Analysis of lines to API-650 Nozzles

Stress analysis of lines connected
to API-650 tank Nozzles
1
In this presentation, we will cover :
• Why the stress analysis of lines connected to
Tank Nozzle is different than stress analysis of
lines connected to Equipment nozzle ?
• How Caesar modelling should be done to

account for this ?
• What are the acceptance criteria for nozzle

load & pipe stress?
2
We start with :

account for this ?

load & pipe stress?
3
Stress analysis of lines connected to
Equipment nozzle has to account in
Caesar Modeling
• Nozzle’s thermal movements, and

• Nozzle flexibility
4
Tank nozzle has to account in Caesar
Modeling
• Nozzle’s thermal movements

• Nozzle rotations due to tank bulging
• Tank settlement
5
Nozzle rotations due to
tank bulging
6
Nozzle rotations due to
tank bulging
.
What is this Tank

Bulging?
7
In “Tank Bulging”, we will cover
• What is Tank Bulging?
• What are its effect on nozzle?
• Why this effect is not there in equipment?
• How this is calculated?
• Sample calculation for shell movements at various
heights due to tank bulging, to see tank profile after
bulging.
• Why it is not calculated for nozzles at higher level?
• Pipe routing guidelines - to minimize effect of tank
bulging. 8
we start with
bulging.
bulging. 9
TANK
Tank is filled with liquid.
This liquid has varying height.
Due to this there is varying liquid

pressure on tank wall.
It has more pressure at bottom.
10
TANK
Due this, tank wall try to expand

more at bottom.
But the bottom plate prevent this

expansion and holds bottom end
of shell in position.
11
TANK BULGING
Due to this, actual shape of tank

is formed like this.
12
Now we will see
bulging.
bulging. 13
TANK BULGING
and its effect on nozzle
Due to this, nozzle on the shell

move radially outward,
and rotates in vertical plane,

depending upon there position.
The nozzle on lower portion

rotates downwards. Whereas
nozzle on upper portion rotates
upwards.
14
Now we will see
bulging.
bulging. 15
This effect is not seen in equipments
• Equipment diameter is small (up to 3m)
Therefore the amount of radial growth is insignificant
16
This effect is not seen in equipments
Therefore the amount of radial growth is insignificant
• Tank diameter is large (10 m to 60 m)
Therefore the amount of radial growth is significant
17
• Equipment pressure is not varying so much
with liquid height
Equipment
has internal
pressure,
not only
pressure
due to fluid
weight.
18
• Equipment bottom is not flat to act like
stiffener to hold shell ends
19
Now we will see
bulging.
bulging. 20
API 650
• Appendix – P : address tank bulging.
• Mandatory for tank with diameters above
36m.
• For tank with dia. 36 m & below : Optional , or
mandatory only if specified by purchaser.
• Petrofac practice : apply for all tanks.
21
API 650 : Appendix – P
Formula for outward radial growth of the shell due to
tank bulging :
W = 9.8x10-6 GHR2x[1-e-βL cos(βL)-L/H]/Et
Where,
G= Design Specific Gravity of Liquid
H = Liquid Height (mm)
R = Nominal Tank Radius (mm)
t = Shell Thickness of Opening (mm)
β = Characteristic Parameter, 1.285/(Rt)0.5 (1/mm)
E = Modulus of Elasticity (MPa)
L = Vertical distance of Opening Centreline
from Tank Bottom (mm)
22
Formula for rotation of the shell due to tank bulging :
θ = 9.8x10-6 GHR2x[1/H-βe-βL[cos(βL)+sin(βL)]/Et
Where,
G= Design Specific Gravity of Liquid
H = Liquid Height (mm)
R = Nominal Tank Radius (mm)
t = Shell Thickness of Opening (mm)
β = Characteristic Parameter, 1.285/(Rt)0.5 (1/mm)
E = Modulus of Elasticity (MPa)
L = Vertical distance of Opening Centreline
from Tank Bottom (mm)
23
Now we will see
bulging.
bulging. 24
A Sample calculation as per
with,
G= Design Specific Gravity of Liquid = 1.0
H = Liquid Height = 10,000 mm
R = Nominal Tank Radius = 11250 mm
t = Shell Thickness of Opening = 16 mm
For varying heights.
25
26
TANK
27
API 650
• Appendix – P :
• Mandatory for tank with diameters above
36m
and
• Applicable / valid for nozzles on lower half of
bottom shell course.
Because, formula does not address varying shell thickness of
tank, which is common practice for tank.
28
Now we will see
bulging.
bulging. 29
Nozzle rotations & movement
due to tank bulging,
for nozzles at higher level.
We do not calculate this for nozzles at higher

