947

Pressure Vess
11. Pressure Vessels and Heat Exchangers

Part B 11
Ajay Mathur

11.1.8 External Local Loads ..................... 951

This chapter is intended to present an overview of
11.1.9 Fatigue Analysis ........................... 951
Pressure Vessels/Heat Exchangers and covers basic
design concepts, Loadings & testing requirements 11.2 Design of Tall Towers ............................ 952
relevant to these equipment. Design criteria, 11.2.1 Combination of Design Loads......... 952
fabrication, testing & certification requirement 11.2.2 Wind-Induced Deflection .............. 952
of various Standards/Codes adopted in different 11.2.3 Wind-Induced Vibrations .............. 952
countries are discussed on a comparitive basis to
11.3 Testing Requirement ............................ 953
bring out similarities of features.
11.3.1 Nondestructive Testing (NDT).......... 953
In order to complete the overview, a brief
11.3.2 Destructive Testing of Welds .......... 953
discussion is provided on commonly used Materials
of construction and their welding practises along 11.4 Design Codes for Pressure Vessels ........... 954
with updates on the on-going developments in 11.4.1 ASME Boiler and Pressure Vessel
this area. Code ........................................... 954
The author is a Mechanical engineering grad- 11.4.2 PED Directive and Harmonized
uatefrom M.S University-Baroda (India) & has over Standard EN 13445 ........................ 954
20 years experience in design and fabrication of 11.4.3 PD 5500 ...................................... 956
Pressure Vessels, exchangers, Skid mounted plants 11.4.4 AD Merkblätter ............................. 958
and Fired Heater modules for Refinery, petrochem- 11.5 Heat Exchangers................................... 958
ical Nuclear & chemical plants in India & abroad.
11.6 Material of Construction ........................ 959
11.6.1 Carbon Steel ................................ 959
11.1 Pressure Vessel – General Design Concepts 947 11.6.2 Low-Alloy Steel ............................ 960
11.1.1 Thin-Shell Pressure Vessel ............. 947 11.6.3 NACE standards ............................ 960
11.1.2 Thick-Walled Pressure Vessel ......... 949 11.6.4 Comparative Standards for Steel..... 960
11.1.3 Heads ......................................... 950 11.6.5 Stainless Steel.............................. 960
11.1.4 Conical Heads .............................. 950 11.6.6 Ferritic and Martensitic Steels ........ 964
11.1.5 Nozzles ....................................... 950 11.6.7 Copper and Nickel Base Alloys ........ 964
11.1.6 Flanges ....................................... 950
11.1.7 Loadings ..................................... 951 References .................................................. 966

11.1 Pressure Vessel – General Design Concepts

Pressure vessels are closed structures, commonly in the 11.1.1 Thin-Shell Pressure Vessel
form of spheres, cylinders, cones, ellipsoids, toroids
and/or their combinations and which contain liquid or When the thickness of the vessel is less than about
gases under pressure. There are various other require- one tenth of its mean radius, the vessel is called
ments such as end closures, openings for inlet/outlet a thin-walled vessel and the associated stresses resulting
pipes, internal/external attachments for support and from the contained pressure stress are called membrane
other accessories. stresses. The membrane stresses are assumed to be uni-
948 Part B PARTMARK
Part B 11.1

σΧ & σΦ

σΧ

σΦ
P

Fig. 11.1 Cylindrical shell with hemispherical heads

form across the vessel wall and act tangential to the will be decided based on hoop stress, which is gov-
surface. erning since its value is double than that of meridional
stress.
Vessels Under Internal Pressure In the case of a spherical shell, the circumferential
The membrane stresses due to internal pressure are de- stress is identical in all directions and its magnitude is
rived from application of equilibrium conditions to an half the hoop stress of a cylinder of the same radius.
appropriate element cut from the vessel shell, which Because the thin shell is assumed to offer no resis-
in turn has been considered as a symmetrically loaded tance to bending, radial stresses, if present, are small
shell of revolution. Figure 11.1 shows the resulting compared to membrane stresses and are generally ig-
membrane stress for a thin-walled cylindrical shell, both nored. This implies that the thin-shell analysis considers
in the axial axis (called the meridional stress) and cir- only biaxial stresses and therefore follows the maxi-
cumferential direction (also called the hoop stress), and mum stress theory, wherein stress failure depends on the
the corresponding formulae are tabulated in Table 11.1. numerical magnitude of the principal stresses ignoring
It can be seen that the thickness of the cylindrical vessel stresses in other directions.

Table 11.1 General vessel formula

Part Stress formula Thickness d
Inside radius Outside radius
Shell
PRm PRi PR0
Longitudinal direction σx = 2d 2SE+0.4P 2SE+1.4P
PRm PRi PR0
Circumferential stress σΦ = d SE−0.6P SE+0.4P
Heads
PRm PRi PR0
Hemisphere σ x = σΦ = 2d 2SE−0.2P 2SE+0.8P
longitudinal stress
=circumflex stress
PDi K PD0 K
Ellipsoidal 2SE−0.2P 2SE+2P(K −0.1)
PDi PD0
2:1 semiellipsoidal 2SE−0.2P 2SE+1.8P
0.885PL i 0.885L 0
100%–6% torispherical SE−0.1P SE+0.8P
PL i M PL 0 M
Torispherical L/r < 16.66 2SE−0.2P 2SE+P(M−0.2)
Cone
PRm PDi PD0
Longitudinal σx = 2d cos α 4 cos α(SE+0.4 p) 4 cos α(SE+1.4P)
PRm PDi PD0
Circumferential σΦ = 2d cos α 2 cos α(SE−0.6P) 2 cos α(SE+0.4P)
 Pressure Vessels and Heat Exchangers 11.1 Pressure Vessel – General Design Concepts 949

Vessels Under External Pressure

Part B 11.1
σΦ for single σΦ resultant
Thin-walled vessels under external pressure fail at
cylinder
stresses much lower than the yield strength due to
the instability of the shell. In addition to the physical
properties of the material of construction at the oper- σΦ due to PS
ating temperature, the principal factors governing the
instability and the critical (collapsing) pressure Pc are P1
geometrical, namely the unsupported shell length L, the A B
shell thickness t and the outside diameter Do . The so- σΦ due to PS
lution to the theoretical elastic formula for the critical
pressure at which the cylinders would collapse under
external pressure depends on the number of lobes at
collapse, which is a cumbersome exercise. To elimi- Fig. 11.3 Resultant stresses in a shrink-fitted multiwall ves-
nate the dependency on the number of lobes, American sel
Society of Mechanical Engineers (ASME) codes have
adopted a simplified procedure using geometric charts for a vessel for internal and external pressure conditions
developed for the critical stress ratio v/s L/Do for dif- are shown in Fig. 11.2a,b.
ferent values of the Do /t ratios. This critical stress ratio From Fig. 11.2, it is evident that the maximum cir-
in turn is related to the collapsing ratio for a particu- cumferential stress occurs at the inside surface of the
lar material in another chart, thus an estimate for the cylinder; however it is tensile in nature for internal pres-
maximum allowable pressure may be determined for sure conditions and compressive in the case of external
a given geometry. Several trials can be made to achieve pressure. As per Lame’s solution,
an optimum solution for shell thickness, and maximum  
PRi2 Ro2
unsupported length (or stiffener spacing) for a given σΦ = (+) 2 1+ 2 , (11.1)
external pressure. Ro − Ri2 Ri
 
PR2 R2
11.1.2 Thick-Walled Pressure Vessel σr = (−) 2 i 2 1 − o2 , (11.2)
Ro − Ri Ri
The membrane stresses used for evaluation of thin shells the magnitude of stress for a thick-walled vessel is
cannot be used for thick-walled vessels (with a thick- a function of the ratio of the outer to the inner radii.
ness greater than their mean radius) subjected to internal Prestressing is a manufacturing technique by which
pressure, since the radial stresses are significant and the compressive stresses are produced in the inside shell
stress distribution in the vessel walls varies across the of a multilayer/multiwall vessel and/or tensile stresses
thickness. The stress distribution across the thickness are created on the outside shell of this vessel operating

a) σΦ (Tensile) b)

