Gp 44-25 Guidance On Practice For Depressurisation 6w2gv

Document No.

GP 44-25

Applicability

Group

Date

DRAFT 3 January 2007

Guidance on Practice for Depressurisation

GP 44-25

BP GROUP

ENGINEERING TECHNICAL PRACTICES

DRAFT 3 January 200

GP 44-25 Guidance on Practice for Depressurisation

Foreword This is the first issue of Engineering Technical Practice (ETP) BP GP 44-25. This Guidance on Practice (GP) is based on parts of heritage documents from the merged BP companies as follows:

BP RP 44-4 37

Guide to Depressurisation. BP Engineering Code of Practice 37 – Guide to Depressurisation.

Copyright  2007, BP Group. All rights reserved. The information contained in this document is subject to the and conditions of the agreement or contract under which the document was supplied to the recipient’s organization. None of the information contained in this document shall be disclosed outside the recipient’s own organization without the prior written permission of BP Group, unless the of such agreement or contract expressly allow.

ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 2 of 38

DRAFT 3 January 200


Table of Contents Page Foreword ........................................................................................................................................ 2 Introduction ..................................................................................................................................... 6 1.

Scope .................................................................................................................................... 7 1.1. General....................................................................................................................... 7 1.2. Objective..................................................................................................................... 7 1.3. General requirements ................................................................................................. 7 1.4. Calculation methods ................................................................................................... 8

2.

Normative references............................................................................................................. 8

3.

and definitions............................................................................................................. 9 3.1. ......................................................................................................................... 9 3.2. Definitions................................................................................................................... 9

4.

Symbols and abbreviations .................................................................................................. 10

5.

General................................................................................................................................ 10

6.

Operational vapour depressurisation ................................................................................... 11 6.1. General..................................................................................................................... 11 6.2. Plant and equipment ................................................................................................. 11 6.3. Pipelines ................................................................................................................... 11

7.

Emergency vapour depressurisation.................................................................................... 12 7.1. General..................................................................................................................... 12 7.2. Vessels, aboveground pipework, and valves............................................................. 12 7.3. Compressors ............................................................................................................ 12

8.

Depressurisation requirements ............................................................................................ 13 8.1. General..................................................................................................................... 13 8.2. Emergency shutdown system ................................................................................... 13 8.3. Emergency shutdown valves .................................................................................... 14 8.4. Emergency depressuring (EDP) system ................................................................... 15 8.5. Emergency depressuring valves ............................................................................... 16

9.

Application of depressurisation systems .............................................................................. 16 9.1. General..................................................................................................................... 16 9.2. Manned production platforms and floating production facilities.................................. 17 9.3. Unmanned production platforms ............................................................................... 17 9.4. Onshore gas/condensate plants................................................................................ 17 9.5. Hydrotreating/hydrocracking reactors........................................................................ 18

10.

Time for depressurisation..................................................................................................... 18 10.1. Fire case................................................................................................................... 18 10.2. Non-fire case ............................................................................................................ 19 10.3. Stopping depressurisation......................................................................................... 19 ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 3 of 38

DRAFT 3 January 200


11.

Methods of depressurisation ................................................................................................ 19 11.1. Uncontrolled depressurisation................................................................................... 19 11.2. Depressurisation by zone.......................................................................................... 19 11.3. Controlled depressurisation ...................................................................................... 20 11.4. Depressurisation flow rates....................................................................................... 20 11.5. Draindown................................................................................................................. 20

12.

Effects of depressurisation................................................................................................... 21 12.1. Auto-refrigeration ...................................................................................................... 21 12.2. Hydrates and ice ....................................................................................................... 22

13.

