Material Selection Handbook
Material Selection Handbook
OVERVIEW
-
Thermal cracking The most prevalent thermal processes are visbreaking and coking, both of which
are applied to residuum.(2)Visbreaking is a mild thermal cracking; coking is a severe thermal cracking.
Fluid Catalytic Cracking (FCC) - Distillates heavier than diesel are fed to this unit to crack them
catalytically, primarily into gasoline. Also produced are light components consisting of four-carbon
molecules, three-carbon molecules, two-carbon molecules, and light gas. The three- and four-carbon
molecule fractions contain 01efins'~)that can be converted to gasoline by alkylation with isobutane. The
two-carbon molecules and lighter gas contain hydrogen sulfide, ammonia, and some hydrogen cya-
nide. Fractions heavier than gasoline, called cycle oils, also are produced.
('1 Naphtha often is confused with naphthene. Naphthenes are cyclic hydrocarbon compounds, whereas, naphtha is
defined as a hydrocarbon mixture with a boiling point from 50°C to 200°C (120°F to 400°F).
Residuum is heavy petroleum from the bottom of a fractionator that has the lighter petroleum products, such as
gasoline, removed by distillation.
13) See section 2.1. no. 2 for definition of olefin.
Hydrocracking - Distillates heavier than diesel are catalytically cracked at high pressure and el-
evated temperature in the presence of hydrogen to produce either gasoline or diesel and lighter prod-
ucts. Generally, the objective is to produce diesel and lighter products, such as two-carbon molecules
and lighter gas, three-carbon molecules, and four-carbon molecules. The two-carbon molecules and
lighter gas contain hydrogen sulfide and some ammonia. Most of the ammonia and an equivalent
amount of hydrogen sulfide are removed with wash water.
Alkylation - Fractions containing light olefins (like FCC three- and four-carbon molecules) are alky-
lated with isobutane using sulfuric acid or hydrofluoric acid as a catalyst to produce gasoline (alkylate).
Polymerization - Light olefins are polymerized to gasoline in a process using a solid catalyst containing
phosphoric acid.
Supporting Processes - The primary support process units are:
1. Amine unit - Light gases are brought into contact with amine to absorb hydrogen sulfide and/or
carbon dioxide. The amine is regenerated by removing hydrogen sulfide and/or carbon dioxide, which
is then fed to a sulfur plant.
2. Sulfur plant - Acid gas (hydrogen sulfide, sulfur dioxide, and carbon dioxide) is fed to a sulfur
plant, where the hydrogen sulfide is converted to elemental sulfur by partial oxidation with air.
3. Sour water stripper - Sour water is fed to a stripper to remove ammonia and hydrogen sulfide. The
wet ammoniahydrogen sulfide overhead is fed to the sulfur plant. In cases where there is an appre-
ciable amount of ammonia, a two-stage stripper is used to produce separate hydrogen sulfide and
ammonia products.
4. Hydrogen plant - Natural gas (or refinery gas) and steam catalytically react at high temperature to
form carbon dioxide and hydrogen. The carbon dioxide is removed by absorption, and the hydrogen is
used in hydroprocessing units.
Figure 1.1 is a simplified block flow diagram for a refinery processing a high sulfur crude oil to obtain
a large number of products.' The crude oil is fractionated in an atmospheric distillation unit at about 345
Waa (50 psia) at temperatures up to about 370°C (700°F). Naphtha and lighter components are produced at
the top of the column (overhead products). The kerosene, diesel, and light gas-oil produced in the middle of
the column are removed from the side of the column (sidecut products). The overhead is condensed; water is
separated from the naphtha; the naphtha is fed to a stabilizer and gasoline splitter to produce butane and
lighter components, light gasoline, and heavy naphtha. The bottom product from the atmospheric column is
fed to a vacuum distillation unit to recover additional light gas-oil and heavy gas-oil as distillate products.
The vacuum unit runs at about 10 kPaa (1.45 psia) and at temperatures up to about 415°C (775°F).
The boiling ranges of the products generally are:
E
Amine
Plant *Fuel Gas
LPG &
Light Gases
3 ! I
I
Gas Piant & B
L
E
Gasoline
-
HCI
Desuiturization
- Jet Fuel
I
Middle Distillates
pheric Kerosene
-I
: A Naptha
97
I: E
I
Desulfurization 7HeatingOil
Hydrogen
Pduction
I E
0
N
Vacuun
Coking
coke
Industrial
Fuel
b
Asphalt
Sour Water
FCC Crude HDS Coking
J J J d h 2 2 -
Sour Water
Corrosives underlined Stripper
Treated Water
Environments in refineries can be broken down into various categories, such as hydrocarbon and sulfur,
hydrocarbon plus hydrogen sulfide plus hydrogen, etc. The materials required for vessels, exchangers, and
so forth are listed as a function of temperature on the materials recommendation sheets in Appendix A. In
general, when different environments are on opposite sides of a piece of equipment (e.g., an exchanger
tube), the most severe service governs. In some cases, a material different from the materials on either side
may be the proper choice. Another more general guide to materials is found in Appendix B. The various
materials-related phenomena are listed as a function of temperature then as a function of environment.
Further information on petroleum refinery materials selection can be found in the bibliography listed at the
end of each chapter.
The corrosion allowances given in Appendix A are the product of maximum anticipated corrosion rate
and the design lives listed in Appendix A. Corrosion allowances are given in millimeters. The typical corro-
sion allowances used in the metric system (e.g., 1.5 nun, 3 mm, etc.) are not numerically equal to the typical
corrosion allowances used in English units (e.g., 1/16 inch, 1/8 inch, etc.). However, they are roughly equiva-
lent. Often temperatures (e.g., 150°C [300"F]) also are only approximately equal, because it usually is more
convenient to use round numbers in each system.
It must be recognized that corrosion is not uniform, and the equipment may be exposed to unanticipated
corrosives. Therefore, in order to have the least expensive equipment consistent with good engineering
practice, it is recognized that some areas may require repair during the design life. Although it is often more
economical to have a minimum first-cost for equipment - consistent with avoiding premature plant shut-
downs - then to use maintenance dollars to obtain the total design life, safety and reliability also must be
taken into account when selecting materials of construction.
A corrosion allowance of 1.5 mm (1116-inch) is the generally accepted minimum for carbon and low
alloy steel equipment (4)(5)(6) consistent with the minimum first-cost approach. Because material normally is
purchased in standard wall thickness, e.g., specifying a 1.5 mm (1116-inch) minimum corrosion allowance
for carbon and low alloy steel usually results in an actual corrosion allowance closer to 3 mm (Win.),
because the actual corrosion allowance is the wall thickness minus that calculated for pressure.
Both clad and weld overlay are used interchangeably to indicate materials of construction. The choice
between roll bond clad, weld overlay, and explosive bond clad basically is a matter of economics. Roll bond
cladding usually is economical when the calculated thickness of the high alloy material is in the order of 13
mm (0.5 in.) or greater. As the cost of high alloy material increases, the thickness where roll bond cladding
becomes economical decreases. When the backing plate thickness exceeds 5 1 mm to 64 mm (2 in. to 2.5 in.),
weld overlay and explosive bonding become more attractive. Explosive bonding often is economical for
thick backing plates, such as those used for tubesheets.
There are, however, some additional considerations. The backing material for roll bond clad plate is
heat treated after the cladding operation. This can be detrimental to high alloy cladding that form undesir-
able phases at the temperatures required to heat treat the backing material.' Thus, weld overlay or explosive
bond clad might be more desirable for these materials. Note, however, there have been some lack of bond
problems reported with explosive bonded materials when the backing plate was too thin, e.g., 9.5 mm (3/8
in.).
Materials selection for centrifugal pumps is contained in API 610, Appendices G and H.3 Minimum
requirements for fired heaters are contained in API 560. Both NACE International Group Committee T-8 on
Refinery Industry Corrosion and the API Corrosion Committee publish minutes of their semiannual meet-
ings. A computer software database of the 1959 to 1994 T-8 minutes is available from NACE International:
2. CRUDE UNITS
Crude oil is a mixture of hydrocarbon molecules of various weights. The arrangements of the carbon
and hydrogen atoms vary significantly. The common arrangements are:
(4) For an internally lined and externally painted equipment, a zero corrosion allowance is often used. For SS and high
alloy material, 0.5 mm (1/32 -in.) is usually a minimum if some corrosion is expected. Where no corrosion is expected on
SS or high alloy material, zero corrosion allowance is sometimes specified.
(5) Heat exchanger tubes are a special case. They are normally specified as a particular gauge, e.g., 12-gauge, 14-
gauge, etc. This usually results in an adequate amount of corrosion allowance for the typical five-year life assumed for
exchanger tubes. This is because exchanger tubes normally are only 19 mm to 25 mm (0.75 in. to 1 in.) in diameter;
thus, the wall thickness required for pressure usually is very small. However, when the gauge is specified, the wall
thickness required for pressure must be determined by the designer to make sure the thickness is adequate for pres-
sure plus some nominal corrosion allowance.This is particularly important if there is a high differential pressure across
the tubes and the unit must operate with loss of pressure on either side. Alternatively, some users specify small corro-
sion allowances ranging from 0.25 mm to 1.5 mm (0.01 in. to 0.065 in.)
("The minimum corrosion allowance for pressure containing parts in TEMA (Tubular Exchangers ManufacturingAsso-
ciation) Class R heat exchangers is 3.0 mm (118 in.).
A typical asphaltic crude oil might be composed of the following constituents:
Constituent Volume %
The constituents in oil that cause corrosion are sulfur compounds, saltwater, inorganic and organic
chlorides, inorganic and organic acids, and organic nitrogen compounds (which form cyanides). The sulfur
compounds found in crude oil are shown in Figure 1.15 As can be seen from the tabulation, a wide variety of
sulfur compounds may be present in crude oil. A crude oil containing more than 0.0014 m3 (0.05 ft3) of
dissolved hydrogen sulfide per 378 liters (100 gal.) of oil is called sour. However, a crude oil containing less
than 0.014 m3(0.5 ft3) of hydrogen sulfide per 378 L (100 gal.) of oil is not corrosive to steel in petroleum
processing equipment.
Faund In
Hydrogen Sulfide H,S Yes Yes Yes Figure 1.2 -Types of sulfur
Mercaptans
Aliphatic RSH Yes Yes Yes compounds in crude oil and
Aromatic RSH Yes Yes Yes distillate^.^
Naphthenic RSH No No No (Reprinted with permission from
Sulfides
Aliphatic R-S-R Yes Yes Yes
“High Temperature Degradation of
Aromatic R-S-R No No No Structural Materials in Envir-
Yes Yes Possibly onments Encountered in the
Petroleum and Petrochemical
Aliphatic R-S-S-H Yes Yes No Industries,”Anti-Corrosion Methods
Aromatic R-S-S-H ? No Yes end Meterials, November 1975,
Thiopheneand /s\ CH
CH Yes Yes Yes
Sawell Publications, Ltd., London,
hornologues II II
HC HC
~
England)
Polysulfides R-S,-S-R ? Yes Yes
The types of organic sulfur compounds shown in the first column in Figure 1.2 thermally decompose
during processing to constituents, such as hydrogen sulfide and mercaptans. If sufficient quantities of or-
ganic sulfur compounds (above about 0.2 wt% to 0.3 wt% sulfur) are present, the crude oil is corrosive to
carbon and low alloy steels at temperatures above about 232°C to 288°C (450°F to 550°F)and up to 455°C
(850°F).Above 455°C (850”F),corrosion rates drop if coking occurs on the walls; otherwise, the corrosion
rates continue to increase. It is important to note that the wt% sulfur limit is for crude oil, not cracked
products. As can be seen in Figure 1.2, a number of sulfur compounds remain in cracked products in addi-
tion to H,S. Foroulis has shown that the aliphatic and aromatic mercaptans and the aliphatic sulfides and
disulfides that remain in cracked products are very corrosive.IO(Conversely, the thiophenes do not decom-
pose, hence, are not corrosive.) These compounds are corrosive in ppm quantities at elevated temperatures
as discussed in Chapter 3, Section 2.7.
Chromium steels, e.g., 1-1/4 Cr-112 Mo, are required in naphthalene above 370°C (700°F) even when
the sulfur is below 0.2 wt%. This is because the decomposition of naphthalene to undesirable products is
7
()’ Naphthenic acids also have been reported in some Far East, African, Russian, Gulf Coast, Indian North Sea, and
Canadian crudes.
Primarily vacuum (heavy) gas oil.
(9) It is well established that naphthenic acids are destroyed in hydrotreater reactors.
(lo) In essence, all that is being done is to say if almost all of the metal loss from corrosion can be accounted for in the
(iron sulfide) scale then sulfidation is controlling. Conversely, if a great deal of the metal loss cannot be accounted for
in the scale, then naphthenic acid is controlling.
hydrocarbon streams that contain water. Therefore, care should be exercised to minimize air getting into
equipment from leakage in seals, air in-leakage in vacuum columns, etc. Corrosion problems in the aqueous
phases are discussed in the following section on overhead systems, in the Chapter 2 section on characteris-
tics of sour water, and in the Chapter 3 section on coolers.
2.3.1 Overview
Figure 1.3 shows a simplified flow diagram and process description of a crude oil distillation unit.] The
crude oil is first desalted to minimize formation of hydrochloric acid in waters condensed from the top of the
fractionator columns and to minimize salt deposits in the preheat section. As the crude oil is heated above
232°C to 288°C (450°F to 550"F), depending on the wt% sulfur,'") corrosion begins to occur due to break-
down of sulfur compounds. Condensing water containing hydrogen sulfide and hydrogen chloride causes
corrosion in the overhead system of the atmospheric column. Corrosion caused by breakdown of sulfur
compounds continues in the atmospheric column; it is worst at the inlet from the fired heater as a result of
turbulence from high vapor velocities and flashing. The cycle of corrosion above 232°C to 288°C (450°F to
550°F) and in the condensing water in the overhead system is repeated in the other two columns as the
bottoms from the preceding column are further distilled.
Figure 1.3 shows a single-drum overhead system. Double-drum systems also are used. The difference
between the two systems is the reflux temperature at the top of the tower. In the single drum system, total
-
Figure 1.3 Crude distillation,three stages.'
(Reprinted with permission from The Role of Stainless Steels In Petroleum Refining,
April 1977, American Iron and Steel Institute, Washington, DC)
In general, the higher the sulfur content, the lower the temperature where corrosion begins to be a problem.
(I1)
However, as shown in Figure 1.13, it is the sulfur release characteristics that determine at what temperature corro-
sion begins to be a problem.
liquid condensation occurs in the overhead condensers. The reflux will be cool and keep the tower top cool.
It is advisable to check the hydrochloric acid dew point vs partial pressure to determine the anticipated
location of corrosion. If direct addition of the cold reflux will cause condensation on the trays, measures
such as spray distribution of the reflux or raising the tower top temperatures should be taken to prevent this.
Addition of neutralizing amines to the reflux to neutralize the acid should not be done because this results in
column tray corrosion from the formation of amine hydrochloride salts. Some refiners recommend limiting
the vapor velocity in the overhead lines to 15 m / s (50 ft/s); however, 37 m / s to 43 d s (120 ft/s to 140 ft/s) is
more common. The vapor velocity is limited to 7.5 m / s (25 ft/s) in the presence of water (two-phase flow).
The initial corrosion control system used in a crude unit is a desalter. Modem desalters accomplish
separation of oil and water electrostatically. The intemals used to accomplish electrostatic separation nor-
mally are of a propriety design. The vessel itself usually is carbon steel. Some refiners specify that the
bottom be cement-lined; however, most refiners specify only a heavy corrosion allowance to protect against
corrosion from the saltwater that collects on the bottom of the vessel. The payout on a desalter is difficult to
establish. Desalters normally are used when the salt content of the crude oil exceeds 76 ppm (20 lbs per
When high reliability of the unit is desired, crude oils with salt contents of 30 ppm to
1,OOO barrels [~tb]).(’~)
38 ppm (8 ptb to 10 ptb) are desalted. Some refiners now feel that a desalter should be used regardless of the
salt content. The goal in desalting is 3.8 ppm (1 ptb) or less in the effluent crude oil. This results in 50 ppm
or less of HCl in the overhead water. Fluctuations in salt content are particularly troublesome to the down-
stream equipment; therefore, the desalter should be designed for the maximum anticipated salt content.
Desalting is used to remove bottoms sludge and water (BSSrW) as well as any brines in the crude oil that
result from lack of settling in the oil field, saltwater from tankers, and emulsified salt brine in the crude oil.
Additives are used to help break the emulsion after the crude oil is heated to about 120°C to 135°C (250°F to
275°F) depending on the viscosity of the crude oil. Wash water, preheated by the effluent water, helps to
dissolve the salts from the crude oil. Although mixing the water and crude oil is important, some systems are
not designed to accommodate this procedure. For example, some systems are not designed for the 207 kPa
(30 psi) pressure drop required across the mixing valve.
Removal of salts from crude oil is important to avoid corrosion and plugging of the overhead system.
Plugging is caused by the formation of salts, e.g., ammonium chloride, from the reaction of neutralizers and
HCl. Saltwater in crude oil usually is similar to ocean water. Sodium chloride is thermodynamically stable;
however, magnesium and calcium chlorides thermally decompose at relatively low temperature^."^) The
decomposition reaction converts about 90% of the MgC1, and about 15% of the CaCl, to metal oxides and
hydrochloric acid in the presence of water vapor (hydrolysis). The hydrolysis takes place in the vapor phase
as the crude oil is heated to be fed to the atmospheric and vacuum towers. Since the hydrolysis takes place in
the vapor phase at temperatures below the temperature where HCl causes a significant acceleration of corro-
sion in the crude oil environment, corrosion from the HCl does not occur until it dissolves in the condensing
water in the top of the towers. Desalting to 3.8 ppm (1 ptb) can keep tower overhead condensates under 50
ppm of hydrochloric acid. A 50 ppm concentration of hydrochloric acid is quite corrosive to carbon steel.
The overhead condensate should not have more hydrochloric acid than 10 ppm. Adding sufficient water to
quench through the dew point before the first exchanger will minimize the hydrochloric acid concentration.
One way to reduce the amount of HC1 is caustic injection after the desalter. The injection amount is
guided by the amount of salt in the overhead condensate. The amount of caustic added usually is limited to
10 ppm (2.6 ptb). Caustic normally is not added when the products of the crude unit will be fed to hydrotreaters
or fluid catalytic crackers because it may deactivate the catalyst. Fluid cokers will foul from premature
coking if more than 19 pprn (5 ptb) of caustic is used. Spent caustic washes can be used if they do not contain
contaminants that cause increased corrosion or plugging. For example, a spent alkylation plant caustic would
add sulfur dioxide, which can react with hydrogen sulfide to cause sulfur plugging. The following steps are
(12) The exception is when the specific gravity of the crude oil is the same as that of the water. Although there are
some recent developments which may allow desalting in this situation, in the past it has not been possible to desalt
when there was not a significant difference in the specific gravity of the crude oil and that of water.
Magnesium chloride thermally decomposes at around 120°C (250°F).
required to avoid problems with a caustic injection system:
1. Use an injection quill"" designed to avoid caustic embrittlement at the injection point.
2. Inject the caustic into a slipstream of crude.
3. Put the crude oil on the tube side of preheat exchangers. This will prevent the caustic from
concentrating in pockets around baffles and floating heads.
4. Avoid overfeeding. Overfeeding caustic, poor mixing, or using a caustic that is too strong causes
furnace tube coking, fouling of crude oil preheat exchangers, and caustic embrittlement.
Caustic injection is used to convert magnesium chloride to sodium chloride since sodium chloride does
not readily hydrolyze to hydrochloric acid. Reducing hydrochloric acid is helpful in both the atmospheric
tower and the vacuum tower overheads.
However, one drawback should be pointed out: when magnesium chloride hydrolyzes, both hydrochlo-
ric acid and magnesium hydroxide are formed. Magnesium hydroxide is stable and insoluble. It will end up
in the coke from a coker or in the total dissolved solids in fuel oil. Sodium chloride in heavy oils can be
cracked in cokers, fluid catalytic crackers, hydrocrackers, and hydrodesulfurizers, and can lead to ammo-
nium chloride plugging problems in coker fractionators, reformer depropanizers, etc. However, plugging is
only a problem if the quantity of chloride is large.
With a single-drum system the top of the tower is subjected to water condensation, which contains
hydrogen chloride, ammonia, and hydrogen sulfide. Corrosion rates can be severe if the chloride content is
high. For this reason, alloy 400 (UNS N04400) has been used as a crude tower lining and as a tray material.
Alloy 400 is resistant to these waters under 120°C (250°F); however, corrosion will occur if the H,S or NH,
levels are too high. Alloy 400 valve trays are not acceptable because the valves wipe the protective scale
from the contact surface, causing severe corrosion. This occurrence has not been reported on cage-type
valve trays or sieve trays. Recent experience indicates 6 Mo super SS (UNS N08367) has worked in over-
head systems where alloy 400 has failed. Alloy C276 (UNS N10276) also has been used where alloy 400 has
failed.
Corrosion protection in a single-drum system has been achieved by neutralizing the water condensate to
a pH of 5.5 to 6.5 and using a filming amine inhibitor. More recently, some refiners have concluded that
proper neutralizing is all that is required. Neutralizers used are ammonia and neutralizing amines. There are
pros and cons associated with each neutralizer. Control of pH with ammonia is difficult unless aqueous
ammonia is used because of the strong effect hydrogen chloride or ammonia have on condensate pH if too
litde or too much ammonia is used. With neutralizing amines it is easier to control the pH over a wide range
of amine concentrations. However, neutralizing amines are soluble in both water and hydrocarbon.
