a nominal oxidation allowance; 1/16 to 1/8"(1.5 to 3 mm) is typical.
Most often, hot lines and equipment are thermally insulated to conserve energy.
Properly insulated and jacketed, hot lines and equipment can be kept in safe
service at temperatures above the oxidation limits of the materials of construction.
Care must be taken to ensure that the process stream chemistry is either non-
oxidizing or that the process-side surface is protected by an insulating or refractory
liner. In such cases, the limiting factor will be the availability of a code maximum
allowable stress.

Alloys containing significant amounts of molybdenum are potentially subject to

catastrophic oxidation. The superaustenitic stainless steels such as Alloy AL-6XN, a
21Cr-25Ni-6.5Mo-N alloy (UNS N08367) are an example. This problem is associated
with the formation of a heavy molybdenum oxide scale, usually as a result of an
improper heat treatment or a severe thermal excursion while in service. Experience
has shown that removal of such scales prior to service (or return to service) will
prevent the problem. Removal usually requires a pickling treatment.

Heat tinting, such as the blue tinge often seen on welds, is a common condition
associated with welds and thermally cut surfaces. For carbon and low alloy steels,
such tinting is usually ignored. However, the subsurface areas of heat-tinted
stainless steels may be significantly depleted of chromium. For demanding
environments, heat tinting is usually removed by mechanical methods such as
grinding or by chemical cleaning or both.
 2.2.4 Sulfidation and Sulfidic Corrosion
Sulfur and sulfur compounds may attack carbon and low-alloy steels at
temperatures above 260°C. Sulfidic corrosion is most often associated with sulfur in
fuel gas system (as organic sulfides and/or as H2S). It can cause severe pitting and
general wastage in carbon and alloy steels at temperatures exceeding 260°C.
Corrosion rates can be estimated using the McConomy and the Couper-Gorman
curves.

2.2.5 Metal Dusting

2.2.5.1 Summary
Catastrophic carburization (metal dusting) is a corrosion phenomena experienced
in process equipment that is exposed to gases containing carbon monoxide (CO),
hydrogen (H2) and other hydrocarbons typically in the range of about 450°C to
800°C.

2.2.5.2 Introduction
Metal dusting can be described as a catastrophic carburization in conditions of
high carbon activity, low oxygen partial pressures in the general temperature
range of 450-800°C. This process produces powdery carbon, metal particles,
carbides, and oxides. Corrosion can be in the form of pitting and general metal
wastage. Equipment in processes that use synthesis (syn) gas containing varying
amounts of hydrogen, carbon monoxide, methane, and water is susceptible to
metal dusting at elevated temperatures.

Metal dusting also is experienced in the heat-treating industry in equipment that

handles items being treated (annealed, carburized, etc.). Oil residue on the items
can mix with gases used during heat treating. This results in the formation of gases
that are chemically favorable for metal dusting. Gas mixtures used for carburizing
can also cause metal dusting if control of chemistry is not managed.

The common compounds in every one of these processes are CO and H2.

2.2.5.3 Phenomenological Aspects of Metal Dusting

As mentioned above, metal dusting is the rapid carburization of metals containing
transition elements, (predominantly iron and nickel) and their alloys. Carburization
occurs as the result of the catalytic dissociation of gaseous carbon compounds,
(the most prevalent being CO), which charge carbon into the metal to form metal
carbides. The supersaturated metal carbides dissociate to produce a filamentary
mixture of iron and/or nickel and graphite. Concurrent reactions at the gas-metal
interface can cause the formation of coke as well. Carbon deposition can or may
occur when the thermodynamic carbon activity of the gas exceeds unity. Gaseous
reactions that typically involved in metal dusting are:

1) CO + H2 ↔ C + H2O a¢ = K1 · pCO · (pH2/pH2O)

2) 2CO ↔C + CO2 a¢ = K2 · (p2CO/pCO2)
3) CH4 ↔ C + 2H2 a¢ = K3 · (pCH4/p2H2)
4) H2O + CO ↔ CO2 + H2 K4 = (pCO2 • pH2) / (pCO • pH2)
5) H2O + CH4 ↔ CO+3H2 K5 = (pCO • p3H2) / (pH2O • pCH4 )
 (a¢) is carbon activity, Kx is the thermodynamic equilibrium constant for the given
reaction and pxy is the partial pressure of the named gaseous compound.

Studies show that metal-dusting reaction rates are governed by gas and gas-metal
kinetics. Factors influencing kinetics vary with gas and metal chemistry, metal
crystallographic structure, (e.g., body-centered vs face-centered cubic), as well
as environmental and other metallurgical factors.

The following micro-processes are involved during the metal-dusting reaction:

a) Carbon is transferred from a carbonaceous gas mixture of a¢ > 1, which

supersaturates the metal at the metal surface.
b) With alloys of iron, adsorbed carbon forms carbides (Fe3C), which when formed
reduce the rate of carbon ingress into the metal.
c) The continued charging of carbon into the metal supersaturates the carbide at
the gas - metal surface. This causes instability and decomposition of the carbides
back to the small pure-metal particles and graphite. Dissociated metal attaches
to the basal planes of graphite that are growing into un-decomposed metal
carbide)
d) The metal diffuses through the graphite lattice structure away from the carbide
and agglomerates into nanometer sized particles, which serve as catalyst sites for
further carbon formation.
e) Transport of iron and nickel atoms through the graphite is faster when the lattice
planes of the graphite formed in the carbide-dissociation reaction are
perpendicularly aligned to the metal surface as opposed to parallel alignment.

2.2.5.4 Influencing Environmental Factors

2.2.5.4.1 Gas Chemistry

What is certain is that carbon-deposition rates are catalytically enhanced by
contact with iron, nickel and/or cobalt in engineering alloys. Altering the gas
chemistry influences the rate of carbon deposition by altering the carbon activity.
Additionally, diffusion kinetics are important in dictating the rates of carbon
penetration into the metal and the diffusion of alloying elements from the metals
to the surface.

Temperature, pressure, gas chemistry, gas velocity, etc. all have been shown to
affect the rate of metal dusting. Thus, it is not surprising that the reaction elements
of metal dusting are still not clear.

2.2.5.4.2 Temperature, Pressure, Gas Flow

 Because they affect gas equilibrium, temperature and pressure influence carbon
activity. In reaction (1) and (2) above, ac increases as temperature decreases and
pressure increases, whereas with reaction (3), a¢ increases as temperature
increases and pressure decreases. As an example of the effect of temperature on
metal dusting, it has been reported in the literature that metal dusting occurs in
equipment at locations of low gas velocity wherein temperatures increased 59 or
decreased 61 into a metal dusting regime. Localized damage is confined to the
hotter or cooler areas. Investigations verify that the rate of carbon deposition varies
depending upon gas and metal temperatures 15s, as well as alloy chemical
composition.

Deposited carbon in the range of metal dusting temperatures tends to be

filamentary and as such, high gas velocities can sweep away the carbon formed
at the metal surface which can influence the rate of carbon deposition on
exposed metal surface. This situation accelerates the rate of penetration
particularly in nickel-based alloys, because fresh metal is exposed continuously at
the surface. Hard particle erosion in gases at high velocities can erode the
protective oxide to continuously uncover fresh metal.

Stresses imposed on equipment surfaces by temperature and pressure gradients

cause fractures in protective oxides. Metal dusting then occurs at the exposed
metal at the fracture site. It is believed that this can cause local metal dusting
penetration sites seen in some of the higher alloyed construction materials.

2.2.5.4.3 Gas Inhibitors

It has been known for a long time and is well established from laboratory
investigations and in industrial applications that additions of sulfidic compounds,
such as hydrogen sulfide will inhibit the metal-dusting reaction. Sulfides adsorb on
the metal surface, and this negates the adsorption of carbon into the metal, thus
minimizing the formation of transition-element carbides. While the sulfidic
compounds are efficient inhibitors of metal dusting, they are potent poisons of
certain catalysts used in producing syngas and other organic compounds.
Competing oxidizing gas reactions that reduce carbon activity are those in which
larger concentrations of water vapor and carbon dioxide are present to shift the
gas equilibrium to produce higher concentrations of oxidizing species. These
reactions favor the formation of metal oxides, which at the surface restrict diffusion
of carbon into the metal. However, the altered more-oxidizing gas chemistry can
be at cross-purposes with maximizing the efficiency of the synthesis gas
manufacturing process.
 3
Chapter

3 CORROSION UNDER INSULATION

3.1 Introduction
Corrosion under insulation (CUI) or "under-lagging corrosion" as it is sometimes
called, is a rather common type of corrosion attack in refinery and petrochemical
plants. This type of corrosion can cause failures in areas that are not normally of a
primary concern to an inspection program. Metal loss corrosion, pitting and
environmental cracking such as chloride stress corrosion cracking could occur
under insulation. Problems have occurred in piping and equipment insulated for
both low- and high-temperature services, for virtually all types of insulation,
including products used for personnel protection and fireproofing. Some of the
worst problems have occurred in insulated piping and equipment subject to
periodic "deluge" testing of fire water systems. Bypass loops, typically hot only
during startup or shutdown, are frequent locations of aggressive external corrosion.

Often, a "vapor barrier" is relied upon to prevent liquid water from penetrating the
outermost layer of insulation. However, even the best such barriers only slow down
the penetration rate. They are permeable to water vapor and are subject to
punctures, slippage, aging, delamination, etc.

In addressing this old and well-known problem of corrosion under insulation, it is

important to give attention to the chemistry of insulation products to be used on
austenitic stainless steel. Insulation products should not be permitted to contain
leachable chlorides.