P2

P1

σr (Compressive)
σr (Compressive) σΦ (Compressive)

Fig. 11.2a,b Stress distribution for a thick shell: (a) internal pressure, (b) external pressure
950 Part B PARTMARK

under internal pressure. The prestresses so created help The formulae for computing the thickness of dif-
Part B 11.1

to neutralize the stress peaks in the existing stress dis- ferent types of heads under both internal and external
tribution, making the material stress almost uniformly pressure are provided in Table 11.1.
throughout the thickness, as shown in Fig. 11.3. This There is a failure phenomenon in the knuckle region
also considerably reduces the required wall thickness. due to tangential stress under compression, which can
Prestressing is carried out by the following methods: occur through elastic buckling (circumferential wrin-
kles) at a stress much lower than the yield strength
• shrink-fitting outer layers over the core, inner
or through plastic buckling. A reliable analysis for
layer(s)
predicting buckling failure has been introduced in the
• wire/coil wrapping
harmonized pressure vessel standard EN 13445, which
• autofrettage
will be discussed later.
A detailed introduction to the basic theory of
membrane stress and its application to commonly en- 11.1.5 Nozzles
countered elements of pressures vessels is presented
in [11.1]. However, a brief discussion on other ves- Nozzles or openings are provided in pressure vessels to
sels components such as heads, nozzles and flanges is satisfy certain process requirements such as inlet or out-
covered below. let connections, manholes, hand holes, vents and drains
etc. These may be located on the shell or head accord-
11.1.3 Heads ing to the functional requirement and could be circular,
elliptical or rectangular in shape. The basic construc-
There are three general categories of heads (also called tion of a nozzle connection consists of essentially short
dished ends): pieces of pipes welded to the vessel wall at an open-
Hemispherical: These dished ends are analyzed as the ing made in the wall. The other side would normally be
thin-walled spherical shells discussed a flanged end suitable for connection to the correspond-
earlier. ing piping or to bolt on the blind cover (as in the case of
Ellipsoidal: Ellipsoidal heads are developed by the a manhole). The complete nozzle may also be formed
rotation of a semi-ellipse and have by rolling or forging to the required shape.
a 2:1 ratio of major R to minor axis In addition to weakening of the vessel wall, the noz-
h. These heads are the most fre- zle opening also causes discontinuity in the wall and
quently used end closures in vessel creates stress concentration at the edges of the open-
design, particularly for internal pres- ing. This is compensated by providing reinforcement
sures greater than 10 bars and also for pads around the nozzle necks, which are suitably at-
bottom heads of tall, slender columns. tached to the vessel wall. Rules are provided in every
Torispherical: Torispherical heads have a meridian pressure vessel code to calculate the reinforcement re-
formed of two circular arcs, a knuckle quirement for all nozzles. At times, additional thickness
section with radius r, and a spherical is provided at the base of the nozzle wall itself; such
crown segment with a crown radius of nozzles are called self-reinforced nozzles.
L. The maximum crown radius equals
the inside diameter, which gives the 11.1.6 Flanges
same maximum membrane stress in
the crown region as in the cylindrical Tall columns are usually constructed in detachable sec-
region. The minimum knuckle radius tions for ease of fabrication, transportation, erection,
is 6% of the crown radius, although assembly, and internal maintenance. Like the nozzles
a 10% knuckle is the most commonly and the piping system, these sections (or heads) must
used. be provided with end flanges with an arrangement for
easy bolting and dismantling as required. A flanged
11.1.4 Conical Heads joint therefore consists of a pair of flange; each is at-
tached to one of the components to be joined and is
A conical head is generated by the rotation of a straight held securely in place by a series of bolt or studs. A gas-
line intersecting the axis of rotation at an angle, α, ket is interposed between the two adjoining flange face.
which is the half-apex angle of the formed cone. The joint must have structural integrity with (zero) min-
 Pressure Vessels and Heat Exchangers 11.1 Pressure Vessel – General Design Concepts 951

imum leakage during service. Several configurations of stresses within the prescribed limits. These loads are

Part B 11.1
flanges in various construction materials and a large va- usually significant at nozzles, the vessel support region,
riety of gaskets are available; the selection is mainly at brackets, lifting lugs, and saddle supports for hori-
dependent upon the service requirements. zontal vessels. Since the contact area of the attachment
The main consideration for the design of a flanged is relatively small compared to the vessel area, a sim-
joint are: plified form of the interface force distribution between
the vessel and the support is assumed. The analysis is
1. to ensure a positive contact pressure at the gasket
based on elastic stress analysis and stress categorization
flange interface to prevent leakage in service. The
is used to compare the resultant stresses.
gasket must be able to withstand the required seal-
This approach is used in annexure G of the PD 5500
ing force,
code and in the Welding Research Council (WRC) bul-
2. the gasket sealing force is to be provided by bolt
letins WRC 107 and 297 used while in the design of the
tightening without overstressing,
ASME code. In 1991 the WRC published another bul-
3. to ensure the structural integrity of flange sections
letin, WRC 368, for the evaluation of shell and nozzle
and minimize flange deflections.
stresses due to internal pressure.
The analytical solution for evaluating localized
11.1.7 Loadings stresses in the shell wall above the saddle supports for
a horizontal vessel is based on the method developed by
Loadings or forces are the causes of stress in pressure
Zick and published in 1951.
vessels. It is important to identify areas where and when
The method used in the harmonized standard
these forces are applied to pressure vessels. The stresses
EN 13445-3 (clause 16) for evaluating nozzle loads is
produced by these loads, which could be general or lo-
based on limit load analysis. This standard provides sep-
cal are additive and define the overall state of stress in
arate rules for calculating line loads for lifting eyes,
the vessel or its component. The combined stresses are
supporting brackets, and saddle supports.
then compared to the allowable stress defined by the
pressure vessel code.
11.1.9 Fatigue Analysis
An outline of the various categories and types of
loadings is summarized in Tables 11.2, 3, respectively.
Vessels undergoing cyclic service and repeated load-
11.1.8 External Local Loads ing may fail in fatigue due to progressive fracture of
localized regions. The behavior of metals under fa-
Stresses caused by external local loads at the points of tigue conditions varies significantly from the normal
attachment to the vessel must be assessed to keep these stress–strain relationship. Damage accumulates dur-

Table 11.2 Categories of loadings

General loads Local loads

General loads are applied more or less continuously Local loads are due to reactions from supports, inter-
across a vessel section nals, attached piping and equipment such as mixers,
platforms etc.

Examples Examples
• Pressure loads – internal or external pressure (de- • Radial load – inwards or outwards
sign, operating, hydrotest and hydrostatic head of • Shear load – longitudinal or circumferential
liquid) • Torsional load
• Moment loads – due to wind, seismic, erection and • Tangential load
transportation • Moment load – longitudinal or circumferential
• Compressive/tensile loads – due to dead weight, in- • Thermal loads
stalled equipment, ladder, platform, insulation, piping
and vessel contents
• Thermal loads – skirt head attachment
952 Part B PARTMARK

ing each cycle of loading and develops at localized structive testing, and in-service inspection of vessels
Part B 11.2

regions of high stress until subsequent repetitions fi- designed for fatigue.
nally result in visible cracks, which grow, join, and All pressure vessel codes have established specific
spread. criteria for determining when a vessel must be designed
Localized stresses at abrupt changes in sections such for fatigue. Each code has adopted a methodology for
as at the head junction or a nozzle opening, misalign- carrying out fatigue analysis based on the use of fatigue
ment, defects during fabrication, and thermal gradients curves obtained from test specimens. The fatigue design
are probable causes for fatigue failure. Special attention rules of some of the pressure vessel codes are discussed
should be paid to manufacturing tolerances, nonde- later.

