Production Planning and Design in The PR
Production Planning and Design in The PR
Production Planning and Design in The PR
M. ASWED
Lecturer, Faculty of Engineering, University of Tripoli, Libya
&
C. RuIz
Lecturer in Engineering (Fellow, Exeter College), Dept. of Engineering Science,
University of Oxford, Parks Road, Oxford, Great Britain
(Received: 28 April, 1974)
ABSTRACT
The effect of several Design Codes on the total cost of pressure vessels is studied. It is
shown that a significant reduction in cost can only be achieved through careful
consideration of the management of production, including the inspection procedure.
Comparative values are given to the costs involved in the various elementary operations
that constitute the complete design, fabrication, erection and commissioning for
conventional vessels.
INTRODUCTION
The pressure vessel industry is almost unique in having a number of rigid Codes
introduced and accepted in order to safeguard the integrity of the product rather
than the efficient use of the fabrication facilities. Such Codes facilitate the task of
designers and thus result in an economy of time and effort during the design stage.
On the other hand, their intrinsic lack of flexibility demands, for example, expensive
inspection procedures which sometimes cannot be substantiated by sound scientific
reasoning. Also peculiar to this situation is the existence of licensing or insuring
authorities, making radical departures from the Codes virtually impossible in many
cases and unjustified by the possible saving in materials in most. This being so, it
follows that any cost reduction must be made within the framework of the Design
Codes. The pressure vessel designer may be free to choose the Code that he con-
siders to be most suitable for a given application and to minimise the cost of the
finished product, but, once this choice has been made, no departure from the Code
205
Int. J. Pres. Ves. & Piping (2) (1974)--© Applied Science Publishers Ltd, England, 1974
Printed in Great Britain
206 M. ASWED, C. RUIZ
rules is possible. A strength evaluation, based on stress analysis and the application
of the general principles of materials science, may be necessary to supplement those
rules that are clearly inadequate but, in most cases, the purpose of such detailed
study is to ensure the safety of the vessel rather than to reduce its cost. It is in fact
doubtful whether any significant cost reduction would be achieved if vessels were
to be designed from first principles only, without the practical experience embodied
in Design Codes.
Another aspect peculiar to the industry is the fragmentation of the production
cycle. The main pressure vessel dimension and design conditions are often fixed by
engineering consultants, on the basis of specifications agreed with customers.
Detailed drawings are then prepared by fabricators and submitted for approval to
the consultants who are ultimately responsible for the whole plant. Even when
vessels are entirely shop-fabricated, a not insignificant effort will be required at site
by those responsible for the assembly of the plant. Throughout the whole process,
independent assessors are called upon to ensure the integrity of the completed
product by performing the inspections and tests specified by customers, consultants,
fabricators and erectors, often operating in different countries. Given this division
of responsibilities, the Design Code fulfils an essential unifying function between
them and the suppliers of materials. Although the volume of the pressure vessel
industry in this respect is small compared with structural and heavy engineering as
a whole, the wide range of operating conditions, e.g. temperature, pressure, fluid
contained, etc., demands a correspondingly extensive range of materials. These, in
turn, are standardised as far as possible in Design Codes.
Finally, in common with other heavy engineering activities, pressure vessels are
usually treated as unique, both in design and in construction. Details, such as
supporting brackets, columns, penetrations, reinforcements, etc. are seldom
standardised, although to do so would probably result in a saving in cost and
design effort. An exception to this remark is the use of standard pipe fittings in
some applications.
To a newcomer--whether an engineer who has had no experience of the pressure
vessel industry, or a nation whose interests have previously been confined to the
selling of raw materials--the aspects we have described are most bewildering. It is,
however, important to acquaint oneself with them if, as is the case of Libya,
technological development based on the petrochemical industry is deemed desirable,
since pressure vessels constitute an essential part of any process plant as well as
the main capital investment.