CUI can be catastrophic in nature or at least have an adverse economic effect in

terms of downtime and repairs. Its extensive occurrence leads to a ASTM STP 880
publication: "Corrosion of Metals under Thermal Insulation", issued in 1983. This
publication reviewed the causes and factors affecting the occurrence and rate of
CUI, field experience with insulation types and control measures including use of
coatings, specifications, system design and inspection. An inspection gap noted
at the time was an "urgent need to develop a nondestructive on-stream
examination (NDE) method to detect corrosion under insulation without the
removal of insulation".

Later, in 1990, a NACE report was issued, NACE Standard RP0198-981, entitled
"Corrosion Under Wet Thermal Insulation" which updates the ASTM STP 880
publication and provides useful guide to preventing corrosion under thermal
insulation.
3.2 Location for CUI

3.2.1 Carbon Steel

Corrosion under insulation will occur at places where water has collected and salty
deposits have formed and will furthermore be dependent upon the operating
temperature. CUI is normally occurred on carbon steels and 300 series stainless
steels. The suspect temperature range for carbon steel is between 0°C and 150°C
wall temperature, and is most severe at about 93°C. The failures are often the result
of localized wall loss and/or general wall thinning.

For all surfaces of carbon steel piping and equipment under insulation, a complete
paint system as is required for bare uninsulated surfaces is specified.

3.2.2 Austenitic SS
CUI in austenitic stainless steel manifests as pitting and/or external stress corrosion
cracking in the temperature range 50-150°C. The stress corrosion cracks usually
develop at a slow rate, but can extend over large surface areas.

Austenitic stainless steel surfaces of piping and equipment under insulation, with
temperatures between 50°C and 200°C and as well as below ambient
temperature (below dew point) should be suitably protected with a non-zinc
containing paint prior to insulating.

Painting under insulation of austenitic stainless steel surfaces of piping and

equipment with temperatures below 50°C or above 200°C is required if:

 The uninsulated austenitic stainless steel surfaces would also be painted.

 The temperature fluctuates within these temperature ranges.
 Below dew point temperatures are likely to occur.

Another method to prevent stress corrosion cracking of stainless steel is by

introducing a layer of aluminum foil between the metal surface and the insulation.
This technique has been developed by ICI and has now been applied for about
15 years. Their experience is good. Where failures have occurred, they were due
to:

 A single wrap of too thick material instead of a double wrap (ICI specify a
30-40 microns thick aluminum foil)
 Leakage of fluids into the insulation at e.g. flanges joints
 Mechanical damage of the foil by puncture, etc.

Not all oil companies have been in favor of this technique because the risk of
aqueous fluids penetrating into the insulation is still considers to exist.

Another important aspect is that metallic cladding around the external surface of
the insulation is selected and applied in such a way that the ingress of water is
 prevented. For cold and dual temperature services (As defined in PTS 30.46.00.32),
it is of important that the type of vapor barrier is properly selected and applied.

3.3 Causes
The prime precursor for pitting and general corrosion of carbon steel and stress
corrosion cracking of stainless steel under thermal insulation is the ingress of water
into the insulation which traps the water like a sponge in contact with the metal
surface. Consequently hygroscopic salty deposits are formed on the metal
surfaces, aggravating the damage. The water can come from rain water, leakage,
deluge system water, wash water, or sweating from temperature cycling or low
temperature operation such as refrigeration units.

Thermal insulation must remain dry to retain its insulating properties and to prevent
corrosion of underlying metal. Saturated insulation is considered extremely
detrimental, even dangerous, because of the potential for corrosive failure of the
piping. Concentration cells, temperature differences, and wetter vs dryer areas
cause corrosion — as do leachable chlorides and other contaminants.
Contamination may also be present on the metal surface before it is coated or
insulated. Once these areas become wet, corrosion can occur.

Absence or damage of an protective coating will aggravate the condition and

serve to promote faster deterioration.

The following areas could be susceptible to CUI:

 Areas exposed to mist overspray from cooling water towers.

 Areas exposed to steam vents.
 Areas exposed to deluge systems.
 Areas subject to process spills, ingress of moisture, or acid vapors.
 Carbon steel piping systems, including those insulated for personnel
protection, operating between 0°C and 150°C. CUI is particularly aggressive
where operating temperatures cause frequent condensation and re-
evaporation of atmospheric moisture, or those operate above 150°C but are
under intermittent service..
 Deadlegs and attachments that protrude from insulated piping and operate
at a temperature different than the active line.
 Austenitic stainless steel piping systems that operates between 50° C and
200°C. These systems are susceptible to chloride stress corrosion cracking.
 Vibrating piping systems that have a tendency to inflict damage to insulation
jacketing providing a path for water ingress.
 Steam traced piping systems that may experience tracing leaks, especially
at the tubing fittings beneath the insulation.
 Piping systems with deteriorated coatings and/or wrappings.
  Locations where insulation plugs have been removed to permit thickness
measurements on insulated piping should receive particular attentions.

3.4 Mechanism
Corrosion under insulation occurs according to a conventional corrosion
mechanism, but requires three ingredients:

o Availability of oxygen.
o High temperature.
o Concentration of dissolved species.

In normal conditions, as the temperature increases, the amount of oxygen

dissolved in solution decreases as the boiling point is reached resulting in reduced
corrosion rates. However, on the surface covered by insulation, a poultice effect is
created which holds in the moisture and essentially makes it closed system. In fact
the measured corrosion rates associated with corrosion under insulation follow
trends to higher corrosion rates commonly associated with only pressurized systems.

Furthermore, in cases where precipitation becomes trapped on the metal surface

by insulation, corrosive atmospheric constituents such as chlorides and sulfuric acid
can concentrate and accelerate corrosion. In some cases, chlorides are present
in the insulation which greatly promotes corrosion of the underlying surface which
it becomes laden with moisture.

3.5 Types of Corrosion Under Insulation

By identifying the types of corrosion that can occur under insulation, the proper
measures can be employed to prevent them. Intruding water is the key problem in
CUI. Special care must be taken during design not to promote corrosion by
permitting water to enter a system either directly or indirectly by capillary action.
Moisture may be external or may be present in insulation.

Corrosion may attack the jacketing, the insulation hardware, or the underlying
piping or equipment. Depending on other factors, chloride, galvanic, acidic or
alkaline corrosion may occur.

3.5.1 Galvanic Corrosion

Galvanic corrosion generally results from wet insulation with an electrolyte or salt
present that allows a current flow between dissimilar metals (i.e., the insulated
metal surface and the outer jacket or accessories). The extent and severity of the
attack on the less noble metal depends not only on the difference in potential of
the two metals, but also on their relative areas.
3.5.2 Alkaline or Acidic Corrosion
Alkaline or acidic corrosion results when an alkali or acid and moisture, are present
in certain fibrous or granular insulation. For hot service above 150°C, most of the
water is driven off. This water vapor may condense at the edge of the insulation,
and dissolve the alkaline or acidic chemicals there, resulting in corrosion of the
aluminum or stainless steel jacketing.

Some alkaline waters with aluminum produce etching and pitting. Pitting can be
severe, especially when chloride ions are present. Insulating cement may also
contain alkaline chemicals and water (while the cement is still drying). Below
150°C, alkaline water may cause corrosion if the substrate or insulated surface is
stainless steel, copper, brass or aluminum. Steel would normally not be affected in
the time needed for the cement to dry. Fresh, potable water is recommended
when mixing insulating cement.

The leaching tests performed on polyurethane foam insulation containing fire

retardant chemicals (i.e., brominated or chlorinated compounds) with distilled
water have led to the formation of aggressive acidic solutions. The same is found
true for phenolic foams. The pHs of the solutions are often two to three. Laboratory
corrosion rates have been shown to be 15-20 mpy. Of the two foams, the phenolic
are by far the more corrosive.

3.5.3 Chloride-Induced Corrosion

Chloride-induced corrosion can be caused by the combination of insulation
containing leachable chlorides with the 300 series austenitic-stainless-steel
surfaces, when moisture is present and temperatures are above 60°C.
Concentration of the chloride ion usually results from the evaporation of rain water,
or of water used to fight fires, or of process water. Stress-corrosion cracking of
insulating jackets often results from airborne salts in coastal regions.

The probability of failure and the speed of crack propagation are governed by the
temperature of the stainless steel and the chloride concentration at the metal
surface. Solutions containing less than 1 ppm are normally considered safe. Below
80°C, levels of 100 ppm are not particularly dangerous if continuous surface-
wetting occurs; but at higher temperatures, lower levels can result in failure.

In practice, it should be assumed that evaporation of the solution will inevitably

occur. Because local concentration of chlorides takes place at the metal surface,
the bulk concentration may be of little importance. Above 200°C, external stress-
corrosion cracking is normally not experienced. The stress required to cause
cracking of stainless steel may result either from fabrication or operation (or
shutdown).

Water entering the insulation and diffusing inward will eventually reach a region of
dryout at the hot pipe or equipment wall. Next to this dryout region is a zone in
which the pores of the insulation are filled with a saturated salt solution - this
includes any chlorides present. When a shutdown or process change occurs and
the metal-wall temperature falls, the zone of saturated salt solution moves into the
metal wall. Upon reheating, the wall will temporarily be in contact with the
saturated solution (e.g., chlorides), and stress-corrosion cracking may begin.
3.6 Prevention

3.6.1 Design
An appropriate protection against external corrosion should already start in the
design phase. The most important design requirement of a good insulation system
is to prevent water ingress into the insulation. However, practice has shown that an
absolute exclusion of water ingress is almost impossible, therefore additional
measures such as painting, etc. are applied. The ingress of water can be prevented
by:

 Installing a continuous welded water-retaining collar around all protruding

parts of vessels and columns.
 Reducing the number of supports and fixings penetrating the insulation to an
absolute minimum.
 Locating valves and flanges in the horizontal part of pipelines instead of the
vertically.
 Placing nameplates that do not interfere with insulation practice.
 Avoiding underground insulated pipes and insulated pipes in low pipe
trenches.
 Installation of gratings on top of columns, vessels and over piping, where
applicable.

