Sie sind auf Seite 1von 581

Corrosion and the FCC Process

SECTION 1 CORROSION AND THE FCC PROCESS

1.0 Overview: Corrosion and the FCC Process..................................................... 1-1

1.1 High Temperature Oxidation .............................................................................. 1-3

1.2 Corrosion by High Temperature Hydrogen Sulfide ......................................... 1-7

1.3 Polythionic Acid Stress Corrosion Cracking.................................................... 1-9

1.4 Corrosion by Ammonium Bisulfide ................................................................. 1-17

1.5 Corrosion and Stress Corrosion Cracking by Amines .................................. 1-21

1.6 Catalyst Erosion ................................................................................................ 1-27

1.7 Carbonate Stress Corrosion Cracking ............................................................ 1-31

1.8 Chloride Stress Corrosion Cracking ............................................................... 1-35

1.9 Corrosion at Injection Points ........................................................................... 1-39

1.10 Stress Corrosion Cracking by Caustic (Sodium Hydroxide) ........................ 1-41

1.11 Wet H2S Cracking .............................................................................................. 1-43

1.12 Corrosion by Naphthenic Acids....................................................................... 1-47

1.13 Interpretation of Corrosion Data...................................................................... 1-49

1.14 Corrosion by Ammonium Chloride.................................................................. 1-55

1.15 Sulfide Stress Cracking .................................................................................... 1-57

1.16 Corrosion of Dead Legs.................................................................................... 1-61

1.17 Corrosion by Ammonium Polysulfide ............................................................. 1-63

1.18 Corrosion at Specification Breaks................................................................... 1-65

1.19 High Temperature Carburization ..................................................................... 1-67

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.0 OVERVIEW: CORROSION


AND THE FCC PROCESS

FCC, or Fluid Catalytic Cracking, is a refining process in which heavy oils of little
commercial value are broken down into lighter, more useful products through the use of
high temperature and a catalyst.

The catalyst used in FCC Plants is a fine powder which moves through the plant in a
fluidized state. In fluidization, a gas is blown through the powdered catalyst at a
velocity sufficient to suspend it in the gas. To suspend the catalyst in the gas requires a
relatively high gas velocity, creating a potential problem of erosion in places where the
solid-gas mix contacts the internal surfaces of the plant equipment. This solid-gas
mixture can then be piped from vessel to vessel as if it were a fluid.

During the FCC process, the fluidized catalyst moves back and forth between the reactor
and regenerator vessels as follows:

In the reactor, the catalyst is contacted by oil at high temperature. The oil is
cracked to form lighter components, which exit the reactor on their way to
distillation. In the cracking reaction, carbon is deposited on the catalyst, which
eventually deactivates it.
Catalyst activity is restored by circulating the catalyst to the regenerator vessel.
Air is blown into the regenerator to burn off the carbon and restore the activity.
The regenerated catalyst is then returned to the reactor.
Flue gas exiting from the regenerator is cooled, and fine catalyst particles not
removed by the regenerators cyclones are removed in electrostatic precipitators,
external cyclones, or stack scrubbers.

Numerous corrosives exist or are formed in the fluid catalytic cracking process:

In the regenerator, metals are exposed to oxidation by hot air and oxygen-
containing flue gas at temperatures of approximately 1300F. These flue gases
also contain sulfur oxides, which can cause severe corrosion when the gas is
cooled below the dew point.
Sulfur compounds in the FCC feedstock are partially converted to hydrogen
sulfide in the reactor. This conversion can cause corrosion at high temperatures.
Nitrogen in the feed is converted partially to ammonia and partially to cyanides.
When the reactor effluent stream components are cooled below the water dew
point, ammonia and H2S are absorbed in the water and react to form corrosive
ammonium bisulfide.
Cyanides absorbed in this water act as potent corrosion accelerators and also
lead to hydrogen blistering, sulfide stress cracking, and other related forms of
cracking damage.
To make matters worse, the reactor effluent also contains carbon dioxide. As a
result, we find bicarbonates or carbonates in many of the process streams that
have been cooled below the water dew point. Under certain conditions, their
presence can cause carbonate stress corrosion cracking.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-1
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Sulfur and nitrogen in the FCC feed are the most important factors that affect corrosivity.
The effect of individual corrosives is discussed on the following pages.

Confidential
Copyright 2003 by ChevronTexaco
1-2 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.1 HIGH TEMPERATURE OXIDATION

Oxygen reacts with steel and other iron-based alloys to form an iron oxide. The type of
oxide that results is temperature dependent. At ambient temperatures, and in the
presence of moisture, a red iron oxide appears, which we all recognize as ordinary rust.
In simplified form, this compound has the chemical formula Fe2O3. At high temperatures
(over 1000F), a different oxide appears. It is hard, black, tends to be quite smooth, and
has the formula Fe3O4.

When oxidation rates are low, the metal looks Higher oxidation rates result in a
black, and it may be difficult to see that scale is thicker, but still metallic-looking
present. Higher oxidation rates result in a thicker, scale, making it difficult to tell that
but still metallic-looking scale, making it difficult to significant attack has occurred.
tell that significant attack has occurred. An
energetic beating with a hammer to remove the
oxide layer reveals the thickness of the scale and the extent of the corrosion. Generally,
we can assume that the scale is about 5 times as thick as the metal that has been
consumed.

When oxidation rates are very high, the scale begins to crack and spall off because of
the internal stresses that build up in it as it forms.

In summary, we can categorize the oxidation of steel as follows:

Below 212F Forms red rust, but only when moisture is present.
212-1000F Essentially no corrosion.
Over 1000F Forms a hard, uniform black oxide.

High temperature oxidation In an FCC Plant, we use air in the regenerator to burn
presents a potential FCC carbon off of the catalyst. During this process, high
Plant problem only in the temperature and the presence of oxygen produce the
regenerator and its flue gas conditions under which oxidation of metals will occur.
system. High temperature oxidation presents a potential FCC
Plant problem only in the regenerator and its flue gas
system.

Oxidation rates at high temperature are affected by time, temperature, and oxygen
concentration. Oxidation follows a parabolic rate law, in which the corrosion rate
decreases with time. In order to attack the metal, oxygen in the environment must first
diffuse through the scale layer. The diffusion rate progressively decreases as the scale
builds up, and corrosion rates drop off. We can express the parabolic rate law
mathematically by saying that the amount of corrosion penetration is proportional to the
square root of time.

If the oxide scale were never disturbed, the corrosion rate would slow gradually over the
life span of the plant. But we cannot rely on this stabilization. As the scale gets thicker,
it may begin to spall off. It may also spall as a result of thermal stress from startups and
shutdowns. Exposure to moisture or flue gas condensate during a plant shutdown may
cause it to change chemically into red rust.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-3
permission of ERTC
Fluid Catalytic Cracking Process August 2003

To allow for these possibilities, we base our corrosion rate predictions on the assumption
that the scale remains intact for 1 year. This is a conservative assumption, but safe. If
you were looking for the most probable rate for a 3-5 year run, rather than a guaranteed
safe rate, you could make the following assumptions:

That the metal penetration over 4 years is twice that of 1 year.


That the average corrosion rate over the period is half of what it is during a
1-year span.

Some years ago, the organization that is now ERTC made a detailed study of oxidation
rates at high temperature. A massive amount of data were collected. Corrosion rates
were correlated on the basis of temperature. To account for the time factor, all rates
were normalized to a 1-year time span. Results of this study are shown in Figure 1.1-A.

The only data available


for this correlation were
rates in an air
environment. The affect
of oxygen concentration
was not taken into
account. In the early
days of FCC, this fact
was important because
regenerators ran at
relatively low
temperatures compared
to todays regenerators.
Also, oxygen
concentrations were as
low as 0.1% as
compared to 20% in air.
But as technology has
improved over the
years, conditions in
FCC regenerators have
become more severe in
terms of temperature
and oxygen
concentration. We now
assume that oxygen
levels in a regenerator
are high enough so that
corrosion rates do not
differ greatly from those
that we see in air at the
Figure 1.1-A. Effect of temperature on same temperature.
oxidation rate of steels.

Confidential
Copyright 2003 by ChevronTexaco
1-4 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

We can protect against high temperature oxidation by using alloys or insulating metal
surfaces to keep them cool. The shell of an FCC regenerator is kept cool and oxidation-
free by using an insulating refractory. (The refractory also protects against catalyst
erosion, which is a subject discussed elsewhere in this document.) Because internal
parts of the regenerator cannot be protected by thermal insulation, we must use alloys
here.

Low-chrome steels like 1-1/4 Cr - 1/2 Mo offer a strength advantage over carbon steel
but are not measurably better in oxidation resistance. 5 Chrome oxidizes at somewhat
reduced rates. 12 Chrome is even better. However, for parts operating at full
regenerator temperatures, we use the austenitic (300 Series) stainless steels such as
Type 304. Cyclones and the hexmesh supporting refractory are typical examples of
stainless steel use in the regenerator.

In selecting the appropriate grade of stainless steel for use in an FCC regenerator, the
potential problem of sensitization and polythionic cracking must be considered. This
topic is discussed elsewhere.

Figure 1.1-A shows the oxidation rates of steel and various alloys in air. The beneficial
effect of the higher alloy content is readily apparent. The chromium content of the steel
determines its resistance to high temperature oxidation. Corrosion rates progressively
decrease as chromium content rises.

HOMEWORK QUESTIONS

1.1 High Temperature Oxidation

1) True or False: The oxide scale on a metal is generally much thicker than the metal
that has been consumed.

2) For carbon steel, at what temperature will metal loss exceed 10 mils in a year?

3) Whats the primary element we add to steels to improve their oxidation resistance?

4) True or False: Oxide layers formed at high temperature are easy to visibly
distinguish from non-oxidized metal surfaces.

5) True or False: For most steels operating at high temperature, a temperature


increase of 100-200F will not significantly increase their oxidation rate.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-5
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
1-6 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.2 CORROSION BY HIGH


TEMPERATURE HYDROGEN SULFIDE

Hydrogen sulfide (H2S) is formed in the FCC reactor by thermal decomposition of


organic sulfur compounds in the plant feed. H2S is corrosive to iron and steel at high
temperature, forming iron sulfide as a corrosion product. It also attacks other metallic
elements such as nickel and copper, forming nickel sulfide and copper sulfide,
respectively. Temperature and H2S concentration control corrosion rates. Changing to
a higher sulfur feed is likely to lead to higher corrosion rates.

H2S is corrosive to steel at temperatures above approximately 500F. In the FCC Unit,
we see potentially corrosive conditions in the reactor; the transfer line from reactor to
main fractionator; the main fractionator itself; and in the fractionator bottoms, sidecut
piping, and exchangers.

Although H2S corrosion rates usually increase with temperature, FCC experience seems
to deviate from this generality. H2S corrosion in reactors occurs at far lower rates than
we would expect. For example, most of our FCC Plants use carbon steel cyclones in
reactors, and these cyclones last for 7 to 10 years at operating temperatures of 900-
960F. Information first published around 1950 indicates that high temperature H2S
attack increases up to about 800F, and then decreases with further increases of
temperature. This information may be true, but we arent sure why. It may be that the
sulfide scale layer becomes plugged with coke, which retards sulfide diffusion through
the scale. Even though the life obtained on steel reactor cyclones is considered
satisfactory, some ChevronTexaco plants have upgraded to Type 405 or 410 steel.

Resistance to high temperature H2S attack is achieved by adding chromium to iron or


nickel-based alloys. Corrosion resistance increases with chromium content:

Of the iron-based alloys, 5 chrome is the lowest to have significantly better


resistance than carbon steel. The 12% chromium steels (Types 405 or 410) are
not corroded by H2S under any conditions likely to be encountered in an FCC
Unit. The 18-8 stainless steels (Types 304, 321, 347) are also totally corrosion-
resistant and are widely used in preference to 12 chrome because of much
superior mechanical properties.
Nickel-based alloys have good resistance to hot H2S if they contain sufficient
chromium. At equal chromium contents, alloys with more iron and less nickel
perform better. The Incoloy alloys are more resistant than the Inconels. Pure
nickel and Monel, a nickel-copper alloy, do not have useful resistance.
Copper-base alloys do not resist H2S at elevated temperature. Aluminum is
resistant but is not used because of poor strength at high temperature.

Resistance to hot H2S can also be achieved by insulating metal surfaces internally to
keep them cool. Typical composition of equipment in an FCC Plant is as follows:

Reactors are usually either internally insulated carbon steel (cold wall) to lower
the shell temperature and protect the shell from attack, or they are 1-1/4 Cr (hot
wall) with refractory applied only for erosion protection.
Reactor cyclones are typically built of carbon steel.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-7
permission of ERTC
Fluid Catalytic Cracking Process August 2003

The reactor effluent line is most often 1-1/4 Cr and shows acceptable sulfidation
rates.
Main fractionator cladding and internals, where alloy, are Type 405 or 410
stainless.
Hot sidecut and bottoms piping from the main fractionator and tubes in the hot
exchangers are 5% chrome steel, sometimes 9% chrome steel.

High temperature hydrogen sulfide corrosion does High temperature hydrogen sulfide
not occur rapidly enough in FCC Units to create the corrosion does not occur rapidly
probability of a catastrophic failure. Attack is easily enough in FCC Units to create the
found by UT because rates are nominal and attack probability of a catastrophic failure.
is quite uniform. A reasonably well-conceived Attack is easily found by UT
inspection program should avoid surprises. because rates are nominal and
attack is quite uniform.

HOMEWORK QUESTIONS

1.2 Corrosion by High Temperature Hydrogen Sulfide

1) True or False: All nickel-based alloys are highly resistant to hot H2S corrosion
(sulfidation).

2) Whats the commonly accepted threshold temperature for hot H2S corrosion?

3) Whats the primary element we add to steels to improve their hot H2S corrosion
resistance?

4) Why might hot H2S corrosion rates of carbon steel in the fractionator bottoms system
exceed those of carbon steel in the reactor even though the reactor runs hotter? Is
carbon steel an acceptable material of construction for hot fractionator bottoms
piping?

Confidential
Copyright 2003 by ChevronTexaco
1-8 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.3 POLYTHIONIC ACID STRESS


CORROSION CRACKING

Polythionic acid stress corrosion cracking affects only the austenitic (300 Series)
stainless steels and related austenitic alloys such as the Incoloys and Inconels.
Cracking can occur with astonishing speed; in laboratory tests we have seen cracks
grow at roughly 1/2 inch per hour. We have not seen major explosions or fires as a
result of polythionic acid cracking, but we have seen many leaks.

In order to understand the principles of polythionic acid cracking, we need to look at a


few of the basic metallurgical concepts:

Recognize that all metals are crystalline solids unless in the liquid or vapor state.
Crystallization does not represent some kind of a defect, as many people think;
nor can a pump shaft which has suffered fatigue be said to have crystallized.
These concepts are simply not true.
When a molten metal solidifies from the liquid state (crystallizes), it forms crystals
or what metallurgists call grains. The grain boundary is where the grains
merge. Impurities in the metal tend to collect in these grain boundaries as the
metal cools and solidifies. Also, if solid metals are welded or heat treated,
certain elements may migrate to the grain boundaries and accumulate there.
With austenitic stainless steels, the material that accumulates in the grain
boundaries is chromium carbide (Cr23C6), the reaction product between carbon,
with chromium in the alloy. Chromium carbide precipitation in the grain
boundaries is called sensitization. Sensitization occurs when a stainless steel
reaches temperatures in the approximate range of 800-1500F. (Experts do not
always agree on the exact upper and lower limits.) Above 1500F the carbon
remains dissolved in the metal. Below 800F temperatures are too low for the
carbon to move around within the metal structure. Heating within this critical
temperature range may occur during welding, heat treatment, or elevated
temperature operation.

The problem with sensitization is that it creates a zone at the grain boundaries
susceptible to selective corrosion. The chromium carbide precipitate does not corrode;
the parent metal immediately adjacent to the carbide corrodes. Next to the carbide
precipitate, much of the chromium is sucked out of the parent metal by reacting with the
carbon to form the carbide, leaving a thin band of low chromium material. This band
corrodes rapidly when exposed to certain acidic corrosives. Figure 1.3-A shows a
photomicrograph of a stainless steel before and after sensitization.

NOTE: Field metallography


can detect sensitization.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-9
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Not Sensitized, 304 SS Sensitized, 304 SS


Magnification 500x 10,000 Hours at 1000F
Magnification 500x
Figure 1.3-A. Photomicrograph of stainless
steel before and after sensitization.

Sensitization itself does not really damage the metal, but it leaves the metal open to
these special forms of corrosive attack:

Intergranular corrosion.
Intergranular stress corrosion cracking.

Figure 1.3-B. Photomicrograph of intergranular corrosion.


The difference between these
two phenomena is simply that
intergranular corrosion attacks
all the grain boundaries,
causing the material ultimately
to fall apart into a pile of sugar-
like metal grains, whereas
intergranular stress corrosion
cracking does not attack all
boundaries. The latter form of
corrosion occurs when residual
or applied stress in the
presence of certain specific
corrosives causes cracks that
penetrate the metal along
sensitized grain boundaries.
The type of attack we see
depends on the specific
corrosive present. Figure 1.3-B shows an example of intergranular corrosion;
Figure 1.3-C illustrates intergranular stress corrosion cracking.

Confidential
Copyright 2003 by ChevronTexaco
1-10 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

It is unclear why intergranular


corrosion occurs in some cases and
intergranular cracking in others.
However, we can relate these forms
of corrosion to certain specific
corrosives. Intergranular corrosion is
a common problem in chemical
plants. In the context of FCC Units,
our concern is intergranular stress
corrosion cracking, and the chemical
agent said to cause it, namely
polythionic acid.

Whether cracking is actually caused


by polythionic acid or by wet sulfur
dioxide (SO2) is a disputed topic.
We can easily create cracking in the
laboratory in sulfurous acid (H2SO3),
which is simply SO2 dissolved in
water. However, the conviction
within the industry is that the cracks
are caused by water solutions of the
polythionic acids.

Polythionic acids are a family of Figure 1.3-C. Photomicrograph of


acids having the general formula intergranular cracking.
H2SxO6. These acids are formed when iron sulfide
scale on the metal is exposed to oxygen (air) in the presence of liquid water. This is not
a problem during normal operation because free oxygen is not present in the
hydrocarbon system (reactor, main fractionator, and downstream), and where oxygen is
present (the regenerator system), temperatures are far too high for liquid water to exist.

However, during shutdowns, equipment may be opened, allowing the sulfide scale,
which forms during normal operation, to contact air. The result is that iron sulfide
oxidizes to iron oxide, liberating SO2 in the process. We can readily see the result:
corrosion-formed scale in the plant (iron sulfide) is black when the equipment is first
opened but soon turns red as it converts to iron oxide. If liquid water is present, we have
an aqueous solution of SO2 (sulfurous acid). This solution further reacts with any
unoxidized scale to form a complex mixture of acidic sulfur compounds including
polythionic acids.

We must remember also that SO2 in water causes polythionic acid cracking. SO2 is a
normal component of regenerator flue gas, as is water vapor. Therefore, simply cooling
the regenerator while it contains flue gas can result in water condensation, absorption of
SO2 in the water, and formation of sulfurous acid which, like polythionic acid, can cause
cracking.

To summarize, there are two conditions which are simultaneously required for
polythionic acid stress corrosion cracking to occur:

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-11
permission of ERTC
Fluid Catalytic Cracking Process August 2003

The metal must be sensitized as a result of carbide precipitation at the grain


boundaries.
The sensitized material must be exposed to polythionic acids or sulfurous acid.

Cracking can be prevented in the following ways:

Prevent sensitization.
Prevent the formation of polythionic acids or wet SO2.

A sure way to prevent cracking is to use alloys which do not sensitize, either during
equipment construction or subsequent operation. The common grades of austenitic
stainless steel differ widely in their resistance to sensitization, as can be seen in the
following chart:

Type 304 Sensitizes during welding along a narrow band next to the weld. It also
sensitizes during heat treatment if it is slowly cooled from the heat
treating temperature through the critical 800-1500F range. And, finally, it
sensitizes during high temperature operation above 800F. To allow a
factor of safety, we consider Type 304 likely to sensitize at temperatures
above 700F.

Type 304L More difficult to sensitize than Type 304 because it must remain within
the critical temperature range for a longer period. Type 304L can be
welded and heat treated without sensitizing. However, it sensitizes on
prolonged exposure above 800F. Like 304, limit 304L to 700F. Above
this temperature, its crack resistance is no better than 304, and it has
lower strength at high temperature.

Type 321 Even better than 304L. It can be welded or heat treated without
sensitizing, subject to certain limits. It can also be used at steady state
operating temperatures as high as 850F. There are two conditions
under which we worry about Type 321 sensitizing:

Long-term service above 850F.


Prolonged heat treatment in the middle of the sensitization range.
This is a problem only with heavy-walled, stainless clad chrome-moly
equipment which see extremely long heat treating times in the
sensitizing range. This is not a concern in FCC Units.

Type 347 The best of the common stainless steelsvery resistant to sensitization.
It can be welded or heat treated and sensitizes only if subjected to
temperatures above 900F for a year or more.

Resistance of these alloys to sensitization is summarized in Figure 1.3-D. This chart is


called a TTS (time-temperature-sensitization) diagram. It shows the times and
temperatures of high temperature exposure at which each of the above-mentioned alloys
sensitize. Note that for each alloy, the sensitization range follows a loop. At
temperatures above or below or at times less than the loop shows, sensitization does
not occur.

Confidential
Copyright 2003 by ChevronTexaco
1-12 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

Curves correspond to susceptibility to intergranular corrosion and polythionic acid SCC. For
conditions within the loop for a given alloy, it will be sensitized. Outside its loop,, the alloy
will not be sensitized. These are average, typical curves to be used for guideline purposes only.

347 Sensitized

304L Sensitized

304 Sensitized

321 Sensitized

Figure 1.3-D. Time-Temperature-Sensitization


(TTS) curves for the commonly-used 300-Series
stainless steels.

One peculiar effect shown by the TTS curves is the


Recovery from sensitization may gradual recovery from sensitization that occurs on
be one of the more important very long exposure in the sensitization range. If we
reasons why we have so little follow a horizontal line across the TTS chart in the
problem with polythionic direction of longer times, we note that sensitization
cracking in regenerators in spite first increases, then peaks out, and eventually
of the fact that they operate well decreases. Given enough time at high temperature,
within the sensitization range for the metal will no longer be susceptible to polythionic
stainlesses. cracking. We speculate that this recovery from
sensitization may be one of the more important
reasons why we have so little problem with polythionic cracking in regenerators in spite
of the fact that they operate well within the sensitization range for stainlesses.

These curves provide guidelines for our use of stainless steel in FCC Units. Clearly an
FCC Plant operates at such high temperatures that any grade of stainless steel we use
may ultimately sensitize in service. Also, Type 347 sensitizes more slowly than other
grades of stainless, and if it does, the sensitization takes longer. A further advantage of
Type 347 is that it recovers its stability (desensitizes) more quickly than the other grades
of stainless. In spite of these advantages, we do not advocate the use of Type 347 in
reactor and regenerator systems.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-13
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Two factors argue against exclusive use of Type 347:

Embrittlement
Experience

Sigma phase formation may cause stainless steels to embrittle eventually at regenerator
temperatures. Sigma is a brittle crystalline phase that forms in stainless steels over long
periods of time at temperatures over 1100F. It results in brittleness and loss of impact
strength, which is most apparent when the plant is down and cool but not when it is hot
and operating. Embrittlement of Type 347 is faster and worse than that of Type 304.

Type 304 has been used successfully for many years without cracking. Prior to the
introduction of high temperature regeneration in the 1960s, there was not nearly as
much stainless steel in FCC Plants than there is now. As use of stainless became more
widespread, Type 304 was the usual choice. There have been rare cases of polythionic
acid cracking, but most of the reported cases have not been in the regenerator itself, but
in equipment external to the regenerator. There have been reports of polythionic cracks
in catalyst withdrawal lines, slide valves, flue gas lines, and expansion bellows. There is
justification for using Type 347 rather than 304 for these equipment items. Cracking of
Type 304 cyclones is rare enough that upgrading to Type 347 has not been considered
necessary. However, we have used Type 347 for cyclone support rods because of
concern about a cyclone dropping if a support rod were to crack.

Type 304 rather than 347 continues to be used for hexmesh and other types of
refractory anchors. We know of one case in which S-bar anchors in a regenerator did
crack, but this is the exception. It occurred during a water wash, and was not in one of
our plants.

Having looked at prevention of polythionic acid cracking through proper materials


selection, consider how cracking can be controlled by using certain operating and
maintenance procedures. To understand these procedures, we must first recognize that
the acids can be formed in several ways:

Regenerator flue gas contains SO2, which can be absorbed in condensing


moisture when the plant is cooled down.
Wash water, used in regenerators to render them dust-free during shutdowns,
also absorbs SO2 from the vapor in the vessel or from the gas that has
permeated the refractory lining.
Iron sulfide scale, created while the plant is operating, oxidizes when exposed to
air. In the presence of moisture, the polythionic acid solutions that cause
cracking are formed. The mechanism through which this result occurs has been
explained previously.

Iron sulfide scale is the usual corrosion product in the reactor system, because of the
presence of H2S. Regenerator scale has been reported by some investigators to be a
combination of sulfide and oxide, although we would normally expect it to be an oxide.
Therefore, media capable of causing cracking is generated in the reactor system by wet
oxidation of sulfide scale. The same problem can appear in the regenerator, but more
likely the acidity comes from water condensation and absorption of SO2 from the flue
gas.

Confidential
Copyright 2003 by ChevronTexaco
1-14 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

Cracking occurs only in the presence of moisture; but an ultra-thin film of adsorbed water
is all that is needed.

One common way to avoid polythionic cracking is to blanket the equipment with
nitrogen. Although many types of process plants employ this technique, it is not
practical in FCC Units. In any case, it does not protect against cracking when the acidity
is due to water and SO2 in the flue gas. It is effective only when the acidity comes from
wet oxidation of sulfide scale.

A second way to protect against cracking is to keep the equipment dry. Unless water is
deliberately introduced, regenerators tend to stay quite dry, perhaps explaining why we
dont see more polythionic cracking there. Also, reactors generally dont contain much
austenitic stainless; most of the alloy used there is chrome-moly steel or 12 chrome SS.
In some FCC regenerators, water washing is practiced to keep dust levels down during
shutdowns, making it easier and more comfortable to work in the vessel. This practice is
not extensively used at ChevronTexaco. Some companies have reported polythionic
cracks as a result of water washing. To prevent this problem, sometimes sodium
carbonate (soda ash) is added to the wash water to neutralize the acids. However,
concern has been expressed that soda ash soaking causes degradation of the refractory
lining, especially when the vessel is heated on startup. One company claims to have
shown by laboratory test that refractory damage by soda ash is not a problem, and we
think that this is probably true.

As far as FCC Units are concerned, we can summarize the polythionic acid cracking
situation as follows:

The major area of concern is the regenerator system.


Internal parts within the regenerator are typically Type 304. Although this alloy
sensitizes, polythionic cracking is not common. Use of 304 is justified by lengthy
experience.
Quite a few cases of polythionic cracking have been reported on stainless steel
expansion joints and slide valves, due to wet conditions resulting from a wet
steam purge. This problem has largely been corrected by using packed and
insulated expansion joints and by changing to internally insulated carbon steel
rather than stainless steel slide valves.
There is no point in considering Types 304L or 321 in the reactor/regenerator
system.
Regenerators are sometimes water washed for dust removal. This practice
increases the chance of polythionic cracking. Risk of damage can be reduced by
adding soda ash to the wash water.
We have used Type 347 cyclone hanger rods in regenerators to avoid polythionic
cracking, but the problem of sigma formation when using 347 probably makes
304 a better choice.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-15
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

1.3 Polythionic Acid Stress Corrosion Cracking

1) The most sensitization resistant 300 Series stainless steel is __________.

2) True or False: Water washing our regenerators during shutdowns will lessen the
possibility of polythionic acid SCC.

3) True or False: Recovery from sensitization is likely to occur more quickly at


regenerator operating temperatures than at reactor operating temperatures.

4) List some reasons for not using more 347 SS in our regenerators.

Confidential
Copyright 2003 by ChevronTexaco
1-16 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.4 CORROSION BY AMMONIUM BISULFIDE

Ammonium bisulfide (NH4HS) is the reaction product of ammonia (NH3) and hydrogen
sulfide (H2S). FCC feedstocks contain both organic sulfur and organic nitrogen
compounds. These compounds decompose in the reactor to form H2S and NH3,
respectively. These gases exit the reactor with the effluent process stream and enter
the main fractionator where they go overhead. In the fractionator overhead system, they
are absorbed in the condensing water, where they react to form ammonium bisulfide.

Figure 1.4-A shows the parts of an FCC Plant where NH4HS corrosion can occur. The
occurrence of bisulfide corrosion depends on the concentration of ammonium bisulfide in
the water phase and on fluid velocity. Experience in hydrocrackers tells us that bisulfide

Note: Watch dead legs and high velocity locations in particular.

Figure 1.4-A. Generic fractionator overhead


system showing areas subject to bisulfide
corrosion.

is not highly corrosive to carbon steel when the concentration is below 2% by weight. It
is rare for NH4HS concentrations to exceed this level in the main fractionator overhead
system. As a result, we do not often see the classical ammonium bisulfide corrosion
pattern, in which extremely rapid corrosion occurs in areas of high velocity (over 20 fps).
It can possibly occur in local areas of plants utilizing a very high nitrogen feed and can
be identified by the fact that it occurs only where velocity and turbulence are high.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-17
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Ammonium bisulfide corrosion is more commonly seen Ammonium bisulfide corrosion


in dead legs and other non-flow areas of FCC Plants. is more commonly seen in dead
Such areas need to be inspected regularly. legs and other non-flow areas of
FCC Plants. Such areas need to
Although dilute ammonium bisulfide does not rapidly be inspected regularly.
corrode steel, it does corrode copper alloys such as
admiralty and 70-30 cupro-nickel. This corrosion occurs at increased rates as pH rises.
The pH in the main fractionator overhead and in the
Plants which have experienced inter/aftercooler system depends largely on the
steadily rising nitrogen levels in amount of ammonia present, which in turn is a
the feed have had to dispense function of the nitrogen content of the plant feed.
with the use of admiralty, which Plants which have experienced steadily rising
in many cases had given good nitrogen levels in the feed have had to dispense with
service in earlier, low-nitrogen the use of admiralty, which in many cases had given
days. good service in earlier, low-nitrogen days.

Cyanides are also formed in the reactor in relatively small amounts. They are potent
promoters of NH4HS corrosion. Use of polysulfide in the plant will remove the cyanides
chemically and lessen corrosion. Cyanides are especially noteworthy because they
contribute to hydrogen blistering and wet H2S cracking. This issue is discussed in
another section of this document. The effect of cyanide becomes progressively more
pronounced as bisulfide concentration and pH rise. Therefore, in plants where nitrogen
content of the feed is high, we see more of the problems associated with bisulfide and
have more need to use polysulfide for control in plants.

Austenitic stainless steels resist bisulfide corrosion. The 12% chromium steels have
reasonably good resistance but may be damaged where cyanide concentrations are
high. The copper-based alloys are rapidly corroded by bisulfide but have some history
of past use as heat exchanger tubes in plants where ammonia concentrations were low
and the pH was near neutral. They are generally not used in plants with a high nitrogen
feed.

In summary, ammonium bisulfide is a common corrosive in the main fractionator


overhead system. It can cause local corrosion and severe fouling under some
conditions. However, in FCC Plants, it is not likely to cause rapid or catastrophic metal
loss. The biggest problem with NH4HS is that the alkaline sulfidic environment it creates
causes potentially dangerous hydrogen blistering and wet H2S cracking.

Confidential
Copyright 2003 by ChevronTexaco
1-18 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

HOMEWORK QUESTIONS

1.4 Corrosion by Ammonium Bisulfide

1) Ammonium bisulfide corrosion in FCC Plants is most commonly seen in __________


__________ and other __________ areas.

2) True or False: The threshold ammonium bisulfide concentration for corrosion of


copper alloys is 2% by weight.

3) True or False: If ammonium bisulfide concentrations in the fractionator overhead


exceed 2% by weight then expect rapid attack of carbon steel at high velocity
(>20 fps) locations.

4) The alkaline, sulfidic environment that ammonium bisulfide creates promotes


potentially dangerous __________ __________ and _____ _____ __________.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-19
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
1-20 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.5 CORROSION AND STRESS


CORROSION CRACKING BY AMINES

Aqueous amine solutions are commonly used to remove hydrogen sulfide and carbon
dioxide from process gases. Our FCC Units use several different amines:
monoethanolamine (MEA), diethanolamine (DEA), and methyl diethanolamine (MDEA).
MEA is the preferred solvent for CO2 removal and DEA is preferred for H2S. MDEA is
used for both. The varying amounts of H2S and CO2 found in sour process gases in
FCC Units determines which solvent is best used in the individual case.

Amine units consist of an absorber, where the acid gas is absorbed in the amine, and a
regenerator, where the acid gas is stripped out. The regenerator may be located in the
FCC Unit along with the absorber, or it may be remote from it. Frequently one
regenerator will handle the amine from absorbers in several different plants.

Amine systems are vulnerable to general corrosion, localized corrosion, amine stress
cracking, sulfide stress cracking, and wet H2S cracking. The latter three are the most
dangerous because they can cause catastrophic failure. For example, in 1984, an
amine absorber at Unovens Lemont, Illinois, Refinery failed suddenly from sulfide
cracking and resulted in 17 deaths as well as several hundred million dollars in damage.
The vessel was 8 1/2 feet in diameter and 53 feet long. The larger half of the vessel was
thrown 2/3 mile by the explosion.
Figure 1.5-A. Effect of temperature on
The corrosivity of amines is a corrosion rates in an amine system.
Corrosion Rate, mpy
function of temperature, amine
loading, velocity, amine
concentration, and
concentration of heat-stable
salts. Other things being equal,
MEA is the most corrosive of
the amines and MDEA is the
least corrosive. Therefore,
MDEA is used at higher
concentrations than other
amines. Currently
recommended concentration
ranges are 15-20% for MEA,
25-30% for DEA, and 45-55%
for MDEA. For corrosion
concerns, make sure that the
total amine strength is being
reported (free plus combined
amine).

In general, temperature is a
Temperature, F
more important factor in amine
corrosion than loading.
Hot, lean amine in the bottom of the regenerator is more corrosive than cooler rich
amine. The effect of temperature on corrosion is shown in Figure 1.5-A.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-21
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Amines remove H2S and CO2 through an acid-base In general, temperature is a


reaction. The amount of acid gas absorbed by the more important factor in amine
amine is measured by amine loading. A loading of corrosion than loading. Hot,
0.5 mols H2S per mol of amine means that one lean amine in the bottom of the
molecule of H2S has been absorbed by 2 molecules of regenerator is more corrosive
amine. Remember that amines are alkaline and H2S than cooler rich amine.
and CO2 are acid. If more acid gas is present than
can be neutralized by the amine, free acid remains in the system and can lead to
corrosion. We usually limit amine loadings to about 80% of theoretical to compensate
for less than perfect liquid-gas contacting, which could result in locally exceeding the
stoichiometric limit.

Exceeding the stoichiometric limit is most likely to cause corrosion in the lower part of
the absorber. More specifically, our experts recommend that we not exceed 0.40 mol/
mol for MEA, 0.50 for DEA, and 0.45 for MDEA.

Piping areas just Piping corrosion is velocity-related. Selective corrosion of


downstream of the welds is common. Liquid velocities in carbon steel amine
control valves cause piping are usually limited to 6 fps. This limit not only helps to
the greatest concern. reduce corrosion but also helps avoid iron sulfide corrosion
product from being eroded off the pipe wall in a finely divided
state. Accumulation of solid particles in the amine will promote foaming. Two piping
areas just downstream of the control valves in the rich amine line from absorber to
regenerator cause the greatest concern:

The level control valve on the absorber. A high pressure drop here will cause
H2S or CO2 to flash, causing severe turbulence and possibly severe local
corrosion if the piping is carbon steel.
The control valve in the rich amine line, just ahead of the regenerator. Corrosion
at this spot has such a lengthy history that it has been our practice for many
years to use stainless steel here.

If an FCC Plant must handle increased amounts of H2S or CO2, the options are to
increase amine concentration to a more corrosive level, increase circulation rates and
thereby raise velocities, or increase loading with a corresponding increase in corrosivity.
If the plant is already running at concentration, velocity, and loading limits, the least
objectionable option is usually to increase amine concentration. However, this decision
merits serious discussions with Materials and Process Engineers.

Heat-stable salts (HSS) are corrosive. They are formed by the reaction between amine
and acids or oxygen in the feed. No firm limit has been set on concentration. Some
companies set a limit of 5%, but this concentration may be too high, as evidenced by
corrosion of a lean DEA pump at Pascagoula in 1992, where the HSS concentration
varied from 2 to 5%. Current practice for ChevronTexaco and much of the industry is to
limit HSS to 10% of amine concentration. We cannot really expect that a true and
universal limit will ever be set. Of the many different heat-stable salts, such as amine
formate, acetate, and so on, the formate salt is the most corrosive. Some of the others
may not be corrosive at all. One adverse effect of HSS is that it chemically ties up some
of the amine, increasing true loading on the free amine that remains for reaction with
acid gas.

Confidential
Copyright 2003 by ChevronTexaco
1-22 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

As discussed above, control of fluid velocity, amine concentration, HSS concentration,


and amine loading are all useful corrosion control measures. Another way to prevent
amine corrosion is to use an austenitic (300 Series) stainless steel. Type 304L is the
grade most commonly used for this purpose. We do not use Type 304, except for items
not requiring welding or heat treatment, because of the possibility of polythionic acid
stress corrosion cracking during shutdowns. We learned this lesson the hard way, as a
result of massive piping failures many years ago at Richmond.

The previously mentioned limits on velocity, concentration, and loading do not apply to
stainless steels.

Stress corrosion cracking (SCC) is another major concern in amine service. MEA has a
long history of causing SCC. For many years we believed that DEA did not cause
cracking, and ChevronTexaco experience appeared to confirm this belief. However, in
1986, the National Association of Corrosion Engineers conducted an industry survey and
concluded that any amine could cause cracking. Of 137 plants surveyed, 22 (16%) had
experienced one or more instances of cracking. Later, they found that quite a few of
these cracks were due to sulfide stress cracking rather than to amine stress corrosion
cracking. However, the survey did show active cracking phenomena in all amine plants
and identified the need for appropriate control measures.

Amine SCC is caused by a combination of applied or residual stress and exposure to


amine solutions. The probability of cracking increases as temperature increases. The
cracks are intergranular and can be identified by metallurgical examination, which will
also determine whether a given crack is amine SCC or sulfide cracking. (See the
section Sulfide Stress Cracking, for a detailed discussion of this phenomenon.)

Because amine SCC occurs under applied or residual stress, it is commonly found in
weld areas. We can thus conclude that the residual stress from welding rather than the
much lower applied stresses caused by service conditions are largely responsible for the
cracks. Few appreciate the fact that residual stresses from welding are of yield point
magnitude. On relatively thin-walled equipment, we have seen cracks on the inside
surface where a clip or ladder support had been welded to the outside.

Amine SCC affects carbon steel and the low alloys but not the austenitic stainless steels.
The cure for amine SCC is thermal stress relief. Company practice is to stress relieve
almost all carbon steel columns, vessels, and piping in amine service, regardless of the
temperature of the specific amine being used. Any repairs should also be stress
relieved. The only exceptions to stress relief should be socket and seal welds in
DEA/MDEA operating below 140F, and fresh amine storage tanks.

Sulfide stress cracking can be a serious problem in amine plants. It is caused by the
presence of applied or residual stress, use of materials of high strength and hardness,
and exposure to H2S-containing media. Steels stronger than 90 ksi yield strength or
harder than Rockwell C22 are susceptible to cracking. Heat-affected zones of welds
may also be hard enough to crack. Sulfide stress cracking can be avoided by the
following:

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-23
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Postweld heat treatment.


Use of steels no stronger than Grade 70.
Maximum weld metal hardness of 200 Brinell.
Control of welding procedures.

Hydrogen blistering can also occur in amine service. Blisters may crack to the surface.
Cracks may run from blister to blister, causing hydrogen-induced cracking (HIC) or
stress-oriented hydrogen-induced cracking (SOHIC). These phenomena are discussed
separately in the section Wet H2S Cracking. Hydrogen blistering is a problem in
columns and vessels but not in piping, unless the piping is made from plate rolled and
welded. We do not know why it occurs in one plant but not another, or why it may be
worse in a given plant during one period of time. We can best prevent it only by using
coatings or cladding.

Inspection procedures for amine piping and equipment are outlined in Inspection
Strategies IS-1 (amine corrosion) and IS-10 (amine SCC). Piping is inspected by a
combination of UT for general corrosion and radiography for preferential corrosion of
welds. Columns and vessels require regular WFMT for detection of cracks that could be
caused by amine SCC, sulfide cracking, or wet H2S cracking.

Additional detailed information on corrosion and inspection issues in amine plants can
be found in the Inspector Training Manual on Amine / SPU / Tail Gas Plants issued in
March 1999.

Confidential
Copyright 2003 by ChevronTexaco
1-24 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

HOMEWORK QUESTIONS

1.5 Corrosion and Stress Corrosion Cracking by Amines

1) Match the amine in the middle with its concentration limits at left and its loading limits
at right:

25-30% MDEA 0.45 mol/mol


15-20% DEA 0.4 mol/mol
45-55% MEA 0.5 mol/mol

2) Which amine is most aggressive at promoting amine SCC?

3) Which cracking mechanisms are possibilities in amine plants?

Sulfide stress cracking


Amine cracking
Polythionic acid
Carbonate cracking
Wet H2S cracking

4) To minimize corrosion of carbon steel, limit amine velocities to _____ fps and limit
the concentration of heat stable salts to _____.

5) True or False: Hot lean amine is generally more corrosive than cool rich amine.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-25
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
1-26 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.6 CATALYST EROSION

Erosion is defined as the loss of metal due to the impact and cutting action of solid
particles. Unlike corrosion which is the loss of metal from chemical reaction, erosion
does not involve any reaction of the metal or refractory with the environment. It is a
mechanical phenomenon. The rate of erosion is influenced by many factors:

Properties of the material being eroded.


Properties of the solid doing the eroding.
Particle size.
Particle velocity.
Particle concentration.
Angle of incidence.

In an FCC Unit, particle size is not a variable that we can control. FCC catalyst is
typically 150-200 mesh as charged to the plant. Some degradation occurs in service,
and the catalyst fines that exit the reactor and regenerator with process gas and flue gas
are smaller. Larger particles are more erosive than smaller ones because they strike
metal or refractory surfaces with more energy.

A harder catalyst ought to cause more erosion than a soft one. However, as Process
Engineers, not Materials Engineers and Analysts, pick the catalyst, we cant control its
properties. And despite our concerns in the past, we have never had any solid evidence
of a problem in this area. Even so, we continue to watch for one.

Particle velocity is also an important consideration. ChevronTexacos internal


information suggests that erosion by FCC catalyst is not serious below 80-100 fps.
Some other oil companies are inclined to favor a threshold closer to 50 fps. In either
case, there is a threshold below which little or no erosion occurs. Once this threshold is
exceeded, metal loss increases as the third or fourth power of the velocity.