Repressurisation.................................................................................................................. 22

Annex A (Normative) Background to the selected depressurisation time ...................................... 23 A.1. Depressurisation purpose .................................................................................................... 23 A.2. Depressurisation systems designed for pool fire exposure................................................... 23 A.3. Depressurisation systems designed to minimize leak size ................................................... 26 A.4. Calculation of depressurisation mass flow rates................................................................... 26 A.5. API 521 guidelines ............................................................................................................... 27 Annex B (Normative) Methods for estimating the minimum wall temperature of depressurised vessels and pipework........................................................................................................... 33 B.1

Proposed methods............................................................................................................... 33

B.2

General assumptions ........................................................................................................... 33

B.3

Vessel contents ................................................................................................................... 33

B.4

Method 1.............................................................................................................................. 33

B.5

Method 2.............................................................................................................................. 33

B.6

Method 3.............................................................................................................................. 33

B.7

Method 4.............................................................................................................................. 33

Bibliography .................................................................................................................................. 33

List of Tables Table A1 - High temperature tensile properties for typical carbon steel (1) ................................... 30 Table A2 - High temperature tensile properties for 18-8 stainless steel (1) ................................... 32

List of Figures Figure A1 - API RP 521 figure on average rate of heating steel plates exposed to open gasoline fire on one side .................................................................................................................... 28 Figure A2 - API RP 521 figure on effect of overheating steel (ASTM A515 grade 70) ................... 29 Figure A3 - Typical carbon steel (SA-515, grade 70) rupture stress versus time to rupture (bibliographical reference [1], Page 20)................................................................................ 29 Figure A4 - Typical carbon steel (SA-515, grade 70) tensile strength and yield stress versus temperature (bibliographical reference [1], Page 16)............................................................ 30 ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 4 of 38

DRAFT 3 January 200


Figure A5 - Typical carbon steel (SA-515, grade 70) internal pressure versus pool fire exposure time to minimize potential for vessel rupture ........................................................................ 31 Figure A6 – 18-8 grade stainless steel (304, 304L) rupture stress versus time to rupture (bibliographical reference [1], Page 20)................................................................................ 31 Figure A7 – 18-8 stainless steel (304; 304L) tensile strength and yield stress versus temperature (bibliographical reference [1], Page 140).............................................................................. 32 Figure A8 – 18-8 stainless steel (304; 304L) internal pressure versus pool fire exposure time to minimize potential for vessel rupture.................................................................................... 33 Figure A9 - Effect of depressurisation on reduction of distance to overpressure effects (e.g., sideon overpressure).................................................................................................................. 33


Page 5 of 38

DRAFT 3 January 200


Introduction This Guidance on Practice (GP) provides guidance on depressurising systems that are within its stated scope and is for use in determining the need for, and design of, specific depressuring systems. This GP refers to National and International Standards that are widely accepted. Codes and Standards of the country where the equipment is manufactured and/or operated should be considered and may be accepted if they can be used to achieve an equivalent, safe, technical result. In any case, statutory and local regulations must be complied with. The value of this GP to its s is significantly enhanced by their regular participation in its improvement and updating. For this reason, s are urged to inform BP of their experiences in all aspects of its application using the “shared learning” folder on the ETP website.


Page 6 of 38

DRAFT 3 January 200

1. 1.1.

Scope General a.

This GP provides guidance on depressurisation as it relates to system design and selection of process equipment and piping. This GP is applicable to new, above ground plant and facilities and may also be used to assess possible hazards in existing systems.

b.

This GP generally covers the depressurisation of connected systems, typically consisting of one or more pressure vessels, heat exchangers, rotating equipment and other equipment items interconnected by pipework as described herein. During depressurisation of these systems:

c.

1.2.

1.3.


1.

Fluid flow may through any or all equipment and pipework in the system being considered.

2.

Fluid may reverse flow direction to its expected flow path during either normal or abnormal operation.

3.

Depressurising flow through centrifugal compressors and other rotating equipment may result in “windmilling” of the equipment, which in turn could impact the temperature reached by the fluid (and could cause damage to the rotating equipment).

This document describes and provides guidelines for the Emergency Depressuring (EDP) system and the Emergency Shutdown (ESD) system as it relates to the EDP along with briefly discussing depressurisation as it relates to maintenance. The EDP and ESD systems are integral parts of the overall operations and safety systems provided for a facility.

Objective a.