11.2 Design of Tall Towers

11.2.1 Combination of Design Loads • National building code (NBC)
• British standard (BS) 6399
The shell thickness of tall columns as computed based
on internal or external pressure is not usually sufficient Code for Seismic Loads
to withstand the combined stresses produced by the op-
• ASCE-7
erating pressure plus weight, and wind or seismic loads.
• UBC/NBC
Combined stresses in the longitudinal direction σL due
• International building code (IBC) 2000
to pressure P, dead weight W and applied moment M,
• Response spectrum data
with W and M taken at the elevation under considera-
tion, are calculated as follows: Tall cylindrical vessels are normally designed to be
self-supporting; they are supported on cylindrical or
1. On the windward side conical skirts with base rings resting on concrete foun-
      dations, firmly fixed to the foundations by anchor bolts.
PD 4M W
σL = + − , (11.3) Detailed analytical methods for computing the thick-
4d π D2 d π Dd
ness of the skirt components and sizing of the anchor
 
(PD/4) + (4M/π D2 ) − (W/π D) bolts can be found in [11.2, 3], which also provide pro-
d= . cedures for other types of supports such as lugs, legs,
allowable stress
(11.4) and rings due to wind and seismic loads.

2. On the leeward side 11.2.2 Wind-Induced Deflection

     
PD 4M W
σL = − − . (11.5) A sustained wind pressure will cause tall columns to
4d π D2 d π Dd deflect with the wind. Most engineering specifications
limit the deflection to 150 mm per 30 m of column
For the design of a particular vessel, the value of the height. The vessel is assumed to be a cantilever beam
moment derived from either the wind or seismic loads firmly fixed to the concrete pedestal and individual
is used in these expressions. This is due to the assump- deflections induced by wind load and moments are cal-
tion that the wind and seismic loads are not expected to culated for each varying section of the vessel using
occur simultaneously and therefore the higher moment the deflection formula for cantilever beams. The total
of the two is considered to be governing. These loads deflection is then calculated using the method of super-
are based on site-specific data, which is obtained from position.
one of the following national standards as applicable to
the installation site. 11.2.3 Wind-Induced Vibrations

Code for Wind Loads Wind-induced vibrations can be caused by vortex shed-
• American Society of Civil Engineers (ASCE) 7, ding, the magnitude of which is dependent on wind
formerly American National Standards Institute velocity and vessel diameter. Vortex shedding results
(ANSI) A58.1 in severe oscillations, excessive deflections, structural
• Uniform building code (UBC) damage and even failure. When the natural frequency
 Pressure Vessels and Heat Exchangers 11.3 Testing Requirement 953

Table 11.3 Types of loadings

Part B 11.3
Steady loads Non-Steady loads
These loads are long term and continuous These loads are short term and variable

Examples Examples
• Internal/external pressure • Shop and field hydrotest
• Dead weight • Earthquake
• Vessel contents • Erection
• Loading due to attached piping and equipment • Transportation
• Loading to and from vessel supports • Upset, emergency
• Thermal loads • Thermal loads
• Wind loads • Startup, shutdown

of a column or stack coincides with the frequency of investigated dynamically for vibrations. Detailed meth-
vortex shedding, the amplitude of vibrations is greatly ods for determination the need for dynamic analysis
magnified. After a vessel has been designed statically, and a method for carrying out this dynamic analysis are
it is necessary to determine if the vessel needs to be provided in [11.3].

11.3 Testing Requirement

11.3.1 Nondestructive Testing (NDT) a penetrant that becomes florescent under ultraviolet
light.
NDT of the raw material, components, and finished ves- Magnetic particle examination, which can only be
sel is important from the safety point of view. The most used on magnetic material, is carried out by passing
widely used methods of examination for plates, forg- magnetic flux with the help of a probe through the part
ings, castings and welds are briefly described below. to be tested. Fine magnetic particles, which are dusted
Radiographic examination is done either by X-rays over the surface, tend to concentrate at the edge of the
or gamma rays. The former has greater penetrating crack. To pick up all the cracks, the area is probed in
power, but the later is more portable. Single-wall or two directions.
double-wall techniques are used for tubular compo-
11.3.2 Destructive Testing of Welds
nents. Penetrameters, or image quality indicators, check
the sensitivity of a radiographic technique to ensure that In contrast to the NDT methods, which are essentially
any defect will be visible. predictive techniques, the mechanical integrity of welds
Ultrasonic techniques use vibrations with a fre- is checked by testing sample test plates called prod-
quency between 0.5 and 20 MHz transmitted to the uction test coupons. The coupons are welded along with
metal by a transducer. The instrument sends out a se- the actual vessel joint (usually a longitudinal seam) and
ries of pulses, which are seen on a cathode ray screen thus are representative of the actual welding techniques
after being reflected from the other end of the member. employed for the vessel. The following tests are nor-
Reflections either from a crack or inclusion in the metal mally carried out on the test piece:
(or weld) can be detected on the screen; based on the
magnitude and position of the signal, the location of the 1. Tensile testing, which includes transverse tensile
flaw can be ascertained. and all-weld tensile tests
Liquid penetrant examination involves wetting the 2. Bend test – transverse bend/side bending
surface with a colored fluid that penetrates open 3. Macro-etching, hardness, impact testing
cracks. After wiping out excess fluid, the surface is 4. Intergranular corrosion test (IGC) for austenitic
coated with a developer fluid, which reveals the liquid stainless-steel material/welds
that has penetrated the cracks. Another system uses 5. Ferrite checking
954 Part B PARTMARK

11.4 Design Codes for Pressure Vessels

Part B 11.4

In modern competitive industry, new process plants are of stringent material testing and more-careful quality
being set up rapidly and existing units are continually control. In addition to detailed design rules division 2
revamped, modernized and refurbished for the chem- provides for discontinuities, fatigue, and other stress
ical, petrochemical, pharmaceutical, energy, refinery, analysis considerations, which are based on maximum
and nuclear industries all over the globe. A variety shear stress theory. Division 2 also contains rules for
of equipment is needed for the storage, handling and vessels with layered construction (multiwalled vessels)
processing of hydrocarbons and chemicals in these
processing plants. Unfired pressure vessels such as stor- Fatigue Analysis
age tanks, Horton spheres, mounded bullets, reactors, Clause AD-160 of division 2 provides several meth-
autoclaves, distillation/fractionating columns, and heat ods for exempting fatigue evaluations. If the cyclic
exchangers are some of the basic components of any operation does not meet all the conditions of AD-160
such process plant. a fatigue evaluation method as per appendix 5 or 6 is
Various codes specifying the requirements for the added. The stress ranges are first determined for the
design, material, fabrication, inspection, and testing specified cyclic operation and then, using fatigue curves
of pressure vessels have been written and adopted as presented in appendix 5, the associated number of al-
national standards in various countries. Most the in- lowable cycles and Miner’s rule are used to determine
ternational pressure vessels have been developed with the life fraction and cumulative damage. Fatigue curves
a higher degree of technical similarity between them. up to 370 ◦ C for carbon/low-alloy steel and 430 ◦ C for
Core area such as vessel class, design criteria and austenitic stainless steel are provided in the code.
requirements for independent inspection and certifica- Division 3, which is a comparatively recent publica-
tion are based on similar (but not identical) guiding tion, provides a state-of-the art code and is intended for
principles. high-pressure applications where fatigue and fracture
A compilation of websites of various organizations, dominate.
associations, technical standards, and current European
Union (EU) legislation relating to pressure equipment 11.4.2 PED Directive and Harmonized
sector is presented in [11.4]. Standard EN 13445
The subsequent subsections present a brief discus-
sion of the key features of some of the international The early phase of development of design codes and
pressure vessel codes. associated legislation in the pressure equipment sector
was done predominantly at the national level. In view of
11.4.1 ASME Boiler and Pressure Vessel Code the substantial economic potential of this sector, a need
was felt in the European community to introduce a uni-
The ASME boiler and pressure vessel code, sec- form and harmonized regulatory framework within the
tion VIII, published by the American Society of Me- EU. The objective of the common legislation mecha-
chanical Engineers, also known as ASME International, nism promulgated through directives was to minimize,
is a widely accepted code in the USA and 80 other if not eliminate, trade barriers between EU member
countries in the world. The code is written by voluntary states for pressure equipment, at the same time meet-
engineering talent and is constantly upgraded by the ing essential safety requirements stipulated by the new
corresponding committee members to include the latest directives.
developments in material and design methodology. Of the several European directives enforced
ASME section VIII is written against a well-defined since 1987, the Pressure Equipment Directive (PED)
theoretical background and is divided into three subdi- (97/23/EC) and the Simple Pressure Vessel (SPV) Di-
visions: 1, 2 and 3. The important design rules of both rective (87/404/EEC) are the two principal instruments
the codes are summarized in the respective appendices for the pressure equipment sector. The approach of
of each code. The contents of the three divisions are or- the directives includes the identification of the prod-
ganized to cover specific pressure ranges, as illustrated uct, prescription of the essential safety requirements
in Table 11.4. (ESRs) to be met by technical standards, demonstra-
As can be seen from the code comparison table, di- tion of conformity and CE marking. Technical standards
vision 2 permits higher working stresses at the expense adopted by the European Committee for Standardiza-
 Pressure Vessels and Heat Exchangers 11.4 Design Codes for Pressure Vessels 955