The subject of this investigation is to study the effect of Design Codes in the
industry as a whole and to identify areas where further study would result in
significant savings. In the past, most investigations have been limited to stress
analysis and strength evaluation problems, in order to assess more accurately the
safety of a vessel rather than to minimise its cost. These aspects are of interest to
specialists, be they consultants, designers or fabricators, usually working in highly
PRODUCTION PLANNING AND DESIGN 207
DESIGN
The process of design has been described by Bickell and Ruiz I who also included a
comparison between the main Design Codes in common use in 1966. 2 A more
recent comparative study between the rules set up by the various Design Codes was
published by Poynor, 12 while a detailed description of the criteria for the selection
of stresses and of the rules for the dimensioning of the pressure vessel components
is found in references 3 and 4. Closely related with the question of allowable stresses
is the characterisation of materials by means of standard specification. Tables
listing equivalent materials have been included in National standards and are
commonly used in industry, s, 9,1 o, 11,13
Tile most important decision when dimensioning a vessel for a given application
consists in the determination of the shell thickness, since this, in turn, fixes the
vessel weight which, as will be seen, is often the only consideration when assessing
the total cost. Although there are slight variations in the formulae proposed by the
various Codes, they do not differ in their approach which is to assume that the
strength of a material can be defined by means of its tensile strength, yield point or
the stress required to produce rupture or a given deformation after a certain time.
The calculated stress under the operating pressure is then maintained below a
certain proportion of the material strength. Table 1 shows the design stress criteria
TABLE 1
DESIGN STRESS FACTORS
B S 1500 -- 4.0 -- --
B S 1515 1"5 2.35 1.5 1.0
ASME viii-1 1 "6 4.0 1-67 1,0
ASME viii-2 1'5 3.0 -- --
AD Merkblatt 1-5 2.4 1"5 --
ANC (France) -- 3.0 -- --
ISD 1.4-1-5 2.4 1.5 --
208 M. ASWED, C. RUIZ
of a few of the major Design Codes, In all.cases, the design stress is the smaller of
the values obtained dividing the mechanical property considered by the
corresponding stress factor.
Figures 1 to 9 illustrate the variation of design stress with temperature for a
number of materials, in accordance with British, American and German specifica-
tions. It is clear that, from the viewpoint of cost alone, the Design Code which
allows the highest stress for a specified material and temperature is to be preferred.
In recent years the general trend has been to increase the allowable design stress,
provided that suitable care is taken in the design of details, the assessment of the
most probable service loading and of the maximum loads to be encountered in
service, backing the design when necessary by means of a stress analysis and
strength evaluation. This trend is clearly represented in the existence, side by side,
of Codes such as BS 1500 and BS 1515 in Great Britain and ASME Section viii,
Divisions 1 and 2 in the USA. A penalty to be paid when using the more advanced
Codes is that the design effort increases, accompanied by a corresponding tighten-
ing up of the quality control. Thus, by comparison with BS 1500, the times required
for the completion of a design in accordance with other Codes are as shown in
Table 2.
TABLE 2
DESIGN TIME
* Shorter time, similar to BS 1515 or ASME viii-1 depending on familiarity with Code.
t Minimum time. May increase to 3 or 4 when a detailed stress analysis becomes necessary.
t--
"""--'%H., AD Z
5 Ik
SA-285-C
Z
I 1 I I I
0 I00 200 300 400 500
Temperature (°C)
tO
O
15
Din 17155 HI BS 1515
E BS 15OH51-23B .... ASME viii-I --0--0--
I I I I I
O I00 200 :500 400 500 600
Temperature (°C)
Fig. 2. Design stress against temperature in accordance with A S M E viii Div. 1 and Div. 2,
A D Merkblatt, and BS 1515 for c a r b o n steel plates.
BS 1515
A S M E viii-I - o - o -
20 A S M E viii-2 . . . . . . .
O
4" AD
¢b
E
15 - - Din 17155 H II
Z
ASME D i v 2 S A - 5 1 5 - 5 5 --o~......
-0 . . . . • . . . . • .... o..,__.. ~,o~,~"
Bs 5oH6 -26 .
-- t , .,. ~ . . . . . . . . . . . . . . . " " ~ :.'~ •L.. A
"~ I0 ~coroon Silicon - • ' •-0~'~-~'"
t-
._O', kil led ) ASME Div. I SA-515-55 "'~-~.
(carbon silicon) ""~.~.
5
O
Z
I I I I I I
0 I00 200 300 400 500 600
Temperature (°C)
Fig. 3. Design stress against temperature in accordance with ASME viii Div. 1 and Div. 2,
AD Merkblatt, and BS 1515 for carbon silicon killed steel plates.
bo
BS 1515
20 ASME viii-I - e - e - -
4"
E BS 1501-224-50B ASME viii-2 . . . . . . .