In line with the above described practices due consideration should be given to
draining facilities. Small penetrations (of e.g. nozzles, instrument tubing) should
drain instead of providing ingress to the possible water. Further it may be
worthwhile to replace such parts with corrosion resistant materials e.g. nickel alloys
or duplex stainless steels.

3.6.2 Coating
The most effective way to control corrosion under insulation is to prevent the
electrolyte from reaching the metal surface with a properly applied coating on the
pipe. This should be supplemented by properly sealing the insulation with a vapor
barrier or weather barrier jacket. Damage to the cladding and deterioration of
sealant, however, often occur in service and if repairs are not carried out soon
afterwards the insulation will become wet.

This has led to the specification of additional precautionary measures such as

painting, using non-absorbent thermal insulation materials and designing
protective systems that permit any ingress water to easily drain out.
 3.6.3 Materials Selection
Careful selection of insulation materials to prevent those that contain high levels of
corrosive impurities such as chlorides is critical to reducing corrosion under
insulation.

For most of the thermal insulation applications either mineral wool, glass wool,
polyurethane foam (alternatively polyisocyanurate foam) or cellular glass ("foam-
glass") is used.

Both PTS 30.46.00.31 and PTS 30.46.00.32 limit the amount of leachable chlorides.
This amount is also specified for the auxiliary materials, such as vapor barriers,
adhesives, mastics, etc.

Developments in insulation materials are so far limited to a glass wool, inhibited with
sodium and silicate ions (in order to prevent external stress corrosion cracking) and
a silicone treated mineral wool (low in water absorption).

The selection of cladding material depends upon factors such as potential fire
hazard, surface temperature, presence of austenitic stainless steels, the location of
the plant.

The current PTS specification on thermal insulation allows:

 Aluminum jacketing of a half hard quality meeting ASTM B 209 M, alloy 1060,
temper H 14
 Aluminized Steel Type 2, coating designation T2-100 in accordance with
ASTM A463.
 Stainless steel in accordance with ASTM A 167, Type 304L or 316L.
 Galvanized steel in accordance with ASTM A 526; hot-dip galvanized to
ASTM A525 coating designation C90.

3.6.4 Use of Water Proofing

Water proofing to prevent the ingress of water from outside sources is another
method. However, it has been shown that sometimes these techniques tend to lock
in moisture which can also increase corrosivity.

3.6.5 Use of Inhibitor

Inhibitors have been tried with varying success since repeated wet / dry cycles
may make inhibitors ineffective. This is an area of opportunity. However, long terms
performance and efficacy must be proven.

3.7 Monitoring and Inspection

Inspecting piping or vessel for corrosion under insulation is persistently a
troublesome exercise faced by the refinery and petrochemical inspectors. As such,
there is a dire need to non-destructively inspect the pipe or vessel while they are in
service. CUI is difficult to find because of the insulation cover that masks the
corrosion problem until it is too late. It is expensive to remove the insulation,
particularly if asbestos is involved.

There are a number of methods used today to inspect for corrosion under
insulation. The main ones are profile radiography, ultrasonic spot readings, and
insulation removal. The other method now available is real-time X-ray. Real-time X-
ray has proven to be a safe, fast and effective method of inspecting pipe in plant
operations. It would be a daunting task to muster the resources needed to tackle
this extensive list of piping with the traditional inspection methods. This is where real-
time X-ray offers a real advantage. Once the damaged areas are identified,
follow-up X-rays and ultrasonic can measure the loss by external corrosion. These
techniques will not however, detect SCC in stainless steels.

It is strongly recommended that suspect plant items, vessels as well as piping, be

included in an inspection and maintenance scheme on the basis of external
corrosion. Scheduled inspections should be conducted to detect :

 damaged metallic cladding

 gaps or unsealed joints
 unreplaced insulation
 leaks or spills
 high and low temperature projections
 presence of salt crusts
 condensation on cold surfaces
 breaks or shrinkage cracks in weather-and vapor barriers

Visual inspection is still the most widely used method of inspection. However, as the
insulation has to be removed or plugs in the insulation have to be cut, the
inspection is mainly limited to potentially critical places. Since CUI tends to be
localized, and, unless the inspection plug is positioned in the precise spot, corrosion
sites may be missed. Furthermore, the plugs can be a source of moisture leakage.

Conventional radiography is an effective method for the determination of CUI.

Disadvantages of the method are its cost and slowness. Flash radiography is a
quick and effective method to assess the presence of corrosion, but the rate of
corrosion can not be determined. So far it has only been used for piping.

For detection of stress corrosion in stainless steel, visual inspection may be followed
by dye-penetrant inspection.

Various inspection techniques to obtain useful information should be evaluated.

Two example of these methods are Infrared scanning and Neutron back scatter.
They can be summarized as follows :

 Infrared Scanning
o Advantages:
o small, light, portable equipment
o fast method
 o suitable for hot and cold insulation
o actual corrosion can be detected
o Disadvantages
o cannot be used in rainy conditions
o experienced operator is necessary
o not intrinsically safe
o plant should be in operation

 Neutron Back Scatter

o Advantages :
o can be used under all weather conditions
o relatively simple
o plant should not necessarily be out of operation
o suitable for hot and cold insulation
o Disadvantages
o two operators are necessary
o relatively slow method
o radioactive source
o not intrinsically safe
o detects only moisture

3.8 Specification
A system to combat corrosion under insulation should in any case be based on
properly prepared specifications.

Two technical specifications on thermal insulation are available in the PTS system :

o PTS 30.46.00.31-Gen.: Thermal Insulation for Hot Services

o PTS 30.46.00.32-Gen.: Thermal Insulation for Cold and Dual-
Temperature Services
Both documents cover the minimum requirements for materials, design,
application, quality control, inspection and maintenance of existing insulation
system.

3.9 Conclusions
The prime cause of "under-lagging" corrosion of carbon steel and stainless steel is
the entry of water into the insulation.

Optimum design of installation systems and careful application of insulation,

sealing and cladding materials will prevent ingress of water. However, other pre-
cautionary measures such as painting and installing draining facilities should
certainly be taken.

These aspects are covered by thermal insulation PTS specifications. A regular

update should take place on the basis on own experience or of external sources,
such as the relevant NACE publication.
 4
Chapter

4 EROSION-CORROSION

4.1 Introduction
Erosion-corrosion affects steam and condensate piping systems and equipment
and items such as steam condensate air coolers, pipes, bends, different elbows,
pumps, valves, etc.

Usually, pure steam is not aggressive from a corrosion point of view. However, if its
condensates are contaminated by CO2 and dissolved oxygen it becomes rather
corrosive. When steam condenses, liquid formation results and the system
becomes actually a two-phase (vapor-liquid) system. Normally, the wetter the
steam (i.e., the greater the amount of condensation), the more corrosive it
becomes.

It is well known that the steam is acidic because of the decomposition of

bicarbonates resulting in carbonic acid formation. At elevated temperatures and
pressures, pH of water also falls to about 6.0. Naturally, temperature and oxygen
content are the factors that will most aggravate the corrosion reactions.

Steam generally passes through piping system at very high velocities because of
economic reasons. The flow velocity is another major accelerating factor in a
steam condensate system. Under this operating condition, steam condensate can
cause erosion-corrosion. This usually occurs when protective surface films or
adherent corrosion products are mechanically removed by flowing liquid
environment. The mode of damage is generally characterized by patterns such as
grooves, waves, rounded holes, valleys and horseshoe-shaped depressions. All
these forms of damage exhibit strong directional pattern consistent with liquid flow
and normally contain smooth-bottomed recesses. At an area away from this
erosion-corrosion portion, only gradual thinning is normally occurred due to the
combined action of the liquid flow and the presence of acidic species.

Impingement corrosion is known as a severe form of liquid erosion-corrosion, which

usually occurs when corrodent or solid particles bombard (impinge) the metallic
surface at an angle (worst case at approximately 90°). Such damages occur
frequently in elbows or turns of tubes or pipes, impellers of turbines, etc. In particular,
high-velocity wet steam can cause some specific forms of impingement corrosion
depicted as steam erosion with rounded surface and showing a large number of
small cones.

In wet steam (two-phase), erosion-corrosion is similar in appearance to that which

occurs in the liquid phase. However, the mechanism is complex and it involves the
electrochemical corrosion process (anodic and cathodic reactions), the
 convective mass transport and mechanical wear effects of abrasion, droplet
impingement, and/or solid particle erosion.

Erosion-corrosion can occur in most carbon steel heat exchangers operating at

high velocity and temperatures. This type of attack usually typified by local
damages up to complete perforation at areas indicated by steam condensate
flow direction. Visual inspection normally reveal a very smooth, nonuniform
asymmetrical, wave-like product on the surface at the flow side, with some liquid
impingement character. No micro structural changes or cracks should be
observed at the damaged areas.

4.2 Mechanism
Condensation of steam normally occurs more in the lower temperature region. In
both low and high temperature regions there is possibly a mixture of liquid phase
with some amount of steam in vapor phase. Steam condensate normally flow at a
relatively flow rate; however, it is supposed that if only a few percent of this flow
was in vapor phase, the velocity would significantly increase and consequently
increase the tendency towards liquid erosion-corrosion.