Table 11.4 Comparison of various divisions of ASME codes vis-à-vis EN13445

Part B 11.4
Section VIII division I Section VIII division II Section VIII division III EN 13445 part 3
First publication < 1940 1968 1997 2002
Units adopted řF and ksi ◦C and N/mm2
Pressure limits Upto 3000 psig No limits, No limits, For gas-ps up to 3000 bar and/or
usually + 600 psig usually from 10 000 psig ps∗ V up to 3000 bar L
For liquid-ps up to 1000 bar
and/or ps∗ V up to 10 000 bar L
Design factor
Tensile strength 3.5 3.0 Ferritic/normalized steel – 2.4
Austenitic steel – 3
Yield strength 1.5 1.5 Yield based with reduc- Ferritic/normalized steel – 1.5
tion factor for yield to Austenitic steel – 1.5 or 1.2
tensile ratio less than 0.7 (depending on rupture elonga-
tion)
Average stress 1.0 1.0 Guidelines for design in creep
for 1% creep in range are under development
100 000 h
Allowable stress Committee and provided in Tables in section II C for individual Designer
calculated by grades and products
Testing groups NDE requirement de- Stringent NDE require- NDE requirement is even Classified from I to IV in de-
pendent on weld effi- ment more stringent than Divi- creasing extent of NDT
ciency factor sion 2
Hydrostatic test 1.3 × design pressure 1.25 × design pressure 1.25 × design pressure Greater of 1.43 × allow-
able pressure and 1.25 ×
temperature-adjusted pressure

Table 11.5 EN 13445:2002 Unfired pressure vessels – a quick reference

Reference Title Contents/special features
EN 13445-1:2002 General Scope, extent of testing with respect to weld joint coefficient, material grouping,
etc.
EN 13445-2:2002 Materials Materials listed include steels with sufficient ductility, cast iron and aluminium.
Permitted or tabulated design stress are not provided.
List of normative annexures are provided which include technical, inspection and
delivery condition.
Material listed in CE harmonized product standard can be used for that product.
Material listed in CE harmonized material standard, if appropriate to the product.
European approval of material (EAM) required for materials not listed in the har-
monized standards.
Particular material appraisal (PMA) required for material not listed in the harmo-
nized standards or approved via EAM.
EN 13445-3:2002 Design See Table 11.6 listing features for individual components.
EN 13445-4:2002 Fabrication Provides weld designs, tolerances, production testing, post-weld heat treatment and
repair requirements.
EN 13445-5:2002 Inspection and testing Specifies nondestructive testing, pressure testing, marking and documentation re-
quirement for noncyclic operation and special provisions for cyclic operations.

EN 13445-6:2002 Requirement for design and fabrication of pressure vessels and pressure parts constructed of spheroidal
graphite cast iron.
CR 13445-7:2002 Supporting standard for guidance on the use of the conformity procedures.
prEN 13445-8 Additional requirement for pressure vessels of aluminium and aluminium alloys.
956 Part B PARTMARK

Table 11.6 A quick summary of EN 13445 part 3

Part B 11.4

Type of analysis Component Special features

Design by formulae Dished ends Additional formulae for knuckle and knuckle buckling
for non-cyclic load-
ing Cones and conical ends Based on limit analysis
(used in clauses 7 to
Opening in shells Pressure area method
16)
(full pressure cycles Opening in flat ends Replacement of section modulus
less than 500 cycles) Flanges Modified Taylor forge method, alternate method also provided based on
limit load analysis carrying out detailed assessment of flange–bolt–gasket
system, useful for joints where bolt loads are monitored
Weld joint efficiency Linked to testing group I/II/III/IV
Tubesheets In addition to traditional method, new method using limit analysis approach
in which edge loads and restraints are addressed, is provided
Allowable tube loads are calculated
Bellows Covers both thin- and thick-walled bellows
Rules are based on Expansion Joint Manufacturer’s Association (EJMA)
standards
Rectangular vessels Covers both unreinforced and reinforced vessels where outside stiffeners
are attached
Non-pressure local loads Local loads are assessed by comparison with allowable loads calculated
based on limit load analysis
Design by formulae Simplified fatigue analysis Fatigue strength calculated using fatigue curves, which are more conserva-
for cyclic loading (clause 17) tive than those used in detailed fatigue analysis for the unwelded region
(used in clauses 17 Table is provided for selecting stress factor for various design configurations
and 18) Detailed fatigue analysis Correction factors depending upon temperature, thickness, mean stress and
(full pressure cy- (clause 18) surface finish are applied on fatigue curves
cles exceeding 500 Cumulative damage calculated as per Miner’s rule
cycles) Guidance provided on bending stress calculations due to misalignment of
weld and recommendation given for weld-toe grinding
Design by analysis Direct route New route based on Eurocode (for steel structure), which overcomes the
(covered by annexes (annex B) shortcomings of the familiar stress categorization method, addresses failure
B and C) modes directly by way of design checks under the influence of actions (all
imposed thermo-mechanical quantities including pressure, thermal and en-
vironmental) using partial safety factor depending on nature of single action
or combination of actions
Stress categorization route Based on categorization of stresses into primary, secondary and peak stress
(annex C) and comparing the same with specified limits, Stress classification table is
provided for given component and its location depending upon the source
load to help list and allocate stress category

tion (CEN/CENELEC), which provide means for the Table 11.5, while Table 11.6 summarizes the essential
user to comply with the ESRs of the PED, are called features of the design rules for various pressure vessel
harmonized standards. components, as covered in part 3 of EN 13445. An ex-
EN 13445:2002 Unfired Pressure Vessel is a major cellent presentation of the background to the rules of
harmonized product standard within the CEN pressure EN 13445 part 3 is available [11.5].
equipment portfolio. The standard utilizes expertise and
best practices from across the European member states 11.4.3 PD 5500
as well other internationally accepted standards. The
adoption of the first issue of this standard, published PD 5500:2000, the specification for unfired fusion
in 2002, was preceded by discussions between experts welded pressure vessels, is a recent replacement for
who took nearly 10 years to achieve a major technical BS 5500:1997. It was issued under the status of a pub-
convergence. lished document (PD) in anticipation of simultaneous
The various sections of EN 13445 along with a brief release of the harmonized standard EN 13445 for pres-
description of the respective contents are tabulated in sure equipment in Europe.
 Pressure Vessels and Heat Exchangers 11.4 Design Codes for Pressure Vessels 957