If the amine chloride concentration is high, the amine chloride will return to the fractionator and cause
chloride corrosion. High amine chloride conc'entrations can occur if there is insufficient water to quench
through the dew point before the first exchanger or in a dual reflux system when the system is designed to
condense water in the second reflux drum. Neutralizers known to form hydrochloride salts include MEA and
morpholine. Fortunately, some neutralizing amines that do not form high temperature salts are now com-
mercially available.
Ammonia or neutralizing amines are injected in the fractionator overhead line so they can react with the
HCl both before and after the dew point of hydrochloric acid in solution is reached. Severe localized corro-
sion can occur at injection points if injection is not done properly. Drip injection of concentrated inhibitor
can result in a sludge layer of concentrated inhibitor on the bottom of a pipe. This inhibitor sludge layer can
cause dissolution of the protective scale on the inside of the pipe at temperatures in excess of 121°C (250°F).
A catastrophic fire, resulting in an entire unit being destroyed, was caused by erosion-corrosion in an area
where scale was removed by drip inhibitor injection. Tests revealed corrosion rates of 0.5 mmly to 1 mm/y
(14) UNS N10276 is used for the quill and the injection line by some refiners.
(20 mils/y to 40 milsly) at 177°C (350°F). To minimize corrosion problems at injection points, the following
are recommended:
1. Use quills to inject neutralizers or inhibitors, particularly for temperatures above 121“C (250°F).
2. Do not inject concentrated neutralizers or inhibitors, i.e., dilute them with water and a slip
stream of naphtha before injection.
3. Inject water through spray nozzles pointed downstream.
4. Locate injection points well ahead of any bends.
5 . Periodically inspect injection points (API Standard RP570, “Piping Inspection Code; Inspection,
Repair, Alteration and Rerating of In-Service Piping Conditions” recommends every three years).
Often, however, neutralization is not accomplished, and severe corrosion from hydrochloric acid still
occurs at the dew point. This is because the first drops of condensate have a very low pH (e.g., pH 1 to 2)
even if the pH of the total condensate is alkaline, e.g., pH 8. The pH is controlled at the overhead receiver
water draw because dew point pH measurement is not feasible. Continuous monitoring of the pH and the
chloride concentration is preferred to make sure overhead corrosion control is maintained One method of
controlling the dew point pH is to recycle water from the drum to the overhead line. This water buffers the
condensate at the hydrochloric acid dew point and also provides water in which the ammonia can dissolve.
A double-drum system operates with a high tower top temperature that is above the dew point of the
water-hydrogen chloride solution. A heat exchange with crude oil or another stream condenses only hydro-
carbon in the first drum. This hydrocarbon, usually called heavy naphtha, is a hot reflux that controls the
tower top temperature. As mentioned previously, neither ammonia nor neutralizing amines should be used
in this system because these amine chloride salts can be carried back to the column in the reflux and cause
severe corrosion.
There should be no water draw from the first drum. The vapors from the first drum travel through
condensers to cool them to the temperature (approximately 38°C [100”F]) required for gas and gasoline
separation. Corrosion can occur in these condensers; therefore, the same corrosion controls used in a single-
drum system should be implemented in a double-drum system.
Materials used in the overhead condensers vary with the source of the cooling water, the amount of
chloride, and the success of inhibitors, pH control, wash water, etc. For refineries with brackish or saltwater
cooling, the use of titanium (UNS R50400)tubes is economical since carbon steel cannot be used. Titanium
is also cost competitive with SS, alloy 400 (UNS N04400), brasses, and copper-nickel alloys in fresh
(treated) water.
If ammonium chloride forms and plugs the hot areas, pitting of titanium, as well as SS, alloy 400,
brasses, and copper-nickel alloys can occur. This has been reported for both air coolers and water coolers.
Brasses such as admiralty brass (UNS C44300) have been used successfully where water side velocities are
controlled to under 2.4 m/s (8 ft/s) and ammonia is not high enough to corrode or crack the brass (pH below
7.2).
Note that admiralty (UNS C44300)can crack in the presence of air, e.g., during downtime, if ammo-
nium chloride is present. Copper-nickel alloys are much more resistant to ammonia but will corrode if the
hydrogen sulfide content is high. Duplex SS, such as UNS S31500 and UNS S32205, as well as ferritic
alloy UNS S44400,have experienced under-deposit corrosion in overhead systems. Conversely, austenitic
alloy UNS NO8904 has worked well. Carbon steel is used only where very careful control is exercised on
cooling water chemistry.
Figures 1.4 to 1.10 apply to carbon steel overhead systems.6-” Figure 1.4 can be used to calculate the
amount of ammonia that must be added to an overhead system to obtain a desired pH when the total sulfur
(H,S + HS-) in solution or the partial pressure of hydrogen sulfide is known. For comparison, the pH of a
pure hydrogen sulfide-water system is shown as a function of hydrogen sulfide partial pressure in Figure
1.5.’ The chloride concentration at 0°C to 120°C (32°F to 250°F) can be estimated from Figure 1.66 if the
hydrogen chloride and water vapor partial pres-
sures are known. Once the pH of the overhead
has been estimated, or measured under pressure,
the corrosion rate can be estimated from Figure
1.7.8
Figure 1.8 shows how the corrosion rate on
carbon steel increases as the temperature de-
crease~.~ The increase in solubility of corrosive
gases with decreasing temperature more than off-
sets the slowing of corrosion processes as a result
of the temperature decrease.
Figure 1.9 lo shows that the corrosion rate ap-
pears to drop with decreasing pH below 5.25 be-
fore the corrosion rate again increases with a de-
crease in pH, as indicated in Figure 1.7.8
Figure 1.10 l 1 is a good example of how metal Figure 1.4 -Total S in solution vs pH.6
loss is reduced by desalting to remove chlorides. (Reprinted with permission from “Crude Unit
However, using the two corrosion rates shown can Overhead Corrosion,” Hydrocarbon Processing, May
1972, Gulf Publishing Co., Houston,TX)
be misleading. Fouling is very important to heat
transfer, unit throughput, and tube bundle life. For
example, after desalting, a 0.08 mm/y (3.2 mpy)
Figure 1.5 -
Solubility of H,S in
water. Partial
pressure of H,S in
vapor vs pH of sol-
ution?
(Reprinted with per-
mission from Handbook
of Chedstty, 6th Edition,
1979, McGMw-HIII P u b
lishing Co., New York, NY)
-
Figure 1.7 (above) Corrosion of steel in
aqueous sulfide solutions. Note the acid
corrosion region because of an excess of
hydrogen ions at pH values below 4.5.*
(Reprinted from “Corrosion of Steel by Sulfides and
Cyanides in Refinery Condensate Water,” Materials
Protection, December 1968)
-
Figure 1.6 (above left) HCI and H,O
partial pressures over HCI depending
on temper-ature, total pressure, and
acid Concentration.’ (Reprinted with
permission from “Crude Unit Overhead
Corrosion,” Hydrocarbon Processing, May
1972, Gulf Publishing Co., Houston, TX)
-
Figure 1.8 (left) Corrosion rate of steel vs
temperature of atmospheric tower sour
condensate water?
(Reprinted with permisslon from “Determination of
Refinery Corrosion Rates Using the Pair Technique,”
TEMPERATURE, F Materials Protection, June 1969.
corrosion rate predicts that a 2.1 mm (0.083 in.) wall
tube will last 25 years (providing wall is not reduced
below the containing capacity of the tube).
However, experience reveals that the scale
formed is about seven times the volume of metal loss,
and the bundle would be completely plugged in four
years, assuming 4.8 mm (3/16 in.) ligaments on the
tubesheet. Square pitch bundles can be mechanically
cleaned, but triangular pitch bundles cannot. Unless
chemical cleaning can be done, triangular pitch
bundles must be retubed once they are plugged.
The 0.24 mm/y (9.3 mpy) corrosion rate (before
desalting) predicts a nine-year life. However, the
bundle would be completely plugged with scale in
one to four years. In both cases an inhibitor must be
used to reduce the corrosion rate and keep the tubes
clean.
‘i i
1 ° 8
I- -I
/
/
0
Figure 1.12 - 0
0
Comparison of sulfur levels
vs carbon steel corr-osion
rates.’2(Reprinted from J. Gutzeit,
“High Temperature Sulfidic /
Corrosion of Steels,” in Process /
Industries Comslon-Theory and
Practice, 1986)
10
,’
/
t- i
.01
0.4
k 0.8 1.2 1.6 2.0
Corrosion rate multiplier
Wt% H2SEvolved
r
0.03
0.03 -
0.02
0.02 -
0.01
0.01 -
0
6 o >
2 450 500 550 600 650 750 O F
Stock Temperature
-
Figure 1.13 Hydrogen sulfide release from -
Figure 1.I4 Corrosion of carbon steel
crudes.13 furnaces tubes.13
The 5 Cr-1/2 Mo alloy commonly used in furnace tubes corrodes at a measurable rate; the design life
usually is 5 to 10 y. Even though metal temperatures are higher than the process temperature in furnace
tubes, corrosion rates are always reported as a function of the process temperature. Therefore, furnace tubes
tend to corrode at higher rates than piping for the same process temperature. Conversely, exchanger tubes
corrode at a lower rate than piping when the process fluid is being cooled. Exchangers that have the cooling
medium on the shell side often corrode in the tube rolls, where the metal is at the hot stream temperature.
Flow velocity also plays a part since the high velocities often encountered in furnace tubes and piping tend
to accelerate corrosion.
Columns in crude units are usually clad with 12 Cr above about 288°C (550°F).Some refiners use clad
above 232°C (450°F) to minimize maintenance and fouling from corrosion products. The 12 Cr cladding is
essentially immune to attack and is used because the design life of columns (which are expensive to re-
place) usually is 20 years. Pressure retaining parts of 12 Cr usually are limited to 343°C (650°F) to avoid
problems resulting from 475°C (885OF)emb~ittlernent.(*~) As mentioned previously, the majority of attack
occurs in the flash zone, where the feed enters the column and partially flashes to vapor.
When naphthenic acids are present in sufficientquantities (see previous discussion), alloys with 0- 12%
chromium, as well as 3041304L (UNS S30400/S30403), 321 (UNS S32100) and 347 (UNS S34700) are
severely attacked. Conversely, molybdenum-bearing austenitic SS, e.g., 3 16/316L (UNS S3 1600/S31603),
3 17/317L (UNS S317OO/S31703), etc., exhibit resistance to naphthenic acid attack. Where hardfacing is
necessary, Stellite 6 (UNS R30006)(16)has been used successfully.
(I5) Loss of room temperature toughness due to precipitation of a chromium rich phase in the 370°C to 510°C (700°F
to 950°F) range.
(I6) 1 C, 4.5 W, 28 Cr, 55 Co.
Type 316 stainless steel
Figure 1.15 - Cage tray caps exposed to oil containing naphthenic acid for six months.
Figure 1.15 shows the severe attack that occurred on 12 Cr while 316 SS suffered no attack in naphthenic
acid service.
Figure 1.16 shows data compiled on the corrosion rate of carbon steel in naphthenic acid containing
oils as a function of temperat~re.'~ In vacuum processing of reduced crude oil (the bottoms from a crude
unit), the most severe corrosion often occurs at approximately 288°C (550°F). The reason is that the
naphthenic acids tend to concentrate in those fluids with true boiling points in the 370°C to 425°C (700°F
to 800°F) range, and the effect of a vacuum (in the vacuum column) is to reduce the boiling point 110°C to
160°C (200°F to 300°F).
Naphthenic acid attack can be minimized by limiting the velocities to 60 d s (200 ft/s) maximum;
however, the velocity limit of 40 m/s (1 30 ft/s) is preferred. Percent vaporization also is a factor. Some
refiners limit vaporization to 60%to minimize corrosion; however, corrosion of 316/316L SS (UNS S31600/
Next Page
1001 r2.54
L
4)
P
1 2 3 5 10
Total Acid Number, mg KOH/g
-
Figure 1.16 Naphethenicacid corrosion of carbon steel.''
S31603) has occurred in vacuum unit transfer lines containing a high naphthenic acid content with 15% to
40% vaporization. To minimize attack, both vaporization and velocity should be kept at a minimum.
The failures of 316/316L (UNS S31600/S31603) in naphthenic acid have been attributed to modem
steel-making practices, which result in molybdenum contents very close to 2%. It is believed that 2.5%
minimum molybdenum is required for naphthenic acid service, at least for California crude oils.(") How-
ever, 316/316L (UNS S31600/S31603)with 2.5% minimum molybdenum usually requires a special order
and usually results in long deliveries. Therefore, when 316/316L (UNS S31600/S31603) is not deemed
satisfactory, 3 17/317L (UNS S31700/S31703) usually is specified.
In existing units that were not alloyed for naphthenic acid resistance but now process crude oils con-
taining naphthenic acid, phosphate ester inhibitors have been reported to be useful in minimizing corrosion
until suitable replacement material can be installed. Also, refiners blend high TAN crude oils with low
TAN crude oils to keep the TAN of a feed below 1.5 mg KOWg. Often refiners blend to a TAN of 0.5 mg
to 1.0 mg KOWg.
When SS piping must be welded to a clad vessel, even small diameter ( 4 1 mrn [2 in.]) nozzles usually
are required to be clad or weld overlaid to avoid a dissimilar weld in the vessel. This is because of the high
stresses imposed (at service temperatures above about 232°C [450"F]) by the dissimilar thermal expansion
and because the hard fusion line in the dissimilar weld can sulfide stress crack in the presence of H,S (See
Section 6.1 in Chapter 2).
Where dissimilar welds are required, e.g., for 12 Cr clad restoration, 309 SS (UNS S30900) filler metal
commonly is used for dissimilar welds for service temperatures below 425°C (800"F), and high nickel [65
Ni-15 Cr-Fe] filler metal is used for dissimilar welds for service temperatures above 425°C (800°F). For
service temperatures above 399°C (750"F), the high nickel filler metal should be protected from high sulfur
environments except in delayed cokers, where severe corrosion has not been a problem because the coke on
(I7) One refiner has reported that currently produced 316/316L (UNS S31600/S31603) is satisfactory for Venezuelan
crude oils containing naphthenic acid.
2
Fluid Coking and Cracking,
Delayed Coking, Alkylation,
Sulfur Plants, and Sour Water Strippers
1. INTRODUCTION
The equipment used in fluid coking and fluid catalytic cracking (FCC) is similar. A catalyst is circulated
in a FCC unit, whereas the coke produced in a fluid coker circulates, providing heat and a particle on which
coke from cracking of the hydrocarbon can form. The products of fluid coking and delayed coking are the
same (i.e., coke and distillate products), but the equipment is physically different. Alkylation of the three-
and four-carbon molecule products from these units is commonly done to convert them to branched chain
compounds that increase the octane of gasoline. The feed to fluid catalytic crackers are gas-oil distillate
(Figure 2.1) or residuum. For delayed cokers and fluid cokers, the feed is residuum.
37
Scrubber Venturi
Fractionator Heater Gasifier Scrubber
-
I [Withdrawal I
Sulfur
Removal
=+
Steam \I
I/
Reactor % Steam
-
Figure 2.2 Fluid coking.2
(Reprinted with permission of Gulf Publishing Co., Houston,TX)
before it flows to the regenerator. In the regenerator, the catalyst is regenerated by the burning of the coke at
about 650°C to 760°C (1,200"F to 1,400"F). Entrained catalyst is again removed from the gases before the
gases leave the top of the vessel. The hot regenerated catalyst passes out of the bottom of the regenerator
where it again is combined with fresh feed plus some recycle from the fractionator.
In fluid coking (Figure 2.2); the reactor also operates at about 480°C to 540°C (900°F to 1,000"F), but
the pressure is 172 Waa to 240 Waa [25 psia to 35 psia]). The residuum feed is injected into the reactor,
where it is cracked thermally in a fluidized bed catalyst. Products other than coke leave the top of the reactor
and are quenched in a scrubber, where residual coke is removed. The coke fines, and some of the heavy
fractions are recycled to the reactor. The lighter fractions are fed to conventional fractionating equipment.
The coke (coke laden catalyst in an FCC) drops to the bottom of the reactor and is circulated to the
heater by steam. In the heater, the volatile products are driven off. A circulating coke stream is sent to the
gasifier, where 95% or more of reactor coke is gasified with steam and air.
The mechanical designs of fluid coker and fluid catalytic cracker reactors and regenerators (called burners
or heaters in fluid cokers) are essentially the same. These vessels are often 7.6 m to 9.2 m (25ft to 30 ft) in
diameter and 24.5 m to 30.5 m (Soft to 100 ft) high. Walls are usually made of 12.5 mm to 25 mm (0.5 in. to
1 in.) thick carbon steel with refractory lining. The refractory lining is in two layers and is applied by a spray
gun technique (hence the term gunite). The first layer against the vessel wall is an insulating refractory
(temperatures in the regenerators often reach 705°C [1,300"F]) held on by Y-anchors (Nelson-studs) on the
wall. A 12 Cr hexmesh (similar to walkway grating) is then attached to the studs, and an erosion-resistant
CATALYST DENSITY (kglm’)
240
220
200
180
160 f
1 140 3
s 120
F
G
0
2> 100
3
80
60
40
20
-
Figure 2.3 Use of erosion-resistant linings
in fluid catalytic cracking units as function of catalyst density and velocity?
refractory is applied over the mesh. Erosion-resistant linings are required because of the fluid catalyst
circulating in the system. The extent of the erosion problem is a function of velocity and catalyst density
(Figure 2.3).3
For the newer high-temperature units, SS fiber-reinforced refractory linings are used because the thermal
expansion of the hexmesh refractory support is sufficient to cause cracking problems in the refractory lining.
The erosion-resistant lining is required to resist wear from the continual cycling of the catalyst through the
two vessels. Smaller units have been made from clad construction without refractory lining. In these cases,
the use of carbon steel is limited to 454°C (850°F)to avoid loss of strength due to long-term graphitization.
Graphitization is the formation of graphite (free carbon) in steel by the decay of iron carbide.
The flow pattern in a fluid coker is similar to that in a fluid catalytic cracker except that the circulating
solids are coke. Part of the coke is burned in the burner or heater vessel, and hot coke is circulated to the
reactor.
The cyclones in both types of units are lined with erosion-resistant refractory. The cyclones in the
reactors usually are made of carbon steel, or C-1/2 Mo; the cyclones in the regenerators are made of either 2-
1/4 Cr-1 Mo,304H (UNS S30409) or 321H (UNS S32109). Oxidation resistance and creep strength, rather
than sulfur attack, govern the selection of the cyclone material.
The fluid cracker and coker vessels are often field-erected because of their large size. This means they
must be hydrostatically tested for the first time in the field; this poses the risk of brittle fracture. Figures 2.4
to 2.6 show a brittle fracture of a fluid coker burner vessel.4 Figure 2.4 shows an overall view of the
collapsed vessel and gives an idea of the extent of the catastrophe. The debris from the failure under the
support structure is shown in Figure 2.5. This gives an impression of the effect the failure had in the immediate
area. Figure 2.6 shows the numerous paths the cracks took on the bottom head where t h ~brittle , fracture
originated. The incident occurred February 17, 1956, in the Avon, California refinery during construction of
a fluid coker unit. The vessel, 11 m (36 ft) in diameter by 23 m (75 ft) high, was made from 16 mm (5/8 in)
thick A285C (UNS K02801) steel. It failed in the knuckle (ASME head) to skirt attachment during hydrotesting
when the ambient temperature was 7°C to 10°C (45°Fto 50°F). Since this failure, most designers have used
only ellipsoidal or hemispherical heads. They require a minimum temperature, typically 27°C (80"F), for
hydrostatic tests of field-erected vessels.
The valves controlling the flow between the reactor and regenerator (or burner) are specially designed
slide valves (box-like valves where the gate slides in guides). They have weld overlay hardfacing (e.g., a
cobalt base alloy containing28 Cr, 4 W, 1C) on the stems, and sliding surfaces and erosion-resistant refractory
lining in the body. The hardfacing weld beads should be deposited parallel to the flow direction to minimize
turbulent flow.
Type 304H SS (UNS S30409) commonly is used for the small high-temperature lines around the reactor
and regenerator. Intergranular stress corrosion cracking (IGSCC)from polythionic acid (*) has been a severe
problem in these lines during shutdown. Presently, a nitrogen and ammonia purge is introduced at 93°C
(200°F) above the dew point during cooldown to prevent attack. Expansion bellows in the regenerated
catalyst standpipe have been subject to failure. Alloy 625 (UNS N06625) has failed due to loss of ductility
-
Figure 2.4 Front
view of collapsed
vessel. Test clos-
ure for vapor line
opening in top
center of top head
at upper left?
(Reprinted with
permission of the
American Petroleum
Institute)
3. DELAYED COKING
In delayed coking (Figure 2.7),' the chargeis fed to the fractionator. The bottom product of the fractionator
is then fed to the heater, where it is heated to coking temperature. A mixture of liquid and vapor from the
coker heater is fed to the coke drum, which operates at about 480°C (900°F)and 345 kPaa (50 psia).
The main reaction in a delayed coker occurs in pairs of vessels called coke drums. Feed temperature to
a coke drum is 490°C to 5 10°C (920°F to 950°F); outlet temperatures are 425°C to 440°C (800°F to 820°F).
Delayed coking is a batch process where the vessels am in coking service for about 24 h then in decoking
service for about 24 h. This causes the vessels to cycle between mom temperature and about 510°C (950°F)
about every 48 h. While the reaction is taking place in one vessel (i.e., while coke is filling the vessel), the
other vessel is being decoked. A drum is decoked by:
The vapor from the overhead of the coke drums is returned to the fractionator. The overhead vapor is
separated into its components in the fractionator, and products are withdrawn. The heavier fractions recycle
to the coke drum in the fractionator bottoms.
