Depressuration Report

Depressurisation Report
NO-HLD-10-AET2-001021
Rev.
Status
Date
Revision memo
00
IFR
25.05.12
ISSUED FOR REVIEW
Issued by
GK/SLT
Checked by
M
Approved by
NOS
This document has been generated by an Electronic Document Management System. When printed it is considered as a for information only copy.
The controlled copy is the screen version and it is the holder's responsibility that he/she holds the latest valid version.
Document number
DEPRESSURISATION REPORT
Document Type : REP
System / Subsystem :
Contractor document number :
4100H11.002
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
00
Rev Date :
Status
IFR
25.05.12
Page 2 of 62
TABLE OF CONTENT
1.
2.
3.
4.
Summary ............................................................................................................................................... 3
Introduction ........................................................................................................................................... 3
Abbreviations ........................................................................................................................................ 4
Depressurisation Criteria....................................................................................................................... 5
4.1.
Process Sectionalisation........................................................................................................... 5
4.2.
Calculation Design Basis .......................................................................................................... 5
4.3.
Initial condition .......................................................................................................................... 6
4.3.1.
Calculation results fire case ............................................................................................... 6
4.3.2.
Equipment and piping integrity during fire............................................................................. 7
5. Depressurisation methodology.............................................................................................................. 7
5.1.
General ..................................................................................................................................... 7
5.1.1.
Pipe....................................................................................................................................... 7
5.1.2.
Equipment............................................................................................................................. 7
5.2.
Fire Case .................................................................................................................................. 7
5.3.
Cold case minimum design temperature ............................................................................... 8
5.4.
Manual Depressurisation .......................................................................................................... 9
6. Depressurisation and Cold temperature result...................................................................................... 9
6.1.
HP Depressurisation ................................................................................................................. 9
6.1.1.
HP3-20-RO0010 1st Stage Separator .................................................................................. 9
6.1.2.
HP4-20-RO0024 2nd Stage Separator ............................................................................... 10
6.1.3.
HP6- 35-RO1025 Gas Export Compressor 1...................................................................... 11
6.1.4.
HP20- 35-RO1225 Gas Export Compressor 2.................................................................... 12
6.1.5.
HP8-31-RO0015 1st Stage Recompression ....................................................................... 13
6.1.6.
HP9-31-RO0036 2nd Stage Recompression...................................................................... 14
6.1.7.
HP10-31-RO0056 3rd Stage Recompression..................................................................... 15
6.1.8.
HP11-31-RO0074 4th Stage Recompression..................................................................... 15
6.2.
LP Derpressurisation .............................................................................................................. 16
6.2.1.
LP1-20-RO0042 3rd Stage Separator ................................................................................ 16
6.3.
Maintenance Depressurisation ............................................................................................... 17
6.3.1.
10-RO-1222 Brent gas well (Typical) To HP....................................................................... 18
7. References .......................................................................................................................................... 19
APPENDIX A - DEPRESSURISATION LOAD SCHEDULE....................................................................... 20
Depressurisation Load Schedule ............................................................................................................ 20
Manual Depressurisation Load Schedule ............................................................................................... 21
Summary of Low temperature calculations............................................................................................. 22
APPENDIX B - NEW*S Documentation ..................................................................................................... 23
This document is the property of COMPANY. It must not be stored, reproduced or disclosed to others without written authorisation from the Company.
DEPRESSURISATION REPORT-REV00.DOC
Document number
Document Type : REP
4100H11.002
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
00
Rev Date :
Status
IFR
25.05.12
Page 3 of 62
1. SUMMARY
The purpose of the depressurisation system is to reduce the pressure in hydrocarbon systems. Depressurisation can
be planned operations, for example before maintenance, or it can be at an emergency in case of fire, to ensure that
the system will not be subject to stresses above design at extreme temperatures. In addition, the depressurisation
shall reduce the hydrocarbon inventory in order to minimize the amount of hydrocarbons fed to a fire in case of a
rupture.
The main segments modelled in the 3D-model have been simulated in NEW*S. For the Gas Export and the 4th
stage Recompression segment further investigation is needed to keep the minimum temperature above -46C. The
blowdown time need to be prolonged or it may be limited by initiating depressurisation at a higher temperature than
the minimum ambient of - 10C. The blowdown rate currently exceeds 10.15 MSm/d, calculations need to be rerun
with longer depressurisation time until 10.15 MSm/d is not exceeded.
Equipment and piping integrity during fire will be part of the detail engineering work.
2.
INTRODUCTION
The main objective of this report is to describe the design of the Hild depressurisation system with respect to:
x
Design rates and capacities to the flare systems.
Sizes and sizing basis for major blowdown devices.
Sizes of depressurisation lines in the flare system.
Uncertainties and areas for further work.
The design is in accordance with all applicable guidelines, standards and company requirements. Where this is not
the case, Company will be notified for review and approval of the deviation.
Reference is made to relevant standards and philosophies listed in section 10 and to the following design
documents:
NO-HLD-10-AET2-001106
Safety Analysis Flow Diagram, Oil Separation
NO-HLD-10-AET2-001107
Safety Analysis Flow Diagram, Oil Export
NO-HLD-10-AET2-001108
Safety Analysis Flow Diagram, Gas Recompression
NO-HLD-10-AET2-001109
Safety Analysis Flow Diagram, Gas Treatment And Export
NO-HLD-10-AET2-001110
Safety Analysis Flow Diagram, Flare, Closed & Open Drain
NO-HLD-10-AET2-001112
Safety Analysis Flow Diagram, Fuel Gas System
NO-HLD-10-AET2-001126
Utility Flow Diagram, Flare System
Document number
Document Type : REP
4100H11.002
3.
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
00
Rev Date :
Status
25.05.12
Page 4 of 62
ABBREVIATIONS
BDV
Blowdown valve
ESDV
Emergency Shutdown Valve
Fire
FPSO
Floating Production Storage Offloading (Vessel)
HP
High Pressure
ISO
International Standard Organisation
KO Drum
Knock Out Drum
LAH
Level Alarm High
LAL
Level Alarm Low
LSHH
Level Switch High High
LP
Low Pressure
LSLL
Level Switch High High
NEW*S
Fluid properties and process simulation program from Bubblepoint AS
PAH
Pressure Alarm High
PDMS
Plant Design Management System from Aveva.
PFP
Passive Fire Protection
PSHH
Pressure Switch High High
PSLL
Pressure Switch Low Low
PSS
Process Shutdown System
PSV
Pressure Safety Valve (Relief valve)
P&ID
Piping and Instrument Diagram
RO
Restriction Orifice
SD
Safety Shut Down
SOP
Settle-out pressure
UFD
Utility Flow diagram
IFR
Document number
Document Type : REP
4100H11.002
4.
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
00
Rev Date :
Status
IFR
25.05.12
Page 5 of 62
DEPRESSURISATION CRITERIA
The ISO 23251 /3/ suggests a depressurisation down to 6.9 barg or 50% of design pressure, which ever is lower,
within 15 minutes. According to NORSOK S-001 /4/ the use of passive fire protection shall be minimized by rapid
depressurisation.
Simultaneous depressurisation is the preferred method for evacuation of hydrocarbons. It is considered as the safest
depressurisation method due to less complex configuration than for staged depressurisation.
Low temperature depressurisation calculations needs to be performed in order to validate the selected minimum
design temperature.
4.1.
Process Sectionalisation
The process is divided into depressurisation sections by sectionalisation valves (SD or ESD valves). Reference is
made to GS EP SAF 261 /5/, chapter 5.1.3 defining the criteria for whether a BDV is required. Ref. Table 4-1.
Table 4-1 Criteria for BDV.
Piping
or
Vessel
BDV required
That cannot be isolated
No
That can be isolated, but cannot No
be exposed to fire.
That can be isolated and can be
exposed to fire:
x Flammable gas
x P > 7 barg and PVgas > 100 bar.m3
x Liquefied HC
x Mgas or Mliq > 2 tonnes of C4 and more
volatile
x Liquid HC
x No
x Two-phase
x P > 7 barg and PVgas > 100 bar.m3
x Toxic gases
x As required for protection of personnel
The depressurisation facilities consisting of an actuated block valve (BDV), orifice plate, pipe expander and a
manual isolation valve (locked open, full bore) in that order, all in upstream system pipe spec.
4.2.
Calculation Design Basis
The depressurisation calculations will be performed in the NEW*S software Version 3.30, 2011 (From
Bubblepoint AS). The software has been validated by Imperial College on the Skarv BP FPSO project and gives
satisfactorily results compared with the BLOWDOWN software (Imperial College). Further it should be noted that
the methodology utilised in the NEW*S software is acknowledged by the API 521 committee (Reference is made
to API 521 /3/ chapter 5.15.2.3. In this section there is made a reference to a more rigorous method. Reference
[141] effectively describes the thermodynamic background for the NEW*S software presented at the GPA
conference in 1993). Reference is made to Appendix B
Document number
Document Type : REP
4100H11.002
4.3.
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
00
Rev Date :
Status
IFR
25.05.12
Page 6 of 62
Initial condition
No credit is taken for fire-fighting equipment. The flash is assumed to be isentropic and compensated by heat from
the wall heated by the external fire load. Initial condition and design criteria for calculating depressurisation loads for
sections exposed to fire are as follows:
x
Temperature, ref. Table 4-2.
Pressure, ref. Table 4-2.
Vessel liquid level, ref. Table 4-2.
Heat input is 100 kW/m2
Backpressure is 1 bara
Combined piping and equipment volumes, wetted areas and metal mass to be used.
For Emergency depressurisation AET will follow the Hild Field Development Basis of Design /6/ section 15.2.15
quoting that the platform shall be designed in accordance with GS EP SAF 261 /5/. AET suggest using PSHH or
settle-out pressure based on PSHH at suction and discharge side of compressor and normal operating temperature.
Ref GS EP SAF 261 section 5.2.2 quoting "The initial pressure to be considered shall be the maximum operating
pressure, which will normally correspond to the PSHH."
This will derogate from GS EP EPC 103 /7/, section 13.2 quoting "The initial pressure will be the system design
pressure/safety relief valve set pressure, except for the compressor systems for which the settle out pressure will be
considered."
Table 4-2 Initial conditions, depressurisation
Case
Purpose
Fire?
Initial
temperature
Initial pressure
Initial liquid level
4.3.1.
Emergency depressurisation
Maximum
total
depressurisation flowrate to
flare
Sizing of depressurisation
orifice and depressurisation
piping
Yes
Normal operating
PAHH / Settle out pressure
LAHH/LL to be checked
Planned depressurisation
Low temperature
Total
depressurisation Minimum
design
flowrate expected upon temperature case
planned shutdown
No
Normal operating
No
Minimum ambient
Normal operating pressure Based