Table 11.7 Summary of the salient features of ASME, PD 5500 and the AD code

Part B 11.4
ASME section VIII division 1 PD 5500 AD Merkblätter
Information Part Summary Section Summary Section Summary
Responsibilities UG-99 Responsibilities listed 1.4 Responsibilities for code Druckbehälter Authorized inspector to
for manufacturer and au- compliance is on manu- VO issue final vessel certifi-
thorized inspector (AI) facturer cation
Certification UG-115 Manufacturer to have 1.4 Code compliance is doc- Druckbehälter Authorized inspector to
to 120 certificate of authoriza- umented by use of form VO issue final vessel certifi-
tion to construct ASME X issued by manufac- cation
stamped vessel turer and counter signed
by AI
Construction – – 3.4 Three categories with HP 0 Four testing groups I, II,
categories different material and III and IV are defined
NDT requirement are
defined
Joint types UW-3 Defines A, B, C, D 5.6.4 Defines A and B welded
category with different joints, with different
NDT requirements; UW- NDT requirements
12 gives joint efficien-
cies
Weld joint de- UW-12, Shows typical weld E.1(1)– Shows typical weld
tails UW-13, joints for guidance and E.1(6) joints for guidance and
UW-16 are not mandatory are not mandatory
Welder’s UW-28, Weld procedure specifi- 5.2, 5.3 WPS, PQR and welder’s HP 2/1 WPS, PQR and welder’s
approval 29, 48 cation (WPS), qualification are qualification are
procedure qualification required required
record (PQR) and
welder’s qualification
are required
Permissible UG-4, UG-4 refers to materials 2.1.2 This sections references W0 to W13 W0 to W13 cover
materials UG-10 specified in ASME sec- British standard mater- for metal- all types of al-
tion II ials lic and N loyed/unalloyed steel,
series for castings, forgings, clad
non-metallic steel bolts and nuts but
material do not include gaskets
Material iden- UG-94 Material for all pressure 4.1.2 Positive material identi-
tification parts is marked and cer- fication is required for
tified for traceability all pressure parts
NDT UW-51 NDT techniques are 5.6.4 Ultrasonic testing (UT) HP 5/3 DIN standards are refer-
techniques to 53, detailed in ASME and radiographic testing enced for all NDT tech-
UW-11, section V (RT) testing both are ac- niques
UW-42 ceptable
Assembly tol- UG-80, Tolerances for circular- 4.2.3 Tolerances for circular-
erance UW-33 ity and alignment are ity and alignment
specified are specified
Pressure UG-99, Test pressures are spec- 5.8.1 Test pressures are spec- HP 8/2, HP 30
testing UW-50 ified along with other re- ified along with other re-
quirements of testing quirements of testing
Heat treatment UCS-56 HP 7/1 to HP
7/4

Though PD 5500 does not currently have formal sta- Fatigue Analysis
tus as a harmonized standard and compliance with its Annexure C of the code presents criteria for exemp-
technical requirement does not qualify presumption of tion from fatigue analysis. A simplified fatigue analysis
conformity to PED, it does include information showing using design curves can be done using conservative es-
how its technical content would comply with the ESR of timates of the stress range due to pressure changes and
the directive. thermal gradients. By using appropriate an design curve
958 Part B PARTMARK

to obtain allowable cycles and satisfying a cumulative requirement for equipment, design, manufacture and
Part B 11.5

damage rule, a simplified analysis can be done. testing, and materials. These regulations are compiled
A detailed analysis is required if the specified crite- by seven trade associations of Germany, who together
rion is not satisfied. Detailed methods for determining form the AD. The AD associations represent a balanced
stresses due to pressure, thermal gradient, and pip- combination of material and pressure vessel manufac-
ing loads based on using stress concentration factors turer’s, operators, employer’s liability insurance, and
are provided in annexure G. Based on the individual technical inspectorates.
stresses thus obtained, the maximum principal stress To a large extent the AD Merkblätter code is based
range for each individual cycle is determined and fa- on Deutsches Institut für Normung (DIN) standards
tigue evaluation is done as per annexure C. and is continuously amended in keeping with technical
The fatigue curves are limited to 350 ◦ C for ferritic progress. The associations has published a regulation
steel and 430 ◦ C for austenitic stainless steels. called AD 2000 conforming to the safety requirement
and other stipulations as laid down in the PED, which
11.4.4 AD Merkblätter has been made compulsory in Europe from May 2002
onwards.
Arbeitgemeinschaft Druckbehälter (AD) Merkblätter A quick summary showing the key technical points
regulations are a set of generally accepted rules of tech- of PD 5500 in comparison with the ASME section and
nology regarding pressure vessels and contain safety AD Merkblätter code are shown in Table 11.7.

11.5 Heat Exchangers

Heat exchangers are devices that transfer heat from plate exchangers are the most commonly used types of
a hot to a cold fluid. They are used extensively in pro- surface exchangers.
cessing plants and are given specific names when they The construction of shell-and-tube exchangers is
serve a special purpose, for example superheaters, con- broadly divided into the shell side and tube side (also
densers, evaporators etc. In a surface-type exchanger, called the channel section). Several small-diameter
the two process fluids are separated by a physical bar- tubes (on the tube side) are attached to larger pressure
rier and the heat is transferred from the warm/hot fluid vessels (known as the shell side), or parts thereof, called
through the barrier to the cold fluid. Shell-and-tube and tube sheets. The tubes are distributed within the tube
Table 11.8 Design rules for tube sheets as per different codes
TEMA ASME section VIII division 1 EN 13445-3
U tube tubesheet
Reference section Section 1999 edition Appendix AA-1 edition 2001 and Clause 13 D-2002 (based on
UHX-12 edition 2002 (manda- Code francais de construction des
tory) Appareils a Pression
(CODAP)/unifired pressure
vessel (UPV) rules)
Assumptions Perforated tubesheet and unperfo- Refined and rational analytical treatment is used after taking into ac-
rated rim not accounted for; the count actual geometry (based on model proposed by F. Osweiller in
effect of tube sheet attachment 2000)
with shell/channel not considered
Ligament efficiency 0.45 ≤ η ≤ 0.60 0.25 ≤ μ ≤ 0.35
Safety factor 2.6 1.5 2.0
Allowable stress Stress classification of division 2 Based on primary and secondary
appendix 4 stress as per appendix C
Remarks TEMA approach leads to lower Higher thickness obtained in ASME rules than UPV/CODAP due to
thickness than ASME due to lower allowable stress
higher ligament efficiency and
high safety factor
 Pressure Vessels and Heat Exchangers 11.6 Material of Construction 959

Table 11.8 (cont.)

Part B 11.6
Fixed/floating tubesheet
Reference section Appendix AA-2 edition 1992 Clause 13 E-2002 (based on
CODAP/UPV rules)
Assumption Stiffening effect of tube bun- Coefficient F is dependent on the Coefficient F is dependent on the
dle and weakening effect of tube stiffness ratio X stiffness ratio X; the value of F
holes are assumed to counterbal- is higher than ASME since tubes
ance each other, coefficient F is are assumed to be uniformly dis-
not dependent on the stiffness ra- tributed over the whole tube sheet
tio X of the axial tube bundle
rigidity to the bending rigidity of
the tube sheet
Remarks TEMA does not provide the same
design margin for all cases, lead-
ing to over-thickness for higher X
and under-thickness for lower X

sheet in a certain pattern, the three most common of Various configurations and designs of shell-and-
which are the triangular, square and rotated triangular, tube exchangers are covered extensively by the Tubular
and square. Exchanger Manufacturers’ Association (TEMA).
Designs for the joint between the tube and tube The principle of mechanical design for most of the
sheet vary widely and are chosen to be compatible with components of the heat exchanger is identical to the de-
the severity of the service conditions. The joint may sign of a pressure vessel. However, the design of the
be expanded, welded, or a combination of both. There tube sheet is typically different because of its construc-
are various constructional details of welded joints, the tional geometry. The tube sheet design rules have been
choice of which is based upon service and environ- rationalized in recent years by various pressure vessel
ment. codes, which are summarized in the Table 11.8.