Early detection and isolation of hazardous releases and reduction of certain hazardous inventories can substantially limit the consequences from an emergency situation such as a major release of flammable materials, hydrocarbons or a fire. An emergency Shutdown (ESD) and emergency vapour space depressurisation system shall be provided in situations where rapid isolation of uncontrolled releases is desirable; to shut off secondary fuel sources that could feed a fire or vapour cloud, and to minimize releases through the use of rapid depressurisation. Coupled with a fire and gas detection system, strategically located and properly designed ESD and EDP valves can significantly reduce exposure from fire and vapour clouds.

b.

These systems do not replace any requirement for providing pressure safety valves (PSV) as required by regulation, and are supplemental to plant pressure relief protection systems.

General requirements a.

This GP specifies BP general requirements for vapour depressurising systems as applied to all sectors: Exploration and Production; Refining and Marketing; Gas, Renewables, Supply and Trading. However, additional sector factors may have to be taken into and reference should be made to any sector specific GP(s) in the ETP library.

b.

Plants, systems, and facilities covered by this GP require some form of operational depressurisation. Depressurisation can also be required for emergency situations. Clause 9 provides the information necessary to select a depressuring system and means of disposal applicable to a particular facility.

c.

Contractors are to develop designs and apply their services in accordance with the principles of this GP as amplified or modified by any accompanying supplementary specification(s). Proposed depressurisation designs, services, calculation methodologies, and the final depressurisation system design shall be subject to general discussion with and written approval by BP. ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 7 of 38

DRAFT 3 January 200

1.4.


Calculation methods This GP provides guidance on calculation methods to be used for determining:

2.

a.

Depressurisation loads.

b.

The effects of depressurisation on the temperature of plant piping.

c.

The effects of depressurisation on the minimum design temperature of vessels.

Normative references The following normative documents contain requirements that, through reference in this text, constitute requirements of this technical practice. The latest edition of the following normative documents apply.

BP GP 24-03 GP 24-10 GP 24-20 GP 24-24 GP 30-35 GP 30-45 GP 30-76 GP 30-80 GP 30-85 GP 43-54 GP 44-70 GP 44-80 GP 62-01 PSS 10 gHSEr

Guidance on Practice for Concept Selection for Inherently Safer Design. Guidance on Practice for Fire Protection – Onshore. Guidance on Practice for Fire and Explosion Hazard Management of Offshore Facilities. Offshore ive Fire Protection. Guidance on Practice for Control Valves and Pressure Regulators. Guidance on Practice for Human Machine Interface for Process Control. Guidance on Practice for SIS – Development of the Process Requirement Specification. Guidance on Practice for SIS – Implementation of the Process Requirements Specification. Guidance on Practice for Fire and Gas Detection. Guidance on Practice for Depressurisation of Pipelines. Guidance on Practice for Overpressure Protection Systems. Guidance on Practice for Relief Disposal Systems. Guidance on Practice for Valves. BP Refining Process Safety Standard No. 10.0 Hydrotreating/Hydrocracking. BP’s Getting HSE Right.

American Petroleum Institute (API) API RP 521

Guide for Pressure-Relieving and Depressuring Systems.

American Society of Mechanical Engineers (ASME) ASME VIII

Boiler and Pressure Vessel Code, Section VIII, Pressure Vessels.

American Society for Testing and Materials (ASTM) ASTM A515

Specification for Pressure Vessel Plates, Carbon Steel, for Intermediate and Higher-Temperature Service.

British Standards Institute (BSI) BSI PD 5500

Unfired Fusion Welded Pressure Vessels.


Page 8 of 38

DRAFT 3 January 200


Energy Institute (Previously known at the Institute of Petroleum) IP 9

Model Code of Safe Practice in the Petroleum Industry, Part 9 - Liquefied Petroleum Gas.

National Fire Protection Association (NFPA) NFPA 59A

Standard for the Production, Storage, and Handling of Liquid Natural Gas.

Other BLOWDOWN™

Software developed by Imperial College in London.

Note The for information on this program is Dr. Stephen Richardson, Department of Chemical Engineering, Imperial College, London United Kingdom SW7 2BY.

3. 3.1.

3.2.

and definitions a.