4.3 Erosion Assessment

At high fluid velocities, damage of protective scale can result in corrosion rates in
excess of those estimated by the deWaard and Milliams method. In the absence
of sand, the velocity at which this may occur is given by the API RP 14E formula.
c
Ve  m

Where
Ve = fluid erosion velocity
rm = gas / liquid mixture density at flowing pressure and temperature
c = empirical constant
[Industry experience to date indicates that for solid-free fluids, values
of c = 100 for continuous service and c = 125 for intermittent service
are conservative].
 5
Chapter

5 MICROBIOLOGICALLY-INDUCED CORROSION

5.1 Introduction
Microbiologically influenced corrosion (MIC) is defined as the corrosion of materials
caused by microorganisms. The combination of unexpected attack and rapid
failure makes MIC a matter of considerable concern in, especially, cooling water
systems. Most MIC occurs in stagnant water systems or the water legs of mixed
phase, quiescent process streams. A typical example would be the pitting
corrosion that occurs in the bottom of a pipeline, before commissioning, due to
microbial activity in residual hydrotest water. It generally occurs at ambient
temperatures but may occur at temperatures as high as 93°C. Most materials of
construction are susceptible.

The microorganisms responsible for MIC are primarily bacteria and fungi. They may
be anaerobes (which will not function in the presence of oxygen) or aerobes
(which require oxygen, usually dissolved, in order to function). Often, MIC is caused
by a mixture of several microorganisms living in colonies. It should be noted that
while strictly anaerobic environments are not common in nature, anaerobes are
commonly found within the anaerobic micro-environments that can develop in
even highly aerated systems. For example, anaerobic conditions can be
established under the slime film formed by some aerobic types of microbes.

A number of organisms are commonly associated with MIC, including sulfate-

reducing bacteria (SRB), sulfur/sulfide-oxidizing bacteria, iron/manganese-
oxidizing bacteria, aerobic slime formers, methane producers and acid-producing
bacteria and fungi.

Microbiologically-induced corrosion arisen through the action of sulphate

reducing bacteria. Such bacteria will thrive under anaerobic conditions but will
survive under aerobic conditions obtaining their nutrient from sulphate in water and
hydrocarbon. The sulphate ultimately becomes sulphide. This will acidify the local
environment where the bacteria are attached to and destabilise protective scales
on the metal. With the protective scales being attacked under acidic condition,
corrosion is further accelerated by enzymes used by the bacteria.

MIC occurs most likely under stagnant conditions and under deposits, on carbon
and stainless steels. However, if bacteria populations are high then sessile sulphate
reducing bacteria can attach themselves to pipewalls and lead to corrosion even
under high flow. The situation is exacerbated in systems where oxygen levels swing
by large amounts.

MIC manifests as a localized, pitting form of attack, following the development of

a surface biofilm. The anaerobic environments often support the growth of biofilms,
 which almost invariably contain SRB, a major cause of MIC. MIC corrosion
generates iron sulfide within the biofilm, which is cathodic to the steel and may
greatly exacerbate corrosion at anodic sites.

5.2 Mechanisms of MIC

Proper diagnosis of the corrosion mechanism would lead to the correct inspection
frequency, location and techniques to be applied, which would lead to a cost
savings, as compared to a misdiagnosis.

Despite efforts made by numerous scientists to determine the mechanism of MIC,

the exact mechanism in which SRB influence the corrosion of metals/alloys is still
subject of controversy. However, it can be simplified as follows:

 Deposits (sand, paraffin, asphaltenes, water scales) accumulate in lower

sections of piping and vessels,
 Planktonic bacteria accumulate in/near deposit accumulations and become
sessile,
 Sessile bacteria grow, producing corrosive chemicals and iron-sulfide scaling,
 Leaks occur when corrosion has significantly degraded pressure containing
capabilities.

Earlier studies in the 30's, have revealed the significant role of SRB in cathodic-
depolarizing the metal surface for accelerated corrosion. These bacteria reduce
inorganic sulfate to sulfide, which may remain as dissolved sulfide, be liberated as
H2S gas, or may react with ferrous ions to produce the characteristic black
precipitate of iron sulfide. The overall reaction can be written as:

4Fe + SO4= + 4H2O → FeS + 3Fe(OH)2

The iron sulfide as a corrosion by-product is cathodic to metallic iron (thus the
creation of galvanic cell) and can cause a continued corrosion. Recent studies
support the earlier reports that bio-precipitated sulfides on metal surfaces serve as
electron acceptors for the cathodic reactions in anaerobic systems and that
corrosion depends more on the corrosion products than SRB activity.

In addition, a specific cathodic reaction in the presence of H2S can also occur
with the following likely reaction path:

2H2S + 2e → 2HS- + H2

SRB is theorized to be able to remove molecular cathodic hydrogen from the metal
surface, favoring the production of H2S. This reaction is thought to be catalyzed by
the enzyme hydrogenase present in SRB and assays for this enzyme have since
been used to assess MIC risk.

Little information is available on the kinetics of MIC. However, it is believed that

bacterial growth includes a period of incubation and it took approximately one
month for bacteria to reestablish themselves on recently cleaned and sterilized
 piping. It is also reported that it took about 75 days for bacteria to produce
corrosion rates in excess of 0.18 mm/y (7 mpy).

5.3 Microbiologically-Induced Corrosion Assessment

In this document, the review of the corrosion mechanism shall be focusing to
identify the possible corrosion under the influence of microbial activity, the
parameters that determine the rate of attack and, if possible the likely rate of
attack.

5.3.1 MIC Susceptibility Assessment

As a first step, a MIC susceptibility assessment study should be performed for all
vessels and process piping. A criticality rating with respect to probability of the
system experiencing MIC should then be established, placing all equipment in their
probable failure risk and the possibilities of re-design and control consideration.

The MIC controlling factors, inspection records (U.T. and B-Scanning survey results),
results of fluid analysis and bacteria analysis should be established as a data base
for a complete susceptibility assessment.

5.3.2 Influencing Factor-Based Susceptibility Assessment

The physico-chemical and biological interactions, which take place between the
biofilm and the environment are very complex. They may be interdependent on
other processes not directly related to biological activity. This makes modeling the
corrosion phenomena that constitute MIC an extremely difficult task. Prediction of
pipe wall penetration rates attributed to MIC-induced localized corrosion is
unreliable, because of the uncertainty of the onset of pitting and whether it
proceeds at a constant rate. Nevertheless, susceptibility to MIC could be assessed
on the basis of details of water chemistry and facilities operational parameters. The
main idea is to link the risk of MIC and the influencing parameters.

Predicting the susceptibility of vessel and piping to MIC is facilitated by considering

1) whether the vessel and piping environment can support microbial activity and
biofilm formation, and 2) how vessel and piping operational parameters affect the
microbiology. The premise is that if an active biofilm can form inside a vessel or
piping, then they may suffer from bacteria-associated corrosion.

Microbial activity is generally limited by extremes of certain parameters. Although

some microorganisms can grow outside the normal limits, they are usually confined
to extreme habitats. The limits for pH and temperature are well established for
bacteria that could occur in oil processing facilities, but tolerances for salinity for
example, are less well understood. The following parameters must be considered
in corrosion rate calculation:

 Presence of Water and Water Wetting

o If the crude, vessel and piping are free of water, there will be no
MIC. MIC is only to be considered at separated flow conditions
(e.g., in stratified flow); if all water is entrained in the oil, the risk of
MIC is significantly reduced.
 pH of Water
o SRB growth is largely confined between pH's of 5 and 9.5.

 Salinity or Total Dissolved Solids (TDS)

o Most SRB will grow best at < 60 g/l TDS. SRB may still grow if TDS > 60
g/l, but this must be confirmed by growth in a nutrient medium with
the same TDS concentration. The suspended solids can be made
up of silicon, sodium, aluminum, chloride and zinc elements.

 Temperature
o Different types of SRB can grow at temperatures between 4 and
110°C. If the temperature is above 45°C, it is necessary to confirm
that SRB isolated from the vessels and piping can grow at the
operating temperature.
o The average temperature of the three platform of ~ 38°C is
conducive for bacteria growth; in fact these temperatures are the
most favorable envelop for the growth.

 Influence of Nutrient
o Nutrient availability directly influences microbial activity and bio-film
formation both qualitatively and quantitatively. Carbon and
nitrogen are the two of the most important nutrients, which can be
readily quantified. Sulfate is also important for SRB. It is the principal
respiratory electron acceptor, which can ultimately have a
significant influence on MIC. There are threshold concentrations
below which these nutrients cannot be effectively utilized by
microorganisms.
o When one of the nutrients essential for microbial activity falls below
the threshold, this activity ceases until more nutrients become
available. It should be noted that a water analysis indicating a
nutrient deficiency does not mean that no MIC can take place, if
previously nutrients were available. MIC associated with a biofilm
that was formed when nutrients were available, but which
subsequently faces a nutrient shortage, may continue for some
time.
o Additionally, the ratio of carbon to nitrogen is important. A ratio of
carbon to nitrogen of approximately 10:1 is required to sustain
microbial growth. The ratio will indicate the limiting nutrient, useful
information when looking at options for controlling microbial activity
(e.g. a production chemical could change the ratio in such a way
as to encourage microbial activity).
o Some bacteria can survive and even grow in the complete
absence of sulfate, they simply use an alternative electron
acceptor, such as nitrate. Apart from sulfate, other oxidized sulfur
species may be utilized, such as thio-sulfate or bi-sulfite. Hence if
there are sources of these species present in the water, then they
may represent an important increase in the potential for sulfide
generation if the level of sulfate is very low.
o Note that the threshold values given for nutrients are based on
rather few data. These values should be substantiated with more
 experimental data to obtain a higher level of confidence since oil
field waters often contain key nutrients close to threshold limits.
o The failure analysis report indicated the presence of appreciable
numbers of planktonic micro-organisms in the offshore trunkline,
mainly in the streams from Baram fields. This report also indicated
favorable conditions for the bacteria to go sessile. It is high possibility
that the bacteria could find their way into other fields as well.
o In summary, the following nutrients should be included in a
susceptibility assessment:
 Sulfate Concentration
SRB growth is severely reduced at sulfate levels <10 mg/l
unless an alternative electron acceptor is present.
 Total Carbon (C)
The source of carbon is from fatty acids such as formate,
acetate, propionate and butyrate. SRB growth is severely
restricted if utilizable carbon is <20 mg/1.
 Nitrogen (as utilizable N)
SRB growth is severely restricted if utilizable nitrogen is <5
mg/1
 Critical C:N ratio
SRB growth is the most prominent if C:N <10.