on
constant
/ Settle out pressure
volume flash from PSV
set pressure to minimum
ambient temperature
Normal operating
LAHH/LL to be checked
Calculation results fire case
Reference is made to Depressurisation Load Schedule attached in Appendix A, which gives a total blowdown rate of
10.15 MSm3/d to the HP Flare and approx 0.16 MSm3/d to the LP Flare.
Document number
Document Type : REP
4100H11.002
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
00
Rev Date :
Status
IFR
25.05.12
Page 7 of 62
According to GS-EP-SAF 261 /5/ it is a good principal to fire insulate liquid containing vessel and piping to reduce
boiling and thereby avoid longer depressurisation time. This principle is foreseen to result in fire insulation on all
piping, pressure vessels, filters, electrical heaters and shell and tube heat exchangers containing liquid.
Passive fire protection is foreseen required on the 2nd Stage Separator and the 3rd Stage Separator. The fluids in
the 2nd Stage Separator and 3rd Stage Separator contain significant quantities of water. As the segments are
depressurised, the vapour phase will not be saturated with water, which subsequently will result in a significant mass
flux of water to the vapour phase as the temperature increases and the pressure drops. These vessels are
recommended to be fire insulated to prevent a steam explosion.
4.3.2.
Equipment and piping integrity during fire
HOLD - will be part of detail engineering work.
5.
DEPRESSURISATION METHODOLOGY
5.1.
General
Sensitivity checks on depressurisation rate have been performed on different years for the 1st Stage
Separator section, worst case is HILD-Gas_Alone-2018-07 will be used for Fire Case
NEW*S will be used for calculations outlined in section 5.2.
NEW*S will be used for low temperature calculations outlined in section 5.3.
5.1.1.
Pipe
The PDMS 3D model will be used to determine pipe work volumes within each segment (limited by sectionalisation
valves). The inner diameter is determined using Piping And Valve Material Specification ref /1/. In addition, 15% of
the calculated pipe work volume is added to include a margin in this early stage of the project. Volumes will be
rechecked as part of detail engineering work.
5.1.2.
Equipment
The equipment volume is based on the Master Equipment List /8/. The end head of all vessels are assumed to be
elliptical. An additional 10% of the total calculated volume is added to include a margin.
Volumes will be updated when supplier information is available.
5.2.
Fire Case
The purpose of the fire case is to calculate the orifice area and to design the upstream and downstream pipe
dimensions of the blowdown line.
Start conditions:
Document number
Document Type : REP
4100H11.002
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
00
Rev Date :
Status
IFR
25.05.12
Page 8 of 62
Normal operating temperature
PSHH or settle-out pressure based on PSHH at suction and discharge side of compressor.
High high liquid level in separators
High high liquid level in scrubbers
Credit for insulation only if PFP, i.e equipment and pipes provided with insulation for other purpose are
regarded as uninsulated.
The heat input is 100 kW/m (Global fire, ref /4/)
Add all calculated blowdown rates (Simultaneous opening of the BDV for all fire zones) and find the total
blowdown rate,
If the blowdown rate exceeds 10.15 MSm/d, calculations to be rerun with longer depressurisation time until
10.15 MSm/d is not exceeded.
As a first estimation, the pressure drop requirement is to reach 6.9 barg within 14 minutes. A time delay of 1
minute is assumed in order to allow sectionalisation valves to close and blowdown valve to open. A total
depressurisation time will then be 15 min, ref GS EP SAF 261 /6/. This default depressurisation time, is for
vessel wall thickness of 25 mm; for thinner walls, the depressurisation time shall be reduced i.e. 3 minutes
for each 5 mm. For thicker walls, the depressurisation time cannot be longer than 15 minutes unless a
specific study is validated by Total.
The depressurisation time of the compressor segments and the TEG Contactor segment needs to be revised when
supplier information with respect to maximum pressure drop per time [bar/min] has been clarified. Currently it is
assumed that these segments will withstand the gradients resulting from an API blowdown.
5.3.
Cold case minimum design temperature
The purpose of the Cold Case is to determine the minimum operating temperature of the blowdown segment and the
minimum operating temperature downstream the orifice.
The low temperature calculation is performed on the main vessel in a process segment, ignoring all piping. This is
regarded to be a conservative approach as the piping holds much more heat capacity in the steel relative to heat
capacity of the process fluid, and would contribute to increase the minimum design temperature if included in the
calculations. The low temperature calculations performed in NEW*S give separate minimum temperatures in both
the fluid and the pipe/vessel wall for both the liquid and vapour phase. The lowest steel temperature reported is
applied as the basis for the minimum design temperature and material selection for a segment. However, it is
evaluated that the depressurisation nozzles on the separators and the connected depressurisation lines will locally
reach a significantly lower temperature due to the good heat transfer caused by a high flow rate throughout the
depressurisation. Therefore, the minimum gas temperature is applied as the basis for material selection and
minimum design temperature for these parts of the segments whenever it affects the material selection.
Low temperature calculations will be performed for high and low pressure segments.
Reference is made to Depressurisation Load Schedule attached in Appendix A for cold temperature results.
Low temperature calculations are performed with the orifice size calculated in fire case and initial conditions are
given below
Start conditions:
x
The initial pressure is based on constant volume flash from settle out PSHH or PSV set pressure and normal
operating temperature to an ambient temperature of -10 C, as stated by Company /6/.
Document number
Document Type : REP
4100H11.002
NA
Pipe work metal and volume excluded.
Use LSLL in vessels. LSHH to be checked as well.
Use the orifice size calculated in Fire Case
5.4.
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
00
Rev Date :
Status
IFR
25.05.12
Page 9 of 62
Manual Depressurisation
Equipment that may be isolated and maintained during operation is equipped with manual depressurisation line to
Flare. These depressurisation lines are equipped with a restriction orifice that shall be sized to keep temperatures
and flow rates within the design limitations of the equipment.
Reference is made to the Depressurisation Load Schedule attached in Appendix A, a separate sheet gives the
overview of maintenance blowdown with rates and corresponding cold temperatures.
6.
DEPRESSURISATION AND COLD TEMPERATURE RESULT
The input data collected from the spreadsheet Hild Equipment volumes.xls is included in a Unit Defined Command
Procedure file (UDC-file) for calculation of pressure profile (time dependent pressure) during depressurisation. An
UDC-file is used by NEW*S to read input data to defined units in NEW*S and is actually a collection of most of the
commands needed to run a depressurisation simulation. In the UDC-file, all lengths are in meter and all times are in
minute.
The simulation gives, besides pressure in the segment with time, blowdown rates with time, time to reach
backpressure from flare and remaining volume and mass of fluid in the segment with time.
Based on experience from previous project the Coefficient of Discharge used for all restriction orifice plates with
critical flow have been CD= 0.83932.
All the UDC and results obtained with the process simulation tool NEW*S, will be found in the folder
\depressurisation simulations\
The folder name will indicate the corresponding RO.
6.1.
HP Depressurisation
6.1.1. HP3-20-RO0010 1st Stage Separator

20-RO0010 is installed on the 1st Stage Separator, and routed to the HP Flare header. Reference is made to P&ID
NO-HLD-10-AET2-001210.
The RO is sized based on volume from the depressurisation simulations\Hild Equipment volumes.xls.
As input for the calculation stream 20010 from HILD-Gas_Alone-2018-07 have been used. Else,
Operating pressure, bara
85
Total vessel volume, m3
63.62
Initial flow rate, kg/h
60103.2
14 minute to reach 7.9 bara

Calculated orifice diameter, mm
38.74
Document number
Document Type : REP
4100H11.002
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
Status
00
Rev Date :
25.05.12
Page 10 of 62
IFR
References are made to NEW*S simulations for orifice calculation see: depressurisation simulations\HP3-20RO0010
All relevant data are included in NEW*S UDC HP3-20-RO0010.udc.txt and simulation result is printed in HP3-20RO0010.doc
Low temperature calculation is performed in NEW*S on the 1st Stage Separator, 10-VZ2001, with an orifice size
giving the same pressure profile as the one calculated in the fire case. Initial conditions and result are:
Operating temperature, C
-10.00000
62.5
Calculated minimum temperatures:

Vapor (nozzle temperature),C
-51.80
Liquid,
-22.53
Minimum metal temperatures:

Vessel (vapour space),
-22.29
Vessel (liquid space),

C
-21.32
Minimum downstream temperature in the flare system at t= 2.45 is
-71.22
Reference is made to NEW*S simulations stored in: depressurisation simulations\Cold temperatures\ HP3_cl-1sep
All relevant data are included in NEW*S UDC HP3_cl-LL.udc.txt and result is printed in HP3_cl-LL.doc
The minimum design temperature of -46C in the segment and a cold sleeve in the depressurisation nozzle on the
Separator with minimum design of -60C is sufficient.
6.1.2. HP4-20-RO0024 2nd Stage Separator

20-RO0024 is installed on the gas outlet of the 2nd Stage Separator, and routed to the HP Flare header. Reference
is made to P&ID NO-HLD-10-AET2-001211.
32
184.53
46236

56.05
Document number
Document Type : REP
4100H11.002
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
Status
00
Rev Date :
25.05.12
Page 11 of 62
IFR
Low temperature calculation is performed in NEW*S on the 2nd Stage Separator, 10-VZ2002, with an orifice size
-10.00000
21.45

-44,92
Liquid,
-14.17

-16.21

C
-13.92
-49.32
Reference is made to NEW*S simulations stored in: depressurisation simulations\Cold temperatures\ HP4_cl- 2sep
All relevant data are included in NEW*S UDC HP4_cl.udc.txt and result is printed in HP4_cl.doc
The minimum design temperature of -46C in the segment and a cold sleeve in the depressurisation nozzle on the
Separator with minimum design of -60C is sufficient.
6.1.3. HP6- 35-RO1025 Gas Export Compressor 1

35-RO1025 is installed on the Gas Export Compressor 1 outlet, and routed to the HP Flare header. Reference is
made to P&ID NO-HLD-10-AET2-001231.
As input for the calculation stream 35003A from HILD-Gas_Alone-2018-07 have been used. Else,
117
33.87
61165,2

29.09
Low temperature calculation is performed in NEW*S on the Gas Export Compressor Suction Scrubber 1, 10VZ3501, with an orifice size giving the same pressure profile as the one calculated in the fire case. Initial conditions
and result are:
-10.00000
92.9
Document number
Document Type : REP
4100H11.002
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
Status
00
Rev Date :
25.05.12
Page 12 of 62
IFR

-51.23
Liquid,
-32.55

-24.94

C
-31.34
-87.96
Reference is made to NEW*S simulations stored in: depressurisation simulations\Cold temperatures\ HP6-ExeA
The results need further investigation. To keep the minimum design temperature of -46C in the segment, the
the minimum ambient of -10 C.
6.1.4. HP20- 35-RO1225 Gas Export Compressor 2

35-RO1225 is installed on the Gas Export Compressor 2 outlet, and routed to the HP Flare header. Reference is
As input for the calculation stream 35003B from HILD-Gas_Alone-2018-07 have been used. Else,
117
29.91
53874

27.3
Low temperature calculation is performed in NEW*S on the Gas Export Compressor Suction Scrubber 2, 10VZ3502, with an orifice size giving the same pressure profile as the one calculated in the fire case. Initial conditions
and result are:
-10.00000
92.9

-52.70
Liquid,
-33.35
Document number
Document Type : REP
4100H11.002
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
Status
00
Rev Date :
25.05.12
Page 13 of 62
IFR
-28.0

C
-31.92
-88.58
Reference is made to NEW*S simulations stored in: depressurisation simulations\Cold temperatures\ HP7-ExeB
6.1.5. HP8-31-RO0015 1st Stage Recompression

31-RO0015 is installed on the 1st Stage Recompression outlet, and routed to the HP Flare header. Reference is
As for thinner walls, the depressurisation time shall be reduced i.e. 3 minutes for each 5 mm. The segment has a
wall thickness of 8mm, so as pr governing document /6/ the blowdown time has to be reduced to less than 5 minutes
to reach 50% of design pressure. As input for the calculation stream 31008 from HILD-Gas_Alone-2018-07 have
been used. Else,
8.1
12.73
4420,2
4.1 minute to reach 5 bara