11.6 Material of Construction

11.6.1 Carbon Steel to the top and is removed; the resultant steel is called
killed steel. A fully killed steel usually contains less
Steel is an iron alloyed with carbon at a level of than 150 ppm oxygen and at least 0.10% silicon. Be-
0.05–2.0%. In addition it contains smaller proportions sides being cleaner due to the formation of fewer oxides
of phosphorus, sulphur, silicon, aluminum, and man- and inclusions, fully killed steels are much easier to
ganese. These steels are known as plain carbon steels, weld due to a lower tendency to outgas during weld-
which are classified as mild-, medium- and high-carbon ing.
steels according to the percentage of carbon.
Mild steel has 0.05–0.3% carbon by weight, Welding
medium-carbon steel has a carbon content of 0.3– Mild-steel electrodes are grouped into those with
0.6% and high-carbon steel has more than 0.5%, up rutile-type flux covering and those with low hydro-
to a maximum of 2%, carbon content. Mild steels are gen flux covering. Rutile-covered electrodes are used
the most versatile materials for the construction of pres- for general fabrication involving thinner sections, lower
sure vessel due to their good ductility and relative ease tensile strength, and in applications where there is no
of forming, rolling, forging, fabrication, and welding. requirement for impact properties. For all other appli-
They are also most suited and economic for applica- cations, where strength, impact properties, and weld
tions where the rate of corrosion is low. During the quality are essential, low-hydrogen-type electrodes are
manufacture of these steels, silicon and/or aluminum employed. The flux covering eliminates sources of po-
are added to react with dissolved oxygen in the molten tential hydrogen and thus minimizes the chances of clod
metal alloy to form a slag of Al2 O3 /SiO2 , which floats cracking.
960 Part B PARTMARK

11.6.2 Low-Alloy Steel 11.6.3 NACE standards

Part B 11.6

Low-alloy steel contains additions of the elements Ni, The National Association of Corrosion Engineers
Cr, Si, Mo, and Mn in amounts totaling less than 5%. (NACE) is a worldwide technical organization that stud-
The added elements improve mechanical properties, ies various aspects of corrosion in refineries, chemical
heat treatment response and /or corrosion resistance. plants and other industrial systems.
The weld ability of low-alloy steels is also good; how- NACE standard MR0175, entitled Sulfide corro-
ever since these steels are hardenable by heat treatment, sion cracking resistant metallic material for oilfield
they do require careful attention to welding procedure equipment, is widely used for applications in sour gas
including pre- and post-weld heat treatment (PWHT) and oil environments. NACE compliance is recom-
for stress relief, which is discussed later. mended in systems where there is a likelihood of sulfide
The workhorse alloys for pressure vessels, ex- cracking due to the presence of a measurable amount
changer and heater tubes, and piping for elevated of H2 S. Since the susceptibility of carbon, low-alloy
temperature service, usually greater than 250 ◦ C, con- and austenitic stainless steel to sulfide corrosion is di-
tain 0.5%–9.0% chromium plus molybdenum. With rectly related to strength and hardness level, it these
the increasing chromium content, resistance to high- standards recommend that the hardness of the afore-
temperature hydrogen attack, and resistance to sulfida- mentioned plate material should be restricted to 22 HRC
tion and oxidation increases. (200 BHN). The cold working of these steels during
forming/rolling shall be less than 5%. Post-weld heat
Developments treatment is to be carried out for carbon steels in the
There has been an increasing trend towards improved case of greater cold working. A few duplex stainless and
toughness properties and temper embrittlement resis- some nickel-based alloys are also acceptable according
tance by restricting levels of impurity elements at the to the NACE criteria, subject to a maximum hardness
ladle. level of 28 and 35 HRC, respectively.
To minimize the temper embrittlement of low-alloy
steel, the phosphorous content is restricted to 0.010% Welding
or lower, while the combined phosphorous–tin content All weld procedures must be qualified to meet the same
is limited to 0.010%. hardness levels standards as specified for the corre-
The effects of the tramp elements have been ad- sponding parent material.
dressed by the Bruscato factors X and J, as defined by
the following equations: 11.6.4 Comparative Standards for Steel

(10P + 4Sn + 5Sb + As) The discussion for carbon steel material and the dis-
X= cussion to follow for other construction materials are
100
generally based on the generic composition of mater-
(all elements in ppm) ,
ials without referring to any code of construction. It is
J = (Si + Mn)(P + Sn) × 104 (elements in %) . difficult to furnish equivalence of any grade of material
from one code to another; at best steels grades can be
For this reason PWHT is a must for all creep- compared based on the closest matching technical re-
resisting Cr–Mo-type alloy steels. PWHT also stabilizes quirements. Tables 11.9, 10 list some of the comparable
and softens the microstructure at the heat-affected zone standards for flat and tubular products commonly used
(HAZ) of the weld. If the base metal is quenched and in fabrication of pressure vessel.
tempered, higher PWHT temperatures can be specified
for improved resistance to creep embrittlement. 11.6.5 Stainless Steel
Fabrication Alloys of iron and carbon with over 12% chromium,
In view of the criticality of preheating, post-heating, which resist rusting under most atmospheric conditions,
and PWHT, it is prudent to employ specialized heat- are called stainless steels. The alloys become more cor-
treatment techniques using electrical resistance pads rosion resistant as the chromium level increases. As
or induction coils,especially for thick piping sec- more nickel is added, several phases are possible and
tions. the alloy may undergo transformation depending on the
 Pressure Vessels and Heat Exchangers 11.6 Material of Construction 961

Table 11.9 Comparative chart for material standards of flat products (plates) for pressurized use

Part B 11.6
Harmonized European standard ASTM DIN BS 1501
Description Standard Grade
General requirement EN 10028-1 SA 20
DIN 17155
Non-alloy and alloy EN 10028-2 P235GH SA 283 Gr C/ HI 161 Gr 360/
steel with elevated-
temperature proper- SA 516 Gr 55 164 Gr 360
ties
P265GH SA 516 Gr 60 H II 161 Gr 400/
164 Gr 400
P295GH SA 516 Gr 65 17Mn4 224 Gr 490
P355GH SA 414 Gr G 19Mn6
16Mo3 SA 204 Gr B 15Mo3 1503-243 B
13CrMo4-5 SA 387 Gr 12 13CrMo4 4 620 Gr 27
10CrMo9-10 SA 387 Gr 22 10CrMo9 10 620 Gr 31
DIN 17102
Weldable fine-grain, EN 10028-3 P275N StE 285 224 Gr 400
normalized
P275NH SA 516 Gr 60 WStE 285 224 Gr 430
P275NL1 SA 662 Gr A TStE 285
P275NL2 EStE 285
P355N SA 537 CL 1 StE 355 224 Gr 490
P355NH SA 662 Gr C WStE 355 224 Gr 490
P355NL1 SA 737 Gr B TStE 355 224 Gr 490
P355NL2 EStE 355
P460N StE 460
P460NH WStE 460
P460NL1 TStE 460
P460NL2 EStE 460
DIN 17280
Nickel alloy EN 10028-4 11MnNi5-3 13MnNi6
steel with low-
temperature 12Ni14 SA 203 Gr D,E,F 10Ni14
properties
X12Ni5 (12Ni19) SA 645 12Ni19
X8Ni9 NT SA 353
X8Ni9 QT SA 553
Weldable fine-grain, EN 10028-5 P355M / ML1/ML2
thermomechani-
cally rolled P420M / ML1/ML2
P460M / ML1/ML2
Weldable fine grain, EN 10028-6 P690Q / QH QL1/QL2 SA 517, SA 533, SA724
quenched and tem-
pered
Stainless steel EN 10028-7 Various gardes SA 240 DIN 17440 BS 1449-2
(refer to Table 11.10) DIN 17441 and BS 1501-3
Note: Common legend for Tables 11.9, 10 is shown under Table 11.10
962 Part B PARTMARK

Table 11.10 Comparative chart for material standards of tubular products for pressurized use
Part B 11.6