~e~.. ~ Din 171.55Ast 45 A~,
_ _ e . . . . o~ e~...~.!, u
~15
v
ASME Otv.'2 SA - ' ' ~ e . . . . ~,".~e~_ rrf
~16-65 (corbon silicon) .~
ASME Div.I S A - 5 1 6 - 6 5 -~11~.°,,~--~'-~
._~ I0
(corbon silicon) "-...~o.~. ~.~
t~
0
1 I I I I
0 I00 200 500 400 500
Temperoture (°C)
Fig. 4. Design stress against temperature in accordance with ASME viii Div. 1 and Div. 2,
AD Merkbtatt, and BS 1515 for carbon silicon killed steel plates.
BS 1515
ASME viii-I - - o - o ~
Din 17155
20 -- 13C.Mo44 ASME viii-2 . . . . . . .
" o..... AD o
E"
E 8S 1501-151-620 "" " ' ' - c1
f5
" "r"~ O ~O~
v ASME Div. 2 S A ~ 5 8 7 - B(AN) "~.. - - - o -
v -O . . . . • " ' ~ O . . . . • ~ "'~"~"~"- • . . . . O ~ . ~ _ ~ O ~
IC.~'2Mo "~. °'"e .... o- ~- o...
~ O "-~. ~ ""- O - - - . O, ~. . ""
t--
I0 -- ASME Div. I S A - 5 8 7 - B ( o ) w......,La"~"'~'~-~"-l'~,,--.--e-eS
-- z
.m 1501-151-620 z
IC.r/2 Mo. Si ""~o (From 1501Pt 1:1970)
£3 \ Z
5
z
I f I I I e~e- !
0 I00 200 500 400 500 600 650
Temperature (°C)
Fig. 5. Design stress against temperature in accordance with ASME viii Div. 1 and Div. 2,
AD Merkblatt, and BS 1515 for CrMo steel plates.
bo
30--
BS 1515
ASME viii-I -q~-e.--
20 _ BS 1501-621 ASME viii-2 . . . . . . .
4"
E ASME Div. 2 S A - 5 8 7 - C o n o~ (No German equivalent ~r
-- I~,Cr ~'2Mo "~'e~----._e__... available ) >
Q. . . . • . . . . • . . . . • . . . . O--.---.
--,v-.. O .... • .... q~,,... "~ r~
ASME Divl S A - 3 8 7 - C (o) "o,,.. ~ " BS 1501-621
¢5
_ ll/4 Cr, P2Mo.Si • . • ~ 0, O..,~t "'
I0
E
I ] t ,. ] I t'% ~', -
0 I00 200 300 400 500 600 650
Temperature (°C)
Fig. 6. Design stress against temperature in accordance with A S M E viii Div. 1 and Div. 2,
A D Merkblatt, and BS 1515 for C r M o steel plates.
BS 1515
ASME viii-I --o-•-
20 -- ASME v i i i - 2 . . . . . . .
ASME Div 2 S A - S B T - D N o
, -. .... ._,..._ _Din 1712 IO Cr. Mo9lO AD
d"
2v4Cr I Mo ..._ "'- ~ - ~ - - " ~ " ' - - ' - ' ~ e ~ ' ~ 7 _ - . - - e . . . . o--o ......
E 15 - - BS 1501-622 ~ . e ~
5:z
.... • ..... • ".
. . . . • ..... o_.~•......._ ~•-X.• Din 17155
¢/) ASME Div. I S A - 5 8 7 - D ( o ) "~•'"'~e.~ IOCr. Mo910 z
:z
I0 -- 2a/4Cr. i MO ° • •-.......•..%_ "".--. ,,.,•
(/)
c- ~ BS 1501-622
i/)
t-~
z
O~ • O,,,,,
I I I I I 1 I
0 I00 200 :300 400 500 600 650
Temperoture (*C)
20 -- ASME viii-I - - o - - o - -
ASME viii-2 . . . . . . .
AD
A 15
Din 17440
E •- i ~ e ' " ~ - o-~.~ xlO Cr Ni N6189 >
E o_.~_~.~..--oe=._~ ' •
- --.~8~...~
- ASME Div. 2 "~--~ --~-*--~-~,#"~.---_: ~.-.~ e~. _._.....--~.-~.