30.8
All relevant data are included in NEW*S UDC HP8-31-RO0015.udc.txt and simulation result is printed in HP8-31RO0015 to 2 bara.doc
Low temperature calculation is performed in NEW*S on the 1st stage Suction Scrubber, 10-VZ3101, with an orifice
size giving the same pressure profile as the one calculated in the fire case. Initial conditions and result are:
-10.00000
5.65

-29.06
Liquid,
-16.27

-10.23

C
-13.44
-30.68
Document number
Document Type : REP
4100H11.002
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
Status
00
Rev Date :
25.05.12
Page 14 of 62
IFR
Reference is made to NEW*S simulations stored in: depressurisation simulations\Cold temperatures\ HP8-1r1
The minimum design temperature of -46C in the segment is sufficient.
6.1.6. HP9-31-RO0036 2nd Stage Recompression

31-RO0036 is installed on the 2nd Stage Recompression outlet, and routed to the HP Flare header. Reference is
to reach 50% of design pressure. As input for the calculation stream 31018 from HILD-Gas_Alone-2018-07 have
been used. Else,
10
7.03
4280
4.2 minute to reach 5 bara

27.3
Low temperature calculation is performed in NEW*S on the 1st stage Suction Scrubber, 10-VZ3101, with an orifice
-10.00000
6.97

-38.81
Liquid,
-18.04

-10.25

C
-13.79
-117.57
Document number
Document Type : REP
4100H11.002
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
Status
00
Rev Date :
25.05.12
Page 15 of 62
IFR
6.1.7. HP10-31-RO0056 3rd Stage Recompression

31-RO0056 is installed on the 3rd Stage Recompression outlet, and routed to the HP Flare header. Reference is
to reach 6.9 barg. As input for the calculation stream 31028 from HILD-Gas_Alone-2018-07 have been used. Else,
23
4.19
4894
4.6 minute to reach 7.9 bara

19.2
References are made to NEW*S simulations for orifice calculation see: depressurisation simulations\ HP10-31RO0056
Low temperature calculation is performed in NEW*S on the 3rd stage Suction Scrubber, 10-VZ3103, with an orifice
-10.00000
15.9

-29.06
Liquid,
-16.27

-10.23

C
-13.44
-39.68
6.1.8. HP11-31-RO0074 4th Stage Recompression

31-RO0074 is installed on the 4th Stage Recompression outlet, and routed to the HP Flare header. Reference is
Document number
Document Type : REP
4100H11.002
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
Status
00
Rev Date :
25.05.12
Page 16 of 62
IFR
wall thickness of 18mm, so as pr governing document /6/ the blowdown time has to be reduced to less than 10
minutes to reach 6.9 barg. As input for the calculation stream 31038 from HILD-Gas_Alone-2018-07 have been
used. Else,
51
29.06
41609
9.4 minute to reach 7.9 bara

36.9
References are made to NEW*S simulations for orifice calculation see: depressurisation simulations\ HP11-31RO0074
All relevant data are included in NEW*S UDC HP11-31-RO0074.udc.txt and simulation result is printed in HP11-31RO0074 doc
Low temperature calculation is performed in NEW*S on the 4th stage Suction Scrubber, 10-VZ3104, with an orifice
-10.00000
38.3

-43.38
Liquid,
-40.12

-14.33

C
-36.63
-51.55
6.2.
LP Derpressurisation
6.2.1. LP1-20-RO0042 3rd Stage Separator

20-RO0042 is installed on the 3rd Stage Separator, and routed to the LP Flare header. Reference is made to P&ID
NO-HLD-10-AET2-001213.
The minimum temperature of -46C in the downstream LP flare system is the restriction when designing the flow
orifice for the 3rd Stage Separator. Calculations performed in NEW*S indicate a maximum rate of 10491 kg/h to be
within the design criteria of -46C for the segment.
Document number
Document Type : REP
4100H11.002
NA
Discipline :
NO-HLD-10-AET2-001021
Revision
Status
00
Rev Date :
25.05.12
Page 17 of 62
PRO
IFR
191.68
10491
Less than 14 minute to reach 1.5 bara

67.02
References are made to NEW*S simulations for orifice calculation see: depressurisation simulations\LP1-20RO0042
All relevant data are included in NEW*S UDC LP1-20-RO0042.udc.txt and simulation result is printed in LP1-20RO0042.doc
Low temperature calculation is performed in NEW*S on the 3rd Stage Separator, 10-20VZ2003, with an orifice size
-10.00000
6.13

-40.69
Liquid,
-11.42

-12.39

C
-11.30
-41.44
Reference is made to NEW*S simulations stored in: depressurisation simulations\Cold temperatures\ LP1_cl-3sep
All relevant data are included in NEW*S UDC LP1_cl.udc.txt and result is printed in LP1_cl.doc
6.3. Maintenance Depressurisation

All calculations of maintenance blowdown valves are stored in the attached folder:
depressurisation simulation\Maintenance manual valve
The folder name will indicate the corresponding FO.
For low temperature calculation, reference is made to:
depressurisation simulations\Cold temperatures\Maintenance manual valve
Document number
Document Type : REP
4100H11.002
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
Status
00
Rev Date :
25.05.12
Page 18 of 62
IFR
6.3.1. 10-RO-1222 Brent gas well (Typical) To HP

10-RO-1222 is installed on the Brent gas well and routed to the HP Flare header. Reference is made to P&ID NOHLD-10-AET2-001200.
The Mach number in the downstream flare system is the restriction when designing the flow orifice on the typical
Brent gas well. Calculations performed in Flare net indicate a maximum rate of 1900 kg/h. Input stream used for the
calculation is 20010, PSHH=94 bara and normal operating temperature is 66.48C.
Document number
Document Type : REP
4100H11.002
7.
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
Status
00
Rev Date :
25.05.12
Page 19 of 62
IFR
REFERENCES
1. NO-HLD-10-AET2-100005, Piping And Valve Material Specification
2. Scandpower, Guidelines for the Protection of Pressurised Systems Exposed to Fire Rev 2,
31.03.04
3. BS EN ISO 23251:2007
4. Norsok S-001- Edition 4, February 2008.
5. GS EP SAF 261, Emergency Shut-Down and Emergency De-pressurisation (ESD & EDP). Rev.02
6. NODOC01-#928856-V8, HILD field Development Basis of Design, Functionality Description &
Operating Requirements - Rev 7.
7. GS EP ECP 103, Process Sizing Criteria. Rev.05
8. NO-HLD-00-AET2-000502 Master Equipment List
9. Determination of Temperatures and Flare Rates During Depressurization and Fire. (Sverre Over,
Ellen Stange and Per Salater, Presented at 72nd annual GPA Convention March 15-17, 1993, San
Antonio, Texas
10. Size Depressurisation and relief Devices for Pressurised Segments Exposed to fire. (Salater,
Overaa, Kjensjord, CEP (AIChe) sept. 2002)
11. NO-HLD-00-GEO-955616, #955616 -Metocean Specification For The Hild Field
12. GS EP SAF 262, Pressure protection relief and hydrocarbon disposal systems. Rev.02
13. NORSOK P-001, Process Design, Ed.5 Sep 2006
14. NORSOK P-100, Process Systems, Rev.03, Feb 2010
15. .NO-HLD-10-AET-1-002000, Material Selection and Corrosion Protection Report
16. P09010-MTA-0044 - METOCEAN REPORT DRAFT FROM ARGOSS
17. NORSOK L-002, Piping system layout, design and structural analysis, edition 3, July 2009
Depressurisation Load Schedule
APPENDIX A - DEPRESSURISATION LOAD SCHEDULE
Document Type : REP

4100H11.002
NA
Discipline :
PRO
Document number
Revision
Status
00
Rev Date :
25.05.12
Page 20 of 62
IFR
NO-HLD-10-AET2-001021
Manual Depressurisation Load Schedule
Document Type : REP

4100H11.002
NA
Discipline :
PRO
Document number
Revision
Status
00
Rev Date :
25.05.12
Page 21 of 62
IFR
NO-HLD-10-AET2-001021
Summary of Low temperature calculations
Document Type : REP

4100H11.002
NA
Discipline :
PRO
Document number
Revision
Status
00
Rev Date :
25.05.12
Page 22 of 62
IFR
NO-HLD-10-AET2-001021
Document number
Document Type : REP
4100H11.002
NA
Discipline :
NO-HLD-10-AET2-001021
PRO
Revision
Status
00
Rev Date :
25.05.12
Page 23 of 62
APPENDIX B - NEW*S DOCUMENTATION
IFR
Safety
Size Depressurization and

Relief Devices for Pressurized
Segments Exposed to Fire
Per Salater,
Sverre J. Overaa
and Elisabeth Kjensjord,
Norsk Hydro ASA
Piping and equipment must withstand

fires without rupturing. This can be
accomplished by properly designing relief
and depressurization systems and using
passive fire protection, when needed.
his article presents the minimum requirements for performing proper depressurization and fire-relief calculations together
with a procedure for sizing depressurization and relief systems for pressurized systems exposed
to fire. An engineering approach for modeling geometrically complex process segments is detailed. This approach excludes the necessity of describing the total
segment geometry in detail. A fire model is described
with its required input parameters. The parameters will
vary for different fire characteristics.
Minimum requirements for calculations

Several simulation tools are available for sizing depressurization orifices and relief valves. Most lack the
necessary physical modeling required. The list below
summarizes the minimum requirements for the design
of depressurization and relief devices for pressurized
systems exposed to fire:
rigorous thermodynamics (multicomponent fluid
model and use of equation of state)
fire modeling (emissivities, absorptivity, temperature, convection, initial flux, duration and size)
segment (vessel, pipe) material properties, i.e., tensile (rupture) strength, heat capacity, conductivity
all are temperature-dependent
This article is extracted from the Norsk Hydro Best Practice on Depressurization and Fire
Relief Design. This Best Practice is a major basis for the international guide, Guidelines
for the Design and Protection of Pressure Systems to Withstand Severe Fires (1) (to be
issued shortly) and the Norwegian guide, Guidelines for Protection of Pressurized Systems Exposed to Fire (2).
38
www.cepmagazine.org
September 2002
CEP
heat-transfer coefficients (boiling, radiation, convection and conduction)

mass transfer between the fluid phases (boiling
and condensation)
fluid flow, i.e., flow-regime calculations (laminar
or turbulent) that are input into heat-transfer and pressure-drop calculations
modeling of the process-segment geometry (system volume, system outer and inner wall areas, weight,
wall thickness, liquid and gas volumes)
insulation (thickness and conductivity)
stress or strain, depending upon the rupture criteria
used, to which all pipes and equipment are exposed.
The fire
Modeling of fires for engineering purposes requires
simplifications compared to the more thorough turbulent-combustion models used in computational fluid
dynamics (CFD). Nonetheless, large-scale tests (3)
have verified that an engineering approach to fire modeling gives wall- and fluid-temperature profiles that are
close to measured ones when choosing the appropriate
input parameters for Eq. 1.
Before proceeding with the fire model, some terminology needs to be defined:
A global fire is a large fire that engulfs the entire
or a significant part of the process segment. A local
fire exposes a small (local) area of the process segment to the fire peak heat-flux. A jet fire is an ignited
release of pressurized, flammable fluids. A pool fire
Local fire
Heat Flux
is the combustion of flammable or combustible fluids spilled and retained on a

surface. The ventilation- and fuel-controlled fires are related to the stoichiometric ratio of air-to-fuel (Figure 1).
Figure 1 is general for both a jet and a pool
fire; the difference being a higher flux for the
jet fire. For a pool fire, the API fire (4) is illustrated as the lower, dashed line to the right.
Note that in the equation API RP 521 uses, increasing the area reduces the flux. The dashed
lines represent the average heat flux. However, when studying the total volume of a fire,
any point on the continuous curve will be
found. A ventilation-controlled fire is to the
left of the peak heat-flux in Figure 1 at a stoichiometry of < 1. The fuel-controlled fire is to
the right, i.e., the stoichiometry is > 1.
Global fuel-controlled
Globalventilationcontrolled
1.0
Stochiometric Ratio
Figure 1. The heat flux from a fire and its relation to the stoichiometric ratio.
The fire equation