level.
30
Why it is not calculated for nozzles at higher level?
Reasons :
For nozzles at higher level,

• Effect of tank bulging on nozzle is insignificant,
31
Maximum Height
radial above this
shell peak is
movement 9.29 m.
is at 710
mm from Which is
bottom. 13 times
(9.29/0.71
= 13) of
0.710 m.
Thus slope in
top portions
would be 13
times less than
slope in bottom
portion.
32
Max.
Slope is
0.654
degree at
bottom.
Whereas
in upper
portion it And there
is 0.051 is
degree at constant
1 m height slope of
(which is 0.022
12.8 times degree
less w.r.t. from
0.654), height 3m
onwards
(which is
30 times
less w.r.t.
0.654).
33
Reasons :

• Effect of tank bulging on nozzle is insignificant
34
Reasons :

• Effect of tank bulging on nozzle is insignificant
• Effect of nozzle movement and rotations on the
piping and nozzle load is insignificant.
Because piping from nozzles at higher level is not
short and rigid.
35
Now we will see
bulging.
bulging. 36
To minimize the effect of tank bulging, on
piping.
Due to bulging, nozzle at lower levels rotates
downward.
This causes pipe to move vertically downwards.
To minimize the amount of this movement :
• Piping shall be rotated through 90° as close to
the tank wall as practical.
(2D spool may be provided to avoid elbow
stiffening due to flanged elbow)
37
In other words we can also say, .
nozzle orientation should be such that, it does not
. directly points towards pipe’s primary direction to
which it has to run.
It should point to 90° to the pipe’s primary

direction to which it has to run.
38
Converts nozzle loading
into favorable direction
As the pipe’s downward movement is restricted
by next pipe support, it causes vertical
reaction load and moment on tank nozzle.
This (rotating piping by 90°), converts the
longitudinal moment acting on tank nozzle to
torsional moment.
This is of minor consequence as nozzle is much
more stronger against torsional loading, than
longitudinal loading.
39
Tank nozzle has to account in Caesar
Modeling
• Nozzle’s thermal movements

• Nozzle rotations due to tank bulging
• Tank settlement
40
In “Tank Settlement”, we will cover
• Why Settlement is not seen in equipments?
• How much Settlement to be considered?
• Pipe routing guidelines - to minimize effect of

tank settlement.
• Use of Spring support
41
we start with

tank settlement.
42
Why Settlement is not seen in equipments
It is possible to design its foundation with large raft (say 10
m), to minimize or have insignificant settlement.
• Tank diameter is large (10 m to 60 m)
It is impractical to design it’s foundation with much bigger raft
than this, to have insignificant settlement.
43
Now we will see

tank settlement.
44
How much Settlement to be considered
IN CASE OF SAND :
• Majority of the total settlement occurs during hydro
test of tank (before piping is connected).
This is generally permanent.
… typically 60%
• For balance 40% of settlement, which occurs after

piping connection, piping needs to be designed.
IN CASE OF CLAY : Progressive settlement.