Particle concentration is obviously an important parameter. The more particles that


strike the surface in a given period of time, the more metal will be removed.

The angle at which the particle strikes the surface clearly has an important effect. We
know instinctively that particles traveling parallel to the surface do not cause much metal
loss. We also expect that particles striking the surface at a 90-degree angle would
cause the worst erosion, which is true for brittle materials. However, for the materials we
use for most equipment (carbon steel, Type 410 and Type 304), damage occurs fastest
at about a 45-degree angle of incidence.

Erosion rates are clearly going to be Although lab tests may show some
affected by our choice of construction differences, in actual practice there is little
material. Although lab tests may show difference in erosion resistance between
some differences, in actual practice there is carbon steel, 1-1/4 Cr - 1/2 Mo, Type 410,
little difference in erosion resistance and the 300 Series stainless steels.
between carbon steel, 1-1/4 Cr - 1/2 Mo,
Type 410, and the 300 Series stainless steels. Harder materials may or may not be

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-27
permission of ERTC
Fluid Catalytic Cracking Process August 2003

better. Figures 1.6-A and 1.6-B show the results of laboratory erosion tests conducted
at ambient temperature and 1000F.

120 Ft/Sec. Ambient Temperature


Erosion Rate, In./Yr.
16 Figure 1.6-A. Relative erosion
Metals
14 rates in FCC catalyst (120 ft/sec.
ambient temperature).
12
10
8
Hard Facings
6
4 Ceramics
2
0
1-1/4 Cr - 304 SS Carbon Stellite 6 Stoody TMT-55 LC1C Silicon Aluminum Silicon
1/2 Mo Steel 2134 Carbide Oxide Nitride

120 Ft/Sec. 1000F


Erosion Rate, In./Yr.
16 Metals
14 Figure 1.6-B. Relative erosion
rates in hot FCC catalyst (120 ft/
12 sec. ambient temperature).
10
8
6
Hard Facings
4
2
0
Carbon Steel 1-1/4 Cr - 1/2 Mo 304 SS Stoody 2134 TMT-55 Stellite 6

Neither heat treatment nor cold work increase erosion resistance significantly. Hard
facing materials are often used. Some perform very well, but not all. Either lab tests or
field experience should guide the selection of a good hard facing alloy. Materials like
Stellite 6 have a good history in FCC Units. The Corrosion Prevention Manual section
on FCC Units gives extensive data on erosion resistance and can be used as a guide to
materials selection.

Confidential
Copyright 2003 by ChevronTexaco
1-28 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

Many hard facing alloys consist of hard particles (tungsten carbide, etc.) in a soft matrix.
There is danger here that the very fine catalyst will erode the matrix material between
the tungsten carbide particles, causing them eventually to fall out. And clearly geometry
plays a role. We are likely to see more rapid damage if the spacing between tungsten
carbide particles is great enough to allow catalyst to impinge the matrix material in
between them.

It is thought, although not well proven, that metals and alloys erode faster at
temperatures that approach their oxidation or sulfidation limits. Oxidation or sulfidation
rates tend to decrease with time because a protective film forms on the metal. The film
acts as a diffusion barrier which stifles attack as it gets thicker. Erosion may remove the
protective film as fast as it forms, allowing the metal to corrode rapidly. Note that in this
case, the metal is corroding, not eroding. It is the protective film that is eroding.

Erosion is a potential problem in the reactor and regenerator vessels, their cyclones,
catalyst and flue gas piping, and in the piping, pumps, and exchangers in the main
fractionator bottoms system. Hard facings provide protection as needed in the main
fractionator bottoms. They are also used to protect selected parts of the catalyst slide
valves. In the other locations listed, protection is usually provided by erosion-resistant
refractories. See sections of the manual on refractories and those on specific equipment
items for a more detailed discussion of where erosion protection is being provided.

Experience tells us that the worst erosion is


Experience tells us that the worst erosion is
likely to occur where we have a small crack
likely to occur where we have a small crack
or gap in a reactor or regenerator internal
or gap in a reactor or regenerator internal
member through which catalyst can be
member through which catalyst can be
propelled at high velocity by a pressure
propelled at high velocity by a pressure
differential. This type of situation has
differential.
caused many serious erosion incidents,
including reactor shell leaks.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-29
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

1.6 Catalyst Erosion

1) A conservative threshold velocity for erosion by FCC catalyst is ______ fps.

2) Which ChevronTexaco Gray Manual contains extensive data on the erosion-resisting


properties of commercially available hard facings? What hard facings are used in
your FCC Plant?

3) True or False: Actual service experience shows a big difference in erosion


resistance between carbon steel, 1-1/4 Cr steel, and 304H SS.

4) List five FCC components that routinely experience at least some erosion damage
during a normal run cycle.

Confidential
Copyright 2003 by ChevronTexaco
1-30 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.7 CARBONATE STRESS CORROSION CRACKING

It has been recognized for many years that alkaline carbonate solutions can cause
stress corrosion cracking. However, it was not until 1983 that this problem was identified
in oil refineries, specifically in FCC Units. Its surprising appearance after many years of
FCC operation was the direct result of changes in plant operating conditions, changes
that altered the chemical composition of the water condensates from and beyond the
main fractionator overhead system.

Overhead vapors from the main fractionator contain ammonia, hydrogen sulfide, carbon
dioxide, cyanides, and water. When the overhead water condenses, it absorbs the other
constituents. The ammonia reacts with H2S to form ammonium bisulfide and with CO2 to
form either ammonium bicarbonate or ammonium carbonate. Whether we form
bicarbonate or carbonate depends on the amount of CO2 present and on the amount of
ammonia. Large amounts of ammonia lead to a more alkaline pH and the formation of
carbonate in preference to bicarbonate. Specifically, both carbonate and a high pH must
be present to generate the conditions necessary for carbonate SCC.

Carbonate cracking occurs only on carbon steel, not on other alloys. Cracking requires
the presence of an applied or residual stress. For this reason, cracks are usually found
in or adjacent to welds. Most often they are in the base metal and run parallel to the
weld, but occasionally they are transverse to the weld. Cracks are intergranular and
oxide-filled. Carbonate cracking can occur at lower stress levels than other SCC
mechanisms. For this reason, prevention of cracking at welds requires special stress
relief procedures. Also, care must be taken to ensure proper support of piping. Cracks
have occurred outside weld areas due to stresses resulting from missing pipe supports.
Figure 1.7-A shows the piping and equipment in the fractionator overhead system that
can be exposed to conditions under which carbonate SCC may occur.

Chemical analysis of overhead waters from FCC main fractionator systems in


ChevronTexaco refineries has shown that they contain the following chemicals:

400 to 40,000 ppm NH3


100 to 16,000 ppm H2S
0 to 25,000 ppm CO2
pH 7.3 to 10.1

Generally, cracking occurs when the pH is greater than 9. However, the critical pH
depends on carbonate concentration. This relationship is shown in Figure 1.7-B. We
have seen cracking occur at carbonate concentrations as low as 100 ppm.

Regular sampling of plant process waters can identify the Regular sampling of
conditions under which carbonate cracking might occur. One plant process waters can
important factor is the nitrogen content of the plant feed. An identify the conditions
increase in feed nitrogen leads to more ammonia and a under which carbonate
higher pH in the fractionator overhead. A pH above 9 can cracking might occur.
lead to SCC.

Conduct regular inspections for detection of carbonate cracks in equipment pictured in


Figure 1.7-A. Inspection procedures are outlined in Inspection Strategy IS-12. Because
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-31
permission of ERTC
Fluid Catalytic Cracking Process August 2003

the threshold stresses for carbonate cracking are low, some refineries continue to
inspect at least some welds even if they have been PWHTd. Richmonds carbonate
cracking inspection program is included in the back of this manual as a reference.

Figure 1.7-A. Generic


fractionator overhead
system showing
areas subject to
carbonate cracking.

Figure 1.7-B.
Carbonate SCC; pH
versus carbonate ion.

Confidential
Copyright 2003 by ChevronTexaco
1-32 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

Since SCC requires the presence of stress, thermal stress relief of piping and pressure
vessels is an effective preventive measure. Where potential SCC conditions exist,
PWHT of both new and repaired equipment is important. Careful temperature control
during PWHT is essential; if temperatures are too low, stresses will not be completely
relieved and cracking may still occur.

Because each plant is different in terms of process chemistry, not all refineries
encounter Carbonate SCC. However, we have seen cracking in several ChevronTexaco
FCC Plants. This problem is a matter of serious concern.

HOMEWORK QUESTIONS

1.7 Carbonate Stress Corrosion Cracking

1) True or False: Carbonate cracking is never found outside weld areas.

2) In which situation is carbonate cracking more likely to occur?

a) pH of 8.6 and carbonate at 1000 ppm or


b) pH of 8.8 and carbonate at 100 ppm

3) True or False: ChevronTexaco requires carbon steel in carbonate services to be


stress relieved at higher temperatures than required by Code.

4) Why does an increase in the nitrogen content of the plant feed increase the
likelihood of carbonate SCC?

5) True or False: Carbonate cracks in piping can be easily detected and sized by shear
wave UT.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-33
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
1-34 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.8 CHLORIDE STRESS CORROSION CRACKING

The austenitic (300 Series) stainless steels resist corrosion because they contain
chromium, which forms a protective film on the metal. In air, this film is chromium oxide.
In most refinery environments it is chromium sulfide. The film is insoluble in most
environments, some strong acids excepted.

Certain chemical elements, especially the chloride ion, disrupt this protective film
possibly because metal chlorides are soluble. Film disruption is usually local and occurs
most readily under the influence of residual or applied stress. Common sources of
residual stress are welding and cold work from forming operations such as tube bending.
When the film ruptures, the chloride ion prevents it from healing and rapid attack occurs
because the metal, now unprotected, acts as an anode. The remaining metal that still
contains the oxide or sulfide film acts as a cathode to form an extremely active galvanic
cell. The result is formation of a crack, which has been generated chemically rather than
mechanically.

Chloride cracking occurs primarily at slightly elevated temperatures. On sensitized


stainless steels, it can occur at ambient temperature. Chloride cracking also requires
the presence of water and cannot occur at temperatures above the water boiling point.

Almost any chloride can cause chloride stress corrosion cracking. In FCC Units,
opportunities for chloride cracking are few, because the biggest use of stainless steel is
in high temperature service, and because not many chlorides are likely to be found in the
plant. We do find ammonium chloride deposits in the top section of main fractionators
and in the fractionator overhead condensers. The main fractionator itself is not a
problem because the top section is always built of carbon steel. (Richmond has seen
substantial corrosion of carbon steel underneath packing, due in large part to chlorides.)
However, any stainless tubes used in the fractionator overhead condensers could crack.

Injection of water ahead of stainless-tubed At any water injection point, the


condensers can also cause SCC if the temperature at quantity of injection water must
the point of injection is above the water dew point be sufficient to allow at least 25%
and if the amount of water injected is insufficient to of it to remain in the liquid phase
avoid a boil-dry condition on the piping or tube after the system reaches
surface. We have seen chloride cracking of stainless temperature equilibrium.
tubes caused by water injection at the compressor
discharge. At any water injection point, the quantity of injection water must be sufficient
to allow at least 25% of it to remain in the liquid phase after the system reaches
temperature equilibrium.

Stainless heat exchanger tubes can chloride crack in cooling water service. Although
Company experience shows that stainless tubes have usually performed well in cooling
water service, there have been occasional failures. These can be avoided in two ways:

Dont use stainless tubes in cooling water service if past experience in your
refinery has not been successful or if chloride content of the water is high. (In a
refinery lacking experience, check with ERTC if chlorides exceed 50 ppm.)
Dont use stainless tubes if there is any problem of water scale formation in the
tubes. Chlorides will concentrate under the scale and crack the tubes.
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-35
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Generally, for stainless tubes in cooling water service, be sure to have high water
velocities (>5 fps) and low metal skin temperatures (<140F). Stainless exchanger tubes
have cracked following attempts to control process temperatures by throttling cooling
water flow.

Steam connections to stainless equipment can be a source of trouble because of


entrained water which may contain chlorides. Plenums, for example, can suffer chloride
cracking if they do not crack first from thermal fatigue. Expansion joints in hot piping
may also crack if steam purged. This is generally attributed to poor steam quality so that
liquid water comes in with the steam. The same thing can happen in steam-purged
stainless steel slide valves, though these are no longer being used.

Chloride cracking can be prevented in several ways:

Use ferritic, duplex, copper based or high-nickel alloys instead of stainless.


Postweld heat treat. (May not be practical in the kind of FCC equipment in which
cracking is likely to occur.)
Control water entrainment in steam to stainless equipment.
When injecting water into the system, be sure you add enough.
Check refinery experience or get ERTC advice before using stainless tubes in
cooling water service.

Confidential
Copyright 2003 by ChevronTexaco
1-36 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

HOMEWORK QUESTIONS

1.8 Chloride Stress Corrosion Cracking

1) True or False: The primary concern area for chloride SCC in FCC Plants is carbon
steel components at the top of the main fractionator.

2) To avoid chloride SCC of 304L exchanger tubes from cooling water, keep water
velocities above _____ fps and tube skin temperatures below _____ F.

3) True or False: Stainless steel wont chloride crack below 140F, even if it is
sensitized.

4) Avoiding boil-dry conditions on SS tubes is key to preventing chloride SCC. At any


injection point, be sure to inject enough water to allow at least _____ % of it to
remain as liquid after the system reaches temperature equilibrium.

5) Will the following design/operating changes increase or decrease the likelihood of


CW-induced chloride cracking in a SS tubed exchanger?

Operations decides to throttle CW flow to control process temperature


Design Engineering decides to put CW on the shellside instead of the tubeside

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-37
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
1-38 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.9 CORROSION AT INJECTION POINTS

Several years ago, Shell Oil experienced an explosion and huge fire in an FCC Plant at
their Norco, Louisiana, Refinery. This catastrophe was caused by a corrosion leak in a
piping ell, just downstream from an aqueous ammonia injection point. It focused API
and NACE attention on the subject of corrosion at injection points. One result is that
API 570 now requires that injection points of all types be inspected at intervals not to
exceed 3 years.

The purpose of this discussion is to look at why injection points corrode and which ones
pose the greatest danger. We define an injection point as a location at which some
outside material such as water, inhibitor, or neutralizer is added to the process stream.
The junction between two process streams is not considered an injection point in the
present context, except under unusual circumstances. For example, a sour water
stream from an FCC distillation system that is withdrawn, recirculated, and reinjected
does have an injection point, even though we could consider it a process stream. The
analyst must use good judgment in deciding what is an injection point and what is not.
When in doubt, inspect the location as if it were.

There are a number of reasons why an injection point may corrode:

The material being injected is corrosive.


A corrosive material may form by combination or reaction of the process stream
and injection fluid.
Turbulence may be created.
Temperatures may change (especially when water is injected into a process
stream which is higher in temperature than the water dew point).
The injected material may impinge on the pipe wall.

The first reported cases involved inhibitor injection lines in crude unit overhead piping.
The inhibitor was a relatively thick, viscous material, corrosive in its concentrated form.
It was being added in small quantities, ran as a thin stream along the vapor line, and
eventually corroded a groove in the line. We have not seen this problem in many years,
because current practice is to dilute the inhibitor with solvent before injecting it.

The Shell Norco explosion shows why injection points may corrode. Here, aqua
ammonia was being injected into a wet, sour line to neutralize acidic constituents. The
injection quill was blunt-ended, so that the injected fluid probably exited as a solid jet.
Furthermore, the injection point was located only a short distance upstream from an
elbow, and the flow direction was vertically downward. The result was a solid stream of
liquid striking the elbow, causing an erosion-corrosion failure. One unique feature of that
failure was that the corrosion formed a ring which caused an oval disc of metal to blow
out. The center of this disc, and the center of the liquid impingement area, showed very
little corrosion by comparison. Following the explosion, many experts speculated and
wrote technical articles about what caused the peculiar corrosion pattern. However, we
see the design and location of the injection nozzle as the primary defect causing the
failure. A spray nozzle, instead of a blunt-ended one, should have been located further
upstream of the elbow. For insight into what constitutes a good design, ask ERTC for a
copy of the water injection system being recommended for hydrocrackers.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-39
permission of ERTC
Fluid Catalytic Cracking Process August 2003

These are the basic guidelines:

Inject liquid through a spray nozzle.


Have an adequate quantity of injection water.
Place the injection well ahead of bends.

Hydrocrackers have experienced several corrosion FCC Plants may use multi-stage
failures at water injection points. Broadly, we blame water injection in which the same
the corrosive ammonium bisulfide for the damage. water is withdrawn from one
Without the proper injection system design, corrosion location and injected in another.
failure could happen in the main fractionator Here, the injected fluid itself is
overhead system of an FCC Unit, because of the corrosive. We must guard against
presence of ammonium bisulfide. FCC Plants may direct impingement against the
use multi-stage water injection in which the same wall of the piping.
water is withdrawn from one location and injected in
another. Here, the injected fluid itself is corrosive. We must guard against direct
impingement against the wall of the piping.

API standards set minimum inspection intervals for injection points. Some locations will
require more frequent checks. We must understand corrosion in our plants to identify
these areas.

HOMEWORK QUESTIONS

1.9 Chloride Stress Corrosion Cracking

1) True or False: API 570 requires that special attention be given to injection points.

2) True or False: One of the tenets of good injection point design is to place the
injection point well ahead of bends.

3) True or False: Use of a blunt-ended injection quill instead of a spray nozzle probably
contributed to the Shell Norco failure.

4) Because of the presence of __________ __________, corrosion can occur at water


injection points in the fractionator overhead system whenever the injected water
impinges on the pipe wall.

Confidential
Copyright 2003 by ChevronTexaco
1-40 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.10 STRESS CORROSION CRACKING BY


CAUSTIC (SODIUM HYDROXIDE)

Sodium hydroxide (NaOH, also called caustic) and other highly alkaline hydroxides
such as potassium hydroxide (KOH) can cause stress corrosion cracking of carbon
steel, low alloys, and stainless steels. There is little difference in crack resistance
between these three classes of materials, except that the higher strength alloys are
more likely to crack.

Corrosion cracking requires stress (either residual or applied), the presence of caustic,
and an elevated temperature. Very strong caustic solutions (50%, for example) will
crack steel at temperatures above 110F. Somewhat lower concentrations such as 20%
require at least 140F to crack. For very dilute solutions, the temperature limit for
cracking is quite high and difficult to establish exactly. In looking at the effect of caustic
concentration, consider the possibility that contact with a hot surface will evaporate
water and increase the concentration in the solution that remains. Caustic is non-volatile
and will not evaporate with the water.

Caustic is used in FCC Units for the treating/purification of Traced caustic lines
certain product steams, most frequently the LPG components. should be stress relieved.
This treatment, typically done at ambient temperature, poses
no special risk of caustic SCC. The only areas of concern are the heat-traced caustic
lines. Welds may crack if the tracing causes metal surface temperatures in excess of
140F. Therefore, traced caustic lines should be stress relieved.

Further caustic SCC problems arise in FCC Plants because traces of caustic may be
present in the feedwater to steam generators. Even small amounts of caustic in the
BFW can concentrate to dangerous levels where water is evaporating and leaving the
dissolved solids behind. In process steam generators, this condition can occur where
tubes are not completely submerged, where steam blanketing occurs, or where heat flux
is extremely high.

Caustic cracking in steam generators can be controlled in two ways, and both need to be
utilized:

The first approach is to design the process steam generator properly.


Generators are usually of U-tube design, and the tubes, therefore, need to be
stress relieved to reduce the probability of cracking. Because it is impossible to
achieve totally effective stress relief on a heat exchanger bundle, there is no
guarantee that cracking will be prevented. Designing the steam generator to
make sure that the tubes are completely submerged is important.
The second approach is to control feed water treatment. Some water treatments
require a high level of alkalinity, forming sodium hydroxide on water evaporation.
However, other water treatments insure that caustic is not formed, even when the
water is evaporated to dryness. Our water-treating consultants must be
consulted on this subject. Any contemplated change in BFW treatment must be
evaluated with respect to the probability of its causing stress corrosion cracking.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-41
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

1.10 Stress Corrosion Cracking by Caustic (Sodium Hydroxide)

1) The threshold temperature for caustic SCC in high concentration caustic is _____ F.

2) SCC is definitely a concern at non-stress relieved welds on heat-traced caustic lines.


Should we also be concerned about corrosion under the heat tracing?

3) Designing steam generators to make sure the tubes are completely __________ is
important for preventing caustic SCC.

Confidential
Copyright 2003 by ChevronTexaco
1-42 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.11 WET H2S CRACKING

The term wet H2S cracking refers to a collection of phenomena capable of causing
cracks in carbon steel. The term originated when an oil company initiated a survey of
pressure vessels in wet H2S service and found that about half of them contained cracks.
Causes of cracking were generally not investigated. Eventually, it became apparent in
the industry that the cause of these cracks was not a single phenomenon, but a
combination of phenomena. Included under the umbrella of wet H2S cracking are the
following:

Sulfide stress cracking.


Hydrogen blistering.
Hydrogen-induced cracking (HIC).
Stress-oriented hydrogen-induced cracking (SOHIC).
Previously undetected fabrication defects.

The section, Sulfide Stress Cracking, discusses sulfide stress cracking. In carbon steel
pressure vessels, it is generally the result of using the wrong welding electrode or
welding procedure. Postweld heat treatment will control sulfide stress cracking, but not
the other forms of wet H2S cracking.

Hydrogen blistering occurs as a result of aqueous corrosion in the presence of H2S.


Hydrogen atoms form in the corrosion reaction. Some of these atoms combine to form
the hydrogen molecule (H2). Others diffuse into the steel. When the atoms reach
internal voids or discontinuities, they combine to form hydrogen molecules (H2).
Although very small hydrogen atoms can diffuse through solid steel, larger molecules
cannot, and they are trapped at the internal cavities in the steel. In time, the
accumulation of hydrogen causes extremely high pressures (many thousands of pounds
per square inch) to develop in these voids and discontinuities. The cavities enlarge and
distort, and perhaps cause a bulge (hydrogen blister) to appear on the surface.
Eventually the blister may crack, either across the center or around the edge.

Hydrogen-induced cracking (HIC) is an advanced stage of hydrogen blistering. It is also


called stepwise cracking or hydrogen blister cracking.

HIC occurs when pressures within multiple overlapping hydrogen blisters cause cracks
to grow from one blister to another. Visual examination of a HIC failure shows that the
cracks grow in steps (blister-to-blister).

The term hydrogen-induced cracking has a dual meaning. In todays oil industry
technology, it describes hydrogen blister cracking as discussed above. In the past,
hydrogen-induced cracking was a general term that defined all mechanisms by which
hydrogen could cause cracks in steelsulfide stress cracking, HIC, SOHIC, and delayed
fracture. For purposes of this discussion, refer to the newer definition.

SOHIC is fundamentally the same as HIC, except that the direction of crack growth is
influenced by residual or applied stress. It is most often found in weld areas, where
residual welding stresses affect the crack orientation.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-43
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Inspection Strategy IS-17 discusses inspection methods to find wet H2S cracks. Wet
fluorescent magnetic particle inspection (WFMT) effectively finds all forms of cracking
named in this section. The extensive use of WFMT inspection created two problems:

1) It revealed fabrication defects that were often harmless, but still had to be
dealt with; and
2) The surface preparation for WFMT leaves the metal more exposed to future
wet H2S cracking.

The current approach for equipment in wet H2S service is to coat vessels in the more
severe wet H2S services. They have been inspected. This not only prevents cracking,
but it makes future inspection much easier; simply check that the coating is OK. For
new equipment, the ideal solution is stainless cladding or possibly Duplex SS, if the cost
is not prohibitive.

Figure 1.11-A shows locations in fractionator overhead system where wet H2S cracking
can occur.

Figure 1.11-B shows the various forms of wet H2S cracking discussed in this section.

Note: Sulfide cracking, a form of wet H2S cracking, can occur anywhere in the
OVHD and light ends recovery system where high strength, high hardness
steels are used (e.g., bolting, valve stems and compressor internals). Limit
hardness or use resistant alloys.
Figure 1.11-A. Generic Fractionator:
Overhead system showing areas subject to
wet H2S cracking.

Confidential
Copyright 2003 by ChevronTexaco
1-44 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

Figure 1.11- B. Different


forms of wet H2S cracking.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-45
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

1.11 Wet H2S Cracking

1) The industry-accepted threshold concentration for wet H2S cracking is _____ ppm of
H2S in water.

2) Which of the below are valid means to minimize and/or prevent wet H2S cracking
and blistering in FCC Plants?

Use stainless cladding


Use high strength steels
Use steels containing high sulfur
Use epoxy coatings
PWHT
Inject ammonium polysulfide
Use clean, HIC-resistant steels
Inject corrosion inhibitors
Increase the nitrogen content of the plant feed

3) True or False: The main fractionator is more likely to experience wet H2S cracking
than the interstage separator.

4) True or False: Seamless pipe and forgings rarely experience hydrogen-induced


cracking.

Confidential
Copyright 2003 by ChevronTexaco
1-46 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.12 CORROSION BY NAPHTHENIC ACIDS

Naphthenic acids are naturally occurring organic acids found in crude oil. In the crude
distillation unit, the acids partition themselves between various cuts in accordance with
their boiling points. The amount of naphthenic acid in a given cut is measured by
titrating an oil sample with potassium hydroxide. The result is expressed as the number
of milligrams of KOH required to neutralize the acid in one gram of oil. This number is
called either neutralization number or total acid number (TAN).

All crudes contain some quantity of naphthenic acid. In Middle Eastern crudes the
amount is low. Certain California and Louisiana crudes, as well as some crudes from
South America and West Africa, contain enough acid to cause corrosion in the crude
unit. If corrosion occurs in the crude unit, we can also expect corrosion in downstream
plants that handle the more acidic cuts from the crude. We usually find that the most
acidic and most corrosive cut from the crude unit is the heavy gas oil. Unfortunately for
FCC people, this cut is a common FCC feed.

Naphthenic acids are quickly destroyed by the very The primary impact of the
high temperatures in the reactor. As a result, acids will be found in piping
naphthenic acid corrosion is not a problem in the that brings hot gas oil from the
reactor itself or in downstream equipment. The primary crude unit to the FCC and in
impact of the acids will be found in piping that brings piping and equipment in the
hot gas oil from the crude unit to the FCC and in piping reactor feed system.
and equipment in the reactor feed system. Note,
however, that because hydrotreating destroys naphthenic acids, we do not see corrosion
in the FCC reactor feed system when the plant processes a hydrotreated feed.
Currently, all ChevronTexaco FCC Units where naphthenic crudes are used are running
hydrotreated feed.

Figure 1.12-A
plots carbon
steel corrosion
rate as a
function of
temperature
and Total Acid
Number.
Consider it an
approximation Figure 1.12-A.
appropriate for Naphthenic acid
piping and heat corrosion of
carbon steel.
exchangers.
Pumps
experience
much higher
rates because
of extremely
high fluid
velocities.
Naphthenic acid
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-47
permission of ERTC
Fluid Catalytic Cracking Process August 2003

corrosion doesnt usually occur at TAN values below 1.5 mg/g, but this limit may be too
high for high-velocity areas such as pumps.

The appearance of naphthenic acid corrosion varies widely. Where velocities are
nominal as in liquid lines, attack is usually uniform, although we may see some tendency
for higher rates at bends. The metal will commonly be covered with a mushy black
scale. Where velocities are high, as in pumps or mixed phase lines, the metal surface
may be wavy or scalloped with high-low areas, and the surface may be clean and bright.

The answer to naphthenic acid corrosion is to upgrade to Type 316L or (better yet)
Type 317L stainless steel. This is a molybdenum-bearing grade of stainless. Grades
not containing moly, such as Types 304, 321, or 347, may or may not corrode. They
tend to be unreliable in their ability to resist naphthenic acid. The straight chromium
steels, such as 5 Cr and 12 Cr, have no better resistance than carbon steel.

We can summarize the entire naphthenic acid picture for FCC Units as follows:

Corrosion occurs only when the TAN of the plant feed is over about 1.5 mg/g.
If it does occur, it will be limited to the reactor feed system.
It will not occur below 350-400F.
Corrosion can be prevented by upgrading to Type 316L, or in the most severe
cases, Type 317L.
Corrosion rates can be relatively high, but with the feeds we have processed so
far, not catastrophic. Inspectors can stay ahead of the corrosion if alerted by the
analyst to an increase in acidity of the plant feed.

As the last statement suggests, if there is any chance that the plant might be processing
non-hydrotreated naphthenic feed, the analyst must monitor the TAN of the plant feed on
a regular basis.

Inspection for naphthenic acid corrosion is described in Inspection Strategy IS-8.

HOMEWORK QUESTIONS

1.12 Corrosion by Naphthenic Acids

1) Except in pumps, naphthenic acid corrosion usually doesnt occur at TAN values less
than _____ mg/g.

2) True or False: Naphthenic acid corrosion is not a concern downstream of the FCC
reactor.

3) What element do we add to stainless steels for naphthenic acid corrosion


resistance?

4) The threshold temperature for naphthenic acid corrosion is _______ F.

Confidential
Copyright 2003 by ChevronTexaco
1-48 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.13 INTERPRETATION OF CORROSION DATA

You can obtain useful corrosion data from handbooks, laboratory tests, plant
inspections, use of corrosion probes or coupons in the operating plant, or information
obtained in other plants. Although this information can be useful, serious errors can
occur if you take the data at face value without giving it careful thought.

Many textbooks or handbooks contain tables of data that indicate the probable
performance of metals and alloys in specific corrosives as a function of temperature and
concentration. Consider this information as a first approximation only. If a material is
shown as acceptable, consider it a candidate for further evaluation. If a material is
shown as unacceptable, do not study it further. Data may be considered unacceptable if
it doesnt consider all the factors that influence corrosion rates. As an example, we built
a chemical plant at Oak Point, Louisiana, a number of years ago. One of the primary
corrosives in the plant was hydrochloric acid. According to the textbooks, Hastelloy B
was an excellent material for HCl service, and pilot plant tests in the research labs
confirmed this. However, when the plant started up, the Hastelloy B equipment was
seriously damaged in a matter of a few months. The problem was that one of the
purchased feedstocks contained a small amount of dissolved ironan oxidizing agent
and a potent corrosion promoter for Hastelloy B. Note that even the pilot plant test
program did not predict the problem, although it might have done so if the tests had
been run using actual purchased samples of reagents from the prospective suppliers.

Laboratory corrosion tests can be as misleading as textbook data. From the example
cited above, it is apparent that both data sources gave the wrong answer. Laboratory
tests are run under fixed and controlled conditions. It is up to the tester to determine if
the test program has considered all important variables. If the test had been run 10F
hotter, would the material have corroded? Was a sensitivity to high velocity corrosion
considered? How about the effects of other chemical constituents or impurities? Do we
get an artificially low corrosion rate because all the corrosives were used up during the
test? These are all considerations for a lab test program. Such tests are best left to
experts. Even orientation of the test specimen can affect corrosion rates. For example,
a vertically oriented test specimen may corrode faster that a horizontal one, because
hydrogen bubbles sweeping upward past the surface set up a convection current that
continues to bring fresh corrosives to the surface.

Exposing corrosion coupons in process plants can provide valuable data, but be careful
in picking coupon location. Consider the example of a crude unit atmospheric overhead
system. Chemical compositions in the overhead system are constantly changing. As
the process stream cools, more water condenses, and the corrosives (H2S, NH3,
neutralizing amine) are absorbed in continuously changing concentrations. The
temperature changes and velocity varies from point to point. Do we want to expose the
coupon in a relatively quiet area, or in an area where the high velocity process fluid
directly impinges on it? And remember that we are dealing with a three-phase fluid
(liquid hydrocarbon, liquid water, and vapor). Which phase is the coupon seeing?

Orientation of the coupon is also important. If inserted transversely into a line, the
highest velocity is in the center of the line and the lowest near the edge. Corrosion rates
are different if the flow is parallel to the coupon surface rather than perpendicular to it.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-49
permission of ERTC
Fluid Catalytic Cracking Process August 2003

This is not to say that corrosion coupons dont provide valuable data. A series of
coupons exposed sequentially in a given location can tell us if operating conditions are
altering the corrosion picture. Coupons of multiple materials exposed in a
hydroprocessing reactor can give valuable insight into the severity of corrosion and the
performance of other alloys. In high-temperature, non-aqueous systems such as this, it
is easy to obtain reliable data because not much except temperature and corrosives (in
this case H2S) concentration affect rates.

Always examine corrosion coupons visually after removing them. If corrosion is uniform,
you can take the measured rates at face value. Look for pitting, which results in a
shorter equipment life than the calculated rate predicts. If a coupon is knife-edged, a
potential velocity problem exists that could cause accelerated attack in turbulent areas.

Corrosion (Corrosometer) probes have been used for about 40 years to rapidly
determine corrosion rates in process plants. They have the same potential as coupons
for providing misleading data. They also have some peculiarities. Corrosion probes
measure corrosion rates electrically by measuring the electrical resistance of a carbon
steel (or other alloy) wire or tube. As corrosion occurs, the cross-sectional area of the
metal decreases, and resistance rises. The change in resistance is converted by
appropriate electrical circuitry into a number that tells you the corrosion rate. An
advantage of the corrosion probe is that it determines corrosion rates quickly and
correlates them with plant operating conditions. Probes have the same disadvantages
as coupons, plus a few other disadvantages. Probe readings are accurate only if
corrosion of the probe element is uniform. If a pit or area of localized corrosion
develops, the probe reading changes very little at first. But as penetration increases, the
reading suddenly begins to skyrocket, reaching infinity as the probe element is cut in
half. When a probe is taken out of service, it needs to be visually examined. Disregard
the data from the last part of the probe life if corrosion is observed to be highly non-
uniform.

Experienced equipment inspectors develop a good understanding of how to interpret


ultrasonic (UT) inspection data. They know that they cannot take the readings at face
value. Some of the considerations they take into account include the following:

Accuracy of readings and data scatter.


Reading errors.
Differences between hot and cold readings.
Localized corrosion.
Number of inspection points.

UT readings are generally accurate to 0.01 inch. This means that if we regularly
measure a line that is not corroding, readings are sometimes correct, sometimes
0.01-inch high, and sometimes 0.01 inch low. In this case, the scatter band in the data
is 0.02-inch wide. If we were to take annual readings and calculate the rate from the last
two readings, we might calculate a rate as high as 20 mpy, or even an increase in metal
thickness. The early days of UT produced unnecessary panic when inspectors who
were following a suspected corrosion location would see a 20 mil (0.02 inch) data scatter
on two consecutive monthly readings. This suggested a 240 mpy rate when there
actually was no corrosion.

Confidential
Copyright 2003 by ChevronTexaco
1-50 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

There are several ways to deal with this kind of situation. First, consider the
consequences if the apparent high rate might be real. If it would pose a potential failure
hazard, you could increase the frequency of UT inspection until the question had been
answered. If the apparent corrosion were occurring in an area of non-uniform wall
thickness, such as an elbow or control valve manifold, check by using radiography. In
any case, follow-up UT using a scanning technique would ensure that you hadnt missed
the corroding spot by an inch or two.

Simple errors in UT readings arent unusual and should Simple errors in UT readings
be expected occasionally, in view of the number of arent unusual and should be
readings taken in a refinery each year. Errors become expected occasionally, in view
more common as shortages of trained Company of the number of readings
personnel require the hiring of contract help. Most taken in a refinery each year.
errors stand out readily. For example, in the following
series of annual readings in the record, the mistake is apparent:

0.38 0.37 0.39 0.55 0.37 0.36 0.37 0.35

Clearly, the 0.55 value is incorrect and should be ignored. Although most computerized
UT data programs calculate the long-term rate for us, human intervention is still
important.

Look at the following sequence of UT readings:

0.41 0.37 0.39 0.55 0.37 0.36 0.37 0.35

It is similar to the one above, except that the first reading is 0.41 rather than 0.38.
Assume that the first reading was taken cold, before initial startup of the plant. We might
ignore the initial cold reading in calculating the long term corrosion rate, after the plant
has been running for a significant number of years. Even in old plants, we see hot
(onstream) readings sometimes, and cold (shutdown) readings at other times.
Inspection records should indicate which readings are hot and which are cold so that the
possibilities of differences can be considered when interpreting the data.

Localized corrosion, which usually occurs in turbulent areas, is common in services


where liquid water is present. It is not nearly as common but still might occur in hot, dry
services. Here are some examples:

HOT, DRY SERVICES:

Corrosion Generally Uniform Corrosion Can Be Localized

Hot H2S/H2 (Hydroprocessing) Naphthenic Acid (Crude Units)


Hot H2S (Crude Units) FCC Fractionator Bottoms (Catalyst Fines)

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-51
permission of ERTC
Fluid Catalytic Cracking Process August 2003

WET SERVICES:

Corrosion Generally Uniform Corrosion Can Be Localized

Sea Water NH4HS (Hydroprocessing)


Cooling Tower Water (May Pit) NH4Cl (Crude Unit Overheads)
Steam Condensate
Wet Sulfur

ChevronTexaco Inspection Strategies offer excellent guidance regarding which services


promote uniform attack and which lead to localized corrosion.

Localized corrosion is the most difficult obstacle for inspectors and analysts; it requires
abilities near extrasensory perception to predict where the attack may occur. Inspectors
and analysts must understand the nature of corrosives and the appearance of attack.
For certain services, such as effluent air cooler outlet piping in hydroprocessing units, we
know that corrosion is non-uniform and we have modified our inspection techniques
accordingly. For example, we use a UT scanning technique supplemented by
radiography in such locations. In other services (usually aqueous), we tend to see
higher corrosion rates at bends or other turbulent areas, but the difference between this
and straight pipe, for example, is not great. The usual inspection technique for this type
of situation is to locate many inspection points at bends and turbulent points. Follow-up
inspection can be scheduled for any spots that appear to be suffering accelerated
corrosion.

Bells ring in the refinery when UT shows a sudden and unexpected low reading.
Consider a hypothetical case based on results of two annual readings. We assume that
there are 20 inspection points on the line, and this is the only one that suggests an
attack. In the first set of readings, the values at point x are as follows:

0.34 0.35 0.34 0.33 0.35 0.25 0.34

In the second annual set of readings, the values at point x are

0.34 0.35 0.34 0.33 0.35 0.34 0.25

We dont know if the second reading is in error or if we have a sudden jump in corrosion
rates. We need to go back to find out. Consider the possibility that the 0.09 inch loss
did not occur uniformly during the 1-year elapsed time between measurements, but may
have taken place during the last month. This would represent 1080 mpy instead of
90 mpy. In the interests of safety, we cannot take the chance that it may be the higher
rate.

Consider the first measurements shown above. Hopefully, we had the good sense to
evaluate the apparent low reading by scanning UT and/or radiography as appropriate,
rather than waiting a year before taking the next reading. If that low reading was
carefully evaluated and found to be erroneous, the inspection record can be marked to
show this. However, we cannot assume an erroneous reading without careful check.

Confidential
Copyright 2003 by ChevronTexaco
1-52 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

The number and location of inspection points needs to be carefully considered. Many
inspection points on non-critical lines that are unlikely to corrode serve no purpose
except to take up inspection time. Consider the frequency of inspections in the same
light. All refineries contain potentially corrosive and dangerous systems that demand
frequent and intensive inspection. On the other hand, many handle things like warm,
dry, non-corrosive oil that will never show measurable corrosion during the life of the
plant. Annual UT checks on 40 different inspection points are unnecessary on such
lines.

HOMEWORK QUESTIONS

1.13 Interpretation of Corrosion Data

1) Name two corrosives in FCCs that OFTEN produce highly localized metal loss.

2) Name two corrosives in FCCs that USUALLY produce uniform metal loss.

3) True or False: All ASNT Level II shear wave UT inspectors perform about the same
when it comes to detecting and characterizing weld flaws.

4) True or False: Results to date on ERTCs test bundle show wide variations in the
capabilities of eddy current inspectors in detecting exchanger tube defects.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-53
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
1-54 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.14 CORROSION BY AMMONIUM CHLORIDE

Chlorides are undesirable constituents in FCC Units. They are present in the plant feed
in trace quantities, and their sources are varied. For example, the FCC catalyst contains
chlorides when delivered to the plant for installation, and injection water used for
corrosion control may also contain some chloride.

In the reactor/regenerator system, we occasionally see chloride cracking, but cases are
rare. They are discussed separately under Chloride Stress Corrosion Cracking. On
occasion, chlorides may also cause corrosion. Such attack is typically confined to the
main fractionator overhead and gas recovery systems.

However, in 2000 Richmond experienced significant chloride-related corrosion in the


main fractionator. The corrosion occurred in the top two beds of packing, and where
water collected in these areas. The column had accumulated many days where it had
been exposed to steam layup conditions, and we believe that played a major role n the
corrosion. Other potential problem areas are:

If the main fractionator overhead vapors are cooled until they reach the water
dew point, ammonium chloride is absorbed in the condensate in preference to
ammonium bisulfide, which is present in much higher concentrations. This
absorption leads to an acidic condition in the initial condensate. A little more
cooling causes absorption of ammonium bisulfide in the water, and corrosivity
quickly diminishes. As a result, this corrosion occurs only where we have the
initial condensate. We have no acidic condition at all, and thus no chloride
corrosion, if we inject water before the overhead vapors are cooled below the
dew point. The amount of water must be sufficient to avoid its total evaporation.
To avoid total evaporation and to aid in heat recovery, ChevronTexaco often
injects water after the first exchanger in the main fractionator overhead.
If we inject water into a hot line, but not enough
to quench the process fluid to a temperature We must be sure that water
well below the water dew point, the injection rates are sufficient to allow at
water may boil dry. We then have a corrosive least 25% of the water to
condition, again due to acidic ammonium remain unvaporized at the
chloride. We must be sure that water rates are injection point.
sufficient to allow at least 25% of the water to
remain unvaporized at the injection point. If water is first injected into the main
fractionator overhead vapor line, the chlorides that cause corrosion under a boil-
dry condition probably come from the overhead vapors. Downstream of the
compressors, the source of corrosion-causing chlorides is different. There, the
chlorides probably come from the injection water that was previously used in the
main fractionator overhead and then recycled to a second injection location.

Clearly, chloride corrosion in FCC Units can be prevented by controlling plant operating
conditions and does not require the use of expensive alloys. Because of the large
amounts of water used in the fractionator overhead system, ammonium chloride
corrosion is seldom a problem.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-55
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

1.14 Corrosion by Ammonium Chloride

1) True or False: Like chloride SCC, corrosion by ammonium chloride affects only the
austenitic stainless steels (e.g., 300 Series SS).

2) True or False: Ammonium chloride corrosion in the fractionator overhead can be


controlled by injecting plenty of water before the overhead vapors reach the dew
point.

3) How do we assess whether enough water is being injected in the overhead to


prevent ammonium chloride corrosion?

4) What elements do we want in stainless steels for chloride pitting resistance?

Confidential
Copyright 2003 by ChevronTexaco
1-56 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.15 SULFIDE STRESS CRACKING

Sulfide stress cracking (SSC) is not a true stress corrosion cracking phenomenon,
because it is not caused by corrosion penetration along the grain boundaries of metal.
Be careful not to confuse SSC with SCC (stress corrosion cracking). The two types of
cracking follow different rules.