The heat flux absorbed by a segment from a fire, qabsorbed
(kW/m2), can be modeled as:
qabsorbed = segment fire T 4 fire segment T 4 segment +
h Tgas Tsegment
Open-fuel-controlled
pool fire
(1)
and more soot attaches to the surface. For more on absorptivity and emissivity, see Ref. 5.
By combining the suggested highest or lowest typical values into the fire equation, the heat fluxes toward a cold segment are found (Table 4 (6)). Typical heat fluxes measured in
large-scale jet fire and pool-fire tests are within the maximum
and minimum values in Table 4. Norsok (7) recommends
using the initial incident heat fluxes as specified in Table 5.
Sizing procedures
The fire-relief and depressurization calculations determine:
size of the relief valves and depressurization orifices
requirements for passive fire protection (PFP)
size of the pipes downstream from the relief and depressurization valves (if any)
The absorbed heat flux will be reduced with increasing

segment temperature, and a steady-state segment temperature will be reached when the heat influx from the fire
equals the heat outflux from the segment.
The view factor, which is not included in Eq. 1, is a
Table 1. Typical flame emissivities for global
scaling factor for the radiative terms. The view factor is
and local fires.
1.0. It equals 1.0 when the segments that absorb radiation
see nothing but an optical, thick flame. Calculation of
Type of Fire
Global Fire
Local Fire
fire
fire
view factors is difficult and a conservative assumption involves use of a view factor of 1.0, which results in Eq. 1.
Ventilation-controlled pool fire
0.60.75
0.70.9
Fuel-controlled pool fire
0.60.75
0.70.8
The incident heat flux is calculated by setting segment =
Jet fire
0.50.75
0.60.75
1.0 and disregarding the segment emissivity term. The
initial incident heat flux from a fire is
calculated by setting segment = 1.0 and
Table 2. Typical temperatures and convective heat-transfer
Tsegment equal to the normal operating temcoefficients for a global fire.
perature (of the cold segment).
2
Type of Fire
Input to the fire equation

The different terms in the fire equation
are combined to achieve the required initial heat fluxes. Tables 1, 2 and 3 suggest
values to be used. The segment absorptivity and emissivity in Eq. 1 are normally
equal and depend upon the nature of the
surface. Typical values are 0.70.9. A
value of about 0.8 is typical for oxidized
surfaces. The value will change as more
Tfire, C
Tgas, C
Ventilation-controlled pool fire 1,0001,050

9501,000
Jet fire
1,0001,150
850950
800900
9501,050
h, W/m K
1530
1530
50125
Table 3. Typical temperatures and convective heat-transfer

coefficients for a local fire.
Type of Fire
Ventilation-controlled pool fire
Jet fire
Tfire, C
1,0501,125
1,0001,050
1,0001,150
CEP
Tgas, C
Equal to Tfire
Equal to Tfire
Equal to Tfire
September 2002
h, W/m2K
2030
2030
100150
www.cepmagazine.org
39
Safety
Table 4. Minimum and maximum heat fluxes calculated by Eq. 1 for the suggested values of the
input parameters toward a cold segment.
Type of Fire
Maximum
Incident Flux,
kW/m2
Minimum
Incident Flux,
kW/m2
Maximum
Absorbed Flux,
kW/m2
Minimum
Absorbed Flux,
kW/m2
326
367
138
187
158
228
121
187
88
124
101
142
306
347
127
171
145
208
98
160
65
92
75
105
Global jet fire

Local jet fire
Global fuel-controlled pool fire
Local fuel-controlled pool fire
Global ventilation-controlled pool fire
Local ventilation-controlled pool fire
optimum location of the systems

Table 5. Initial incident heat fluxes against "cold" segment (7).
sectionalization valves
minimum design temperature for Type of Fire
Global Fire
Local Fire
the flare system (if any) and for the
(Average load) (Maximum point
load)
pressurized segment.
The minimum design temperature may Pool fire, enclosed area, ventilation-controlled
200 kW/m2
130 kW/m2
influence the materials selection of the
Pool fire (crude), open or enclosed area,
100 kW/m2
150 kW/m2
system under evaluation. The design usu- fuel-controlled
ally begins by considering carbon steel or
Jet fire
250 kW/m2
stainless steel as the material of construction. However, temperature calculations
may result in the need to use a different grade of steel, for
culations. This is to allow for future tie-ins and expected
example, replacing a normal carbon steel with one suited
project growth. However, by increasing the design cafor low temperatures.
pacity of flare system, less PFP will usually be required.
Although a local fire has a higher heat flux than a
Depressurization-orifice-sizing procedure
global fire, the global fire normally exposes the pressurPrior to running depressurization calculations, the folized segment to the largest flux of heat energy, due to the
lowing must be established:
larger area encompassed by a global fire. Hence, the
the fire scenarios (jet fire, pool fire, local fire, global
global-fire parameters determine the rupture pressure. On
fire, etc.); define the initial heat flux, the duration and the
the other hand, the local fire has the highest heat flux, so
size (extent) of the fire(s)
its parameters determine the rupture temperature of the
the criteria for unacceptable rupture, which are usually
process segment.
one or more of the rupture pressures, the released
Valves and flanges are not accounted for in the procedure.
flammable/toxic fluid at rupture, and the time to rupture
Also, it does not consider the mitigating effects of active fire
the time from the start of a fire until depressurizaprotection (such as a deluge). The sizing criteria are set to
tion is initiated
avoid an unacceptable rupture that could escalate the fire.
the physical properties ultimate tensile strength
Step 1: Perform an initial estimate of the size of all de(UTS), heat capacity and thermal conductivity at elevated temperatures (up to 8001,000C) for the materials of
Nomenclature
construction used in the depressurization segments
the depressurization segment geometry (system volh
= convection heat-transfer coefficient of air/flame in contact
with segment, W/m2K
ume, wall area, weight, etc.).
=
absorbed heat flux from the fire, W/m2
q
Once the above data are assembled, follow the iteraabsorbed
T
=
flame temperature, K
fire
tive procedure in Figure 2. The goal of this depressuriza= temperature of air/flame in contact with segment, K
Tgas
tion design is to limit the use of passive fire protection
Tsegment = segment temperature (time-dependent), K
(PFP) by depressurizing as fast as possible, while remaining within the discharge capacity of the flare sysGreek letters:
tem. PFP should be avoided due to the risk of the consesegment = segment absorptivity, dimensionless
= flame emissivity, dimensionless
fire
quences of undetected corrosion under insulation, and
segment = segment emissivity, dimensionless
the additional installation and maintenance costs in
= Stefan-Boltzmann constant = 5.67 10-8 W/m2K4
curred by PFP systems.
axial
= longitudinal stress, MPa
When designing a new plant, it is not recommended to
= hoop stress, MPa
hoop
consider using the entire flare-system capacity in the cal= equivalent stress (Von Mises), MPa
Von Mises
40
www.cepmagazine.org
September 2002
CEP
Required information prior to blowdown Iteration

- Description of the fire scenarios (type of fire, duration,
heat fluxes, size)
- Blowdown section geometry (system volume, area, weight, etc.)
- Ultimate tensile strength at elevated temperature of
materials in the blowdown section
- Manual or automatic blowdown, i.e., time delay for start of
depressurization
- Acceptance criteria for rupture
Step 1:
Estimate the size of all orifices and calculate the pressure profile and
flare rates for all segments. Use the fire with the largest heat
input (kW). No PFP in this initial iteration.
Reduce the
size of the
orifice.
(Step 1)
Is the flare system
capacity utilized (when
adding all of the simultaneous
blowdown rates
together)?
No
(Step 1) Evaluate to increase the blowdown rate,

preferably for the most hazardous blowdown section.
Step 2:
Add insulation if required. Calculate the process segment pressure
profile. Use the fire with the largest heat input (kW).
Tip: Do several calculations with varying amounts of fire insulation.
Yes
No
In case of
any of the "ORs"
(Step 1)
Is the blowdown
rate less than
maximum
l-dP/dt l
Yes
Step 3:
Calculate the wall-temperature profile for all pipes and equipment.
Use the local fire with the highest heat flux (kW/m2).
Step 4:
Use the temperature profiles from Step 3 to calculate the rupture pressure
for all pipes and equipment. Compare with the pressure profile from
Step 2 (Step 1 in the first iteration).
Acceptance criteria:
- Pipe rupture pressure
- Equipment rupture pressure
- Released fluid at rupture
- Time to rupture
- No rupture
Step 5:
Are the acceptance
criteria for rupture
met?
No
Step 6:
Decide which pipe/equipment to fire insulate
or
Increase orifice diameter if available
capacity in the flare system, or reduce system
volume by relocation of sectionalization valves
or increase the flare system capacity or change
material quality or increase wall thickness.
Yes
Step 7:
Calculate the minimum design temperature (low-temperature design
temperature) of the blowdown section and the flare system tail pipe.
Figure 2. Follow this sizing procedure

to design depressurization orifices.
Is the
minimum design
temperature
acceptable?
No
Start depressurization at a higher

temperature (or change material).
Yes
The design of this section blowdown orifice
and fire insulation requirements is finished.
CEP
September 2002 www.cepmagazine.org
41
Safety
pressurization orifices, using the capacity of the flare system in the calculations. A recommended first estimate is an
orifice diameter that takes the pressure below the unacceptable rupture pressure within the typical time to rupture. The
initial pressure should be the highest normal operating pressure or an equalization pressure (settle-out pressure) for a
compression segment. A global fire should be used. The typical time-to-rupture can be set at that interval it takes to reach a
600800C wall-temperature, depending upon the wall thickness. A value of 510 min is typical for a dry wall exposed to
a medium-heat-flux jet-fire, with no depressurization.
One way of improving the safety of the plant is increase
the rate of depressurization, as the hazardous aspects of the
segment increase. The total blowdown rate can be kept unchanged by increasing the depressurization times of the
less hazardous segments. A segment containing large
amounts of light liquids (e.g., condensate or liquefied
petroleum gas (LPG)), those that will result in boiling-liquid expanding-vapor explosions (BLEVEs) are regarded as
a particularly hazardous section. In any case, there may be
limitations on maximum pressure gradients for items such
as compressors or gaskets.
Step 2: Add insulation, if required, and simulate the
pressure profile during depressurization when exposing the segment to the global fire. For the first iteration
only, omit this step and go to Step 3. A global fire will add
heat to the fire-exposed area without PFP. The initial pressure in this calculation should be equal to the highest normal operating or settle-out pressure. Credit for insulation
should be given only for PFP. Piping and equipment with
insulation used for purposes other than for PFP should be
regarded as uninsulated.
Unrealistic backpressure in the flare system may result
in a too-rapid simulated depressurization. The orifice backpressure should be based on the time-dependent simultaneous depressurization rates.
If a depressurization segment is 100% fire-insulated,
then the integrity of the insulation and supports usually determines the maximum allowable depressurization time,
which is typically 3060 min. Account for the integrity of
the insulation by extending the depressurization time for a
100% fire-insulated section. The reduced depressurization
rate for this section is used to allow for the increase of the
rate from a most-hazardous depressurization-section. A reduced depressurization rate may increase the fire duration,
if a leak in this section is the source of the fire.
Step 3: Simulate the temperature profile for all
piping and equipment in the depressurization segment when exposed to the local peak-heat flux. A jet
fire is normally used in these calculations, but the local
load for a pool fire should be used if the segment will
not be exposed to a jet fire. All piping means all pipes
with different diameters, pressure classes and/or material qualities. The temperature profile for one particular
pipe usually is rather insensitive to pressure changes
42
www.cepmagazine.org
September 2002
CEP
within a segment, i.e., the temperature profiles from the