Harmonized European standard ASTM DIN BS

Description Standard Grade
Non-alloy EN 10216-1 P195 SA 53 Gr A/SA 106 Gr A BS 3059-1 Gr 320
for general
use P235 SA 53 Gr B/SA 106 Gr B DIN 1629 St 37 BS 3601 Gr 360
P265 DIN 1629 St 44 BS 3601 Gr 430
DIN 17175 BS 3059-2 BS 3602 BS 3606
Carbon/ C−Mn Exchan-
alloy ger tube
Unalloyed EN 10216-2 P195GH SA 179 – Cold-drawn tubes Gr 320
and alloyed for exchanger
for elevated SA 192 – Boiler tubes
temperature
P235GH St 35.8 360 360 400
P265GH SA 210 Gr A1 – Boiler/ St 45.8 440 430 440
superheater tube
SA 210 Gr C 17Mn4
Tube Pipe
16Mo3 SA 209 T1 SA 335 P1 15Mo3 243 243
SA 213 T2 SA 335 P2 BS 3604
(ferritic)
10CrMo5-5 SA 213 T11 SA 335 P11 621 621
13CrMo4-5 SA 213 T12 SA 335 P12 13CrMo44 620 620
SA 213 T21 SA 335 P21
10CrMo9-10 SA 213 T22 SA 335 P22 10CrMo9 10 622 622
X11CrMo5 SA 213 T5 SA 335 P5 625 625
X11CrMo9-1 SA 213 T9 SA 335 P9 12CrMo195 620-470 620-470
X10CrMoVNb9-1 SA 213 T91 SA 335 P91 X10CrMoVNb91 91
DIN 17179
Unalloyed EN 10216-3 P275NL1/NL2 WStE/TStE 285
and alloyed,
fine grain P355N/NH EStE/StE 355
P355NL1/NL2 WStE/TStE 355
P460N/NH EStE/StE 460
P460NL1/NL2 WStE/TStE 460
ASTM 333 ASTM 334 DIN 17173 BS 3603 carbon/alloy
Unalloyed EN 10216-4 P215 NL Gr 1 Gr 1 TTSt35 N
and alloyed,
low P265 NL Gr 6 Gr 6 430 LT
temperature
12Ni14 Gr 3 Gr 3 10Ni14 530 LT
12Ni14+ QT Gr 3 Gr 3 10Ni14 530 LT
X10Ni9 Gr 8 Gr 8 X8Ni9 509 LT
X10Ni9 + QT Gr 8 Gr 8 X8Ni9 509 LT
Legend for steel grade to EN series
G – Other characteristics follows, N – Normalized condition, H – Elevated temperature property required, M – Thermo mechanically rolled,
QT – Quench and tempered, L1 – Low temperature property, impact testing at −50 ◦ C, L2 – Special low temperature property, impact testing
at −50 ◦ C, enhanced requirement
 Pressure Vessels and Heat Exchangers 11.6 Material of Construction 963

actual content of Cr–Ni–Fe and C. Gamma or austenitic creep strength is also necessary, casting alloys such as

Part B 11.6
stainless steel is an iron alloy containing at least 18% HK-40 and IN657 have been developed.
Cr and 8% Ni with carbon up to 0.10%. The austenitic Duplex stainless steels were developed to com-
structure provides a combination of excellent corrosion bine the attractive properties of ferritic and austenitic
resistance, oxidation, and sulfidation resistance with stainless steels. In simple terms, the ferrite could be
high creep resistance, toughness and strength at temper- said to provide mechanical strength and stress corro-
atures up to 550 ◦ C. The 18Cr–8Ni stainless alloys form sion resistance, while the austenite provides ductility
a series known as the American Iron and Steel Institute and the two combine to produce a fine-grained, two-
(AISI) type 300 series, with varying amounts of carbon, phase microstructure with high strength and good
molybdenum, and titanium added. corrosion resistance. Of the many alloying elements
The 18–8 series have good formability besides be- Cr and Mo enhance the formation of ferrite, while
ing readily weldable without stress relief; however Ni and N stabilize the austenite. Resistance to pitting
they can be hardened by cold working. These steels and crevice corrosion in chloride environments is in-
are susceptible to grain-boundary chromium carbide creased, expressed by the pitting resistance equivalent
precipitation, known as sensitization, when subjected PREN = %Cr + 3.3 × %Mo + 16 × %N.
to a temperature range of 535–800 ◦ C. To prevent This number is used to rank materials according to
sensitization, low-carbon grades (C < 0.03%) and sta- their expected resistance to pitting corrosion. A 23% Cr
bilized grades with added columbium or titanium are Mo-free grade would have a PREN value of about 25.
used. Regular Mo-alloyed duplex grades have a PREN value
The traditional grades of stainless steel that are ex- of 30–36. Steels having a PREN value higher than 40
tensively used in the fabrication of chemical plant and are normally defined as super-duplex stainless steels.
equipment are AISI 304 type and the 2% molybdenum-
bearing 316 types along with their low-carbon versions Welding
304L and 316L or the stabilized grades 321 (with Ti) or Duplex stainless-steel welds with matching compos-
347 (with Cb) where intercrystalline corrosion is to be ition filler material show high ferrite levels, which has
avoided. low toughness and poor ductility. Therefore all welding
Higher-alloyed grades such as AISI 309 and 310 consumables for duplex materials are over-alloyed with
have a higher chromium content that makes them suit- nickel, which allows more austenite to form so that the
able for high-temperature applications such as furnace ferrite level in welds is lowered and the welds have good
liners, preheaters and column trays. ductility and corrosion resistance. It is recommended
to achieve a ferrite content of approximately 22–70%;
Developments equivalent to 30–100 FN (ferrite number). In addition to
Newer grades of stainless steel have been developed the ferrite count, a corrosion test in ferric chloride is also
to overcome the limitations of low 0.2% proof stress, carried out as per ASTM G48, which gives a good as-
sensitivity to stress and pitting corrosion especially in sessment of the corrosion resistance of the weld metal.
chloride media, inadequate corrosion in reducing me- As per this test, the critical pitting temperature (CPT) is
dia, and preferential attack on the ferrite phase in strong specified at 22 ◦ C for duplex and 35 ◦ C for super-duplex
oxidizing media. steel.
Nitrogen alloyed steel such as AISI 304LN and Preheating of duplex material is not required except
316LN have been developed with the addition of 0.2% where heavy loads on high-ferrite-containing welds
nitrogen, resulting in improved proof stress by about may cause cracking. Post-welding solution annealing
15%. Nitrogen alloyed steel finds wide application in is required only in cases where the resultant weldment
the transportation and storage industries. has deteriorated due to detrimental phase transforma-
For reactors, strippers, and condensers in urea ser- tion and/or has high ferrite levels.
vice steels with higher chromium and nickel contents Root passes of nitrogen-alloyed stainless steel are
with nil ferrite have been developed. Modified compo- welded with higher-alloyed filler, to compensate for the
sitions of AISI 310 steel such as SANDVIK 2RE69 and influence of nitrogen.
Assab 725 LN have been developed for strong oxidizing Urea-grade stainless steels and steels in high-
conditions in fertilizer plants. temperature applications are welded with matching
For elevated-temperature applications in furnaces, composition electrodes enriched with 4–5% manganese
and hydrocarbon and steam reformers, where higher to counteract the tendency for microfissuring.
964 Part B PARTMARK