45
How much Settlement to be considered
• It is more at the center of tank, and typically

50% at the edge of tank.
(Since our nozzles and tank roof are connected /

supported on shall, that is on outer edge of tank, we
need to consider the settlement at outer edge of
tank.)
46
Data to be obtained from civil for each tank
• Total long term settlement.

• Settlement that will occur during construction
and hydrotest of tank.
• Recovery (if any) following construction and
hydrotest of tank.
• Further settlement, after hydrotest of tank,
(at the edge of the tank).
47
Sample Data from civil for each tank
Settlement at Settlement at
Centre of Tank. Edge of Tank.
Tank no.
40% of total
Settlement at
Edge of Tank.
48
Now we will see

tank settlement.
49
To reduce effect of tank settlement
• Keep first support away
50
Now we will see

tank settlement.
51
To reduce effect of tank settlement
• Keep first support away
• Large dia. piping with large settlement may call
for spring support
however avoid spring support
- because accidental draining of line
will cause excessive upward force on nozzle
52
If Spring support is used
• WNC (Weight with No Content) load case is required (since it will be
liquid line).
• Nozzle should be OK in OPE & WNC case.
- Spring should be under tuned.

(to pass nozzle in OPE & WNC case)
-This will increase nozzle load in normal operating case, but will reduce load
in WNC case.
53
This can
be done
by
specifying
Operating
load in
Caesar.
54
You can see in this example,
. all piping is rotated through 90°,
and first support is away from the

nozzle.
55
Now we will see :

account for this?

load & pipe stress?
56
To model in Caesar
• Settlement
along with
• Nozzle’s thermal movement

• Nozzle movement due to tank bulging
• Nozzle’s local flexibility
57
First we will concentrate on
• Settlement
along with
• Nozzle’s thermal movement

• Nozzle movement due to tank bulging
• Nozzle’s local flexibility
58
For nozzles on tank bottom
to account for
Nozzle’s thermal movement
• Do not model tank, to account for its nozzle

movement.
• In-stead specify total nozzle movement &

rotations (at tank shell junction),
[due to thermal growth and tank bulging].
59
How to calculate total nozzle movement,
due to thermal growth and tank bulging?
• API-650 formula calculates nozzle
movements due to tank bulging only.
• Petrofac’s in-house developed spread
sheet available in e-BMS, readily
calculates total movement due to tank
bulging and thermal growth.
60
How to use this
• First copy e-BMS excel file with tank &
nozzle name.
• Within excel file, copy sheet for different
temperature cases,
such as T1 (Max. Design), T2 (Oper.),
T3 (Min. Design)
61
To apply this in Caesar
• Input this in Caesar as
D1  T1 (Max. Design)
D2  T2 (Max. Oper.)
D3  T3 (Min. Design),
62
Now for combining this with Settlement
• Input Settlement in Caesar as
D4  Settlement
63
Max. Oper. Temp.
• I would just like to add here,
ADCO commented us to use Max. Oper. Temp.,
in place of Normal Oper. Temp.,
for OCC load cases (e.g. T2+WIND).
• From Process we got confirmation that Ope.
Temp. listed in line list are Max. Oper. Temp.
64
Max. Oper. Temp.
• As per SHELL DEP 01.00.01.30-Gen.
“DEFINITION OF TEMPERATURE, PRESSURE AND
TOXICITY LEVELS”
4.1.2 Maximum operating temperature (MOT)
• The MOT is the highest temperature which provides
sufficient flexibility for the control of the intended
operation.
• In many cases this flexibility is not required and in
those cases the MOT is equal to the OT.
• The MOT is determined by the process engineer in
consultation with the process control engineer.
65
Load cases for “before settlement”
W+P1+T1+D1  OPE, at Max. Design Temp.