In this GP the term ‘approve’, as applied to BP, is used if BP does not wish a design to proceed unless certain features have been agreed in writing with a contractor or supplier. This does not imply that all details in a written document have been considered by BP and does not affect the design responsibilities of the contractor or supplier.

b.

Throughout this document, the words ‘should’, ‘shall’ and ‘must’, when used in the context of actions by BP or others, have specific meanings. For the purposes of this GP, the following and definitions apply: 1.

“Should” - is used if a provision is preferred.

2.

“Shall” - used if a provision is mandatory.

3.

“Must” - is used only if a provision is a statutory requirement.

Definitions Blowdown Can be used interchangeably with ‘depressurisation’ but more commonly taken to mean a rapid relieving of all system pressure down to atmospheric or low pressure levels. BP BP p.l.c. and their associates. Controlled Depressurisation An instrument controlled depressurisation generally giving a lower maximum flow rate over a longer period of time than uncontrolled depressurisation. Depressurise or Depressurisation To reduce the internal pressure of process equipment. Draindown Draining of liquid from vessels or storage to remove the source of flashing liquid that may feed a fire. Emergency Depressurisation The ability to rapidly depressurise, by release of gas, a plant or part of a plant in an emergency. ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 9 of 38

DRAFT 3 January 200


Operational Depressurisation The ability to depressurise equipment, by release of gas, to permit decommissioning and maintenance operations. Yellow Shutdown On an offshore production facility, the production process from wellhead valves to export lines is shutdown and isolated. Other systems remain in operation. This type of shutdown normally includes controlled depressurisation of the production facilities.

4.

Symbols and abbreviations For the purpose of this GP, the following symbols and abbreviations apply:

5.

EDP

Emergency Depressuring

EDPV

Emergency Depressuring Valve

ESD

Emergency shut-down

ESDV

Emergency shut-down valve

LPG

Liquefied petroleum gas

MCR

Main Control Room

PAS

Process Automation System

PSD

Process shut-down (i.e. not ESD)

SI

Systeme International d' Unites

SIL

Safety Integrity Level

SIS

Safety Instrumented System

General A depressurisation system is a means of reducing the pressure in a process plant or pipeline below the normal operating pressure. The main reasons for this are: • • • • •

For maintenance and inspection. To reduce the failure potential of pressure containment equipment for scenarios involving over-temperature from a fire or exothermic/runaway process reactions. To minimise release of fuel that may be feeding a fire or could be ignited. To minimise the uncontrolled release of hazardous gases. To mitigate the effect (i.e. size and/or duration) of a jet fire.

Like pressure relief, depressurisation is generally to a flare or remote vent. It is normal to be able to utilise part or all of the relief system for depressurisation. a.

Disposal is normally to a flare, but an atmospheric vent may be pursued for emergency discharge in rare instances when the project can demonstrate flaring is not an option and GP 44-80 conditions for venting with the required management approvals are obtained. ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 10 of 38

DRAFT 3 January 200


Vents must also follow applicable local regulations, discharge to a safe location (refer to API 521, Section 5), and have appropriate dispersion analysis conducted. b.

Depressurisation systems may be installed to cover operational and/or emergency situations and may be automatically or manually initiated.

c.

Depressurisation systems shall be developed in accordance with the strategy and associated performance standards for managing major hazard events in GP 24-03 and either GP 24-10 (for onshore facilities) or GP 24-20 (for offshore facilities) as appropriate.

d.

GP 44-70 and GP 44-80 contain further guidance on both system design and required process components for depressurisation systems. Dependent upon the nature of the contained fluid, depressurisation can cause low temperatures (auto-refrigeration) which can form large volumes of liquid condensate, plugging of the depressurisation system due to solids being formed (freezing) or the need for more expensive materials to avoid brittle fracture issues.

6. 6.1.

Operational vapour depressurisation General Operational depressurisation is often required for the shutdown of machinery and as preparation for plant inspection and maintenance. The maximum rate of depressurisation is influenced by: • • •

The reduction in temperature due to auto-refrigeration (see clause 12.1), The disposal system (e.g., flare) capacity, and The pressure rating of the depressuring system and connected equipment.