11.6.6 Ferritic and Martensitic Steels Maintaining the preheat after welding (also called
Part B 11.6

post-heating) for a specified period (generally 300–

Ferritic steels are chromium–iron stainless steel with lit- 350 ◦ C for 30 min) in some cases (say for pipe
tle or no nickel and form a body centric structure unlike thicknesses greater than 12 mm with a chromium con-
the face-centered austenitic steel. When ferritic steels tent of 2–7%) also helps to reduce hydrogen levels,
are modified by heat treatment, they become hardened thereby preventing cold cracks and stress corrosion
and form martensitic steels. cracking.
Martensitic steels derive their excellent hardness Martensitic steels can be welded but caution needs
from the high levels of carbon added to their alloy. to be exercised as they will produce a very hard and
The most commonly used martensitic steel is ASTM brittle zone adjacent to the weld. Cracking in this zone
type 410 stainless steel used for column tray and tower can occur (particularly in thicker sections) and therefore
lining in crude service for refinery applications. The preheating and PWHT is recommended.
increased carbon level in 410 steel results in a much Though ferritic steels are less prone to cracking
harder martensitic cutlery steel or tool steel type 420. due to their lower strength and non-hardenability, the
By increasing the percentage of chromium, transfor- weldment suffers from excessive grain growth, sensiti-
mation hardening is suppressed, as in ASTM types 430 zation and a lack of ductility. Due to the excessive grain
and 446, which are essentially ferritic. These steels growth problem, only thin-gauge sheets are generally
are resistant to chloride stress corrosion cracking; used. Filler material can be of either a similar composi-
however they are subject to ductile–brittle tempera- tion, or alternatively an austenitic grade can be used to
ture embrittlement, thereby restricting their mechanical help weld toughness and increase ductility.
properties. Table 11.11 lists some of the comparable standards
The limiting values for X and J factor are usually for stainless-steel grades commonly used in the fabrica-
specified for welding consumables, although it would tion of pressure vessel.
be preferable also to restrict these for the base material.
A step-cooling simulation treatment is performed Developments
on higher-thickness quench and tempered plates to de- Since the ferritic grades do not possess good welding
termine susceptibility to embrittlement phenomenon in properties, hybrid utility ferritics such as 3CR12 with
terms of meeting the specified shift in the 40 ft − lb controlled martensite (dual phase steel) have been de-
charpy V-notch (CVN) transition temperature. veloped to overcome these welding difficulties. Using
new steel-refining techniques, along with the addition of
Welding titanium or niobium, it has been possible to develop ex-
With increasing base and weld material strength and tremely corrosion-resistant grades such as superferritic
hardenability, hydrogen diffusibility in the weldment stainless steel.
is kept below 5 ml/100 g of deposited weld metal. The
flux covering of all electrodes is of the low-hydrogen 11.6.7 Copper and Nickel Base Alloys
type and employs binders that give high resistance to
moisture absorption. Prior to use, they are dried or Brass
baked at the manufacturer’s recommended tempera- Brasses are commercially produced with varying per-
ture. centages of copper and zinc to provide a range of
Although there are several theories for determining properties depending on the end-use requirements.
the optimum preheat temperature, the common indus- Admiralty brass, which is widely used for tubes in
trial practice is to use carbon equivalent as guidance to water-cooled condensers for low water speeds, is an al-
select the temperature, as shown here loy brass containing 71% Cu, 28% Zn, and 1% Sn. To
prevent dezincification, small amounts of arsenic, phos-
CE Temperature ◦ C phorous or antimony are added.
< 0.4 50 At high water speeds and when seawater contains air
bubbles, aluminum brass containing 2% Al is more suit-
< 0.55 100
able due to the formation of a protective film. For tube
< 0.70 150 plates in condensers and exchangers, it is usual prac-
< 0.8 200 tice to use high-zinc brasses, hot-rolled Muntz metal
< 0.9 250 (60% Cu, 40% Zn) or Naval brass (60% Cu, 39% Zn,
 Pressure Vessels and Heat Exchangers 11.6 Material of Construction 965

Table 11.11 Comparative chart for various stainless-steel grades

Part B 11.6
Structure Hardenability ASTM 240 EN 10028-7
Grade UNS No. Grade Number Analysis built up from
basic type
Austenitic Hardenable by 304 S 30400 X5CrNi18-10 1.4301 Cr 18% + Ni 8% basic
cold work type
304L S 30403 X2CrNi18-10 1.4303 304 with low carbon
308 S 30800 X2CrNiMo8-14-3 1.4432 Higher Cr and Ni for
more corrosion and scal-
ing resistance
309 S 30900 X2CrNiCuWN25-7-4 1.4501 Still higher Cr and Ni
310 S 31000 X5CrNi25-21 1.4335 Highest Cr and Ni (Cr
25% + Ni 20%)
316 S 31600 X5CrNiMo17-11-2 1.4401 Mo added for corrosion
resistance
316L S 31603 X5CrNiMo17-12-2 1.4404 316 with low carbon
316N S 31603 X2CrNiMoN17-11-2 1.4406 316 with nitrogen added
for low-temp. service
321 S 32100 X6CrNiTi18-10 1.4541 Ti added to avoid carbide
precipitation
347 S 34700 X6CrNiNb18-10 1.4550 Cb added to avoid carbide
precipitation
Martenstitic Hardenable by 410 S 41000 X12Cr13 1.4006 Cr 12% basic type
heat treatment
420 S 42000 X46Cr13 1.4034 Higher C, cutlery applica-
tion
431 S 43100 X4CrNiMo16-5-1 1.4418 Higher Cr and Ni added
for improved ductility
Ferritic Non- 405 S 40500 X6CrAl13 1.4402 Al added to Cr 12% to
hardenable prevent hardening
430 S 43000 X6Cr17 1.4016 Cr 17% basic type
442 S 44200 Higher Cr to resist oxi-
dation and sulfidation at
446 S 44600 higher temperature

Precipitation- Age- 17-7 PH S 17700 X7CrNiAl17-7 1.4568

hardening hardenable
steel
14-8MoPH S 13800 X8CrNiMoAl15-7-2 1.4532

and 1% Sn) to take advantage of higher tensile strength, Cu–Ni Alloys

although the two-phase structure of these alloys cannot Alloys of copper and nickel have historically been used
be satisfactorily inhibited against dezincification. in saltwater condensers as they show better resistance
to saltwater than brasses. Increasing nickel content was
Bronze found to be beneficial and 30% Ni alloy was adopted as
Bronze is a tin alloy of copper with other elements such the standard for naval vessels. The addition of iron and
as aluminum added for additional properties. Because Mn was found to improve resistance to impingement
of the hardening effect of tin, hot-rolled bronze plates attack.
have greater strength than brass plates and therefore These alloys are used as tubes for heat exchang-
can be used for tube plates and channel material for ers, saltwater pipelines, and hydraulic lines as well
exchangers. as for several applications in marine and offshore
966 Part B PARTMARK

platform services. Cupronickel tubes are superior to Welding

Part B 11

brass in terms of better resistance to dezincification Gas tungsten arc (GTAW), gas metal arc (GMAW), and
for applications involving higher metal temperature shielded metal arc (SMAW) processes are most com-
of water-cooled exchangers. They are also excellent monly used for welding brasses and bronze. Whereas
materials for tube plates and, because of their good thin gauges are welded with GTAW using zinc-free
formability and weldability, they can be used in sheet fillers, heavier gauges are joined by a GMAW/SMAW
form for the fabrication of heat-exchanger shells and process using zinc-free silicon bronze or aluminum
water boxes. Monel, a nickel-copper alloy (67% Ni, bronze fillers/electrodes. Zinc-free fillers are prescribed
30% Cu) has good resistance to saltwater, and to hy- since the evolution of zinc fumes makes the weld
drochloric and hydrofluoric acid under nonoxidizing porous and affects visual observation of the welding
conditions. Therefore they are excellent material for process; moreover zinc fumes are extremely hazardous
cladding and trays in towers handling acid vapors. to health. Argon and helium, either individually or in
Other Nickel-based alloys are classified as: combination, are used for shielding in the case of the
GTAW/GMAW process.
1. Chromium bearing as in Inconel 600, and Hastel-
Welding consumables for welding of all Cu–Ni-
loy C-22 and C-276;
based alloys are available with compositions matching
2. Containing chromium and molybdenum such as In-
the specific parent material with the generous addition
conel 625, Hastelloy B, Incoly 825;
of manganese and/or niobium, which are added inten-
3. Precipitation hardening alloys such as Monel K-
tionally to give resistance to hot cracking and to raise
500, and Inconel 817.
hot strength. Most of these consumables are often used
These alloys show excellent resistance to pitting, for dissimilar metal welding between the nickel base
stress corrosion cracking in chloride environments and most steels or between other ferrous and nonferrous
and retain strength at elevated temperatures. There- alloys.
fore they are excellent candidates for exchanger tubes, Carbon and silicon are controlled to low levels to
heat-transfer plates for plate heat exchangers (PHE), minimize detrimental precipitates in the weld metal and
and pressure coils for steam/hydrocarbon reformers, HAZ for electrodes with specifications matching with
naphtha-cracking furnaces etc. in the petrochemical and some of the high-molybdenum alloys such as Hastel-
fertilizer industries. loy C276 and Hastelloy B.

References

11.1 J.F. Harvey: Theory and Design of Pressure Vessels 11.4 C. Matthews: Engineer’s Guide to Pressure Equip-
(Van Nostrand Reinhold, Amsterdam 1985) ment – The Pocket Reference (Professional Engi-
11.2 D.R. Moss: Pressure Vessel Design Manual (Gulf, neering, Suffolk 2001)
Houston 1987) 11.5 G. Baylac, D. Koplewicz (Eds.): EN 13445 Unfired
11.3 H.H. Bednar: Pressure Vessel Design Handbook (Van Pressure Vessels – Background to the Rules in Part
Nostrand Reinhold, Amsterdam 1986) 3 Design (UNM, Paris 2002), (Issue 2 download from
www.unm.fr)