SA-240-547 -8 . . . . . • ..... e__i_~.. ~ BS 1501-547 $ 4 9 ('3
S _-,,,~._
/"
tZ~
-- ASME Div. I SA 2 4 0 - 5 4 7 ~ %
SA 2 4 0 - 5 4 8 e~
J J I I I I
0 I00 200 500 400 500 600
Temperature (°C)
Fig. 8. Design stress against temperature in accordance with A S M E viii Div. 1 and Div. 2,
A D Merkblatt, and BS 1515 for austenitic chromium-nickel (niobium stabilised) steel plates
( 1 8 C r - 1 0 N i - M n - N b (8 10 × C % ) . .
BS 1515
20 - A S M E viii-I --e--e-
ASME viii-2 0
fi-
E
--Din 17440 x I0 CtNi.Ti 189 AD ¢3
-I
:Z
---~'~ 11~-7"------e--"~o~_ A S M E DIv. 2 S A - 2 4 0 - 5 2 1
-- 0~'~11~.~ • ~ • ~ •-"-'~•-- __• __..~ •~•,....,. >
"~ I0 - - 8 S 1501 :.'521 $ 4 9 "Q~-.-. , ~--o 1,...,: o~ Z
c " ~-----• ----_ .... •- - ,. _ Din 7 4 4 0 xlOCr Ni T1189
--- - - u - - . ~ II,~.-. -- - '~------~Nu---- "
•~ ' • ~ - ~ q~.__. • (quenched)
I I I I I I z
0 I00 200 300 400 500 600
Temperoture (°C)
"-.-4
218 M. ASWED, C. RUIZ
TABLE 3
RECOMMENDED MINIMUMVESSELTHICKNESS
(weakening factor) given by Bickell and Ruiz, 2 while a more complete study was
reported by Chukwujeku.S Rules only differ in essence for isolated openings and
branches while they all are virtually identical when dealing with groups of openings
in close proximity. This means that the relative reduction of the total weight, and
hence of the cost of the vessel of a Design Code demanding less compensation for
isolated openings than another, is seldom as important as it would appear at first
sight. It should be noted that the requirements of the BS 1515 and AD Merkblatt
are similar and less demanding than those of the ASME viii and BS 1500 Codes.
There are situations where the judicious departure from a Code is beneficial. For
instance, a significant saving in weight with respect to the conventional ASME viii
flanges may be achieved by using self-sealing flangeless closures, as described by
E1 Magrissy and R u i z 6 and illustrated in Fig. 10. 7 Even when designing within the
scope of the ASME rules, some freedom is available for the optimisation of a given
configuration. One must, however, recognise a natural reluctance on the part of the
users and of insurance organisations to accept radical departures from a widely
accepted Code and to demand considerable supporting evidence whenever such
departures are suggested by designers or fabricators. The costs involved can only be
estimated for each particular problem.
I0
.o
o
L~
Tapered skirt t=l.5T
J I I I i I
0 0.05 O. I O. 15
T/R
Fig. 10. Saving in weight with respect to ASME viii flanges when using self-sealing flangeless
closures. 7
I
Quotations
Fabricators,prices
and deliveries
I
Analysis of ]
quotations
I
Vessel Dept.checks Vendors submit Vendors prepare Select vendor
J drawingsand drawings for detailed and ptace order
J, returns to vendors approvaI shop drawings
Dispatch ]
vessel
~_ ~ ~-
P,
)
~repare construction 0
ssue of vessel drawing
"tl ) ('1
0
plate .~quad check o
o.
C)
~,onstruction issue of
vessel drawings
F,"
fob. drawings ffl
data book
SR
1
[~Nlteh
222 M. A S W E D , C . R U I Z
attachments require a shorter time or are purchased as finished products. The total
fabrication varies a great deal, depending on the type of vessel. Time would be
saved by using standard components but to do so requires the setting up of stocks
which would increase the capital investment. At present, most vessels are designed
and made as 'one-off' units and no serious attempt to standardise details is made.
This situation is a result of the lack of feedback of information from the fabricator
to the designer. Lacking this information, designers often have no basis for the
selection of preferred sizes of openings, mode of compensation, position, type and
dimensions of supporting brackets, ltjgs or skirts, etc., thus creating an unnecessary
variety of solutions when only a few would be sufficient.