 Impact of Flow Rate

Flow rate will directly influence the nature of biofilm formation and the rate of
nutrient delivery. As flow rate increases, biofilms become less bulky and more
adherent. However, above a certain threshold, the initiation of biofilm
formation is significantly limited. The transition zone is considered to be
between 2 and 3 m/s. At the other extreme, stagnation is often associated
with the severest MIC incidents. Low flow rates not only promote deposit
accumulations, but can also result in oil/water separation. This produces an
effective increase in water cut in the bottoms of piping and vessels. Both
deposits and higher water cuts promote increased bacteria levels.

This condition may exist in certain section within the facilities where low velocity
regimes are expected to prevail. A piping and vessel's flow history could be an
important parameter in determining current status of the equipment and
subsequent risk of MIC. Semi-stagnation condition can be experienced at the
dead legs, closed drains, etc.

No flowrate could be used for actual calculation from the three platforms,
although it can be easily calculated. However, lack of other relevant data has
made this calculation meaningless. A significantly higher flow velocities can
be an inhibitory for biofilm formation and MIC, but may make them susceptible
to erosion or erosion-corrosion.
 Impact of Debris/Suspended Solids
Presence of debris exacerbates MIC through the creation of stagnant
conditions in the bottom of the vessel and low velocity piping, where biofilms
can be established.

Knowing the level of suspended solids is very important; relatively, the higher
the level, the less effective the inhibition will be. This is because adsorption of
 inhibitor to suspended solids and scale deposits will reduce the chemicals
availability for the intended target.

 Dissolved Iron
Dissolved iron trending may be used to estimate the extent of MIC. It is
important to have the analysed water stream of the three. Presence of iron
sulfide, FeS can be indicated by the dissolved iron content and when H2S is
present.

 Impact of Oxygen Ingress

The highest biofilm-associated corrosion rates are seen when oxygen
periodically accesses the sulfide-rich, anaerobic region within SRB biofilms.

 Impact of a Biocide Treatment

Biocides are used to control unwanted microbial activity. If a biocide is to be
used, the risk of a failure due to MIC should be significantly reduced, or even
eliminated. It is essential that the biocide is selected carefully and the
treatment regime is appropriate and applied properly and consistently. For
practical purposes, the inclusion of biocides as a parameter in a predictive
scheme would need to be based on the assumption that the correct biocide
has been selected and applied in such a way that it can function as intended.
With this in mind, it is possible to make some simple rules regarding the type of
biocide and the application dose and frequency for different types of water
chemistries, water cuts, etc.

 Impact of Operational History

The following operational history parameters have an influence on MIC:
 Operational age of vessel and piping in years
 Total operational downtime: Long periods of stagnation promote
MIC.

 Impact of Pigging Frequency

Pigging can potentially have a significant effect on biofilm and MIC. The more
frequent the pig runs, the less time biofilm has to recover. A pigging frequency
of at least once every two weeks will be very effective in controlling biofilm.
No detailed data are available on different pigging intervals versus biofilm
recovery rates or MIC, although there is some published experience of the
impact of pigging as a treatment for MIC. The extent to which pigging is
effective in controlling biofilm-associated corrosion depends on how well a
corrosion process was established prior to the pigging program. Since pigging
has a significant impact on the occurrence of MIC, the frequency should take
a prominent place in the prediction module.

 Other Parameters
Other parameters may also play a role in MIC. An example is production
chemicals. Production chemicals can impact MIC in a number of ways. They may
be a source of nutrients to stimulate microbial activity. They may interact with and
reduce the efficacy of biocides being used to treat a bio-fouled system. Corrosion
inhibitors could act to reduce MIC in some circumstances. Phosphorus and
nitrogen-containing chemicals have been shown to act as nutrient sources. Since
production chemicals are often used continuously in oil transport pipelines, this
parameter needs to be addressed in a future predictive module.
Examining all the above parameters and knowing the reliable values can lead one
to confidently predict the rate of metal wastage due to MIC. In a simple expression,
the MIC corrosion rate calculation can be based on the following equation:

CR =C x Fp (5)
with
F = f1 x f2 x ........fn (6)

where C is a constant (C=2 mm/y), the f's are factors for the various influencing
parameters, and p is a power law index (0.57). An overview of the factor (f) values
are given in Table 5.3.2

fn Criteria Factor (f) Factor (f)

when true when false
f1 pH between 5 and 9.5? 1 0.001
f2 Total Dissolved Solids (TDS) <60 g/l? 1 0.2
f3 If TDS>60 g/l, do SRB grow? 0.2 0.0001
f4 Temperature (T) between 10 and 45 °C? 1 0.2
f5 If T>45 °C, do SRB grow? 1 0.2
f6 Sulfate > 10 mg/l? 1 0.2
f7 Total Carbon (C) from fatty acids >20 mg/l? 1 0.2
f8 Nitrogen (as utilizable N) > 5 mg/l? 1 0.2
f9 C:N ratio <10? 1 0.4
f10 Flow velocity < 1 m/s ~1
f11 Flow velocity = 2 m/s ~0.6
f12 Flow velocity = 2.5 m/s ~0.1
f13 Flow velocity = 3 m/s ~0.01
f14 Fluid Velocity > 4 m/s ~0.001
f15 Debris on the bottom of piping and vessel 2 1
f16 Pigging frequency: never ~1
f17 Pigging frequency: 13 wks ~0.3
f18 Pigging frequency: 4 wks ~0.001
f19 Pigging frequency: 1 wk ~0.0001
f20 Prolonged oxygen ingress > 50 ppb 5 1
f21 Oxygen content < 50 ppb 1 0.001
f22 Biocide routinely used? 0.2 1
f23 Operational history:
f24 • age of equipment < 0.5 yr 1
f25 • age of equipment > 0.5 yr & downtime = 1 wk ~1
• age of equipment > 0.5 yr & downtime = 50 2
wks

Table 5.3.2: Overview of Factors for Various Influencing Parameters in MIC

 6
Chapter

6 CORROSION IN BOILER SYSTEM

6.1 Introduction
Boiler tube failures result from fireside corrosion, over firing, steam blanketing,
waterside fouling and long or short term over heating. A number of mechanisms
have been identified as causes of boiler waterside corrosion including caustic, acid
corrosion and chelant/polymer corrosion mechanisms. Corrosion on the external
tube surface has also been documented. External corrosion mechanisms involving
sulfide and sulfate salts are also part of damage mechanisms normally
encountered in boiler.

6.2 Normal Boiler Conditions

In normal boiler water conditions, magnetite (Fe3O4) is the preferred high
temperature iron oxide form. Well crystallized (unhydrated) magnetite forms a
dense layer resulting in excellent passivation and thus stifling corrosion. The reaction
sequences for the formation of magnetite follows the well-known mechanisms
originated by Schikorr:

Fe + 2H2O ↔ Fe2+ + 2OH- + H2↑ (1)

Fe2+ + 2(OH)- ↔ Fe(OH)2↓ (2)
3Fe(OH)2 ↔ Fe3O4 $ + 2H2O + H2↑ (3)

In this process, the formation of the ferrous hydroxide (Reaction 2) is the rate
determining step. Therefore, the solubility and stability of this reaction product is the
key to corrosion protection in the boiler. The resultant formation of magnetite
causes passivation and maintain corrosion rate at a very low rates.

Under good boiler operating conditions, the oxidation of iron to magnetite at the
metal surface is slow because the magnetite forms a fine, tightly adherent layer
with good protective properties. The film generally displays good adhesive strength
in part because the thermal coefficients of linear expansion for magnetite and
steel are very similar. Therefore, varying heat load and surface temperature do not
cause undue stress between the film and the underlying metal surface.

Waterside internal chemistry involves a balance between deposition and corrosion

and tend to be more aggressive, except in conditions with better
dispersants/chelants. Often corrosion may be induced by concentration of
corrosive species under deposits, i.e., no deposits, no concentration, no corrosion.
6.3 Damage Mechanisms in Boiler System
The integrity of boiler tube metal can be adversely affected by numerous
phenomena related to a boiler's firing rate dynamics, water chemistry, and design
and age of the unit. The deleterious phenomena include many types of
microstructural degradation processes from tube metal overheating, usually due
to excessively high heat flux and/or the presence of insulating, waterside deposits,
as well as various corrosion reactions and mechanically-related failure
mechanisms.

Because modern, high pressure utility boilers typically utilize demineralized make-
up and employ stringent water chemistry controls, deposition-related, metal
overheating failure mechanisms will not be discussed at length here. However,
several, diverse, corrosion and failure mechanisms can occur in all high pressure
and lower pressure units, and boilers operating at steady state. Such problems can
be especially trying on high pressure utility boilers operating under sliding pressure,
as the specific demands put on them can lead to a relative prevalence of certain
metal corrosion reactions and associated tube degradation problems.