Because of the thermal cycling experienced by the coke drums, thermal fatigue is a problem. There is
some indication from recent work that controlling the initial quench rate may greatly reduce the thermal
fatigue problem. The use of C-lE Mo for coke drum vessels is controversial. Cracks do develop in these
vessels because of thermal fatigue; however, some refiners have blamed some of the cracking on poor
'SO'F
--.,
GAS OIL
r!
f
HEATER
rl n
STM = STEAM
FEED
-
Figure 2.7 Delayed coking.'
(Reprinted with permission of the American Iron and Steel Institute)
toughness of C-1/2 Mo. Although toughness does not affect fatigue crack growth (fatigue crack growth is a
ductile fracture process), brittle fracture will occur in fewer thermal cycles in a low-toughness steel than in
a high-toughness steel. This is because high toughness steel can tolerate a Iarger fatigue crack than a low-
toughness material.
Since C-112 Mo plate is supplied in the as-rolled condition in the thickness normally used for coke
drums (about 25 mm [ 1 in.] thick), the toughness is usually poor. To obtain improved toughness at minimal
cost, one should specify that the C-1/2 Mo be made to fine-grain practice. Many users also specify normalizing,
which significantly increases toughness. Most coke drums have been made of 1 Cr-112 Mo materials for
about the past 15 years because of the controversy over C-1/2 Mo and the higher allowable stress of 1 Cr-l/
2 Mo. After the proper toughness requirements are specified, e.g., 205 at -18°C (15 ft-lbs at 0°F) for C-1/2
Mo, the ultimate choice of material should be based on cost.
Delayed cokers usually are clad with 12 Cr to avoid high-temperature sulfur attack. For cladding, type
410s SS (UNS S41008) is preferred over 405 SS (UNS S40500) because the higher chromium in 405 SS
causes problems with forming and with 475°C (885°F) embrittlement (discussed in Chapter 1).
Type 309 SS (UNS S30900) normally is used for clad restoration on 12 Cr clad vessels. However,
severe cracking has occurred when 309 SS (UNS S30900) was used for clad restoration on coke drums. This
is because of the thermal cycling of materials with the large differences in coefficients of thermal expansion,
namely the 309 SS (UNS S30900) weld metal ,the 12 Cr clad, and the low-alloy steel backing. High nickel
(65 Ni-15 Cr-Fe) filler metal performs significantly better than 309 SS (UNS S30900) for clad restoration in
coke drums because of the lower coefficient of thermal expansion, and the coke protects the 65 Ni- 15 Cr-Fe
filler metal from accelerated sulfur attack. Some refiners grind both the inside (clad restoration) and outside
(base metal welds) to minimize stress raisers that accelerate thermal fatigue. It is important to note that coke
drums are the only high-temperature sulfur-containingenvironmentin which the 65 Ni- 15 Cr-Fe alloy materials
normally are used.
FRESH
ACID_
i
l
WATER
B
r
U
;
STM
9
i
RERUN COLUMN
cw HEAVY
ISOBUTANE
t
1
IS0 RECYCLE CW-COOLING WATER
STM-STEAM
-
Figure 2.8 Sulfuric acid alkylation.'
(Reprinted with permission of the American iron and Steel institute)
Although there have been reports of good performance of 65 Ni- 15 Cr-Fe filler metal in coker fractionators
up to 400°C (750"F), presumably because of protection by the coke that occurs in coke drums, severe
sulfidation has occurred in a crude fractionation tower when 65 Ni-15 Cr-Fe type electrodes were used
above 288°C (550°F). Furthermore, 65 Ni-15 Cr-Fe filler metal is more expensive and more prone to weld
cracking than 309 SS (UNS S30900), normally used except in coke drums.
4. ALKYLATION PLANTS
In sulfuric acid (effluent refrigeration) alkylation (Figure 2.8),' the feed and the recycle sulfuric acid are
charged to the contactor. The reaction takes place at about 4°C to 10°C (40°F to 50°F) at 345 kPaa (50 psia)
in the presence of 98% sulfuric acid. The sulfuric acid in the effluent of the reactor is removed in the acid
settler, the trap and flash drum, the caustic wash, and the water wash before the alkylate is fed to the
fractionators.Sulfuric acid in alkylation plants is handled primarily in carbon steel as long as the concentration
of the sulfuric acid does not fall below about 80%; however, most operators require 89% minimum. Since
corrosion of carbon steel in concentrated sulfuric acid is accelerated over 50°C (120"F), electrical heat
tracing, rather than steam tracing, is used to prevent freezing in cold climates.
Most refiners specify that the acid piping be sloped so it is self-draining. If the velocity exceeds 0.6 m/
s to 0.9 m/s (2-3 ft/s) for carbon steel, alloy 20 (UNS N08020) is used. Although 316L (UNS S31603) can be
used up to 1.2 d s e c (4 ft/sec) without accelerated erosion-corrosion,the use of alloy 20 (UNS NO8020) is
much more common. Where erosion-corrosion is a problem on alloy 20, e.g., on mixers, alloy C276 (UNS
N10276) or Stellitem 21 (UNS R30021 [0.25 C, 27 Cr, 5 Mo, 2.8 Ni, 65 Co]) hard facing is used. Small
valves (less than 152 mm [6 in.] in diameter) usually are made of alloy 20 (UNS N08020), and larger valves
are made of steel with alloy 20 trim (stem and seats).
Significant corrosion has been reported in dead legs and socket welds in sulfuric acid systems; therefore,
these should be avoided. In addition, hydrogen grooving has occurred on the upper portions of horizontal
tanks, manways, nozzles, and piping on surfaces exposed to liquid sulfuric acid. The grooving is the result of
hydrogen from the corrosion reaction non-uniformly disrupting the protective iron sulfate film. The
phenomenon can be prevented by maintaining a minimum velocity of 0.3 m / s (1 ft/s) or by use of 3 16L
(UNS S31603) or alloy 20 (UNS NO8020). For more information on hydrogen grooving, see Brubaker and
TatnalL5
When proper treatment of the effluent of the reaction section is not maintained, fouling and corrosion
from SO, combining with water in the overhead can occur. A pH below 6 in the water from the accumulator
boot indicates corrosion is occurring. Steel cases and high silicon cast iron impellers often are used for
pumps; however, they only last five to ten years. Critical pumps often are made of alloy 20 (UNS N08020)
(cast equivalent, ASTM A351 Grade CN-7M). For additional information on materials for handling and
storage of concentrated sulfuric acid at ambient temperatures, see NACE Standard RP0391, “Materials for
Handling Storage of Concentrated (90% to 100%) Sulfuric Acid at Ambient Temperatures.”
The feed to a hydrofluoric acid alkylation unit is desiccant-dried, then sent to the combined reactor
settler (Figure 2.9).* The reaction takes place at 32°C to 38°C (90°F to 100°F) at 1,725 Waa (250 psia) in
the presence of 90% hydrofluoric acid. The effluent from the combined reactor settler is fed to the main
fractionator. The hydrofluoric acid goes overhead with the light ends and is condensed and collected in the
accumulator. Part of the condensed overhead fluid is recycled to the feed to the combined reactor settler, part
is used for reflux to the main fractionator, and the remainder is fed to the hydrofluoric acid stripper. The
overhead of the stripper is returned to the main fractionator overhead condenser. The bottom product of the
stripper is caustic washed to remove all traces of hydrofluoricacid. The bottom product of the main fractionator
NOTES:
1. Shaded areas are where Yonelm 400 is used.
2. Hydrogen blistering or cracking has occurred
3. Corrosion reported.
often is fed to a debutanizer fractionator column.
Carbon steel is the primary material of construction for equipment used when the concentration of
hydrofluoric acid is above 70% to 80%. The locations where hydrogen blistering, cracking, or corrosion has
been reported, as well as where alloy 400 (UNS N04400) is required, are shown in Figure 2.9.lIt is important
to thoroughly dry equipment after a shutdown to avoid corrosion caused by pockets of water left in the
equipment diluting the acid so it is corrosive to carbon steel. Drying is aided by specifying that the acid
piping be sloped so it is self-draining. It has been reported that limiting the sum of the copper, nickel, and
chromium residual elements to 0.02%maximum will minimize corrosion of carbon However, galvanic
corrosion problems have been reported in this low residual element carbon steel when it is coupled to weld
metal or base metal containing high quantities of residual elements.
For small diameter piping, screwed connections with PTFE tape are preferred over socket or seal-welded
threaded connections to avoid crevice corrosion. Hydrogen embrittlement of high strength steels has been a
problem; steels of tensile strengths exceeding a specified minimum of 414 MPa (60 ksi) and hardnesses over
200 Brine11 usually are avoided. Limiting the arsenic in the fresh acid to 25 ppm often is done to minimize
the potential for hydrogen embrittlement. Flange bolts may be ASTM A193 B7, although B7M is preferred;
but valve bonnet bolts should be ASTM A193 B7M or alloy 400 (UNS N04400)because of leakage. Alloy
K-500 (UNS N05500) bolts have failed where the hardness exceeded HRC 30. Alloy 400 (UNS N W 0 0 )
spiral wound gaskets filled with PTFE or graphite with PTFE inner rings usually are specified. Where
corrosion of flange faces has occurred, alloy 400 (UNS N04400) overlay usually is specified for repair.
Weld slag on the surface of carbon steel welds has been reported to cause accelerated attack. Therefore,
whenever possible, slag-free processes such as gas-tungsten arc should be used. Alternatively, special care
should be taken to ensure slag is removed.
In higher temperature areas (e.g., the acid reboiler and the bottom of the acid regenerator), alloy 400
(UNS N04400) is used. Butt-welded construction is preferred even in small diameter piping to avoid crevice
corrosion. Alloy 400 (UNS N04400) is subject to stress corrosion cracking (SCC) in hydrofluoric acid if air
gets into the system. As a precaution, some refiners specify stress relief of solid alloy 400 (UNS N04400).
Valves are carbon steel with alloy 400 trim and PTFE packing. Spot chemical testing of alloy 400 usually is
specified because it is easily confused with austenitic SS. For additional information on materials for handling
hydrofluoricacid, see NACE Publication 5A17 1, "Materials for Receiving, Handling, and Storing Hydrofluoric
Acid."
5. FRACTIONATION
After the reactions have taken place in the catalytic crackers, cokers, and alkylation plants, the effluents
are separated by fractionation to obtain the various products desired. The materials for the fractionation
equipment are selected using basically the same criteria as those used for crude units, as discussed in Chapter
1 (i.e., carbon steel below 288°C to 315°C [550"Fto 600"FI and 5 0-112 Mo piping, and 12 Cr clad vessels
above). Because the temperatures in coker fired heaters are higher than in crude units, furnace tubes and
furnace transfer lines in delayed cokers usually are 9 Cr-1 Mo.
Coking and catalytic cracking produce cyanides and hydrogen sulfide in the water condensed in these
processes. Very little water is produced in the alkylation process. The traces of acid from the alkylation units
as well as cyanides (greater than 20 ppm) from coking and catalytic cracking cause corrosion problems in
the tops of columns, in the overhead lines, and in the drums where the aqueous phase condenses. Cyanides
are very active if the pH is in the alkaline range. Cyanides accelerate corrosion by converting normally
protective iron sulfide films into soluble ferrocyanide complexes. Alkaline conditions usually occur because
the nitrogen in the streams that produces the cyanide also produces ammonia.
The corrosion problems caused by these produced waters are hydrogen blistering '-" and sulfide stress
cracking (SSC)"' in carbon steel and pitting and stress corrosion cracking of admiralty (UNS C44300)
exchanger tubes. The blisters in carbon steel are caused by the hydrogen generated in the corrosion process
recombining at voids and inclusions in the steel. The most severe hydrogen blistering has been found in
deethanizer columns in fluid catalytic cracker gas plants. The tendency to form blisters can be monitored by
putting hydrogen probes on the system. A probe is mounted on the outside of the equipment. The current
measured by the instrument is directly related to the hydrogen flow through the wall. Water (to dilute the
corrosive constituents), liquid polysulfide' (to react with and remove cyanides), and inhibitors have been
used to avoid blistering. However, polysulfide additions have caused fouling and accelerated corrosion in
stagnant areas and may increase the likelihood of carbonate cracking in FCC overheads. Therefore, poly sulfide
additions are only used when other control methods are not effective. In some catalytic crackers, sufficient
excess oxygen is used in the regenerator to cause the sulfides to convert to polysulfides. As a result, corrosion
in the sour water from these units is minimal.
Studies of corrosion in oil and gas fields indicate that SSC becomes a problem when the partial pressure
-
Figure 2.1 0 Sour gas systems.*
0 ) The formation of blisters on or below the metal surface from excessive internal hydrogen pressure.
Brittle fracture by cracking under the combined action of tensile stress and corrosion in the presence of water and
(4)
of hydrogen sulfide exceeds 0.35 kPaa (0.05 psia) and the total pressure exceeds 450 kPaa (65 psia). Typical
refinery practice takes into account only the partial pressure, or the ppm of H,S, i.e., not the total pressure,
when determining the potential for SSC. Figures 2.10 and 2.1 I s can be used to predict where SSC will occur
as a function of total system pressure vs hydrogen sulfide concentration. McIntyre and Boah have suggested
a more conservative definition of sour service in which they ignore total pressure as a criterion, then define
sour service as a function of the pH of the water phase and the H,S partial pressure as f01lows:~
pp H,S > 0.3 kPaa (0.05 psia) if the water phase pH > 4.0
pp H,S > 0.05 Waa (0.008 psia) if the water phase pH < 4.0
(5’API Standard RP-942 has been discontinued because it covered the same topic as NACE Standard RP0472.
and Controls to Prevent In-Service Cracking of Carbon Steel welds in P- 1 materials in corrosive Petroleum
Refining Environments"), and NACE Publication 8x194. NACE Standard RP0296 contains guidelines for
detection, repair, and mitigation of cracking in wet H,S (SSC) in existing vessels. NACE Publication 8x194
summarizes current thinking on requirements for pressure vessels in wet H,S service in refineries. RP0472
is limited to welds in carbon steel equipment in refinery service. It requires a maximum of 200 Brinell for
carbon steel welds, whereas MR0175, based on oil and gas field experience, requires a maximum of HRC 22
(HB 241) for carbon steel. MR0175 also has requirements for other materials used in H,S service. The
difference in hardness requirements for carbon steel may have arisen because submerged arc and other
automatic welding processes commonly are used for refinery equipment, whereas most carbon steel oil and
gas field equipment is welded by the manual metal-arc process, which tends to deposit softer welds than do
automatic welding processes. MR0175 is specified for refinery equipment, such as valve trim, springs, and
internal bolting. For example, floating head bolting is required to be ASTM A193 B7M, where B7 normally
would be used when water and H,S are present.
Under certain conditions, steels with hardnesses lower than 200 Brinell will crack in H,S service. Crack-
ing has been reported in low-temperature separators in hydrocrackers and hydrodesulfurizers, in fluid cata-
lytic cracker and fluid coker overhead equipment, and in amine H,S absorbers. Refiners are now specifying
PWHT in these services. This phenomenon is referred to as hydrogen-induced cracking (HIC) or stress-
oriented hydrogen-induced cracking (SOHIC). HIC is similar to blistering, except that after the blisters are
formed, they join up to form internal fissures parallel to the metal surface. SOHIC is the growth of HIC
cracks due to the influence of applied or residual stress. It is believed that PWHT reduces potential for HIC
and SOHIC by reducing the driving stress for crack propagation and by increasing the resistance to crack
propagation by increasing the notch toughness. See Chapter 5, Section 8.2, for additional discussion of HIC.
In addition to PWHT, some steel companies are suggesting limiting sulfur in steel to 0.001% to mini-
mize failures. Niobium-vanadium steels are used so the high strength can be maintained without the need for
high carbon. Conversely, there have been SSC problems with steels of normal carbon contents containing
microalloying elements. PWHT at 635°C (1,175"F) minimum is required to lower the heat-affected zone
hardness below 248 HVN when the Nb + V is greater than 0.01%. In addition, the effect of PWHT on both
hardness and residual stress should be verified for Nb-V steels. For example, Nb-V steel that had been
PWHT'd at 593°C (1,100"F) minimum failed in amine service due to stress corrosion cracking. Subsequent
laboratory tests revealed that relief of significant residual stress did not occur until the PWHT temperature
reached a minimum of 676°C (1,250"F).
Calcium treating the steel and limiting the oxygen to less than 10 ppm have been suggested as ways to
minimize HIC and SOHIC. Recent experience indicates calcium treatment may be detrimental to HIC resis-
tance and that so called "HIC-resistant steels" have poorer resistance to SOHIC than steels not made to HIC-
-
Figure 2.12 Sour water
stripper (non-acidified).'
(Reprinted with permission of
the American Iron and Steel
institute)
I TREATED WATER
STM - STEAM
TO DISPOSAL
resistant requirements." Furthermore, HIC cracking has been reported in electroslag remelted steel, which is
essentially inclusion free. Thus, at this writing the jury is still out on whether the additional cost for so called
"HIC-resistant steels" can be justified.
SSC also has been reported in the narrow fusion line between carbon or low alloy steel dissimilar weld
when it is welded with an austenitic electrode and not post-weld heat-treated. Hardnesses in the fusion zone
have been found to be in the 300 to 400 Brine11 range in the as-welded condition.
The sour or foul water from all units in a refinery is collected and sent to a stripper for purification
before it is discharged into the environment. A typical process flow diagram of a non-acidified condensing
sour water stripper is shown in Figure 2.12.' In this unit, the sour water enters the stripper vessel through the
feed bottoms exchanger. The hydrogen sulfide and other corrosive gases are removed by the heat supplied
by the reboiler connected to the bottom of the stripper vessel. The overhead stream is condensed then collected
in the reflux drum. The liquid from the reflux is recycled to the stripper, and the gases are either burned or
fed to a sulfur plant.
The main corrosion problems in a non-acidified condensing sour water stripper occur in the overhead
system. Exchanger tubes and headers in the overhead condenser often are titanium. Grade 2 (UNS R50400)
is used for tubes and Grade 3 (UNS R50550)for headers. It usually is not possible to electrically isolate the
titanium from dissimilar metal (i.e., carbon steel, SS, etc.) piping by insulating flanges. Therefore, there is a
potential for embrittlement from hydriding of the titanium when the metal temperature is in excess of 80°C
(176°F). For this reason, titanium headers often are specified when titanium tubes are used. Should any
hydriding occur, it will be primarily in the thicker headers, thus giving more time to be discovered during the
recommended periodic inspection for hydriding.
In low pH environmentsat temperatures above 110°C(225"F),the high strength Grade 12 (UNS R53400)
offers somewhat better resistance to corrosion in tight crevices and under deposits. However, Grade 12
(UNS R53400) is more susceptible to hydriding than Grade 2 (UNS R50400). The line from the condenser
to the reflux drum often is 316L SS (UNS S31603), and the reflux drum often is clad with 316L (UNS
S31603). The reflux pump often is alloy 20 (UNS N08020). In very corrosive waters, such as those containing
phenols or large quantities of salts, alloy C276 (UNS N10276) is used. The velocity in carbon steel piping is
limited to 6 m/s (20 ft/s) in two-phase flow and 15 m/s (50 ftfs) in vapor lines. Carbon steel towers have been
satisfactory in most units where the overhead is kept above 80°C (180°F). Two API surveys (API Publications
944 and 950) concluded that the location and severity of corrosion vary with the type of unit, as follows: I 2 . l 3
Acidified: severe corrosion in the feed section, bottom section, and stripper tower,
-
Figure 2.13 Sulfur plant. *
(Reprinted with permission of Gulf Publishing Co., Houston, TX)
sulfide is partially burned at about 1,370"C(2,500"F)and 103 kl'aa (15 psia) in the reaction furnace to form
sulfur dioxide; next the mixture of H,S and SO, is passed through a waste heat boiler then passed over
catalyst beds at about 260°C (500°F)and 103 Waa (15 psia) in the converters. Sulfur is condensed from the
effluent of successive converters, then solidified in pits.
7.2 Materials
The effluent from the reaction boiler is handled in 310 SS (UNS S31000) above 425°C (800"F), 321 or
347 SS (UNS S32100 or S34700) above 260°C (500"F),and carbon steel below 260°C (500°F). The acid
gas (H,S, CO,, and S0,)are kept above the dew point by, e.g., steam tracing; otherwise, severe corrosion of
carbon steel lines will occur. Molten sulfur is handled in steam-traced steel or aluminum. At the discharge to
the pits, oxygen causes severe attack on steel, so the discharge end of the steel line often contains a short
piece of alloy 20 (UNS N08020).
REFERENCES
1. “The Role of Stainless Steels in Petroleum Refining,” American Iron and Steel Institute, April
(1977).
2. Refinery Process Handbook, Hydrocarbon Processing September ( 1968).
3. Private communication
4. A.G. Harding, E.F. Ehmke, “Brittle Failure of a Large Pressure Vessel,” API Div. Rejning, vol. 42,
no. 3 (1962): p. 107.
5 . S.K. Bmbaker, R.E. Tatnall, “Materials Selection and Design to Minimize Hydrogen Grooving,”
Materials Performance May (1995): p. 59.
6. H.H. Hashium, W.L. Valerioti, “Effect of Residual Copper, Nickel and Chromium on the Corrosion
Resistance of Carbon Steel in Hydrofluoric Acid Alkylation Service,” CORROSION/93, paper no. 623
(Houston, TX: NACE International, 1993).
7. E.F. Ehmke, “Polysulfide Stops FCCU Corrosion,”-Hydrocarbon Processing, 60,7 (1981): p. 149.
8. NACE Standard MROl75 (latest revision), “Sulfide Stress Cracking Resistant Metallic Material for
Oil Field Equipment” (Houston, TX: NACE).
9. D.R. McIntyre, J.K. Boah, “Review of Sour Service Definitions,” Materials Performance August
(1996): p. 54.
10. L.M. Smith et al., Part 1, D.R. McIntyre, “Technical Forum: Review of Sour Service Definitions,”
Materials Performance April (1997): p. 66.