Sulfide stress cracking is caused by hydrogen generated by aqueous corrosion of metal.


Hydrogen diffuses into the metal in the form of hydrogen atoms (H0) rather than
hydrogen molecules (H2). If the metal is subjected to applied or residual stress, the
stresses tend to concentrate at the tips of minor flaws or discontinuities. The hydrogen
in the metal diffuses to the points of stress concentration. Visualize high stress pulling
the metal atoms farther apart, making more room for the hydrogen. Once the hydrogen
diffuses into this area, it stiffens the metal by reducing the ability of the metal atoms to
move relative to each other under stress. Stress causes the metal to crack rather than
stretch by yielding. Sulfide cracks move in a series of sudden, sharp bursts. The time
between bursts is determined by the time it takes hydrogen to migrate to the new crack
tip.

Similar types of failures can be caused by other corrosive media. Hydrofluoric acid (HF)
is one example. The general phenomenon is called hydrogen stress cracking (HSC).
Sulfide stress cracking is a special case of HSC in which the agent causing the damage
is H2S.

SSC can occur on carbon steel, high-strength low alloys like 4140, chrome-moly steels,
and the 12 Cr and 17 Cr alloys. Duplex alloys also sulfide crack but not as readily. The
austenitic materials, such as the 300 Series stainless steels, Inconels, and Incoloys, do
not sulfide crack under any conditions we are likely to encounter. Metallurgically, crack-
sensitive materials are those with a ferritic structure as opposed to an austenitic
structure.

Soft steels, unlike hard steels, are not susceptible to Cracking does not occur on
SSC (steels in this context referring to all materials materials lower in yield
with a ferritic structure). In general, cracking does not strength than 90 ksi or softer
occur on materials lower in yield strength than 90 ksi or than Rockwell C22.
softer than Rockwell C22. Note that strength and
hardness are directly related; a steel of 90 ksi yield strength has a hardness of about
Rockwell 22.

Defining crack susceptibility by strength and hardness is straightforward for materials of


homogeneous properties, such as a valve stem or pump shaft. It becomes more
complicated when we consider weld in plate or pipe. Here, the weld metal has a
different hardness than the base metal. The weld heat-affected zone has both a hard
and soft band adjacent to the weld. In theory, we should be able to prevent cracking by
keeping base metal, weld metal, and weld heat-affected zone below Rockwell 22. In
practice, however, the hard band in the weld HAZ is so narrow that conventional
hardness measuring techniques cannot find it. The practical answer in dealing with
welds is as follows:

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-57
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Avoid steels stronger than Grade 70.


Limit the hardness of the weld metal to 200 Brinell.
Be careful when selecting welding procedures.

Postweld heat treatment (PWHT) is a potent weapon for control of SSC. It reduces
residual stress and softens the hard heat-affected zone of the weld. It is now common to
specify PWHT for all vessels in sour service. In some situations PWHT may not
accomplish the desired objective totally. Steels of high carbon equivalent (i.e., high in
carbon, moly, or certain trace elements) do not respond to heat treatment. Consult with
Materials Engineers when buying a new column or vessel for sour service.

On ordinary carbon steels, we see sulfide cracking only at welds. Cracks in high
strength steels may occur anywhere, welded or not. An example is SSC of 12-chrome
valve stems, which are normally very hard. To solve this problem, use an 18-8
(austenitic) stainless steel.

Gas compressors can be another source of trouble. In centrifugals, we have avoided


blade failures in sour services by limiting hardness to Rockwell 22. However, this
approach cant be used for valve springs in reciprocating compressors, where numerous
failures have occurred. The answer here is to use a resistant alloy.

Even though a material is susceptible, cracking occurs only when we have a corrosive
medium capable of corroding steel and charging the metal with hydrogen while doing so.
This process requires an aqueous solution; it does not occur in the hot and dry sections
of the plant. In FCC Units, we may see it in the main fractionator overhead system and
in all downstream equipment exposed to wet, sour conditionsincluding the gas
recovery section. We are particularly concerned about systems in which condensed
water containing ammonia, H2S, and cyanide are present. The higher the
concentrations of these corrosives, the more likely we are to have SSC. When cyanides
are present, as they are in all FCC Units, higher pH (more NH3) increases the tendency
to crack.

Also see the section of this document covering Wet H2S Cracking. This term covers a
variety of failure mechanismsincluding sulfide stress crackingall of which can be
found in wet, sour services. Also see Inspection Strategy IS-17 on wet H2S cracking.

Confidential
Copyright 2003 by ChevronTexaco
1-58 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

HOMEWORK QUESTIONS

1.15 Sulfide Stress Cracking

1) Sulfide stress cracking (SSC) occurs only on high strength, high hardness steels.
Cracking does not occur on materials lower in yield strength than _____ ksi or softer
than Rockwell C_____.

2) True or False: Sulfide stress cracking generally occurs much faster than the other
forms of wet H2S cracking (e.g., HIC, SOHIC).

3) True or False: SSC has caused failures of B7 floating head bolts and 410 SS tray
support bolts.

4) True or False: Even Grade 60 carbon steel can SSC in weld heat affected zones if
its welded wrong.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-59
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
1-60 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.16 CORROSION OF DEAD LEGS

Dead legs are defined as piping segments that normally see no flow but remain
pressurized during plant operation. Examples are vents, drains, instrument bridles on
vessels, and closed bypasses on exchangers or control valves. Temperature and fluid
composition can be different in a dead leg than in the main process line.

One cause of dead leg corrosion is the condensation of moisture from a process fluid
above the water dew point in the main flow line. This problem has occurred in resid
hydrotreaters, where water condensation and pick-up of chlorides in the water causes
chloride stress corrosion cracking of non-flow branch connections in hot stainless steel
piping.

A second cause of dead leg corrosion is the accumulation of stagnant water from a
process fluid containing a separate water phase. The presence of liquid water in a
process fluid may not cause corrosion of the main flow lines if the concentration of
corrosives is low or the metal is washed regularly by the non-corrosive oil phase.
However, different conditions exist in dead legs:

The water phase may puddle so that the metal is in contact with water at all
times.
Corrosive constituents may concentrate in the water, increasing corrosivity.
Fouling materials, such as scale and sludge, may accumulate in dead areas.
Iron scales (iron sulfide or iron oxide) can form a galvanic cell with carbon steel if
water is present, causing corrosion of the steel under deposits.

Another form of dead leg corrosion can occur when certain chemical compounds form
corrosives as a result of thermal decomposition. The decomposition may be slow, but
concentrations can reach high levels because there is no opportunity for them to escape
from the system. Examples are as follows:

Hot oil lines (generally 500F+) can slowly generate hydrogen sulfide by
decomposition of sulfur compounds. For this to occur, the temperature must be
high enough for H2S to be corrosive, and hot enough to cause thermal
decomposition.
Chloride injection lines in Rheniformers may corrode where they connect to the
reactor feed line. This is caused by the thermal
decomposition of organic chloride to form HCl. We cannot predict with
certainty which dead legs will
In FCC Plants, the most likely place for dead leg corrode and which will not.
corrosion is in fractionator overhead lines, where However, experience
aqueous ammonium bisulfide is present. We cannot eventually helps us identify the
predict with certainty which dead legs will corrode and more dangerous services.
which will not. However, experience eventually helps Dead legs should be eliminated
us identify the more dangerous services. Dead legs wherever they are not
should be eliminated wherever they are not necessary. necessary. Dead legs require
Dead legs require regular inspection. regular inspection.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-61
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

1.16 Corrosion of Dead Legs

1) Dead legs should be __________ wherever they are not necessary.

2) Where are good locations to look for dead leg corrosion in FCC Plants?

Confidential
Copyright 2003 by ChevronTexaco
1-62 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.17 CORROSION BY AMMONIUM POLYSULFIDE

Ammonium polysulfide (APS) is used in FCC main fractionator overhead systems to


control hydrogen blistering and related phenomena (HIC, SOHIC, and sulfide stress
cracking). It reacts chemically with cyanides, which are potent promoters of corrosion
damage. To a lesser extent, it reduces corrosion caused by ammonium bisulfide by
reinforcing the sulfide film on the metal and making it more resistant to removal by high
velocity fluids. Sodium polysulfide, sometimes used in hydroprocessing units, is not
used in FCC, because it lacks high reaction rates with cyanides.

Dilute solutions of polysulfide are not corrosive to carbon steel or the stainless steels.
However, they do corrode copper-base alloys. Polysulfide solutions become corrosive
to steel when low pH or oxygen cause the polysulfide to decompose and form elemental
sulfur. It is wet sulfur rather than the polysulfide itself that is corrosive.

We see this problem almost exclusively in water injection lines to which polysulfide has
been added. However, we do not see it in the main process line in an FCC Unit
because here we are adding polysulfide to a high pH, ammonia-containing fluid. Also,
any oxygen in the injection water has already been consumed by reacting with the
polysulfide.

In order to control corrosion, water for diluting polysulfide must be low in oxygen.
Maximum allowable O2 content is 15 ppb. Additionally, the pH should be in the range of
8 to 10.

The source of dilution water must be carefully selected to insure that it is free of oxygen
and high enough in pH. Stripped sour water has the advantage of being free of oxygen.
It may or may not be high enough in pH, depending on how completely the ammonia has
been stripped. Stripped water may also contain dissolved salts like ammonium chloride
that depress the pH to the acid side. Steam condensate is always suspect as a water
source because it is often contaminated with oxygen from vacuum leaks in surface
condensers. The pH may also be low due to dissolved carbon dioxide. Both stripped
sour water and steam condensate must be checked for dissolved iron which precipitates
as iron sulfide on contact with either the polysulfide or H2S in the main process stream.
The iron sulfide can cause serious fouling.

A recently issued ammonium polysulfide treatment guide is included in the back of this
manual as a reference. Additional details on APS handling and application are included
in the FCC Overhead Corrosion Control document, also in the back of this manual.

Stainless steels are not corroded by polysulfide solutions, even if the polysulfide has
decomposed from low pH or the presence of oxygen.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-63
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

1.17 Corrosion by Ammonium Polysulfide

1) Water for diluting polysulfide should have a maximum oxygen content of _____ ppb
and a pH of _____ to _____.

2) True or False: Stripped water is preferred to steam condensate for dilution water.

3) True or False: Copper-based alloys resist corrosion by polysulfides.

Confidential
Copyright 2003 by ChevronTexaco
1-64 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.18 CORROSION AT SPECIFICATION BREAKS

Plant piping systems use a variety of pipe classes to accommodate changes in


temperature, pressure, or corrosivity. For example, if temperatures in a flow stream drop
when the fluid cools, it may be economical to change from stainless steel pipe to carbon
steel because there is less corrosivity at lower temperatures. In this case, we would
change from a stainless steel pipe class to a carbon steel class. The point at which the
change occurs is called a specification break (or spec break).

A likely place to look for corrosion is immediately adjacent A likely place to look for
to the spec break point, on the lower alloy or lesser- corrosion is immediately
corrosion-allowance side. Corrosion may occur if the spec adjacent to the spec break
break was not made at exactly the right place. It is difficult point, on the lower alloy or
to predict the conditions that might exist at the spec break, lesser-corrosion-allowance
especially conditions that might change years later. side.
Youll usually find a valve at the spec break point; it separates a flow system from a non-
flow or intermittently flowing system. Valves at the spec break point should conform to
the piping standard for the higher alloy or higher-corrosion-allowance pipe class.

In an FCC Unit, look for corrosion where you find a class change between Cr-Mo or
stainless steel and carbon steelin the main fractionator sidecut and bottoms lines, for
example. Drain lines and other non-flow connections to Cr-Mo piping are often carbon
steel and require regular inspection at the spec break point. Hot oil lines are especially
suspect. Above approximately 500F, high-sulfur oils slowly crack to liberate corrosive
H2S.

Heat exchanger bypass lines are always suspect. If the process fluid is being heated, a
leaking by-pass valve leaks cold fluid around the exchanger, and corrosion rates are
decreased. But if the fluid is being cooled, a leaking by-pass causes higher-than-normal
temperatures in the bypass and can cause accelerated corrosion. One of the worst fires
in Company history, at the Richmond Isomax in 1989, resulted from a materials failure
caused by an erroneous temperature prediction in an exchanger bypass line.

Although an FCC main fractionator does not have an area that fits the exact definition of
a spec break, it is similarly suspect. This column is usually clad with 12 chrome in the
hot zone. Inspectors must watch for corrosion on the bare carbon steel immediately
above the termination of the cladding.

Inspection Strategy IS-24 discusses corrosion at spec breaks and outlines inspection
techniques appropriate to the situation.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-65
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

1.18 Corrosion at Specification Breaks

1) Valves at spec breaks should conform to the piping standard for the __________
__________ or __________ __________ __________ pipe class.

2) True or False: Heat exchanger bypass lines are usually a good place to look for
spec break-associated corrosion problems.

Confidential
Copyright 2003 by ChevronTexaco
1-66 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and the FCC Process

1.19 HIGH TEMPERATURE CARBURIZATION

At high temperatures, metals and alloys can absorb carbon from the surrounding
atmosphere to form metal carbides, a process called carburization. The extent to which
carburization occurs depends on temperature, environment, and type of metal.
Carburization can be constructive, providing hard, wear-resistant surfaces on
mechanical parts, or it can be destructive, as in furnace tubes in ethylene furnaces. In
FCC Units, we classify it as a phenomenon of minor concern. Carburization involves a
series of steps:

Carbon (coke) is deposited on the metal surface. Unsaturated hydrocarbons


(olefins) in the surrounding atmosphere tend to promote damage. These olefins
exist in large quantities in the FCC reactor environment.
The carbon reacts with the metal to form metal carbides.
Diffusion of carbon and/or metal atoms within the metal control the extent to
which the reaction product (metal carbide) penetrates the metal. This diffusion is
a function of temperature and material composition.
As the metal carbide layer forms, it is under a high compressive stress because it
occupies a greater volume than the base metal. This layer is prevented from free
expansion by the restraint of the unaffected metal underneath. It may bulge
away from the underlying metal and flake off, gradually reducing metal thickness.

Although we havent examined affected areas metallurgically, visual observation points


to carburization as the cause of the slow thinning of FCC reactor cyclones.

Carbon pick-up in the base metal as a result of carburization may also make the base
metal more difficult to weld. Higher carbon will result in a hard and brittle weld zone.

The effect of alloying elements on carburization depends greatly on circumstances. In a


reducing environment, chromium may cause more rapid carburization. However, in an
oxidizing or sulfidizing environment, chromium, because of its greater affinity for sulfur or
oxygen than carbon, may retard carburization. This characteristic of chromium can be
useful in actual practice. In ethylene furnaces, for example, deliberate addition of H2S
minimizes carburization of stainless steel furnace tubes. In FCC reactors, carbon steel
cyclones may lose metal by carburization, but the 1-1/4 Cr reactor shells do not
carburize significantly. Regenerator internals do not carburize significantly in units
running with complete combustion. With incomplete combustion, CO in the regenerator
can carburize 300 series stainless steel.

In summary, we largely ignore carburization in FCC We largely ignore carburization


Units. Even though using 1-1/4 Cr for reactor in FCC Units. Even though
cyclones could reduce carburization, the far greater using 1-1/4 Cr for reactor
maintenance and repair problems associated with cyclones could reduce
Cr-Mo steels makes it a poor choice. carburization, the far greater
maintenance and repair problems
associated with Cr-Mo steels
makes it a poor choice.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
1-67
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

1.19 High Temperature Carburization

1) True or False: Carburization can make weld repairs to carbon steel reactor cyclones
more difficult.

2) True or False: 300 Series stainless steels are immune to carburization.

3) Under what situation is carburization a concern for regenerator internals?

Confidential
Copyright 2003 by ChevronTexaco
1-68 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

SECTION 2 METALLURGICAL AND REFRACTORY


CONSIDERATIONS IN FCC PLANTS

2.0 Overview: Metallurgical and Refractory Considerations in FCC Plants ....... 2-1

2.1 Sigma Phase Embrittlement............................................................................... 2-3

2.2 Refractory Linings............................................................................................... 2-5

2.3 Welding Issues .................................................................................................... 2-9

2.4 High Temperature Graphitization .................................................................... 2-15

2.5 High Temperature Creep .................................................................................. 2-17

2.6 Brittle Fracture and Autorefrigeration............................................................. 2-21

2.7 Fatigue ............................................................................................................... 2-29

2.8 885 Embrittlement and Creep Embrittlement ................................................. 2-35

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

2.0 OVERVIEW: METALLURGICAL AND REFRACTORY


CONSIDERATIONS IN FCC PLANTS

Section 1 of this manual describes the many types of corrosion that can occur in FCC
Plants. High temperatures, corrosive liquids and gases, and erosive solids create
environments in which serious metal loss can occur. High temperatures also result in
metallurgical problems that we do not see in most other process plants.

One important cause of equipment damage is thermal expansion. When temperatures


approach 1300F in the regenerator, for example, internal parts expand and contract,
imposing high stresses where structural restraints prevent free movement. High
stresses due to temperature differentials are also imposed on certain piping and
equipment parts. At the temperatures we see in reactors and regenerators, materials
distort and may eventually fail due to creep. Instead of being elasticstretching and
then returning to their original shape and size on coolingthey are permanently
deformed. They may stretch to the point of in-service failure. In the early years of FCC
operation, our trial-and-error attempts at solving creep problems resulted in changes in
reactor or regenerator internal configurations during almost every turnaround.

We encounter several phenomena in FCC Units that can result in embrittlement. Welds
in carbon steel and low alloys may suffer from graphitization, in which the carbon in the
weld metal coalesces into graphite particles. At reactor temperatures, coke (carbon)
deposits slowly react with the metal, resulting in what metallurgists call carburization.
Metal is lost as this brittle surface layer gradually flakes off. In addition to graphitization
and carburization, stainless steel welds are also subject to the formation of a brittle
material called sigma phase.

Graphitization and carburization are often confused. Graphitization and sigma phase
formation are metallurgical phenomena; they involve structural changes that occur in
metal exposed to high temperature, changes that do not involve interaction with the
surrounding atmosphere. On the other hand, carburization (discussed in Section 1 of
this manual) is a corrosion phenomenon and does involve a reaction with the
surrounding atmosphere.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-1
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
2-2 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

2.1 SIGMA PHASE EMBRITTLEMENT

When a molten metal solidifies, its metal atoms line up in distinctive patterns. This
process, called crystallization, can produce cubes, hexagons, or other complex shapes.
Each shape has a metallurgical name. Iron and steel form a crystal we call ferrite. The
300 Series stainless steels form a different crystal shape called austenite.

Primarily, chemical composition determines crystal shape. Iron and steel, as well as
steels that contain only chromium as an alloying element, are generally ferritic.
Stainless steels high in nickel are austenitic. Those with lower amounts of nickel can be
both ferrite and austenite; these are the duplex stainless steels. The 300 Series
stainless steels are austenitic, and the 400 Series are (with a few exceptions) ferritic.
The duplex stainlesses have no equivalent Series.

Heating and cooling, with variations in time and temperature, can cause changes in the
crystal structure. New crystal shapes may form in addition to the common ferrite or
austenite. For example, chromium carbide crystals may form in the grain boundaries,
resulting in sensitization. Sensitization can cause corrosion problems and is discussed
in Section 1 of this manual. Another type of crystal that may form is sigma phase, the
subject of this chapter. Sigma phase is a brittle constituent that can damage the
mechanical properties of the alloy and produce a problem called sigma phase
embrittlement.

Sigma embrittlement is primarily associated with the 300 Series (austenitic) stainless
steels. It is an iron-chromium compound that can occur in steels containing more than
approximately 17% chromium. Sigma embrittlement forms when the steel is heated for
extended periods in the range of 1100-1700F. The time required depends on
temperature and metal composition, forming fastest approximately at the mid-point of
this temperature range.

Wrought 300 Series stainless steels (sheet, plate, barstock) are entirely austenitic.
Sigma phase forms slowly in these alloys. But weld metal for these steels typically
contains 3-10% ferrite, the amount needed to avoid hot cracking during welding.
Because sigma forms much more rapidly from ferrite than from austenite, weld metal
embrittles much more seriously than the base metal.

Certain chemical elements tend to promote sigma embrittlement. Types 316 and 347
embrittle more rapidly than Type 304. However, 304 does not come with a guarantee
against this problem. Sigma embrittlement failures of 304 components in FCC
regenerators have been reported.

Brittleness due to sigma phase formation is most Brittleness due to sigma phase
serious when the metal is cold. It tends to formation is most serious when the
disappear when the metal is heated above metal is cold. It tends to disappear
approximately 1000500F and to reappear on when the metal is heated above
cooling. Thus embrittlement is not likely to cause an approximately 1000500F and to
on-stream failure. However, where service reappear on cooling.
temperatures are high enough to cause sigma
formation (i.e., in the regenerator system), be careful when performing maintenance
work. Try to minimize shock loadings that might trigger a brittle fracture.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-3
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Also, sigma embrittled stainless steel is susceptible to cracking during welding. Special
weld procedures that minimize residual stress may be used. In some cases it is also
possible to perform a de-embrittling heat treatment prior to welding.

Stainless steel castings frequently contain ferrite in large amounts. As a result, they
sigma-embrittle rapidly in high-temperature service. A cast stainless slide valve, for
example, could become badly embrittled. We solved this problem by moving to
internally-insulated carbon steel slide valves.

Sigma embrittlement is controlled in several ways:

Limit the ferrite content of weld metal. A minimum ferrite content of about 3% is
necessary to insure weldability, but the maximum should not exceed 10%. This
is our primary method of sigma control!
Try to avoid shock loadings when the metal is cold.
Use Type 304 steel in the regenerator system in preference to other grades. In
the past, we have made one exception to this rule: to reduce the chance of
polythionic acid cracking, we have often specified Type 347 for cyclone hanger
rods. However, operating experience does not fully support this exception; 304
hanger rods have performed satisfactorily at many ChevronTexaco and
competitor locations.

HOMEWORK QUESTIONS

2.1 Sigma Phase Embrittlement

1) True or False: Sigma forms more quickly in the austenite phase than the ferrite
phase of 300 Series stainless steels.

2) True or False: Sigma phase doesnt cause brittleness at reactor-regenerator


operating temperatures and is primarily a concern only when the metal is cool.

3) Sigma formation requires prolonged exposure in the _____ F to _____ F


temperature range.

4) True or False: Stainless steel casting are more resistant than wrought stainless
steels to sigma formation.

5) Limit the ferrite content in stainless steel welds to _____ %.

Confidential
Copyright 2003 by ChevronTexaco
2-4 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

2.2 REFRACTORY LININGS

Refractory linings are used in the FCC reactor/ Refractory linings serve two primary
regenerator system and its internal parts, such purposes: protection against erosion
as cyclones and associated piping. They serve caused by moving catalyst and
two primary purposes: protection against protection against high temperatures.
erosion caused by moving catalyst and
protection against high temperatures. In any given location, we may need erosion
resistance, thermal insulation, or both. What we need determines the type of refractory
we use.

The refractories consist of an aggregate and a binder. The aggregate comprises 60 to


80% of the refractory weight. It may be a very light, porous material for insulation or a
very hard and dense material for erosion resistance. If we need both insulation and
erosion resistance, we have two choices. We can use a dual-layer process in which a
light insulating refractory is placed against the vessel wall, and a second layer of dense,
erosion-resistant refractory is placed over it. Or we can use a single-layer material of
intermediate-insulating and erosion-resistance ability. The dual-layer refractories
provide better erosion resistance and insulation, but they are much more expensive and
difficult to maintain. Therefore, we use the intermediate-density, single-layer materials
wherever experience assures us that they are adequate.

The refractory binder is an inorganic material, usually a calcium aluminate cement. In


most cases, the refractory material is mixed with water and then put in place, where it
sets by reacting with water. A few refractories use a binder containing phosphoric acid.
Ordinary portland cement is not used as a binder because it deteriorates rapidly at
temperatures above 800F. The refractories we use must be able to withstand 1400F,
to which we generally add a safety margin of 200F. The high-temperature refractories
we use are generally rated for 2000F+.

Refractory selection and use is covered in more detail in the FCC Refractory Guidelines
Report included in the back of this manual, and periodically updated on the FCC Best
Practices Website. Refractories are classified as Types A through I, respectively.
These types are defined in Standard Drawing GB-N1268 in the Refractory Gray Manual.
Obviously, selecting the correct refractory requires that we balance the insulation and
erosion-resistance needs for the specific application. If the refractory is too soft, it may
erode quickly, exposing the underlying metal to erosion and overheating. If it is too hard,
it may not protect the underlying metal from overheating and distress from local thermal
expansion.

Refractory linings do not adhere to the metal surface and must be anchored in place.
Thin (typically 1") erosion-resistant linings are usually anchored by 12 chrome or type
304 stainless steel hexmesh that is welded directly to the underlying metal. More
recently, S-bars, K-bars, hex-cells, and curl anchors have been introduced. These
minimize breakage of hexmesh where it must be forced to fit surfaces of extreme
curvature. (See Refractory Manual for details on how to attach the hexmesh and
S-bars.) We use punch tabs for extending liners around corners, and U-tabs for lining
plate edges. Dual-layer linings, involving an erosion-resistant layer over an insulating
refractory, also use hexmesh. But in this case, the hexmesh is held away from the
underlying surface by studs that are as thick as the insulating layer. We are moving

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-5
permission of ERTC
Fluid Catalytic Cracking Process August 2003

away from the dual component hexmesh lining systems because of the maintenance
problems that hexmesh produces.

A single-layer, insulating refractory lining or a refractory lining with both insulating and
erosion-resistant characteristics must be relatively thick. A thickness of 4 to 5 inches is
common. Such linings are usually placed by either gunning, vibration casting, or using
self-leveling refractory materials. This is in contrast to the very high density erosion-
resistant linings, which are rammed or hand packed in place. Support is provided by
anchors, which come in various sizes and shapes. Details on anchor types, spacing,
and attachments can be found in the Refractory Manual. Footed "wavy Vee" anchors
are preferred, and their length should be 75% of the refractory thickness.

Choice of anchor material depends on the application. Almost all anchors or hexmesh
are Type 304 steel, with one exception. 12 chrome (Type 410 steel) is used for
hexmesh in the reactor and reactor cyclones, where it is welded directly to the
underlying steel to support a 1" erosion-resistant lining. The choice of 12 chrome here is
dictated by thermal expansion considerations. The 12 Cr has a coefficient of thermal
expansion close to that of carbon steel or 1-1/4 Cr. Therefore, thermal expansion is less
likely to break the attachment welds. For a dual-layer lining, where the hexmesh is on
studs, away from the vessel wall, Type 304 is probably a better hexmesh choice than
12 Cr.

Refractory linings provide huge cost savings by permitting us to use less expensive
materials. Without a lining to keep it cool, the shell of the regenerator would have to be
made of stainless steel. We cannot keep cyclones cool because they have hot process
gas on all sides. However, we can and do prolong their life greatly by using high-density
refractory linings to protect against erosion. Later sections of this manual deal with the
refractory requirements of specific equipment. Additionally, see the section on erosion.

In summary, we can classify refractories for the reactor/regenerator system as follows:

Low Density Typically thick. Used for thermal insulation. Poor resistance to
erosion.

Intermediate Density Used as a compromise where we need both erosion resistance


and thermal insulation. Typically thick.

High Density Used for erosion resistance only. Not a good thermal insulator.
Typically thin (1"). These are usually phosphate-bonded and
rammed in contrast to others that are hydraulically-bonded and
gun-applied. They are usually hexmesh supported (in either
single or dual layer) in contrast to others that are supported by
Vee anchors.

The light or intermediate-density materials that are supported by anchors rather than
hexmesh may develop cracks due to thermal expansion and contraction. In the worst
case, the cracks may pull apart, leaving a gap in the refractory. Or cracks may intersect,
causing a piece of refractory to fall out. These problems can be minimized to a great
extent by metal fiber reinforcement. Addition of stainless steel fibers to the refractory
mix holds the refractory together, even when cracked.

Confidential
Copyright 2003 by ChevronTexaco
2-6 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

The length of time that an FCC Unit can run The length of time that an FCC Unit
between turnarounds may well be determined by can run between turnarounds may well
the durability of the refractory linings. To be determined by the durability of the
prolong lining life, get expert advice during refractory linings.
refractory installation and repair. Also get expert
advice on dryout procedures. Inspection and testing are of utmost importance. (Again,
see the Refractory Manual on this subject.) We know that plant operators want to get
the plant back on-stream as quickly as possible; that is their job. However, careful
refractory installation and dryout pay dividends by extending the plants run length.

HOMEWORK QUESTIONS

2.2 Refractory Linings

1) Where is a good place to use independent (such as S-bars, curl anchors) anchors
instead of hexmesh?

2) The length of wavy Vee anchors should be about _____ the length of _____ % of the
insulation thicknesses.

3) Insulating and intermediate density refractories are typically applied _____ inches
thick. Erosion-resistant refractory is typically _____-inch thick.

4) Where are 12 Cr anchors used instead of the usual 304 SS? Whats the reason for
using 12 CR at this location?

5) Whats the purpose of adding stainless steel fibers to refractory?

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-7
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
2-8 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

2.3 WELDING ISSUES

Background

Most refinery piping and equipment are fabricated by welding. Welds need special
attention and inspection. Welding alters the structure of materials used to construct
equipment and the joint design may lead to stress intensification. Changes in
metallurgical structure caused by welding can affect corrosion and mechanical
performance of the materials. For example, welded stainless steels can be susceptible
to intergranular corrosion and stress corrosion cracking. Improper welding of carbon
and low alloy steels can produce local hardening, which can reduce toughness or render
materials susceptible to hydrogen sulfide cracking.

By Code and ChevronTexaco requirements, welding must be performed according to


qualified welding procedures. Procedures for welding piping and equipment are
normally qualified under the requirements of ASME Section IX. Structural welding
procedures are usually governed by AWS D1.1. These procedures set parameters
(such as welding current and voltage, electrode size, and travel speed to make a sound
weld) and specify requirements for preheat and postweld heat treatment. Preheat and
postweld heat treatment conditions are determined by the construction Code, e.g.,
ASME B31.3, and by ChevronTexaco standards, which are sometimes more restrictive.

To avoid qualifying a separate procedure for each combination of base metal and filler
metal, Section IX groups materials of similar welding characteristics in P-Numbers (base
metal properties), F-Numbers (electrode and weld rod usability), and A-Numbers (weld
metal composition). Paraphrasing the Code, these assignments are based on
comparable characteristics where this can logically be done. These assignments do not
imply that a material can be indiscriminately substituted for another in the same P-, F-, or
A-Number. Metallurgical and service conditions must be considered. For example,
Type 317L stainless steel could be welded with ER308L from a procedure qualification
standpoint, but the combination is not appropriate from a corrosion standpoint.

Even the alloys and stainless steels used in refineries that are readily weldable require
special attention, as discussed in the following sections.

Welding Cr-Mo Steels

Chromium-molybdenum steels find wide application within refineries because they resist
hydrogen attack, have good high-temperature strength, and resist sulfide corrosion. The
most commonly used Cr-Mo steels include 1-1/4 Cr - 1/2 Mo, 2-1/4 Cr - 1 Mo, 5 Cr -
1/2 Mo, and 9 Cr - 1 Mo. In Section IX, the 1-1/4 Cr - 1/2 Mo steels are classified as P-
No. 4; the other steels are P-No. 5.

Because of their high hardenability, welding these materials requires more care than
welding carbon steel. That is, their alloy content may allow metallurgical reactions
during welding that produce hard and sometimes brittle phases, such as martensite and
bainite. These hard phases reduce toughness and ductility of the materials and increase
susceptibility to hydrogen embrittlement.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-9
permission of ERTC
Fluid Catalytic Cracking Process August 2003

The detrimental phases occur because of rapid cooling during welding, and the
possibility of their formation increases with increasing alloy content. You can retard the
cooling rate by preheating the weld joint. The higher the alloy content of the steel and
the thicker the piece, the higher the preheat. Table 2.3-A lists the recommended
preheat temperatures for various Cr-Mo steels. Preheating is not only good practice, but
is also required by construction Codes and ChevronTexaco Standards.

Table 2.3-A. Minimum preheat temperatures for Cr-Mo steels.


Material Minimum Preheat Temperature F
1 Cr - 1/2 Mo 300
1-1/4 Cr - 1/2 Mo 300 (350 for Piping)
2-1/4 Cr - 1 Mo 300 (350 for Piping)
3 Cr - 1 Mo 350
5 Cr - 1/2 Mo to 9 Cr - 1 Mo 450

Because Cr-Mo steels are susceptible to hydrogen embrittlement, low-hydrogen welding


practices and consumables are mandatory. Weld electrodes must be kept warm and dry
in a hotbox until they are used in order to prevent pickup of moisture from the
atmosphere. Prior to welding, the joint should be thoroughly cleaned and dried. When
repair welds or modifications are made to materials already in service, the area of the
weld must be cleaned.

Preheat should be maintained throughout the welding sequence. For Cr-Mo steels
containing more than 2% Cr, maintain the preheat until the postweld heat treatment
(PWHT) begins. The weld should be completed without interruption, but if it must be
interrupted, use intermediate heat treatments or special precautions to insure slow
cooling.

Postweld heat treatment relieves the stresses caused by welding and tempers (softens)
the hard phases that may have formed. The minimum temperature is determined by the
grade of steel and the minimum time by the thickness of the joint. Recommended
temperatures and times are given in Table 2.3-B. It is a ChevronTexaco practice to
postweld-heat-treat all welds in Cr-Mo steels, although exceptions are sometimes
allowed for steels containing up to 2-1/4 Cr if special welding techniques are used. In
FCCs, this exception could be a major benefit in vessel repairs. However, consult ERTC
before taking action. You can check the effectiveness of the heat treatment by hardness
measurements. ChevronTexaco requires a maximum hardness of 215 Brinell (HB) on
completed Cr-Mo welds.

Table 2.3-B. Postweld heat treatment of Cr-Mo steels.


Material Temperature, F Minimum Time, Hr
1 Cr - 1/2 Mo 1300-1375 2
1-1/4 Cr - 1/2 Mo 1300-1375 2
2-1/4 Cr - 1 Mo 1300-1375 2
3 Cr - 1 Mo 1300-1375 (1325-1400 for Piping) 2
5 Cr - 1/2 Mo to 9 Cr-1 Mo 1325-1400 2

Confidential
Copyright 2003 by ChevronTexaco
2-10 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

Hot-wall reactors are typically constructed of


PWHT of FCC vessel repairs requires
1-1/4 Cr - 1/2 Mo steel. Repairs, therefore,
careful temperature control and
usually require postweld heat treatment. We
disconnection of all external
usually give a column or vessel a full
connections to the vessel that might
encirclement heat treatment. This treatment lets interfere with free thermal expansion.
the vessel expand uniformly in an upward
direction while its hot, thus minimizing stresses due to dissimilar thermal expansion. But
heat treatments in FCC reactors present special problems. First of all, heat treating a
band all the way around a reactor is expensive and time consuming. Secondly, the top
part of the reactor is so heavy that it cannot support itself adequately if a band of metal
around a lower part is hot, soft, and subject to high temperature creep. The vessel top
must be supported by a crane during the treatment of such a band.

Therefore, it is preferable to heat only a spot instead of fully encircling the vessel.
However, this can create high thermal expansion stresses around the periphery of the
spot and may lead to bulging if not properly engineered. Local heat treatments are
normally avoided for stress corrosion cracking services where it is important to reduce
residual stresses to a minimum. But for the 1-1/4 Cr FCC reactors and fractionators, our
main goal is to soften the weld and HAZ, so a local spot PWHT is allowed. Always
consult ERTC for assistance when considering PWHT of large Cr-Mo vessels and
equipment. No matter which method you use, control the temperature carefully, and be
sure to disconnect all external connections to the vessel that might interfere with free
thermal expansion.

Refractory anchors can be welded to 1-1/4 Cr without subsequent PWHT since the
welds are small and non-pressure containing. For new construction, refractory anchor
welds on Cr-Mo material are typically PWHTd, because it is easy to do for new
construction.

Welding Austenitic Stainless Steels

The normal metallurgical structure of 300 Series stainless steels is the non-magnetic
face-centered cubic phase, austenite. During solidification from liquid, another phase
may formmagnetic body-centered ferrite. Mechanical work and heat treatment after
solidification remove the ferrite from wrought austenitic stainless steels, but the ferrite
remains in cast products and weld metal.

Some ferrite is needed in the weld metal of austenitic stainless steels to prevent hot
crackingthe minimum amount required cannot be defined precisely, and it depends on
the alloy. ChevronTexaco experience indicates that a minimum ferrite number of 3 is
satisfactory for 300 Series stainless steels. The ferrite number is an arbitrary
standardized number that approximately equals the percent ferrite in the weld metal. A
number of instruments (e.g., ferritescope) can be used to measure the ferrite number of
weld metal, but the most common way to determine it is from the chemical composition
of deposited weld metal. ChevronTexaco welding standards require chemical analysis
of the filler metal used for production. This analysis can be used to calculate the ferrite
number from the Schaeffler diagram or, preferably, the Delong or Welding Research
Council diagrams. See the Welding Manual for these diagrams.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-11
permission of ERTC
Fluid Catalytic Cracking Process August 2003

In addition to retarding hot cracking, ferrite affects austenitic stainless steel weld metal in
other ways. It can reduce toughness at cryogenic temperatures. As a result, impact
testing of weld metal is required when the design temperature is less than -20F. Ferrite
can also transform to sigma phase when the service temperature exceeds 1100F.
Sigma phase is a hard, brittle constituent that can reduce toughness of stainless steels if
present in sufficient quantities. The ChevronTexaco practice of restricting the maximum
ferrite number to 10 lessens the probability that a harmful amount of sigma will form.

Type 347 stainless steel is frequently used at elevated temperatures because of its high-
temperature strength and resistance to sensitization. However, this alloy is more
sensitive to weld cracking than other austenitic stainless steels. Many cracking
problems reported with Type 347 have been hot cracking in the weld metal; however,
modern practice has almost eliminated this type of cracking by controlling the chemical
composition of filler metal. Carbon, columbium, silicon, phosphorous, and sulfur must be
kept as low as feasible, consistent with strength and stabilization requirements. This is
particularly an issue for field repair welds.

Another problem is cracking in the heat affected zone (HAZ) after postweld heat
treatment. Welds in Type 347 are frequently postweld-heat treated to minimize the risk
of SCC and/or lock in resistance to sensitization. In highly restrained welds, this
treatment may lead to cracking, usually only in joints thicker than 3/4 inch. To detect
reheat cracking, you need to perform an acceptance test such as radiography after
completion of postweld heat treatment.

Weldments may also crack if the material has been sigma embrittled due to long-term
exposure. Procedures that minimize residual stress and distortion may be beneficial. In
some cases it may be possible to perform a heat treatment that restores some of the
ductility, but special procedures are required. Consult a welding specialist or materials
engineer.

Dissimilar Metal Welds

It is sometimes necessary to make welds between dissimilar metals, such as carbon


steel and stainless steel. Most metals found in refineriescarbon steel, low alloy steels,
stainless steels, and nickel alloysare metallurgically compatible. They can be welded
to one another successfully as long as their differences are considered when developing
the welding procedure and selecting the filler materials. Some materials, such as
titanium and aluminum, cannot be joined to other metals by fusion welding. Special
techniques are required in these cases.

Here are practices to follow when joining carbon steels to low alloy steels, or when
joining low alloy steels of different compositions:

Use the filler metal corresponding to less alloyed material.


Use the preheat required for the more highly alloyed steel for joints between
dissimilar ferritic materials (carbon and low alloy steels).
Use the postweld heat treatment temperature for the higher alloy material;
however, the temperature must not exceed the lower critical temperature for
either steel.

Confidential
Copyright 2003 by ChevronTexaco
2-12 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

To avoid exceeding the lower critical temperatures, ChevronTexaco recommends a


maximum postweld heat treatment temperature of 1325F for welds between carbon
steels and Cr-Mo steels. This temperature is at the low end of the range used for Cr-Mo
steels. Therefore, the softening of heat affected zones of Cr-Mo steels by postweld heat
treatment may be less than expected when this lower postweld heat treatment
temperature is used.

Several factors must be considered when welding carbon and low alloy steels to
austenitic stainless steels. Type 309 stainless or the nickel-based Inco 182 are
recommended filler metals for these joints. Using these materials avoids the excessive
weld hardness resulting from carbon steel or low alloy filler metal. High hardness welds
are susceptible to hydrogen sulfide cracking.

When the service temperature exceeds 600F, Inco 182 is the preferred filler material for
welds between stainless steel and carbon or low alloy steel. Its thermal expansion
behavior is intermediate to that of the two other materials. The intermediate thermal
expansion reduces thermal stresses across the joint and lessens the possibility of
thermal fatigue. Also, carbon diffusion from the carbon or low alloy base metal to the
high Cr weld metal at service temperatures above 900F can result in a phenomenon
known as dissimilar metal weld cracking. Again, the Inco 182 filler metal helps minimize
this.

The postweld heat treatments required for carbon and low alloy steels normally sensitize
stainless steels other than the low carbon L grades (304L, 361L, etc.). You can avoid
this problem by buttering the ferritic component with Inco 182. In this process, several
layers of Inco 182 are deposited on the weld end of the carbon or low alloy steel to
produce a minimum thickness of 3/16-inch after machining. The part is then given the
required postweld heat treatment and the weld is completed between the stainless and
the Inco 182 without subsequent PWHT.

When making carbon steel to chrome-moly welds, we must not exceed 1325 (the lower
critical temperature for carbon steel). This may mean the chrome-moly weld will be
harder than ideal. Contact your materials engineer or ERTC for help in such situations.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-13
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

2.3 Welding Issues

1) True or False: E7018 is the right electrode to use to weld 304 SS to carbon steel.

2) Whats a possible consequence of spot stress relieving an FCC reactor weld repair?

3) For which steels should preheat be maintained until PWHT begins?

4) The hardness limit for carbon steel welds is _____ Brinnel and for Cr-Mo steels is
_____ Brinnel.

Confidential
Copyright 2003 by ChevronTexaco
2-14 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

2.4 HIGH TEMPERATURE GRAPHITIZATION

In ordinary carbon steel, the carbon exists largely as Graphitization is rarely serious
iron carbide (Fe3C). When the steel is exposed to very except in the heat-affected zones
high temperatures, the iron carbide decomposes to of welds. It is more serious in
form ferrite (iron) and graphite (carbon), a pipe than it is in plate.
phenomenon called graphitization. Graphitization
can have a serious effect on the properties of the metal, depending on where the
graphite forms and how the particles are shaped.

Graphitization is rarely serious except in the heat-affected zones of welds. It is more


serious in pipe than it is in plate. In plate, the graphite forms near-spherical globules,
which often leave the plate properties unaffected. In pipe, however, the graphite can
form a chain-like array in the weld heat-affected zone, creating a plane of weakness that
can lead to sudden failure under stress or impact.

To minimize the risk of graphitization, we normally limit carbon steel to a maximum


temperature of 800F and limit carbon - 1/2 Moly steel, which performs better, to a
maximum of 850F. Chrome-moly steels such as 1-1/4 Cr - 1/2 Mo do not graphitize;
chromium contents of 1/2% or greater provide immunity.