W+P1+T2+D2  OPE, at Max. Ope. Temp.
W+P1+T3+D3  OPE, at Min. Design Temp.
W+P1  SUS
build the load cases as above, for scenario before settlement.
66
Load cases for “before settlement”
W+P1  SUS
It may be noted that we do not add bulging
movements to SUS case (with tank filled).
The reason is, in SUS case, we include only primary
loads.
Nozzle movement due to bulging (or in fact any
movement), is secondary load.
67
Load cases for “after settlement”
W+P1+T1+D1+D4  OPE, at Max. Design Temp.
W+P1+T2+D2+D4  OPE, at Max. Ope. Temp.
W+P1+T3+D3+D4  OPE, at Min. Design Temp.
W+P1+D4  SUS
(Where, D4  SETTLEMENT)
build the load cases as above, for scenario after settlement.
68
Now incorporating nozzle flexibility
along with settlement,
and nozzle rotations in Caesar
• In Caesar define Nozzle at nozzle-shell

junction node.
69
Enter node on
nozzle pipe piece,
at shell junction
.
Select “API 650”
.
Select “Nozzle”
Enter node on shell , at
shell junction
Nozzle flexibility is
inserted by Caesar
between these nodes
Input other data required by Caesar

to calculate nozzle flexibility
No input field for nozzle orientation in which
nozzle stiffness's will be inserted
Caesar insert nozzle stiffness's in direction of
element
70
Now to enter shell displacements and Enter “Tank node” (node on shell ), which
rotations. we have entered in “Nozzle” input data
.
Select
“Displacements”
Enter D1, D2, D3, nozzle

movements due to temp. & bulging.
Enter D4, Tank settlement
71
. If we are modelling this nozzle.
Then this is not a radial nozzle.

.
It is along X axis, but has offset.
If we model this nozzle along

X, then nozzle stiffness will be
inserted along X, not in radial
direction.
Therefore model small element (equal

to shell thickness) in radial direction,
and input nozzle on this element.
Shell movement & rotations

calculated are also in radial
direction, and needs to be
resolved in X & Z axis.
72
.
We will have nozzle orientation plan.
73
.
Where offset will be specified.
Tank outer
radius we know.
Therefore we
can calculate
radial direction.
74
Resolving radial shell thickness and radial shell movements & rotations,
is a tedious job.
Therefore enter shell

thickness and shell
movements & rotations along Observe the nozzle
nearest principal axis (Z axis orientation here,
in this case). before rotation.
Then click here to rotate that

element.
75
.
76
.
.
That is because
shell thickness is
rotated (on which
nozzle is specified).
Now observe the
nozzle orientation
here, after rotation.
This also rotated
shell displacements
and rotations.
77
• In this case,
while checking the nozzle loads also,
we have to get load & moments
about radial & lateral axis
(not X & Z).
78
For this, you need to select
.
entire model, and rotate it
about Y axis, by the angle of
. nozzle (in this case 9.2°), so
that radial direction of nozzle
is along nearest principal axis
(Z axis in this case).
79
.
Now observe the

nozzle orientation
here, after rotation.
80
Sign convention for nozzle
rotations :
As per API-650,
θ +ve is
upwards
and,
θ -ve is
downwards
But when it comes to Caesar convention, it is not always +ve is upwards. 81

Caesar follows Right hand rule.
• Put the thumb of your right hand along the
positive axis
• The direction your fingers curl is positive
rotation about that axis
82
rotations
when θ is +ve, : θ is +ve,
Then when
Rz is also +ve, Rz is -ve,

Y RZ (-) θL (+)
Y
θL (+)
Z X RZ (+)
Z X
RZ (+)
θL (-)
Z
RZ (+)
But, if nozzle
θL (-) RZ (-) Z
If nozzle is in RZ (+) is in -X
+X direction, NOZZLE AXIS : (+X) NOZZLE AXIS : (-X) direction,
when θ is +ve, θL (+) RX (+) Y Then when θ is +ve,
Rx is also +ve, RX (-)