11. M.S. Cayard et al., “An Exploratory Examination of the Effects of SOHIC Damage on the Fracture
Resistance of Carbon Steels,” CORROSION/97, paper no. 525 (Houston, TX: NACE, 1997).
12. API Publication 944, Survey of Materials Experience and Corrosion Problems in Sour Water
Equipment (Washington, DC: American Petroleum Institute, 1972).
13. API Publication 950, “Survey of Construction Materials and Corrosion in Sour Water Strippers -
1978 (Washington, DC: American Petroleum Institute, 1983).
BIBLIOGRAPHY
API Publication 942, “Recommended Practice for Welded, Plain Carbon Steel Refinery Equipment for
Environmental Cracking Service,” (Washington, DC: American Petroleum Institute).
API Recommended Practice 75 1, “Safe Operation of Hydrofluoric Acid Alkylation Units, 1st edition
(Washington, DC: API, June 1992).
Bergman, D.J., and G.W.G. McDonald, “Solving Problems in HF Alkylation Units,” Corrosion 17, 4
(1961): p. 9.
Bradbury, W.A. “Monolithic Linings for Refinery Service,” Petroleum Refiner 33, 3 (1954): p. 167.
Copkson, H.R., and C.F. Cheng, “Stress Corrosion Cracking of Monel in Hydrofluoric Acid,” Corrosion
12, 12 (1956): p. 647t.
Dean, S.W., G.D. Grab, “Corrosion of Carbon Steel Tanks in Concentrated Sulfuric Acid Service,”
Materials Pelformance 25, 7 (1986): p. 48.
T.F. Degnan. “Effect of Velocity on Life Expectancy of Steel Pipelines in Commercial Strengths of
Sulfuric Acid,” Corrosion 15, 6 (1959): p. 326t.
Dobis, J.D. et al., “Survey Reveals Nature of Corrosion in HF alky units,” Oil and Gas Journal 93, 10
(March 6, 1995): p. 63.
Ehmke, E.F. “Polysulfide Stops FCCU Corrosion,’’ Hydrocarbon Processing 60, 7 (1981): p. 149.
Foroulis, Z.A. “Mechanism of Refinery Corrosion by Aqueous Sour Water Condensates,” 6th Middle
East Corrosion Conference (Manama, Bahrain: Bahrain Society of Engineers, January 1994): p.
17.
J. Gutzeit, “Corrosion of Steel by Sulfides and Cyanides in Refinery Condensate Water,” Materials
Protection (Dec. 1968): p. 17.
J. Gutzeit, “Process Changes for Reducing Pressure Vessel Cracking Caused by Aqueous Sulfide
Corrosion,” Materials Protection (May 1992): p. 60.
Hansen, R.D. “FCC Slide-Valve Problems Can Be Remedied,” Oil & Gas Journal 8, 16 (April 18,
1983): p. 139.
R.L. Hildebrand. “Sour Water Strippers - A Review of Construction Materials,” Materials Perform-
ance 13, 5 (1974): p. 16.
Holmberg, M.E. and F.A. Prange, “Corrosion in Hydrofluoric Acid Alkylation,” Znd & Eng. Chem.
37, 11 (1945): p. 1,030.
INCO Corrosion Engineering Bulletin CEB-1, “Resistance of Nickel and High Nickel Alloys to Corro-
sion by Sulfuric Acid,” (Suffern, NY: INCO, 1983).
INCO Corrosion Engineering Bulletin CEB-5, “Corrosion Resistance of Nickel-Containing Alloys in
Hydrofluoric Acid,” Hydrogen Fluoride and Fluorine (Suffern, NY: INCO, 1968).
Kmetz, J.H., and D.J. Truax, “Carbonate Stress Corrosion of Carbon Steel In Refinery FCC Main Frac-
tionator Overhead Systems,” CORROSION/90, paper no. 206 (Houston, TX: NACE, 1990).
Mason, J.F. Jr. and C.M. Schillmoller, “Corrosion in Sour Water Strippers,” Corrosion 15, 7 (1959): p.
36.
McGowin, T.S. and R.A. White, “Coke Drum Fracture Experience,” APZ Div. Refining, 52, 3 (1972):
p. 778.
Memck, R.D. “Refinery Experience with Cracking in Wet H,S Environments,” Materials Performunce,
27, 1 (1988): p. 30.
Moore, K.L. “Alloy Stops Corrosion in Fluid Coker,” Petroleum Refiner 39, 5 (1 961): p. 1,949.
Neumaier, B.W. and C.M. Schillmoller, “What to Do about Hydrogen Blistering,” Petroleum Refinery
36, 9 (1957): p. 3 19.
“Summary of Questionnaire Replies on Corrosion in HF Alkylation Units,” Corrosion 15, May
(1959): p. 33.
Thornton, D.P. Jr., “Corrosion-Free AF Alkylation,” Chemical Engineering 77, 14 (July 13, 1970):
p. 108.
Weil, N.A., and F.S. Rapasky, “Experience with Vessels of Delayed Coking Units,” APZ, Division of Re-
fining, 38, I11 (1958): p. 214.
3
Hydroprocessing,
Catalytic Reforming, and Flue Gas
1. INTRODUCTION
Some sulfur compounds are cracked to hydrogen sulfide and removed in the crude unit. The sulfur
compounds remaining in the processed crude oil must be removed if the lighter petroleum fractions are to be
reformed into products, such as gasoline, etc. Desulfurizing also is necessary (1) to meet the sulfur limits
placed on kerosene and diesel products, (2) to prevent the sulfur from “poisoning” catalytic reformer cata-
lyst, and (3) to reduce sulfur dioxide emissions from the fluid catalytic regenerator flue gas.
2. HYDROPROCESSING UNITS
2.1 General
The most common hyroprocessing units are hydrodesulfurizers (Figure 3.1)’ and hydrocrackers (Figure
3.2).’ The primary objective in hydrodesulfurizing is to remove sulfur compounds and nitrogen by convert-
ing them to H,S and NH,,respectively. The amine scrubbing of the recycle gas shown in Figure 3.1 is not
always done. In addition, water wash sometimes is recycled from the separator. The hydrocracking process
converts gas oils to middle distillates as well as converts sulfur compounds and nitrogen to H,S and NH,,
respectively. Amine scrubbing of the recycle gas sometimes is done. In addition, water wash sometimes is
recycled from the separator. The shell and tube reactor effluent trim coolers shown in both Figures 3.1 and
3.2 are used only occasionally. The feed to hydrodesulfurizers is much heavier material than that is fed to a
hydrocracker. In both types of units the feed is first passed to exchangers, where it is heated by the reactor
effluent.
The feed next goes to a fired heater then to the reactors at about 370°C to 400°C (700°F to 750°F). The
pressures in hydrodesulfurizersnormally range from 2,415 Waa to 13,790 kPaa (350 psia to 2,000 psia); the
pressure in the hydrocracker ranges from about 10,340 Waa to 20,685 kPaa (1,500 psia to 3,000 psia).
The sulfur (and nitrogen in a first-stage hydrocracker) is removed by adding hydrogen to the feed,
heating this mixture in a furnace, and passing it over a catalyst at high pressure in reactor vessels. The
molecules are broken down in the reactor, and the released sulfur reacts to form hydrogen sulfide and
mercaptans. In addition, the organic nitrogen compounds are converted to ammonia and some hydrogen
cyanide (in first-stage hydrocrackers). In hydrodesulfurizers, the reactor effluent is cooled through a series
of exchangers, then sent to a high-pressure separator vessel, where the gas is taken off the top to a unit to
remove the hydrogen sulfide. The gas, with most of the hydrogen sulfide removed, is recycled to the feed.
The liquid from the high-pressure separator is passed through a letdown valve to a low-pressure separator.
The liquids from the low pressure separator are then fed to a fractionator.
In a hydrocracker, the effluent of the first stage also is cooled and sent to a high-pressure separator. In
55
Figure 3.2 - Hydrocracking.‘
(Reprinted with permission from “The Role of Stainless Steels in Petroleum Refining,”
American Iron and Steel Institute April, 1977)
Figure 3.3(b)
-
(left) Photomi-
crograph showing
fissuring in weld
metal. Nital etch,
1oox.2
Figure 3.3(c)
(below) -
Photomicrograph
showing structure
of parent metal.2
-
Figure 3.3(a) (above) Cross section of
exchanger. Note cracking in weld metal at right
(light colored area) of specimen.2
many of the hydrocracking processes, the liquid from the high-pressure separator is then fed to a heater and
to a second stage reactor. The effluent from the second stage follows the same scheme as in a hydrodesulfurizer,
i.e., exchange cooling, high- and low-pressure separation, then fractionation. The bottom product of the
fractionator is recycled to the feed of the second stage hydrocracker.
For materials selection, hydrocrackers are treated the same as hydrodesulfurizers, particularly in the
first stage. From a materials standpoint, the demarcation between low pressure units (hydrodesulfurizers)
and high pressure units (usually hydrocrackers) is 4,480 kPaa (650 psia).
When steel is exposed to hydrogen at high temperatures, the hydrogen enters the steel, forming a con-
centration gradient through the steel, as indicated in Appendix C. If insufficient carbide-forming alloying
elements (e.g., Cr, Mo, Nb) are present or the microstructure is in the wrong condition from improper heat
treatment, the hydrogen that has entered the steel reacts with the carbon in the steel to form methane. Since
the methane molecules are too large to diffuse through the steel lattice, they cause microfissures to form.
These fissures eventually combine to form cracks, as shown in Figure 3.3(a).* Figures 3.3(b)*and 3 . 3 ( ~ ) ~
show the microstructures of fissured and unattacked base metal, respectively. Note the absence of the dark
(pearlite) constituent in Figure 3.3(b) that is present in Figure 3.3(c). A significant amount of hydrogen
attack can occur before the fissures combine to form cracks and before any visible bulging or change in
thickness occurs. Thus, this is a very insidious form of attack.
Carbon steel can be used until the temperature exceeds an alloy’s limit in API Publication 941.3 Figure
3.4 (opposite page) shows the temperature vs hydrogen partial pressure limits in API 941 for carbon and
alloy steel^.^ Solid lines indicate where internal decarburization and fissuring will occur. Dashed lines indi-
cate where the less harmful surface decarburization will occur.
Hydrogen partial pressure, megapascals absolute
1,500
800
1,400
1,300
700
1,200
r 1,100
6 0 0.-
8
c
E-
-
U
0
0
fY 1,000
ff : - f
0
500 8h
f
5 800
u
P
z
400 z
700
600
300
500
400 200
300
0 500 1,000 1,500 2,000 2,500 3.0% 5.W 1.W 9.W 11.000 t3,WO
I
Sc.lachanps I
Hydrogenpartial pressure, pounds per square inch absolute
Copyright0 1967 G A. Nelson Reprodualon rights granted aby author 10 API This Figure was revised by API m 1969. 1977. 1983. and 1990
Legend
Surfacedecarburiratlon
Internal decarburization
(Hydrogen attack)
-
Notes:
1. The limits described by these curves are based on service experience orginally collected by G.A. Nelson and on addltlonal
information gathered by or made available to API. Carbon 1.0 Cr 2.0 Cr 2.25 Cr 3.0 Cr 6.0 Cr
2. Austentitic stainless steels are generally not decarburized in hydrogen at any temperature or hydrogen pressure. steel 0.5 Mo 0.5 Mo 1.00 Yo 0.5 Mo 0.5 Mo
3. The limits described by these curves are based on experience with cast steel as well as annealed and normalized steels. Satisfactory o o n o o v
Hydrogenattack 0 A 4
SurfaceDecarburlzation. R 8 (8, I
See comments a m a m d o
-
Figure 3.4 Operating limits for steels in hydrogen service to avoid decarburization and fi~suring.~
(Reprinted with permission of the American Petroleum Institute)
API 941 was derived from laboratory work done by Naumann in Germany in the early 1940s.’ Since
then, it has been revised periodically. based on operating experience and test data supplied by various petro-
leum refining companies. Although the curves originally contained a safety factor on the data, the numerous
revisions based on new data have led some companies to apply safety margins, e.g.. 30°C (50°F) to the
curves. Because API 941 is periodically revised, it is important to verify that the latest revision is used.
Because experience with C-1/2 Mo steels (particularly in catalytic reformers) has been poor, API 941 has
removed C-1/2 Mo from the design curve and produced a separate “experience” curve for C-1/2 Mo steels
(Figure 3.5).3 API 941 also contains information about three failures of 1-1/4 Cr-1/2 Mo that occurred in
zones below the published limits for this material.
Because of the numerous failures of C-1/2 Mo material, most refiners specify a minimum of 1 Cr-1/2
Mo when API 941 indicates that the limits for carbon steel have been exceeded. Research on C-1/2 Mo
materials has indicated that some failures can be attributed to the lack of (or improper) heat treatment after
forming or welding. Heat treatment should be a minimum of 650°C ( 1,200”F)to prevent hydrogen attack in
C- 1/2 Mo steels. Hattori has shown that the onset of hydrogen attack in C- 1/2 Mo can be predicted provided
one knows the hydrogen partial pressure, the temperature, and the carbide morphology.’
When high-pressure hydrogen is to be in contact with metals for only a short period, carbon steel may be
used at higher temperatures tean indicated by Figure 3.4.’ Figure 3.6 gives the permissible times carbon
steel may be used as a function of temperature and hydrogen partial pressure.3High temperatures can be
tolerated for short periods because hydrogen attack has an incubation period. Because hydrogen attack is
cumulative, the total anticipated time at temperature for the life of the equipment must be used as a basis for
iI
i!
References
A. AMOCO Oil Company, private communication to APl Subcommittee on Corrosion, 1960.
8. A.R. Ciuffreda and W.D. Rowland, “Hydrogen Attack of Steel in Reformer Service,” Proceedings, 1957, Vol. 37, American Petroleum Institute,
New York,pp. 116-128.
C. C.A. Zapffe, “Boiler Emhrittlement,” Transoflions ofthe ASME, 1944, Vol. 66, pp. 81-126.
D. R.E. Allen, RJ. Jansen, P.C. Rasenthal, and F.H. Vitovec, ‘The Rate of Irreversible Hydrogen Attack of Steel at Elevated Temperatures,”
Proceedings, 1%1, Vol. 41, American Petroleum Institutute, New York, pp. 74-84.
E. L.C. Weiner, ‘‘Kinetics and Mechanism of Hydrogen Attach of Steel,” Corrosion, 1%1, Vol. 17, pp. 109-115.
F. JJ. Hur, J.K. Deichler, and G.R. WorreU, “Building a Catalytic Reformer,” Oil and Gas Journal, October 29,1956, No. 78, pp. 103-107.
G. F.K. Naumann, “Influence of Alloy Additions to Steel Upon Resistance to Hydrogen Under High Pressure.” Technkche Minc*tngen Knrpp, 1938,
VOI. 1, NO 12, pp. 223-234.
H. M. Hasegawa and S. Fujinaga, “Attack of Hydrogen on Oil Refinery Steels,’’ Tersu To Hagane, 1%0, Vol. 46, No. 10, pp. 1349-1352.
1. T.C. Evans, “Hydrogen Attack on Carbon Sleek,” Mechanical Engineering, 1948, Vol. 70, pp. 414-416.
J. Air Products, lnc., private communication to API Subcommittee on Corrosion, March 1960.
K. API Refinery Corrosion Committee Survey, 1957.
L. 1. Class, “Present State of Knowledge in Respect to the Properties of Steels Resistant to Hydrogen Under Pressure,’’ Stahl and Eken, August 18,
1960, Vol. 80, pp. 1117-1135.
Figure 3.6 -Time for incipient attack of carbon steel in hydrogen service?
(Reprinted with permission of the American Petroleum Institute)
2.3 Sulfidation by Hydrogen-Hydrogen Sulfide Mixtures
When the temperature exceeds 288°C (550°F) in hydrogen-hydrogen sulfide mixtures, severe corrosion
occurs on carbon and low-alloy steels. Corrosion rates of various materials as a function of mole o/o hydro-
gen sulfide (based on only the constituents in the vapor phase) vs temperature for both gas oil and naphtha
streams are shown in Figures 3.7(a) through 3.7(i).6Although desulfurization takes place in the first stage of
hydrocrackers, the feed to many second stage hydrocrackers still contains sufficient sulfur compounds to
require that materials be identical to those in the first stage hydrocrackers. Before any hydrogen enters the
process stream, 5 Cr-1/2 Mo and 12 Cr clad steels can be used above 288°C (550°F) provided the feed does
not have a lot of free H,S. When hydrogen is present, 18 Cr-8 Ni SS usually is selected; stabilized grades
commonly are used to prevent intergranular attack by polythionic acid during downtime. See Section 2.5 for
the discussion of attack by polythionic acid.
The 12 Cr alloy occasionally is used up to 345°C (650°F) when the hydrogen sulfide concentration is
less than 1 mol%. Aluminized (diffusion-coated aluminum) carbon and low-alloy steel also display excel-
lent resistance to hydrogen sulfide in these environments; however, attack can occur at breaks in the coat-
ings. Aluminizing also is costly and is subject to damage during mechanical or acid cleaning. Aluminizing
is, therefore, usually selected primarily to reduce fouling from scale formation rather than to protect from
corrosion.
-
Figure 3.7 lsocorrosion
curves for carbon, low alloy,
and stainless steels as a
function of mole percent
H,S and temperature!
(Reprinted from “Computer
Correlations to Estimate High
Temperature Hydrogen Sulfide
Corrosion in Refinery Streams,”
Materials Protection, January
1971)
-
Figure 3.7(a) Carbon steel,
naphtha diluent.g
10
1.0
(I)
cy
z
0.1
X
w
4
0
I
0.01
0.001
400 600 800 1,000
TEMPERATURE. *F
-
Figure 3.7(b) Carbon steel, gas oil diluent.6
t A 10
OIL OATA.
LETTERS ylow COMPANY
0
400 600 800 1000
TEMPERATURE, O F
-
Figure 3.7(c) 5% Cr steel, naphtha diluent.6
1
Lines represent American
oil company data.
Letters show company
supplying data.
Numbers are reported
mils per year.
ATYPICAL DATA
UJ
(Y
f
Yr 0.1
l - ~
w
a
i
0
L
0.01
No Corrosion
-
Figure 3.7(d) 5% Cr steel, gas oil diluent.6
10
1.0
*
w
0.1
0 -01
0.001
400 800 aoo 1000
TEMPERATUUE. 'F
-
Figure 3.7(e) 9% Cr steel, naphtha diluent.6
QD
(Y
x
*
w
4
0
E L
-
-
0.01 - --
E- --
I
LINES REPRESENT AMERICAN
OIL COMPANY DATA'.
--
- NUMBERS ARE REPORTED
WLS PER YEAR.
LETTERS SHOnr COMPANY N O CORAOSt0IY
-
.- SUrrCVlNG DATA -
- -
0.001 I
400 600 800 9000
T E M P E R A T U R E , 'F
-
Figure 3.7(f) 9% Cr steel, naphtha diluent.
-
Figure 3.7(9) Cr stainless steel, naphtha diluent.6
1.0
(I)
(Y
S
*
w
0.1
A
0
E
0.01
0.001 L
400 600 800 1000
TEMPERATURE, "F
-
Figure 3.7(h) 12% Cr stainless steel, gas oil diluent.
-
Figure 3.7(i) 18-8 stainless steel, naphtha diluent.6
Next Page
-
Figure 3.70) 18-8 stainless steel, gas oil diluent.6
4
Hydrogen, Methanol, Ammonia,
GasTreating,Hydrodealkylation,
Polymerization, Phenol, Solvent Treating
1. INTRODUCTION
Hydrogen, methanol, and ammonia plants are very similar. Methane or naphtha feed stock is first des-
ulfurized then combined with steam in a reformer furnace. Hydrogen and carbon dioxide are produced in the
reformer furnace at about 820°C (1,500”F)as the starting point for all three processes.
In a hydrogen plant (Figure 4.1),’ the process gas (hydrogen and carbon dioxide) from the reformer
furnace is cooled to about 450°C (850°F) in a quench steam generator, then cooled further to about 370°C
(700°F) and sent to a shift converter, where additional hydrogen is formed. The process gas is again cooled
then fed to a pressure swing adsorption (PSA)TMunit, a hot potassium carbonate absorption system, a
monoethanolamine (MEA) absorption system or a SulfinolTMunit to purify the hydrogen by removing the
carbon dioxide.
-
Figure 4.1 Hydrogen’
(Reprinted with permission from “Refinery Process Handbook,”Hydrocarbon Processing,
Houston,TX: Gulf Publishing, September 1968)
91
-
Figure 4.2 Ammonia.’
(Reprinted with permission from “Refinery Process Handbook,” Hydrocarbon Processing,
Houston,TX: Gulf Publishing, September 1968)
In an ammonia plant (Figure 4.2),’ the synthesis gas from the reformer furnace is fed into a secondary
reformer vessel in which air is added through a burner to create outlet vessel temperatures in the order of
980°C (1,800”F).The outlet of the secondary reformer vessel is cooled in a quench steam generator and sent
to a shift converter; this is followed by a carbon dioxide removal system. The purified nitrogen from the air
added in the secondary reformer vessel and hydrogen synthesis gas is fed to a methanator to convert residual
oxides of carbon back to methane (which is inert in the ammonia conversion); the gas is then compressed to
about 20,700 kPaa (3,000psia). The compressed synthesis gas is fed to an ammonia converter vessel. As the
synthesis gas passes over catalyst beds, ammonia is formed. The ammonia product is then cooled and refrig-
erated to separate out impurities.