These temperature limits are typically applied to pressure-containing parts. For vessel
internals, we can risk graphitization, because a fracture will not result in a leak
especially if the component in question is constructed of plate rather than pipe. For
example, reactor cyclones are made of carbon steel. The risk and consequences of
graphitization of the carbon steel are justified because of the ability to perform repairs
without PWHT.

Key points to remember about graphitization:

Dont worry about graphitization below 800F.


Consider the possibility of graphitization when selecting materials for FCC
Unitsbut only in carbon steel pressure-containing parts operating over 800F.

For carbon steel, the phenomenon takes years to occur at temperatures of 800-950F,
and at 950-1050F it could occur in a matter of months. Extended periods at
temperatures above about 1050F can soften carbon steel and reduce its strength below
acceptable limits.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-15
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

2.4 High Temperature Graphitization

1) To prevent graphitization limit carbon steel to _____ F.

2) What element do we add to steel to prevent graphitization?

3) True or False: Graphitization is primarily a concern in the heat-affected zones of


pipe welds.

Confidential
Copyright 2003 by ChevronTexaco
2-16 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

2.5 HIGH TEMPERATURE CREEP

Metals stretch when subjected to stress. At low temperatures, time isnt a factor;
stressed below the yield point, the metal springs back elastically to its original size when
the stress is removed. Stressed above its yield point, the material deforms permanently.
If the stress remains constant, however, there is no further deformation.

At high temperatures, on the other hand, applying stress causes the metal or alloy to
stretch and to continue doing so as long as the load is applied. This phenomenon is
called creep, and it eventually causes the material to fail.

Metal and alloy information usually includes either creep rate or amount of stress
causing rupture in a given time period (either 10,000 hours or 100,000 hours in most
cases). When designing equipment for high temperature service, we use these
measures as a guide to allowable stresses. When designing equipment for lower
temperature service where creep isnt a factor, we base allowable stresses on yield or
ultimate strength of the material.

Furnace tubes provide an example of high-temperature design based on creep. Our


usual basis for furnace tube design is 1% elongation in 100,000 hours. Most tube
materials stretch far more than 1% before failure occurs (HK-40 is an exception). We
use 5% stretch as the maximum allowable, beyond which the tube should be retired or
have its remaining life carefully analyzed. The 5% stretch limit is also applicable to
components other than furnace tubes.

Much equipment in an FCC Unit operates in the creep rangehot-wall reactor vessels
(1-1/4 Cr), carbon steel reactor cyclones, and stainless steel regenerator cyclones, for
example. Internally-insulated carbon steel reactors and regenerator shells do not
operate in the creep range. Creep becomes significant in carbon steel above 800F,
1-1/4 Cr above 900F, and 300 Series stainless steels above 1100F.

Small increases in temperature can affect creep life Small increases in temperature
significantly. A regenerator part designed to have a can affect creep life significantly.
10-year life at a temperature of 1300F, for example,
would fare differently at different temperatures:

1100F...............................1900 Years
1150F...............................400 Years
1200F ..............................100 Years
1250F ..............................30 Years
1300F ..............................10 Years
1350F ..............................3 Years
1400F ..............................1 Year

At 1500F, the life of the component would be about 1 month. A rough rule of thumb is
that within the creep range, a 50 temperature change affects the rupture life by 34x.

In regenerators or cold-wall reactors, creep failures can occur on the shells if the
insulating refractory fails. Hot spots or bulging indicate refractory failure. External

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-17
permission of ERTC
Fluid Catalytic Cracking Process August 2003

refractory-lined patches or unlined steam boxes have been applied in many cases as
temporary repairs when refractory failures have occurred.

Reactor and regenerator internals, which cannot be insulated, operate in the creep
range. Because much of the stress on internal members is due to forces generated by
thermal expansion, failure prediction has been largely a matter of trial and error in the
past. Modern stress-analysis techniques can remedy this problem.

Figure 2.5-A shows a typical curve of creep as a function of time. Stage 1 creep lasts a
short time and generally does not occur in stainless steels. Most creep occurs in
Stage 2. At Stage 3, deformation occurs quickly and failure is imminent.

Figure 2.5-A. Typical tensile creep curve.

1 Cr and 1-1/4 Cr equipment operating above 900F can also suffer from a phenomenon
known as creep embrittlement. This is essentially low creep strength and ductility in
the heat affected zone of welds, which can result in creep failures with little apparent
macro deformation. This affects higher strength Class 2 grades more dramatically, and
is one reason why we prefer the lower strength Class 1 grades for most equipment,
including FCC reactors. See Section 2.8 for more detail.

Confidential
Copyright 2003 by ChevronTexaco
2-18 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

HOMEWORK QUESTIONS

2.5 High Temperature Creep

1) True or False: For equipment already operating in the creep range, temperature
increases of only 50-100F will likely have little affect on creep life.

2) List some FCC components which operate in the creep range.

3) If fissures are visible in a component operating in the creep range, is there much
creep life left?

4) True or False: Any measurable stretch in a component is an indicator that failure is


imminent.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-19
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
2-20 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

2.6 BRITTLE FRACTURE AND AUTOREFRIGERATION

Background

When designing or testing metallic vessels and piping, we assume that the metals are
ductile, which means that if the metal is stressed beyond its ability to deform elastically,
it deforms plastically. It elongates or increases in size as it decreases in wall thickness,
failing after considerable noticeable deformation. The assumption is incorrect for cast
irons, for certain metals that undergo a ductile-brittle transformation at low temperatures,
and for metals that embrittle when exposed to high temperatures. Cast irons are brittle
at all temperatures and are rarely used for pressure-containing parts in refinery service.

Many common metals used in petroleum refineriesmost 300 Series stainless steels
(304, 316, 321, 347, etc.), copper alloys, nickel alloys, and aluminum alloysdo not
undergo a ductile-brittle transition at low temperatures. If 300 Series stainless steels
contain significant amounts of ferrite, they can lose some toughness at low
temperatures. At low temperatures, carbon and low alloy steels (also 12 Cr steels)
become brittle; they fail without plastic deformation at stresses at or below the elastic
yield strength.

Figures 2.6-A and B show the abrupt change in fracture toughness with temperature,
as measured in the Charpy test. Carbon steel equipment needs to be above its 15 ft-lb
transition temperature in order to be safe from brittle fracture.

Figure 2.6-A. Charpy V-notch energy absorption of A36 steel.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-21
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Figure 2.6-B. Charpy V-notch energy


absorption of A516 Grade 70 steel.

Factors Affecting Brittle Fracture

The occurrence of brittle fracture of an engineering structure such as a pressure vessel


depends on three factors:

Presence of a flaw.
Applied tensile stress.
Material properties.

The interdependence of these factors is shown in Figure 2.6-C.

Per ASME code, brittle fracture is not considered a problem when the applied stress is
less than 8 ksi. Because the presence of defects is significant in determining whether
brittle fracture will occur, it is important to identify and evaluate flaws and defects.

The effect of temperature on toughness of steel is shown in Figures 2.6-A, B, and C.


Thickness also affects properties-the metallurgical microstructure of thicker materials is
not as resistant to brittle fracture as thinner sections of the same material. This is
because thick materials have less enhancement of properties from forming and heating.
In addition, thickness is related to the effect of stress; thicker sections have more
inherent restraint. Conversely, brittle fracture is not usually considered a problem for
steel thickness below about one-tenth inch.

Confidential
Copyright 2003 by ChevronTexaco
2-22 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

Figure 2.6-C. Interdependence of


factors affecting brittle fracture. ASME Code and
Industry Approach
to Brittle Fracture

Both pressure vessel


Susceptible Stress
and piping codes
Material Intensity
require impact
testing of specific
BF materials if the metal
temperature is below
a certain point.
Because pressure
Identification and Evaluation of Flaw
vessels are often
Flaws is an Important Part of thicker than piping,
Avoiding Brittle Fracture the major concern
about brittle fracture
is with vessels.
Figure 2.6-D is a
schematic depiction of the type of curves provided by the ASME Code and API RP 579
(see below). These curves show which materials can be safely used at various
temperatures without any special impact test requirements. Note that both the type of
material and the material thickness is important.

Most Plate Steels

Most Forgings,
Piping, and Fittings

Example Only
See the Corrosion
Prevention &
Metallurgy Manual for
Actual Curves.

Figure 2.6-D. Similar to UCS-66 from


ASME Section VIII, Division 1.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-23
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Minimum Design Metal Temperatures,


Minimum Acceptable Temperatures,
Minimum Pressurizing Temperatures

Currently the ASME Code only relates to fabrication of new vessels or piping. A new
API document, RP 579 Recommended Practice for Fitness for Service, contains an
industry standard for brittle fracture assessments of existing equipment. RP 579 uses
the same methodology as the Code, but some of the terminology differs, and it is
intended to apply to existing equipment.

The Code uses the Minimum Design Metal Temperature (MDMT) to designate the
lowest temperature at which the equipment can safely operate under significant stress
(more than 8 ksi). RP 579 uses the term Minimum Acceptable Temperature to mean
essentially the same thing. For years ChevronTexaco has used the term Minimum
Pressurizing Temperature, or MPT.

The following table shows the maximum pressure that may be applied to pressure
vessels when they are below the MDMT or MPT.

ASME Design Code Year Built Maximum Pressure Below MDMT


Section VIII, Div. 1 Pre-1999 40% Max. Allowable Working Pressure
Section VIII, Div. 1 1999 & Later 35% Max. Allowable Working Pressure
Section VIII, Div. 2 All 25% Max. Allowable Working Pressure

Embrittlement Mechanisms Can Affect the MDMTs of Alloy Equipment

Any time we encounter an embrittling mechanism which can shift the ductile to brittle
transition temperature of a material, we must carefully consider what is the appropriate
Minimum Acceptable Temperature (or Minimum Pressurizing Temperature). For
pressure vessel steels, the most relevant embrittling mechanisms both affect low Cr
steels:

Temper embrittlement.
Carbide precipitation embrittlement.

Temper embrittlement is caused by the diffusion of tramp elements such as P, Sn, Sb,
and As to the metal grain boundaries. This results in an upward shift in the ductile-to-
brittle transition temperature. The effect is greatest in early generation (pre-1970s) high
strength 2-1/4 Cr steels, where the transition temperature shift can be as much as 300F
if they are operated above 650F. The effect is much less severe in 1 Cr and 1-1/4 Cr
steels and in low strength Class 1 materials. 1 Cr and 1-1/4 Cr steels also suffer from a
carbide precipitation embrittlement phenomenon if operated for extended periods above
900F.

For these reasons ChevronTexacos practice is to assign MDMTs for 1 Cr and 1-1/4 Cr
equipment as follows:

For all equipment <3/4 thick, set the MDMT at 50F.


For equipment 3/4 thick, select the MDMT from either Curve A or Curve B of
UCS 66, as follows:

Confidential
Copyright 2003 by ChevronTexaco
2-24 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

If operated above 750F, use Curve A.


If operated less than 750F (rare), then use Curve A for annealed and for low
strength Class 1 steel, and use Curve B for normalized and tempered or for
high strength Class 2 steel.

For 2-1/4 Cr steel, consult with an experienced metallurgist, and Section 300 of the
Corrosion Prevention Manual.

Hydrotesting Warrants Special Precautions

Many of the brittle fractures experienced by the industry have occurred during hydrotest.
ASME Code recommends that the mean metal temperature of Division 1 equipment
during hydrotest be at least 30F warmer than the MDMT for the vessel. For Code
Division 2 equipment, the added 30F is required. This added measure of safety is due
to the fact that the hydrotest exposes the vessel to higher than design pressure (1.5
times the MAWP for Section VIII, Division 1 vessels, and 1.25 times Design Pressure for
Division 2 vessels), during which the risk of brittle fracture is increased. The mean metal
temperature is typically controlled by controlling the temperature of the hydrotest water.

ChevronTexaco has generally not added the + 30F margin to the hydrotest temperature
if the MDMT has been established directly from the UCS-66 curves. However, our
practice is to apply the 30F when the MDMT is established by either of the following
methods:

Grandfathering based on past successful service experience (API RP 579,


Method 2A).
Credit for reduced operating stress or pressure (API RP 579, Method 2C).

If the hydrotest water is required to be greater than 120F, then it is important to take
precautions to prevent burns to personnel in the event of a leak or rupture.

Autorefrigeration

Heat converts liquid to vapor. The heat of a stove, for example, boils water in a tea
kettle. Some liquids (including water) vaporize at atmospheric temperature, the heat
required coming from the liquid itself. During this process, the liquid cools, a
phenomenon called autorefrigeration. If the liquid has low vapor pressure, as water
does, the effect is minimal. A dish of water left on the kitchen sink cools itself to a few
degrees below the temperature of the surrounding atmosphere. On the other hand,
liquids of high vapor pressure, such as ethane, propane, butane, and anhydrous
ammonia, cool to very low temperatures by autorefrigeration.

Fluids with this property can be used as Autorefrigeration in the wrong


refrigerants. Unfortunately, autorefrigeration in the place at the wrong time can cause
wrong place at the wrong time can cause materials materials failures through low
failures through low temperature embrittlement. temperature embrittlement.
Our goal is to construct piping and equipment out of steels that remain ductile under all
possible operating conditions. We must consider not only normal operating temperature,
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-25
permission of ERTC
Fluid Catalytic Cracking Process August 2003

but also the possible cooling effects of autorefrigeration. The autorefrigeration


temperature is defined as the temperature the contents of the vessel would reach if the
vessel were depressured to 40% of its MAWP (for ASME VIII, DN. 1 vessels). Get
assistance from a process engineer to make this calculation. The autorefrigeration
temperature need not always be used as the minimum design metal temperature,
because the loss of pressure that causes autorefrigeration also reduces stresses on the
vessel. However, failures have occurred where equipment was re-pressured while still
cold from autorefrigeration. The steel had adequate toughness under the lower stress
conditions during depressuring but not when full pressure was applied.

Clearly, autorefrigeration is a concern only when handling liquids of high vapor pressure.
For practical purposes, these liquids include ethane, propane, and butane, as well as the
other C2, C3, and C4 hydrocarbon liquids. We need not consider autorefrigeration when
handling these compounds as vapors, because vapors do not evaporate.

In refinery plants, deethanizers, depropanizers, and debutanizers must be evaluated for


potential autorefrigeration problems. Equipment must be constructed of steel grades
that remain ductile under autorefrigeration (depressurizing) conditions. Normalized
A516 or A537 is typically used.

Figure 2.6-E. Isomax TKC steam


generator brittle fracture during hydrotest.

Lesson From a Brittle Failure of a ChevronTexaco Vessel

Failures during hydrostatic test periodically occur when there is a lack of awareness of
the hazard of brittle fracture. An example is the failure of a carbon-1/2 Mo steel heat
exchanger channel during routine hydrostatic test (see Figure 2.6-E). This vessel had

Confidential
Copyright 2003 by ChevronTexaco
2-26 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

been in service for 18 years. The failure took place at 50F and 3400 psig (test pressure
was to be 4800 psig). Impact testing was not done at the time of purchase. The Charpy
V-notch impact strength of a sample from the failed material was 3 ft-lb. Fortunately, no
one was injured; however, start-up of the plant was delayed while the channel was
repaired. The lesson to be learned is that the potential for brittle failure should be
carefully assessed before hydrostatic testing is performed. For more information on
preventing brittle fractures, refer to Section 520 of the ChevronTexaco Pressure Vessel
Manual.

HOMEWORK QUESTIONS

2.6 Brittle Fracture and Autorefrigeration

1) Whats the MPT for 4 thick A516-60 in a situation where vessel records contain no
information on the heat treatment of the material?

2) True or False: Setting MPTs per Code curves takes into account metallurgical
embrittlement that occurs in service.

3) True or False: You should always check the MPT of a vessel before authorizing a
hydrotest.

4) The occurrence of brittle fracture depends on three factors what are these factors?

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-27
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
2-28 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

2.7 FATIGUE

Background

Fatigue is the most common type of mechanical failure of rotating equipment in a


refinery. It occurs when cyclic stress causes a crack to initiate and then propagate.
Fatigue is a ductile fracture process at the root of the notch. When a fatigue crack
propagates, a material with higher notch toughness can tolerate a larger crack before
brittle fracture than a material with lower notch toughness.
Figure 2.7-A. Beach marks
The source of fatigue on a fatigue fracture face.
stress can be
mechanical (e.g.,
vibration of a line
attached to a
compressor) or
thermal (e.g., cyclic
heating of a restrained
part). Fatigue is
usually divided into
two typeshigh-
stress/low-cycle
(<~103 cycles) and
low-stress/high-cycle
(>~103 cycles).
Fracture faces of low-
stress/high-cycle
fatigue failures often
have beach marks
that indicate the later
crack propagation
cycles and a
featureless area
where the part could
no longer support the
load and the final
failure occurred. Figure 2.7-A shows a fatigue failure exhibiting beach marks. This
appearance has been erroneously called crystallizing. (All metals are made of crystals;
they cant crystallize because they are already crystalline.)

Figure 2.7-B illustrates the appearance of round bars that failed in bending under
various loading conditions. Under the optical microscope, fatigue usually appears as
single cracks without branching. Under the electron microscope, you can see fatigue
striations on the fracture face representing each cycle. The characteristic striations are
used to establish that a failure is indeed due to fatigue.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-29
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Figure 2.7-B. Fracture appearances


of fatigue failures in bending.

Figure 2.7-C. Schematic illustration of the fatigue


Endurance and behavior of ferrous and nonferrous alloys.
Fatigue Limit

Carbon, low alloy


steels, and
titanium have
endurance
limitsa cyclic
stress below
which fatigue
failure does not
occur. The limit
is about half the
tensile strength.
Conversely, 300
Series stainless
steels,
aluminum,
copper, and
nickel alloys do
not have an
endurance limit.
This means that given enough cycles, such materials eventually fail at any cyclic stress
range. You can use a material that does not have an endurance limit under cyclic
loading conditions only if the number of cycles is limited to test data indications of what it
can withstand without fracture. This is called the fatigue limit. Figure 2.7-C is a

Confidential
Copyright 2003 by ChevronTexaco
2-30 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

schematic illustration of fatigue curves on a plot of stress versus number of cycles,


showing both fatigue limits and endurance limits.

Factors Affecting Fatigue Life

Stress concentration is an important factor in fatigue. Notches at sharp corners (such as


shaft keyways), rough weld surfaces, and so on greatly shorten fatigue life. Conversely,
using generous radii in corners and blend grinding to remove sharp corners at weld toes
can greatly increase fatigue life. The fact that the fatigue limit is about half the tensile
strength would indicate that the higher the strength, the higher the fatigue strength. As
you can see from Figure 2.7-D, this is true only for polished specimens. Because most
structures have notches, there is rarely any advantage The best solution to fatigue is
in using a higher strength material to solve a fatigue proper design. Fatigue is not a
problem. The best solution to fatigue is proper design. metallurgical problem.
Fatigue is not a metallurgical problem.

Figure 2.7-D. Relationship


between the endurance limit and
ultimate tensile strength of
various steels.

Stress relieving a structure does not usually affect fatigue life; however, shot peening a
part to put residual compressive stresses on the surface increases fatigue life.

Following are common locations where fatigue failures occur in refineries:

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-31
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Heat Exchangers

Tube vibration as a result of flow


patterns is a common cause of Tube-Sheet
fatigue failures. Many such Tube Failure
failures have occurred on alloy
tubes, which are usually Vibrating Tube
purchased with thinner wall
thickness. Figure 2.7-E shows a
typical location where failures
initiate at a tube-to-tube sheet
joint. Fatigue failures, although
rare, can be found in the middle
between two baffles. There are
also typical patterns on the tube
sheet layout depending on the
Rolled Tube-Joint
baffle location, etc., where fatigue Failure
failures occur. Figure 2.7-D. Typical failure
regions at the tube-to-tube sheet
Branch Connections joint for vibrating tubes.

Failure of valved branch connections on large lines is common in refineries. For


this reason, such connections should be designed properly and regularly
inspected.

Reciprocating Pumps and Compressors

Reciprocating pumps and compressors are significant sources of alternating


stresses and, therefore, a common area to find fatigue failures.

FCC Reaction Mix Lines

In FCCs, thermal fatigue is more prevalent than in In FCCs, thermal fatigue is


most processes. Reaction mix lines for example, more prevalent than in most
have fatigue cracked, especially at miters. processes.
Examples include Richmond (1992), Pascagoula
(1999), and Hawaii (2003). The source of the fatigue stress is the differential
thermal growth between the reactor overhead and the fractionator inlet nozzle. A
high stress is placed on the mix line each time the reactor temperature is cycled.
After sufficient cycles, fatigue cracks initiate and grow.

Confidential
Copyright 2003 by ChevronTexaco
2-32 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

HOMEWORK QUESTIONS

2.7 Fatigue

1) The characteristic markings often visible on fatigue fracture surfaces are often called
__________ marks.

2) Name some potential locations for fatigue failures in FCC plants.

3) True or False: Fatigue problems are usually best solved by changes in metallurgy.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-33
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
2-34 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Metallurgical and Refractory Considerations in FCC Plants

2.8 885 EMBRITTLEMENT AND


CREEP EMBRITTLEMENT

885 Embrittlement

400 Series stainless steels are subject to embrittlement in the 700-1000F range. It is
called 885F embrittlement because it occurs most rapidly at this temperature. As with
sigma phase, it is also caused by the precipitation of a phase (in combination with the
precipitation of carbonitrides). Most 300 Series stainless steels do not suffer from 885F
embrittlement; however, the ferrite phase in weld metal and castings and in duplex
stainless steels is subject to 885F embrittlement. The higher the chromium content, the
more susceptible the alloy. For example, Type 410S stainless steel is marginally
susceptible, Type 405 is slightly more susceptible, and 430 is very susceptible. 885F
embrittlement prevents the use of 400 Series SS in many pressure-containing, high-
temperature environments.

Creep Embrittlement

Creep embrittlement is an entirely different phenomenon. It affects only 1 Cr and 1-1/4


Cr steels, and results in the significant lowering of creep rupture ductility, in the
temperature range of about 9001100F. Occurring in 1 Cr and 1-1/4 Cr steels, the
effect is to cause cracks to develop with little or no bulging or deformation of the metal,
especially at notches or other stress raisers. The phenomenon is related to grain growth
and carbide precipitation in the weld heat-affected zone. It has only a small effect on
fracture toughness at room temperature. The phenomenon has been responsible for a
number of cracking incidents around 1 and 1-1/4 Cr - 1/2 Mo nozzles in catalytic
reformers.

The lower strength versions of the 1 Cr and 1-1/4 Cr materials are less susceptible to
creep embrittlement (as well as potential toughness problems). It is always a good idea
to consult with a materials engineer when dealing with alloys, but especially seek help if
you must deal with the higher strength (e.g., class 2) alloys. We avoid using the higher
strength 1 Cr or 1-1/4 Cr alloys when possible.

There have only been a few instances of creep embrittlement in 1-1/4 Cr FCC
components, perhaps because they are not highly stressed. If our 1-1/4 Cr reactors
were designed to Section VIII, Division 2 (i.e., with higher allowable stresses) instead of
Division 1, then creep embrittlement would probably be more of an issue.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
2-35
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

2.8 885 Embrittlement and Creep Embrittlement

1) True or False: 885 embrittlement does not occur below 800F or above 900F.

2) True or False: 885 embrittlement is primarily a concern with the 400 Series stainless
steels.

3) True or False: No 1-1/4 Cr components in FCCs operate at the temperature ranges


where creep embrittlement is a concern.

Confidential
Copyright 2003 by ChevronTexaco
2-36 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

SECTION 3 CORROSION AND METALLURGICAL


DETERIORATION IN FCC PLANTS

3.0 Overview: Corrosion and Metallurgical Deterioration in FCC Plants............ 3-1

3.1 FCC Feed Piping.................................................................................................. 3-3

3.2 The FCC Reactor ................................................................................................. 3-5

3.3 The FCC Regenerator ....................................................................................... 3-11

3.4 Catalyst Piping .................................................................................................. 3-15

3.5 Reaction Mix Line.............................................................................................. 3-17

3.6 Main Fractionator .............................................................................................. 3-19

3.7 Flue Gas Coolers............................................................................................... 3-21

3.8 Fractionator Overhead Condensers ................................................................ 3-23

3.9 Main Fractionator Bottoms System................................................................. 3-27

3.10 Main Fractionator Overhead System............................................................... 3-29

3.11 Cooling Water System ...................................................................................... 3-33

3.12 Intercoolers and Aftercoolers .......................................................................... 3-35

3.13 Gas Recovery Section ...................................................................................... 3-37

3.14 Expansion Joints............................................................................................... 3-41

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

3.0 OVERVIEW: CORROSION AND


METALLURGICAL DETERIORATION IN FCC PLANTS

This section of the training manual leads you through the FCC plant and discusses the
corrosion and metallurgical deterioration that typically occurs in specific FCC
components. Sections 1 and 2 of this manual cover the phenomena you should
understand before you begin this section on equipment deterioration.

Get a copy of a PFD for your plant to help guide you as you read this section of the
manual. Make notes on your PFD to assist you with your inspection planning activities.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-1
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
3-2 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

3.1 FCC FEED PIPING

Offplot piping from other process units to the FCC and onplot lines from the plot limit to
the reactor are typically constructed of carbon steel. We rarely see significant corrosion
in these lines.

Only the combination of hot feed (over 350F) and a high Rates of attack will be highest
neutralization number (TAN) in the oil need concern us in areas of turbulence, such
here. Under these conditions, naphthenic acid corrosion as piping bends, control valve
is a possibility. Fired heaters, which preheat the FCC manifolds, and pumps.
feed, provide the temperatures needed for corrosion.
Currently, only Richmond and El Segundo do not use fired heaters for feed preheat. We
discussed the characteristics of naphthenic acid corrosion in an earlier section of this
document. Such corrosion is velocity-sensitive. Rates of attack will be highest in areas
of turbulence, such as piping bends, control valve manifolds, and pumps.

As mentioned earlier, naphthenic acid corrosion can be prevented by upgrading to


Type 316 steel for castings and Type 316L steel for piping and other items requiring
fabrication by welding.

Refineries that run naphthenic crudes and do not hydrotreat the FCC feed must be
checked periodically for the Total Acid Number (TAN) of the plant feed. A TAN in
excess of 1.5 mg KOH/g indicates impending naphthenic acid corrosion.

HOMEWORK QUESTIONS

3.1 FCC Feed Piping

1) True or False. Non-hydrotreated FCC plant feeds with TANs in excess of 1.5 mg
KOH/g will always promote corrosion regardless of their temperature.

2) Refineries that run __________ crudes and do NOT __________ their FCC feed
must periodically check the TAN of the plant feed.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-3
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
3-4 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

3.2 THE FCC REACTOR

Several different Fluid Catalytic Cracking processes have been used within
ChevronTexaco. Traditionally, most of our units have been the Esso Model IV type, with
side-by-side reactor and regenerator. In these plants, the catalyst flows from the reactor
to the regenerator and back through U-bends. Flow is regulated by pressure balance
and catalyst aeration rates. The cracking reaction, not surprisingly, takes place in the
reactor.

ChevronTexaco has been modifying many FCC Units to use the UOP design. In this
design, spent and regenerated catalyst standpipes and risers replace U-bends. Slide
valves control flow: the regenerated catalyst slide valve controls the reactor
temperature, and the spent catalyst slide valve controls the reactor catalyst level. Most
important, however, is that the cracking reaction now takes place in the riser, which
brings both regenerated catalyst and feed into the reactor. The reactor is little more than
a disengaging space to allow separation of catalyst and hydrocarbon vapor. It is smaller
than the old reactors, and the internals differ greatly.

Each
conversion to a
UOP-type
design has
resulted in a
different
configuration;
there is no
such thing as a
standard
configuration in
our FCC Units.
Figures 3.2-A
and 3.2-B
illustrate the
Esso and UOP
designs.

Historically,
reactors have
been divided
into hot wall
and cold wall
designs. The
hot wall
reactors have
shells of 1-1/4
Cr - 1/2 Mo
steel. They
Figure 3.2-A. Typical Esso Model IV. may be

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-5
permission of ERTC
Fluid Catalytic Cracking Process August 2003

refractory lined
entirely or
selectively for
erosion
resistance. In
this case, the
conventional
refractory has
been a single
1 thick layer of
AA-22
supported by
12 chrome
hexmesh. The
amount of wall
lining in our
earlier reactors
varied from
plant to plant.
Some were
entirely lined.
Others had
lining only in
the dense bed
area. We
discovered
later that lining
was
unnecessary in
some areas,
and we
removed it,
leaving the wall
Figure 3.2-B. ChevronTexaco
bare. Burnaby UOP FCC Unit..

Cold wall reactors have carbon steel shells. Because of the lower design temperature
allowed by ASME Code, they have to be internally insulated. Castable refractories with
good insulating characteristics are very soft and easily eroded. But hard, erosion-
resistant refractories are poor insulators. Therefore, we used dual-layer linings in the
early years. This dual-layer lining consisted of an insulating layer of soft refractory
against the shell that was protected from erosion by a 1" thick, hard layer of AA-22,
supported by 12 Cr hexmesh. The hexmesh was held away from the shell by metal
studs that were as thick as the insulating layer. This type of lining proved expensive to
install and difficult to maintain. As repairs become necessary, some of the dual-layer
lining is being replaced with a single, thick layer of intermediate-density castable,
supported by vee anchors rather than hexmesh. This type of lining doesnt provide as
much insulation as the light-weight insulating refractory nor as much resistance to
erosion as the AA-22, but it is generally proving to be satisfactory. The FCC Refractory
Guidelines Report included in the back of this manual provides specific refractory

Confidential
Copyright 2003 by ChevronTexaco
3-6 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

selection guidelines for all FCC components based on ChevronTexaco and industry
experience.

As new reactors are being built to the UOP design, we are using the hot wall design with
a 1-1/4 Cr shell and 1 of erosion-resistant AA-22 lining. Although we believe that some
parts of the reactor shell need no protection against erosion, we accept this requirement
as a part of the standard UOP design package.

As noted above, some of the earlier reactors were built of carbon steel and others of
1-1/4 Cr - 1/2 Mo. The question of which material to use was eventually resolved in
favor of the 1-1/4 Cr. This alloy has some disadvantages, however:

1-1/4 Cr usually requires postweld heat treatment (PWHT) following construction


or repair. We have had several cases in which erosion has perforated the shell
and local patching was required.
Sometimes we must cut into the shell, even remove the top head, to install new
vessel internals. Then we have to decide if and how to perform stress relief.
Heat treating a spot on the shell is not recommended because the metal
surrounding the spot remains cold, restrains the heated area, and prevents free
expansion. A bulge may appear in the shell at the spot being heat treated. A full
encirclement stress relief is recommended because it allows free expansion of
the metal in an upward direction. However, the ring of heated metal may not be
strong enough to support the weight of the top head and its attachments. We
have seen both success and failure in our attempts to heat treat locally.

Clearly, the repair and heat treatment of a 1-1/4 Cr Clearly, the repair and heat
reactor shell requires a careful review. ERTC pressure treatment of a 1-1/4 Cr reactor
vessel and welding experts must be consulted. Based shell requires a careful review.
on recent work by the Edison welding Institute (EWI), ERTC pressure vessel and
ChevronTexacos historical requirement to PWHT all welding experts must be
1-1/4 Cr repair welds may be relaxed in certain consulted.
situations, including some FCC vessel repairs. The
EWI work showed that acceptable, good toughness welds can be made without PWHT,
using special, controlled-deposition techniques. For now, the decision not to PWHT
1-1/4 Cr repairs must be handled on a case-by-case basis. Be sure to consult ERTC on
this issue.

Current practice is to make Typical reactor cyclones in our FCC Units have been
reactor cyclones of 1/2 thick constructed of 1/4 thick carbon steel with an erosion-
carbon steel with AA-22 lining resistant internal lining of AA-22 refractory. The lining
supported by 12 Cr hexmesh. is supported by 12 Cr hexmesh. The carbon steel
cyclones suffer a slow metal loss due to carburization.
In some plants, cyclones have been upgraded to 12 chrome. Although this alloy resists
carburization, it may embrittle at operating temperature, making subsequent
maintenance and repair more difficult. Current practice is to make reactor cyclones of
1/2 thick carbon steel with AA-22 lining supported by 12 Cr hexmesh.

Although carbon steel cyclones may suffer from very slow carburization, the 1-1/4 Cr
reactor shells apparently do not. Chromium in the alloy retards carbon diffusion into the
metal, slowing attack to a negligible rate.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-7
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Grids in the old Model IV units and stripping section baffles are typically constructed of
carbon steel. This material has proven satisfactory.

Although the reactor environment is very high in hydrogen sulfide, we have seen a
surprising absence of H2S corrosion on carbon steel and 1-1/4 Cr. We are not sure why.
Both temperature and H2S concentration lead us to expect relatively high corrosion
rates. Others have suggested that H2S corrosion rates rise with temperatures up to
about 850F and then decrease. Case-history data to support this theory was presented
as long ago as 1974. The common manifestation in FCC Units has been local corrosion
at cool spots in the reactor shell or reaction mix line, where attachment of external
supports caused local heat loss and a moderately depressed local metal temperature.
We suspect that the drop in rates at very high temperatures is due to coke formation and
protection by the coke layer.

Mechanical distress or local erosion causes most reactor Mechanical distress or local
damage. Internals are often complicated in design, erosion causes most reactor
making it difficult to allow them free thermal expansion on damage.
heating and cooling. Cracks or weld defects can lead to
rapid erosion if catalyst is present and there is a pressure differential across the crack or
defect. Catalyst can then flow at high velocity through the defect, quickly enlarging it.
We have also seen reactor shell leaks because catalyst flowing rapidly through the
defect has impinged on the shell.

Cyclones and dip legs characteristically suffer from erosion damage. Repair of cyclone
refractory and base metal and repair of dip legs and trickle valves is required on
occasion.

Our new reactors of UOP design differ so much from one another and from the old
Model IV designs that general discussion of internal materials (other than cyclones) is
difficult. However, carbon steel and low-chrome steels predominate in these designs.
Corrosion is not a significant factor in materials selection here. The factors that need to
be considered are mechanical properties, ease of fabrication, and need for local erosion
resistance.

Confidential
Copyright 2003 by ChevronTexaco
3-8 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

HOMEWORK QUESTIONS

3.2 The FCC Reactor

1) True or False: Cool spots on the reactor shell and reaction mix line are probably
more likely to experience H2S corrosion than those areas seeing full reactor
temperature.

2) Why do we use carbon steel for reactor cyclones even though its subject to
numerous degradation mechanisms including carburization, graphitization, and H2S
corrosion? How thick should our carbon steel cyclones be?

3) True or False: ChevronTexaco and the rest of our industry requires that all weld
repairs to 1-1/4 Cr reactors be stress relieved.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-9
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
3-10 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

3.3 THE FCC REGENERATOR

Our first FCC Unit went on-stream in 1951 (Salt Lake) and the latest in 1963
(Pascagoula). Regenerator temperatures in those days were low, around 1150F, and
demands on construction materials were not as severe as todays. These early plants
were Model IV design. They used a perforated grid, under which the air was introduced.
The spent catalyst riser (SCR) and overflow well (OFW) entered through the bottom of
the vessel and passed upward through the grid. Because thermal expansion
considerations made it impractical to weld SCR and OFW to both regenerator shell and
grid, expansion bellows were used where these lines passed through the grid. To
accommodate thermal-expansion differences between grid and shell, flexible members
(grid seals) were used between shell and grid.

Today, the basic Model IV design remains only in Hawaii. Each regenerator is
essentially unique within the ChevronTexaco FCC family, and each has its own design
and materials problems. Despite the differences, certain basic principles provide a
unifying materials philosophy. The potential problems we must consider are as follows:

Materials exposed to the full regenerator temperature must have adequate


strength to maintain structural integrity at that temperature.
The alloys used must resist high temperature oxidation.
Materials not having sufficient strength or oxidation resistance at full regenerator
temperature must be thermally insulated with refractory linings.
Both metals and refractory linings must be capable of resisting catalyst erosion.
The severity of erosion varies from place to place within the regenerator.
Metals must not suffer unacceptable metallurgical changes that would lead to
early failure.

The following paragraphs illustrate how we accomplish these objectives in the


regenerator.

Regenerator shells are commonly constructed of carbon steel. Internal refractory linings
are used to keep the shell cool enough to avoid an unacceptable loss of strength. These
linings help protect against erosion where needed. They also protect against oxidation,
a factor that has become increasingly important as regenerator temperatures have risen
over the years. The refractory lining must be a good insulator because its primary
purpose is to keep the shell cool. But insulating refractories are soft and easily eroded
by moving catalyst. Erosion-resistant refractories, on the other hand, are poor
insulators. The use of a dual-layer refractory in the early years of FCC solved this
problem. The inner layer was light, soft, and a good thermal insulator. A second
refractory layer, dense, erosion-resistant, and about an inch thick (usually AA-22) was
applied over the first. This erosion-resistant layer was supported by stainless steel
hexmesh, which was held away from the shell by studs as thick as the insulating layer.
In early plants, these studs were carbon steel. As regenerator temperatures rose, we
began experiencing stud failures by oxidation, failures that led to loss of the entire
refractory system. To solve this problem, we switched to stainless steel studs.

As our knowledge of refractory requirements and the refractories themselves improved,


single-layer, intermediate-density refractories became popular for use on regenerator

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-11
permission of ERTC
Fluid Catalytic Cracking Process August 2003

shells. These dont insulate as well as the light insulating refractories, nor do they
provide as good an erosion barrier as AA-22. They do represent a good compromise,
however. Substantial cost savings and ease of application make them a good choice.
The typical system now used on regenerator shells is a single-layer refractory that is
applied by gunning and supported by 304 SS wavy V-studs. The FCC Refractory
Guidelines Report included in the back of this manual provides specific refractory
selection guidelines for all FCC components based on ChevronTexaco and industry
experience.

Grid temperatures are lower than those in the catalyst bed above the grid because
regeneration air introduced below the grid has a cooling effect. Therefore, we can use
lesser alloys for the grid than for some of the other internals. Early plants used grids of
1 Cr or 5 Cr steel. Grid seals were typically Type 405 stainless. Overflow wells and
spent catalyst risers were also low alloy. As regenerator temperatures increased over
time, we gradually upgraded materials to 300 Series stainless steels, first for overflow
wells and risers, and recently for grids.

The newer (UOP) regenerator designs do not have air grids. All internals operate at a
temperature high enough to require the use of stainless steel (304H). Regenerator
design varies from plant to plant. UOP specifications, which we follow and consider to
be generally satisfactory, govern materials selection.

As with other regenerator components, the cyclones in early plants were usually carbon
steel, although carbon-1/2 moly was used in a few cases. These had reasonable life at
regenerator temperatures up to about 1150F but had to be upgraded as operating
temperatures rose. The material of choice today is Type 304H stainless. Cyclones are
internally protected by a 1 thick erosion-resistant refractory liner, almost always AA-22
supported by Type 304 hexmesh. Recently, some plants have used S-bar anchors
instead of hexmesh, especially for repairs. These offer an advantage on curved
surfaces, especially those with a small radius of curvature where bending hexmesh to fit
is difficult. The top few feet of dip legs in the cyclones may be lined with refractory also.

Cyclone refractory is subject to gradual thickness loss due to erosion. Repair by


buttering has not usually proven satisfactory because patches dont stick. It is usually
more economical to replace the entire cyclone than to replace only the hexmesh and
refractory system. Currently we do so when refractory and hexmesh loss reaches 50%
of thickness. If refractory is repaired, remove all damaged refractory and repack the
hexmesh with all new refractory.

Although cyclone hanger rods were often made of lesser alloys in the early years, 300
Series stainlesses have long been the material of choice. Type 347 stainless was often
used in the past, but Type 304H is now recognized as being more reliable.

The selection of the proper grade of austenitic The selection of the proper grade of
(300 Series) stainless steel for regenerator austenitic (300 Series) stainless steel
service is a matter of no small importance. Key for regenerator service is a matter of
characteristics of the various grades of stainless no small importance.
steel are as follows:
All things considered, Type 304H has
proven to be the best choice.
Confidential
Copyright 2003 by ChevronTexaco
3-12 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

Of the common stainless grades, Type 316 has the best high temperature
strength. However, we dont use it because molybdenum in the alloy makes it
susceptible to catastrophic oxidation at high temperature.
Type 304H is most often the material of choice. It has good strength at high
temperature and good oxidation resistance. It does sensitize at high
temperature, however, and can be susceptible to polythionic acid stress
corrosion cracking during shutdowns, especially if the regenerator is water-
washed. The risk of cracking can be minimized by adding soda ash to the wash
water.
Type 304L shares the disadvantages of 304H and also has poor strength at high
temperature.
Type 304 can be used in place of 304H for those applications where strength is
not especially important. The problem with 304 is that high-temperature strength
depends on the carbon content of the alloy; high carbon means high strength.
The carbon content of 304 can be as high as 304H or as low as 304L. We use
304H rather than 304 to assure high-temperature strength, which is essentially
the same thing as buying 304 to a specific minimum carbon content.
Type 347 stainless has the best resistance to sensitization and polythionic acid
cracking of any stainless steel. Its major shortcoming is its susceptibility to sigma
phase embrittlement on prolonged high-temperature exposure.

All things considered, Type 304H has proven to be the best choice. Experience has
shown that polythionic acid cracking failures in the regenerator are rare and that
resistance to sigma phase embrittlement is a more important consideration. At
ChevronTexaco, we dont usually water wash regenerators for dust. If we did, we might
see more polythionic cracking, but probably not enough to change our recommendation.

HOMEWORK QUESTIONS

3.3 The FCC Regenerator

1) True or False: Todays material of choice for regenerator cyclones is 304H.

2) Match the stainless steel at left with the primary reason we dont use it in
regenerators at right.

316 Insufficient high temperature strength


304L Sigma embrittlement
347 Catastrophic oxidation

3) True or False: When repairing erosion-damaged refractory in the regenerator, you


can expect the repair refractory to adhere well to the refactory that has been in
service.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-13
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
3-14 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

3.4 CATALYST PIPING

The original Esso Model IV units used U-shaped lines (U-bends) to carry catalyst from
reactor to regenerator and back again. Some units still have this piping arrangement.
Although the lines contained slide valves, these valves were not used for control.
Instead, catalyst flow was controlled by the pressure balance between vessels.

Our newer UOP designs do not use U-bends. The catalyst flows downward through a
sloped standpipe until the lift medium (air, in the case of the regenerator, and steam in
the case of the reactor) is introduced. The catalyst then flows up in a straight line (riser)
into the vessel. Slide valves in the standpipes are used to control the catalyst flow. The
spent-cat slide valve controls reactor catalyst level and the regenerated-cat slide valve
controls reactor temperature.

In most of our early plants, U-bends were constructed Vibration-cast, single-layer


of carbon steel, with a dual-layer refractory lining refractory with stainless steel
supported by hexmesh. One or two of our smaller needles is the preferred material
plants had bare carbon steel U-bends. These for catalyst lines today, not only
managed to survive during the early days when in plants having U-bends, but
regenerator temperatures were low, but most had to also in the new UOP designs
be upgraded as temperatures increased over the having standpipes and risers
years. As refractory technology improved, the instead of U-bends.
refractory system of choice became the single-layer,
intermediate-density refractory, supported by V-studs. The quality of this lining was
further improved by reinforcing it with stainless steel needles and by applying it by
vibration casting rather than gunning. Vibration-cast, single-layer refractory with
stainless steel needles is the preferred material for catalyst lines today, not only in plants
having U-bends, but also in the new UOP designs having standpipes and risers instead
of U-bends. See the FCC Refractory Guidelines Report in the back of this manual for
specific recommendations.