6.3.1 Corrosive Concentration Mechanism

In boiler feed water, suspended solids (mainly iron and copper oxides) can be
deposited at the heat transfer surfaces. These deposits are porous in nature and
offers no protection. Hot surfaces cause the concentration of the deposits true
what is called "wick boiling". In this phenomena, boiler water permeates the porous
deposit by capillary action through small pores like a liquid permeating a wick;
steam then escapes through larger pores (channels) leaving non-volatile solutes
behind. These solutes then concentrate beneath the deposit. The formation of a
steam bubble further concentrates boiler solids. Feedwater solids concentrate in
the boiler relative to blowdown rates. Boiler water solids may concentrate an
additional 2000 times at the heat transfer surface as a result of this "concentrating
film" produced from non-boiling equilibrium. The formation of a steam bubble
further concentrates boiler solids. The concentrated solutes are always very
corrosive to the steel substrate. This corrosive species concentration could lead to
various types of corrosion mechanisms.

For example, in high pressure systems with demineralized water make up, failures
due to waterside corrosion at heat transfer surfaces in operating boiler units are
commonly associated with caustic or acid phosphate attack resulting from the
over concentration of sodium hydroxide or acid phosphates either under deposits
or in crevices.

6.3.2 Caustic Corrosion

Caustic corrosion is sometimes referred to as "caustic attack or "caustic gouging".
Corrosion of this type generally results from fouled heat transfer surfaces and the
presence of an active corrodent (sodium hydroxide) in the boiler water.
Concentrated solutions of alkali occur in situations where the normal washing of
tube metal is restricted after steam bubble release. The concentrated caustic
formation occurs at hot transfer surfaces in boiler.

Once local caustic concentrations are reached such that caustic attack occurs,
the corrosion can proceed to failure in a very short time. Caustic corrosion results
 in irregular wall thinning or gouging of the tube water side surface. Areas subjected
to caustic attack typically show smooth, rolling contours surrounded by encrusted
boiler water solids and crystalline dense oxides. The oxides, however, are not
protective. Particles of metallic copper may also be embedded in the deposit
layer.

Failures due to caustic attack are caused by metal loss. The damage progresses
to failure when the tube wall thins to a point where rupture occurs locally. The
microstructure does not change and the tube metal retains its ductility.

Caustic attack can also lead to other types of failures. One form is hydrogen
damage which results from hydrogen liberated in the corrosion process diffusing
into the metal. The hydrogen can then react with iron carbides in the metal to
produce methane which develops pressure leading to the formation of
microcracks and diminishing tube strength. This type of attack rarely occurs at
pressures below 7 MPa (1000 psig).

Another form of metal damage in which caustic is involved is stress corrosion

cracking. In this case, the combination of concentrated caustic and tensile stresses
may cause continuous, intergranular cracks to form in the metal.

Stress corrosion cracking may occur in areas where boiler tubes are rolled into
boiler drums or headers. The occurrence of stress corrosion cracking is diminishing
due to improved boiler water treatment programs, such as coordinated
phosphate, and boiler design.

6.3.2.1 Concentrating Film Corrosion Mechanism - Departure from Nucleate Boiling

Caustic in boiler water can also cause 'Concentrating Film Corrosion Mechanism',
especially those units operating under a sliding pressure regime with demineralized
make-up and minimal blowdown, where the steam is used to drive a turbine, and
phosphate is utilized for internal treatment chemistry.

Stratification of the steam-water mixture will lead to steam blanketing and

subsequently produces high concentrations of caustic. Steam blanketing can
occur when the fluid velocity is insufficient to maintain turbulence and produce
thorough phase mixing. Direct removal of iron from the surface of the tube in an
area where a very high concentration of sodium hydroxide is present can cause
localised attack.

The high concentration of sodium hydroxide is caused by the departure from

nucleate boiling (DNB) at the surface. Poor circulation leading to DNB can produce
concentration mechanisms at the surface of a boiler tube. If the circulation is
inadequate to sweep away steam bubbles forming at the tube surface, the
bubbles begin to coalesce, producing a film or blanket of steam on the surface.

When units operating with sliding pressure shed load, unique and damaging effects
can be caused in many sections of the boiler. Oxygen corrosion can occur in the
economizer because dissolved oxygen generally increases in the feedwater cycle.
As air in-leakage and the pounds of oxygen through the condenser are constant,
if the feedwater flow is reduced with the pressure, the dissolved oxygen content
increases in direct proportion to the feedwater flow reduction. Iron oxide transport
to the unit could also result from economizer oxygen corrosion.

The way that steam is generated at the metal surface during sliding pressures can
also lead to tube metal damage. As the internal circuit is essentially at equilibrium
during steady state operation, a sudden reduction in pressure can release steam
bubbles from certain sections in the lower part of boilers not normally associated
with primary steam generation, such as mud drums, headers, or lower waterwall
tubes. Such sudden pressure reductions and the associated steam release can
cause localized occurrences of erosion-corrosion, especially along the tops of
horizontal or inclined tubes, like floor tubes or nose tubes. Erosion-corrosion can be
defined as the relative acceleration in metal loss due to the movement of the
corrosive fluid at the metal surface. When steam suddenly flashes at the boiler tube
surface, the protective oxide film can be eroded due to the force of the steam
bubble eruption, and the subsequent re-oxidation of the surface layer effectively
contributes to a loss in tube wall thickness. Non-vertical tubes repeatedly damaged
in such a manner can have longitudinal grooves produced along the top axes. The
tube wall thickness reductions can subsequently lead to stress-rupture by raising
the hoop stress on the tube.

In association with this type of erosion-corrosion mechanism, a concentrating film

corrosion mechanism can also contribute to wall loss, as corrosive levels of certain
boiler water solids can concentrate at the metal surface during steam blanketing.
Steam blanketing occurs after incomplete steam/water separation in horizontal or
inclined boiler tubes, and it is often associated with film boiling. A concentrating
film corrosion mechanism can also readily occur at the onset of departure from
nucleate boiling (DNB). DNB is the point when water vaporization no longer
produces discrete bubbles, but rather, the continuous film of water vapor is
produced. Whereas localized deposition of concentrated boiler water constituents
does normally occur at the tube metal surface during the formation of steam
bubbles, even during nucleate boiling, the deposits should fully re-dissolve in the
boiler water when the steam bubble leaves the tube wall. At the point of DNB, the
generation of steam bubbles is greater than the ability of the boiler water to rinse
and re-dissolve the concentrated solids. Concentrating mechanisms, such as a
DNB situation or the presence of waterside deposits, and phosphate hideout, allow
for rapid, sustained increases in the levels of corrosive solids at the surface, and the
release and reforming of the acid phosphate film during pressure fluctuations can
be a significant factor in boiler tube integrity, as boiler water solids can become
entrained at the tube surface. The protective metal oxide layer and the tube wall
itself can be corroded by underdeposit acid corrosion or caustic attack. Similarly,
steam can form in the areas between the metal surface and an overlying deposit
and create locally increased levels of very corrosive materials.

Sodium hydroxide added to the boiler water to maintain proper alkalinity can
concentrate and cause caustic corrosion as per the following reactions:

Fe3O4 + 4NaOH → 2NaFeO2 + Na2FeO2 + 2H2O (4)

FeO + 2NaOH → Na2FeO2 + H2↑ (5)

Alternatively, acid phosphate salts can become entrained at the tube metal
surface, leading to an underdeposit, low pH - phosphate corrosion mechanism.
Phosphate hideout, as described above, is the accretion of phosphate salts at
 tube surfaces under high load conditions, which causes an increase in pH and a
decrease in phosphate in the bulk boiler water. If the concentrated, acid salt
deposits are not fully rinsed away upon load reduction, i.e., during sliding pressure
operation, localized corrosion can occur in the underdeposit areas.

Whenever the tube wall thickness is reduced through an erosion-corrosion or other

corrosion mechanism, the hoop stress on the affected tube is effectively increased.
When the tube hoop stress from internal steam pressure equals the strength of the
tube material, the boiler tube fails from stress-rupture.

Another way in which sliding pressure operation of a boiler can contribute to tube
metal deterioration is through the specific, physical and thermal stresses imparted
on the tubes from extensive cycling, All boilers that experience numerous episodes
of start-up and shut-down, as well as those operating with sliding pressure, can be
subject to such stresses. Tubes with welded membrane walls are particularly
susceptible to stress-assisted corrosion. This can be attributed to the constraint on
the tubes and the differences in the coefficients of thermal expansion for steel and
iron oxide. The internal oxide layers can be locally disrupted in the areas of the
membrane wall welds. Specific corrosion can subsequently occur in the
unprotected areas, and the numerous cycles of expansion and contraction can
produce deep pits and failures. Phosphate has been detected in the bases of the
pits and cracks, and it is postulated that the phosphate salt detected is similar to
the acid salt that is formed during phosphate hideout, thus contributing to a
corrosion fatigue mechanism.

Sliding pressure can also contribute to thermal fatigue tube failures due to
irregularities in the boiler water flow caused by rapid changes in pressure. When
the tube is starved of its internal coolant, even for very short periods, slight degrees
of short-term, metal overheating can result. After numerous episodes of such short-
term overheating, circumferential cracking can result, a tell-tale sign of thermal
fatigue damage.

Lastly, the stresses put on tube surfaces can contribute to localized anodic areas
and sites for pit initiation. This is especially true of older units that have experienced
numerous cycles and frequent load changes. Such tubes (can exhibit breaks in the
oxide layer at repeatedly stressed locations, which ultimately results in pitting
corrosion.