The importance of good management in the fabrication stage is further empha-
sised by the fact that only about 30 to 40 ~ of the total fabrication time may actually
be spent in performing an operation. For the rest of the time, the vessel simply takes
up space in the workshop, waiting for the delivery of components or for an
inspection to be performed. Inspection is, in fact, a source of delay, since it is a
process that runs concurrently with design and fabrication, as shown in Fig. 14,
and it requires the presence of personnel not directly concerned with the production.
The sequence illustrated in Fig. 14 need not be the same in all practical cases; some
of the tests may be dispensed with and others may be replaced or occur at alternative
times, for instance, a pneumatic test has been shown as an alternative to a hydraulic
test and no provision has been made for instrumented pressure tests, with strain
gauges or brittle lacquers, although such tests are becoming more frequent as
vessels of complex geometry are replacing conventional designs.
TABLE 4
VARIATION OF INSPECTION TIME OR EFFORT WITH VESSEL THICKNESS,
MATERIAL AND DESIGN CODE
Reference: 1 in thick, mild steel vessel to BS 1500 (Index 1)
Parameter change Index
Thickness (2 in) 1.5 2.0
Stainless steel 1.5
High yield steel 2.0
Change of Code: ASME viii-1 1.3
BS 1515 1-5
ASME viii-2 1+8
AD Merkblatt 1.5 2-0
Mild steel vessels, made to a simple Code such as BS 1500 and having a thickness
of less than 1 in, naturally require less inspection than those vessels made to more
exacting specifications and employing more expensive materials or materials more
difficult to fabricate or weld. Table 4 gives an indication of how the inspection
time or effort varies with material, thickness and Design Code.
PRODUCTION PLANNING AND DESIGN 223
\ examination
J
.~ Exam. of tube a t t a c h m e n t ~
to tube plates
Final exam. and check of to Q Witness welder procedures and
c~imensions
dn~"-wtn~.~fim-~::~T / welderqualificationtes'ts
5 .
~ E .....f cmrculor I~Ilr Dye penetrant crack detection tests
-- \ seams including
N,,~olignment J
Exam of longitudinal seams including
alignment
j=i
224 M. A S W E D , C. R U I Z
For instance, O& could be the welding of a seam of length L, depositing a volume
V of weld metal. Two types of resource Rj could be used, either an automatic sub-
merged arc machine or an operator with a manual welding set. The time Tj would
obviously depend on the method chosen and also on the number of units available,
i.e. the number of automatic or manual sets being used, so that we may write:
rj = - rj(R> N)
where Ri indicates the type of resource and N~ the number of units of that resource
deployed. Similarly, the cost per unit time will depend on both the type of resource
and how many units are being used, i.e.
Cj = - Cj(Rj, Nj)
Finally, the total cost becomes:
n mi
In general, as the number of resources in use increases, the time will decrease and
the cost per unit time will increase, so that the total time will be reduced while the
total cost may remain unchanged. In each particular case, a detailed study will be
required since the requirement may be to minimise time or to minimise cost even
when this may imply an increase in the fabrication time. Normally the number of
stops and the type of resources are fixed for a given vessel.
No waiting periods have been considered when setting up the above equations.
These periods may be required when resources are deployed to cover a number of
vessels in order to ensure the full employment of these resources, or they may arise,
as unwanted delays, as a result of late deliveries or lack of synchronisation
between inspection, designers and fabricators. The following example provides an
indication of the time and effort required for the completion of a typical, medium-
sized vessel, 4 m diameter, 33 m long and 16 m m thick.
13-14 5 4.2
14-16 3-4 3.3
16-17 20 16-7
17-18 2-3 2.5
18-19 8 6.7
20--21 8 6"7
782 100
13-14 4 3
14-15 8 6
15-16 16 12
16-17 8 6
17-18 1 0"8
18-19 8 6
19-20 2 1'5
20-21 not applied
21-22 16 12
COST ANALYSIS
An accurate estimation of the cost is only possible after detailed drawings have
been produced and fabrication and inspection charts have been drafted. It is,
however, essential to obtain an approximate estimate at an earlier stage and this
may be done by noting that costs depend on the weight of the vessel and on its
capacity. From a total of some 50 vessels, it was found that:
Cost (£) = A x (Weight expressed in kg)
and also:
Cost (£) = B x (Capacity expressed in m3)" (4)
The best fit to the data available from vessels under construction or recently
completed was provided by taking A = 0.2 and B = 300 with n = 0.632. The
largest deviations between the data and the predicted cost from these equations
could be accounted for and were equivalent to -I-50 ~. Most of the data--about
70 ~ - - f e l l within a _+20 ~ band of the predicted results.