The U-bend system had slide valves that did not control. Many of the early regenerated-
cat slide valves had cast or wrought Type 304 stainless steel bodies and erosion-
resistant (AA-22) refractory linings on parts exposed to flow. Steam was often used as a
purge to keep catalyst from collecting in the valve body. If the valve body was not
externally insulated to keep it hot, water condensation could form at the end of the valve
remote from flow.

The combination of water and sulfur oxides from the process gas created an aqueous
acidic condition. This acidic condition, along with temperatures high enough to sensitize
stainless steel, often led to polythionic acid stress corrosion cracking of the valve bodies.
Cast stainless steel slide valves generally did not polythionic crack but, depending upon
ferrite content, were susceptible to sigma phase embrittlement. A high ferrite content
(>8%) in the casting increased the rate and extent of embrittlement. Cracking of
stainless steel slide valves, by either polythionic acid or sigma embrittlement, can be
prevented by taking the following steps:

Do not insulate cast stainless slide valves. Heat loss keeps the metal below the
1000F threshold for sigma embrittlement.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-15
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Insulate wrought stainless valves to prevent water condensation.


Purge valves with nitrogen rather than steam.

All problems associated with stainless slide valves can be avoided by using internally
insulated carbon steel or low alloy slide valves instead. This is the currently
recommended practice.

In our newer plants of UOP design, slide valves are used to control. We recommend
using carbon steel slide valves with internal insulating and erosion resistant refractory
lining. Use the hexmesh and AA-22 system to protect those parts subject to erosion.
Pressure drop across the valve is important in determining how long the valves will last
before they fail from erosion. At the 3-4 psi pressure drop used in our current UOP-
designed plants, the typical slide-valve life is 3-5 years. However, Pascagoula is having
success (i.e., minimal erosion) with pressure differentials as high as 46 psi. In any
event, track pressure drop carefully and inspect valves each shutdown to correlate
erosion damage with pressure drop.

Spent-catalyst slide valves in the U-bend systems operate at much lower temperature
than regenerated-cat slide valves. Spent-catalyst slide valves were typically constructed
of a low alloy steel such as 1-1/4 Cr. We used a refractory lining on those internal parts
that required protection against catalyst erosion.

HOMEWORK QUESTIONS

3.4 Catalyst Piping

1) Refractory for catalyst piping should be applied by __________ __________ rather


than gunning.

2) Name some of the mistakes we made which led to polythionic cracking of the
stainless regenerated catalyst slide valves we used to use in our FCC Units.

3) True or False: The higher the pressure drop across the slide valve, the quicker it will
fail from erosion damage.

Confidential
Copyright 2003 by ChevronTexaco
3-16 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

3.5 REACTION MIX LINE

The reaction mix line carries the product of the cracking reaction from the reactor to the
fractionator. We have used a variety of alloys in this service over the years, the choice
depending on our perception of the corrosion or metallurgical problems likely to be
encountered.

Materials used in the past include 1 Cr - 1/2 Mo, 1-1/4 Cr - 1/2 Mo, 5 Cr - 1/2 Mo, and
Type 321 stainless. The assumption that high-temperature hydrogen sulfide corrosion
could occur led to the use of 5 chrome and Type 321 in this service. But later
experience has shown that hot H2S attack is not a problem here and that we can base
our materials selection only on the need for strength and resistance to high temperature
graphitization. The material of choice is now 1-1/4 Cr - 1/2 Mo steel.

As described previously, high temperature graphitization is a metallurgical change that


can occur in carbon steel above 775-800F and in carbon - 1/2 moly steel above 850F.
Damage occurs over long periods of time and is limited to the weld area. Carbides in
the steel gradually change to graphite, which may form a chain-like brittle precipitate,
leading to a plane of weakness and possible brittle fracture. Alloys containing 1%
chromium or more are not subject to graphitization and thus are appropriate for the
reaction mix line.

The term graphitization can be misleading. Do not confuse high-temperature


graphitization of steel with the so-called graphitization (or graphitic corrosion) of cast iron
that occurs in sea water and a few other services. Cast iron contains large amounts of
graphite. When cast iron corrodes, the iron disappears, leaving the graphite behind.
Graphitization of carbon steel in FCC Units or other hot services is a totally different
phenomenon.

In the chapter on reactors, it was noted that H2S Local attack could occur in the reaction
corrosion of carbon steel peaks at around 850F mix line at cool spots where heat is
and drops at higher temperatures. Local attack sucked out by external supports. This
could occur in the reaction mix line at cool spots possibility, however remote, should
where heat is sucked out by external supports. influence the design of piping supports
This possibility, however remote, should and the location of UT/RT inspection
influence the design of piping supports and the points.
location of UT/RT inspection points.

Reaction mix lines are susceptible to fatigue cracking, especially at miters. The source
of stress causing fatigue is the differential thermal growth between the reactor overhead
and the fractionator inlet nozzle. High strains are placed on the piping each time the
reactor is cycled. For new construction, the use of induction bends instead of mitered
bends avoids this problem.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-17
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

3.5 Reaction Mix Line

1) Reaction mix lines are susceptible to __________ cracking, especially at


__________.

2) The steels used for reaction mix lines must contain some __________ to resist
graphitization.

3) True or False: Because of hot H2S corrosion, 5 Cr is now the material of choice for
reaction mix lines.

Confidential
Copyright 2003 by ChevronTexaco
3-18 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

3.6 MAIN FRACTIONATOR

Products of the cracking reaction exit the reactor at temperatures in the 950-1000F
range. They pass, without cooling, into the main fractionator column where contact with
cooler fluids lowers their temperature. Most main fractionators have a bottoms
temperature in the 650-750F range and an overhead temperature of 200-250F.

The typical main fractionator is constructed of carbon steel, with 12 chrome (Type 405,
410, or 410S) cladding and internals where temperatures exceed 550-600F. The
purpose of this cladding is to resist high temperature H2S attack. Corrosion of
12 chrome in the main fractionator has been negligible.

When selecting a base metal for the main fractionator, you A metal temperature above
must consider the possibility of graphitization. 800F at the feed nozzle
Graphitization can occur on carbon steel where normal and immediately adjacent
operating temperature exceeds 800F, but, because it shell should cause concern
requires time to develop, we dont worry about occasional for carbon steel base metal.
temperature excursions. The area of actual concern is
small; temperatures exceed the 800F threshold for
graphitization only in the immediate feed inlet zone of the column. Using 1-1/4 Cr -
1/2 Mo steel base metal in the lower part of the column can provide immunity to
graphitization. Equipment in older plants is not protected in this manner. A metal
temperature above 800F at the feed nozzle and immediately adjacent shell should
cause concern for carbon steel base metal. Note that the presence of 12 Cr cladding
does not protect the underlying metal against graphitization.

Process fluids entering the main fractionator contain catalyst fines, which often cause
local erosion in the column bottoms system, a subject discussed elsewhere. Although
minor local erosion can occur in the lower part of main fractionator column, it is generally
not a serious problem.

The upper part of the column is subject to ammonium chloride corrosion if the vapor-line
temperature is low enough for wet conditions to exist. If conditions are dry, there is no
corrosion. However, salt deposition has occurred in some plants. Chloride salts can
lead to significant corrosion, especially in the upper third of columns with packing. In the
past, plant operators have cleared the column of salt fouling by adding water to the
reflux to take the salt out in the top sidecut. We cant prevent ammonium chloride
corrosion if the column runs wet; the necessary alloys are extremely expensive. One
effective solution is to raise the overhead vapor line temperature to the point at which the
column runs dry.

Wet H2S cracking or carbonate Theoretically, wet H2S cracking or carbonate cracking
cracking can occur in the top of can occur in the top of the main fractionator if the
the main fractionator if the column runs wet. However, we have not seen this
column runs wet. problem in the past.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-19
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

3.6 Main Fractionator

1) Name three corrosion or cracking mechanisms that can occur at the top of the main
fractionator if the column runs wet.

2) Where might graphitization be a concern in the main fractionator?

Confidential
Copyright 2003 by ChevronTexaco
3-20 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

3.7 FLUE GAS COOLERS

Flue gas from the regenerator is cooled in shell and tube heat exchangers by passing
downward through the tubes. Boiler feedwater on the shellside cools the hot flue gas by
generating steam. Shell, tubes, and tube sheet are carbon steel. A thick refractory
lining on the top tube sheet and metal or ceramic ferrules in the top end of the tubes
protect against erosion and overheating.

Rapid tube failures occurred in some of our earlier flue gas coolers. Attack occurred on
the outside of the tube, several inches below the top tube sheet, resulting in a narrow
groove that circled the tube. In some cases, tube life was less than a year.

Detailed investigation showed that two factors were responsible:

Less than perfect water treatment left oxygen in the boiler feedwater. In at least
one plant, oxygen came from an unblanketed feedwater tank that allowed
intermittent oxygen entry. Oxygen analysis had difficulty finding the oxygen, and
oxygen scavengers had difficulty controlling it.
The design of the flue gas cooler was an even more important factor in these
failures. Steam blanketing produced a stagnant liquid level, with several inches
of tube at the top running dry and hot. Corrosion occurred at the liquid level.
Design modifications aimed at preventing steam blanketing eliminated corrosion.
For example, a few plants switched to conical-top tube sheets. In general,
redesign of the shellside internals, especially the baffles, forced boiling water up
against the top tube sheet and eliminated the vapor pocket.

These modifications have largely eliminated the problem of flue-gas cooler corrosion.
Some plants have used a profile radiography procedure, using a source in one tube and
film in others, to inspect for external grooving at the top of the tubes. Other locations
and some of our competitors have used remote field eddy current inspection.

Another potential tube failure mechanism is localized erosion, especially for locations
using slinkies in their tubes to improve heat transfer. Hawaii experienced an erosion
failure which occurred approximately 6 years after installing slinkies in their flue gas
cooler. Both eddy current and IRIS can be useful tools for finding tube damage due to
erosion.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-21
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

3.7 Flue Gas Coolers

1) Design modifications aimed at eliminating __________ __________ have essentially


eliminated corrosion in flue gas coolers.

2) How can we inspect for the tube OD grooving corrosion that has historically occurred
in flue gas coolers?

Confidential
Copyright 2003 by ChevronTexaco
3-22 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

3.8 FRACTIONATOR OVERHEAD CONDENSERS

Fractionator overhead condensers are shell and tube units with cooling water on the
tube side. Shells are carbon steel and generally have a long life. However, they are
subject to sulfide stress cracking, wet H2S cracking, and carbonate cracking. These
phenomena are discussed in those sections of this manual which deal with specific
corrosion media and are discussed further under Fractionator Overhead System.

We have tried a variety of tube materials in fractionator


Although we have tried a
overhead condensers and have had limited success with
variety of tube materials in
conventional low cost tube materials such as carbon steel
fractionator overhead
and admiralty. Were not sure what is causing the
condensers, we have had
corrosion. Is it ammonia corrosion? H2S attack? Acidic
limited success.
ammonium chloride damage? Should we worry about
stress corrosion cracking by ammonia or chlorides? What
are the problems associated with waterside corrosion?

Some of our earlier plants used bimetallic tubes, a tube extruded over a tube. The inner
tube was made of admiralty, to resist cooling water corrosion. The outer tube was
carbon steel, to control attack by ammonia and/or ammonium bisulfide. These tubes did
not perform well, especially in the hot bundles. We now believe that the corrosion here
was caused by the acidic conditions that exist when ammonium chloride is present
under dew point conditions. See Chloride Corrosion for a detailed discussion.

The next generation of materials consisted of carbon steel tubes in the cold bundles and
admiralty tubes in the hot bundles. We thought that carbon steel would resist the highly
alkaline conditions caused by absorption of ammonia in the cold-shell condensing water,
and admiralty would provide enough protection in the hot shells where condensed water
was less alkaline or perhaps even acidic. The pH variation in the overhead system is
due to the combined presence of HCl, ammonia, and H2S in the main fractionator
overhead vapors. As the vapors are cooled to the dew point, the first water to condense
preferentially absorbs the HCl, and this water is acidic. As more water condenses, the
NH3 begins to be absorbed, and pH rapidly rises to an alkaline value. This alkaline
water pulls H2S into solution, making ammonium bisulfide (NH4HS).

These second generation materials, mixed admiralty and carbon steel, were used with
reasonable success at both Richmond and El Segundo for years. Our other FCC Units
got acceptable life from admiralty tubes in all bundles. The admiralty life span was
controlled primarily by the nitrogen content of the plant feed; more nitrogen meant
greater ammonia production in the reactor and higher pH levels in the fractionator
overhead.

Over time, the nitrogen content of feedstocks has risen in most plants, and the life of
admiralty tubes has deteriorated. Tubes have failed from general corrosion as well as
ammonia stress corrosion cracking. Upgrading from admiralty to 70-30 cupro-nickel,
which has better resistance to high pH and is not subject to ammonia SCC, was
somewhat successful. More recent upgrades used either Type 304L stainless steel,
titanium or Duplex stainless steel. Type 304L was successfully used at Richmond where
the cooling water was historically low in chlorides and non-scaling, but the situation at
Richmond is changing with the use of reclaimed water. A high chloride content in the
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-23
permission of ERTC
Fluid Catalytic Cracking Process August 2003

water or deposition of water scale, with resulting concentration of chlorides beneath the
scale, can lead to chloride pitting or stress corrosion cracking. Also, stainless tubes in
the hot bundles can chloride crack due to ammonium chloride under dew point
conditions.

Two of our plants, El Segundo and El Paso, have long used titanium tubes. Titanium is
thought to be immune to corrosion by any combination of corrosives likely to be present
on either process side or cooling water side. Hydriding is a problem of some concern
with titanium. It occurs when hydrogen atoms generated by corrosion reactions enter
the titanium and react to form hydrides. With formation of enough hydrides, the titanium
becomes brittle and susceptible to mechanical failure. Due to galvanic effects, hydriding
is most prevalent in titanium exchangers with non-titanium tubesheets. High
temperatures and pH extremes also accelerate hydriding. Hydrided titanium bundles
require careful handling during down times.

To our knowledge Duplex stainless tubes have not been used in ChevronTexaco FCC
overheads. However, several competitors, including Exxon, have reported success with
Duplex in this service. Duplex is generally intermediate in cost; more expensive than
304L or admiralty but less expensive than titanium.

This discussion of overhead condensers is incomplete without a brief mention of cooling


water corrosion. The following information applies to other water-cooled exchangers in
the FCC Unit.

Copper-based alloys such as admiralty and 70-30 cupro-nickel have been used
successfully to control cooling water corrosion. Carbon steel, used by many
plants inside and outside of ChevronTexaco, requires careful control of the
cooling water treatment. Performance of carbon steel varies widely from one
geographic area to another and from plant to plant within a given geographic
area. Before choosing carbon steel for a new exchanger, the Materials Engineer
must review carefully both local experience and the effectiveness of water
treating.
The austenitic (300 Series) stainlesses have long lives in cooling water service,
provided that the water is low in chloride and that scaling does not occur on the
water side. Duplex stainless steel has better resistance to chloride pitting and
cracking than 300 series stainless steel.
On the process side, titanium resists both alkaline corrosion caused by ammonia
or ammonium bisulfide and acid attack under dew point conditions. Titanium
also has excellent resistance to cooling water corrosion and can be used in high
chloride-containing waters up to 220F for Grade 2 and 350F for Grade 12.

Confidential
Copyright 2003 by ChevronTexaco
3-24 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

HOMEWORK QUESTIONS

3.8 Fractionator Overhead Condensers

1) True or False: As nitrogen in the plant feed increases, admiralty tubes in fractionator
overhead condensers are more likely to corrode.

2) Why would admiralty perform better in the hot overhead bundles versus the cool
ones?

3) Titanium is probably immune to corrosion by any combination of corrosives likely to


be present on either the process side or cooling water side, but __________ is a
problem of some concern.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-25
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
3-26 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

3.9 MAIN FRACTIONATOR BOTTOMS SYSTEM

The hot bottoms stream from the main fractionator exits the column at temperatures
between 650-700F. When temperatures are in this range, materials selection must
consider the possibility of high-temperature H2S corrosion. Additionally, the fractionator
bottoms contain catalyst fines in significant quantities. These solids can cause erosion
in areas of high velocity and can cause plugging in non-flow lines or areas of low
velocity.

Although H2S corrosion is possible when temperatures Corrosion rates are highest
exceed 500F, it is more commonly seen at temperatures at the hot end (the column
above 550F. Corrosion can occur wherever the H2S end) of the bottoms system
concentration exceeds 1 ppm. Corrosion rates are highest and decrease as the process
at the hot end (the column end) of the bottoms system and stream is cooled.
decrease as the process stream is cooled.

High-temperature hydrogen sulfide corrosion can be controlled by using chromium-


containing alloys. A 5% chromium steel is quite adequate for piping in the fractionator
bottoms system. The 12% chromium steels (Types 405, 410, or 410S) as well as the
300 Series stainless steels (Types 304, 321, 347) have excellent resistance; they can be
used for exchanger tubes and cladding on exchanger shells and channels. The choice
between materials is a simple matter of cost and availability. 5 Chrome is available as
pipe and sometimes tubes, but usually not as sheet or plate. The 12 chrome and 18-8
stainlesses cost more than 5 chrome, are available in all forms, and are delivered more
quickly.

Look for erosion in the Fractionator bottoms systems can be eroded by catalyst fines
bottoms pumps, control suspended in the oil. This slurry is erosive at high velocities
valve manifolds, and (over 25-50 fps) and at highly turbulent locations, but not at
check valves. the rather low velocities (5-10 fps) likely to exist in the majority
of the piping. Therefore, we look for erosion in the bottoms
pumps, control valve manifolds, and check valves.

ChevronTexacos Fractionator Bottoms Best Practice (BP) discusses erosion control in


the bottoms systems in detail. This BP is accessible on the ChevronTexaco Intranet
under the Worldwide Refining Best Practice network. For many years, our practice was
to construct the fractionator bottoms pumps of one of the 12 Cr steels (e.g., CA6NM).
Erosion problems eventually led to the use of special coatings on the base metal.
However, special hardfacings are difficult to apply to 12 Cr steels, and weld repair, when
erosion damage eventually occurred, was also very difficult. Because of these
difficulties and because erosion, not corrosion, is the primary damage mechanism for
pumps, our current practice is to use the carbon steel pump cases with an applied
coating of Jet Kote II or equivalent tungsten carbide hard facing. The recommended
impeller and wear ring material is carbon steel, with a boron diffusion coating. Because
we may find better coatings in the future, Analysts should check for progress in this field
before specifying a coating for a new pump.

We have also used Lawrence Pumps to control erosion. These pumps have an erosion-
resistant replaceable liner of high chrome cast iron. They cost about twice as much as
the hard-faced CA6NM pumps but last twice as long. One problem, however, is that the
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-27
permission of ERTC
Fluid Catalytic Cracking Process August 2003

liner is brittle and subject to failure by thermal shock. These pumps also require a very
slow warm-up when being placed in service.

Pump speed affects the rate at which a pump erodes. Current practice is to limit speeds
to 1800 rpm maximum. Higher-speed pumps have short life spans.

12 Chrome trim has generally been satisfactory in the fractionator bottoms system for
valves that do not throttle or control. In those services where better materials are
required, we have used Stellite and tungsten carbide successfully. Control valve
selection is still under study, but recommendations will eventually be incorporated into
our Best Practice document.

HOMEWORK QUESTIONS

3.9 Main Fractionator Bottoms System

1) What are the current Best Practice recommendations for materials for our
fractionator bottoms pumps?

2) What type of trim is used in valves which are used for throttling or control?

Confidential
Copyright 2003 by ChevronTexaco
3-28 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

3.10 MAIN FRACTIONATOR OVERHEAD SYSTEM

The important corrosives in the main fractionator overhead system are ammonia (NH3),
hydrogen sulfide (H2S), carbon dioxide (CO2), cyanides (CN-), and traces of chlorides.
Individually or in combination, they are capable of causing the following types of
corrosion damage:

Ammonium bisulfide corrosion of carbon steel.


Ammonia corrosion of copper alloys.
Ammonia stress corrosion cracking of copper alloys.
Chloride cracking of stainless steels.
Chloride corrosion of carbon steel and stainless.
Sulfide stress cracking of high-strength carbon and low alloy steels.
Wet H2S cracking of carbon steel.
Carbonate stress corrosion cracking of carbon steel.
Cyanide acceleration of all forms of damage to carbon steel.

Each of these has been discussed individually in the section of this manual that deals
with specific corrosives. For the purposes of this discussion, we limit our definition of the
main fractionator overhead system to the fractionator overhead condenser and reflux
drum, compressors, interstage separators, and the compressor intercoolers and
aftercoolers. Also, to keep this chapter to a manageable length, the following topics are
discussed in separate chapters:

Downstream gas recovery system.


Fractionator overhead condensers and inter/aftercoolers.
Ammonia corrosion and stress corrosion cracking of copper alloys. (Because
these types of corrosion occur only on cooler and condenser tubes in the
fractionator overhead system, they are discussed in applicable heat-exchanger
chapters.)
Chloride corrosion and stress corrosion cracking.

The nature and severity of corrosion damage depends to a great extent on the chemical
composition of the main fractionator overhead vapors. We can assume that we always
have H2S. By itself, H2S is only slightly soluble in water and forms a weak acid that is
mildly corrosive. Unfortunately, the FCC fractionator overhead system also contains
ammonia, CO2, cyanides and other compounds that interact with the H2S to greatly
increase its corrosivity.

Ammonia is alkaline, and the H2S and CO2 are acidic in aqueous solution. When we
have condensing water, the NH3 reacts with the H2S and CO2. Ammonia reacts with the
strongest acid first, and when that has been consumed, any remaining ammonia
consumes the next strongest acid. In sequence, the NH3 first reacts with CO2 to form
ammonium bicarbonate, then with H2S to form ammonium bisulfide, and finally with
ammonium bicarbonate to form ammonium carbonate. Bicarbonate is harmless.
Bisulfide can be corrosive and can cause both sulfide stress cracking and wet H2S
cracking. Carbonates cause carbonate stress corrosion cracking. Cyanides are potent
accelerators of all forms of corrosion associated with bisulfide.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-29
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Our studies of carbonate stress corrosion cracking show it Our studies of carbonate
likely to occur only in those plants where nitrogen content stress corrosion cracking
of the feed is high. Conversion of nitrogen to ammonia show it likely to occur only in
results in high pH and enough ammonia to convert those plants where nitrogen
harmless bicarbonate into potentially damaging content of the feed is high.
carbonate. Review of plant data tells us that cracking
occurs at a pH above 9 when carbonate concentration is
over 100 ppm. At a higher carbonate concentration (400+ ppm), cracking may occur at
a pH in the 8-9 range.
Carbonate SCC has
occurred in all our We have seen more carbonate cracking in recent years,
refineries to date except possibly because we use promoted catalysts, which increase
Salt Lake. conversion of CO to CO2. Carbonate SCC has occurred in all
our refineries to date except Salt Lake.

WFMT inspection can detect carbonate cracking in vessels and exchanger shells, and
shear wave UT can detect it in piping. Such inspections should be conducted regularly,
as discussed in ChevronTexacos Inspection Strategy IS-12. Richmonds carbonate
cracking inspection program for piping is included in the back of this manual. Thermal
stress relief can prevent cracking, but the heat treatment must be done carefully;
threshold stress for cracking is very low. New carbon steel piping and equipment for the
main fractionator overhead system must be stress relieved, and PWHT should be used
following any weld repair. The procedure for proper PWHT of carbon steel welds in
carbonate service is detailed in the Welding Gray Manual.

Ammonium bisulfide corrosion of carbon steel can occur in the fractionator overhead
system if bisulfide concentration is high enough. Look for it where bisulfide
concentration of the water phase is high (over 2% by weight) and where fluid velocities
are over 20 fps (bends, control valve manifolds). Ammonium bisulfide solutions also
cause local attack in dead legs. See Inspection Strategy IS-2 for additional details on
this subject. Cyanides accelerate ammonium bisulfide corrosion. Polysulfide injection
neutralizes cyanides and inhibits bisulfide corrosion whether or not cyanides are present.

Ammonium bisulfide solutions also cause sulfide stress


Ammonium bisulfide solutions
cracking and all the various forms of wet H2S cracking
also cause sulfide stress
(hydrogen blistering, HIC, SOHIC). Cyanides cracking and all the various
accelerate such attack, and polysulfide injection retards forms of wet H2S cracking
it. See the preceding discussion of these phenomena. (hydrogen blistering, HIC,
Also review Inspection Strategy IS-17. Current new
SOHIC). Cyanides accelerate
construction recommendations and inspection and
such attack, and polysulfide
repair guidelines for FCC wet H2S service vessels are
injection retards it.
included in the back of this manual. (See MEEs
website for updates.)

Some important facts to remember are these:

Sulfide stress cracking can be prevented by avoiding materials harder than HRC
22 or stronger than 90 ksi yield strength. For welds in ordinary carbon steel, limit
weld metal hardness to 200 Brinell.

Confidential
Copyright 2003 by ChevronTexaco
3-30 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

PWHT, applied in this part of the plant for prevention of carbonate cracking, also
prevents sulfide stress cracking. It does not prevent hydrogen blisters, HIC, or
SOHIC. Coating or cladding are the best ways to prevent these forms of
damage.
WFMT inspection finds sulfide cracks as well as HIC, SOHIC, and cracked
hydrogen blisters.
Hydrogen blisters can be detected visually or by straight beam UT.
Continuous on-stream water wash is a valuable tool for minimizing all forms of
corrosion, cracking, or blistering caused by ammonium bisulfide. This topic is
discussed in more detail later in this chapter. Adding ammonium polysulfide to
the wash water provides additional benefit. Filming amines have also been used
in these systems for control of corrosion, but experience has been mixed.

We have used on-stream water wash for years to control corrosion in the fractionator
overhead system. Water washing is beneficial because it dilutes the corrosives to a
non-corrosive concentration and removes them from the system. Water wash schemes
continue to evolve. In 1985, Richmond experienced blistering upstream of the cyanide
scrubber column. Calculations confirmed that reverse cascading the water from the
interstage system to the overhead condensers caused a build-up of contaminants
(ammonia, hydrogen sulfide, carbon dioxide, and cyanides) and, therefore, increased
corrosion problems in the system. This was not a problem in most FCC Units where the
APS injection was maintained essentially full time to neutralize the cyanides, but
Richmond was having difficulty keeping their APS pumps in operation due to plugging
and mechanical problems.

To counteract this problem, Richmond implemented a parallel cascade system. By


adding another pump off the overhead accumulator, sour water was pumped from the
overhead accumulator forward to the interstage condensers and the cyanide scrubber
column, and also was recirculated directly to the overhead condensers. Fresh APS was
added in two placesin the overhead condenser loop and in the stream going to the
cyanide scrubber and interstage system. Spent APS solution was drawn off at the
interstage separator drum and at the overhead accumulator. There was no recycle of
water from the interstage system to the overhead condensers.

El Segundo continued to use the reverse cascade system. El Paso North (now shut
down), Philadelphia, and Port Arthur 1241 also used reverse cascade flow. Pascagoula
had a hybrid system.

A reverse cascade system is less complicated to operate because the water is easily
pressured back from the high pressure side back to the low pressure side, requiring
fewer pumps. But the reverse cascade design becomes detrimental from an overall
corrosion standpoint if APS injection is not reliable and cyanides build up in the water
wash system.

In systems where polysulfide is not reliably injected, a parallel cascade water wash
system is definitely preferred. We now recommend a parallel cascade arrangement
when designing a new system. Before converting an existing reverse cascade system to
a parallel cascade system, however, the cost consideration of a new pump and re-routed
piping must be taken into account.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-31
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Ammonium polysulfide can be added to the wash water for even better control of
corrosion, cracking, and blistering. Some of our FCC Units use polysulfide. Use of
polysulfide may or may not be justified, depending on the severity of corrosion and
cracking. Where it is done, special requirements regarding dilution-water quality must
be followed. Guidelines for proper handling and storage of APS are included in the back
of this manual.

A discussion of bisulfide corrosion would be incomplete without reference to the


possibility of corrosion at water injection points. This is a topic of vital importance, as the
massive explosion at the Shell Oil refinery in Norco, Louisiana, a few years ago proves.
Key practices to avoid such disasters are as follows:

Use a spray nozzle at the injection point to prevent direct impingement of water
against the pipe wall.
Inject at least 8 pipe diameters upstream of any bends.
Inspect at a minimum of every 3 years.

For additional information on overhead corrosion control see the FCC Overhead
Corrosion Control Guidelines report included in the back of this manual, and also
available on the ChevronTexaco Intranet (see the Intranet for the most current revision).
See Inspection Strategy IS-20, and API 570 Piping Inspection Code for injection point
inspection practices.

HOMEWORK QUESTIONS

3.10 Main Fractionator Overhead System

1) Our FCC overhead water injection points should be at least __________ pipe
diameters upstream of any bends.

2) What problems do cyanides cause in the overhead system?

3) True or False: Brass, stainless steel, and carbon steel can all experience some form
of stress corrosion cracking in the overhead system.

Confidential
Copyright 2003 by ChevronTexaco
3-32 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

3.11 COOLING WATER SYSTEM

All process plants must deal with corrosion caused by cooling water. In FCC Units, the
heat exchanger tube materials commonly used for resistance to water corrosion are
often incompatible with the process fluids. Piping, valves, pumps, and exchanger
channels in cooling water service are almost always carbon steel and have reasonable
life. A number of materials have been used for exchanger tubes with varying degrees of
success:

Carbon steel
Epoxy-coated carbon steel (e.g., Saekaphen)
Copper alloys (usually admiralty, occasionally 70-30 Cu-Ni)
300 Series stainless steels
Duplex stainless steels
Titanium

Exchanger tubes typically require better materials than other water-handling


equipment, for two reasons:

Tubes are thinner and tolerate less corrosion before they reach minimum
acceptable thickness.
Tubes run at high temperatures; generally, corrosion rates increase as the
temperature rises.

Some plants use carbon steel exchanger tubes in cooling water service successfully.
Other plants are less successful with this material. Two factors affect success rate:
quality of available cooling water and past history of successful water treatment. Before
specifying carbon steel for exchanger tubes in cooling water service, check historical
experience with steel in the refinery and in the plant. Also, recognize that carbon steel is
susceptible to failure if water treatment is ignored. Epoxy-coated carbon steel tubes
have good cooling water corrosion resistance as long as the coating is intact. Coating
can be damaged by re-rolling tubes. Therefore, tubes should always be seal-welded
prior to coating. Only straight-through tube bundles can be coated, not U-bends.

Copper alloys such as admiralty and 70-30 cupro-nickel have extremely good resistance
and are unlikely to fail quickly from cooling water corrosion. Cupro-nickel is superior to
admiralty in this regard. However, copper alloys are vulnerable to stock-side corrosion
from ammonia or ammonium bisulfide, and admiralty can stress-corrosion crack in
ammonia.

The austenitic (300 Series) stainless steels also have good Avoid stainless if
resistance to water corrosion, but they can be vulnerable to chloride content of the
pitting and stress corrosion cracking due to chlorides. Of the cooling water is high.
stainlesses, Type 304 is the usual choice; other grades offer
few advantages over 304. Before specifying a 300 Series stainless, check past
experience in the plant and in the refinery. Avoid stainless if chloride content of the
cooling water is high. 50 ppm chloride is a safe limit and up to 500 ppm chloride may be
tolerable if water side velocities are high (>5 fps) and tube skin temperatures are low
(<140F). It is wise to avoid stainless steels if you have water-side scaling caused by

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-33
permission of ERTC
Fluid Catalytic Cracking Process August 2003

the cooling water and conditions in the exchanger, or if water is on the shellside. Tube
wall temperatures under the scale deposits are often high. Chloride-containing water
permeates the scale. On reaching the tube surface, the water evaporates, leaving the
dissolved solids, including the chlorides. The chlorides concentrate under the scale. In
a cooling water containing only 10 ppm chloride, chloride levels can rise to thousands of
parts per million, resulting in chloride cracking of stainless steels.

Duplex stainless steels are seeing increased use in Duplex stainless steels are seeing
cooling water service throughout the industry, increased use in cooling water
although ChevronTexaco has yet to use these service throughout the industry.
materials in FCC overheads. They offer improved They offer improved resistance to
resistance to chloride pitting and cracking versus 300 chloride pitting and cracking
series stainless and cost less than titanium. versus 300 series stainless and
cost less than titanium.
Titanium has achieved wide acceptance in cooling
water service. Its use in FCC fractionator overhead and gas recovery systems has
produced no stock-side corrosion problems to date, although hydriding is still a potential
concern in some cases. Probable corrosion rates of titanium are so low that we can use
extremely thin-walled tubes; 18 or 20 gauge tubes are common, minimum thickness
being determined by mechanical requirements rather than corrosion. (By comparison,
carbon steel tubes are usually 14 gauge, and copper alloy or stainless steels are
16 gauge.) You can avoid potential problems by having a Materials Engineer review
your choice of tube sheet material and technique you
Titanium has achieved wide intend to use for rolling the strong, thin-walled titanium
acceptance in cooling water tubes. Baffle spacings for titanium tubes often need to
service. Its use in FCC be less than for other alloys. Because the thinner
fractionator overhead and gas gauge and lower modulus makes titanium vibrate much
recovery systems has produced more readily than other alloys, the baffle spacing must
no stock-side corrosion be correct to avoid fatigue cracking. See the Heat
problems to date. Exchanger Gray Manual for details.

HOMEWORK QUESTIONS

3.11 Cooling Water System

1) True or False: Copper-nickel is more resistant that admiralty to SCC.

2) __________ spacings for titanium tubes often need to be less than for other alloys.

3) True or False: Epoxy-coated steel exchanger tubes should be seal welded rather
than rolled.

Confidential
Copyright 2003 by ChevronTexaco
3-34 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

3.12 INTERCOOLERS AND AFTERCOOLERS

Compressor intercoolers and aftercoolers have suffered from corrosion in most of our
FCC Units. The corrosives present and the materials used closely parallel those in main
fractionator overhead condensers. Experience varies from plant to plant; the economical
solution to the corrosion problem for one plant does not necessarily fit all plants.

Carbon steel tubes, used in some plants, have two limitations. They may not be suitable
for the local cooling water. More often, however, steel tubes fail from ammonium
bisulfide corrosion. The corrosion is more likely to occur in the intercooler where
bisulfide concentrations are higher than in the aftercooler. Injecting polysulfide, which
acts as a corrosion inhibitor, ahead of the intercooler improves the life of steel.

Admiralty has a spotty performance history in this service. Admiralty has a spotty
If nitrogen content of the plant feedstock is high, performance history in this
admiralty has a very short life. High nitrogen content service. If nitrogen content
results in more ammonia, higher ammonium bisulfide of the plant feedstock is high,
concentrations, and higher pH levels in the inter/ admiralty has a very short
aftercooler systemall factors detrimental to admiralty. life.
The 300 Series stainless steels have good stock-side resistance in this service. There
are two concerns with this class of material:

If water is injected ahead of the intercooler at rates too low to avoid a boil dry
condition on the tube wall and high chloride concentrations on the metal surface,
the result can be acidic chloride corrosion or stress corrosion cracking.
Pitting or chloride cracking may occur on the cooling water side. This problem is
discussed in the chapter on cooling water systems. Cooling water quality and
quantity may change. For example, a change from high quality water to
reclaimed water or a serious local water shortage can have a serious effect on
the life of stainless tubes.

Titanium tubes are a good choice for intercooler or aftercooler service, if less costly
materials are not performing satisfactorily. The unalloyed grade of titanium (Grade 2) is
quite satisfactory here because of the low service temperatures. Thin-walled tubes (18
or 20 gauge) can be used; expected corrosion rates for titanium approach zero. Consult
a Materials Engineer regarding selection of an appropriate tube sheet material and
proper tube-rolling techniques for titanium.

Duplex stainless may also be a good choice for intercooler or aftercooler service.
Competitors have reported success with this material but it hasnt been tried within
ChevronTexaco.

This discussion of intercoolers and aftercoolers has dealt entirely with tube materials.
Steel shells and other exchanger parts are also subject to carbonate cracking, sulfide
cracking, and wet H2S cracking. These phenomena have been discussed at length in
the fractionator overhead system section.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-35
permission of ERTC
Fluid Catalytic Cracking Process August 2003

HOMEWORK QUESTIONS

3.12 Intercoolers and Aftercoolers

1) True or False: The materials and corrosion problems for these exchangers closely
parallel those of the fractionator overhead condensers.

2) True or False: Ammonium bisulfide corrosion of carbon steel is more likely in the
intercoolers than the aftercoolers.

Confidential
Copyright 2003 by ChevronTexaco
3-36 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

3.13 GAS RECOVERY SECTION

The gas recovery section of the plant is subject to the forms of corrosion that we see in
the main fractionator overhead system.

However, rates of corrosion and severity of damage are generally less because of the
lower concentrations of corrosives. Much of the corrosive material is removed with the
draw water from reflux drums and KO pots. Even so, the following forms of metal
damage are of concern in the gas recovery section:

Wet H2S corrosion


Wet H2S cracking
Ammonium bisulfide corrosion
Sulfide stress cracking
Hydrogen blistering
Ammonia corrosion (of copper alloys)

We find corrosion in the gas recovery section only where We find corrosion in the gas
liquid water is present. The major corrosives (ammonia recovery section only where
and hydrogen sulfide) generally follow the same route liquid water is present. The
through the plant as the water. Gas recovery systems major corrosives (ammonia
vary from plant to plant in design. We can identify the and hydrogen sulfide)
areas most likely to experience corrosion by tracing the generally follow the same
route of the corrosives. The severity of attack in these route through the plant as
systems depend upon the extent to which the corrosives the water.
are removed in the main fractionator system, through
such techniques as continuous water wash.

Some columns in the GRU have water draw trays where water accumulates and is
drawn off. These areas often have higher corrosion rates than the rest of the column.

The forms of attack listed above have been discussed previously. Control measures are
as follows:

Reduce hydrogen blistering by using coatings or cladding for construction of


columns and vessels. Removing H2S, NH3, and cyanides by water washing and
polysulfide injection (in the main fractionator overhead system) is also very
helpful.
Control sulfide stress cracking and wet H2S cracking by carefully reviewing
welding procedures to avoid overly-hard weld metal and heat-affected zones.
Control sulfide stress cracking by avoiding the use of overly hard or strong
materials such as hardened 12 chrome tray bolting.
Choose materials for heat exchanger tubes carefully, especially in overhead
condensers. For guidance, check on past experience both in the plant under
consideration as well as in other company plants.

Brittle fracture as a result of autorefrig- Brittle fracture as a result of autorefrigeration


eration can occur in certain items of can occur in certain items of equipment in the
equipment in the gas recovery section. gas recovery section. See the earlier

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-37
permission of ERTC
Fluid Catalytic Cracking Process August 2003

discussion on this subject. Very light liquid hydrocarbons, such as ethane, propane, and
butane, evaporate if the system is depressured. The heat of vaporization comes from
the liquid itself and from the equipment containing it. The result is a rapid drop in
temperature that can chill the equipment below the ductile-to-brittle transition
temperature and lead to brittle fracture. This cooling is especially hazardous in LPG
systems. Ideally, the equipment has been constructed of materials that remain ductile at
autorefrigeration temperatures. However, the Analyst must verify that the plant is safe
by checking materials, establishing their minimum allowable temperatures, and
determining probable autorefrigeration temperatures. If necessary, the Analyst can then
impose operating restrictions when the equipment is depressured.

Corrosion rates of steel within the gas recovery section tend to be low. Equipment may
be dirty; even a little corrosion by wet H2S generates considerable sulfide scale. The
scale may be pyrophoricmay catch fire on exposure to air. Check past experience in
the plant to see if there is any history of fire. Packed columns, operating under wet
conditions, are especially vulnerable if the packing is carbon steel. The large surface
area of the packing can result in enough sulfide scale to destroy the column if the sulfide
catches fire.

Some columns in FCC Units have an unusual and undesirable tray design. Instead of
the usual flat trays sitting on support rings welded to the shell, they have dished trays.
These look roughly like dished pressure vessel heads and are installed in a dished-
downward configuration. They are attached to the shell by welding. Scale and corrosive
material can accumulate in the hidden crevice between shell and tray and accelerate
corrosion. The area is difficult to inspect by any reasonable means.

Thermosyphon reboilers are often used on GRU columns, and can have some unique
problems. These heat exchangers normally have steam on the shellside that heats and
often boils the tubeside process fluid to induce tubeside flow.

Corrosion products from the column can accumulate on the tubeside and result in loss of
heat transfer and under deposit corrosion. On the shellside there are two common
problems. High velocity steam at the inlet can impinge on tubes and cause erosion.
This is prevented by installing deflection devices such as rod baffles or a perforated
impingement plate. CO2 from the steam can also build up in the vapor space of the
shell. It is typically concentrated in low flow regions that can exist at the top of the
bundle or under the horizontal baffles. The typical solution is to continuously vent a
small plume of steam from the shell, which allows the CO2 to escape.

Confidential
Copyright 2003 by ChevronTexaco
3-38 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

HOMEWORK QUESTIONS

3.13 Gas Recovery Section

1) True or False: The gas recovery section is where we should be most concerned
about brittle fracture resulting from autorefrigeration.

2) True or False: From a corrosion and inspection standpoint, dished column trays are
preferred to flat trays.

3) To find corrosion in the GRU, follow the flow of __________ through the plant.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-39
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
3-40 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

3.14 EXPANSION JOINTS

Much of the FCC equipment operates at very high temperatures. If thermal expansion of
the metal components is not adequately taken into account in the design, high stresses
will develop in the materials. A common way to accommodate thermal expansion is to
include expansion joints in the design, using a system of one or more bellows. In the
past, many of our expansion joint systems used bellows contained in the vessels,
notably the regenerator. When we had tears and other problems with the bellows it
could impact the efficiency of the unit, but rarely would it cause the unit to shutdown.
With the newer generation UOP designs, the expansion joints are an integral part of the
pressure containing system. If there is a leak in these joints it means we are leaking hot
catalyst and hydrocarbon to the atmosphere. Such failures are certainly a safety issue,
and most likely will shutdown the unit.

Expansion joint failures represent one of the greatest threats to the reliability of modern
FCC Unitsboth in ChevronTexaco and the industry. We are developing our
understanding of how the modern large FCC expansion joints work (and dont work!).
We are working to improve the design, maintenance, and monitoring/inspection
practices of these expansion joints. The summary given in this section represents our
current thinkingplease contact the FCC/Alky Best Practices website under GRKM or
ERTC Materials and Equipment Engineering when faced with inspection plans,
monitoring plans or other expansion joint issues.

Expansion Joint Design

See Figure 3.14-A for a schematic of an expansion joint showing some of the
terminology. We leave the fundamental design of the expansion joint system to the
manufacturers in terms of how many convolutions are required, the thickness of the
material, limiting rods, etc. However, we (and the piping engineering contractor) should
carefully review the design. It is extremely important to communicate the expansion
needs and process conditions to the manufacturer. Richmond failed an expansion joint
and had a series of related problems when the engineering contractor miscalculated the
amount of expansion that the expansion joint needed to accommodate.