first iteration can be kept throughout the whole iteration
procedure. A final update of the temperature profiles
must be performed prior to the last iteration.
Step 4: Calculate whether or not rupture occurs. Determine the stress or strain that all pipes and equipment are
exposed to for the temperatures and pressures seen during
the depressurization (Calculated in Steps 1 or 2, and Step
3) and determine whether the segment will rupture.
Two failure (rupture) criteria are often used: the maximum
stress or maximum strain (% elongation). The maximum
stress criterion is usually the UTS. Rupture strain is a matter
of definition. Strain calculations require finite-element modeling of the system, which is usually not performed during
this step-wise method. Such calculations should be performed for verification purposes during the final design.
The suggested approach is to calculate the stress
from the internal pressure and add extra stress (margins)
when calculating the longitudinal stress. The stresses of
importance for a pipe are the hoop stress caused by internal pressure, and the longitudinal stress. The longitudinal stress is the sum of axial stresses due to pressure;
the weight of the pipe, valves, fittings, branch pipes,
etc; stress due to reaction forces exerted on the pipe by
pressure; and stress due to thermal elongation of the
pipe. The equivalent stress (von Mises) is the stress to
be compared with the temperature-dependent UTS to
determine whether rupture occurs. The hoop stress,
hoop, is equal to:
hoop =
Pressure Outer dia.

2 Wall thickness
(2)
The longitudinal stress, long, is given by:

long = 1/2hoop + x
(3)
The equivalent stress is given by:

Von _ Mises = hoop 2 + axial 2 hoop axial
(4)
The term x in Eq. 3 represents all stress except for that

set up by the pressure. A piping engineer should be consulted when determining the value of x.
It is recommended that the UTS by reduced by 20% or
more, depending on the reliability of the UTS data. The
20% is a safety margin. Reduce the wall thickness by accounting for the mill tolerance. It must be assumed that the
lower mill tolerance is delivered. Reduce the strength by
including the weld factor. Again, a piping engineer should
be consulted.
Step 5: Check the rupture pressure against the acceptance criteria. If all piping and equipment in the de-
Required information prior to relief-calculation iteration

- Description of the fire scenarios (type of fire, duration,
heat fluxes, size)
- Relief segment geometry (system volume, area, weight, etc.)
- Ultimate tensile strength at elevated temperature of
materials in the relief section
- Acceptance criteria for rupture
Add insulation if required. Calculate the process segment

pressure profile. Use the fire with the largest heat input (kW).
Calculate the wall-temperature profile for all pipes and equipment.

Use the local fire with the highest heat flux (kW/m2).
Use the temperature profile to calculate the rupture pressure for

all pipes and equipment. Compare with the actual pressure.
Acceptance criteria:
- Pipe rupture pressure
- Equipment rupture pressure
- Released flammable fluid
at rupture
- Time to rupture
- No rupture
Are the
acceptance
criteria for rupture
met?
No
Decide which
pipe/equipment
to fire insulate.
the flare-system tail-pipe. This depressurization calculation should be performed without fire input to the section.
All planned types of insulation (not
only fire insulation) should be taken
into account. The initial temperature
should be the minimum ambient temperature or minimum operating temperature, whichever is lower. The initial
pressure is calculated from a cooldown
of the system down to the start temperature, prior to depressurization. The
cooldown calculation should be performed using the trip pressure from the
highest shutdown pressure. The minimum temperature in the flare tail-pipe
should be calculated with the depressurization segment as the only source to
the flare system.
Other considerations
Some key ones to note are:
The loss of bolt pre-tensioning due
to bolt elongation as a result of increasThe design of this section fire-relief valve
and fire-insulation requirements is finished.
ing temperature is important when
studying flange failure. The piping engineer should be consulted on this. Flanges
Figure 3. Sizing procedure for fire-relief valves is similar to that for orifices.
are recommended to be fire-insulated.
Lines in the flare system having
no flow during a fire depressurization (e.g., downstream
pressurization segment meet the acceptance criteria, then
pressure-control and pressure-relief valves) are usually
the fire insulation is completed. Go to Step 7 for low-temfire-insulated because: they are thin-walled and can heat up
perature calculation, otherwise go to Step 6 and add in inrapidly; the depressurization gas flowing in the flare syssulation. Alternatively, go back to Step 1 and increase the
tem does not cool these pipes; and the flare system will be
size of the orifice or increase the flare system capacity.
pressurized to a value near its design pressure, at the same
Step 6: Decide which piping/equipment to fire-insutime that the pipe temperature is high.
late. If any run of piping or piece of equipment does not
meet the acceptance criteria, then add PFP to one or more
Fire-relief-valve sizing procedure
of these components. It is recommended to add PFP to
Sizing of the fire-relief valves (Figure 3) should be
the corrosion-resistant pipe with the largest diameter. But,
performed with the same minimum requirement as specif there are pipes that are already insulated for reasons
ified at the beginning of the article. Also, the procedure
other than PFP, these should be fire-insulated first.
closely follows that for sizing depressurization orifices,
The reasons for choosing the pipe with the largest diwith the exception, of course, that the pressure will inameter are it is the most critical with respect to reaction
crease until the relief valve opens. The size of the relief
forces and pressure waves when it ruptures, and it will revalve should be such that the minimum relief-rate
quire the largest amount of insulation per length. Large
equals the liquid boil-off and gas-expansion rates. This
pipes are also cheaper to paint and insulate (per unit area)
avoids a pressure increase above the set pressure of the
than smaller ones. The reason for insulating the corrorelief valve.
sion-resistant pipes first is to avoid insulation on metals
A fire-relief valve will usually not protect a pressurthat corrode more easily. When partially insulating pipes,
ized system against rupture if the gas-filled part of the
it is preferable to add the covering on an area where the
system is exposed to fire. The fire-relief valve will norpossibility of a fire is largest and where inspection of the
mally protect against rupture if the flame is exposed to
insulation and pipe can be performed easily.
the wetted wall only when the boiling liquid on the inStep 7: Calculate the design low-temperature limitaside keeps the wall temperature at a reasonably low
tion of the depressurization segment (this is known as
value. For multicomponent mixtures, the temperature
the minimum design-temperature calculation) and in
CEP
43
Safety
Antisurge Valve
Blowdown
Orifice
To the Flare System

Blowdown
EV
Process
Stream
Sectionalization Valve
EV or XV
PSV = Pressure Safety Valve

EV = Emergency Valve
XV = Sectionalization Valve
To the Flare System
PSV
Cooler
Compressor
EV or XV
Process Stream
The real geometry of any type of process segment (e.g., the above system drawn in continous
lines) is transformed into a hypothetical cylindrical vessel (the vessel below in continuous line).
The hypothetical vessel is used in calculation of system pressure during depressurization and relief.
The diameter and length are increased until the volume equals the volume of the original system
(use the dominating diameter of the original system). The liquid level is adjusted to match the
liquid volume of the original system, "Add or subtract" wet and dry areas to match the wet and
dry area of the original system. Set a wall thickness equal to the dominating wall thickness of
the original system and adjust the weight until the weight matches the original sysem weight.
Hence, we have a hypothetical segment where the system area and weight do not fit the
hypothetical cylinder, but do fit the original system.
Figure 4. Modeling
the real process segment

as a hypothetical one
simplifies sizing procedures.
Blowdown valves open in

fire situations.
Process Stream
Scrubber
EV or XV
Sectionalization valves close

in emergency situations.
PSV
Flare System
Blowdown
EV
Flare System
Blowdown
Orifice
Length
Di
will increase as the lighter components evaporate, and

the wall will eventually reach the rupture temperature,
even if it is liquid-filled.
Modeling process segments

Here, we present an approach to modeling complex
depressurization and relief segments. The method models
the complete complex geometry by creating one hypothetical segment that represents the total system volume
and heat-transfer areas, and several sub-segments that
represent the real geometries of the segment.
The hypothetical segment is used for calculation of the
system pressure during depressurization or relief. The subsegments are used for calculation of the temperature response of each piece of piping and equipment within a process segment. The sub-segments are modeled with the
same geometric information that is required for a wall-temperature-response calculation namely, the sub-segment
wall thickness, segment or pipe diameter and inside fluid.
44
www.cepmagazine.org
September 2002
CEP
Hypothetical segment used for system pressure

The hypothetical segment is modeled with the real system volume, system outer-area, with and without PFP, system inside-area in contact with the gas, system inside-area
in contact with the liquid, and system weight (of piping
and equipment).
The hypothetical segment is modeled as a cylinder.
This shape is specified with a diameter equal to the mostdominating (volume) diameter of any item of piping or
equipment in the original segment. The associated wall
thickness for this diameter should be used. The length of
the cylinder is set such that its volume equals the volume
of the original segment. The liquid level is adjusted to obtain the actual liquid volume. The hypothetical cylinder
now represents the correct system volume. However, the
outer area of this cylinder must be corrected to the realsystem outer area by subtracting or adding gas and liquid
area to the segment. Therefore, heat transfer with the surroundings (fire, ambient air) will be modeled over the real
area. The system outside and inside areas may be different