θL (+) Z X Rx is -ve,
RX (-)
Y
θL (-)
RX (+)
Z X RX (+) X
X
RX (+) θL (-)
Similarly, if But, if nozzle
nozzle is in -Z NOZZLE AXIS : (-Z) NOZZLE AXIS : (+Z) is in +Z
direction, direction,
83
This can be presented in form of table:
Nozzle rotation in
API-650 Caesar Nozzle Axis Input of Nozzle rotation in Caesar
Nomenclature
θL (+) +X + RZ
-X - RZ
+Z - RX
-Z + RX
+X - RZ
θL (-)
-X + RZ
+Z + RX
-Z - RX
84
rotations :
It is better if
this chart &
table is put
on a paper,
and
circulated to
stress team
/ out side 3rd
party like
GRE
vendor, to
avoid any
mistakes.
85
Now we will see :

account for this ?

load & pipe stress?
86
Acceptance criteria for:
• Nozzle load
• Pipe Stress
87
Acceptance criteria for Nozzle loads :
• For nozzle loads, of-course it should be within

allowable limit.
• In earlier version of API 650

(up to 10th Edition – 2003 with Addendum-3),
it had section P.3 in Appendix-P,
which was based on WRC 297
88
Acceptance criteria for Nozzle
loads :
• Petrofac have in-house spread sheet in eBMS
to check nozzle load
as per Appendix-P.3 of earlier version of API-

650
89
• In present version of API 650
(11th Edition – 2007),
in Appendix-P,
Section P.3 is Deleted in its Entirety.
- not included in any other section,
- No alternate method suggested.

90
This leaves us the option to check nozzle
• as per WRC 297 (P.3 was based on WRC 297)
OR
• To continue using method as per P.3 of

previous version of API 650.
91
• Under section 1.0 Scope of API 650
It is stated that
[under clause 1.1.22 of 11th Edition – 2007]
“An alternative or supplemental design may
be agreed upon by the Purchaser or
Manufacturer.”
92
Based on this, we can specify
• WRC 297
OR
• Appendix P.3 of previous version of API 650

[10th Edition – 2003 with Addendum-3].
93
Acceptance criteria for:
• Nozzle load
• Pipe Stress
94
Acceptance criteria for Stress :
ASME B31.3
CHAPTER II : DESIGN
Para 301 defines various temp, pr., forces, and

considerations to be accounted in design.
95
Sub Para 301.8 states,

Effect of pipe support movement due to
- Thermal expansion
- Settlement
- Tidal movement
- Wind sway
to be accounted in design.
96
Para 319 : Piping flexibility, Sub Para 319.2.1(c) states,
Movement due to earth settlement,
since it is a single cycle effect,
will not significantly influence fatigue life.
A displacement stress range greater than
that permitted by para. 302.3.5(d)
{i.e. SA = f [1.25(Sc + Sh) – SL ] }
may be allowable
if due consideration is given
to avoidance of excessive localized strain and end
reactions.
97
However it does not specify exact value

to which it can be allowed.
98
ASME Section III, Division 1, Subsection NC
(Rules For Construction Of Nuclear Facility Components)
Para NC-3653.2(b) states,

for effect of single non repeated anchor movement
i MD / Z ≤ 3Sc
Where, MD = resultant moment due to any single non-repeated

anchor movement (e.g., predicted building settlement)
99
Thus NC code specifies limit as 3Sc.