In addition to handling the expansion/contraction of the system, the bellows of the


expansion joint must be able to operate over many cycles that is, not fail by fatigue.
Following proprietary and Expansion Joint Manufacturers Association calculation
formulas usually give fatigue lives that far exceed what we need. Expansion joints
should be able to last 10-20 years, or possibly even longer.

Squirmwhich is what happens when you try to twist an expansion jointcan be a


major problem with as little as 1.4 degree of twist. You can also have in-plane squirm,
where the bellows tend to bulge out in one direction. Bellows work well only for axial
movements. To accommodate some lateral movement, you will often have two bellows.

Pressure thrust drives much of bellows design. An analogy is a soda can left in the
sunit will tend to bulge both in the middle, and at the ends. If you were to put a
bellows in the middle of the can, the can would get longer due to the pressure thrust.
Most refinery applications need to restrain these pressure thrusts. Tie rods are used to
overcome the pressure thrustthey are always in tension. Usually there are at least
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-41
permission of ERTC
Fluid Catalytic Cracking Process August 2003

three tie rods, to keep the ends in plane. Limit rods are like tie rods, except the nuts are
backed off so there is a gap. The purpose of limit rods is to limit the movement within
design parameters. Typically the nuts of limit rods are tack-welded so that you cannot
adjust the nuts.

Figure 3.14-A. Expansion joint design.

BAND / COLLAR

ROOT RING

CONVOLUTION (CON)
EXPANSION
JOINT COVER

BELLOWS LINER
ROOT RING/
EQUALIZING RING
TIE ROD
LUG

GUSSET

WELD END

CREST
SIDE WALL

ROOT

CON HEIGHT

THICKNESS
TANGENT
CUFF POINT PITCH

Convolution Nomenclature

Gimbaled expansion joints provide complete freedom of movement, rather like a CV joint
on a car. However, they are the most expensive.

Confidential
Copyright 2003 by ChevronTexaco
3-42 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

Pantograph linkagesthe big devices which look like scissors joining the two bellows of
many FCC expansion jointsare used to ensure that the pipe segment between the two
bellows always stays in the middle of the system. They have slotted center hinges to
take the dead weight, especially when filled with catalyst.

Each convolution of a bellows is supposed to handle its fair share of the total movement.
Equalizing rings and centralizing rings are often used to avoid any ring from getting too
much compression. These are rings of various designs placed between each
convolution. The difference between the two is that the reinforcing ring is in contact with
the bottom of the convolution. The industry is currently debating the merits of these
rings.

Special FCC IssuesDouble Ply DesignTemperature Monitoring

Because FCC expansion joints are exposed to very high temperatures and catalyst
erosion, they have some special challenges.

Most modern FCC bellows are made from Inconel 625LCF (Low Cycle Fatigue)
material, rather than the 300-series stainless steels of the past. The intent is to gain
high temperature strength and fatigue resistance, while also improving lower
temperature acidic condensation corrosion resistance. The Inconel 625LCF forms a
brittle phase when exposed to temperatures of 1200-1400F which should not be a
problem at operating temperatures, but severely reduces ductility at room temperature.
This is unlikely to be a practical problem, because if we expose our bellows for long at
those temperatures they will probably fail anyway first by thermal fatigue or simple
overload.

To avoid overheating the bellows they are protected by internal insulation and refractory.
Currently a major issue is the best design to avoid catalyst erosion and keep this
insulation in place. Recently a major El Paso expansion joint failure was caused when
(apparently) the seala stainless steel woven wire mesh ropecame out, allowing the
pillow of ceramic fiber insulation to also come out. This led to overheating and failure
of both plies of the two-ply bellows.

Another failure mode reported by the industryalthough to our knowledge not a


significant problem as yet with the Inconelis acidic condensation corrosion of the
bellows. If the bellows is insulated so well it actually runs too cool, or during
turnarounds, then acids can form and corrode the thin bellows material. Some
manufacturers go so far as to recommend external heating blankets to control the
temperature in the optimum 450-800F range. We have not gone as far as that, but the
external rings covering the bellows are important to prevent driving rain from cooling the
bellows too much (as well as avoiding other problems).

Because the FCC expansion joints are so critical, they actually use two plies of thin
materialin essence, two bellows on top of each other. A pressure-measuring device
(either a gauge, or a pop-up button) is placed between the two plies. The idea is that if
the inner ply fails, we will see the pressure increase and plan to replace the system at
the next planned shutdown. Unfortunately in the El Paso case noted above, there was
less than a year between the first ply failure and the outer ply failure.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-43
permission of ERTC
Fluid Catalytic Cracking Process August 2003

Temperature monitoring has only been used for our most recent expansion joint
installations, at Salt Lake. The outer ply of the Salt Lake bellows will have several
thermocouples installed, so that the temperature can be easily checked. This
inexpensive option should be used on all future FCC expansion joints.

Bellows ManufacturingShop QA

Today large bellows manufacturers start by forming a tube with an autogenous


longitudinal GTAW weld. Plenishing (pressing) the weld puts cold work into it and
eliminates stress concentrators. The convolutions are then formed, most often by roll
forming. In roll forming the tube is spun within three wheels, with the middle wheel
pushing out the convolution.

After the bellows are formed, the end is welded to the pipe portion of the expansion joint
assembly. This weld is critical and difficult. It is also a location where acids could
condense in service and cause corrosion.

We should always call for 100% dye penetrant (PT) of the longitudinal weld, and the end
attachment weld. Sometimes radiography (RT) is done before or after the forming, but
we doubt this provides much additional quality. A hydrotest is usually done on the
system, which we support. Finally, we strongly urge that a moderate pressure test be
conducted by using the pressure gauge system between the two plies. Discuss with the
manufacturer an appropriate pressure level. This may be typically 10-40 psig.

On-Line System Evaluation..Monitoring

On-line inspection of the bellows portion of the expansion joints is difficult because of the
covers. Where visible, the convolutions should be free of distortion, with none bumped
out or squished together. While best practices for bellows inspection have not yet
been developed, the temperature should be monitored at least once per quarter, as
should the pressure gauges. If thermocouples are not available, then the temperature
can be checked by infrared, or even by running your hand along the system, an inch or
so above the cover. It is not unusual to have the bellows section running 100-200F
hotter than the pipe, but if it is running 300-500F hotter, then this indicates problems.

Be sure to check the external restraint rods and hardware in general for any signs of
distortion. Check the location of the growth markers relative to the design. Mark or
record the hot condition locations. Check that the spring cans are not at the extreme
(neither 100% when hot nor 0% when cold). Check for signs of distress at the Y of the
reactor and regenerator standpipe (but there should be signs of movement). Also check
the thrust blocks.

Inspections During Shutdowns

Check anything that you could not see while the unit was hot. Especially check the
refractory, and the insulation. Make sure the refractory seal is sound. If it is, then we
need not check the ceramic fiber pillow insulation. However, if the refractory seal is
missing, then we have the challenge of trying to look at the pillowthere has been some
success with using borescopes by tenacious inspectors.

Confidential
Copyright 2003 by ChevronTexaco
3-44 Energy Research and Technology Company
To be reproduced and used only in accordance with written
permission of ERTC
Corrosion and Metallurgical Deterioration in FCC Plants

Look for corrosion, particularly at the attachment weld. PT this weld if possible. We do
not think a PT examination of the bellows longitudinal weld after service is of much
benefit, but some people do it.

Check the markers, seal pots, etc., as above in the cold condition, relative to the design
positions.

On-Line Repairs

On line repairs are difficult. It may be possible to install 180-degree segments over
leaking bellows, to run to the next shutdown this would be done by specialists for this
type of temporary repair. It involves two GTAW long seam welds, which will not be of
high quality. You will need at least 2.5" of clearance between the bellows and the
external hardware.

Another possible option is to install a clam shell repair over the system, or even box-in
the bellows. With box type repairs we would want to do a careful flexibility analysis.

HOMEWORK QUESTIONS

3.14 Expansion Joints

1) If expansion joints run too hot, we are concerned because __________.

2) If expansion joints run too cold, we are concerned because __________.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company
To be reproduced and used only in accordance with written
3-45
permission of ERTC
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Vu-Graph Index
1. Process Overview
2. Metallurgy Basics
3. High Temperature Oxidation
4. Corrosion by High Temperature H2S
5. Carburization
6. Corrosion by Naphthenic Acids
7. Corrosion by Ammonium Bisulfide
8. Graphitization
9. Creep
10. Corrosion and SCC by Amines
11. Carbonate SCC
12. Caustic SCC
13. Corrosion and SCC by Chlorides
14. Polythionic SCC
15. Embrittlement Mechanisms
16. Wet H2S Cracking, Blistering, and Sulfide Stress Cracking
17. Welding
18. Brittle Fracture
19. Polysulfide and Inhibitor Injection
20. Overhead Corrosion Control Best Practices Summary/
Hydrogen Probes
21. Fatigue
22. Expansion Bellows/Thermal Expansion
23. Refractory
24. Corrosion at Injection Points
25. Corrosion of Dead Legs
26. Interpretation of Piping Corrosion Data
27. Corrosion at Spec Breaks
28. Cooling Water Corrosion/Boiler Water Corrosion
29. Catalyst Erosion Confidential
Copyright 2003 by ChevronTexaco
Materials and Energy Research and Technology Company
To be reproduced and used only in accordance with written
Equipment Technology Unit permission of ERTC

6/30/03 Heidersbach 030204-Section 00


Process Overview
Senior Analyst and Inspector Training
August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-1
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Process Introduction
Safety Topic 15 Min.
Introductions, Overview, and Expectations (Breaks) 30 Min.
I. Chevron FCC Units Schematics of Current Reactor-Regenerator 30 Min.
Configurations Potential Trouble Spots for Erosion
II. Current Reactor-Regenerator Operating Conditions Effects on Erosion 30 Min.
and Reliability
III. Expansion Bellows Discussion 15 Min.
IV. Main Fractionator GRU Corrosion Wash Water/APS/Inhibitor Usage 30 Min.
and Locations
V. GRU Column Issues 15 Min.
VI. Main Fractionator Bottoms System Fouling/Coking Problems 15 Min.
VII. Causes of Unplanned FCC Shutdowns and Best Practices That Address 15 Min.
FCC Reliability
VIII. What Else Are We Doing to Change Things? 30 Min.
A. Lower Fractionator Overhead Temperature
B. Increased Throughput
C. + Improved Monitoring
D. + Lower Feed Sulfur and Nitrogen Content
What Are Your Concerns?
IX. Questions ?
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-2
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Generic FCC Flow Diagram

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-3
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Chevron FCC Units


Year Typical Feed
Location Type of Unit Started Rate, MBPCD Modifications
El Paso, South UOP Stacked 28 UOP Vortex Disengager, Regen
Modifications 1991
El Segundo (Old) Esso Model IV 1954 65 Chevron Riser 1985
El Segundo (New) UOP Vortex Disengager, Slide Valve 1996 65
Control, Dome/Pipe Grid in Regen.
Hawaii (Old) Esso Model IV 1960 20
Hawaii (New) UOP External Riser, Slide Valve 2003 20
Control, Grated Stripper
Pascagoula (Old) Esso Model IV 1963 62 Chevron Riser 1980
Pascagoula (New) UOP Vented Riser, Slide Valve 1991 65 Pipe Grid in Regen 1997
Control
Richmond (Old) Esso Model IV 1959 65 Chevron Riser 1979
Riser Modified 1983
Richmond (New) UOP Vented Riser, Slide Valve 1995 70
Control, Combustor Regen, 2
Catalyst Coolers
Salt Lake (Old) Esso Model IV 1951 12
Salt Lake (New) UOP Vented Riser, Slide Valve 1994 13
Control
Burnaby UOP External Riser, Slide Valve 1984 17 UOP Catalyst Cooler 1993
Control, Combustor Regen
Pembroke Texaco Reactor and Regenerator Circa 1980 90 Numerous

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-4
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Typical FCC Reactor-Regenerator


Operating Conditions

Hawaii Pascagoula Salt Lake Richmond El Segundo


Before After Before After Before After Before After Before After
Modifications Modifications Modifications Modifications Modifications

Feed Rate, BPD 20,000 20,000 65,000 68,000 12,900 12,600 64,000 65,000 65,000 60,000
(70,000 (65,000
Design) Design)
Conversion, LV % 86 92 71 74 77 79 69 80.2 77 78
Reactor
Temp., F 980 1000 965 1,000 970 980 955 1,010 975 990
Pressure, psig 12 12 19 20 21 18 16 22 16 15
Regenerator
Dense Bed Temp., F 1,290 1,225 1,205 1,310 1,280 1,285 1,305 1,320 1,280 1,270
Dilute Temp., F 1,300 1,250 1,230 1,365 1,300 1,320 1,370 1,360 1,260 1,310
Pressure, psig 10.5 10.5 18 16 18 14 13 27 13 19
Air Rate, MSCFM 54 54 97 121 19 21 134 185 135 140
Cat. Circ., T/M 18 26 51 47 8 8 40 51 53 56

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-5
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Chevron Hawaii
FCC
G9900856

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-6
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Chevron Burnaby
UOP FCC Unit
G9700498

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-7
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Chevron Pascagoula
FCC Unit
G9700201

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-8
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Chevron Salt Lake


FCC Unit
G9700199

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-9
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Chevron Richmond
FCC Unit
G9700200

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-10
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Chevron El Segundo
FCC Unit
G9700198

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-11
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

El Paso South -
UOP Stacked
FCC With
Vortex
Disengager
G9700197

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-12
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Cyclone Erosion
G010634

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-13
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Piping Erosion
G010635

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-14
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

FCC Slide Valves


G011524

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-15
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Unpacked Expansion Joint Detail

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-16
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Packed Expansion Joint Detail


Cold Wall

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-17
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Introduction to
Expansion Joints

Expansion Joint
Nomenclature and
Failure Mode
Locations

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-18
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-19
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Seal Rod Description

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-20
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003
Generic Fractionator Overhead System Showing Hydrocarbon
and Process Water Streams and Areas Most Subject to Bisulfide
Corrosion

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-21
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

APS Water Wash - Parallel Cascade


System

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-22
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Hawaii FCC Overhead


Filming Amine Injection Points

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-23
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Deethanizer Debutanizer System

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-24
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Typical Fractionator Bottoms System

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-25
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Unplanned Shutdowns of
ChevronTexaco FCCs
1994 - YTD 1982 - 93
All Chevron FCC 1994-2002 # of S/Ds Days % Days # of S/Ds Days % Days
Cause Down Down Down Down
BC Not Included in Summary
Refractory/Internals 15 172.5 28.4 2 82 7.9
Leak 11 123.8 20.4 16 106 10.2
Rotating Eq 38 95.8 15.8 16 236 22.7
S/U Procedures 9 73 12.0 10 210 20.2
Utilities 29 41.7 6.9 6 4 0.4
Main Fract = Fouling/Coking 10 29.75 4.9 0 0 0.0
Circ Problems 2 16.25 2.7 2 2 0.2
SSD/Inst 20 17.25 2.8 0 0 0.0
Other 4 19 3.1 6 40 3.8
Expn Joints 2 16 2.6 4 66 6.3
COB/Prhtr/FGC/HEX 1 3 0.5 4 4 0.4
Esp 0.0 2 4 0.4
Slide Valves 0.0 6 90 8.7
Reactor/Regen Grid 0.0 4 196 18.8
Total 141 608.05 11.3 78 1040
Pascagoula Bottoms Pump in 94 1 53
Acts of God, Mrktg Caused, Other Units 4 86
Unplanned Days Down Per Year 69.6 86.7 **
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-26
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Causes of Unplanned FCC Shutdowns

1994 - YTD

50
% of Total Downtime

40

30

20
29
10 20 16 12
7 5 3 3 3
0
Refractory/Internals Rotating Eq Utilities Circ Problems Other

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-27
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Best Practices That Address FCC


Reliability
Refractory Best Practice

Cyclone Reliability Best Practice

Reactor/Regenerator Internals Design Review

Expansion Joint Best Practice

Startup Procedures
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-28
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Critical Variable Monitoring for Fixed


Equipment Integrity
Risk Ranking*
Variable to be Likelihood/ Data
Monitored Significance Severity/Risk Status Location Action Required Owner
V-70
Regenerator
Variables
V-70 Regenerator Regenerator internals are T304HSS and are 2/3/3 Not normal practice to wash None N/A
Shutdown susceptible to sensitization at operating regenerator. Impractical to keep
Procedures temperatures. During shutdowns, the internals dry, potential for caustic SCC if
should be kept dry or polythionic acid stress not all soda ash solution
corrosion cracking is possible. If water wash is removed.
necessary, use soda ash solution to minimize the
possibility of cracking.
V-70 Skin V-70 is refractory-lined CS. Hot spots indicate 1/2/1 Design limit of 650F on CS Inspection Consider increasing IR SRS
Temperature damage to refractory and can lead to overheat, shell, 300-350F typical. Records frequency before major S/D to
erosion, corrosion and/or exceed mechanical watch for spots.
design limit of regenerator. Perform quarterly IR inspection
to monitor for hot spots.
V-70 Primary and If velocity is too high, get excessive cyclone 1/2/1 Computer calculates primary Honeywell Ask Mike McKee if thinks this McKee
Secondary erosion. Possible catalyst carryover damaging cyclone velocity. needs to be alarmed.
Cyclone Inlet expander. Primary cyclone design limits:
Velocities 42-70 fps, 49-67 fps typical.
Currently not alarmed.
V-70 Heat Up If heat up V-70 too quickly, can damage 1/3/2 COD: Heat up 200F/hr max. Need to check with RAD on LAF/
and Cool Down refractory during dryout. If cool too quickly, Cool Down 100F/hr max. refractory heat up and cool RAD
Rates can thermally shock refractory. Refractory manufacturers also down rates- fix COD table as
set dryout schedule; may be needed.
V-70 heating and cooling rates are also limited more restrictive.
by expander heating and cooling rates, which Calculated. No alarms.
may be more restrictive (3F/min. above
600F).
V-70 Regenerator Major increases can accelerate damage to 1/2/1 Design limit: 1450F Chevmon None N/A
Temperature regenerator refractory and internals. Excess of COD UL: 1350F continuous, 50TI724A
CO will lead to regenerator afterburn and 1450F excursions.
potential overheat. Alarmed @ 1385F

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-29
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Best Practices That Have Effectively


Reduced Incidents
Bottoms System Best Practice

Overhead Corrosion Control

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-30
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Turnaround Support Package


FCC Pacesetter Turnarounds
Philosophy and Practices
FCC Shutdown Guidebook
Best Practice Team Support
Intensive Process Review
Work Scope Review
Expert involvement
Execution Support
Post T/A Technical Review

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-31
6/24/03 Heidersbach 030204-Sect 01
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Future Changes That May Effect


FCC Reliability

A. Lower Fractionator Overhead Temperature


B. Increased Throughput
C. + Improved Monitoring
D. + Lower Feed Sulfur and Nitrogen Content

What are your concerns?

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 1-32
6/24/03 Heidersbach 030204-Sect 01
Metallurgy Basics

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-1
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Metallurgy Basics

(See new section in Corrosion Prevention and


Metallurgy Manual)

We will discuss:
The nature of metals
Characteristics of common classes of metals
Heat treatment
Some commonly used pressure vessel steels

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-2
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Nature of Metals
Metals behavior is largely determined by its:
Crystal structure and phase
Chemical composition
The crystal structure refers to the highly ordered
way the major atoms of the steel are arranged

Example: Steel at room temperature has the iron


atoms arranged in the body centered cubic crystal
structure (called ferrite)

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-3
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Nature of Metals
Most metals used in refining are actually alloys, and
contain more than one element
Carbon steel is an alloy of iron, carbon, and usually
manganese and silicon
Low alloy steel usually contains 1-5% Cr, Ni, Mo, or
other elements to improve the strength, toughness, or
other properties
Most metals we use involve more than one phase
Iron carbides in steel provide strength
Sulfide or oxide inclusions create weak spots for
mechanical or hydrogen damage

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-4
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Formation of Grains and Grain


Boundaries in Metals
G9900914

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-5
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Ferrite-Pearlite Microstructure of
Carbon Steel
G9900907.jpj

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-6
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Ferritic Steels: Characteristics


Includes carbon steels, Cr-Mo alloy steels, and
400 Series stainlesses
Are magnetic
Exhibit ductile-to-brittle transition temperature
Most are hardenable by heat treatment or welding
Heat treated to relieve stress and reduce hardness
Immune to chloride SCC
Susceptible to cracking by wet H2S, carbonate,
amine

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-7
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Austenitic Steels: Characteristics


Includes most 300 Series stainlesses and high Ni steels
Are non-magnetic, but welds and castings have some
magnetism (ferrite)
No ductile-to-brittle transition; good toughness at low
temperature
Are not hardenable by heat treatment (a few PH SS are
exceptions)
Heat treated for stress relief, for corrosion resistance, or
to soften (especially after cold working)
Susceptible to chloride SCC
Resist cracking by wet H2S, carbonate, amine
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-8
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Comparison: C & C-Mn Steel Versus


Cr-Mo Steels
Characteristic C&C-Mn Steels Cr-Mo Steels

Hardening by Heat Treatment or Low High


Welding
Preheat for Welding Heavy Sections Only Always

PWHT Required >1-1/2 Thick, or for Always, With Few


Environment Exceptions
In-Service Embrittlement No (Rare) Yes
650-1050F
Allowable Stress at 1000F <3 ksi 7-10 ksi

Maximum Code Temperature, F 1000 (800F for 1200


Graphitization)
H2 Attack Limit at 750 psia, F 500 1000

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-9
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Heat Treatment Terminology


Carbon and alloy steels
Annealing: Heating a steel to about 1650F so that a single phase
develops, with all of the elements dissolved in it, followed by a very
slow cool (furnace cool) to prevent hardening. This is the softest
and lowest strength state.
Normalizing: Heating carbon or low alloy steels to about the same
temperature as annealing, but cooling fairly rapidly in air.
Improves toughness of steels.
Quenching: Rapidly cooling from annealing temperatures, such as
by using oil or water sprays. Results in higher hardness and
strength.
Tempering: Typically done after quenching at temperatures similar
to those used for PWHT (e.g., 1300F for 2.25 Cr) to relieve the
stresses and soften the material).
Post-Weld Heat Treatment (PWHT): Performed after welding to
reduce residual stresses and, for ferritic steels, to temper (soften)
the weld zone and improve its toughness.
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-10
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Heat Treatment Terminology (Contd)


Stainless steels
Solution annealing: Most often our industry uses the term to
refer to heat treatment of stainless steels (say, at 2100F)
with controlled cooling to provide maximum corrosion
resistance.
Stabilize annealing: Sometimes called thermal stabilization.
A heat treatment of some stainless steels to stabilize or
lock in desirable carbides. This allows the SS to be used
effectively at high temperatures.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-11
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003
G010553

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-12
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Nominal Stress; Strain Diagram


G9900904

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-13
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Charpy V-Notch Energy Absorption of


A516 Grade 70 Steel
G010551

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-14
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Typical ASTM Specifications for Materials


Material Plate Pipe Tubing Forgings Bars Castings
Cast Iron A48
A278
Carbon Steel A285 A53 A161 A105 A575
A515 A106 A179 A181 A576
A516 A671 A210 A663
A672 A214 A675
A691
Carbon - 1/2 Mo A204 A335 A161 A182 A217
A209
1 Cr - 1/2 Mo, 1-1/4 Cr - 1/2 Mo, & 2-1/4 Cr - 1 Mo A387 A335 A213 A182 A217
5 Cr - 1/2 Mo A387 A335 A213 A182 A217
12 Cr, CA 15 A240 A268 A182 A479 A217
Austenitic Stainless 304, 304L, 316, 316L, 321, A240 A312 A213 A182 A479 A351
347, 310, CF3, CF3M, CF8, CF8C, CF8M A358 A249 A744
HK40 & HP Modified A297
Duplex Stainless 2205 A240 A790 A789 A182 A276
Super Stainless AL6XN B688 B675 B676
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-15
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Typical ASTM Specifications for Materials


Material Plate Pipe Tubing Forgings Bars Castings
Incoloy Alloy 800 B409 B407 B407 B408
Incoloy Alloy 825 B424 B423 B163 B425
904L B625 B673 B674 B649
B677 B677
Alloy 20 B463 B464 B468 B462 B472 A351
Copper B152 B42 B75
B111
Admiralty Brass B111 (Grades B, C, D)
Naval Brass B171 B124
70-30 Cu-Ni B171 B467 B111
B608 B395
Titanium B265 B337 B338 B381 B348 B367
Monel B127 B165 B163 B164
Inconel 625 B443 B444 B444 B446
Hastelloy C276 B575 B622 B622 B574 A494
Ni Resist A436
Aluminum B209 B241 B234 B247 B211
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-16
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Chromium Steels Typical Chemistry


and Strength
Chemistry Tensile Yield
Grade/ Strength, Strength,
Material Class C Mn Cr Mo Ni V Cu N Ti B ksi ksi
AISI 4140 0.4 0.9 1.0 0.2 90-200 60-175
AISI 4340 0.4 0.7 0.8 0.2 1.8 110-220 100-200
SA387* 12 0.15 0.5 1.0 0.5 55 (65) 33 (40)
SA387* 11 0.15 0.5 1.2 0.5 60 (75) 35 (45)
SA387* 22 0.12 0.5 2.2 1.0 60 (75) 30 (45)
SA387* 5 0.15 0.5 5.0 0.5 60 (75) 30 (45)
SA387* 9 0.15 0.5 9.0 1.0 60 (75) 30 (45)

Note: Balance is Fe in All Cases

* Tensile and yield values listed are minimum for Class 1 plate. Values in parenthesis are
for Class 2 plate.
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-17
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Ferritic, Martensitic, and Duplex Stainless


Steels Typical Chemistry and Strength
Common Name Tensile Yield
(UNS Number) Cr Ni Mo N C Other Strength, ksi Strength, ksi
4101 (S41000) 12 0.75 0.15 70 35

410S1 (S41008) 12 0.75 0.08 60 30

4052 (S40500) 13 0.6 0.08 0.2 Al 60 25

4302 (S43000) 17 0.1 65 30

440A1 (S44002) 17 0.6 0.7 105 60

22053 (S31803) 22 5.5 3.0 0.14 0.02 90 65

Notes:
1. Martensitic
2. Ferritic
3. Duplex

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-18
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Austenitic Stainless Steels Typical


Chemistry
Common Name
(UNS Number) Cr Ni Mo Cu N C Other

304L (S30403) 18 8 0.035


316L (S31603) 17 11 2.1 0.035
317L (S31703) 19 13 3.1 0.035
321 (S32100) 18 11 0.08 Ti = 5 x (C + N) (0.70% Max.)
347 (S34700) 18 11 0.08 Cb + Ta = 10 x C (1.1% Max.)
HK-40 26 20 0.40
HP Mod. 26 35 0.40 1.2 Cb
310 (S31000) 25 20 0.25
Alloy 20 Cb3 (N08020) 20 33 2.1 3.5 0.07 Cb = 8 x C (1.0 Max.)
AL-6XN (N08376) 20.5 24 6.2 0.2 0.02
309 (S30900) 23 13 0.20

Note: Balance is Fe in All Cases Confidential


Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-19
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Common Precipitation Hardening Stainless


Steels Typical Chemistry and Strength
Common Name Chemistry Tensile Yield
(UNS Number) C Cr Ni Mo Cu Other Strength, ksi Strength, ksi

17-4PH (S17400) 0.05 16.5 4 4 0.4 Al 1501 1351

17-7PH (S17700) 0.06 17 7 1.2 Al 1752 1552


ASTM Grade 6603 0.04 15 25.5 1.2 2.1 Ti 140 95
(S66286) 0.006 B
0.25 V

Notes:
1. Precipitation Hardening Temperature: H 1100 (F)
2. Precipitation Hardening Temperature: H 1050 (F)
3. ASTM A 453 for Bolting, A638 for Forging

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-20
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

High Nickel Alloys Typical Chemistry


and Strength
Common Name Tensile Yield
(UNS Number) Cr Ni Mo Cu C Other Strength, ksi1 Strength, ksi1
Monel 400 66 31 0.1 70 25
(N04400)

Incoloy 800 21 32.5 0.05 0.4 Ti 65 25


(N08800) 0.4 Al

Incoloy 825 21 42 3 2.3 0.03 1 Ti 85 35


(N08825)

Inconel 600 15.5 76 0.08 8 Fe 80 30


(N06600)

Inconel 625 21.5 61 9 0.05 3.7 Cb 120 60


(N06625) 4 Fe

Hastelloy C-276 15.5 57 15.5 0.005 4V 115 52


(N10276) 5.5 Fe

Note:
1. Annealed Condition
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-21
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Copper Alloys Typical Chemistry


and Strength
Common Name Chemistry Tensile Yield
(UNS Number) Cu Zn Sn Al Other Strength, ksi Strength, ksi
Admiralty Brass 70 29 1 0.05 As 48 18
(C44300)
Naval Brass 60 39 0.7 58 25
(C46400)
90-10 Cu-Ni 83 1.5 Fe 44 16
(C70600) 10 Ni
70-30 Cu-Ni 69 0.7 Fe 54 20
(C71500) 30 Ni
Aluminum Bronze 90 7 3 Fe 76 33
(C61400)
Copper 99.9 32 10
(C11000)

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-22
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Material Temperature Limits, F


Carbon C- 1-1/4 Cr - 2-1/4 Cr - 5 Cr - 12 Cr 18 Cr - 25 Cr -
Steel 1/2 Mo 1/2 Mo 1 Mo 1/2 Mo (410) 8 Ni (304) 20 Ni (310)
Strength (3000 psi) 990 1075 1135 1150 1115 1100 1275 1320
Oxidation (10 mpy Loss) 1050 1050 1100 1175 1200 1400 1500 2050
Graphitization (Welded Only) 800 850
Temper Embrittlement 700-1050
885 Embrittlement 650-950
Sigma Embrittlement 1100-1700 1100-1700
Hardening on Cooling 1330 1350 1375 1425 1425 1450
Hydrogen Damage
(H2pp, 750 psi) 500 500 1000 1100 1150
(H2pp, 2000 psi) 460 460 650 850 1125
Caustic Embrittlement Stress 110 110 110 110 110 110 200 200
Corrosion Cracking
Chloride Stress Corrosion 140 140
Cracking
Sulfidation (Hot H2S Corrosion) 500 500 500 500 650
Hot H2-H2S Corrosion (0.1 psia H2S) 500 500 500 500 500 750 950 950
Onset of Creep 800 900 900 900 900 950 1100 1100*
Sensitization 750-1500 750-1500
Naphthenic Acid Corrosion 350 350 350 350 350 350 350 350

*Cast 25 Cr - 20 Ni (HK40) is 1400F

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-23
6/24/03 Heidersbach 030204-Sect 02
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Allowable Stresses for Plate Steels


ASME Section VIII Division 1
25

C- Steel; SA516-70
1.25Cr-0.5Mo; SA387 Gr 11, Cl 2
20 2.25Cr-1Mo; SA387 Gr. 22, Cl 2
12Cr; SA240 T410
SA240 T304
SA 240 T347

15
Stress (ksi)

10

0
0 200 400 600 800 1000 1200 1400 1600
Temperature (F)

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 2-24
6/24/03 Heidersbach 030204-Sect 02
High Temperature Oxidation

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 3-1
6/24/03 Heidersbach 030204-Sect 03
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

High Temperature Oxidation


Characteristics
Requires oxygen and high temperatures generally greater than 1000F
Oxide scale is black and metallic looking, making it difficult to visually
determine that attack has occurred
Thermal cycling will spall protective oxide scales and accelerate corrosion
penetration
Most prevalent in the regenerator and its flue gas system
Prevention
Insulate to lower metal temperature, or
Use alloys containing high chromium 304H is used for most non-insulated
regenerator internals
Inspection
Energetic beating with a hammer will remove oxide scales and reveal damage
UT

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 3-2
6/24/03 Heidersbach 030204-Sect 03
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Oxidation Failure of Carbon Steel Hexmesh


Support Studs From a Regenerator

Corroded studs from Irving FCC regenerator


Corroded ends are located at top of picture
PR 951230-15

Magnification 1.25X
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 3-3
6/24/03 Heidersbach 030204-Sect 03
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Approximate Temperatures Above Which


Significant Oxidation Occurs for Various Alloys

Material Grade Oxidation Temperature, F


Carbon Steel B 1000
Carbon - 1/2 Mo Steel T1 or P1 1000
1-1/4 Cr - 1/2 Mo Steel T11 or P11 1100
2-1/4 Cr - 1 Mo Steel - 3 Cr T22 or P22 1150
1 Mo Steel T21 or P21 1150
5 Cr - 1/2 Mo Steel T5 or P5 1200
7 Cr - 1/2 Mo Steel T7 or P7 1200
9 Cr - 1 Mo Steel T9 or P9 1250
18 Cr - 8 Ni Steel 304 or 304H 1500
18 Cr - 8 Ni Steel 316 or 316H 1500
18 Cr - 8 Ni Steel 321 or 321H 1500
18 Cr - 8 Ni Steel 347 or 347H 1500
Ni - Fe - Cr Alloy 800H 1700
25 Cr - 20 Ni 310/HK-40 1900
25 Cr - 35 Ni HP-Alloys 1950

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 3-4
6/24/03 Heidersbach 030204-Sect 03
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Effect of Temperature on Oxidation Rate


of Steels
G9900870

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 3-5
6/24/03 Heidersbach 030204-Sect 03
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Effect of Temperature on
Oxidation Rate of Steels
G9600549

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 3-6
6/24/03 Heidersbach 030204-Sect 03
Corrosion by High Temperature H2S

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 4-1
6/24/03 Heidersbach 030204-Sect 04
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Sulfidation (H2S) Corrosion


Characteristics
For carbon steel occurs above 500F, whenever H2S >1 ppm
Corrosion rates are very temperature sensitive (for carbon steel we start off cliff at 550F) --
above 800F rates actually decrease
Furnace tubes are a serious risk because metal temperatures are not accurately known
Corrosion is usually uniform but watch highly turbulent areas and hot spots
Scale is pyrophoric
In FCC's the fractionator and the fractionator bottoms and sidecut equipment are likely
locations for hot H2S corrosion (the reactor does not experience high sulfidation despite
very high temperatures)
Prevention
Upgrade to 5 Cr - 1/2 Mo, 9 Cr - 1 Mo, or stainless steel
Insulate to reduce metal temperature
Clad vessels with stainless (usually 405 or 410S)
Inspection
See Inspection Strategy IS-32
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 4-2
6/24/03 Heidersbach 030204-Sect 04
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Corrosion of Furnace Tubes in Light


Hydrocarbon Service (700-2000 ppm H2S)
G9900868

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 4-3
6/24/03 Heidersbach 030204-Sect 04
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

FCC Sulfidation Corrosion (Reference Only


Not Valid for Estimating Actual Corrosion Rates)
G9900905

Corrosion of Steels in Crude Oil Containing 1.5% Sulfur Confidential


Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 4-4
6/24/03 Heidersbach 030204-Sect 04
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Sulfidation: Hawaii


Fractionator Bottoms Steam Generator
2003 FCC shutdown
CS tubesheet
5 Cr tubes
Fractionator bottoms
cooled from 610F to
~400F though 4-pass
exchanger
Sulfidation corrosion
of tubesheet face on
first pass only
Tubes not damaged

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 4-5
6/24/03 Heidersbach 030204-Sect 04
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Sulfidation: Hawaii


Fractionator Bottoms Steam Generator

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 4-6
6/24/03 Heidersbach 030204-Sect 04
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Effect of Si on Sulfidation
Small amounts of Si greatly reduce sulfidation rate
(as little as 0.10%)
A106 pipe is Si killed (deoxidized), but A53 pipe is
not
Result: Sulfidation may be localized to an
individual component
Recommendation for CS >500F: Inspect every
component in piping system once to confirm they
are corroding at the same rate; thereafter, just
monitor a few (as usual)

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 4-7
6/24/03 Heidersbach 030204-Sect 04
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example: Effect of Si on Sulfidation


El Paso intermediate reflux from
FCC fractionator line
Piping installed in 1957 leaked
in 1988
Temperature crept up to 620F
from design of 610F
Elbow was monitored and
showed low corrosion
Straight run of piping failed after
RT0112715 - M.A. Colombo, weld repairs
FCC Intermediate Reflux
Line Cracking Unit No. 2 Si content of pipe 0.02%
El Paso 12/07/1988
Original corrosion rate 5-10 mpy
Corrosion rate 20 mpy

Si content of elbow 0.17%


Corrosion rates 5-10 mpy

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 4-8
6/24/03 Heidersbach 030204-Sect 04
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

El Paso Intermediate Reflux From


FCC Fractionator Line

RT0112715 - M.A. Colombo,


FCC Intermediate Reflux Line
Cracking Unit No. 2
El Paso 12/07/1988

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 4-9
6/24/03 Heidersbach 030204-Sect 04
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example: Effect of Si on Sulfidation


Salt Lake coker gas oil
Locations that were
monitored showed low
corrosion
Three sections of A53
straight piping corroded
2-3x faster than A106
Temperature crept from
design of 580F to 650F
to above 700F during
the last 2 years
RT0201401 - K. Heidersbach, Salt Lake Refinery
HGO Line Failure 06/12/2002
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 4-10
6/24/03 Heidersbach 030204-Sect 04
Localized Sulfidation Corrosion Sneaks Up on
Salt Lake Refinery
Regular Monitoring of Line did not Prevent Fire

A leak last March in the coker fractionator heavy gas oil line at Salt Lake
reminds us that sulfidation corrosion of carbon steel can vary
tremendously from component to component, and depends greatly on
the Si content of the steel.

This line was regularly monitored over its 30 year life in accordance
with recommended practices, including gauge points at several elbows,
flanges, and a couple of straight runs. Corrosion rates at the gauge A106 Pipe -
points were in the 3-5 mpy range, but rates were about 5 times higher
Si killed
in two sections of the piping circuit that were not being monitored.

The two sections that corroded faster were made to ASTM A53, which
does not require that steel to be Si-killed. All other sections of this line
met the requirements of ASTM A106, which requires at least 0.1% Si.
This small difference in silicon content accounts for the dramatic
difference in sulfidation corrosion rate. Unfortunately, it is not possible
to detect Si levels in steel through a portable metal analyzer.

For new construction, we limit carbon steel to a maximum temperature


of 550 F when H2S is present. We are working to determine how best
to apply the high Si grades in our piping classes to optimize the
performance of carbon steel as temperatures approach the 550 F limit.
In Salt Lakes case, both the H2S content and the temperature of the
coker fractionator gas oil line had crept up over the years, and left the A53 Pipe -
carbon steel vulnerable to corrosion. Not Si killed
Contact Krista Heidersbach (510-242-1877) or Ned Niccolls
(510-242-4539) for more information.