for a high-pressure system (typically, 100150 bar), with
large amount of small-diameter piping (usually, 3 in. and
below), and should similarly be corrected to the real system inside area. This effect is largest for pipes with a corrosion allowance, that is, those made of carbon steel, generally. The dominating (by weight) metal quality should
be specified.
Unique geometry used for wall-temperature calculations. The temperature response of a pipe or piece of
equipment exposed to fire depends upon the metal wall
thickness, metal properties (e.g., heat capacity and conductivity) and the thermal mass of the inside fluid. Each
combination of these must be calculated individually to
determine the possible different wall-temperature responses of the system.
The exact inside and outside diameters (I.D.s and
O.D.s) should be used. The I.D. will determine the thermal mass of the inside fluid. The O.D. will enable calculation of the fire-exposed area. Any length can be used,
since the sub-segment volume and mass are proportional
to the length. The actual thickness, minus a corrosion allowance, should be used. The actual thickness is the
nominal wall thickness, minus the allowable tolerance of
the pipe thickness (the mill tolerance). The corrosion allowance should be accounted for after reduction by the
mill tolerance. Typical values of the mill tolerance are
1.53 mm for carbon steel, with the larger value being
Literature Cited
1. Institute of Petroleum, Guidelines for the Design and Protection
of Pressurized Systems to Withstand Severe Fires, Inst. of
Petroleum, U.K. (to be issued shortly).
2. Scandpower Risk Management AS, Guidelines for Protection of
Pressurised Systems Exposed to Fire, Scandpower AS, Norway
(www.scandpower.com/?CatID=1071) (May 13, 2002).
3. Stange, E., et al., Determination of Temperatures and Flare
Rates During Depressurization and Fire, paper presented at 72nd
Annual Gas Processors Association Convention, San Antonio, TX
(1993).
4. American Petroleum Institute, Guide for Pressure-Relieving and
Depressurizing Systems, API Recommended Practice (RP) 521,
4th ed., API, Washington, DC (1997).
5. Incropera, F. P., and D. P. DeWitt, Fundamentals of Heat and
Mass Transfer, 4th ed., John Wiley, New York (1996).
6. Health and Safety Executive, Joint Industry Project on Blast and
Fire Engineering of Topside Structures, OTI 92 596/597/598,
HSE, U.K.,(1991).
7. Norsok Standard, Technical Safety, S-001, Rev. 3, (www.nts.no/
norsok) (2000).
Acknowledgment
We would like to thank Erik Odgaard, Jan A. Pappas and Geir Johansen (Norsk Hydro, safety and piping discipline) for their helpful discussions.
more common. The mill tolerance is specified on the

pipe data sheet. When this tolerance is not specified by
the pipe supplier, consult a piping engineer.
The correct metal or alloy composition must be used to
obtain the correct material properties (i.e., heat capacity,
conductivity and the UTS). If the UTS is not available at
elevated temperatures, a preliminary tensile curve can be
made based on the UTS at 20C; this is usually available.
When the strength at an elevated temperature is known for
a material that is close in physical properties (the same
family of materials, such as two different carbon steels),
then the percentage difference in the UTS at 20C is kept
at all temperatures. That is, the new tensile strength curve
should have the same shape as the known curve. When the
tensile strength at elevated temperatures is not known for a
material in the same family, then the percentage difference
between the UTS and yield at 20C is reduced linearly between 20 and 1,000C. The UTS should be released by at
least 30% (not 20%) in such an approximation. When the
UTS is equal to the yield strength in the above calculation,
then the UTS should be set equal to the yield strength at
higher temperatures.
Concluding remarks
We believe that the calculation of longitudinal stress
represents the major challenge when performing depressurization and fire relief design. Modeling of time-dependent
fire characteristics (the extent and heat flux) also represents
a challenge, since the plant layout is usually unknown during the design stage. Yield and UTS data at temperatures
above 400500C often do not exist for the materials used
in the system.
PER SALATER is a principal engineer in Norsk Hydro ASA (N-0246 Oslo,
Norway; Phone: +47 22 53 76 91; Fax: +47 22 53 95 37; E-mail:
per.salater@hydro.com). He has ten years of experience as a process and
system engineer for Norsk Hydros North Sea oil and gas facilities. His areas
of expertise are system design, heat-exchanger thermal design and design
of depressurization systems. He has with Sverre Overaa co-patented
(PCT/NO99/00123, U.S. Patent 09/673467) a design that eliminates the
flare system for oil-and-gas processing facilities and replaces it with a
blowdown header connected to storage. He holds an MSc in mechanical
engineering from the Norwegian Institute of Technology, Trondheim.
SVERRE J. OVERAA is a principal engineer at Norsk Hydro ASA (Phone: +47 22
53 81 00; Fax: +47 22 53 27 25; E-mail: Sverre.J.Overaa@hydro.com). He is a
member of the GPA Technical Section F Research Committee. Overaa has 18
years of experience as a process and systems engineer for Norsk Hydros
North Sea oil-and-gas facilities and is currently head of technical systems
for the Ormen Lange project under development by Norsk Hydro. His areas
of expertise are simulation, fluid properties and system design. He has with
Per Salater co-patented PCT/NO99/00123, U.S. Patent 09/673467.
ELISABETH KJENSJORD is a process engineer with Norsk Hydro ASA (Phone:
+47 22 53 81 00; Fax: +47 22 53 27 25; E-mail: elisabeth.kjensjord@
hydro.com). She is involved in process design of Norsk Hydros oil-and-gas
facilities and is presently engaged in the development of the Grane oil field
located on the Norwegian Continental Shelf in the southern part of the
North Sea. She holds an MSc in chemical engineering from the Norwegian
Univ. of Science and Technology.
CEP
45
Pipes exposed to medium sized jet fires - rupture conditions and models for
predicting time to rupture
Author:
Per Salater
Co-author: Sverre J Overa
Norsk Hydro ASA, Oslo, Norway
Presented at FABIG, London and Aberdeen, January 2004 and Houston, March 2004
Introduction
This article presents data from experiments on pressurized steel pipes exposed to medium sized
jet flames together with engineering models for predicting time to rupture.
In 1996 and 1997 Norsk Hydro, Norways second largest oil company, performed several
medium scale jet-fire tests on pressurized pipes. Pipes were pressurized with nitrogen to
approximately 85-90% of the design pressure of the pipes and then exposed to a 170 - 190 kW/m2
jet-fire (incident heat flux). The pressure was kept constant during the tests, and internal wall and
gas temperatures were measured until the pipes ruptured. Some of the test results are presented
and discussed in this article.
The tests were performed to verify Norsk Hydros engineering methods /1/ for calculation of time
to pipe rupture. The Norsk Hydro methods have later been included in the Institute of
Petroleums Guidelines for the design and protection of pressurized systems to withstand severe
fires /2/ and the Norwegian Guideline for Protection of Pressurized Systems Exposed to Fire
/3/. The tests, together with several cold (without fire) depressurization tests of both large process
systems and small vessels, have confirmed that Norsk Hydros process simulation tool NEW*S
/4/ is an engineering tool that reproduces experimental depressurization and fire tests. Norsk
Hydro has since 1992 used NEW*S in the design of depressurization orifices and fire relief
valves on its oil and gas installations. NEW*S has also been used to determine the required
amount of fire insulation and the minimum design temperatures for pipes, vessels and flare
systems. At present NEW*S is used in the design of the Ormen Lange onshore facility, a
gas plant (70 MSm3/day) to be built on the west cost of Norway with start up in 2007. NEW*S
supports most of the requirements given in reference /2/ and /3/.
The jet fire test rig
The tests were performed by SINTEF (Trondheim, Norway). The test rig is illustrated in Figure
1. See also Photo 1. The jet-fire was a 6 MW propane flame. The propane flow rate was 0.13 kg/s
and the incident heat flux from the jet-fire was between 170 and 190 kW/m2 at its highest and
150 kW/m2 in average for the entire circumference of the axial center of the pipe. In comparison,
the local, peak heat flux of the severe jet-fires in reference /2/ and /3/ are 250 kW/m2 for flow
rates between 0.1 and 2 kg/s and 350 kW/m2 for flow rates larger than 2 kg/s. The fires in
reference /2/ and /3/ are optical thick flames, i.e. there is no radiant heat exchange between the
pipe (object) and the cold surroundings outside the flame. When comparing the experiments with
the calculated temperatures, it can be concluded that an optical thick flame was achieved for the
small diameter pipes ( d 8), but not for the large diameter pipes ( t 10).
~ 6 meters
~ 1 meter
3.5 meters
Test pipe
Propane lance
Jet fire
Figure 1:
Piperack
The test rig
Photo 1:
The test rig
The length of the test pipe sections was 3 meter. The pipe was free to expand on the pipe rack,
hence no stress due to thermal expansion was created and the pipe stress was primarily due to the
pressure.
The jet-fire was an open fire. A windshield wall was placed on one side of the test rig. The jetfire was approximately 1 meter in diameter at the pipe location. The flame length was
approximately 6 meters and the distance from the propane lance to the pipe was approximately
3.5 meters.
The fire tests were carried out outdoors and the ambient conditions (varying wind and
temperature) influenced to some degree the fire heat flux.
For all the tests the inside wall temperature was measured at 4 locations over the circumference
of the axial center of the pipe inner wall as illustrated in Figure 2. The center of the pipe was also
the center of the jet-fire. These 4 temperatures are reported in Figure 3 to 9. A pressure control
valve was installed on all the pipes, and the pressure was kept constant ( r 3 bar) during the tests.
3 mete
rs
"Top"
Gas top
Pipe and flame centre

O
90
"Back"
180O
Flame direction
0O
"Front"
270O
Gas bottom
"Bottom"
Piperack
Figure 2:
Location of thermocouples. The location of the inner wall thermocouples was

identical for all the tests.
For half of the tests several more thermocouples were installed. In these tests the internal gas
temperature and inner wall temperatures were measured in the axial direction from the center of
the pipe. 6 out of 16 tests are presented in detail in this article.
Table 1:
Tests that are presented in detail in this article.

Pipe material Pipe design pressure
Pressure during test
Pipe dimension
(barg)
(barg)
(inch)
Test 1
Carbon steel
100
1 *)
10
Test 2
Carbon steel
50
46
4
Test 3
Carbon steel
450
98
8
Test 4
Carbon steel
50
48
10
Test 5
Carbon steel
100
82
10
Test 6
22 Cr Duplex 100
87.5
10
*) Test performed to measure the temperature difference between the inner and outer wall
Temperature difference between the inner and outer wall
The temperature difference between the inner and outer wall was measured only in one initial
test. This test was run at atmospheric pressure on a 10, 100 bar carbon steel pipe. The
temperature gradient was 3.4oC/mm at its maximum at the location where the highest temperature
was measured. At this temperature gradient the radial heat flux in the wall is calculated to 130135 kW/m2. The total heat flux was somewhat higher since there was heat transfer also in the
axial direction and in the circumferential direction.
The radial temperature gradient is presented in Figure 3. It increases rapidly up to 3oC/mm, then
there is a further increase up to 3.4oC/mm before it starts to decline. The increase from 3 to
3.4oC/mm may be due to a change in wind (flame) condition. The temperature difference would
have declined towards zero if the test had been prolonged until the wall temperature had
stabilized, i.e when the net heat input to the pipe (incident heat from the fire minus re-radiation
from the pipe) became zero. As can be seen from Figure 3 the temperature was still increasing
when the experiment was terminated.
Inner wall temperature
Radial temperatur gradient in wall

800
3.5
700
Temperatur (Celsius)
Kelvin/mm
3
2.5
2
1.5
1
600
500
400
300
200
100
0.5
0
0.0
1.0
2.0
3.0
Time (minutes)
Figure 3:
4.0
5.0
6.0
0.0
1.0
2.0
3.0
4.0
5.0
Time (minutes)
Differential temperature between inner and outer wall + inner wall temperature,
test #1
6.0
Temperatures
The tests illustrated in Figures 4, 6, 8, 9 and 10, all terminated with a pipe rupture and the time to
rupture is at the end of each curve.
The presented wall temperatures are measured at the inner wall of the pipe. The outer wall
temperature is higher than the presented temperatures, but the difference will, as explained above,
approach zero as the temperature increase pr time (dT/dt) approaches zero. Hence, the measured
inner wall temperature at rupture is close to the outer wall temperature in most experiments. The
calculated temperatures are mean wall temperatures, i.e. average temperature in the radial
direction. The calculated temperatures are presented together with the measured temperatures in
the figures.
Temperature (C)
4", 50 bar design pressure, Carbon steel