In a methanol plant (Figure 4.3),’the synthesis gas goes from the reformer furnace to a heat recovery
section, where it is cooled to room temperature. The synthesis gas is then compressed from 5,170 kPaa to
10,345 kPaa (750 psia to 1,500psia) and fed to the converter vessel through preheat exchangers. Methanol
is formed as the gas passes over catalyst beds in the converter vessel at 205°C to 315°C (400°F to 600°F).
The methanol product is then cooled and fed to separators, then to fractionators to complete the purification.
2. MATERIALS OF CONSTRUCTION
The front end section of hydrogen, methanol, and ammonia plants is shown in Figure 4.4.*The second-
ary reformer is used only in an ammonia plant. The feed gas is desulfurized in carbon steel equipment. When
Reforming Synthesis Distillation
Converter
Separator
flash Vessel %!dT Refining
Column
Steam
I I I I
I 1 Pu eto
Fu2GasSystem
DEISUlfUrbed
Naphtha
~ ~ s ~ ~ !Fuel
t
l Gas
b aor Naphtha
n
-
Figure 4.3 Methanol.' (Reprinted with permission from "Refinery Process Handbook,"
Hydrocarbon Processing, Houston,TX: Gulf Publishing, 1968)
QUENCH STEAM
DESULfllRmR DRUM REFORMER FURNACE GENERATOR SHlFT CONVERTER
r E;&-&&- -1
I REFORMER I
I I
FEEDOAS
-
Figure 4.4 High temperature front end section of reforming plant.*
the metal temperature exceeds 425°C to 455°C (800°F to 850"F), 1 Cr-1/2 Mo or 1-1/4 Cr-1/2 Mo is used to
avoid long-term deterioration of the mechanical properties by graphitization. Preheat coils in the top of the
reformer furnace usually are 2-1/4 Cr - 1 Mo up to 650°C (1,200"F) metal temperature and 304H (UNS
S304009) for metal temperatures above 650°C ( 1,200"F). Caustic stress corrosion cracking from solids can
occur in the steam preheat coils if solid carryover is excessive (see Chapter 1, Section 3.7). The inlet connec-
tions to the steam methane reformer furnace tubes are either 1-1/4 Cr-1/2 Mo (595°C [ 1,10O"F] maximum)
or 2-1/4 Cr-1 Mo (650°C [ 1,200"FI maximum).
The methane (or naphtha) and steam are converted to hydrogen and carbon monoxide along with some
carbon dioxide over a nickel catalyst in the HK-40 (UNS 594204) or HP modified, also called CE20N (UNS
592802) primary reformer furnace tubes. CE20N (UNS 592802) has largely replaced HK-40 because of
superior stress-to-rupture strength (e.g.. 12.55 MPa [ 1.82 ksi] vs 8.3 MPa [ 1.2 ksi] at 980°C [ 1,800"FI). Skin
temperatures on these tubes are about 980°C ( 1,80O0F),and the outlet process temperature is about 820°C
(1,500"F). Sulfur content in the fuel gas is limited to 2,000 ppm to 3,000 ppm hydrogen sulfide to avoid
accelerated oxidation of the outside of the tubes. The tubes are centrifugally cast. They have been used in the
as-cast condition which includes about 2.4 mm (3/32 in.) dross and unsoundness on the inside diameter.
Currently, most tubes are bored on the inside to remove the dross and unsoundness and machined on the
outside. Since the tubes are operating in the range where sigma phase (a brittle Fe-Cr compound) forms, the
Cr, Ni, and C are "balanced' to minimize sigma phase embrittlement. The welds must be blasted to remove
all residual weld slag; otherwise, the residual weld slag can form a eutectic with the metal oxides, which
results in catastrophic oxidation.
The outlets of the primary reformer furnace tubes are connected to either a refractory-lined steel or
(occasionally) an alloy 800H (UNS N08810) outlet header with alloy 800H "pigtails." Pigtails are tubes
(about 25.4 mm [l in.] in diameter) connected to a reducing cone or a side boss at the bottom of the centrifu-
gally cast tube. They are called pigtails because they were originally made in a double loop configuration to
compensate for thermal expansion. More advanced designs have eliminated the need for the loops. Prema-
ture failure of alloy 800 (UNS N08800) pigtails has occurred because of too fine a grain size (smaller in size
than ASTM No. 5); however, these problems can be avoided by specifying alloy 800H. Some refiners prefer
single-row reformer tubes to minimize the thermal stresses on the pigtails.
For temperatures over 650°C ( 1,20O"F), 65 Ni-15Cr-Fe filler metal, such as INC082TM(UNS N06082)
or INCO ATM(UNS W86133), should be used (although INCO A has a somewhat lower creep strength than
INCO 82). Neither INCO 92TM(UNS N07092) nor INCO 182TM(UNS W86182) should be used above
480°C to 5 10°C (900°F to 950"F), because they embrittle when exposed to high temperatures. In addition,
INCO 182 has a significantly lower creep strength than either INCO 82 or INCO A. Weld filler metal from
other sources should be examined very carefully because some filler metals are subject to "green rot" (pref-
erential oxidation of chromium that occurs about 730°C [ 1,35O"F],resulting in rapid deterioration). For the
same reason, alloy 600 (UNS N06600), either wrought or cast, should not be used above 730°C (1,350"F) in
this service.
A transfer line connects the primary reformer to the quench steam generator in a hydrogen plant and to
the secondary reformer in an ammonia plant. The secondary reformer in an ammonia plant is connected to
the quench steam generator by another transfer line. Transfer lines normally operate at 788°C to 980°C
(1,450"F to 1,800"F) and usually are made of either alloy 800H (UNS NO88 10)or refractory-lined carbon
steel. Above about 820°C ( 1,50O0F),the combination of low strength and high thermal expansion of metals
makes refractory linings attractive; however, refractory linings can develop hot spots from cracks and some-
times can deteriorate due to condensation of corrosive gases at the metal wall.
The secondary reformer in an ammonia plant is a carbon steel vessel with a dual-layer refractory lining.
Internal temperatures reach about 1,090"C (i,OOO"F) from burning as a result of air added through a burner
at the top of the vessel to the feed gas (hydrogen, carbon monoxide, carbon dioxide, and steam). The burner
is a refractory-lined device subject to failure if not carefully designed. Quench steam generators have refrac-
tory-lined inlet channels and tube sheets. Tubes often are made of carbon steel because the heat transfer from
the steam on the outside of the tube is markedly better than that from the synthesis gas inside the tube. As a
result, the metal temperature closely approaches that of the steam. The inlet ends of the tubes are protected
from the inlet gas by ferrules, usually 310 SS (UNS S31000) with insulation between the ferrule and the
tube. The tube material should be selected according to the maximum anticipated metal temperature and to
API 941. The outlet channels usually are made of low-alloy steel selected by using API 941.
After the synthesis gas leaves the quench steam generator, it goes through a shift converter to convert
more of the synthesis gas to hydrogen and the carbon monoxide tocarbon dioxide. Some ammonia is formed
in the shift converter when nitrogen is present. Alloy selection is based on API 941 until the synthesis gas is
cooled below the dewpoint (usually about 160°C [325"F]). When wet carbon dioxide condenses out of the
synthesis gas, severe corrosion of carbon and low alloy steel results, particularly in turbulent areas. Type
c
5.5
5.0
4.5
4.0
E
3.5
0.7 1.4 2.1 2.8 3.5 4.2 5.6 7.0 14.1 21.1
3.0
1 10' i!O 3b 4b Sb 6'0 $0 Id0 260 3dO
-
Figure 4.5 Relationship between pH, carbon dioxide partial pressure,and estimated
corrosion rate on carbon steel at 38°C (100°F).
304L S S (UNS S30403) normally is used to resist this attack. Figure 4.5 shows the relationship between pH
and carbon dioxide partial pressure.* The higher corrosion rates shown are calculated from the deWaard-
Lotz equation, which was developed from laboratory data. The lower corrosion rates represent typical values
experienced in the field after a corrosion film has formed and steady state conditions exist.
As can be seen in Figure 4.5, the corrosion rate of carbon steel at 38°C (100°F) increases as the carbon
dioxide partial pressure increases.2For temperatures below 60°C to 70°C (140°F to 160"F), the corrosion
rate in wet carbon dioxide is activation-controlled and, therefore, not accelerated by turbulence. Above the
60°C to 70°C (140°F to 160°F) range, the corrosion rate is diffusion-controlled and, therefore, is signifi-
cantly accelerated by turbulence. The following rates, measured in a two-phase, gas-water system contain-
ing carbon dioxide at a partial pressure of 125 kPaa (18 psia), illustrate the effect of turbulence and the effect
of alloying elements in reducing corrosion caused by turbulence:
lMTrademark
-
STRIPPER CO, STRIPPER OVERHEAD OVERHEAD
REBOILER CONDENSER ACCUMULATOR
CO, TO ATMS
4
(I1 b
FROM HIGH
END SECTlO
METHANATOR
TO COMPRESSOR
00,ABSORBER
-
Figure 4.6 Carbon dioxide removal section of reforming plant.2
require stress relief for all equipment exposed to amine solutions regardless of process temperature. An
amine solution is defined by API 945 as concentrations of amine over 2 wt%.
A study by Foroulis of stress corrosion cracking (SCC) of carbon steel in potassium carbonate solutions
revealed that SSC does not occur in the absence of carbon dioxide but does occur in carbonatehicarbonate
mixtures produced in carbon dioxide absorption.6 The strong tendency to stress corrode occurs in the -0.85
to -0.55V(SHE) potential range. Use of potassium metavanadate in concentrations greater than 0.5 wt% (as
NaVO,) can prevent SCC. Conversely, arsenite inhibitor/activators promote SCC. Oxygen, usually added to
maintain the inhibitor in the active (oxidized) state, will minimize the tendency for SCC. Even though proper
inhibition theoretically prevents SSC, most operators require stress relief for process temperatures above
60°C (140°F). This is because SSC can occur rapidly if loss of passivity occurs due to loss of inhibitor or
overheating.
Typical acid gas loading for amines are as follows:3
Threaded connections should be avoided in acid gas amine solution because the turbulence created by
the threaded area causes severe erosion-corrosion of the threads. Where turbulent areas cannot be avoided
by design [e.g., pumps and control valves or equipment to be used where the velocity exceeds 2.4 m/s (8 ft/
s), 304L SS (UNS S30403) should be used. Reboiler tubes should be made of 304 (UNS S30400) or 316
(UNS S31600). If seal welding is required, 304L (UNS S30403) or 316L (UNS S31603) should be used.
Some operators limit the reboiler steam inlet temperature to 150°C (300°F).
Overheating of carbonates and amines in reboilers has resulted in violent gas evolution from the solu-
tion on the shell side leading to severe erosion-corrosion, particularly of carbon steel tube sheets and in the
reboiler discharge (vapor) line. Although austenitic stainless steel tubes also have been used in reclaimers.
there is at least one reported case of chloride SCC of these tubes. Regenerator overhead condensers often are
carbon steel. In older units there often is sufficient amine carryover (at least 0.5%) so that corrosion in the
overhead is inhibited. In newer units carryover is minimized and corrosion, when it occurs, is handled by
addition of corrosion inhibitors or use of stainless steel tubes.
The PSATMunit operates at 10°C to 38°C (50 "F to 100°F) and undergoes 14-minute pressure cycles as
part of the normal operating cycle. The only material problem presently known is fatigue cracking, which is
accelerated by the presence of hydrogen; this occurs as a result of the cycling of the vessels. Therefore, stress
raisers should be avoided in the equipment design.
After the hydrogen is purified, it is ready for use in a refinery hydrogenation process. In an ammonia
plant, the hydrogen-nitrogen mix is sent to an ammonia converter (Figure 4.7),* which requires a startup
heater. Since the material in the heater will be exposed to hydrogen only for a short period, the time-depen-
dent curves in API 941 should be consulted when selecting an alloy for the heater tubes. As mentioned in
Chapter 3, section 2.2, because hydrogen attack is cumulative, the total time the material will be at tempera-
ture must be used as a basis for alloy selection.
The ammonia reaction takes place in an internal 304 SS (UNS S30400) basket in which the temperature
is about 480°C (900°F). Even though the pressure is about 20.7 MPag (3,000 psig), the converter wall
(usually multilayer) does not often require alloy materials to resist hydrogen because cool gas is circulated
on the outside of the basket. Conversely, when the converter wall is solid alloy or the inner layer of a
multilayer vessel is exposed to high temperature, alloy materials are required either for strength, to resist
hydrogen attack, or both. The outer layers of a multilayer vessel are vented to the atmosphere so hydrogen
attack is not a problem. The outlet connection usually is hot enough to require chrome steel alloys to resist
hydrogen attack. Nitriding also should be considered above 400°C (750°F). Using a nitriding allowance
(usually 1.5 mm to 3 mm [ 1/16 in. to 1/8 in.]) is all that is usually required; however, alloy 600 (UNS
N06600), which is resistant to nitriding, is used for basket screens and occasionally for overlaying very high-
temperature parts.
After the ammonia leaves the converter, it is cooled and purified. Although stress corrosion cracking has
occurred in liquid anhydrous ammonia, it has not been a problem in the process plant because no oxygen is
introduced until the ammonia gets into the storage equipment. Ammonia is stored at -33°C (-28°F). It used to
be thought that this temperature was too low to cause SCC; however, cracking has been observed recently in
ammonia storage vessels. Therefore, stress relief of these vessels is being specified. To avoid SCC of carbon
steel equipment used for shipment and subsequent storage, at least 0.2 wt% water is required.
Methanol plants are very similar to ammonia plants; high-temperature reformer furnaces and high-pres-
sure (multilayer) converters are used. High corrosion rates on carbon steel occur in the 260°C (500°F) range
in the carbon monoxide, carbon dioxide, and hydrogen environments found in methanol plants; therefore,
corrosion-resistant alloys are required in this range. Figure 4.8 (page 100) shows corrosion rates of 11 alloys
as a function of temperature in the 50-50 carbon dioxide-hydrogen mixtures common in methanol plants.' In
addition, metal dusting has been reported in the process boiler of a methanol plant. See the following section
for a discussion of metal dusting.
TM Trademark
COMPRESSOR STARTUP AMMONIA ~~~~ LEloOwN
HYDROGEN HEATER CONVERTER COOLER MUM
PIANT
PROWCT
4
I
I IT
(L
1 I I TO RECYCLE
COMPRESSOR
REMOVAL- A
SECTION
STORMjE
-
Figure 4.7 High pressure section of reforming plant.2
The feed stock in hydrodealkylation units is heated to about 650°C (1,200"F) in a preheat furnace prior
to entering the reactor. Above about 590°C (1,100"F) a phenomenon known as metal dusting or catastrophic
carburization occurs on all alloys that are otherwise suitable for the temperature conditions. The attack is
very rapid and takes the form of round bottom pits. The surface of the remaining metal is heavily carburized.
A small quantity of sulfur (0.05 wt% to 0.5 wt%) in the form of hydrogen sulfide or mercaptan added to the
feed will prevent attack. Aluminizing also has been used to prevent attack.
Polymerization units use phosphoric acid as a catalyst in the reactor. Because solid phosphoric acid
catalysts do not cause corrosion, carbon steel can be used. However, liquid phosphoric acid is very corrosive
to carbon steel so corrosion-resistant materials are required where liquid phosphoric acid exists. At a phos-
phoric acid concentration of 100%, 304L SS (UNS S30403) is satisfactory to 80°C (120"F), and 316L SS
(UNS S31603) is required from 50°C to 107°C (120°F to 225°F).
Phenol is produced by the oxidation of cumene and is followed by cleavage of the oxidation product of
phenol and acetone. Solid or clad 304L SS (UNS S30403) is required to resist corrosion in the oxidation
vessel, and solid or clad alloy 20 (UNS N08020) often is used in the cleavage vessel.
Two common solvent treating processes are solvent deasphalting and solvent treating. In solvent
deasphalting, propane, butane, or a mixture of the two, is used to dissolve all hydrocarbons but asphalt. In
general, no corrosion occurs from the process side in these units. Conversely, in solvent treating with fur-
fural, some corrosion occurs when the furfural is mixed with water and in the portions where furfural is
heated above 215°C (420°F). Where these conditions exist, 304L SS (UNS S30403) is used below 70°C
(160"F), and 316L SS (UNS S31603) is used above 70°C (160°F). Where there is a potential for chloride
SCC of SS, red brass (UNS C2300), or 70 Cu-30 Ni (UNS C71500), is used at temperatures up to about
93°C (200"F), and alloy 400 (UNS N04400) is used above, Methyl tertiary butyl ether (MTBE), which is
used as an octane enhancer in gasoline, is not corrosive to carbon steel, SS, or aluminum. Conversely, none
of the common elastomers, i.e., Viton,T'nitrile, etc., are suitable for MTBE. Some refiners have had success
with Kalrez,TMand other proprietary materials. Specially formulated urethanes have been used for vapor
seals in storage tanks where PTFE has failed. VitonTMhasworked in blends of up to 15% MTBE.
Testing temperature"C Testing temperature"C
660 3600
600 2600
k 540 5. 1600
5
400
a 600
-a $ 600
-
*'
d
2 360
420 i 520
c
g
440
-s
f 300 360
&
6 240 u 280
100 2 00
.
a
I20
120 .
a
120
I20
t I00
0
g I00
80 f 80
z
i 60 6 60
e L
.- s 'O h
'r,
40
E& 20 2
8
20
0 0 0 0
3I
Testing temperature, F Testing temperature, F
Metal C Mn si Cr Ni other
1 ............ 0.09 ... ... ... ... 0.016 P
-
Figure 4.8 Effect of temperature and pressure on corrosion rate of several steels by 50-50
CO-H,. Steels were manufactured in Germany and tested in pilot plants at the I.G.
FarbenindustrieAmmonia Werke, Mer~eburg.~ (Reprinted with permission from
Handbook, 8th ed., Vol. 1, Metals Park, OH: ASM)
hfe&??/S
REFERENCES
“API Survey Shows Few Amine Corrosion Problems,” Petroleum Refiner, 46, 11 ( 1968): p. 28 I .
API Recommended Practice 95 1, “Guidelines for Avoiding Corrosion and Cracking Problems in Amine
Units” (Washington, DC: American Petroleum Institute).
Atkins, K.T.G., D. Fyfe, and J.D. Rankin, “Corrosion in Carbon Dioxide Removal Plant Towers,”AIChE
Safety in Air and Ammonia Plants, 16 (Sept. 1974): p. 39.
Avery, R.E., and H.L. Valentine, “High Temperature Piping Systems for Petrochemical Processing (INCO),”
Chemical Engineering Progress, 64, 1 (968): p. 89.
Bagdasarian, A.J., et al., “Stress Corrosion Cracking of Carbon Steel in DEA and ‘ADIP’ Solutions,”
Materials Performance 30,5 (199 1): p. 63.
Banks, W.G., “Corrosion in Hot Carbonate Systems,” Materials Protection 6, 1 1 (1967): p. 37.
Baumert, K.L., et al., “Hydrogen Attack of Carbon - 0.5 Molybdenum Piping in Ammonia Synthesis,”
Materials Performance 25,7 (1 986): p. 34.
Bienstock, D., and J.H. Field, “Corrosion Inhibitors for Hot-Carbonate Systems,” Corrosion 17, 12 (1961):
p. 87.
Bienstock, D., and J.H. Field, “Corrosion of Steels in Boiling Potassium Carbonate Saturated with Carbon
Dioxide and Hydrogen Sulfide,” Corrosion 17, 7 ( 1961 ): p. 89.
Blanc, C., et al., “Amine-Degradation Products Play No Part in Corrosion of Gas-SweeteningPlants,” Oil
and Gas Journal 80,46 (Nov. 15, 1982): p. 128.
Cherrington, D.C., and A.R. Ciuffreda, “HOWto Design Piping Systems for Hydrogen Service,” Oil and
Gas Journal 65,21 (May 22, 1967): p. 102.
Ciuffreda, A.R., and B.N. Greene, “Hydrogen Plant Shutdowns Reduced,” Hydrocarbon Processing 5 1,
5 (1972): p. 113.
Ciuffreda, A.R., and B.E. Hopkinson, “Survey of Materials and Corrosion Experience in Reformer Hydro-
gen Plants,” API Division of Refining 52,3 (1972): p. 549.
Dawson, J.J., “Behavior of Welded Pressure Vessels in Agricultural Ammonia Service, WeldingJournal
35,6 (1956): p. 568.
Deegan, D.C., and B.E. Wilde, “Stress Corrosion Cracking Behavior of ASTM A517 Grade F Steel in
Liquid Ammonia Environments,” Corrosion 29,8 (1973): p. 310.
deWard, C., U. Lotz, and D.E. Milliams, “Predictive Model for CO, Corrosion Engineering in Wet Natural
Gas Pipelines,” Corrosion 47, 12 (1991): p. 976.
Dingman, J.C., et al., “Minimize Corrosion in MEA Units,” Hydrocarbon Processing 45,9 (1966): p. 285.
Dokuchaev, V.A., and Y.I. Archakov, “Study of the Stress Corrosion Cracking Tendency of Steel St. 3 in
Hydrogen Sulfide-Saturated Monoethanolamine,” Kontrol, Remont Zashch. Rorroz. Neftezavod.
Oborud. (1982): p. 60.
Dunn, C.L. et al., “First Plant Data from Sulfinol Process,” Hydrocarbon Processing 44,4 (1965): p. 137.
Edeleanu, C., ed., Materials Technology in Steam Reforming Processes (London, UK: Pergamon Press,
1966).
Foroulis, C.A., “Stress Corrosion Cracking of Carbon Steel in Hot Potassium CarbonateA3icarbonate
Solutions,” Boshoku Gijursu 36, 11 (1987): p. 689.
Gutzeit, J., and J.M. Johnson, “Stress Corrosion Cracking of Carbon Steel Welds in Amine Service,”
Materials Performance 25,7 (1986): p. 18.