6.3.2.2 Common Mechanism in Corrosion by Caustic

Under alkaline conditions, magnetite which is a copolymer of the hydroxides of iron
(III) and iron (II) can dissolve in a very complex manner involving a number of
equilibrium reactions. It can just dissolve as iron hydroxide or it can also be
considered as a homogeneous solid solution of FeO in Fe2O3. In this system, the
bonds of the divalent iron are more easily hydrolyzed. Therefore, the iron (II)
hydroxide system (aqueous) is considered to be in solution equilibrium with
magnetite as follows:

Fe3O4 + 2H2O ↔ Fe(OH)2 + 2FeOOH (6)

Under highly alkaline conditions, the ferrous hydroxide in solution may react as
follows:
 Fe(OH)2 + OH- + Fe(OH)3- ↔ HFeO= + H2O (7)
Fe(OH)2 + 2OH- + Fe(OH)4= ↔ FeO2= + 2H2O (8)

Caustic attack occurs in this manner through activation of the carbon steel surface
by removal of the magnetite or inhibiting its formation. These conditions lead to the
formation of a velvet-black, finely crystalline, reactive magnetite. It has low,
adherence and practically no protective effect.

As a result, the magnetite film is dissolved in the form of soluble ferrate ions as the
equilibrium illustrated in Reaction 6 is driven to the right via the removal of ferrous
hydroxide. In addition, further formation of magnetite is inhibited as the equilibrium
is driven to favor soluble ferrate ion formation instead of magnetite.

Concentrated solutions of hydroxyl ions can be found at boiler tube surfaces under
conditions where normal washing of the tube surface with boiler water is restricted
due to either the presence of deposits or the development of a film of superheated
steam at the surface of the tube.

6.3.3 Acid Corrosion

One mechanism for acid corrosion involves the thermally induced reaction of
permanent hardness salts in an operating boiler. Magnesium salts, for example,
can, in the absence of proper treatment (such as in a concentration boiling cell),
result in reduction in boiler water pH to the range of pH 4.0 to pH 4.5 by the reaction:

MgCl2 + 2H2O ↔ Mg(OH)2 + 2HCl (9)

This, and similar effects with calcium salts, can produce significant surface damage
following the general reaction:

[Fe: Fe2O3: Fe3O4] + HCl ↔ FeCl2 + H2O + H2 ↑ (10)

Condenser leakage, where hardness salts deplete the phosphate and hydroxide
ions, also can cause acid corrosion. In these cases, hydrochloric and sulfuric acid
can form. Hydrochloric acid can also form when the chloride ion finds its way
underneath a deposit where hydroxide ions have been depleted by ferrous ions
formed by a corrosion cell.

6.3.4 Overuse of Chelant Can Cause Corrosion

Chelants, particularly ethylene diamine tetra acetic acid (EDTA), (that is used in
boiler water treatment especially in conjunction with polymeric dispersants) can
also cause waterside corrosion if it is overused.

In the chelation process, it has been proposed that EDTA forms a sufficiently strong
complex with Fe+2 and Fe+3 such that precipitation with hydroxide is minimized. A
complex of Fe+2(EDTA) is the most probable, especially under boiler reducing
conditions which limits the oxidation of ferrous to ferric iron. However, the ferrous
ion solubility is limited by the high pH typically encountered.
 In industrial boilers, chemical species which interact or complex with ferrous iron
inhibit the formation of ferrous hydroxide (Reaction 2). Chelants, when used at high
residual levels, may inhibit magnetite formation by complexing ferrous iron,
interrupting the magnetite formation process.

Additional corrosive conditions, such as oxygen ingress, low or high pH and high
turbulence, can accelerate chelant corrosion by increasing general corrosion and
the subsequent attempt to form magnetite. The presence of excess chelant
residual complexes the ferrous iron formed and interferes with Reaction 2.

6.3.5 Excessive Use of Polymer Can Cause Corrosion

Polymer is also used for boiler water treatment exclusively to complex traces of
residual hardness salts or iron and subsequently maintaining these species either
dispersed or in solution in order to prevent deposition at the heat transfer surfaces.
However, in boilers where excessive concentrations of polymer are employed or
concentrated due to local effects, corrosion can result because, in a similar fashion
as in the case of chelant, the polymer can complex soluble ferrous ions and thus
preventing or retarding the formation of a protective magnetite film. This occurs
especially in high purity or demineralized boiler water makeup systems or when a
corrosive condition exists, such as oxygen attack, low or high pH, or high
turbulence.

6.4 External Corrosion and Erosion

There are a number of external corrosion mechanisms which result in the reduction
of tube wall thickness. One of the most common mechanisms, often called "the
high temperature corrosion mechanism", involves the deposition of alkali sulfate
based salts on the surface.

Na2O + SO3 ↔ Na2SO4 (11)

3Na2SO4 + Fe2O3 + 3SO3 ↔ 2Na3Fe(SO4)3 (12)
10Fe + 2Na3Fe(SO4)3 ↔ 3Fe3O4 + 3FeS + 3Na2SO4 (13)

The reaction of the alkali sulfate with ferric oxide at the tube surface produces the
complex intermediate trisodium iron trisulfate, which is generally molten at the
prevailing temperature. This molten sulfate reacts directly with ten moles of iron on
the tube surface to produce magnetite and sodium sulfate which is available to
react again with ferric oxide producing the trisodium iron trisulfate complex. This
cyclical corrosion mechanism can result in very rapid corrosion of the external
surface of the tubes in the higher temperature zones of the boiler, such as
superheaters and upper furnace wall tubes.

A second mechanism for external corrosion involves the reaction of hydroscopic

acidic sulfate salts with water vapor during boiler shutdown to produce an acidic
medium on the tube surface, which results in accelerated corrosion of the tube. A
third mechanism involves the decomposition of sodium pyrosulfates (Na2SO3 x
SO3) to yield SO3 which reacts with ferric oxide at the surface to product magnetite
(Fe3O4) and ferrous sulfide (FeS).
6.5 Corrosion control
The control of corrosion in a boiler environment is based on maintaining conditions
which enhance passive film formation.

Passivating agents, such as phosphate, may mitigate chelant attack by inhibiting

the corrosion process.

6.6 Monitoring and inspection

6.6.1 Hydrogen Monitoring

The hydrogen study clearly indicated increased hydrogen generation at higher
chelant concentrations indicating increased corrosion of waterside surfaces at
higher chelant concentrations. A hydrogen study was recommended to confirm a
reduction in hydrogen generation and the absence of a correlation between
hydrogen value (corrosion rate) and polymer concentration in the boiler water
over the normal range of polymer concentration.

6.6.2 Videoprobe
A videoprobe inspection of the boiler unit was recommended during the next
annual boiler shutdown to see if evidence of improved boiler passivation could be
observed.
 7
Chapter

7 CORROSION IN COOLING WATER SYSTEM

7.1 Introduction
Cooling water can be a very destructive force to the mechanical integrity of the
equipment that uses it. This is supported by the fact that most cooling water systems
have oxygenated water circulating in them, which is an aggressively corrosive
medium. Cooling water related failures can still occur despite the use of good
water treatment programs and expensive corrosion prevention measures. If a new
cooling water system is brought on-line with no pre-treatment, then localized
corrosion can begin before the treated water has a chance to establish a
protective film. Once localized corrosion has begun, the micro-environment inside
the pit is often sufficiently different to the bulk solution to render ordinary inhibition
ineffective.

The common fallacy stating that as long as the water chemistry meet certain
guidelines (e.g. residual chlorine level), equipment failure will not occur is not
always true. Consideration should be taken for the size of the equipment, wide
variety of metallurgy used in cooling water treatment, water flow rate/stagnation
that is relevant to bacterial growth, thermal gradient that cause deposition,
support of HE tubes, location of the chlorine feed station, and the consumption
rate of the chlorine, just to name a few, for a better avoidance of cooling water-
induced damages.

7.2 Common Degradation Mechanism to Cooling Water System

Problems in cooling water system are generally classified into the following three
categories: corrosion, scale and biofouling. These problems can cause reduced
equipment life, reductions in operating efficiency, leaks what may allow process
fluid and vapors into the environment, etc.

The importance of cooling water system integrity on plant reliability and overall
plant operating costs, has been the driving force to combat corrosion and
deposition problems in cooling water systems. The reliability of cooling water
systems can be compromised by fouling from a variety of sources, including
general and localized corrosion, and microbiologically influenced corrosion (MIC).
In cooling water systems, pitting, crevice and/or underdeposit corrosion are the
most commonly observed forms of localized corrosion. A recent survey shows that
more than 70% of cooling water side failure of carbon steel heat exchangers are
due to pitting corrosion [Field Usage Data on Soft-Steel Heat Exchangers in Cooling
Water Environments, Japanese Society of Chemical Engineers, Chemical
Equipment Materials Committee, Corrosion Subcommittee, October 1990.].
Corrosion, for example, can affect support columns at joints or rivets inside a
cooling tower through crevice corrosion and bi-metallic corrosion

This to indicate that localised corrosion is more of a serious and the difficult
degradation mode in cooling water system than the general corrosion. If the mild
steel heat exchangers (typical thickness = 2 mm) only suffered from general
corrosion attack within the refinery acceptable standard of 3 mpy (or 0.075 mm/yr),
the service life would be longer than 20-30 years. However, this is not always the
case, indicating that other mode of attack is in operation.

Another prevalent type of corrosion in cooling water system is microbiologically

induced corrosion (MIC). Active biological corrosion is considered a direct
chemical interaction of organisms with materials to produce new corrosion
chemistries and the marked acceleration of preexisting electrochemical corrosion
reactions. Area effects, temperature, electrode polarization, flow, oxygen
concentration, electrolyte conductivity, and pH may all influence electrochemical
potentials, and hence, the rate of corrosion. Another manner by which this type of
biological corrosion can become a problem is through the formation of acids or
ammonia, directly attacking the base substrate. This overall form of attack usually
creates intense localized areas of damage. In the passive mode, on the other
hand, the attack is usually more extensive. Here, biological matter acts as a shield
under which corrosion concentration cells may develop. In cooling water systems,
it usually accounts for the greatest amount of metal loss.