⁻ This includes SIF
⁻ The stress is only due to anchor movement (settlement),
(does not include stress due to weight, temperature or
pressure)
This should be spelt out in project stress analysis design
basis, as this is not part of ASME B 31.3.
100
That is, if Load cases for “before settlement” are
L1 = W+P1+T1+D1  OPE, at Max. Design Temp.
L2 = W+P1+T2+D2  OPE, at Max. Ope. Temp.
L3 = W+P1+T3+D3  OPE, at Min. Design Temp.
L4 = W+P1  SUS
101
, and Load cases for “after settlement” are :
L5 = W+P1+T1+D1+D4  OPE, at Max. Design Temp.
L6 = W+P1+T2+D2+D4  OPE, at Max. Ope. Temp.
L7 = W+P1+T3+D3+D4  OPE, at Min. Design Temp.
L8 = W+P1+D4  SUS
(Where, D4  SETTLEMENT)
102
Then :
L8 – L4 ≤ 3Sc
Where,
L8 = W+P1+D4  SUS, after Settlement
L4 = W+P1  SUS, before Settlement
103
Qualification or acceptance criteria for Stress :
However,
⁻ The tank nozzle allowable are generally of such a low
value, that it becomes governing criteria (not pipe
stress)
⁻ When you qualify to tank nozzle, pipe stress value
will be very low, as compared to allowable value.
⁻ It will even pass in normal displacement stress range
criteria itself.
104
That is :
L1 – L7 ≤ f [1.25(Sc + Sh) – SL ]
Where,
L1 = W+P1+T1+D1  OPE, at Max. Design Temp.,
giving highest nozzle position
L7 = W+P1+T3+D3+D4  OPE, at Min. Design Temp.,
giving lowest nozzle position.

105
RECAP
106
Thus today we have seen :

account for this ?

load & pipe stress?
107
Under this, about “Tank Bulging”, we have seen
• A Sample calculation for shell movements at various
bulging.
bulging. 108
And about “Tank Settlement”, we have seen

tank settlement.
109
Then we have seen :

account for this (nozzle movements due to
tank bulging, and settlement, along with
nozzle flexibility)?

load & pipe stress?
110
Questions ?
111
Thank You
112

Analysis of lines to API-650 Nozzles

Uploaded by

Copyright:

Available Formats

Analysis of lines to API-650 Nozzles

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Analysis of lines to API-650 Nozzles

Uploaded by

Copyright:

Available Formats

Stress analysis of lines connected

to API-650 tank Nozzles

• How Caesar modelling should be done to

• What are the acceptance criteria for nozzle

• How Caesar modelling should be done to

• What are the acceptance criteria for nozzle

• Nozzle’s thermal movements, and

• Nozzle’s thermal movements

What is this Tank

Tank is filled with liquid.

This liquid has varying height.

Due to this there is varying liquid

It has more pressure at bottom.

Due this, tank wall try to expand

But the bottom plate prevent this

Due to this, actual shape of tank

Due to this, nozzle on the shell

and rotates in vertical plane,

The nozzle on lower portion

We do not calculate this for nozzles at higher

For nozzles at higher level,

For nozzles at higher level,

For nozzles at higher level,

It should point to 90° to the pipe’s primary

• Nozzle’s thermal movements

• Why Settlement is not seen in equipments?

• How much Settlement to be considered?

• Pipe routing guidelines - to minimize effect of

• Use of Spring support

• Why Settlement is not seen in equipments?

• How much Settlement to be considered?

• Pipe routing guidelines - to minimize effect of

• Use of Spring support

• Why Settlement is not seen in equipments?

• How much Settlement to be considered?

• Pipe routing guidelines - to minimize effect of

• Use of Spring support

• For balance 40% of settlement, which occurs after

IN CASE OF CLAY : Progressive settlement.

• It is more at the center of tank, and typically

(Since our nozzles and tank roof are connected /

• Total long term settlement.

• Why Settlement is not seen in equipments?

• How much Settlement to be considered?

• Pipe routing guidelines - to minimize effect of

• Use of Spring support

• Why Settlement is not seen in equipments?

• How much Settlement to be considered?

• Pipe routing guidelines - to minimize effect of

• Use of Spring support

- Spring should be under tuned.

. all piping is rotated through 90°,

and first support is away from the

• How Caesar modelling should be done to

• What are the acceptance criteria for nozzle

• Nozzle’s thermal movement

• Nozzle’s thermal movement

• Do not model tank, to account for its nozzle

• In-stead specify total nozzle movement &

W+P1+T1+D1  OPE, at Max. Design Temp.