Hall, A.M., et al., “Some Properties of HF, HH, HK and HN Alloys,” ASME Publication 68-PVP-6 (1968).
Ihrig, H.R., “Attack of Hydrogen-Nitrogen Mixtures on Steels,” Znd. & Eng. Chem. 41, 11(1949):p. 2 3 16.
Kobrin, G., and E.S. Kopecki, “Choosing Alloys for Ammonia Services,” Chemical Engineering 85,
28 (Dec. 18, 1978): p. 115.
Lang, F.S.,. and J.F. Mason Jr., “Corrosion in Amine Gas Treating Solutions,” Corrosion 14,2 (1958):
p. 10%.
Loginow, A.W., and E.H. Phelps, “Stress Corrosion Cracking of Steels in Agricultural Ammonia,” Corro-
sion 18, 8 (1962): p. 229t.
Market, J.W., “Metal Dusting in the Process Boiler of a Methanol Plant,” American Institute of Chemical
Engineers Annual Meeting, paper no. 94b (November 25-30, 1984).
Moltley, J.R., and D.R. Fincher, “Inhibition of Monoethanolamine Solutions,” Materials Protection 2, 10
(Aug. 1963): p. 27.
Moore, K.L., “Corrosion of Furnace Tubes by Residual Welding Slag,” Corrosion 16, 1 (1960): p. 26t.
Moore, K.L., “Corrosion Problems in a Refinery Diethanolamine System,” Corrosion 16, 10 (1960): p. 111
.Moran, J.J., et al., “Behavior of Stainless Steels and Other Engineering Alloys in Hot Ammonia Atmos-
pheres, Corrosion 17,4 (1961): p. 115 (191t).
Parkins, R.N., and Z.A. Foroulis, “Stress Corrosion Cracking of Mild Steel in Monoethanolamine
Solutions,” Materials Performance 26, 1 (1988): p. 19.
Pease, G.R., “Corrosion of Nickel-Chromium,Iron Alloys by Welding Slags,” WeldingJournal 35,9 (1956):
p. 469-S.
Polderman, L.D., et al., “Degradation of Monoethanolamine in Natural Gas Treating Service, Oil and
”
1. INTRODUCTION
Refinery and production piping usually are designed to ASME B3 1.3. Atmospheric storage tankage is
designed to API 650. Low-pressure tankage is designed to API 620. The design of pipelines containing
liquids is covered by ASME B3 1.4.Guidelines for gas transmission pipeline design are contained in ASME
B31.8;however, because of public concern over the large amount of stored energy in gas transmission lines,
the U.S. Department of Transportation (DOT) has issued regulations for gas lines (Code of Federal Regula-
tions, Title 49,Part 192). There also are regulations governing liquid lines (Code of Federal Regulations,
Title 49,Part 195).
For refineries and production equipment, pipe is purchased to an ASTM or API specification, and very
rarely are supplementary requirements added. Supplementary requirements for pipelines almost always are
included because pipelines consist of miles of identical material; specifications can thus be tailored to the
needs of the job in dimensions, chemistry, and mechanical properties. When the pipeline is to be purchased
in large tonnages, supplementing the API specification usually entails little or no cost penalty when com-
petitive bidding is involved.
2. LINEPIPE
2.1 General
1. UOE (one pass of submerged arc welding on each side of a long seam),
2. Electric resistance welded long seam,
3. Seamless, or
4. Helical seam (most commonly made using one pass of submerged arc welding on each side).
To minimize damage during handling, minimum practical wall thickness (Table 5.1) has been estab-
lished by some organizations. Other organizations limit the diameter to wall thickness ratio 100 to 105.
105
Table 5.1
Minimum Practical Wall Thickness for Thin Wall Pipe
Minimum Wall
Diameter Thickness(’)
mm (inches) mm (inches)
58.8(2.315) 3.2/3.9(0.12510.154)
114 (4.5) 3.9 (0.154)
219 (8.625) 4.0 (0.156)(3)
273 (10.750) 4.8(0.188)
324 (12.750) 4.8(0.188)
356 (1 4) 5.2(0.203)
406 (1 6) 5.2(0.203)
457 (1 8) 5.6(0.219)(3)
508 (20) 6.4 (0.250)
610 (24) 6.4(0.250)
660 (26) 6.4(0.250)
762 (30) 6.4(0.250)
813 (32) 7.1 (0.281)
914 (36) 7.9(0.312)
965 (38) 7.9(0.312)
1,016(40) 8.7(0.344)
1,067(42) 8.7(0.344)
1,219(48) 9.5(0.375)
The UOE process for double-submerged arc welded (DSAW) line pipe is schematically shown in Figure
5.1 .I Plate is prepared then formed into a “U.” It is then pressed round into an “0,” submerged arc welded
and expanded, which is where the “E’ comes from. Figure 5.2 shows a typical cross section of a double-
submerged arc weld.’ A minimum of weld passes (in this case, two) is used to maximize production. UOE
pipe is available in 406 mm to 1,625 mm (16 in. to 64 in.) diameters. It is the most commonly used pipe,
particularly in large diameters, because of its high reliability when the long seam is properly inspected. The
cost varies because the larger the diameter, the more cost-competitive the pipe. Hydrostatic test records
reveal that only about one failure occurs every 800 km (500 miles) of test.
The process for manufacturing electric resistance welded (ERW) pipe is shown schematically in Figure
5.3.’ In this process, coils rather than plate are formed, then continuously electric resistance welded. Figure
5.4’ shows a typical cross section of an ERW weld. In contrast to the submerged arc weld, the ERW weld is
very narrow. The dark area between the weld and the remainder of the pipe is heat-affected zone; only the
light area in the very center is the weld. ERW line pipe is available in 100 mm to 1,220 mm (4 in. to 48 in.)
diameters with a 14 mm (0.562 in.) maximum wall thickness. However, it is only cost-competitive in 152
“)The minimum the mills will offer for submerged arc-welded pipe is 6.4 mm (0.250 in.)
( a ) Not practical to electric weld.
0) Limited by welding at and below this thickness.
General Description
UOE pipe is made from plate. After being subjected to aulnmatic ultrasonic inspection, each plate is trimmed and
beveled on the side edges and then transferred to a U-ing press where it is formed into a U-shape. It is subse-
quently formed into a round shell on an 0 - i n g press. The straight seam is submerged arc welded from both the
inside and outside. Mechanical expansion turns the pipe into an exactly round, straight section. Each length of
pipe is hydrostatically lested before final inspection. The defect-free quality of Kawasaki UOE pipe is assured by
both advanced inspeclion equipment, including full-length weld automatic ullrasonic tester (AUT), and qualified
inspection personnel with wide experience.
-
Figure 5.1 UOE pipe.'
(Reprinted with permission from Kawasaki Tubular Products, 1978, Kawasaki Steel Corporation, Kobe, Japan)
OUTSIDE
INSIDE
-
Figure 5.4 High frequency weld.2
(Reprinted with permission from Sixth Symposium on Line Pipe Research, October 1979.
Research sponsored by the Pipeline Research Committee, American Gas Association, Arlington,VA)
General Description
Electric resistance weld pipe is made from strip in coil form. After being uncoiled, flat strip is progressively rounded as it passes
through a series of vertical and horizontal forming rolls before welding. Small-diameter pipe is welded by the high frequency
induction method and medium-diameter pipe by the high frequency electric resistance method. Continuously welded pipe is cut
to the specified lengths by a flying cutoff machine, and each length goes through straightening, hydrostatic and inspection equipment
before it becomes a finished product with the desired diameter.
Hook cracks - Caused by nonmetallic inclusions at the edge of the material, which follow flow lines of
the material as the weld is pushed together during welding (Figure 5 3 . ’
Penetrators - Areas of no weld caused by: ( 1 ) nonmetallic materials on the edge to be welded that are
not squeezed out, (2) arcing of electrical contacts, (3) insufficient or excessive upset during welding, or
(4) short-circuiting of welding current due to burrs on welding surfaces (Figure 5.6).’
Stitch welds - Intermittent fusion caused by short repetitive variations in the welding heat (Figure
5.7).4
-
Figure 5.7 Stitch weld (intermittent
fusion) in ERW weld.4
(Reprinted with permission from API Bulletin 5TI,
Figure5.6 -Through wall penetratorsin ERW “NondestructiveTesting Terminology,”9th ed., May
weld.3 1987, American Petroleum Institute,Washington, DC)
111
The process for manufacturing seamless line pipe is shown schematically in Figure 5.8.' In this process,
a heated metal cylinder is pierced then formed into the proper diameter and wall thickness. Seamless pipe is
available up through 660 mm (26 in.) diameter, but normally is used only up to 405 mm (16 in.) diameter.
Although seamless line pipe is the most expensive (15% to 25% above ERW), it is the most reliable. Even
so, failures do occur. Hydrostatic test records reveal that about one failure occurs every 800 km (500 miles)
of test. Some mills produce seamless pipe that has a significantly higher failure rate; it is thus important to
ensure adequate quality control.
2.1.4 Helical Seam Line Pipe
The process for manufacturing helical seam line pipe (previously called spiral weld) is schematically
shown in Figure 5.9.’As in ERW pipe, steel strip in coil form is used, but the coil is formed into a spiral. The
seams are welded by the DSAW process as in UOE pipe. Some of the early prejudice against helical seam
pipe resulting from unravel failures that occurred many years ago has been overcome. It is available from
152 mm to 2,540 mm (6 in. to 100 in.) diameters, but the smallest diameter commonly used for line pipe is
610 mm (24 in.). Problems sometimes are encountered in bending and lining up. Strict adherence to dimen-
sional tolerances during manufacture can significantly reduce costs of line-up, clamping, and fabrication. As
a result, large quantities of helical seam pipe have been used.
Induction-welded long seam pipe (similar to ERW) normally is used for 50 mm to 100 mm (2 in. to 4
in.) pipe. Continuous (furnace) butt weld pipe is available in 20 mm to 115 mm (0.75 in. to 4.5 in.) pipe;
however, it is never used in critical applications because of its very low reliability.
API specification 5L is the principal line pipe specification. It combines the previous API 5L, 5LX, and
5LS. The selection of a grade depends on the strength-to-weight ratio and the pumping costs. ASTM speci-
fications (e.g., A38 1) often are used for compressor station pipe, particularly for extruded headers.
The low srrengrh grades in API 5L are A and B. Grade A has a 207 MPa (30 ksi) minimum yield
strength, and Grade B has a 241 MPa (35 ksi) minimum yield strength. The seamless and welded (ERW,
DSAW, GMA, or FBW) manufacturing processes are permitted.
The exrra strength ( X ) grades in API 5L range from X42 to X80 grades. The number after the “X’
refers to the specified minimum yield strength, i.e., X42 means line pipe with 290 MPa (42 ksi) mini-
mum yield strength. The seamless and welded (ERW, DSAW, and GMA) manufacturing processes are
permitted by API 5L. Normally, all grades are cold expanded to achieve dimensional uniformity and
increased strength from strain hardening. API X52, one of the most commonly used grades, is about the
maximum strength material that can be obtained by cold expansion without alloying additions, although
alloying additions also are permitted. In the past 10 to 12 years, 90% of the large-diameter line pipe has
been X60 to X70 grades. Recently, Coulson reported on the use of X80.5 Small amounts of grain-
refining alloying additions are used to obtain the desired strength and impact properties of these materi-
als. In general, a minimum preheat (depending on composition and conditions) is required for welding
the API X60 and higher-yield strength grades. Matching strength welding electrodes are available for
welding the round seams.
The helical seam grades in API 5L range from A to X70. Cold expansion normally is not performed;
however, some mills have installed and used expanders successfully.
As mentioned previously, specifications to supplement the API requirements normally are written to
tailor the line pipe to the specific operating conditions. Appendix D contains two examples of line pipe
specifications. Typical requirements are summarized below:
1 . Require either vacuum degassing or argon bubbling during steel making to enhance floating of inclu-
sions out of the molten steel. This both improves notch toughness and resistance to hydrogen-induced
cracking.
2. Limit cold expansion to 1.5% to avoid degradation of impact properties.
General Description
Spiral weld pipe is made from strip in coil form. After the strip is uncoiled, it proceeds to a
trimmer and then to forming rolls that shape it into spiral form. The spiral seam is
-
submerged arc welded automatically and continuously first on the bottom of the rotating
pipe from the inside, then on the top from the outside. The welded pipe is cut to the specified
lengths by a flying cutoff machine. X-ray and ultrasonic inspection ensures consistently high
quality spiral weld pipe.
-
Figure 5.9 Helical seam pipe.'
(Reprinted with permission from Kawasaka Tubular Products, 1978, Kawasaki Steel Corporation, Kobe, Japan)
3. Require the hydrostatic test to be at 90% to 100% of the specified minimum yield strength to maxi-
mize reliability. Mills often will pay for the pipe that fails in the field even when the pipe is tested at
100%of the specified minimum yield strength.
4. Require radiography of the end of long seams and of all jointer welds (when permitted) per API I 104.
5. Perform ultrasonic examination of long seams at the mill.
6. Require magnetic particle. liquid penetrant, or individual ultrasonic testing of 150 mm (6 in.) of the
inside (sometimes the end 150 mm [6 in.] of the outside as well) for X60 and stronger pipe.
7. Prohibit repair of parent pipe after cold expansion because the heat of welding reduces the strength
gained from the cold expansion.
8. Require any repairs to be made using preheat and low-hydrogen electrodes to avoid underbead crack-
ing. (Historically, a Vickers hardness of 350 has been considered the threshold for underbead cracking
when cellulosic electrodes are used.)
9. When repair is permitted, require radiography for through-wall repairs, magnetic particle examina-
tion for other repairs, and re-hydrostatic test for all repaired pipe.
10. Require radiography per API 1104 after forming for skelp welds in helical seam pipe.
11. Require fracture toughness testing based on the design conditions. (See Section 7 in this Chapter).
12. Require residual magnetism to be less than 30 gauss to avoid welding problems.
13. In designing, limit weldolets to 100 mm (4 in.) maximum to minimize stress concentrations and
avoid excessive heat input, which destroys the benefits of cold expansion.
Fittings, including extruded multiple outlet headers, usually are made from material similar to pipe in
chemical composition and heat treated to attain mechanical properties that match the pipe. Quenching and
tempering frequently are necessary to meet fracture toughness requirements, particularly in the 414 MPa (60
ksi) and higher yield strength grades. Typical specifications governing pipe fittings and valves are:
ANSI B16.5 All (also valves) Standard ASTM materials, dimensions, and
pressure temperature ratings
M SS- SP-44 Flanges Grades F36 through F70, dimensions, chemis-
try and mechanical properties, but no impacts
MSS-SP-75 All welded fittings Grades Y42 through Y70, full materials
except flanges specification
(includes headers)
ASTM A38 1 Welded pipe (4) Grades Y35 through Y65, full materials
(starting form for specification
welded headers,
tees, and elbows)
ASTM A694 Forged flanges, Grades F42 through F65, full material specifi-
fittings, and valves cation, but no impacts required
ASTM A707 Forged flanges Yield strength 42 ksi (289.6 MPa) through
517.1 MPa (75 ksi), full material specification
(4)Usually fabricated from HSLA plate, such as A633, Grades C and E or A737.
Other standard specifications exist for low-carbon, manganese, molybdenum, niobium and low-carbon nickel,
chromium, molybdenum, and copper high-yield strength steels. Valves are either cast or fabricated from plate.
Specific requirements are covered in API 6D. Radiography of weld ends and repair welds often is specified.
4. ALLOYING ELEMENTS
Requirements are placed on alloying elements in line pipe steel to control strength, weldability, and
notch toughness.
Carbon usually is limited to 0.20% (or lower, e.g., 0.016%) in API X60 to X70 pipe and 0.26% (or
lower) in API X52 pipe to enhance weldability. Steels with 0.08% maximum carbon (low-carbon mar-
tensite) are used to avoid the need for PWHT in heavy sections.
Silicon usually is specified as 0.13% to 0.33% in the United States when notch toughness is required.
Semi-killed (no minimum silicon) steel is still used in the United States for low-strength steels.
Manganese often is limited to 0.8% to 1.6%. With less than 0.8% manganese, there is too little
manganese to tie up sulfur, although, theoretically, a minimum of 0.4% manganese is needed. Also, a
manganese-to-carbon ratio Of 4: 1 minimum is required for good notch toughness.Weldability decreases
with a manganese content above 1.6%. Some specifications limit the manganese-to-siliconratio to 3: 1
minimum to maximize weldability .
Nitrogen, if possible, should be limited to 0.009% maximum; however, this usually is possible only
with basic oxygen steel. The limit on nitrogen usually is 0.012%. Although some nitrogen is desirable to
increase strength, poor base metal and heat-affected zone notch toughness and weld cracking occur as a
result of nitride formation when nitrogen is too high. For nitrogen-bearing grades, a 120°C (250°F)
preheat for welding usually is required. Where specifications permit nitrogen in excess of 0.01%, the
vanadium-to-nitrogen ratio should be specified as 4: 1 minimum to avoid embrittlement.
Aluminum is a deoxidizer and a grain refiner. It also combines with nitrogen to prevent strain age
embrittlement. Some specifications require the aluminum-to-nitrogen ratio to be 2: 1 or greater to mini-
mize strain-age embrittlement. If aluminum exceeds about 0.05%, it can result in low toughness in weld
metal when diluted into the weld metal by high-heat input process, such as submerged arc welding. The
low toughness is a result of acicular ferrite formation in the weld. Therefore, aluminum usually is lim-
ited to 0.05% maximum.
Vanadium and niobium (columbium) are the most commonly added grain refiners. Many users limit
the sum of vanadium and niobium to 0.10% maximum. Vanadium provides the best increase in strength
for the amount added. Vanadium and niobium cause strengthening from grain refinement and increase
the strength by forming carbides and nitrides. Niobium also causes strengthening from precipitation
hardening. Unless the niobium is greater than 0.025%, induction bending may drop the yield strength
below the minimum in API X60 and higher grades. Niobium additions usually are limited by cost
because double inoculation often is required. Vanadium additions are limited because of poor low-
temperature properties, temper embrittlement, and underbead cracking when the concentration becomes
too high. PWHT will lower the fracture toughness of vanadium-bearing steels and weld metal, but only
reduces toughness of niobium-bearing weld metals.
Titanium and zirconium occasionally are used as grain refiners. Titanium also is a carbide former and
usually is limited to 0.020% to 0.030%. Titanium significantly increases the notch toughness up to
0.020%. However, titanium is harmful to notch toughness when the percentage exceeds 0.040%. The
zirconium content usually is limited to less than 0.2% to minimize martensite and massive upper bainite
formation.
Molvbdenum causes the formation of fine acicular ferrite in preference to polygonal ferrite in low-
carbon steels. The molybdenum content usually is less than 0.25% to limit martensite and massive upper
bainite formation.
Rare earth metal (REM) additions and calcium-argon blowing are used to control sulfide amounts and
shapes to increase notch toughness. Rare earth metals are cerium, lanthanum, and misch metal. Misch
Kh
\
Mlnlmum Preheat
and Interpass
Temperature, "C
-
Figure 5.10 The permissible
maximumcarbon equivalentfor
butt welds in line pipe using
cellulosic (EXXlO) electrodes
based on minimum preheat and
-20 interpass temperature,pipe wall
20 thickness,and heat input.6
mm 60
4 M
(Reprinted from Micro AIloying 75 -
Proceedings of Micro Alloying 75
Conference, October 1, 1975, Union
I I 1 I II I I Carbide Corp., Metals Division, New
30 20 10 7.5 0.51 0.43 0.38 0.35 York, NY)
HEAT INPUT, W/cm CARBON EQUlVALENT
Mn C r + M o + V + N i + C u )
(C+-+
6 5 15
metal is a mixture of elements with atomic numbers of 57 to 7 1 containing about 50% cerium. Usually,
1.5 times the sulfur level is added. Problems with gas metal-arc welding have been reported when REMs
exceed 0.02%.
Sulfur is kept low (e.g., 0.005% maximum) to enhance notch toughness and resistance to hydrogen-
induced cracking. However, when the sulfur is 0.001% or below, weldability problems have occurred in
low-heat input processes because of poor fluidity of the weld puddle. Sometimes the sulfur content of
the weld filler metal is specified in the 0.01 range to offset the low sulfur in the pipe.
As discussed in the above section on chemistry, carbon and, to a lesser extent, other alloying elements
have a significant effect on weldability. The effect commonly is expressed as the carbon equivalent (CE),
defined as either:
(1) C E = C + M n or (2) C + M n + l C r + M o + V ) + J N i + C u )
4 6 ' 5 15
Preheat is not required for grades up to API X52 when the CE is less than 0.5 to 0.55 (by definition [ 11).
Definition (2) is commonly used in specifications. Limiting definition (2) CE to 0.43 for API X60 to X70
grades usually is required. The maximum carbon equivalent, based on preheat and interpass temperature,
wall thickness, and heat input (travel speed), can be estimated from Figure 5.10.6 For example, for 0.42 CE
and 25 mm (1 in.) wall pipe, a preheat of 20°C (68°F) would be required for a 12 KJ/cm heat input; however,
if the heat input was reduced to 9 KJ/cm, then a 60°C (140°F) preheat would be required.