The addition of supplemental fuel gas to the flared gas may be required to maintain the minimum heating value required for proper flare operation when depressuring vessels containing inert gas or other gas with a low heating value. 6.2.

6.3.

Plant and equipment a.

Sections of plant or individual plant items containing hydrocarbon gas and/or liquid should normally be isolated, depressurised, drained, and purged to allow access for inspection and maintenance.

b.

Offshore, operational depressurisation should use the same valve as emergency depressurisation, if installed, to minimise weight, minimise leakage to flare, and provide a means of testing the valve. For manual depressurising the valve auto position should be overridden by local key lock to prevent the depressurisation logic becoming over complex.

c.

In onshore applications, the system may be the same as offshore; although a dedicated operational depressurisation valve is often used.

Pipelines Pipeline depressurisation is not covered within this GP; refer to GP 43-54 for pipeline depressurisation guidelines. Occasionally depressurisation of pipelines may be required to allow for inspection or repair. Generally, pipeline depressurising times are extensive, measured in days rather than minutes, making the pipeline unavailable for transportation; which may be particularly important when more than a single operator is reliant on the pipeline availability.


Page 11 of 38

DRAFT 3 January 200

7. 7.1.


Emergency vapour depressurisation General Emergency vapour depressurisation is generally used to avoid incident escalation, thereby reducing risk to personnel and limiting property or environmental damage. Examples are:

7.2.

a.

Reducing the failure potential of pressure containment equipment for scenarios involving over temperature from a fire, exothermic or runaway process reaction.

b.

Minimising the release of fuel that may be feeding a fire or could be ignited.

c.

Minimising the potential uncontrolled release of hydrocarbons or hazardous gas.

d.

Mitigating the effect (i.e., size, impact, and/or duration) of a jet fire.

Vessels, aboveground pipework, and valves Vessels and aboveground pipework are normally protected from overpressure in the fire case by relief valves. However during a fire, high pressure vessels or systems can heat-up rapidly and rupture at pressures below vessel design pressure or relief valve set pressure. The rate of temperature increase and hence the decrease of vessel strength is more rapid for gas filled systems since the rate of heat transfer and thermal capacity for a gas are less than a liquid. The same applies to the gas space of vessels that could be subject to flame impingement. a.

Provision may be made to insulate the vessel vapour space, or apply external water for cooling, or to depressurise the vessel by means of a vapour depressurising system. When dealing with vapour, the objective of a depressurising system should be to keep the internal pressure of the exposed vessels and piping below the rupture pressure as the yield stress of the wall reduces due to overheating.

b.

Depressurisation may also be used to minimise uncontrolled release from a vessel or pipework resulting from a leak such as a blown gasket or leaking valve. Detection of the gas followed by depressurisation minimises the possibility of ignition. If ignition has already occurred, depressurisation on fire detection limits fuel supply to the fire.

7.3.

c.

The platform topsides section of a subsea pipeline, including the pig launcher/receiver and downstream pipework to the pipeline ESDV at the riser/topsides interface, should be depressurised through the normal platform system. The platform design shall ensure that the ESDV and downstream pipeline are fire protected, so the pipeline inventory can be maintained.

d.

Deluge systems are designed to minimise the temperature rise of equipment in and surrounding a fire. ive fire protection may be considered for vessels particularly at risk in a fire, in addition to or instead of deluge systems. Refer to fire protection GP category 24 for more specific information and GP 30-85 on fire and gas detection.

Compressors a.

Blocking in and depressurisation of centrifugal compressors is required on seal oil failure. Depressurisation is required within the hold-up time of the seal oil overhead tank to prevent gas escape via the seals.

b.

For compressors with seals not relying on a seal oil system to contain the gas, the compressor may be considered as part of the depressurisation system.


Page 12 of 38

DRAFT 3 January 200

8. 8.1.

Depressurisation requirements General a.

ESD and EDP systems shall be reliable, failsafe, and based on proven design concepts utilizing a SIS designed in accordance with GP 30-76 and GP 30-80. The SIS shall be independent of the PAS though operators may view the ESD/EDP status through a common human machine interface. See GP 30-45. The SIS for an ESD/EDP system shall comply with the SIL determined during Front-End Engineering Design and detailed design. The ESD and EDP systems shall have a simple interface with plant operators to allow a safe shutdown and, if required, depressuring without plant operators having to consider many alternatives.

b.