Test run at 46 barg
1000
900
800
700
600
500
400
300
200
100
0
Top
Calculated
Calculated
Wall top
Wall front, 100
mm left of centre
Wall bottom
Rupture
Figure 4:
Back
Bottom
6
8
10
Time (min)
12
Wall back
14
16
Temperature profiles for the 4 carbon steel pipe of 50 bar design pressure
operating at 46 barg, test # 2
For all the experiments the highest heat flux and wall temperature was located at position 0o (see
Figure 2), i.e. at the side of the pipe facing towards the source of the jet fire. The second highest
temperature was in most experiments at the top of the pipe (90o) and the lowest at the bottom of
the pipe (270o). In test 4 (see Figure 6) it is probable that the jet-fire was lower compared to the
other tests since the bottom temperature is the second highest. The circumferential position of the
highest heat flux would have moved if the pipe had been moved closer to the jet flame, hence it
cannot be concluded that the highest wall temperature always will be towards the source of the jet
fire.
Temperature (C)

Test run at 98 barg
1000
900
800
700
600
500
400
300
200
100
0
Calculated
Front
Back
Calculated
Bottom
Wall top
Top
Wall front
Wall bottom
Wall back
0 2 4 6 8 10 12 14 16 18 20 22 24 26 28 30 32
Time (min)
Figure 5:
operating at 98 barg, test # 3. This pipe did not burst.
The temperature increase is highest at the start of the experiments and then decreases until the
pipe ruptures or the flame is shut off. For the small diameter pipe in test # 2 the temperature
increase is initially 7.4oC/sek, for test # 4 5oC/sek and for test # 3 1.8oC/sek. The difference is
approximately a linear function of the pipe outer area-weight ratio. Table 2 summarizes the
measured wall temperature responses. At 6-700oC the Ultimate Tensile Strength (UTS) the
material strength - of typical steel material is largely reduced.
Table 2:
Summary of measured wall temperature responses
Temperature
Time to raise the wall temperature for a 175 kW/m2 jet fire
[min:sec]
4, 50 bar,
10, 50 bar, 10, 100 bar,
10, 100 bar,
8, 450 bar,
carbon steel
carbon steel 22Cr Duplex
carbon steel
carbon steel
*)
200oC
0:23
0:42
0:54
1:04
2:18
o
400 C
1:02
1:57
2:12
3:00
4:44
600oC
2:10
4:06
4:14
6:22
8:27
o
800 C
5:14
8:00
17:03
*) This was a 190 kW/m2 fire

Test run at 48 barg
1000
900
Calculated
Temperature (C)
800
Front
Top & back
700
Bottom
600
Calculated
Wall top
Gas top
500
Wall front
Wall bottom
400
Gas bottom
300
Wall back
Gas top
Gas bottom
200
Rupture
100
0
0
10 11 12 13
Time (min)
Figure 6:
The highest internal gas temperature is measured at the top of the horizontal pipe, the lowest at
the bottom (see Figure 6). The temperature difference is as high as 200oC, indicating a large free
convection heat transfer inside the pipe. This temperature stratification is seen in all the
experiments, and is also what we have measured in our several cold blowdown experiments,
some of them presented in /5/. The free convection heat transfer term must be included in the heat
transfer calculations.
The fire model used in the wall temperature profile calculations is:
Qnet = Ds HfV Tr4 + h(Tf - Ts(t)) - HsV Ts(t)4
(1)
The first term is the radiation heat from the flame to the pipe, the second term is the heat
convection from the flame gases and the last term is the re-radiation from the pipe.
Qnet
Hs
Ds
Hf
V
Tr
Tf
net (absorbed) heat transfer to the pipe (W/m2)

emissivity of the pipe material (-)
absorbtivity of the pipe material (-)
emissivity of flame (-)
Stefan-Boltzmanns constant = 5.6710-8 (W/m2 K4)
radiation temperature of flame (K)
flame gas temperature (K)
Ts(t)
h
surface temperature of the material (K)

convective heat transfer coefficient (W/m2K)
This fire model is equal to the fire model in /1/, /2/ and /3/. When comparing calculations with
measured temperatures, parameters for equation 1 were chosen to fit a jet-fire with initial incident
heat flux of 175 kW/m2 and an initial absorbed heat flux of 145 kW/m2. The parameters used are:
x
x
x
x
x
Flame radiation temperature

980oC *)
Flame gas temperature
980oC *)
Flame emisivity
0.7
Pipe emisivity and absorbtivity
0.7
Convective heat transfer coefficient
80 W/m2K
*) The fire was premixed for the duplex pipe test. 1020oC is used in the calculation of this
test, resulting in a 191 kW/m2 jet-fire (incident heat).
These parameters are within the range of the typical values suggested in /1/ and /2/. In Figure 7
the resulting net (absorbed) heat transfer to the pipe is given as a function of pipe temperature.
Note that Equation (1) also can be used to reproduce the open pool fire used in API RP 521. This
is discussed in /1/.
Net (absorbed) Heat

Flux [kW/m2]
The fire model used to reproduce the

experimental medium scale jet fire
175
150
125
100
75
50
25
0
0
100
200
300
400
500
600
700
800
900
1000
Wall temperature [Celsius]
Figure 7:
Net heat flux to pipe using the above fire parameters
As seen in Figures 4, 5, 6, 8 and 9, the calculated wall temperatures with the above chosen
parameters are close to the measured temperatures for the small diameter pipes, but too high for
the large diameter pipes at temperatures above 600oC. This is partly due to the fact that the fire
model assumes an optical thick flame, i.e. there is no radiant heat exchange between the pipe
(object) and the cold surroundings outside the flame.. For the 10 pipes, the radiation to the cold
surroundings seem to dominate more and more above 600oC. For the small diameter pipe this is
not the case to the same degree since the diameter is small compared the diameter of the jet-fire.
Another reason is that the flame temperature increases towards the center of the flame. For the
10 pipe the average flame temperature at the wall location was colder than for the 4 test. The
highest measured wall temperature was observed on a 1 pipe, the second highest on a 2 pipe.
Temperature (C)
Although the fire temperature at the wall location varied for the different pipe sizes, the same fire
temperature is used in all the wall temperature calculations. By changing the fire temperature, a
calculated wall temperature closer to the measured temperature will be achieved for the 10
pipes. A constant fire temperature is used here since this normally is the case in engineering
calculations.

Test pressure = 82 barg
1000
900
800
700
600
500
400
300
200
100
0
Calculated
Front
Back
Bottom
Wall top - failed
Gas top
Wall front
Wall bottom
Wall back
Gas bottom
Gas top
Rupture
Figure 8:
Calculated
Gas bottom
10 12 14 16 18 20 22
Time (min)
Temperature profile for the 10 carbon steel pipe of 100 bar design pressure
10
Temperature (C)
10", 100 bar design pressure, 22Cr Duplex

Test run at 87.5 barg
1000
900
800
700
600
500
400
300
200
100
0
Calculated
Top
Back
Calculated
Bottom
Wall top
Wall front, 100mm
left of center
Wall bottom
Wall back
Rupture
6 7
9 10 11 12 13
Time (min)
Figure 9:
Temperature profile for the 10 duplex pipe of 100 bar design pressure, operating
at 87.5 barg, test # 6
Time to rupture
Photo 2:
During and after test (50 bar, 10, carbon steel)
The time to rupture for a specified stress level is highly dependant on the pipe wall thickness see equations (2) to (4). The pipe wall thickness is usually not equal to the nominal wall thickness
since pipes are specified with a mill tolerance. Carbon steel pipes normally have additional
corrosion allowance. The carbon steel pipes in the experiments were specified with 3 mm
11
corrosion allowance. The pipes were new and the entire corrosion allowance was intact. The
duplex pipe had no corrosion allowance.
The mill tolerance is often specified as r 12.5% of the nominal wall thickness, and the pipe
producers are allowed to utilize this tolerance and thus produce a pipe thickness that can be as
low as 87.5% of the nominal wall thickness. The lower end of the wall thickness is normally seen
for the expensive stainless steel materials, typically -10% since the pipe producers need some
margin to 12.5% themselves. For carbon steel the mill tolerance varies between r 12.5%.
For all the experiments the pipe ruptured at the location of the highest temperature, i.e. at location
0o (see Figure 2). Measured time to rupture is presented in Table 3. Also presented is calculated
time to rupture when using the highest measured wall temperature, varying wall thicknesses and
the pipe failure criteria described below.
Table 3:
Time to rupture measured and calculated. The measured temperatures are used
in the calculations **).
Pipe material
Carbon steel, 4
50 bar, Test # 2
Carbon steel, 8
450 bar, Test # 3
Carbon steel, 10
50 bar, Test # 4
Carbon steel, 10
100 bar, Test # 5
22 Cr Duplex, 10
100 bar, Test # 6
Measured
Calculated
using the
nominal wall
thickness
Calculated
wall thickness
reduced with 10%
mill tolerance
(Minutes)
(Minutes)
(Minutes)
Calculated
wall thickness
reduced with 3
mm corrosion
allowance
**)
(Minutes)
Calculated
wall thickness
reduced with 3 mm
corr. allowance and
10% mill tolerance
(Minutes)
12.8
12
7.5
3.3
>25 *)
>25 *)
>25 *)
>25 *)
>25 *)
10.1
9.2
8.6
7.4
6.6
18.8
18.8
15.2
13.6
12.4
10
8.7
7.9
Not applicable
Not applicable
*) Did not burst and rupture is not calculated.

**) The measured temperature profile would have been steeper if the corrosion allowance had
been removed. Table 4 presents time to rupture when using the calculated temperature profile for
a pipe without corrosion allowance.
The time to rupture is calculated as proposed in /3/ where the stress is calculated from equation
(4) and compared with the material Ultimate Tensile Strength (UTS). In /3/ the recommended
pipe failure criteria is when the equivalent stress is larger than the UTS.
( p OD)
V hoop
(2)
Hoop stress:
2w
( p OD)
V axial
V ext V displ
Longitudinal stress:
(3)
4w
Equivalent stress:
V hoop 2 V axial 2 V hoop V axial
(4)
12
where
V hoop
V axial
V ext
V displ
is the hoop stress (MPa)
p
OD
w
is the pressure (MPa)

is the outer diameter (meter)
is the wall thickness (meter)
is the axial stress (MPa)

is the longitudinal stress due to external loads (weight of pipe and valves etc ), (MPa)
is the longitudinal stress due to thermal expansion (MPa)
During all the tests V ext has been close to zero since the pipe span was only 3 meter. V displ was
zero.
Figure 10 illustrates the calculated time to rupture using the specified failure criteria from /3/ and
the temperature measured at location 0o (see Figure 2). The UTS is reduced as the temperature is
increased. The UTS is taken from /3/ and corrected for the specified UTS from the pipe
datasheet. The procedure for correcting the UTS is specified in /3/. The pipe rupture is predicted
where the curves cross.
As seen in table 3 and the above temperature measurements, a small diameter low-pressure pipe
with the corrosion allowance intact can withstand higher temperatures due to greater wall
thickness. By using the failure criteria from /3/, the rupture temperature for a 4, 50 bar pipe is
reduced from 920oC to 730oC and the time to rupture is reduced from 12.8 minutes to 3.3 minutes
when the wall thickness is reduced with the corrosion allowance. It is therefore important not to
include the corrosion allowance when performing fire depressurization and relief design since the
corrosion allowance will eventually not be there. Note that the importance of the corrosion
allowance is reduced with increasing design pressure (wall thickness). As an example rupture
condition was not reached for the 1 and 2, 50 bar pipes operating at 45 bar both pipes with
the corrosion allowance intact and reaching a wall temperatures above 1000oC at the end of the
test.
From the curves in Figure 10 one can see that when the UTS curve is steep around the time of
rupture, a small error in the calculated pipe stress is of minor importance to the estimated time to
rupture. If however, the UTS decreases slowly with time prior to rupture, a small error on the
pipe stress causes a large error in the estimated time to rupture. Also, as rupture takes place at
high temperature, correct UTS data is more important at high temperature than at low
temperatures.
13
Stress in pipe v.s. UTS