A refinement to the CE is the parameter Pc,, which has been developed to predict hydrogen cracking
susceptibility as it relates to weldability. The Pcmchemistry parameter, applicable to most low-alloy steels
having carbon contents equal to or less than 0.12 wt%, is defined as:
Po,,= C + Si + LMn + Cu + Cr) + Ni + &+ V + 5B"'
30 20 60 15 10
where:
He, = 0.50 Hf for cellulosic electrodes
He, = H, for low-hydrogen electrodes
H, = diffusible hydrogen content per lOOg of fused metal
KI = Stress concentration factor
- 3.5 for the root pass of a double V-groove
- 1.5 for the root pass of a simple V-groove
sw = nominal stress acting on the weld metal, Kgf/mm'= ksiA.42
- 0.050 R, for R, less than 20 Sy
- Sy + 0.005 (R,-20 Sy)for R, greater than 20 Sy
sy = yield strength, kgf/mm2
R, = intensity of restraint (kgf/mm') = 69 h
h = thickness (mm)
After Phmis calculated, the time necessary to cool a weld to 100°C (212"F), called critical T100, is
calculated where:
CT = 52-1,Ol l/(T)0.5
+ 74.2 exp (0.00054 T)
where:
CT = Critical preheat temperature (degrees C)
T = TlOO cooling time to 100 degrees C (212°F) in seconds
A third carbon equivalent, CEN, has now become popular. The CEN is defined as:
where: F = 0.75 + 0.25 than (20 [C - 0.12]), thus F = 0.54 for 0.06% C and 1.00 for 0.21% C
(5) Boron less than 0.001 wt% can be left out of the Pcmcalculation.
FINISHINGTEMPERATURE. "F 5. PROCESSING LINE PIPE
6. BENDING
curred as a result of copper contamination on the surface of induction bent pipe; therefore, copper bending
shoes should be avoided.
It is important to control carefully bending parameters, such as bending speed, temperature, and water
quench. Specifications usually include tests to ensure mechanical properties are not degraded below the
minimum requirements and surface inspection, e.g., magnetic particle inspection to ensure no defects have
initiated. As mentioned previously, niobium must be greater than 0.025% to prevent a drop in yield strength
during induction bending. Many organizations use the International Pipe Association’s Voluntary Standard
for Induction Bending of Pipe (IPA-VIBS-86)as a basis for a bending specification.
The limits on cold bending of line pipe are contained in Paragraph 841.231 of ASME B31.8. As with
cold expansion discussed previously, the maximum strain permitted is limited to a very low level to avoid
loss of toughness, etc. For carbon and low alloy steel other than microalloyed steels, such as line pipe, the
strain usually is limited by stress corrosion cracking considerations.For example, NACE htemational MR0175
limits the strain to 5% to avoid sulfide stress cracking (SSC). The strain in a bend can be calculated from the
following equation:
eo= 1
-
2R +1
I
t
where:
eo= the strain on the outside fibers
R, = the inside radius of the bend
t = the thickness of the pipe
120
-
Figure 5.1 4 Fracture characteristics observeredfrom line pipe through the transition
temperature. (Reprinted with permission from Sixth Symposium on Line Pipe Research, October 1979,
Research sponsored by the Pipeline Research Committee, American Gas Association, Ariington,VA)
7. FRACTURE
Long pipeline fractures such as the one shown in Figure 5.13*can occur in either a brittle or a ductile
manner. The fracture appearance and the speed of fracture of ductile and brittle modes are quite different (as
shown in Figure 5.14).’Various standards have minimum requirements for fracture control. The common
ones in the United States are:
API 5L Appendix SR 5 covers Charpy V-notch testing on pipe with diameters of 115 mm (4.5in.) or
greater. The average shear value of three specimens shall not be less than 60%, and the all heat average
shall not be less than 80%.
API 5L Appendix SR 6 requires 40% minimum shear in a drop-weight tear test on 80%of the heats for
508 mm (20 in.) diameter or larger, API X52 or stronger pipe.
The requirements for Canada are contained in Canadian Standards Association - 2184.Code require-
ments are, of course, subject to change. In addition to the United States and Canada, a number of other
countries have regulations containing specific requirements. Most users specify more than the minimum
toughness required by the governing codes or regulations.
Brittle fractures of line pipe typically follow a sinusoidal path, as shown in Figure 5.15.’A high-speed
photograph of brittle fracture (Figure 5.16)’reveals the advancing tip is well ahead of any significant defor-
mation on the pipe; the crack travels faster than the energy in the pipe can be released. The longest recorded
brittle fracture was 13 km (8.1 miles). It occurred in 1960on a 760 mm (30in.) diameter API X56 line being
gas tested at 60% yield. Research by Battelle reveals that brittle fractures can be avoided in steels other than
quenched and tempered ones by specifying 85% average shear area in a drop-weight tear test at the mini-
mum service temperature.@)This is considerably more stringent than API 5L Appendix SR 6.
Figure 5.17 shows that the % shear in a DWTT only differs slightly from that observed in full scale
tests.’ Meeting the 85% shear criterion also assures that propagating ductile fracture will be arrested. Since
the DWTT is applicable only to the pipe base metal, Charpy V-notch test criterion (applicable to both weld
'0 20 40 80 80 1 0
DWTT PERCENT SHEAR
-
Figure 5.17 Relationship -
Figure 5.16 Photo of brittle fracture from high speed
between DWrr percent shear movie (3,000 frame&)? (Reprinted with permission from Sixth
and full-scale pipe percent Symposium on Line Pipe Research. October 1979, Research
sponsored by the Pipeline Research Committee, American Gas
shear of the fracture surface.* Association, Arlington, VA)
(Reprinted with permission from Sixth
Symposium on Line Pipe Research.
October 1979, Research sponsored by
the Pipeline Research Committee,
American Gas Assoctetion, Arlington,
VA)
-
Figure 5.18 Photo of
ductile fracture from high
speed movie (3,000framed
s).2 (Reprinted with permission
from Sixth Symposium on Line
Pipe Research, October 1979.
Research sponsored by the
Pipeline Research Committee,
American Gas Association,
Arlington, VA)
metal and pipe material) commonly is used to
assure that ductile fracture will not propagate.
The Charpy V-notch approach is discussed in
the next section.
C,= 0.0873U2,(Rt)”A,
Figure 5.21 -The Charpy V-notch energy needed to produce fracture arrest (after Maxey).6
-
(Reprinted from Micro Alloying75 Pmceedlngs of Micm Alloying 75 Conference, October 1,1975, Union Carbide
Corp., Metals Division, New York, NY)
16 mm (5/8 in.), compensation in the form of increased energy absorption requirements often is specified for
thick pipe. It has been recommended that the value calculated from the Battelle equation be increased by
30% when the applied stress is more than 350 MPa (50.8 ksi). Furthermore, the validity of the Battelle
equation for X80 material has been questioned. Ongoing research indicates that new criteria of specific
fracture energy, crack tip opening angle and specific DWlT energy should be used in lieu of Charpy V-
notch energy when designing against ductile fracture in X80 pipe.
When a submarine pipeline is to be installed using a lay barge, the yield strength usually is used for the
stress in equations for calculating the fracture toughness requirements. In addition, when the longitudinal
stress due to laying exceeds 3/4 of the minimum specified yield strength of the pipe, the minimum yield
strength of the round seam weld metal has been required to be 34.5 MPa to 68.9 MPa (5 ksi to 10 ksi) above
the minimum yield strength of the pipe. This prevents concentration of the strain in the round seam weld
metal, which has resulted in fractures occurring during pipe laying.
Work sponsored by the AISI Committee of Large Diameter Line Pipe Producers yielded an equation
similar to Battelle’s.’OThe committee points out that the impact energy required by these equations is insuf-
ficient to stop propagation when splits occur. Splits are parallel to the plane of the original plate surface and
generally include groups which form a chevron pattern. Splits occur in controlled rolled steel but have not
been found in quenched and tempered steels. To date, the exact metallurgical mechanism that causes split-
ting has not been established, but it is associated with low finishing temperatures and high levels of me-
Appendix A
SCOPE
A53, A106
A67 1, A672
A69 1
A214, A179
A192,AZlO I A105, A181
A266,A234
1 1
A29, A575
A57667qp63
A216
A790
8675
B407
B423
B673.8677
8464,8474
842
Admiralty Brass (C44300)
Naval Brass (C46500) 8171
70-30 Cu-Ni B171 8467,8608
Titanium Gr.2 (R50400) B265 B337
Alloy 400 (NO4400) B127 B165
Alloy 625 (NO6623 8443 B444
Alloy C276 (N010276) B575 B622
Ni Resist TP2 (F41002)
Aluminum B209 B24 1
General Guidelines for Materials Selection
and Corrosion Allowances
1. Materials typically used for the following environments are contained in Tables A-2 through A- 17. It
is important to note that the materials listed are typical and that there can be exceptions.
General - hydrocarbon with low sulfur contents, non-corrosive steam and water
Hydrocarbon plus sulfur greater than 1 wt%
Hydrocarbon plus sulfur greater than 0.2 wt% plus naphthenic acid
Hydrocarbon plus sulfur between 0.2 wt% and 1.0 wt%
Hydrocarbon plus hydrogen
Hydrocarbon plus hydrogen and hydrogen sulfide
Sour water and desalter water
Carbonate
Low pressure wet carbon dioxide
High pressure wet carbon dioxide
Amine
Acid gas
Liquid sulfur
Untreated, aerated water
Caustic
Valve trim
3. The numerals after a material designation indicate the minimum nominal corrosion allowances in
millimeters (mm) as follows:
VESSELS 4 CS+15* -.
L.
1-114Cr + 3" 4
me
5 (Note 7) 4 cs i-i14cr* +
Exchanger T u h (Note 1,35) 4 cs 1-114CP
CaSe 4
(S-1) (Note 33)
lmpelkr 4 cs
I
NOTES
'See Note 2,26
"See Note 8
tSee Note 9
TABLE A-3
OPERATINGTEMPERATURE
100-
38
290.
lba
3pO-
lb
490
200
. 590
260
690
315
7p
330
I 6p
4i5
900
4hO
1,opo
540
1,100
590
l.?W
650
1,300
700
1,400
760
F
C
VESSELS
(NW 7)
ExchmgmTubes (Note 1)
FURNACETUBES (Note 3)
Impeller
MOTES
'see Now 2,28
"Sae Note 18.29
Noh, 8
TABLE A-4
-
450 -
850
455
,-I
230
TubeahaeQ ol FH C
-
B.mer (Note 7)
-
-cs
-cs
cs
-1
--
L 1
Same as Above
318L"
316L"
b
PIPING 4C
-S + 1.5 (Note 5) 318L + 1.5" D
Caw -cs --
A
~IBTI+I
318"
3m
P
6 - 11
Impeller -c1 316" P
NOTES:
'Sw Note 2
".Sea Note 29
TABLE A-5
SERVICE I OPERATING TEMPERATURE
Hydmarbon * S ( 0 . 2 - 1 W t x ) 100 200 yo 490 - , -800,
s q o 600_ , 700 - 900. - 1,900 l,!W 1,700 1,300 1,400 F
C W Units. Fluid Catelybc i8 100 WiO 200 260 315 370 425 480 540 590 650 700 760 C
Cracken. COkeN. etc.
VESSELS
iXCHANOER SHELLS AND CHANNELS 4 CS + 3 (Notes 5.9) b+CS Clad wiM 2.5 mm 1Xr" or Solid 5Cr + 3 4
Impeller
I
NOTES:
"See Note 18. 28. 29
TABLE A-6
SERVICE OPERATING TEMPERATURE
H y d m + Hx (Ha< 0.01 NI%) loo 2w 3P- 490 9 0 600 790 8po -spo_I.poo- ,
1.100 l.?W 1,Qoo 1,too F
-
ppHa 0.7 3.6 MPU (I00 PI.) - i8 lbo I I50 200 260 345 370 425 480 540 590 650 700 760 C
If X3.6 MPlr (500 ph),U.N& 20
(NOIlCWOSlvr)
Hydrogen Plants. Catatybc
Reforman. &. -
500 700 455 800 980
260 370
VESSELS cs +3 b +(Note 20) +0 I-INC~ +3 d-.
18-8+1ocRef IumdCS' -b
cs+3
cs
b c ( W e 20)
b t ( N o t e 20)
b t(Note 20)
+*
+
+0
(Note 9) 0
1-1NCr
l-lNCr+ 3
1-1/4Cr
18-8+1wRef I W C S '
18-8
;
Barbs (Note 7)
cs
b t ( m 20) +I
b +(Note 20) -b 0
1-114Cr
I-1NCr
18-8
18-8orAlloy8WH
Impeller
NOTES
'See Note 15
Ref. = Refractwy
TABLE A-7
OPERATING TEMPERATURE
100
d0
2po
100 '
390
150
490
200
f260y - :
600
315
-
790
,
370
I
890
425
,
900
-: - 1,900
400
.
540
-
1.]00
*
590
I
l,?W
.
650
-1,
. p- 1.f00
700
.
760
I F
C
-
925
1-1NCr" or 2-114CP
VESSELS
321 or 347
I
b
C
-S
'I " 321 or 347 b
case
Impeller
NOES
'See Nole 5
"See Note 20
SERVICE OPERATING TEMPERATURE
Sour Water 100 200 3po 490 spa- 6W 790 8pO 900 l,Q00 1,100 1.?00 1,QOO 1,fOO F
Dwlbr Water 58 IbO 150 I 200 260 3i5 370 425 400 540 590 650 700 760 C
VESSELS
I+$ pp c 70 k P u (10 p a ) M 1,OOO
PPm
H2S pp > 70 kPM (10 paia) M 1.000
ppm
Trap and Inlamah(W4)
Balllea (Note 7)
ExchangerTuba (Note1)
p
pw < 70 kPM (10 pdr)
p
p w > 70 kPaa (10 paia)
U n m W water otherside
SERVICE OPERATINGTEMPERATURE
wwplmt,
Gar TreaUng Plant. etc.
VESSELS
BalMr (Note 7)
PIPING
FURNACETUBES (Note 3)
Case
lmplkr
NOTES
154
N
0
%
z,
0
+
s1
n
r-
x
b
r
N
TABLE A-15
SERVICE
Untmatod,*mw Water
-
I50
VESSELS -CS + 2.5 mm BO Cu-10 Ni Clad 4
Impeller
NOTES:
*& N& 27
TABLE A-1 6
Caurlk
NaOH, KOH
- -
-
200 75 200
+
TCS+.-:
VESSELS -cs + 1 5' +Alloy4Oo+lS ccs +3'4.h Alloy400+15 L
,
Trap and lntemals ( M e4)
4
Allay400
Alloy400+ 1 5
Alloy400 -
C C S '
C c s + 3.
+CS'+
-*
-b M
4
4
-w
Alloy 400
400+ 1b5- ,
Alloy 400 b
Barnes (Note 7)
1-
CS' Alloy 400 4. C-CS' --., 4 Alloy 400 b
4
Exchanger Tubes (Note 1) -CS' Alloy 400 b +CS' -b 4 Alloy 400
Impeller -CI'
(S-1)
Y
. h +CI'
(S-1)
-4
(Note 34)
Alloy400 _____*
NOTES:
'See Note 10
Table A 4 7 - VaIveTrim
1. General Service
a) Treated Water
>3.4 MPag (500 psig) in single phase flow Stellite 6TM(UNS R30006)
4. Amine
a) <100"C (200°F) 12 Cr
6. Caustic
2. Above 400°C (750"F), use silicon-killed (not aluminum-killed) carbon steel. Above 440°C (825"F),
use 1 Cr-1/2 Mo or 1-1/4 Cr-1/2 Mo.
3. Materials selection for furnace tubes in non-hydrogen containing hydrocarbon plus sulfur services is
based on the assumption that the metal temperature is 50°C (100°F) higher than the internal fluid tempera-
ture. To avoid excessive oxidation the outside skin temperatures should not exceed 530°C (1,000"F) for
carbon steel, 650°C (1,200"F) for 1-1/4 Cr to 9 Cr and 930°C (1,700"F) for 18-8 SS. When austenitic SS
is specified for temperatures above 530°C (1,00O"F), "H' grades should be used, and cold forming should
be prohibited where creep strength governs the allowable stress unless it is followed by solution anneal-
ing. Cold forming of carbon steel furnace and boiler tubes, e.g., cold bends, should be prohibited unless
followed by stress relief.
5. Use only silicon-killed steels for temperatures above 232°C (450°F). For areas where sour water
collects, see Table A-8.
6. Where not covered by TEMA and the material specified for both sides is the same, corrosion allow-
ance should be 0.75 x the sum of the corrosion allowances for each side up to 6 mrn (1/4 in.) maximum.
Where not covered by TEMA and alloy requirements for two sides are different and solid alloy tube sheet
is used, use corrosion allowance for higher alloy side as the total corrosion allowance. Where cladding is
required on the tubeside, the minimum thickness of cladding should be 10 mm (3/8 in.) so that at least one
of the grooves for rolling will be in the cladding.
7. Baffles should be 6 mm (1/4 in.) minimum thickness; no other corrosion allowance need be used.
8. Where 1-114 Cr-112 Mo is specified, 1 Cr-112 Mo also may be used. Do not use 1 Cr-1/2 Mo in
hydrogen service if hydrogen partial pressure is greater than 0.7 MPaa (100 psia) above 480°C (900°F).
Do not use 1 Cr-1/2 Mo in general service above 530°C (1,000"F).
9. A corrosion allowance of 3 mm (1/8 in.) should be used on carbon steel and low alloy steel exchang-
ers since it is standard for TEMA Class R.
10. Stress relieve carbon and chrome steel welds and cold bends in amine service regardless of service
temperature. For all concentrations of carbonate solutions and in concentrations of caustic up to 30%,
stress relieve for service temperatures above 60°C (140°F). For 30% to 50% caustic, the service tempera-
ture where stress relief is required decreases from 60°C (140°F) to 48°C (118°F). Welded tubing does not
require heat treatment in addition to that required by the ASTM specifications. Rolled tube-to-tube sheet
joints do not require stress relief.
11. Use 12 Cr for valve trays and valves. Sieve trays and stationary bubble cap trays may be made of
carbon steel.
12. For control valves and other areas of high turbulence (velocity >2.5 m / s [ 8ft/s]) e.g., downstream of
control valves, the rich carbonate inlet of a carbonate regenerator, reboiler tubesheets, baffles, etc., use
304 SS (UNS S30400) plus 1 mm (1/32 in.) corrosion allowance. Do not use miters; long-radius elbows
are preferred. Piping specificationsusually contain other limitations on miters.
13. Hardness of completed carbon and low-alloy steel welds should not exceed 200 Brinell. Valve trim
should be 18 Cr-8 Ni SS and meet NACE Standard MR0175. For piping, if the ammonium bisulfide
concentration exceeds 4% or the product of the mol% H,S and the mol% NH, exceeds 0.05, use materials
recommended for ppH,S >70 kPaa (10 psia).
14. Use alloy 400 (UNS N04400) valve trim for caustic service above 93°C (200°F).
15. The metal temperature should not exceed that where carbon steel starts to lose its resistance to hydro-
gen attack.
16. If the service temperature will exceed 480°C (900"F), check with a materials engineer.
17. Stress corrosion cracking of 300 series SS may result if solids carry-over exceeds 1 ppm.
18. Do not use type 405 SS (UNS S40500)above 343°C (650°F). When welding is anticipated, use 410s
SS (UNS S41008) rather than 410 SS (UNS S41000).
19. For U-bends, heat treat the entire 18 Cr-8 Ni SS tube at 1,OOO"C (1,850"F) minimum after bending.
Where a 1OOO"C (1850°F)minimum heat treating temperature is not practical, 900°C ( 1,650"F)may be
used. Use 321 (UNS S32100) or 347 SS (UNS S34700) if only the U-bends can be heat treated.
20. To choose between carbon or alloy steels in hydrogen service, see API 941. The corrosion allowance
given in the table must be applied.
21. See Table A-10 for recommended materials for the top of regenerator column and overhead system.
22. Severe corrosion may occur if lines are not kept above the dew point.
23. Welded assemblies must be heat treated at 900°C (1,650"F) for 4 hours after completion to prevent
polythionic acid cracking during downtime.
24. Unless more restrictive velocities are specified, the maximum velocity should not exceed 6 m / s (20 ft/
s) in mixed phase flow piping where the ammonium bisulfide concentration exceeds 4% or the product of
the mol% H,S and the mol% NH, exceeds 0.05. Use 2205 (UNS S32205) for tubes and 309L (UNS
S30903) clad headers in mixed phase flow where the ammonium bisulfide concentration exceeds 4% and
2205 (UNS S32205) piping where the ammonium bisulfide concentration exceeds 8%. For control valves
where high turbulence occurs due to a pressure drop of greater than 3.8 MPag (2,000 psig), use StelliteTM
6 (UNS R30006) valve trim and 304L (UNS S30403) piping for 10 diameters downstream of the control
valve.
25. Do not use nickel or cobalt base alloys, e.g., alloy 600 (UNS N06600), alloy 400 (UNS N04400),
Colmonoym, etc.; StelliteTM(UNS R30006) may be used.
26. The 1.5 mm (1/16 in.) corrosion allowance should be used where the vessel is externally coated
(painted) and either internally lined or where experience has shown that the corrosion rate is essentially nil
e.g., pure hydrocarbons, dry steam, etc.
27. The choice between brass and copper nickel alloys is contingent upon ammonia content and tempera-
ture of process side. Brass should not be used when the pH due to ammonia exceeds 7.2. Copper nickel
alloys should not be used if the sulfides in the water exceed 0.007 m g L Admiralty (UNS C44300) is
usually limited to a process inlet temperature of 177°C (350°F) for cooling service. Aluminum bronze
(UNS C61400) or copper nickel alloys often are used when the process inlet temperature is 177°C to
232°C (350°F to 450°F). Aluminum bronze (UNS C61400) has a somewhat higher tolerance for hydrogen
sulfide than copper nickel alloys.
28. For lines and equipment handling catalyst, use refractory-lined steel or hard facing on the indicated
alloy. Hard facing is not required for vertical pipe runs.
29. Use solid 5 Cr or 12 Cr clad for hydrocarbons containing over 1 wt% sulfur above 290°C (550°F) and
for crude oils containing 0.1 wt% to 1.O wt% sulfur above 340°C (650°F) unless there is operating
experience or hydrogen sulfide evolution data to indicate where the break between carbon steel and alloy
should be.