Identifying Piping Systems At Risk for Suldation Corrosion


Identify all carbon steel piping operating above 550 F and Once it is established that all components are corroding at the same rate,
containing a few ppm or more H2S. then a small number of gauge points can once again be used to represent
the entire piping circuit.
Inspect every piping component in each circuit to identify any
that may be corroding at a higher rate. Contact Ned Niccolls for guidance on dealing with specific cases.

finally reaching the full 100% of design pressure to the reactor quicker, reducing the warm-up effects at lower temperatures. For guidance on
once the MPT is attained. This increase in time by a shift or more in some cases. the new policy, or to determine if it applies to
pressure mirrors the increase in fracture your reactor, contact Reactor Metallurgy
These new practices only apply to those reactors
toughness that the steel undergoes as the Specialist Dave Cooke (510-242-5452), BP Team
with MPTs of 250 F or above, which are
temperature is raised. Thus, a constant margin of Leader Jeff Johns (801-539-7345), or see
specifically listed in the BP document. They are
safety against fracture is maintained throughout BP #12 on the Hydroprocessing web site:
currently not applicable to reactors with lower
the heating and cooling process. Earlier www.ric841.chevrontexaco.net/hydro/hydroBP
MPTs because of concerns of possible hydrogen
pressurization means that heat can be transferred /BP.htm

Proprietary Information Authorized Use Only 33


Carburization

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 5-1
6/24/03 Heidersbach 030204-Sect 05
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Carburization
Characteristics
Carbon from the process combines with iron in the steel at the metal
surface
Typically occurs at high temperatures under coke deposits
Bulging and flaking of carburized layer can lead to gradual metal loss
Probably contributes to metal loss of carbon steel reactor cyclones
1-1/4 Cr shells do not carburize significantly
Carburization can make steel cyclones more difficult to weld repair
Carburization can occur on 300 Series regenerator internals if
combustion is incomplete
Overall, not usually a problem in FCC plants
Prevention
Chromium additions retard carburization in oxidizing or sulfidizing
environments but not in reducing environments
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 5-2
6/24/03 Heidersbach 030204-Sect 05
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Carburization Depth of Type 304 SS as a Function


of Time and Temperature (Carbon Activity = 1.0)
G9900862

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 5-3
6/24/03 Heidersbach 030204-Sect 05
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Metal Dusting of Type 321


Stainless Steel

From the Cedar Bayou


Hydrotreater Feed
Furnace Convection
Section

RT0109733 - R.A. Semmel, Metal Dusting of Return


Bends in BA-651 Feed Furnace Heater

Micro Showing Change in Carburization With Depth

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 5-4
6/24/03 Heidersbach 030204-Sect 05
Corrosion by Naphthenic Acids

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 6-1
6/24/03 Heidersbach 030204-Sect 06
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Naphthenic Acid Corrosion


Characteristics
Corrosion rate increases with total acid number (TAN) and temperature
Carbon steel, Cr steels, and 300 Series SS are all susceptible, though carbon
steel is most susceptible
Threshold temperature is about 350F
Naphthenic acids are destroyed in the reactor so corrosion is not a problem
downstream of the reactor
Corrosion is highly sensitive to velocity and turbulence
Is a concern only when refineries run naphthenic crudes and don't hydrotreat
the FCC feed
Prevention
Limit TAN via blending
Use molybdenum containing stainless or alloys (316, 317, Inconel 625)
Inspection
See Inspection Strategy IS-8
RT and/or UT scanning are typical
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 6-2
6/24/03 Heidersbach 030204-Sect 06
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Naphthenic Acid Corrosion of


Carbon Steel
G9900881

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 6-3
6/24/03 Heidersbach 030204-Sect 06
Corrosion by Ammonium Bisulfide

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 7-1
6/24/03 Heidersbach 030204-Sect 07
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Ammonium Bisulfide (NH4HS) Corrosion


Characteristics
Occurs only if liquid water is present
Corrosion rate depends primarily on concentration of NH4HS and fluid
velocity 2% NH4HS is an accepted threshold for steel corrosion.
Dilute NH4HS can corrode copper alloys.
Temperature doesn't strongly affect corrosion rates
Corrosion can and has occurred in oil-water-gas mixtures, wet oil, wet
gas, and sour water corrosion can be highly localized
NH4HS promotes wet H2S cracking and hydrogen blistering
Cyanides greatly increase corrosion

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 7-2
6/24/03 Heidersbach 030204-Sect 07
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Ammonium Bisulfide (NH4HS) Corrosion


(Contd)
Prevention
Use alloys (300 Series SS is often a good choice below 150F; titanium
and duplex SS perform well but are more costly than 300 Series SS)
Keep velocities at 10-20 fps and minimize turbulence. Also, eliminate
dead legs.
Use polysulfide or inhibitors to chemically lessen corrosion
Actively monitor the system to detect if the plant is moving into a
corrosive regime
Inspection
See Inspection Strategy IS-2
For monitoring use hydrogen probes; watch N2 in feed and other CPVs

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 7-3
6/24/03 Heidersbach 030204-Sect 07
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Generic Fractionator Overhead System Showing


Areas Most Subject to Bisulfide Corrosion
G9900866

Note: Watch Dead Legs and High Velocity Locations in Particular


Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 7-4
6/24/03 Heidersbach 030204-Sect 07
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Flattened area of admiralty brass tube


section. Flow markings marked with
arrow.
Cross-section of admiralty brass tube
section. Thinned area marked with arrow.
PR 951230-1, 2

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 7-5
6/24/03 Heidersbach 030204-Sect 07
Graphitization

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 8-1
6/24/03 Heidersbach 030204-Sect 08
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Graphitization
Characteristics
Occurs in carbon and carbon - 1/2 moly steel chromium containing low alloy
steels are immune
Occurs after long-term exposure above 800F for carbon steel and above 850F
for C - 1/2 Mo
Is rarely serious except in weld hazs of pipe where it can create a plane of
weakness susceptible to sudden failure (e.g., Port Arthur spent cat line failure)
Reactor cyclones, the fractionator inlet nozzle and adjacent shell, and any
location where the internal thermal insulation for carbon steel is damaged are
possible locations for graphitization
Prevention
Limit temperature, or
Use chromium containing steels
Inspection
RT
Shear wave UT
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 8-2
6/24/03 Heidersbach 030204-Sect 08
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Temperature, C
Temperature, F

538
G9900903

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 8-3
6/24/03 Heidersbach 030204-Sect 08
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Graphitization: Weld HAZ of


Port Arthur Spent Cat Line

Scattered Graphite Nodules are in the Basemetal and Weldmetal


RT0108321 C.A. Shargay,
Cracked Spent Catalyst Line
Port Arthur FCCU 1241
951341-1

06/06/1985
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 8-4
6/24/03 Heidersbach 030204-Sect 08
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Graphitization:
Weld HAZ of Port Arthur Spent Cat Line

Graphitization Graphitization
Along Weld Along Weld
100X 50X

Graphitization Graphitization
Nodules in Nodules in
Basemetal Basemetal
100X 50X

RT0108321 - C.A. Shargay,


Cracked Spent Catalyst
Line Port Arthur FCCU 1241
06/06/1985
951341-1

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 8-5
6/24/03 Heidersbach 030204-Sect 08
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Graphitization: Bend Tests


From Port Arthur Spent Cat Line

Uncracked Specimen

Cracked Specimen
RT0108321 C.A. Shargay,
Cracked Spent Catalyst Line
Port Arthur FCCU 1241
06/06/1985
951341-2

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 8-6
6/24/03 Heidersbach 030204-Sect 08
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Graphitization: Weld HAZ

Severe
Eyebrow Type
of Graphitization
Adjacent to Weld
PR 951030-1

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 8-7
6/24/03 Heidersbach 030204-Sect 08
Creep (Creep Exercise)

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-1
6/24/03 Heidersbach 030204-Sect 09
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Creep
Characteristics
Results from long-term exposure to elevated temperatures creep limits vary
for different materials, generally increasing with chromium content
Creep starts as tiny voids, which grow and link to form fissures and cracks
most materials show bulging before failure
Occurs where temperatures and stresses are highest
Furnace tubes are designed for long life but fail prematurely due to overfiring,
coking, flame impingement, or scaling
In boilers, steam generation and water wall, tubes operate below the creep
temperature and should last indefinitely superheater tubes often operate in
the creep range
Prevention
Proper furnace operation (check by closely monitoring tube skin temperatures)
Reduce metal temperature with thermal insulation
Alloy upgrades
Inspection
See Inspection Strategy IS-34
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-2
6/24/03 Heidersbach 030204-Sect 09
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Minimum Temperature for Creep

Material Minimum T for Creep


Carbon Steel 800F
1-1/4 Cr 1 Mo, 2-1/4 Cr 1 Mo * 900F
5 Cr 1/2 Mo
9 Cr 1 Mo 950F
T-347 SS 1100F

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-3
6/24/03 Heidersbach 030204-Sect 09
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Typical Tensile Creep Curve


G96001603

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-4
6/24/03 Heidersbach 030204-Sect 09
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Creep Voids
950625-26

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-5
6/24/03 Heidersbach 030204-Sect 09
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Examples of Creep: Hawaii FCC Reactor


Primary Cyclone Hanger Rods
Four 5 Cr - 1/2 Mo reactor
cyclone hanger rods failed
Failures were due to creep
and brittle fracture
High loads on rods
resulted from coke buildup
28 years of service
Operating temperature 950-
1000F

RT0112692 - R.M. Thompson, Primary Cyclone Hanger


Rod Failures - Hawaii FCC Reactor 09/13/1988
PR 951230-16

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-6
6/24/03 Heidersbach 030204-Sect 09
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Replacement Criteria
Replacement is condition based, not scheduled
Actual life usually much different than 100,000-hour
design life
25F has ~2X effect
Corrosion/erosion rate varies from design
Stress (pressure) usually less than design
Material variability (mean versus minimum)
Remaining life calculations are approximations only,
and are usually conservative

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-7
6/24/03 Heidersbach 030204-Sect 09
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Inspection for Creep


Tube replacement
3-5% bulging (CS, low alloy)
Thinning projected to reach Tmin.
Remove sample for lab testing (Omega)

Structural components (cyclones, etc.)


Based on cracking, thinning, distortion
Remove samples for lab testing

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-8
6/24/03 Heidersbach 030204-Sect 09
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Creep, Fatigue, and


Creep Embrittlement Interaction
Effects are synergistic
Creep is worse if stress is cyclic
Fatigue is worse at high temperatures
Failures occur at stress concentrations and pre-
existing defects (welds, corners)
For 1 Cr and 1-1/4 Cr, creep embrittlement magnifies
the effects, particularly for high strength material

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-9
6/24/03 Heidersbach 030204-Sect 09
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Creep Life of Steel


(Especially Furnace Tubes)
Calculations
API RP-530 Curves
Exercises

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-10
6/24/03 Heidersbach 030204-Sect 09
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Hoop Stress Due to Operating Pressure

Internal Pressure Hoop Stress (S)

P Do Where: S = Hoop Stress (ksi)


S= 1 x 103 P = Operating Pressure (psig)
2 t Do
t
= Outside Diameter (In.)
= Minimum Wall Thickness (In.)

Example: P = 2450 psig


Do = 7.0 In.
t = 0.6 In.

2450 7.0 3
S= 1 x10 = 13.1 ksi
2 0.6 Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-11
6/24/03 Heidersbach 030204-Sect 09
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Maximum Operating Skin Temperature


ChevronTexaco practice is to use design life of
100,000 hours
Use API Recommended Practice 530 figure for
material of interest
Follow line for calculated operating stress on right
axis across to intersection with 100,000-hour design
life curve
Follow vertical line down to design temperature on
bottom axis
Subtract 25F for gas-fired heaters, or 50F for oil-
fired heaters
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-12
6/24/03 Heidersbach 030204-Sect 09
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Medium-Carbon Steel, ASTM A 53 Grade B


(Seamless), A 106 Grade B, A 210 Grade A-1
G010650

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-13
6/24/03 Heidersbach 030204-Sect 09
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

9 Cr - 1 Mo Steel, ASTM A 213 T9,


A 335 P9, A 200 T9
G010651

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-14
6/24/03 Heidersbach 030204-Sect 09
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Types 347 and 347H Stainless Steel,


ASTM A 213, A 271, A 312, A 376
G010649

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-15
6/24/03 Heidersbach 030204-Sect 09
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Creep Exercises
1) What is the maximum operating skin temperature
for gas-fired 347 stainless steel tubes with the
dimensions given previously (S = 1.31 ksi)?
2) Do the same exercise as (1), but for carbon steel
3) Suppose you have a T-347H SS tube operating at
10 ksi and 1200F how much would your tube life
be reduced if you increase the pressure (stress) by
10%?
4) For the same T-347H tube operating at 10 ksi and
1200F how much would your tube life be
reduced if you now increase the temperature by
10%?
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 9-16
6/24/03 Heidersbach 030204-Sect 09
Corrosion and SCC by Amines

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 10-1
6/24/03 Heidersbach 030204-Sect 10
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Amine Corrosion
Characteristics
Corrosion can be both general and localized (watch
HAZs, downstream of control valves)
Corrosion is dependent on:
Temperature
Amine loading (limit to ~0.50 mole acid gas/mole DEA, 0.40
for MEA, and 0.45 for MDEA)
Velocity (limit to 6 fps for carbon steel)
Amine concentration (limit is ~30 wt % DEA, 20% for MEA,
and 50% for MDEA)
Heat stable salts (limit is 10% of amine concentration) -- see
BP for HSS removal at Amine/SRU website

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 10-2
6/24/03 Heidersbach 030204-Sect 10
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Amine Corrosion (Contd)


Historical vulnerable corrosion areas include:
Regenerator bottoms
Reboiler outlet nozzles and return piping
In and downstream of pumps
In and downstream of pressure letdown valves
Absorber sour gas inlet area
Prevention
Stay within parameters listed above
Upgrade to 300 series stainless steel
Inspection
See Inspection Strategy IS-1
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 10-3
6/24/03 Heidersbach 030204-Sect 10
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Corrosion Rates
Increasing
temperature has a
marked effect on the
corrosion rates in an
amine system for
removing acid gases
(reference only)
G9900876

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 10-4
6/24/03 Heidersbach 030204-Sect 10
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Corrosion Related to Turbulence and


Velocity in Amine Piping
PR 9801731-1, 2

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 10-5
6/24/03 Heidersbach 030204-Sect 10
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Stainless Steel Equipment Items in


Amine Plants
G9801103-A

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 10-6
6/24/03 Heidersbach 030204-Sect 10
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Amine Stress Corrosion Cracking


Characteristics
Actual cracking agents are contaminants in amine
Occurs at non-stress relieved carbon steel welds
Occurs at temperatures above 100F and probability
of cracking increases as temperature increases
Cracks are intergranular and difficult to distinguish
from other potential cracking mechanisms in DEA
plants (e.g., wet H2S, SSC) (can tell difference
metallographically)
Cracking is more likely in MEA than in DEA or MDEA

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 10-7
6/24/03 Heidersbach 030204-Sect 10
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Amine Stress Corrosion Cracking


(Contd)
Prevention
Stress relieve all carbon steel welds including attachment
welds (socket and seal welds in DEA/MDEA <140F and
fresh amine storage tanks are the only exceptions)
External welds can cause ID cracking
Upgrade to stainless steel (watch chloride SCC)
See API 945 for additional information
Inspection
See Inspection Strategy IS-10
WFMT and/or shear wave UT are typical detection
methods, but SWUT is not very effective
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 10-8
6/24/03 Heidersbach 030204-Sect 10
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Stress Relief Guidelines for Newly


Constructed Amine Plants
Stress Relieve?
MEA DEA and MDEA
Absorber Yes Yes
Regenerator Yes Yes
Pressure Vessels in Main Circulating Loop1 Yes Yes
Butt Welded Piping Yes Yes
Socket and Seal Welds Yes >140F
Storage Tanks
Shell No No
Heating Coils Yes Yes
Notes:
1. External welds often create through-wall stresses sufficient to cause ID
SCC. Hence, any external weld should be treated as an internal one.
2. Operators of existing plants can consult with ERTC Materials and
Equipment Engineering to formulate long-term plans on how to bring
their plants up to these new standards.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 10-9
6/24/03 Heidersbach 030204-Sect 10
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Inspection Requirements for


Non-PWHTd Carbon Steel Piping
Temperature MEA DEA, MDEA
Below 140F 20% of All Butt Welds by Not Required
SWUT Every 5 Years
Above 140F As Above 20% of All Butt Welds by
SWUT Every 5 Years

Notes:
1. PWHTd carbon steel and 300 series SS piping do not require special
inspection for amine SCC.
2. Over time, inspect the majority of all the welds in each circuit. It is not
required that all the welds be inspected as long as a representative
number in each circuit is inspected.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 10-10
6/24/03 Heidersbach 030204-Sect 10
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Amine SCC:


El Segundo C-1500 Lean DEA Line

ID View of the Pipe With Cracks Branchy Transgranular Cracks Indicate


Coming From the Welded Threadolet Stress Corrosion Cracking
Magnification 200x
G9900101a, G9900101b

Non-Stress Relieved Welded


Threadolet Small Welds Can Result RT9804150 - P.W. Hunter, El Segundo C-1500 Lean DEA
in Through-Wall Amine SCC Line - Material Failure Analysis 01/01/1901

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 10-11
6/24/03 Heidersbach 030204-Sect 10
Carbonate SCC

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 11-1
6/24/03 Heidersbach 030204-Sect 11
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Carbonate Stress Corrosion Cracking


Characteristics
Usually occurs at non-stress relieved carbon steel welds
Can occur outside weld areas if high stresses exist (e.g., poorly
supported pipe)
Usually requires carbonate (CO3=) greater than 400 ppm and pH
greater than 8.5
FCC main fractionator overhead systems are the only location
where carbonate cracking is known to occur
Increasing the nitrogen content of the feed will increase the
likelihood of carbonate cracking
Cracks are branchy, intergranular, and oxide-scale filled
Has occurred in all refineries except Salt Lake
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 11-2
6/24/03 Heidersbach 030204-Sect 11
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Carbonate Stress Corrosion Cracking


(Contd)
Prevention
Piping
Stress relieve WELL (use higher stress relief temperatures than
typically used for CS)
Upgrade to 304L or 316L
Ensure adequate support and proper fit-up to minimize stresses
outside weld areas
Vessels
Stress relieve WELL
Stainless cladding
Epoxy coatings
Inspection
Shear wave UT for piping (detection and sizing is difficult)
WFMT for vessels
See Inspection Strategy IS-12 Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 11-3
6/24/03 Heidersbach 030204-Sect 11
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Carbonate SCC (pH Versus


Carbonate Ion)
G9900874

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 11-4
6/24/03 Heidersbach 030204-Sect 11
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Generic Fractionator Overhead System Showing


Areas Most Subject to Carbonate Cracking
G9900866

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 11-5
6/24/03 Heidersbach 030204-Sect 11
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

El Segundo Fractionator Overhead

El Segundo Fractionator Overhead Showing Equipment Repaired


G9900867

or Replaced Due to Carbonate SCC or Wet H2S Cracking Confidential


Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 11-6
6/24/03 Heidersbach 030204-Sect 11
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Company Requirements for Postweld


Heat Treatment
1 1
Materials and Service Vessels & Tanks Large Piping Small Piping
Carbon Steel
2
Hot Caustic Yes Yes Yes
2
Cold Caustic Heating Coils Only Only if Traced Only if Traced
MEA Over 100F Yes Yes Yes
MEA Below 100F Note 3 Only if Traced Only if Traced
DEA & MDEA Over 100F Yes Yes No
DEA & MDEA Below 100F Note 3 Only if Traced No
Potassium Carbonate Over 150F Yes Yes No
Potassium Carbonate Below 150F Note 3 Only if Traced No
4 4
Carbonate in FCC Plants Yes Yes Yes
Anhydrous Ammonia Consult M&E Eng. Consult M&E Eng. No
5
Sour Service Yes Seamless Pipe No No
Welded Pipe
Consult M&E Eng.
6
Other Per ASME VIII Per B31.3 or B31.1 No
Carbon Moly Steel Same as Carbon Steel But Same as Carbon Steel But Same as Carbon Steel
Do Not Use Do Not Use But Do Not Use
7
Chrome Moly Steels Yes, per ASME Code & MS Yes, per B31.3/31.1 and Yes, per B31.3/31.1 and
Specifications PIM-MS-4772 PIM-MS-4772
Stainless Steel
8 8 8
Types 304 & 316 No No No
Types 304L, 316L, 321 Consider on a Case-by- Generally Yes, But Not for Only in Known Chloride
Case Basis; Also See Services Below 150F, and Cracking Service
Comments Under Large Not for Chemical Plants Except PWHT is
Piping Where Sensitization is a Generally Performed to
Greater Concern First Block Valve
9
Type 347 <3/4 Same as 304L, <3/4 Same as 304L, Only in Known Chloride
316L & 321 316L & 321 Cracking Service
>3/4 Consult M&E Eng. >3/4 Consult M&E Eng. Except PWHT is
Generally Performed to
First Block Valve
Alloy 20, Incoloy & Inconel Alloys Consult M&E Eng. Consult M&E Eng. Consult M&E Eng.

Notes:
1. Large piping is butt welded pipe 2 & larger. Small piping is socket welded pipe 1-1/2 & smaller.
2. Hot caustic is over 140F for concentrations below 30% (wt) and over 110F for concentrations above 30% (wt).
3. PWHT on-plot vessels. For tanks, PWHT is required for the heating coils only.
4. PWHT all FCC main fractionator overhead piping at 1150-1250F. Furnace PWHt is preferred to obtain the lowest residual
stresses; however, if local PWHT is used, special consideration should be given to the PWHT procedure including
heating band width, insulation thickness, and runout length of insulation. See Local PWHT Band Width for Pipe Welds
in Section 151.
5. PWHT is strongly recommended because it has been shown to reduce wet H2S cracking even though it may not always
eliminate it.
6. For other environments such as HF and nitrate, see the Corrosion Prevention Manual for specific recommendations
regarding PWHt.
7. Chevron MS specifications are more restrictive than the Codes.
8. The use of standard carbon grades (0.08C max.) of 304 and 316 SS is discouraged because of the sensitization which
can occur either during welding or other heating in the 800-1600F temperature range (i.e., PWHT). Sensitization
compromises the intergranular corrosion resistance and cracking resistance of the material.
9. Tpe 347 SS can be sensitive to cracking during PWHT but is less sensitive for thicknesses of and under because of
less restraint. Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 11-7
6/24/03 Heidersbach 030204-Sect 11
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Local PWHT of Piping Welds


G9900865

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 11-8
6/24/03 Heidersbach 030204-Sect 11
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Band Width and Thermocouple


Requirements
Local PWHT of Piping <0.5" Thick in Horizontal Piping Runs
BW = 2 x {[2.06 x (R x t)1/2] + 1}
Where: BW = Band Width
t = Pipe Wall Thickness
R = Pipe Radius to Mid-Wall
Nominal Pipe Minimum Heating Minimum Number of Minimum Number of
Size, Inches1 Band Width, Inches Control Thermocouples Monitoring Thermocouples4
2 3

3/4 - 1 4 1 1
1-1/2 - 3 6 1 1
4-6 8 1 1
12 10 2 1
18 11 2 2
24 12 2 2
26 14 2 2
30 14 2 2

Notes:
1. This table covers extra strong (<1/2-inch) and thinner pipe schedules.
2. Two inches of ceramic fiber insulation (one or two layers) should be used to cover the resistance
heating band and adjacent pipe for a minimum of 9 inches on each side of the band.
3. Locate the control thermocouples on the center of the weld with the first at 12 o'clock and the
second (where required) at 6 o'clock.
4. Locate the monitoring thermocouples 90 degrees from the control thermocouples.
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 11-9
6/24/03 Heidersbach 030204-Sect 11
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Carbonate Cracking: El Segundo


Main Fractionator Overhead Piping
Discharge of accumulator pumps
Several leaks were found
Piping was stress relieved
Installed in 1988 missing pipe
supports
Inadequate support gave external
applied stresses of 7 ksi
Total stress at point of failure 12 ksi
10 ksi minimum needed
Leaks were found on discharge
piping of two other pumps in
January of the same year, also
replaced in 1988
RT0112702 P.J. Risse, El Segundo FCC Main
Fractionator Overhead Leaks in P-101/P-101A Discharge
Piping Carbonate Stress-Corrosion Cracking Material
Failure Analysis 11/30/1995 Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 11-10
6/24/03 Heidersbach 030204-Sect 11
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Carbonate Cracking: El Segundo


Main Fractionator Overhead Piping

200X

8X

RT0112702 P.J. Risse, El Segundo FCC Main


Fractionator Overhead Leaks in P-101/P-101A
Discharge Piping Carbonate Stress-Corrosion
Cracking Material Failure Analysis 11/30/1995 500X
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 11-11
6/24/03 Heidersbach 030204-Sect 11
Caustic SCC

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 12-1
6/24/03 Heidersbach 030204-Sect 12
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Caustic Stress Corrosion Cracking


Characteristics
Carbon and stainless steel are susceptible wherever
residual or applied stresses are high and caustic is
present
Occurs at temperatures above about 110F (carbon steel)
or 200F (stainless steel)
In FCCs, the main area of concern is tubes of process
steam generators where caustic is allowed to concentrate
and heat-traced caustic lines in the caustic treating
section of the plant

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 12-2
6/24/03 Heidersbach 030204-Sect 12
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Caustic Stress Corrosion Cracking


(Contd)
Prevention
Proper design of steam generators to provide complete
submergence of tubes is the best prevention
Stress relieve U-tubes -- this reduces risk, does not
eliminate it
Carefully review proposed changes to BFW treatments
with respect to increasing risk of caustic cracking
Stress relieve all heat-traced caustic lines
Inspection
See Inspection Strategy IS-13
MT, PT, and shear wave UT are common detection
methods, eddy current tubes
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 12-3
6/24/03 Heidersbach 030204-Sect 12
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Relation of Temperature and Concentration


of NaOH to Cause Cracking NACE DATA
G9900884+

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 12-4
6/24/03 Heidersbach 030204-Sect 12
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Recommended Limits of CS
and SS in NaOH Service
Temperature, F
240
Area C Area C CS and
220 SS Not Adequate

200

180

160
Area B Stress Relief Recommended

140

120

100 Area A Stress Relief Not Required

80
0 10 20 30 40 50
G9900895

Wt % NaOH
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 12-5
6/24/03 Heidersbach 030204-Sect 12
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Corrosion of Carbon Steel by


Sodium Hydroxide
G9900880

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 12-6
6/24/03 Heidersbach 030204-Sect 12
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Caustic SCC:


Richmond Caustic Regenerator
Trays repeatedly fell even after
being tack welded into place
Shell and internals are Monel
10-12 wt % caustic at 250F
Tray weld cracking caused by
overloading
Bubble cap cracking caused by
caustic SCC
Bubble caps were not annealed as
required for Monel in caustic
service
El Segundo had same problem,
RT0107749 J.D. Petrozziello, Failed No. 1 Tray in
C-270 Caustic Regenerator - Richmond FCC solution was to change to sieve
10/05/1983 trays
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 12-7
6/24/03 Heidersbach 030204-Sect 12
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Caustic SCC: El Segundo Fractionator


Bottoms Recycle Cooler Floating Head Bolts
Cracks found during
inspection of K-Monel bolts
Shellside 150# steam
Bolts met fabrication
requirements
Threads formed before heat
treatment
Heat treatment is annealed
and aged (not cold formed
and aged)
Maximum hardness Rockwell
C32
Following maximum torque
RT0112706 J.E. Price, Cracked Floating Head Bolts - E102A
guidelines should prevent
FCC Fractionator Bottoms Recycle Cooler 02/18/1987 repeat failures
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 12-8
6/24/03 Heidersbach 030204-Sect 12
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Caustic SCC


El Segundo Fractionator Bottoms
Recycle Cooler Floating Head Bolts

Cracks
Initiating at
Threads (9X)

RT0112706 - J.E. Price,


Cracked Floating
Head Bolts - E102A
FCC Fractionator
Bottoms Recycle
Cooler 02/18/1987

Intergranular
Scale Filled
Cracks (150X)
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 12-9
6/24/03 Heidersbach 030204-Sect 12
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Types of Stress Corrosion Cracking That


Occur in Refineries
Type Susceptible Alloys Conditions for Cracking Fixes
Chloride 300 Series Stainless Steels A. Evaporative Heat Flux Which Produces Cl- A. Stress Relieve
Deposits B. Use Ferritic Stainless Steels
B. Usually Above 140-150F C. Use High Nickel (>30-40%) Alloys, Titanium,
C. Low pH is Bad Copper Alloys
Polythionic 300 Series Stainless Steels A. Shutdowns A. Use 321 SS or 347 SS
(Sensitized) B. Water and Sulfide Scale B. Shutdown Precautions: (1) Soda Ash Wash
and (2) N2 Blanket
Caustic 300 Series Stainless Steels and A. Hot Deposits A. Stress Relief
Carbon Steel B. Use Monel
MEA, DEA, Carbon Steel A. Usually >140F A. Stress Relief
and MDEA B. MEA More Severe Than DEA, MDEA B. Use Stainless Steel
Sulfide Carbon Steel, Alloy Steels, and A. H2O and H2S A. Substitute Resistant Alloy
Hardenable Stainless Steels B. Hardness >RC 22 or 90 KSI YS B. Control Hardness
-
C. NH3 (High pH) + CN- Are Very Bad C. Remove CN With Polysulfide
Ammonia Brasses (Cu-Zn) A. Ammonia + O2 + H2O A. Remove O2 and Use Acid Washes at
Shutdown
B. Use Alloys Without Zinc (e.g., Cu-Ni)
Carbonate Carbon Steel A. FCC Overhead; Not Well Understood A. Stress Relieve Well
B. Narrow Electrochemical Potential Ranges B. Upgrade to 304L, 316L
Moves With Temp., pH
Wet H2S Carbon Steel Vessels; Not A. H2S Level >50 ppm and H2O A. Use Coating or Cladding
Seamless Pipe B. More Likely if Cyanides or Ammonia Present
-
B. Remove CN With Polysulfide

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 12-10
6/24/03 Heidersbach 030204-Sect 12
Corrosion and SCC by Chlorides

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-1
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Chloride Corrosion and Pitting


Characteristics
Can cause severe corrosion and pitting of CS, 300 Series SS,
and even high alloys
Usually occurs on heat exchange surfaces at initial condensation
In FCCs, primary area of concern is fractionator overhead
system exchangers
Also seen behind packing in cool portions of fractionator
Prevention
Make sure 25% of water remains unvaporized at the injection
points
Alloy up should not be required in properly operated
fractionator overhead systems
Inspection
See Inspection Strategy IS-6
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-2
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Change in Water pH Upon Evaporation/


Condensation
G9900898

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-3
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Carbon Steel in Unagitated


Hydrochloric Acid
Corrosion Rate of
Steel, mpy
1000
900
800
700 pH Versus
600 Acid Concentration
500 HCl
pH Concentration
400
0 3.5%
1 0.35%
300 2 0.035%
3 35 ppm
4 3.5 ppm
200

100
90
80
70
60 Hydrogen
Evolution
50 Begins
40

30

100F
20

75F
10
0 1 2 3 4 5 6
Confidential
pH Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-4
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Alloys for Hydrochloric Acid Service


Materials in Numbered Zones Have Reported Corrosion Rates of <20 mpy

ZONE 1 ALL ZONES (Including 5)


Temperature,
C CN-7M (1) (3) (6) Platinum
120 400 Alloy (2) (3) (6) Tantalum
Boiling Point Copper (2) (3) (6) Silver (3) (6)
100 Nickel 200 (2) (3) (6) Zirconium (3) (6)
90 Silicon Bronze (2) (3) (6) B-2 Alloy (3) (6)
4 5 Silicon Cast Iron (7) Molybdenum (3) (6)
80 Tungsten
70C = 158F 70 Titanium (Gr. 7) NOTES:
Titanium (Gr. 2) (4) 1. <2% at 25C
60 2. No Air
3 ZONE 2 3. No FeCl3 or CuCl2
50
2 Silicon Bronze (2) (6) 4. <10% at 25C
40 Silicon Cast Iron (7) 5. <5% at B.P.
1 6. No Chorine
30
ZONE 3 7. Cr-Mo Alloy
20 Silicon Cast Iron (7) 8. <0.05% Concentration
0 10 20 30 40
Concentration HCl, % ZONE 4
400 Alloy (2) (3) (8)
Tungsten
Titanium (Gr. 7) (5)

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-5
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Resistance to Chloride Pitting


Pitting Index
(% Cr + 3.3 x % Mo + 16 x % N)

(Hastelloy C-276), (Inconel 625), (Hastelloy C-22) 50-70

Hastelloy G, AL-6X, Carpenter 20 Mo-6, Duplex 2507 40-43

Increasing Resistance
904L, Duplex 2205 34-35

Incoloy 825, Alloy 20 Cb-3 28-31

Incoloy 800, (Inconel 600), 316L SS 16-25

304L SS, 321 SS 17-18

410 SS, 430 SS 12-17

Carbon Steel and Cr-Mo Steels 0-5

Note: Pitting Index Not Valid for Alloys in Parentheses


Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-6
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Chloride Stress Corrosion Cracking


Characteristics
Affects only austenitic stainless steels -- primarily
300 Series
Occurs primarily at slightly elevated temperatures --
only sensitized stainless will crack below 140F
In FCCs, primary concern area is 300 Series SS
exchanger tubes in the fractionator overhead
system -- chlorides can come from ammonium
chloride on the process side or chlorides in cooling
water
Cracks are tight, transgranular, and branchy
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-7
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Chloride Stress Corrosion Cracking


(Contd)
Prevention
Inject plenty of water ahead of exchangers (at least
25% should remain in liquid phase)
Don't throttle cooling water -- keep velocities >5 fps
and tube skin temperatures <140F
Alloy up
Stress relief helps but totally effective stress relief of
exchanger bundles is difficult
Inspection
See Inspection Strategy IS-14
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-8
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Test Data of Different Stainless Steels

A Compilation of
Practical Experience
and Laboratory
Test Data of
Different Stainless
Steels in Neutral
Aerated Aqueous
Chloride Environments

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-9
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Effect of Nickel Content on Chloride Cracking Susceptibility


of Cr-Ni Alloys in Boiling Magnesium Chloride
Several commercial alloys are shown below; resistance to cracking
increases with nickel content.

Increasing
Resistance to
Cracking
G9700551

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-10
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Appearance of Chloride Stress Corrosion


Cracking

Typical Chloride Stress Corrosion Cracking in Austenitic


PR 950625-6

Stainless Steel Cracks Are Transgranular and Branchy


Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-11
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Chloride Corrosion:


Richmond C-90 Fractionator
Unexpected because column normally runs dry
(25-30F above dew point at top), but
Fractionator was exposed to water for 35 days
during 7/99 unplanned shutdown while steaming to
relief
Possible that salts hydrated on the run causing
some of the damage and majority of damage done
when salts dissolved during the 7/99 shutdown
Dissimilar materials (galvanic corrosion) also
contributed to high corrosion rates

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-12
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Chloride Corrosion:


Richmond C-90 Fractionator
Corrosion primarily in shell of packed Beds #7 and
#8, plus tray bottoms and troughs, and sidecut draw
piping
Shell often showed Zorro pattern of grooves due
to T-410 SS packing pressed onto CS wall, acidic
liquid dripping over it
Packing itself was thinned and pitted
Severe pitting of T-410 SS tray bottoms and trough
weld areas
CS draw piping was thinned and pitted over lower
portion
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-13
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Chloride Corrosion:


Richmond C-90 Fractionator
Fixes
Improved weld details and covered bimetallic welds
with Belzona 1591
Restored most wasted areas (also conducted a FFS)
Installed a trial patch of Belzona 1591 under the
packing
Expect more repair needed in future shutdowns

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-14
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Chloride Corrosion:


Richmond C-90 Fractionator

Portion of the Package Material From Bed #7 and a


G010171

Segment of the Drip Tube From Bed #8 Confidential


Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-15
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Chloride Corrosion:


Richmond C-90 Fractionator

Close Up View of the


Perforation Drip Tube is
Vertically Oriented

Reverse Side Close Up View


Corrosion Patterns
Characteristic of Hot
G010174

NH4Cl Corrosion
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-16
6/24/03 Heidersbach 030204-Sect 13
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Chloride Corrosion:


Richmond C-90 Fractionator
G010173

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 13-17
6/24/03 Heidersbach 030204-Sect 13
Polythionic SCC

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 14-1
6/24/03 Heidersbach 030204-Sect 14
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Polythionic Acid SCC


Characteristics
Affects only in sensitized 300 Series stainless steel exposed to
sulfur-derived acids. In FCC regenerators these acids can form
from mixing of water and SO2 during shutdowns.
For equipment running hot and dry it's only a concern at
shutdown, but it can occur quickly
In FCCs, areas of greatest concern are regenerator internals
and stainless equipment just external to the regenerator (e.g.,
bellows)

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 14-2
6/24/03 Heidersbach 030204-Sect 14
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Polythionic Acid SCC (Contd)


Prevention
Not a problem if no water or oxygen present
For regenerator internals, recovery from sensitization after long
exposure and ChevronTexaco's practice of not water washing
at shutdown helps limit incidents
If the regenerator is water washed, add soda ash to the wash
water (see NACE RP0170 for soda ash wash guidelines)
Stainless equipment external to the regenerator has cracked
(e.g., slide valves and expansion bellows). Materials and
designs changes, such as elimination of wet steam purging and
use of cold wall valves, has addressed these problems.
Inspection
See Inspection Strategy IS-15
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 14-3
6/24/03 Heidersbach 030204-Sect 14
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Time-Temperature-Sensitization (TTS) Curves


for Common 300-Series Stainless Steels

Curves correspond
to susceptibility to
intergranular
corrosion and
polythionic acid
SCC. For
conditions within
the loop for a
given alloy, it will
be sensitized.
Outside its loop,
the alloy will not be
sensitized. These
are average, typical
curves to be used
for guideline
purposes only.
G9601605

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 14-4
6/24/03 Heidersbach 030204-Sect 14
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Sensitization

321 and 347 Stainless Steels Can Sensitize at FCC Operating Temperatures
304L and 316L Can Easily Sensitize at FCC Operating Temperatures
304H, 304 and 316 Can Sensitize From Welding or PWHT Alone
G9800578

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 14-5
6/24/03 Heidersbach 030204-Sect 14
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Sensitization

Not Sensitized, 304 SS Sensitized, 304 SS


Magnification 500X 10,000 Hours at 1000F
Magnification 500X

Note: Field Metallography Can Detect Sensitization


PR 950625-21, 22

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 14-6
6/24/03 Heidersbach 030204-Sect 14
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Polythionic Acid Stress Corrosion


Cracking
PR 950625-13

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 14-7
6/24/03 Heidersbach 030204-Sect 14
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Intergranular Corrosion, 316 SS


PR 950625-20

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 14-8
6/24/03 Heidersbach 030204-Sect 14
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Polythionic Stress Corrosion Cracking:


Irvings FCC Regenerated Catalyst Slide Valve

304H

Most cracking occurred inside the bonnet area near the


seam weld. No refractory lining is present in the bonnet
area. Some cracking was also noted near the inlet flange
attachment underneath refractory lining.
G9900861

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 14-9
6/24/03 Heidersbach 030204-Sect 14
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Polythionic Stress Corrosion


Cracking: Richmond FCC Flue Gas System

Located on waste heat steam


generator drainpipes
304 SS expansion bellows
RT9922564 E.E. Jolly, Richmond Refinery E-60
Replaced with Inconel 625LCF Waste Heat Steam Generator Drainpipe Expansion
Bellows SCC Failure - Material Failure Analysis
G9900980a,b

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 14-10
6/24/03 Heidersbach 030204-Sect 14
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Polythionic Stress Corrosion


Cracking: El Segundo Feed Nozzles
12 low alloy nozzles total
304 stainless steel liners
Multiple severe cracking found
along liners
Regenerated catalyst line
between reactor and slide valve
Nozzles carry hydrocarbon feed
at 600-700F to catalyst at 1300F
Steam fed into nozzles when not
in use
Cracks found in 1981 and 1983
RT0112711 C.L. Vincent, Feed Nozzle Insert Failures in same location at
Polythionic SCC El Segundo FCC 08/30/1983
Richmond were due to catalyst
RT0112708 K.K. Kirkham, Failed Feed Nozzle
erosion
G9900980a,b

Insert, El Segundo FCC 09/14/1981

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 14-11
6/24/03 Heidersbach 030204-Sect 14
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Stress Corrosion Cracking (SCC)

Susceptible Tensile
These principles apply to Material Stress
all SCC mechanisms
including:
Polythionic acid SCC SCC
Chloride SCC
Caustic SCC
Amine SCC
Carbonate SCC Aggressive
Environment

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 14-12
6/24/03 Heidersbach 030204-Sect 14
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Stress Corrosion Cracking - Class


Exercise

Carbonate Chloride Polythionic Caustic Amine


Susceptible Metals

Environment Needed
(Temperatures and
Chemicals)

Stress Sources (List


All Possibilities)

Fixes

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 14-13
6/24/03 Heidersbach 030204-Sect 14
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

What is Reliability?
Good Design
Quality Construction
Excellence in Operations
Excellence in Maintenance
Surveillance/Analysis

Reliability is the Sum of These Five Elements


Therefore, We All Own Reliability

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 14-14
6/24/03 Heidersbach 030204-Sect 14
14-15
6/24/03 Heidersbach 030204-Sect 14
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Types of Stress Corrosion Cracking That


Occur in Refineries
Type Susceptible Alloys Conditions for Cracking Fixes
Chloride 300 Series Stainless Steels A. Evaporative Heat Flux Which Produces Cl- A. Stress Relieve
Deposits B. Use Ferritic Stainless Steels
B. Usually Above 140-150F C. Use High Nickel (>30-40%) Alloys, Titanium,
C. Low pH is Bad Copper Alloys
Polythionic 300 Series Stainless Steels A. Shutdowns A. Use 321 SS or 347 SS
(Sensitized) B. Water and Sulfide Scale B. Shutdown Precautions: (1) Soda Ash Wash
and (2) N2 Blanket
Caustic 300 Series Stainless Steels and A. Hot Deposits A. Stress Relief
Carbon Steel B. Use Monel
MEA, DEA, Carbon Steel A. Usually >140F A. Stress Relief
and MDEA B. MEA More Severe Than DEA, MDEA B. Use Stainless Steel
Sulfide Carbon Steel, Alloy Steels, and A. H2O and H2S A. Substitute Resistant Alloy
Hardenable Stainless Steels B. Hardness >RC 22 or 90 KSI YS B. Control Hardness
-
C. NH3 (High pH) + CN- Are Very Bad C. Remove CN With Polysulfide
Ammonia Brasses (Cu-Zn) A. Ammonia + O2 + H2O A. Remove O2 and Use Acid Washes at
Shutdown
B. Use Alloys Without Zinc (e.g., Cu-Ni)
Carbonate Carbon Steel A. FCC Overhead; Not Well Understood A. Stress Relieve Well
B. Narrow Electrochemical Potential Ranges B. Upgrade to 304L, 316L
Moves With Temp., pH
Wet H2S Carbon Steel Vessels; Not A. H2S Level >50 ppm and H2O A. Use Coating or Cladding
Seamless Pipe B. More Likely if Cyanides or Ammonia Present
-
B. Remove CN With Polysulfide

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 14-16
6/24/03 Heidersbach 030204-Sect 14
Focus on selecting 300
series stainless steels
Research and Technology

An M&EE Newsletter Supplement June, 1998


for high temperature
refinery services

O ne aspect of refinery metallurgy is selecting How does this sensitization occur? When certain stainless
the correct 300-series stainless steel for high steels are exposed to temperatures greater than about 700F, the
temperature refinery environments. The dissolved carbon combines with the chrome in the stainless steel
primary corrosion concerns in high to form chrome carbides. This happens most rapidly at the grain
temperature refinery services are sulfidation corrosion boundaries of the material (see Figure 3). So, immediately next
and naphthenic acid corrosion. to the grain boundaries, the stainless steel will be depleted in
Ironically, virtually any type of 300-series stainless steel will chrome and since its the chrome that gives the stainless steel
resist sulfidation corrosiontypes 304, 304L, 316, 316L, 317, its corrosion resistance, we now have a zone thats sensitive
317L, 321, and 347 all do a good job. This is not so for naphthenic to corrosion or cracking.
acid corrosion (NAC) where only the molybdenum containing Sensitization is of no real concern at operating temperature
stainless steels (e.g., 316L, 317L) provide corrosion resistance since strength, mechanical properties and high temperature
more on NAC later. corrosion resistance are really not effected. However, sensitized
So what is the problem? The problem is that during high stainless is sensitive to corrosion or cracking if it is exposed to
temperature service, some of the above stainless steels will moist conditions.
become sensitized, which means they become susceptible to Moist conditions are almost always present during shutdowns
corrosion and cracking during shutdown conditions. unless special precautions are takenand even then, it is difficult
to keep conditions dry all the time. To make matters worse, the
plants we are concerned with have sulfur
Figure 3 contamination. The combination of sulfur scales
Sensitization (or sulfur coming out of insulation or refractory)
and water forms sulfur acids, also called polythionic
Carbides acids. These acids can cause very rapid (like minutes
or hours) cracking of sensitized stainless steel.
So what can we do to avoid sensitizing the
stainless steel? One obvious step is to keep the
carbon in the stainless steel low, so that it will not
combine as much with the chrome. This is the idea
Chromium
Depleted Zones behind the low carbon, or L grades of stainless,
which limit the carbon content to less than 0.03%.
Unfortunately, while this is good enough to avoid
sensitization during the short time of welding, it
Grain is not good enough to avoid sensitization if the
Boundaries
material is exposed to high temperatures (>700F)
for many hundreds of hours.
This is where the stabilized stainless steels come
321 and 347 Stainless Steels DO NOT Sensitize at Operating Temperature
(Furnace Tubes = Possible Exception) in. Types 321 and 347 are stabilized stainless
304L and 316L Can Sensitize at Operating Temperatures steels. Stabilized stainless steels add either titanium
or columbium (also called niobium) to their chemistry
304 and 316 Can Sensitize From Welding or PWHT

Figure 3: Sensitization of stainless steels Proprietary Information


Authorized Use Only
carbon left at all. So you can then expose the stainless to high
temperatures forever and it will not sensitize. Testing by Amoco
showed that thermal stabilization worked perfectly for Type
347 SS and, to the extent you can trust any metallurgist, we
guarantee that properly thermally stabilized 347 stainless will
never sensitize at any time-temperature combination.
Why do we hold for 4 hours at 1650F to thermally stabilize?
Because we have always done it that way! We actually think
that a 1650F +/- 25F temperature hold for something less than
4 hours would work, but we do not have the data to prove it. If
a vendor claims they can do it effectively at less time, we need
to insist on corrosion test data to backup the claim.
What if we weld in the fielddo we need to restabilize the
furnace tube after welding? Unfortunately, yes. The very high
temperatures of welding puts all the carbides back into solution,
so we need to start over and thermally stabilize the weld region.
The set-up (heating coils, insulation, etc.) is just the same as for
stress relieving.
Does thermally stabilizing the furnace tubes hurt the creep
strength of the stainless? Good questionwe do not think so.
Literature data, vendor data, and some fairly long recent testing
Figure 4: Polythionic acid stress corrosion cracking
of sensitized stainless steel by Chevron on type 347 furnace tube material indicates that any
mix. Type 321 stainless uses titanium while Type 347 uses reduction in creep strength is very small. To be conservative, we
columbium. When heated, the carbon in the stainless will combine currently suggest for design purposes to reduce the Larsen Miller
with the titanium or columbium more readily than it will with Parameter in the API 530 design curves by 1%. This means that
the chrome. Therefore the chrome is left alone, the material does the design metal temperature for a typical hydroprocessing furnace
not sensitize (it is stable), and everyone is happy. for a given tube thickness would be reduced by about 25F. (Heck,
you shouldnt be running the furnace that close to the edge
Well, almost. It turns out that given enough time at a fairly
anyway!)
narrow high temperature band, even Type 321 will sensitizeand
even the super sensitization resistant Type 347 will sensitize at What about naphthenic acid corrosion (NAC)? Excluding NAC,
very long times (like years) in the right high temperature exposure we could summarize stainless steel selection for resisting high
band. Is this a practical problem for us? temperature sulfidation corrosion as follows:

In most cases, no. Type 321 SS will not sensitize below about Temperature Steel Selection
850F. Piping rarely sees temperatures above 800F, and we have Up to ~700F Use 304L; 316L or 317L are also OK
found that 321 SS does well for hydroprocessing piping, for 700 to ~850F Use 321
example. This is fortunate, because 347 is more expensive, and
Above 850F, Use thermally stabilized 347SS
more difficult to fabricate, especially in thicknesses above about
and particularly
3/4". But for items like furnace tubes, which do see metal
furnace tubes
temperatures above 850F, we must go to the best material for
sensitization resistance (i.e., 347). Unfortunately, even 347 could But when NAC is an issue the above rules dont apply. We must
sensitize at furnace operating conditions. Though rare, we have use the molybdenum containing stainless steels, preferably 317L,
experienced polythionic stress corrosion cracking of 347 furnace to resist NAC attack regardless of operating temperature and
tubes. accept the risk of polythionic acid cracking at shutdowns as the
lesser of two evils. Shutdown precautions to prevent moisture
But dont give upwe have one more trick. If we thermally
contact become the primary defense against polythionic acid
stabilize Type 347 stainless steel furnace tubesby heating
cracking in this case.
them to 1650F for 4 hoursthen the mobility of the carbon
is such that it just zips around like crazy, searching for any For more information on selecting stainless steels for high
available columbium soon there will be essentially no free temperature refinery services, contact any company materials
engineer. They all love this topic.