4", 50 bar, carbon steel, 46 barg operating pressure
Stress in pipes v.s. UTS

200
200
175
175
Rupture
Rupture
150
125
Stress (MPa)
Stress (MPa)
150
100
75
125
100
75
50
50
25
25
0
0
10
12
14
0
0
Time (minutes)
6
8
Time (minutes)
10
12

10", 100 bar, 22Cr Duplex, 87.5 barg operating pressure

200
200
175
175
Rupture
150
Stress (MPa)
150
Stress (MPa)
125
100
75
125
100
75
50
50
25
25
0
0
0
Figure 10:
8
10 12
Time (minutes)
14
16
18
20
10
Time (minutes)
Calculated rupture compared with actual rupture when using the failure criteria
from /3/. Rupture is calculated when the curves are crossing. Actual rupture is at
the far right end of the horizontal curves.
Legend for figure 10:

- Blue (declining curve) = UTS as a function of measured temperature.
- Red (lowest horizontal) = Stress in pipe, corrosion allowance intact, wall thickness is not
reduced with the mill tolerance, i.e. the nominal wall thickness.
- Yellow (second lowest horizontal, highest horizontal for duplex) = Stress in pipe,
corrosion allowance intact, wall thickness reduced with the mill tolerance.
- Turquoise (second highest horizontal) = Stress in pipe, wall thickness reduced with the
corrosion allowance (3mm), wall thickness is not reduced with the mill tolerance.
- Green (highest horizontal) = Stress in pipe, wall thickness reduced with the corrosion
allowance (3mm) and the mill tolerance.
12
14
From Figure 10 one can see that the calculated time to rupture is either correct or slightly
conservative for the carbon steel pipes for which the pipe stress was most likely given by the red
curve (the lowest horizontal curve). For the duplex pipe the yellow pipe stress curve (the highest
horizontal) is the most likely curve, and the calculation is still good though more conservative.
Revision 2 of the Norwegian Guideline for protection of Pressurized Systems Exposed to Fire
/3/ will be available in February 2004. The stress equations will be revised in this issue. The
revised equations give slightly better calculated results. The guideline is available at no cost from
Scandpower Risk Managements web page, www.scandpower.com/?CatID=1071
The measured and calculated temperatures in Table 3 and Figure 10 are based on pipes with the
corrosion allowance intact. If the temperature calculations are performed for the same pipes but
with the corrosion allowance removed, the temperature profile will be steeper and the time to
rupture shorter, see Table 4.
Table 4:
Pipe
Time to rupture for pipes without corrosion allowance. The temperature profile is
calculated after removing 3 mm wall thickness.
Calculated
Calculated
175 kW/m2 jet-fire
250 kW/m2 jet-fire *)
[Minutes]
[Minutes]
Carbon steel, 4
1:50
1:10
50 bar, Test # 2
Carbon steel, 10
3:30
2:10
50 bar, Test # 4
Carbon steel, 10
7:40
4:30
100 bar, Test # 5
*) The jet-fire heat flux recommended in reference /2/ and /3/ for an optical thick jet-fire of leak
rate between 0.1 and 2 kg/s.
From this one can conclude that automatic and rapid depressurization at fire detection is essential
to avoid pipe ruptures for thin walled pipes, i.e. pipes with design pressure up to 100 bar.
Spurious fire detection must be avoided by a proper design of the fire detection system.
15
Geometry of pipes after rupture
Photo 3:
Typical pipe geometries after rupture. None of the pipes that ruptured created pipe
fragments. Left: 50 bar, carbon steel, 10. Right: 100 bar, carbon steel, 10.
Table 5:
Test
Pipe geometry after rupture (approximate numbers).

Bend angle
Reduced wall
Size of hole
thickness at rupture relative to the
pipe diameter
location (%)
o
Carbon steel, 4
14
33
~ 6xD
50 bar, Test # 2
Carbon steel, 4,
17o
72
~ 3xD
100 bar, Test # 7
22 Cr Duplex, 10
90o
61
~ 6xD
100 bar, Test # 6
The carbon steel pipe for test 7 is of a more ductile material than test 2. The carbon content is
0.12% and 0.17%, respectively. In test 7 the ductile behaviour was observed as a sudden increase
in diameter just before rupture. This was also seen as a reduced system pressure in the last few
seconds.
For all the tests that ended with a rupture, the size of the hole was more than 2 times the internal
diameter of the pipe. Hence, when calculating the amount of gas released through the hole this
must include flow from both directions towards the leak hole. A good estimate of the initial leak
rate is two times the flow rate through an orifice with a diameter equal to the internal pipe
diameter.
A repeat test was run for all pipes. The rupture temperature and geometry was in most cases
unchanged. None of the pipes that ruptured created pipe fragments.
The 10 duplex pipe rupture caused stronger impact on the test rig than the 10 carbon steel
pipes.
16
An engineering approach to the calculation of time to rupture
A to the point engineering approach to fire relief and depressurization calculation is presented
in /1/. For a more comprehensive description of the topic see /2/ or /3/ of which guideline /3/ is
more useful in engineering calculations since it gives more recommendations than /2/ and stress
equations and material data are also presented.
In brief, the engineering procedure is:
x Determine the fire scenario (type of fire, heat flux, fire duration and extent of fire)
x Determine the system volume, areas and weight
x Determine the acceptance criteria for rupture (what is unacceptable, i.e. maximum
pressure at rupture, minimum time to rupture, maximum release of hydrocarbons).
x Determine the UTS at elevated temperatures for all pipe and equipment materials in the
system.
x Calculate the pressure profile of the system. Use this pressure profile in the stress
calculations.
x Calculate the temperature profile for all the pipe dimensions and equipment in the
depressurization/relief segment. Use this temperature profile in the calculation of the
maximum allowable stress (UTS). The pressure and temperature calculations should be
calculated simultaneously to speed up the calculations. The temperature calculations must
include the heat transfer to and the thermodynamic (expansion or compression) of the
inside fluid
x Compare the pipe stress with the maximum allowable stress for all pipe dimensions. The
same to be performed for the equipment.
x If a rupture is calculated and this is an unacceptable rupture, change the design (add fire
insulation, increase size of depressurization orifice, change material, increase system
design pressure or combine any of these measures) and update the pressure profile if
necessary.
Since 1992 Norsk Hydro has used the in-house process simulation tool NEW*S when performing
depressurization and relief calculations. NEW*S is also used to determine the fire duration by
performing simultaneous depressurization of a segment to the flare system and to the surrounding
through a leaking point. NEW*S is also used when determining minimum design temperatures of
depressurization segments. The Norsk Hydro philosophy has always been to verify the models
used through experiments, preferably controlled, large scale experiments. These experiments
have shown the importance of employing correct thermodynamic and heat transfer models when
performing depressurization calculations. The experiments and 12-15 year of design experience
(offshore and onshore plants) have enabled us to determine what is important and what is less
important. A lot of this experience has been made available in /1/, /2/ and /3/. We have seen from
our latest offshore and onshore projects that by using these guidelines, an unexperienced but
competent process or safety engineer with some piping knowledge, can produce results with good
quality if given a simulation tool that can carry out correct pressure and temperature calculations.
Acknowledgement
The authors would like to thank Elisabeth Kjensjord for helpful input and comments.
17
References
/1/
/2/
/3/
/4/
/5/
Size Depressurization and Relief Devices for Pressurized Segments Exposed to Fire,
Salater, Overa and Kjensjord, CEP (AIChe), (September 2002)
Guidelines for the design and protection of pressurized systems to withstand severe
fires, The Institute of Petroleum, (March 2003)
Guideline for protection of Pressurized Systems Exposed to Fire, Rev 1, Norsk Hydro,
Statoil, Scandpower Risk Management, (May 13 2002). Available at no cost from
www.scandpower.com/?CatID=1071
Developing, implementing and verifying an Engineering Tool, Wilson, Overa, Stange
and Majeed, 69th Annual GPA Convention, (March 11-12, 1991)
Determination of Temperatures and Flare Rates During Depressurization and Fire,
Overa, Stange and Salater, paper presented at the 72nd Annual Gas Processors Assosiation
Convention, San Antonio, TX, (1993)

Depressuration Report

Загружено:

Сведения о документе

Авторское право

Доступные форматы

Поделиться этим документом

Поделиться или встроить документ

Параметры публикации

Этот документ был вам полезен?

Это неприемлемый материал?

Авторское право:

Доступные форматы

Depressuration Report

Загружено:

Авторское право:

Доступные форматы

Depressurisation Report

ISSUED FOR REVIEW

Design rates and capacities to the flare systems.

Sizes and sizing basis for major blowdown devices.

Sizes of depressurisation lines in the flare system.

Uncertainties and areas for further work.

Safety Analysis Flow Diagram, Oil Separation

Safety Analysis Flow Diagram, Oil Export

Safety Analysis Flow Diagram, Gas Recompression

Safety Analysis Flow Diagram, Gas Treatment And Export

Safety Analysis Flow Diagram, Flare, Closed & Open Drain

Safety Analysis Flow Diagram, Fuel Gas System

Utility Flow Diagram, Flare System

Emergency Shutdown Valve

Floating Production Storage Offloading (Vessel)

International Standard Organisation

Knock Out Drum

Level Alarm High

Level Alarm Low

Level Switch High High

Level Switch High High

Fluid properties and process simulation program from Bubblepoint AS

Pressure Alarm High

Plant Design Management System from Aveva.

Passive Fire Protection

Pressure Switch High High

Pressure Switch Low Low

Process Shutdown System

Pressure Safety Valve (Relief valve)

Piping and Instrument Diagram

Safety Shut Down

Utility Flow diagram

Table 4-1 Criteria for BDV.

Calculation Design Basis

Temperature, ref. Table 4-2.

Pressure, ref. Table 4-2.

Vessel liquid level, ref. Table 4-2.

Heat input is 100 kW/m2

Table 4-2 Initial conditions, depressurisation

Initial liquid level

Normal operating pressure Based

Calculation results fire case

Equipment and piping integrity during fire

HOLD - will be part of detail engineering work.

NEW*S will be used for calculations outlined in section 5.2.

Normal operating temperature

High high liquid level in separators

High high liquid level in scrubbers

The heat input is 100 kW/m (Global fire, ref /4/)

Cold case minimum design temperature

Pipe work metal and volume excluded.

Use LSLL in vessels. LSHH to be checked as well.

Use the orifice size calculated in Fire Case

DEPRESSURISATION AND COLD TEMPERATURE RESULT

6.1.1. HP3-20-RO0010 1st Stage Separator

Total vessel volume, m3

Initial flow rate, kg/h

14 minute to reach 7.9 bara

Operating pressure, bara

Calculated minimum temperatures:

Minimum metal temperatures:

Vessel (liquid space),

6.1.2. HP4-20-RO0024 2nd Stage Separator

Total vessel volume, m3