It is possible to modify eqns. (3) and (4) to take into consideration changes in
prices by noting that:
P R O D U C T I O N P L A N N I N G A N D DESIGN 227
CONCLUSIONS
TABLE 5
COST BREAKDOWN OF CARBON STEEL VESSEL DEPENDING ON DESIGN CODE
The total fabrication costs, including the cost of materials, fabrication, direct
labour, inspection and overheads account for about 70 ~ of the total cost. Design,
on the other hand, varies between 20 ~o of the total for the traditional BS 1500 Code
to 32 ~ for the more sophisticated. ASME viii-Div. 2. The figures quoted for design
costs following the remaining Codes may be reduced through familiarity in their
228 M. ASWED, C. RUIZ
use to a value of about 20 9/0 of the total. This means that the difference in the total
cost of vessels designed and made in accordance with any of the Codes listed, with
the exception of the ASME viii-Div. 2, would be negligible.
The cost breakdown shown in Table 5 may be extended to include the effect of
material and vessel thickness. The use of a low or high alloy steel instead of the
conventional carbon steel implies an increase in cost not only due to the higher
cost per unit mass of the material itself, but also due to the fact that the fabrication
is usually more difficult, welding and forming require greater care and hence add up
to the total cost. On the other hand, higher design stresses may be possible, resulting
in thinner, higher vessels. It can be shown that similar conclusions are valid.
ACKNOWLEDGEMENTS
The information contained in this paper has been collected from a large number of
sources, including most of the major pressure vessel-manufacturers and designers
in Great Britain. It is not possible to list all the persons and organisations whose
help is gratefully acknowledged, but the writers wish to express their thanks in
particular to Mr K. Birch (Ralph Parsons Co. Ltd), Mr G. Lorraine and Mr. J.
Round (Clarke-Chapman), Mr H. Butler (International Combustion Ltd), Mr J.
Bullock (Robert Jenkins & Co. Ltd), Mr M. J. Kemper and Mr Turnbult (APV
Co. Ltd), Mr R. Thompson (Power-Gas Ltd), Mr H. Heward (British Engine),
Mr Sergio Menicatti (Snam Progetti), Mr J. Russell (Motherwell Bridge Ltd),
Mr R. D. Kerr (Babcock & Wilcox Ltd), Mr P. Bramhili (British Steel Corporation)
and Mr S. Nicholson (Whessoe Ltd).
The support provided by the National Oil Company of Libya and the University
of Tripoli is gratefully acknowledged.
REFERENCES
1. BICKELL,M. B. and RuIz, C. Recent trends in pressure vessel design, 2nd Commonwealth
Welding Conf. London, I. Welding, 66, pp. 294-313.
2. BICKELL,M. B. and RuIz, C. Pressure vessel design and analysis, Macmillan. 1967.
3. BS-PD 6437, ,4 review o f design methods given in present standards and codes and design
proposals ]br nozzles and openings in the pressure vessels, BS. PD-UDC 66. 023: 621. 642-225.
4. CmPPERFIELD,W. R. et aL Design criteria of boilers and pressure vessels, Proc. 1st Inter.
Conf. PV Technology, ASME/KIVI, Delft, 1969.
5. CHUKWUJEKWU,S. E., Reinforcement method for flush nozzles in pressure vessels. Loc. cit.
(Ref. 4).
6. EL MAGRISSY, I. and Rulz, C. Flangeless closure joints: experimental validation of an
unconventional design, Proc. 2nd Int. Conf. PV Technology, ASME, San Antonio, USA,
1973.
7. EL MAGRISSY,I. and RuIz, C. Flangeless closure joints for pressure vessels, Chemical Process
Engineering (November, 1971).
8. ESSO. Permissible substitutes for USA specification materials. ESSO Basic Practice
(September, 1966).
PRODUCTION PLANNING AND DESIGN 229