Regardless of the sulfur content, when hydrocarbons have a TAN above 1.5 mg of KOWgm and the
temperature is above 230°C (45OoF), use 316L (UNS S31603) - except use 317L (UNS S31703) for
California crude oils. For castings, use CF8M (UNS 592900) or CG8M (UNS 593000) provided the ferrite
content is 8% minimum.
30. When 18 Cr-8 Ni SS is specified, any grade may be used; however, unstabilized regular carbon
(0.08% carbon maximum) grades usually are not used for operating temperatures above 425°C (800°F).
For temperatures above 425°C (SOOOF), stabilized grades should be used if there is a possibility of inter-
granular attack during downtime. For 321 (UNS S32100) or 347 SS (UNS S34700) in thickness in excess
of 13 mm (1/2 in.), restricted chemistry (0.04% to O.O5%C, 0.015 max. P & max. S, Crmbl.6) should be
used. In addition, 347 (UNS S34700) should be limited to 19 mm (3/4 in.) maximum to avoid problems
associated with welding. Types 309 (UNS S30900), 310 (UNS S31000), 316 (UNS S31600), 321 (UNS
S32100), and 347 (UNS S34700) should be used with caution for operating temperatures above 600°C
(1100°F) because of the possibility of sigma phase embrittlement.
31. When experience shows that coking will occur above 455°C (850"F), 1-1/4 Cr-1/2 Mo can be used up
to 590°C (1,100"F) and 2-114 Cr- 1 Mo can be used up to 650°C (1,200"F). Conversely, if coking does not
occur or if high velocities occur that will prevent coke lay-down, SS is required. The choice between 410s
(UNS S41008) and 18-8 depends on the anticipated loading. This is because 410s will lose some room
temperature ductility due to 475°C (885°F) embrittlement. Thus, 410s should be limited to parts with
relatively low stress levels.
32. Designations in parenthesis, e.g. (S-1) are API 610 materials classes.
33. For water service from 120°C to 175°C (250°F to 350°F) use class S-5. For water service over 175°C
(350°F) or boiler feed water over 100°C (200"F), use class C-6.
35. Experience has shown that carbon steel tubes will only give economical life if water treating and
corrosion inhibitors additions are carefully controlled on a continual basis.
36. When ammonia is present in the stripper overhead system, ammonium bisulfide forms. This requires
3 16L for reflux piping.
Table A-I 8
Chemical Compositions of Some Common Stainless
Steels, Special Stainless Steels and High Nickel Alloys
*CPI (Critical Pitting Indcx) or PRE, (Pitting Resistance Equivalent, nitrogen included) = %Cr+ 3.3 x (%Mo + %W)+ 16 x %N
General Guidelines for Materials for
Low-TemperatureServices
2. Materials listed in Table A- 19 are selected based on minimum requirements for operation with respect
to brittle fracture at indicated temperatures in accordance with the requirements of the following codes:
ASME Boiler and Pressure Vessel Code, Section VIII, Division 1 (UCS-66) and Division 2 (AM
204)
ASME B3 1.3 Chemical Plant and Petroleum Refinery Piping (323.2)
3. These code requirements are considered the minimum. Additional testing requirements or more
stringent requirements than those required by the codes may be necessary, depending on the specific
circumstances.
4. The minimum operating temperature should include cold startups at low-ambient temperatures where
applicable (e.g., mining equipment which normally does not require warm startups) and upset conditions.
5 . Select low-temperature steels for fracture-critical structural members designed for tensile stress levels
greater than 40 MPa (6 ksi), and specify a minimum Charpy V-notch impact energy absorption of 27
joules (20 ft-lbs) for base metal, heat affected zones, and welds when the structures are exposed to low-
ambient temperatures. Fracture-critical members are those tension members whose failure would have a
significant economic impact.
6. When materials requiring impact testing are used for welding, impact tests should be conducted on the
base metal, weld metal, and heat affected zones.
7. Materials for atmospheric storage tanks should be selected in accordance with API 650.
8. Materials for low pressure storage tanks should be selected in accordance with API 620, Appendix R.
TABLE A-1 9
MINIMUM PLATE!? STRUCTURAL FORGINGS 6
DESIGN SHELLS AND HEADS I TRAYS NONPRESS PIPING TUBING FITTINGS CASTINGS FASTENERS
--TE a
--
C
57
F
135
4PPLICATIONS FOR ASME A36 4PI 5L GR. B, NO A105 A216
BOILER AND PRESSURE OR A53 OR. B, SPECIAL OR A234 GRADE WCB
VESSEL CODE, SECT. Vlll A283 A106 GR. B REQUIRE-
MENT
0 0 0
DMSIONS 1 AND 2: OR A671
-6 20 FOLLOWFIGURE USC-88 FOR NON- A53 GR.B A105 A352 A193
FOR DMSION 1 AND CRITICAL SEAMLESS A234 GRADE LCB GRADE 87
FIGURE AM-218.1 FOR AS. A106 GR. B A727 OR A758 (NOT IMPACT
DMSION 2 FOR w r L OTHERWISE OR A671 @ 00 TESTED)
-17 0 SELECTIONAND IMPACT SAME AS A106 GR. B A707 GR. L l
TEST REQUIREMENTS. PLATE FOR OR A524 (ONLY). A727 @ @ A194
PIPING 0 OR A758 0@ GRADE 2H
-30 -20 PREFERRED MATERIAL: SAME AS VESSEL A333 OR. 6 A334 A350 GR. LFl A352 A193
A 516 (ALL GRADES) EXCEPT NO SPEC- GRADE 6 OR LF2. A420 GRADE LCB GRADE 87
'lAL REQUIREMENTS OR. WPL6 OR A194
oooa3 FOR BOLTED
TRAYS 40GA
@ 043 GRADE 2H
6
-46 -50 FOR ASME SECTION VIII, SAME AS VESSEL SAME AS A333 A334 A352 A320
DIVISIONS 1 a 2 USE ~ fYPE304 PLATE OR GRADE 7 GRADE 3 A350 GRADE LC1 GRADE L7
A203 OR A537 AS APPLI- OR ALUMINUM PIPING GRADE LF3
TO -74 OCI
CABLE FOR MAT'LS AND . A194
-100 OF
IMPACT TEST GRADE 4
REQUIREMENTS
-60 -75 A203 A352
GRADE D GRADE LC2
<Po A333
GRADE 3
TO -74CI-1OOF
A352 GR. LC3
A522 TP. @ A351 A320. GR. 08
-100 -150 A353 OR
: :2A312 43 ---
----
A334
GRADE 4
a
-----
A553 TYPE 1 A420 (WPLB)
- -AT2- -
A194, OR. 8
TYPE 304 A213 -----
ANNEALED
TP. 304 8 ---TP304 6 ----
GR. F304 @-
-------
8209 AL ALLOY 506Y5456
-----
ALUMINUM
---- 82471361 (6061)
8211
TP 2024-T6
-200 -325 A240 TYPE 304 @ @ 4182. F 304 @ @ SAME AS
------- --- --- ---
4403,WP304- - A351 ABOVE (I.E. GR. 8
--
-255 -425
8209 AL ALLOY 5083/5456 8241 @-
ALLOY 6061
8234 @ B247/36-
ALLOY 6061
-
ALLOY 6061 f'ww OR 2024-T6)
@@
NOTE: THE NUMBERS IN CIRCLES CORRESPOND TO FOOTNOTE NUMBERS ON THE FOLLOWING PAGES
Notes for Table A-19
1. Bolted trays: No special requirements
Welded trays: No special requirements up to and including 13 mm (1/2 in.) thick. Same as vessel
shell above 13 mm ( 1/2 in.) thick.
2. All weld seams in materials requiring impact tests per ASME Section VIII, Divisions 1 and 2, Figures
UCS-66 and AM218.1 (regardless of the governing code) should be 100% radiographed and magnetic
particle inspected.
3. For forgings in thicknesses greater than 25 mm (1 in.), consider restrictive carbon content, e.g., 0.32
maximum, to enhance weldability. A105 forgings are not permitted for tubesheets or shell rings per the
scope of the materials standard.
4. In general, carbon, low alloy and high alloy steels may be used at design metal temperatures down to
-46°C (-50°F) without impact testing under the following (exempt) conditions:
a. ASME Section VIII, Div. 1 Impact tests are not required when the
paragraphs UCS-66 and UCS-67 intersection of minimum design metal
(and Figures UCS-66 and temperature (MDMT)* and nominal material
UCS-66.1) thickness lies on or above applicable
material curve in Figure UCS-66, except
*Note: The MDMT at which impact testing is mandatory for:
impact testing would
otherwise be required may (1) All material thicknesses greater than
be reduced in accordance 102 mm (4 in.) for welded construction.
with Figure UCS-66.1 when (2) All material thicknesses greater
the stress in tension is than 152 mm (6 in.) for non-welded materials
less than the maximum with a MDMT less than 49°C (120°F).
allowable design stress.
It also may be reduced 15°C Also exempt from impact tests are:
(30°F) under UCS-68, if
postweld heat treatment is (1) ANSI B16.5 or B16.47 femtic steel flanges
performed when not otherwise with a MDMT not colder than -30°C (-20"F),
required by code. (2) All UCS materials less than 2.5 mm
(0.098 in.) thick and UCS nuts provided
such CS materials are used at MDMTs
not colder than -46°C (-50"F),
(3) All P-No. 1 Group 1 or 2 materials 25
mm (1 in.) and less provided the vessel is
hydrostatically tested, has an MDMT
between 343°C (650°F) and -29°C (-20"F),
and shock or cyclic loading is not a control-
ling condition. See UG-20 for additional
exemptions.
(4) Materials listed in Table UG 84.3 are
exempt from additional tests provided the
MDMT is warmer than the test temperature.
Note: Welding Procedure Qualification Tests
must include weld and heat affected zone
impact tests unless specifically exempted by
paragraph UCS-67.
b. ASME Section VIII, Div. 2 Impact tests are not required when the
Paragraph AM 218 intersection of MDMT and nominal material
(and Figure AM-218.1) thickness lies on or above applicable
material curve in Figure AM-2 18.1,
except impact testing is mandatory for:
c. ASME B31.3, Section 323.2.2 Impact tests are not required when the
(and Table 323.2.2) MDMT is below -29°C (-20°F) but at
or above -46°C (-50°F) and both the
maximum operating pressure does not
exceed 25% of the maximum allowable
design pressure at ambient temperature
and the combined longitudinal stress
(from pressure, dead weight and displace-
ment strain) does not exceed 41 MPa
(6,000 psi).
5. Type 304 (UNS S30400) is listed because it is the least costly of the acceptable materials. Other 300
series SS may be needed due to considerations other than low temperature. For example, low carbon
grades are desirable for seacoast environments to avoid intergranular stress corrosion cracking during the
periods when the material is not at cryogenic temperature.
In general, austenitic SS materials are exempt from impact testing at temperatures of -254°C (-425°F) and
higher with the following exceptions:
Applicable Code Summa? of Rules
a. ASME Section VIII, Div. I Grades other than 304,304L, 3 16,316L and 347
Para. UHA-51 (UNS S30400, S30403, S31600, S31603 and
S34700) are not exempt from impact tests at
temperatures of -200°C (-325°F) and higher if
they are materials: 1) with allowable carbon
contents in excess of 0.10%; 2) in cast form; 3)
which have not been solution heat treated; or 4) in
the form of weld metal unless they are otherwise
exempted by Paragraph UHA-5 1.
b. ASME Section VIII, Div. 2, Same exceptions from impact tests as Division 1
Para. AM-213 except types 316 and 316L (UNS S31600 and
S3 1603) are not in first category exempt from
testing down to -254°C (-425°F). UHA 213.1 has
exemptions for stress intensities less than 41.4
MPa (6,000 psi).
c. ASME B31.3, Section 323.2.2 Essentially the same exceptions from impact tests
(and Table 323.2.2) as ASME VIII, Div. 2 with slight variations.
See Table A- 18 and check all product forms;
because, for some types, the lowest exempt
temperature varies with product form or material
specification.
6. Impact testing of ASTM A193 Grade B7 studs (but not ASTM A194 Grade 2H nuts) is required by
ASME Section VIII, Div. 1 for temperatures below -40°C (-40"F), by ASME VIII, Div. 2 for temperatures
below -29°C (-20"F), but not by ASME B31.3 above -46°C (-50°F) ifthe material is quenched and
tempered.
7. Where design temperatures are not lower than -46°C (-50"F), impact testing is not required on thin
materials: less than 2.5 mm (0.098 in.) under ASME VIII, Div. 1 and less than 2.5 mm (0.098 in.) under
ASME VIII, Div. 2 (see paragraphs UCS-66(d) and AM204.2).
8. Impact tests of aluminum are required only hider ASME B3 1.3 for service below -270°C (-452°F).
Notched tensile tests to prove ductility are required for wrought products for service below -270°C
(-42°F). Notched tensile tests to prove ductility are required for cast products for service below -200°C
(-325°F) by ASME Section VIII, Division 1.
10. Impact tests are required for austenitic SS castings by ASME Section VIII, Divisions 1 and 2 and
under ASME B31.3 for castings in the non-solution annealed condition.
11. ASTM specification A20 lists impact properties generally achievable using standard mill practices.
12. Material should be specified to be in the normalized condition when used in this temperature range.
13. Transverse Charpy V-notch impact testing shall be specified as a supplemental requirement.
14. For postweld heat treatment requirements, see:
a. ASME Section VIII, Division 1, Paragraph UCS-56.(')
b. ASME Section VIJI, Division 2, Paragraph AF-402.
c. ASME B31.3, Table 331.1.1.
15. For fracture-critical tension members, Charpy impact tests may be required.
('1 Reductions in the minimum postweld heat treatment temperature (permitted by Table UCS-56.1) should not be
allowed for materials which must meet the fracture toughness requirements of Figure UCS-66. Postweld heat treatment
should be considered for pressure vessels with walls 25.4 mm (1 in.) or thicker to minimize the possibility of brittle
fracture during hydrostatic test. In some cases, this will allow a 15°C (30°F) reduction in impact testing exemption
temperature (see ASME VIII, Div. 1, UCS-68).
General Guidelines for Materials for Wear
and Abrasion Resistance
1. Materials selection for wear and abrasion resistance should be based on service performance records.
Materials listed below are common materials used for abrasive service.
2. For some components, toughness as well as abrasion resistance is required. For example, ditch teeth
should have a minimum Charpy V-notch impact energy of 27 joules (20 ft-lbs) at the design temperature
and a minimum hardness of HRC 50.
1') HB 550-650
(3)HB 450-650
(4) HB 550-600
(5) HB 250-450
INDEX
Biocides 20
Boiler feedwater 32 173
Brittle fracture 40 41 43 74
120 171
Calcareous deposits 20
Calcium treatment, and HIC 49 127
Carbonate stress corrosion cracking 42 97 132
Carbon dioxide
high-pressure wet 153
low-pressure wet 152 159
removal of 94 151
in water 172
Carbon equivalent 116
Carbon steels
in alkylation plants 44 46
corrosion allowances 174
in crude units 12
in hydrocracker air coolers 75
in hydroprocessing units 57
isocorrosion curves 61 77
naphthenic acid corrosion 17
in reforming plants 94
in regenerators 38
in steam 30
in sulfur plants 51
water corrosion 21 25 29
Carburization, catastrophic. See Metal dusting
Cast iron, water corrosion 20
Catalysts 37 55
Catalytic cracking. See Fluid catalytic cracking
Catalytic reforming 1 81
Cathodic protection 124 130 132
Caustic injection 9
Caustic service 158 159
Chlorides 83 173
Chlorination, and seawater corrosion 25
Chromate inhibitors 27
Cladding. See also Roll bond cladding
on hydroprocessing reactors 71
for restoration 43
vs. weld overlay 5
Coal tar enamel 129 132
Coatings. See also Organic coatings
for underground piping 129
Coke drums 42
Coking. See Delayed coking; Fluid coking
Condensate. See Steam and condensate
Condenser tubes, under velocity conditions 23
Continuously regenerated catalytic reforming 81
Coolers 75
Cooling water
materials selection for 30
systems 19
Copper and alloys
condenser tubes, under velocity conditions 23
water corrosion 22 26 30
Corrosion allowances 4 21 143 174
Corrosion fatigue cracking 32
Corrosion inhibitors. See Inhibitors
Corrosion rates 13 20 23 95
100 133 134
See also Isocorrosion curves
Cor-ten 84
cracking; Thermal fatigue
CRCR. See Continuously regenerated catalytic reforming
Crevice corrosion 25
Critical pitting temperature 24 25
Critical preheat temperature 117
Crude oil
atomic structure 5
corrosive constituents 6
desalting 9 14
distillation unit 8
neutralization number 7
refinery processing of 2
Crude units 5
corrosion in 8
equipment and piping 14
overhead systems 8
Cyanides, corrosive effect 46
Cyclones 39
Failures
amine absorber 96
fluid coker burner vessel 40 41
FBE. See Fusion-bonded epoxy
FCC. See Fluid catalytic cracking
Feed-effluent exchangers and piping 74
Ferrite, delta. See Delta ferrite
Fiber-reinforced plastic 27 30
Filler metals
for clad restoration 43
for dissimilar welds 18
Fittings, line pipe 114
Flare tips 85
Flue gases 84
Fluid catalytic cracking 1 37
Fluid coking 37
Fractionation equipment 46 81
Freshwater, corrosion in 20
Furfural 99
Furnace tubes 144
Fusion-bonded epoxy 129 130 132
Galvanic corrosion 25 27
Gas lines, dehydration of 124
Graphitic corrosion 22 39 171
Gunite 38
Hardfacing 16 31 40 84
Heat exchangers 5 19 74 144
Heat treatment, postweld. See Postweld
heat treatment
HIC. See Hydrogen-induced cracking
Hook cracks 110
Hydriding, of Ti tubes in seawater 26
Hydrocarbon environments,
materials selection 144 159 171
Hydrochloric acid, in crude units 10
Hydrocracking 2 37 55
Hydrodealkylation units 99
Hydrodesulfurizers 55
Hydrofluoric acid alkylation 45
Hydrogen attack 57
Hydrogen blistering 47 127
Hydrogen flaking 71 72 73
Hydrogen-induced cracking 49 80 127 194
Hydrogen partial pressure calculations 175
Hydrogen plants 2 91 92
Hydrogen sulfide
release from crudes 16
removal of 94
in sour systems 46 173
and storage tank corrosion 134
Hydroprocessing units 55
Hydrostatic testing 40 132 183 190
Hydrotreating 1
Inhibitors 10 22 27 32
96 124 128
Injection
caustic 9
of inhibitors 10
quills 10
Inspection 32 110 183 190
Intergranular stress corrosion cracking 74 85
Internal corrosion, of line pipe 124
Isocorrosion curves 61 77
Iso-pH curves 77
J factor 73
Kalrez 99
Knockout pots 96
Langelier’s Index 20
Line pipe
bending 118
brittle fracture 120
chemistry control 115
ductile fracture 121 122
electric resistance welded 106 108 187
external corrosion 128
fittings and valves 114
helical seam 112 113
internal corrosion 124
manufacture 105 181 188
materials selection 144
processing 118
seamless 111
specifications 112 114
underground 128
UOE 106 107 181
wall thickness, minimum 106
weldability 116 194
Linings. See Refractory linings
Liquid metal embrittlement 29
Low-alloy steels
in hydroprocessing reactors 71
isocorrosion curves 63
Low-temperature service 164
Magnesium chloride 10
Materials selection
alkylation plants 44
catalytic reformers 81
coke drums 42
cooling water 30
and corrosion allowances 143
criteria 141
crude units 5
flare tips 85
flue gas scrubbers 84
general guidelines 4 141
hydrogen, methanol, and ammonia plants 92
hydroprocessing units 55
line pipe 105
low-temperature services 164
reactors and regenerators 38
refineries 171
sour water strippers 46
for specific environments 143
sulfur plants 51
tanks 134
utilities 19
wear and abrasion resistance 170
MEA. See Monoethanolamine
Mercaptan 99
Mercury, and gas corrosion 126
Metal dusting 83 98
Methanol plants 91 92 98
Methyl orange alkalinity 20
Methyl tertiary butyl ether 101
Microbiologically influenced corrosion 20
Monoethanolamine 172
absorption system 91
MTBE. See Methyl tertiary butyl ether
Naphtha
defined 1
heavy 11
Naphthenic acid corrosion 7 17 146
Neutralization, in overhead systems 10
Neutralization number 7
Nickel and alloys
chemical compositions 163
for expansion bellows 40 42
for pigtails 94
Penetrators 110
pH
and carbonate SCC 42
control of 10 12
and corrosion rate of carbon steel in seawater 26
and erosion-corrosion 32
and hydriding 26
and hydrocracker corrosion 78
iso-pH curves 77
and Langelier Index 20
Reactors 38 71
Refineries
materials selection 171
processes 1
Refractory linings 38 83 84 85
94 171
Regenerator vessels 38
Residuum 1 37 84
Roll bond cladding 4
Underground piping
cathodic protection 130
coatings for 129
external corrosion 129
stress corrosion cracking 131
UOE line pipe 106 107 181
Urethanes 99
Utilities, materials selection for 20
Vacuum columns 19
Valves
in crude units 19
line pipe 114
in regenerators 40
Valve trim, materials selection 159
Vanadium pentoxide 84
Velocity, and water corrosion 23 26
Vessels. See also Regenerator vessels
materials selection 144
Visbreaker columns 19
Viton 99
dissimilar welds 18
DSAW line pipe 106 107 108 181
duplex SS welds, delayed cracking in 27
ERW line pipe 106 108 187
in hydroprocessing reactors 71
line pipe weldability 116 194
postweld heat treatment 49 71 80
weld metal, fissures in 57
Wet steam. See Steam and condensate
X factor 73
Zinc coatings 22