The following factors shall be considered in determining the maximum rate of depressurisation:

c.

8.2.


1.

Reduction in temperature due to auto-refrigeration (see 12.1) from depressurisation.

2.

Process flare capacity requirements.

3.

Pressure rating of the relief/ vent system and connected systems.

4.

Impact on pipe s.

5.

Impact on flare pilots (e.g. potential for extinguished pilots due to high depressurisation rate).

6.

Noise induced by depressurisation.

In case of an emergency, the ESD and EDP systems should perform, as a minimum, but not be limited to, the following functions: 1.

Stop selected inlet and outlet hydrocarbon streams by closing dedicated ESDVs.

2.

Stop flow of incoming thermal energy or heat sources within ESD zone (Such as fuel sources, furnaces, and steam to reboiler, if any).

3.

Stop selected drivers on pumps and compressors. Some facilities such as lube oil and seal oil system for compressors, turbines, and lighting system are not stopped or tripped.

4.

Stop outlet liquid hydrocarbon streams by closing ESDVs on vessels requiring inventory containment, and trip associated pumps.

5.

Enable opening of dedicated EDPVs.

Emergency shutdown system The PSD/ESD system normally comprises a hierarchy of shutdown levels (i.e., individual plant equipment, a process skid or module, a localized geographic plant area [often referred to as a zone], and the entire facility). The extent of a process or plant shutdown and the respective initiating conditions are generally defined as part of the initial design. Some factors influencing the extent of an emergency shutdown include, but are not limited to, equipment operating pressure rating, operating philosophy, flare restrictions, and available fire fighting facilities. Manual ESD activation is at the discretion of an operator and is typically accomplished via hard-wired manual switches located in the MCR and at strategic positions throughout the facility. Note that ESD valves may be shared by other shutdown systems. Automatic activation of a shared ESD valve by other systems is allowed; however, manual ESD valve activation shall override any automatic process shutdown. ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 13 of 38

DRAFT 3 January 200

8.3.


Emergency shutdown valves a.

In determining ESDV locations, consideration should be given to equipment prone to failure or containing inventories of materials that present significant fire, vapour cloud, or toxic exposures as well as operational requirements associated with the process. Failures of rotating equipment, fired heaters, expansion ts, and loading operations are recognized as high frequency and in some cases having high consequence loss exposures. The GP 24 series referenced in clause 2 provides further guidance.

b.

Final ESDV locations should be reviewed by process hazard analyses and design hazard reviews that address the impact of isolation on operations. Some typical ESDV locations are potentially located at: • • • • • • • • • • •

Pressurised and refrigerated storage tanks. Feed gas lines. Storage tanks containing highly corrosive or toxic materials. Vessels containing large hydrocarbon inventories. Pumps and manifolds for ship, rail, or truck loading and unloading. Discharge lines on charge pumps. Turbine fuel systems. Process pumps. Fired heater process lines and fuel pumping systems. Refrigeration systems using a flammable medium. Hydrocarbon loading lines.

c.

ESDV location(s) and use shall be optimised based on the individual system configuration. The minimum hydrocarbon liquid level contained in vessels and tanks shall: be specific to the type of fluid, be determined by regulations and consequence analysis, and require that an ESDV be installed.

d.

ESDVs on accumulator and process vessels should be placed as close as possible to the vessel outlet flange.

e.

ESDVs for process area isolation shall be located at the edge or boundary of the process area being isolated. Upon activation, the ESDV shall stop the flow of inlet and outlet process streams, hazardous utility streams, and fuel supply to the affected area.

f.

It is important that the location and number of not only ESDVs but also depressuring valves consider: 1.

Check valves or other flow restrictions that can impede depressurisation.

2.

Packed vessels where packing can be entrained by depressuring.

3.

Other equipment such as compressors that can impede or affect depressurisation. Note in some cases the depressurisation path may be opposite to the normal flow path and/or at a significantly higher flow rate than normal.

g.