Proprietary Information Authorized Use Only


Embrittlement Mechanisms

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 15-1
6/24/03 Heidersbach 030204-Sect 15
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Sigma Embrittlement
Characteristics
Occurs rapidly in the ferrite phase of 300 Series SS castings
and weld metal when exposed to 1100-1700F
Can occur in wrought products over very long period of time
Sigma reduces toughness at temperatures below 1000F
Sigma phase makes weld repairs more difficult
316 and 347 SS embrittle more rapidly than 304 SS
Preventing Failures From Sigma Phase Embrittlement
Limit ferrite content of weld metal to 10%
Don't use 300 Series SS castings in 1100-1700F FCC service
(limit ferrite if castings must be used)
Avoid shock loadings when embrittled material is cool (<1000F)
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 15-2
6/24/03 Heidersbach 030204-Sect 15
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

885F Embrittlement
Environment: Operation at temperatures in 750-950F range

Mechanism: Embrittlement of chromium stainless steels


(400 Series)

Effects: Ductile to brittle transition temperature increases


12 Cr trays may be brittle; difficult to straighten
or weld repair

Note: This embrittlement mechanism prevents use of


400 Series stainless steels for pressure-containing
components in the hotter locations of FCC plants
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 15-3
6/24/03 Heidersbach 030204-Sect 15
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Creep Embrittlement
Environment
High temperatures: >800F for 1-1/4 Cr
Stress concentration and high constraint
Located at weld HAZ
Mechanism
Coarse grain region of weld HAZ has poor creep
ductility

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 15-4
6/24/03 Heidersbach 030204-Sect 15
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Creep Embrittlement (Contd)


Effects
Cracks along fusion line of welds with high constraint
Materials selection criteria
Low alloy Cr-Mo steels (<2% Cr) have poor creep
ductility
Low strength (Class 1) and higher purity steel
improves performance
Alloys which have higher creep ductility such as
2-1/4 Cr - 1 Mo are a common upgrade
Creep embrittlement damage to FCCs has been
limited

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 15-5
6/24/03 Heidersbach 030204-Sect 15
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Mechanism
In the coarse grained HAZ of welds, carbide
precipitates in the interior of grains leave the grain
boundaries as the weakest link for creep
When strain is imposed upon this region, all the
deformation occurs at the grain boundaries, so
failure occurs with little macroscopic ductility
The source of strain is usually the relaxation of
residual stresses during PWHT and subsequently
during long-term service at elevated temperature
Strain is concentrated at the susceptible HAZ region
by geometric effects, such as nozzle or attachment
details, at steep weld profile or undercut, etc.
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 15-6
6/24/03 Heidersbach 030204-Sect 15
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Contributing Factors
High strength Class 2 material, or C content 0.15%
or higher
Weld details with high constraint
Heavy sections (naturally high constraint)
Stress concentrations (toes of nozzle and
attachment welds, miters, stiffeners)
Impurity element segregation (similar to temper
embrittlement; modern steels are more resistant)
High residual stress (low temperature PWHT, high
constraint, heavy sections)
Operating temperature 800-1100F Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 15-7
6/24/03 Heidersbach 030204-Sect 15
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Occurrence and Inspection


Most prevalent in Cat Reformers, particularly at
nozzles and supports on hot wall reactors, and long
seam welds in piping
In FCCs, creep embrittlement has occurred at reaction
mix line supports, stiffener rings, and perhaps support
attachments on hot wall reactors
The coarse grain HAZ region will generally not fail if it
is subjected to a simple membrane stress, with no
stress concentration -- piping girth welds and vessel
seam welds have not generally had problems
Piping welds dont need inspection except for miters,
stiffeners, and long seams in 1 Cr and 1-1/4 Cr
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 15-8
6/24/03 Heidersbach 030204-Sect 15
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Repair and New Construction Guidelines


Use lower strength Class 1 materials (piping
always comes this way)
For Class 2 materials or when cracks have been
found and creep embrittlement is suspected:
Use weld joint details with low constraint; including
blend grinding the weld toes
PWHT per ChevronTexaco policy at 1300-1375F
Control the composition of filler metals and
basemetals, particularly the C content

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 15-9
6/24/03 Heidersbach 030204-Sect 15
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Creep Embrittlement:


Pascagoula Rheniformer Reactor

Cracks on a
Cat Reformer
Reactor Inlet
Nozzle

RT0112854 - C.L. Vincent, Creep Embrittelment Cracks R-1501 & R-1502 Hot Wall Rheniformer Reactors Pascagoula

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 15-10
6/24/03 Heidersbach 030204-Sect 15
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Creep Embrittlement:


Pascagoula Rheniformer Reactor

Micrograph of a
Section From the
Boat Sample Near
the End of the
Crack -- Crack is
in the Coarse
Grain Region of
the HAZ

RT0112854 - C. L. Vincent, Creep Embrittelment Cracks R-1501 & R-1502 Hot Wall Rheniformer Reactors Pascagoula
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 15-11
6/24/03 Heidersbach 030204-Sect 15
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Example of Creep Embrittlement:


Pascagoula Rheniformer Reactor

Micrograph of the
HAZ Area From the
Previous Image --
Cracks Are Along
Prior Austenite
Grain Boundaries

RT0112854 - C. L. Vincent, Creep Embrittelment Cracks R-1501 & R-1502 Hot Wall Rheniformer Reactors Pascagoula

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 15-12
6/24/03 Heidersbach 030204-Sect 15
Wet H2S Cracking, Blistering, and
Sulfide Stress Cracking
Senior Analyst and Inspector Training
August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-1
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Wet H2S Cracking


Will first discuss Causes & Effects of wet H2S
cracking, then
Will discuss how we Manage wet H2S cracking (next
section)

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-2
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Wet H2S Effects on Steels


Effects
Hydrogen blistering
Hydrogen-Induced Cracking (HIC) or Stress-
Oriented, Hydrogen-Induced Cracking (SOHIC)
Sulfide Stress Cracking (SSC)
Mechanism
Hydrogen is charged into steel by wet H2S corrosion

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-3
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Wet H2S Effects on Steels (Contd)


Characteristics
Occurs where both liquid water and H2S are present -- 50 ppm
H2S is a commonly accepted threshold but cracking can occur
at lower concentrations
Significant corrosion and metal loss are NOT required for
cracking to occur
Blistering, HIC, and SOHIC are primarily a concern for carbon
steel plate; damage is less common in carbon steel pipe and
tubing and 300 Series SS is immune
Blistering and HIC/SOHIC rarely produce through wall leaks;
however, damage can progress rapidly under the right
conditions
SSC occurs in hard zones of carbon steel welds and in low
alloy and 12 Cr steels, which are inherently hard
SSC often produces through wall cracking -- known to be a
dangerous mechanism Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-4
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Wet H2S Cracking: What Causes It?


Blistering
Hydrogen atoms from corrosion on the surface of
the metal migrate into the metal. They collect at
laminations or other voids.
Hydrogen that collects at internal laminations or
voids combine to form hydrogen molecules. This
dramatically increases the pressure, forming a
blister.
Cracks often grow from the edges of the blister

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-5
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Formation of Hydrogen Blisters by


H2S Corrosion
A clean steel surface normally catalyzes the combination of atomic hydrogen formed by corrosion to molecular
hydrogen, but if H2S is the corrodant, the combination at the metal surface is greatly retarded and atomic hydrogen
diffuses into the steel.

The buildup of an iron sulfide layer at the metal surface slows the rate of corrosion, and even if the steel has an
internal defect, blistering may not occur. However, if cyanides are present along with the H2S, the ferrous sulfide layer
may be dissolved, and hydrogen diffusion continues, creating high internal pressures at the defect, and as the steel
eventually yields, producing blisters.

H0 + H0 + H2 (At Defect) With cyanides present, FeS scale


dissolves:
With FeS layer, reaction FeS + 6CN- Fe (CN)6 + S= ==
slows and blistering may
not occur. and corrosion continues, with
formation of hydrogen blisters.
G9901105

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-6
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Wet H2S Cracking: What Causes It?


(Contd)
Hydrogen-Induced Cracking or Stress-
Oriented, Hydrogen-Induced Cracking (HIC, SOHIC)
HIC or SOHIC also often called stepwise cracking. In all cases the buildup
of hydrogen (as described earlier with blistering) is involved.
Stress plays a role, but the pressure buildup from the hydrogen itself can
create significant stress.
Perspective: Some cross-country pipelines are designed to carry stresses at
90% or more of the material yield strength. Definitely would be a problem in
sour service.
For refinery vessels, ASME Code requires the applied stress to be less than
67% of yield (typically is much less). PWHT (stress relief) will reduce stresses
well below this, but even with PWHT the combined applied, residual welding,
and discontinuity stresses could easily add up to 50% of the yield at (say)
nozzles.
Lab testing and field experience shows that SOHIC can occur at 30% of yield,
definitely at 50% of yield. So, while PWHT is clearly good, we cannot count on
it alone to prevent cracking.
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-7
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Wet H2S Cracking: What Causes It?


(Contd)
Sulfide Stress Cracking (SSC)
SSC is sometimes also called hydrogen stress
cracking or hydrogen embrittlement cracking
SSC is caused by hydrogen getting into high
strength, high hardness (greater than 200 Brinell)
steels
SSC causes rapid, almost immediate, cracking.
Cracking will fade away once it reaches softer metal,
unless by then the crack is so large it will tear the
remaining metal to form a potentially catastrophic
failure

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-8
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Summary of Wet H2S Effects on Steel


G9601611

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-9
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Generic Fractionator Overhead System Showing


Areas Most Subject to Wet H2S Cracking

Note: Sulfide cracking, a form of wet H2S cracking, can occur anywhere in the OVHD and light ends recovery
G9900892

system where high strength, high hardness steels are used (e.g. bolting, valve stems, and compressor
internals). Limit hardness or use resistant alloys.
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-10
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Different Forms of Wet H2S Cracking


G9900875

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-11
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Wet H2S Cracking in


ChevronTexaco FCC Plants
CHEVRONTEXACO INSPECTION RESULTS PER CRTC WET H2S CRACKING DATABASE
Sorted From Highest to Lowest in Cracking Severity
# % With Cracks % With Minor % With No
Vessel/Service Inspected Requiring Weld Repair Cracking Cracks
Interstage KO Drum 10 50% 30% 20%
Sponge Absorber 11 45% 36% 18%
Interstage Cooler 6 33% 33% 33%
Deethanizer 10 30% 40% 30%
Deethanizer Feed Condenser 8 25% 25% 50%
Depropanizer O/H Condenser 12 25% 42% 33%
Debutanizer Accumulator 8 25% 63% 13%
Main Fractionator 5 20% 20% 60%
KO Drum 5 20% 60% 20%
Deethanizer Feed Separator 12 17% 42% 42%
Fractionator Reflux Drum 13 15% 38% 46%
Depropanizer Accumulator 14 14% 29% 57%
Fractionator O/H Condenser 51 14% 22% 65%
Depropanizer 9 11% 67% 22%
Debutanizer O/H Condenser 34 6% 21% 74%
Debutanizer 11 0% 73% 27%
Light Gas Oil Stripper 1 0% 100% 0%
Depropanizer Feed Bottoms 1 0% 100% 0%

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-12
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Effect of Changing Process Conditions


on the Likelihood for Wet H2S Cracking

Crack Initiation
and Growth
Hydrogen Flux Into Steel

Process Upset
New Vessel

Crack Growth

Protective Scale

Time

Source: NACE Corrosion 1998, Paper 394


G9900115

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-13
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Hydrogen Flux in Richmond WWT NH3


Stripper Accumulator Drum V-820
0.70

0.60

0.50
Lost APS
0.40
Hydrogen Flux DC Voltage, mV

0.30
Vapor
0.20

0.10

-0.00

-0.10

-0.20

-0.30
Liquid
-0.40

-0.50
G9801099-4

01/24/97 01/30/97 02/06/97


22:36:59 MM/Day/Year 12:04:06
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-14
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Hydrogen Flux (J) Versus Time for a


Refinery Deethanizer Tower
2.0
1.8
Hydrogen Flux, Moles/cm2/Day

1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0
0 5 10 15 20 25 30 35
Time, Days
(Two Peaks Correlate With Loss of Water Wash, One Peaks Cause Was Not Determined)

Source: NACE Corrosion 1998, Paper 394 Confidential


Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-15
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Wet H2S Inspection Results for


390 PWHTd and 817 Non-PWHTd Vessels

(Significant Cracking is Cracking in Excess of the Corrosion Allowance


Which is .125 Typically)
50
43
Percent Vessels With Cracking

42

40

30

20 15.8

9.7
10

0
PWHT'D Vessels PWHT'D Vessels Non-PWHT'D Non-PWHT'D
w/Cracks w/Sig. Cracks Vessels w/Cracks Vessels w/Sig.
Cracks

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-16
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Percent of Vessels in Wet H2S Service With Cracks


Greater Than 3/8 Deep or 50% of Wall Thickness

Results of 1100 Inspections


3.5
Percent of Vessels With Deep Cracks

3 Fab Defects
Wet H2S Cracks
2.5

1.5

0.5

0
Nozzles Seams Attachments Not Reported
Crack Location

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-17
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Accelerated Preferential Corrosion at Grit-


Blasted Welds in an Amine Regenerator

WFMT Blast Zone

Preferential Corrosion at Grit Blast Area Weld Overlay Repairs

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-18
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Hydrogen Blister and HIC Cracking


A214 CS Amine Regenerator Overhead Fin-Fan Exchanger Tube - El Paso
G9900109

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-19
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Hydrogen Blisters
PR 950625-17

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-20
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Hydrogen Blister With HIC


PR 950625-3

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-21
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

HIC Damage

HIC Damage

SSC + HIC Damage


G010570.jpg, G010571.jpg

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-22
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Comparison of Amine and


Wet H2S Cracking
Amine SCC Wet H2S Cracking
Susceptible Carbon Steel, Both Plate and Carbon Steel Vessels; Rare in
Alloys Pipe Seamless Pipe
Hard Steels and Alloy Steels Are
Susceptible to SSC
Conditions Usually Above 140F for DEA Can Occur at Any Temperature
for Cracking and MDEA; Ambient and Above Encountered in Amine Plants
for MEA; More Likely as Does Not Require Residual Stress
Temperature Increases But Deep Cracking is Very Rare
Residual Stress Required But Outside of Nozzle and Seam
Even Small Welds (e.g., Welds
Attachment Welds) Can Be Is Usually More Severe Where
Susceptible Corrosion is Severe
Cracking PWHT of Carbon Steel Welds is Corrosion Barriers (Cladding,
Prevention Effective Coating) Are the Most Effective
Using Stainless Steel Cladding Prevention; PWHT is Not Effective
or Pipe Provides Immunity Stainless Cladding Provides
Limit Temperature Immunity
Limit Hardness to Avoid SSC

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-23
6/24/03 Heidersbach 030204-Sect 16
Managing Wet H2S Cracking

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-24
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Managing Wet H2S Cracking


We will discuss:
Basic Strategy
Refinery Wet H2S Cracking Program
Materials for Wet H2S
Use of Coatings
Use of Crak Stop
Inspection

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-25
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Managing Wet H2S Cracking:


Basic Strategy
1. Never allow hard welds in sour service, and be very careful of
high strength materials. This is to prevent the dangerous
Sulfide Stress Cracking (SSC) mechanism.
a) Do not allow yield strengths above 90 ksi, or hardness
above HRC 22. Carbon steel welds must be less than
200 Brinell.
b) If higher strength materials are needed, NACE has
standards for materials in sour service. The old NACE
MR01-75 will be replaced by NACE MR0103-2003 for
Downstream applications.
2. Identify the vessels with greatest risk. Focus on those.
a) Very few actual through-wall leaks.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-26
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Managing Wet H2S Cracking:


Basic Strategy (Contd)
3. Wherever possible (especially for new construction), the
highest risk vessels should have 300 series SS clad or weld
overlay. Carefully consider solid Duplex SS vessels.
a) If SS clad or WOL is not possible, then use organic
coatings or thermal spray.
b) If even coatings are not possible, then inspect
frequently and accept higher risk.
4. Use inspection to support the effort, but recognize costs and
limitations. Minimize future inspections but keep people safe.
a) Internal WFMT is effective but costly, can lead to
surprises, and exposes fresh metal to a burst of
hydrogen activity.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-27
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Refining Wet H2S Cracking Program


Based upon industry experience, lab studies, and
ChevronTexaco experience:
One of the largest databases in industry, some 1500
inspection records
CT program consistent with NACE RP 0296-96,
Guideline for Detection, Repair, and Mitigation of
Cracking of Existing Petroleum Refinery Pressure
Vessels in Wet H2S Environments
Core recommendations given on the Wet H2S
Process Flow Diagrams. August 2, 2002, guidelines
are latest (on website). Revising now, with
FENT/MRENT guidance.
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-28
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Refining Wet H2S Cracking Program


(Contd)
PFDs have shaded areas where inspection is recommended, while
inspection is actually discouraged outside the shaded areas. PFDs
identify Category 1 vs. Category 2 vs. Category 3 vessels (very
severe versus moderately severe service vs. low severity).
No special inspection needed at all if the vessel is SS clad
Generally vessels in the shaded areas should be inspected in
6 years. Changing! (See below)
Coated vessels should be inspected in 10 years
Inspection recommendations are:
100% of nozzle welds
100% of previously weld repaired areas
At least 20% of seam welds, focusing on:
Conical transition joints
T-joints
Head to shell welds Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-29
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Refining Wet H2S Cracking Program


(Contd)
Category 2 equipment can get credit for good behavior to
lengthen inspection from 6 to 10 years if:
No history of cracking deeper than 1/8 or longer than 1,
based upon at least two past inspections, and
Process conditions have not changed substantially (see
details in 8/02/02 update), and
The plant SME and materials engineer agree
Changing! See Below
Check website/ERTC for latest guidelines. Managing wet H2S
cracking is difficult and expensive, and we are looking for
improvement opportunities.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-30
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Proposed Changes: Refining Wet H2S


Cracking Program
Strategy:
Focus on Category 1 vessels. Get those in an
inherently safe position where we do not need to rely
on inspection.
Diminish (eliminate) opening Category 2 vessels
simply for wet H2S inspection
Use common sense -- regardless of category, deal
aggressively with vessels with significant cracking

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-31
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Proposed Program Changes (Contd)


Category 1:
WFMT bare vessels within six years (no change). FENT
Meeting: Unless vessel will be coated, use EC assuming we
gain good confidence. If left bare, at least use Crak Stop.
Default position is to internally re-inspect every
6 years, but rigorous MOC process may extend good
actor vessels with external SWUT
PWHT will favor extending
Some especially crack-prone vessels (deethanizers)
will be difficult call; focus on those and develop a
disposition plan
After initial inspection, use EC for internal
inspections
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-32
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Proposed Program Changes (Contd)


Category 2:
After one thorough internal inspection, no further
inspection required for good actor vessels
May use either WFMT or Eddy Current (assumes
FENT endorses EC, and tools/training are in place)
Plant SME may always overrule
Category 3:
Inspection discouraged

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-33
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Proposed Program Changes (Contd)


Cracks longer than 1 or deeper than 1/8 puts the
vessel in bad actor status (no change)
Any uncoated weld repairs shall place the vessel in
bad actor status. If weld repair was coated but not
PWHTd, may still be bad actor status; especially if
in amine or carbonate cracking service.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-34
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Materials for Wet H2S Service


300 Series stainless cladding (or solid stainless) is
the best prevention. Absolutely recommended for
Category 1 vessels.
For bare steel, use good quality (clean) SA-516
normalized carbon steel, with the chemistry
requirements noted on the Wet H2S Process Flow
Diagrams. These requirements are becoming
industry practice and should add little cost.
Avoid nozzles with reinforcing pads (pads interfere
with inspection)

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-35
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Materials for Wet H2S Service (Contd)


HIC-resistant steels with ultra-low sulfur content
and other modifications are not immune to wet H2S
cracking, and the additional cost is rarely justified
PWHT does not prevent wet H2S cracking, but can
significantly reduce the risk of major failures.
Especially important for weld repairs, especially if
not coated.
Remember to limit weld hardness (200 BHN for CS,
215 BHN for low alloy steel)
Follow NACE guidelines for sour service materials

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-36
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Coatings for Wet H2S Service


Since 1991 we coated over 100 vessels. (All
numbers approximate.)
Eight failures, three partial success (coating
basically did the job but not perfect), nine known
successes. The rest have no indication of problems
but not inspected yet.
Four of the failures were Belzona, described below.
Most or all of the non-Belzona failures are believed
to be application-related.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-37
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Wet H2S Coatings (Contd)


Coatings applications (approximate):
Pascagoula: 65 (21 Belzona)
Richmond: 41 (37 Belzona)
Salt Lake: 6
Hawaii: 4 (1 Magnum CR, other Belzona)
El Segundo: 3
Belzona Failures:
SRU: early application, not post-cured
E-1020 Channel and we think #5 H2S column severely
overheated (badly charred)
V-1120: Uncertain, believe damaged during >400F
steamout
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-38
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Coatings Tests Underway


Steamout Testing
Company Coating 300F 365F
Ameron Amercoat 91 (Amercoat V-9, Valspar V-9) Excellent Excellent
Carboline Phenoline 1205 Fair Fair
Sigma Sigma 7409 / 7436 Excellent Excellent
Sigma Sigma 7409 / 7435 / 7436 Excellent Excellent
Belzona 1391 Excellent Excellent
Belzona 1522 Excellent Failed
Belzona 1591 Fair Fair
Curran Curralon Excellent Excellent

Atlas Cell Testing (Three-phase Water, Hydrocarbon, CO2-H2S Gas)


All Passed 180F Atlas Cell Test Except Curralon Not Tested Yet

Amine Exposure Testing - No Results Yet

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-39
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Coatings for Wet H2S Service


Economics drives the decision to use coatings, or not.
Blasting for WFMT inspection leaves the surface suitable for
coatings. Coatings defer the need to inspect (10 years versus
6 years), and the future inspection is simpler.
Typically the entire vessel is not coated, only weld repair areas,
nozzles, and stripes along pressure seams:
Vessels up to 1 thick: Stripe at least 3 each side of weld
centerline
Vessels greater than 1 thick: Stripe at least three times the
vessel thickness each side of weld centerline, up to a
maximum required of 6 each side
See EG-5014, Internal Coatings for Vessels in Wet H2S
Service
See Belzona Coatings Users Guide by Tom Wisherop, on
Richmond Refinery or ERTC/MEE/Corrosion websites
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-40
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Coatings for Wet H2S Service (Contd)


Conventional Coatings: Thin film epoxy. Not suitable for amine service, and
take a long time to cure unless (carefully) heated to
about 90F. Properly applied these coatings can
give good service up to about 180F. Lab tests also
show quite good steamout resistance to 150#.
Specialty Coatings: Belzona brand high performance epoxy. Some
grades suitable for amine service. Belzona cures
much more quickly than conventional coatings, and
generally have somewhat higher temperature
resistance, but not all grades can resist 150#
steamout. All grades have application issues.
Thermal Spray Coatings: Used to a limited extent by ChevronTexaco and
some competitors. Not yet proven, appears to have
had some success and some failures. Potentially
overcomes temperature and steamout concerns.
Early results from lab tests show that properly
selected and applied thermal spray very effectively
stops hydrogen diffusion under wet H2S type
conditions.
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-41
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Coatings for Wet H2S Service (Contd)


Relative costs for a 7 diameter, 16 tall vessel, 96 of weld coated
with a stripe 6 wide (Richmond, 2003)

Specialty Coating
Conventional Epoxy (Belzona 1591) Thermal Spray
$15,000 Not Forced $23,000 $10,000-$15,000 (?)
Air Cure
$18,000 With Force
Air Cure

Note: These costs include surface preparation, but do not


include maintenance costs of opening the vessel
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-42
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Coatings for Wet H2S Service (Contd)


Typical recoat and cure times for epoxy phenolic coatings
always consult manufacturer data sheets for specifics

Surface Between Final Cure


Temperatures Coats, Min. (Before Immersion)
50F (10C) 72 Hours NR
60F (16C) 36 Hours 14 Days
70F (21C) 24 Hours 10 Days
80F (27C) 18 Hours 6 Days
90F (32C) 12 Hours 96 Hours
100F (38C) 6 Hours 48 Hours
110F (44C) 3 Hours 36 Hours
120F (49C) 2 Hours 24 Hours

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-43
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Coatings for Wet H2S Service (Contd)


Specialty Coatings
Belzona 1391: Hand-applied, two-component system. Not resistant to
amines. Not easy to apply well and within proper thickness
range, but is more forgiving than Belzona 1591.
Manufacturer claims good up to 265F and withstands
steamouts up to 400F; Mat Lab tests show good
performance in steamout tests up to 150# (365F).
Belzona 1591: Hand-applied higher temperature version of Belzona 1391.
Resists amines. More difficult to apply. Manufacturer
claims good up to 365F; Mat Lab steamout tests show
good performance up to 150# (365F).
Belzona 1522: Spray-applied two component system. Resists amines.
Easier to apply well than hand applied versions, but
requires special heated application equipment.
Manufacturer claims good up to 300F; Mat Lab steamout
tests show good performance at 300F but major cracking
at 150# (365F). Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-44
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Coatings for Wet H2S Service (Contd)


Thermal Spray Coatings
Electric Arc: Sometimes called Twin Arc. High deposition rates,
cheaper. Arc struck between two wires or rods,
atomizing gas sprays the melted particles onto the base
metal.
High Velocity Powder is fed into the jet stream of combusted fuel,
Oxyfuel which sends the semi-molten particles at hypersonic
(HVOF): speeds onto the base metal. Produces very dense
coatings. Sealer not required.
High Velocity A hybrid of the two processes described above used
Continuous by Metalspray United. HVCC is a development of their
Combustion earlier Com-Arc process. HVCC creates molten metal
(HVCC): by an arc struck between two wires, then uses
combusted fuel to propel the particles at hyperionic
speed onto the base metal. Not clear if a sealer is
required.
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-45
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

ChevronTexaco and Industry Experience


With Thermal Spray Applicators
ChevronTexaco Sealant
Applicator [C] or Industry [I] Application Process Top Coat Bond Coat Applied? Purpose Comments
Metalspray C TCO Amine Flash Drum HVCC (See Hastelloy 316 Stainless Yes Wet H2S Coated in 2000. In good
United (1F303.1) Note B) C-276 Steel-Like Corrosion/HIC condition after 3 years. Some
Material (See Damage limited blister growth that has
Note C) since stopped.
Metalspray C TCO Regenerator HVCC Hastelloy 316 Stainless Yes Amine Corrosion [11] 1D302.2 and 1D302.1
United (D304) and Contactors C-276 Steel-Like in High Velocity were coated in 2000.
(D301, 1D302.2 and Material Regions 2D302.2 was coated in 2001.
D302.3)
No internal inspection of any
of these vessels. External
inspection with UT gauging of
1D302.2 is favorable except
for the zone immediately
opposite an inlet nozzle where
it is suspected that coating
has been lost.
Metalspray C TCO Regenerator and HVOF Stellite 6-Like None No Localized Erosion Vessel coated in 2002 for the
United Contactor 2D302.2 Material Corrosion first time. No inspection since.
Opposite a Nozzle
With No Flow
Diffuser

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-46
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

ChevronTexaco and Industry Experience


With Thermal Spray Applicators (Contd)
ChevronTexaco Sealant
Applicator [C] or Industry [I] Application Process Top Coat Bond Coat Applied? Purpose Comments
Metalspray C Burnaby Amine First Application Inconel 625 316 Stainless Yes Amine Corrosion Coating was first applied in
United Regenerator [4] by Com-Arc Steel-Like in High Velocity 1997. Original corrosion rate
Operating at About (See Note D) Material Regions was near 40 mpy. After
255F, 15 psig Subsequent thermal spraying, no corrosion
Applications by was detected from external
HVCC Since UT gauging.
1998
Not all areas known to have
high corrosion rates could be
coated during the turnaround.

During the internal inspection


in 1998, coating was found to
be in good shape in general,
but the uncoated area
adjacent to the coating had
corroded in a step-wise
fashion. Some minor spalling
of the coating was noted along
the edges between coated
and uncoated areas.

Inspection in 2002 revealed


similar damage. It is planned
to have the remainder of the
high corrosion areas coated in
the 2004 turnaround.
Metalspray I Amine contactors at Electric Arc Hastelloy Not Known, Not Not Known, Not Wet H2S Coated in 1995. One out of
United Shell Canada's Caroline C-276 Mentioned in Mentioned in Corrosion two contactors had localized
Gas Complex Documentation Documentation 15 ft2 of damage.
Metalspray I Amine Regenerator at Electric Arc Hastelloy Not Known, Not Not Known, Not Wet H2S Coated in 1995. Has been in
United Amine Contactors at C-276 Mentioned in Mentioned in Corrosion service without incident.
Shell Canada's Caroline Documentation Documentation
Gas Complex
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-47
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

ChevronTexaco and Industry Experience


With Thermal Spray Applicators (Contd)
ChevronTexaco Sealant
Applicator [C] or Industry [I] Application Process Top Coat Bond Coat Applied? Purpose Comments
Thermal C Richmond Sour Water Electric Arc Aluminum No No Carbonate This vessel has first thermal
Spray Tank Cracking spray coated, then painted
Industries over to minimize the
(TSI) consumption rate of the
aluminum. A sealer was not
applied prior to painting, so
upon startup, the paint
bubbled off and had to be
refurbished. Has not been
internally inspected since
then.
TSI C Carter Creek Gas Plant, Flame Spray Thermotec No No Wet H2S Damage All three vessels were
C-1501 Desuperheater 18997 [See inspected in 2002. C-1501
Condenser Note A] and V-4402 were sprayed in
V-4401 L.P. Flare K.O. 1996. V-4401 was a new
V-4402 H.P. Flare K.O. vessel sprayed ~1999.

During a 1999 inspection, the


coating in C-1501 was found
to be blistered. The substrate
was corroded, coating did not
come off.

No damage to the coating in


V-4401 was noted.

The V-4402 coating inspected


in 2002 was in good condition
with the exception of partial
spalling on the manway cover
along the edge between the
coating and the gasket.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-48
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

ChevronTexaco and Industry Experience


With Thermal Spray Applicators (Contd)
ChevronTexaco Sealant
Applicator [C] or Industry [I] Application Process Top Coat Bond Coat Applied? Purpose Comments
TSI I Sulfinol Reboiler Flame Spray Thermotec Wet H2S corrosion Date that spraying was done
is unknown but data shows at
least 3 years without failure.
TSI I Sulfinol Contactor Flame Spray Thermotec Wet H2S Date that spraying was done
Corrosion is unknown, but it is known
that complete disbondment
occurred after 3 years.
F.W. Gartner C Salt Lake LPG Sphere HVOF Hastelloy No No HF Corrosion This vessel was first coated in
(T-5065) C-276 1997. General blistering
occurred. No blisters actually
opened up. Some large
blisters had cracks in them.
Substrate under blisters did
show signs of rusting,
although other areas still had
the original grit-blast pattern.

Manway had worst damage


isolated to the edge where the
Monel liner lip ends.
Evaluation of the coating chips
did not reveal general
permeation. Vessel had been
baked out at 1000 F prior to
coating.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-49
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

ChevronTexaco and Industry Experience


With Thermal Spray Applicators (Contd)
ChevronTexaco Sealant
Applicator [C] or Industry [I] Application Process Top Coat Bond Coat Applied? Purpose Comments
Initially by C Pascagoula Exchanger HVOF Hastelloy Not Known, But Not Known, But Ammonium A coating was originally
Randall Shells C-276 Not Likely Not Likely Bisulfide applied by Randall Long and
Long Then within a year it failed. Coating
F.W. Gartner was of poor quality and the
wrong chemistry. FW Gartner
was asked to repair the
damage.

The shell is no longer


available for inspection. Don't
know if any internal inspection
was performed. Don't have
documentation that follows-up
on FW Gartner performance.
F.W. Gartner C Stratco Contactor Originally HVOF Tantalum Not Known, But Not Known, But Shaft Corrosion The shaft was coated while it
Shafts (91% Sulfuric Not Likely Not Likely operated at the El Paso
Acid at 45F With iC4 Refinery, then the shaft was
and C4) (Originally transferred to the Hawaii
owned by El Paso Refinery. Coating was in
Refinery, later used by overall good condition except
Hawaii Refinery.) for the edge where part of the
coating was breached.

Coating lasted at least a year


with the Hawaii Refinery, but I
wasn't able to find out how
long it was used by the El
Paso Refinery.

Note A: Composition is approximately: 22.0% Fe, 20.0% Mo, 10.0% Ti, 4.5 % W, 0.5% V, and the balance is Ni.
Note B: HVCC = High Velocity Continuous Combustion. Uses a wire feedstock propelled at sonic velocities by a special nozzle design.
Note C: A confidentiality agreement is required before the exact composition would be disclosed to ChevronTexaco, but the 316 like has chromium content comparable to Type 316 stainless steel and
is given the Metalspray United internal designation of UTEX 5-322. The Stellite-like material's composition is unknown, but is expected to have comparable amounts of chromium.
Note D: Com-Arc is a hybrid of electric arc and HVOF processes. It involves shorting two wires to create molten droplets while hydrocarbon fuel combustion propels the particles to the surface.
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-50
6/24/03
4/16/03 Heidersbach
Heidersbach 030204-Sect
030069-Sect 16
07b 50
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Coatings for Wet H2S Service (Contd)


Key Steps for a Quality Conventional Coating Job
Follow EG-5014; see Coatings Guide on website
During shutdown planning, allow adequate time for
coating inspection, rework, and curing
Surface prep is key; should be done by coating
applicator. Prior to blasting, check samples to determine
if chlorides require high pressure water blasting.
When using forced air curing, be very careful on
temperature control!
Use experienced NACE qualified third-party or in-house
inspectors to closely follow the job. (Contact Paul Hunter
for recommendations.) They will check among other
things: surface prep, wet and dry film thickness,
application technique. Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-51
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Coatings for Wet H2S Service (Contd)


Key Steps for a Quality Specialty Coating Job
Follow EG-5014; see Coatings Guide on website
Remove excessive weld reinforcement that would
cause coating buildup beyond the acceptable range.
Surface prep is key; should be done by coating
applicator.
Very important that the coating be applied within the
thickness range for that particular product
Use experienced NACE qualified third-party or in-
house inspectors to closely follow the job. (Contact
Paul Hunter for recommendations.)

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-52
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Coatings for Wet H2S Service (Contd)


Key Steps for a Quality Thermal Spray Job
Follow EG-5014
Application selection is critical; refer to EG-5014 and
contact ERTC (Lawrence Cheung) for latest
recommendations
Surface prep is key; should be done by coating
applicator
Use a sealer for the electric arc process

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-53
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Use of Crak Stop in Wet H2S Service


Crak Stop is a proprietary corrosion inhibitor from
InterCorr that minimizes the burst of hydrogen activity
when fresh metal is exposed to wet H2S conditions.
Crak Stop is most often used after a vessel has been
cleared for WFMT inspection (exposing fresh metal), but
we do not subsequently coat the vessels. Crak Stop has
shown to be effective under laboratory conditions; we do
not have carefully controlled field experiments.
Pascagoula Refinery has used Crak Stop.
Crak Stop is available as oil or water soluble. The oil
soluble version seems to stick better; mineral oil is
used as the carrier.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-54
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Use of Crak Stop in Wet H2S Service


(Contd)
Theory of Crak Stop
Without Inhibitor With Inhibitor

Hydrogen Flux
Hydrogen Flux

H to Start Cracks H to Start Cracks

H to Keep H to Keep
Cracks Growing Cracks Growing

Time in Wet H2S Time in Wet H2S

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-55
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Inspection Tools for Wet H2S


Wet Fluorescent Magnetic Particle Testing (WFMT) is
the most sensitive at detecting surface-breaking
cracks. But it requires a near-white blast, leaves the
metal exposed to fresh attack, and it tends to detect
many insignificant indications. A good choice if the
vessel will be subsequently coated.
Ultrasonic Testing (UT), done from outside the
vessel, can be reasonably sensitive, especially if
automated, but is highly dependent upon operator
skill. Not able to inspect reinforced nozzles.

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-56
6/24/03 Heidersbach 030204-Sect 16
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Inspection Tools for Wet H2S (Contd)


Alternating Current Field Measurement (ACFM) and Eddy
Current Testing (ET) can be used in lieu of WFMT if the
procedures and technicians are properly qualified. UT
may be required to accurately determine depth of cracks
detected by ACFM or ET. ACFM or ET requires only that
loose scale be removed, the protective sulfide scale is
left alone. ACFM and ET also tend to naturally ignore
small, insignificant indications.
Acoustic Emission Testing (AET) is a still-emerging
technique that in theory can detect active cracking, such
as when the vessel is tested during cooldown or
overpressure. Interpretation is extremely complex, and
currently ChevronTexaco does not have confidence in
AET for wet H2S inspection.
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 16-57
6/24/03 Heidersbach 030204-Sect 16
Welding

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 17-1
6/24/03 Heidersbach 030204-Sect 17
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Welding Issues
Follow ChevronTexaco standards for preheat and
PWHT they are sometimes more restrictive than
code
Extensive PWHT including special procedures are
required to prevent carbonate SCC
Watch ferrite contents in stainless steels
3-10% for all other 300 Series
Stainless steel may be sigma embrittled from service
1100F
Watch dissimilar welds if done improperly, could
easily lead to failures due to environmental cracking,
thermal expansion differences, etc. Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 17-2
6/24/03 Heidersbach 030204-Sect 17
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Welding Issues Repair of 1-1/4 Cr


FCC Reactors
Full encirclement stress relief is standard, but is
costly and time-consuming, and weight of top head
and attachments makes this tricky
Spot stress relief is less costly but must be carefully
engineered and controlled consult with ERTC
specialists on required thermal gradients to avoid
bulging
Temper bead procedures are available which can be
applied without PWHT, but careful control and
procedure qualification required
Consult ERTC on all 1-1/4 Cr weld repairs in FCC
plants, especially if cracking has occurred
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 17-3
6/24/03 Heidersbach 030204-Sect 17
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Minimum Preheat Temperatures for


Cr-Mo Steels

Material Minimum Preheat Temperature, F


1 Cr - 1/2 Mo 300

1-1/4 Cr - 1/2 Mo 300 (350 for Piping)

2-1/4 Cr - 1 Mo 300 (350 for Piping)

3 Cr - 1 Mo 350

5 Cr - 1/2 Mo to 9 Cr - 1 Mo 450

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 17-4
6/24/03 Heidersbach 030204-Sect 17
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Postweld Heat Treatment of


Cr-Mo Steels

Material Temperature, F Minimum Time, Hr


1 Cr - 1/2 Mo 1300-1375 2
1-1/4 Cr - 1/2 Mo 1300-1375 2
2-1/4 Cr - 1 Mo 1300-1375 2
3 Cr - 1 Mo 1300-1375 2
(1325-1400 for Piping) 2
5 Cr - 1/2 Mo to 9 Cr - 1 Mo 1325-1400

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 17-5
6/24/03 Heidersbach 030204-Sect 17
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Issues With Dissimilar Metal Welds


Dilution of weld by less alloyed basemetal
Not a problem for CS & low alloy (Cr-Mo); use CS filler
For CS or low alloy to SS or high alloy, use over-
alloyed filler metal (309, Inco)
PWHT may not remove residual stress
Dissimilar metal weld cracking if T > 900F for long
times; use Inco to weld SS to CS or low Cr

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 17-6
6/24/03 Heidersbach 030204-Sect 17
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Refractory Anchor Welding


New Construction: Weld to Cr-Mo before PWHT (and
preheat)
Repairs: Preheat Cr-Mo for welding, but no need to
PWHT refractory anchor welds
Weldmetal Selection: Prefer Inco 182 for dissimilar
metal welds (309 SS less desirable)

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 17-7
6/24/03 Heidersbach 030204-Sect 17
Brittle Fracture

Senior Analyst and Inspector Training


August 2003

Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

ChevronTexaco 2003 Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 18-1
6/24/03 Heidersbach 030204-Sect 18
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Brittle Fracture
As they get colder, carbon and low alloy steels undergo a
ductile-to-brittle transformation and become susceptible to
brittle fracture
Susceptibility to brittle fracture is dependent on:
Material Properties Wrought, heat treated (e.g., normalized,
quenched, and tempered) and fine grained steels are less
susceptible than cast, as rolled or coarse grained steels. Newer
steels are less susceptible than older dirty steels.
Applied Stress Brittle fracture is not a concern when applied
stress is less than 8 ksi. The greatest concern for brittle fracture is
with pressure vessels and tanks, not piping components. Vessels
operating near full design pressure are more susceptible than
those operating well below design.
Presence of Flaws Flaws concentrate local stresses and can
greatly increase susceptibility to fracture
Temperature Lower temperatures increase risk
Confidential
Copyright 2003 by ChevronTexaco
Energy Research and Technology Company

Materials and Equipment Technology Unit


To be reproduced and used only in accordance with written
permission of ERTC 18-2
6/24/03 Heidersbach 030204-Sect 18
Senior Analyst and Inspector Training
Fluid Catalytic Cracking Process August 2003

Charpy V-Notch Energy Absor