Although influential, chemical corrosion in cooling water is not as common, or

prevalent, as other types of corrosion. Corrosion through an acidic solution is usually
caused by an upset condition, such as an acid spill, process leak, airborne
contamination, an overfeed of a water treatment chemical, etc. As a general rule,
more acidic solutions have a greater tendency to corrode iron and other metals.
Other factors which influence acid corrosion are metallurgical content,
temperature, water turbulence, surface fouling, corrosion-product formation,
metal-ion concentration, and chemical treatment. In contrast, corrosion in an
alkaline solution is even less common in cooling water systems. Under normal
conditions, caustic solutions produce little corrosion on steel, stainless steel, cast
iron, nickel and nickel alloys. Many precipitated minerals, such as those produced
from side reactions, e.g. phosphates, carbonates and silicates, actually reduce the
rate of corrosion.

Another most prevalent and common types of corrosion found in cooling water
systems is galvanic, or bi-metallic, corrosion. For this type of corrosion, certain
criteria must be met. For one, an electrochemical interaction of two or more metals
with a sufficiently distinct galvanic potential is needed. In other words, at least two
metals with varying degrees of electrochemical charge are needed through
which an electrolyte, the second crucial point, may impart an ionic current. For this
current to allow the exchange of electrons between the two metals, a physical link
must be established for the corrosion to take place. Once these points are
established, corrosion will take place at the location of the more active metal - the
anode. Various scenarios and possibilities exist whereby the corrosion rate may
increase or decrease. Galvanic corrosion may even occur in stainless steels, for
instance.
 As is evident, corrosion in cooling water system can take forms. Either way,
corrosion and its associated effects place a great strain in the proper operation of
a cooling water system.

7.3 Corrosion Control

For corrosion control, the cooling water make-up is treated with polyphosphate,
together with copolymer for scale control.

It is normal to experience that despite a very conscientious effort to monitor and

control the treatment chemical levels, the corrosion rate on carbon steel heat
exchangers still ranges from < 2 mpy to 10 mpy, and heat exchangers still need to
be pulled regularly to remove the scale build-up. However, premature failures still
happen to equipment that use cooling-water, and some of them are not directly
related to the effectiveness of the water treatment program.

For cooling water treatment, a combination of phosphate-based and polymer

chemicals is normally used to control corrosion and scale (principally calcium
phosphate) deposition, and also as dispersant for iron oxide particles. For example,
a combination of ortho-, polyphosphate, phosphonate, azole and polymer for
cooling water treatment can yield cause reduction in corrosion rates as low as less
than 1.5 mpy (0.038 mm/y) for carbon steel and less than 0.3 mpy (0.008 mm/yr) for
copper or brass alloys. Effective deposit control is also easily maintained in the bulk
cooling water and on heat transfer equipment for most routine cooling water
conditions. (9) R. C. May, et. al., U. S. Patent #4,303,568, "Corrosion Inhibition
Treatments and Methods," December 1, 1981.

7.3.1 Use of Biocide

Water treatment, usually with biocides, is the most common approach for
mitigation of MIC based upon its ease of use. However, the use of a biocide
treatment alone is rarely effective for a heavily fouled system and can aggravate
a corrosion condition that was established initially by microbiological activity but
which may be proceeding with no further microbiological involvement. In such
cases, the use of a biocide, especially an oxidizing biocide, may prevent further
infestation but will provide alternative and stronger half-reactions that fuel the
corrosion activity. Biocide treatments alone are only likely to be effective when
microbes actively participate in the corrosion process (e.g., by catalysis of the
reduction of oxygen, or by cathodic depolarization in anaerobic systems), or when
microbes provide alternative reactions that exacerbate corrosion. In summary,
biocide treatments are a very effective preventive measure against MIC, however,
the condition of the corrosion process must be well understood before biocide
treatment is used to remediate an apparent MIC condition.

Microbial colonization of metallic surfaces will produce an impediment to flow

(generally minor) or to heat transfer and can produce aggressive, localized
conditions leading to pitting. Any general type of treatment must be sufficiently
aggressive that small, well established local conditions are remediated. Such
treatments are unlikely to be effective for MIC without serious side effects. Similarly,
the nature and purpose of biofilms is self-preservation. As such, mature biofilms are
highly resistant to treatments that contact only the outermost layer of the film.
 Chlorine is the most common micro-biocide used to treat the water. Due to some
significant value of Total Bacteria Count, the free chlorine (or hypochlorite) level is
normally kept at relatively high levels of 0.5 – 1 ppm. At the same time, the pH is
kept at a range of 7.0 to 7.3 to minimize the scaling problems that are common in
many operation units. Chlorine is by far the most cost effective. It is much less
expensive than non-oxidizing biocides and certain stabilized bromine treatments.

Unfortunately, chlorine does not react selectively with only biological matter. Other
constituents found in a cooling stream can also react with chlorine. With regard to
water treatment chemicals, phosphonate and azole degradation in the presence
of chlorine has been a significant concern. Phosphonate degradation not only
decreases the calcium carbonate inhibition properties of a cooling water
programs but can contribute additional orthophosphate leading to either a
deposit control problem both from calcium phosphate and calcium carbonate or
a more expensive program by requiring additional polymer and phosphonate to
alleviate deposition. Azole destruction by chlorine leads to ineffective corrosion
control for copper and its alloys. Additionally, copper concentrations will increase
in a cooling water when corrosion is not properly controlled, directly impacting
carbon steel corrosion control. The copper ions plate-out on low carbon steel
causing severe galvanic corrosion. This corrosion usually takes place in the form of
pitting and is quite destructive. It is the most serious problem experienced in
stabilized phosphate treatment programs in which chlorine is used for microbial
control.

7.4 Carbon Steel

7.4.1 General Corrosion in condenser

It is common to experience corrosion on the tubesheets and the internal surface of
the heads of a condenser with significant damage (i.e. spalling) on the thick-film
epoxy type lining. General corrosion can be observed in the tube-to-tube sheet,
normally adjacent to the seal welds, and also in the base metal. The condenser
has carbon steel (CS) tubesheets and heads but 316L stainless steel (SS) tubes.
Cooling water is in the tube side.

7.4.2 Lined Carbon Steel Failure

Carbon steel lined with high-bake phenolic can experience pitting at holidays in
the lining. The pits and corrosion in the tubes were due to the high residual chlorine
and low pH in the cooling water as a result of coating failure. There were also
coating damages and corrosion on the tubesheet, but the leaking tubes were
more dramatic during the leak test.

7.4.3 Prevention
The chlorine residue in the cooling water was also recommended to be maintained
at no higher than 0.2 gm/L, and the pH between 7.5 – 7.8. The flow rate of the
cooling water also needed to be increased to 3 ft/sec or above.
7.5 Carbon Steel in Cooling Tower
Cooling tower water is corrosive to carbon steel equipment and the corrosivity
depends on the water treatment, process conditions and outside environment. The
corrosion rate of CS in tower water is estimated to be as high as 10 mpy (0.25
mm/yr). The corrosion rate can also be enhanced by the galvanic effect of the
stainless steel tubes and the carbon steel tubesheets and heads.

7.5.1 Prevention
Coatings and sacrificial anodes are normally used for corrosion protection on
carbon steel HE.

However, coatings can be damaged by various reasons like poor application, high
process temperature, or erosion. High operating temperature can also cause the
coating failure. Since there were no anodes to protect the tubesheets and heads,
corrosion started as soon as the coating failed.

This problem may solved by weld cladding the tubesheet surface with 316L SS, and
the installation of magnesium alloy sacrificial anodes.

7.6 Stainless Steel Corrosion

Stainless steel like 316L used for tube can experienced corrosion pitting and stress
corrosion cracking, started from OD where cooling water is on the shell side.
Chloride in cooling water is the cause of SCC on the SS316L tube. In this case, the
contributing factors are (1) an undersized cooler, which caused higher
temperature on the shell side that promoted deposition. (2) Intermittent service
conditions when cooling water flow was not continuous, which lead to the boil out
of the water (with the solutes left behind), (3) poor water treatment, (4) low
blowdown cycle # on the cooling tower water.

SS316L cooler tube can also suffer corrosion fatigue if the tube is located at areas
where it can suffer shock or vibration from water hammering in the cooling water
system.

7.6.1 Prevention
Recommendations to avoid these problems are (1) Design an adequately sized
sample cooler for the service. (2) Use a close-loop glycol based cooling water
system or keep the cooling water continuous running. (3) Keep the cooling water
outlet temperature to be below 60°C. (4) Disassemble the cooler and clean the
interior on a regular basis (e.g. Yearly).
7.7 Monitoring

7.7.1 Bioprobe for MIC

Many of the same methods used to monitor system performance or system
corrosivity can be used to monitor microbiological effects. Commonly used
corrosion monitoring methods include corrosion coupons, electrical resistance
probes, linear polarization resistance methods, and advanced methods such as
electrochemical noise or electrochemical impedance spectroscopy. Prototypic
materials exposed to typical water chemistries and treatments should provide an
indication of the effectiveness of the treatment after some treatment period. These
methods are good for monitoring corrosion but are not specific to MIC.

System performance monitoring, either from pressure drop devices, from operating
system heat exchangers or from model heat exchangers in sidestreams, can
provide a useful indicator that system surfaces are becoming fouled, however,
they do not provide a complete picture of the source of the fouling and provide
no information on corrosion.

As an alternative, an electrochemical probe can be installed to provide

information on the activity of microbiological fouling on metallic surfaces in real
time. A primary objective of the probe is to provide and monitor the presence of
an active biofilm on metal surfaces. By continuously monitoring the probe, the
plant inspector can collect real time information on the condition of
microbiological fouling. The probe permits the operator to plan inspections, modify
system operation, scheduling cleaning, adjusting the timing and dosage of
chemical treatments or mitigation activities based upon measured biofilm activity
on a surface.