Generally, an ESDV shall be a tight shut-off, “fail close” (on loss of signal or power source), air/pneumatic operated block valve. An ESDV and its accessories shall be of fire safe design if located inside a fire zone.

h.

LNG service ESDVs should be located in accordance with NFPA 59A. For applications not subject to NFPA 59A, ESDVs should not be located at long distances from the process unit just to avoid installation in the fire zone without considering the consequences of having a potentially larger hazardous material release.


Page 14 of 38

DRAFT 3 January 200

8.4.


i.

In general, ESDVs should be located outside of buildings housing hazardous processes or hazardous utility equipment.

j.

If necessary, a hydraulic actuating system in lieu of air may be used. If hydraulic valves are selected, these shall be equipped with a secured supply of actuating fluid and back-up system and be protected as necessary against potential hazards.

k.

In general, the use of spring return air operated valves shall be selected for the ESDVs at first from reliability and maintainability points of view. However, for large size valves, etc., which require large torque, the application of double acting air cylinder type or hydraulic type, etc. shall be investigated with consideration of constructability and maintainability. If double acting air cylinder or hydraulic valves are used as ESDVs, an air bottle or hydraulic accumulators sized to provide an independent air or hydraulic supply for at least three (3) cycles, i.e. Close-open-close-open, to provide motive energy to move the valve to its fail safe position shall be provided. Additionally, valve and actuator components, including wiring, air supply, etc, should be protected against potential hazardous exposures. Inherently failsafe actuation is preferred.

l.

ESD valves shall not be provided with handwheels. ESD valve shall be equipped with open/close position limit switch and the open/close indication shall be displayed on the PAS.

m.

The scope and requirements to provide online testing of the ESDVs shall be developed during the design to meet the required integrity levels. Control valves shall not be used as ESDVs, unless specifically justified and failure of the control loop cannot cause a demand on the ESD function.

n.

The type of ESDV shall be defined dependent on the service requirements and size of the valve. Further guidance on ESD valve selection is provided in GP 62-01.

Emergency depressuring (EDP) system a.

The scope of details of EDP shall be developed during design. EDP for each skid, process area, or the total plant, if applicable, shall be enabled only after activation of the ESD system. Activation of EDP is typically accomplished via hard-wired manual switches that are located in the MCR and, for some offshore installations, automatically on confirmed detection of fire or gas release.

b.

The EDP system shall have adequate venting capacity to achieve reduction of stress in selected equipment affected by fire to a level at which stress rupture is not an immediate concern. In addition, it should be designed to enable minimization of fuel inventory that might otherwise aggravate a fire and to minimize the uncontrolled release of flammable or toxic gases.

c.

EDP system (once activated) shall be able to reduce the pressure of the system to the pressure and maximum duration provided in Annex A.

d.

The design of piping and equipment shall consider the temperature reached during autorefrigeration. Vapours released during depressuring shall be vented to the flare system or other BP approved location.

e.

The depressuring system shall consider, on a case-by-case basis, automated “liquid blowdown” capabilities. The advantages and disadvantages of liquid blowdown shall consider that the depressurisation rate depends on the boiling rate of the liquid. Significantly longer depressurisation times are likely required compared to cases where the vessels are only filled with gas. Removing liquid from a vessel exposed to a fire eliminates the cooling effect of the wetted surface area and the equipment heats up quicker, increasing the failure potential due to overheating. ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 15 of 38

DRAFT 3 January 200


In a non-fire case, liquid blowdown frequently reduces auto-refrigeration temperature effects. 8.5.

9. 9.1.

Emergency depressuring valves a.

Depressuring valves shall, in general, be tight shut-off (to avoid loss of hydrocarbons during normal operation) air operated block valves. See GP 62-01.

b.

The EDPV is typically designed to “fail close” or fail in the last position. However, the failure position of each valve shall be reviewed during the design hazard review. Should the EDPVs fail open, operational upsets due to malfunctioning valves or inadvertent opening of the valves should be considered as-well-as any single mode failure resulting in the EDPVs simultaneously opening and potentially exceeding flare capacity.

c.

Use of a single control valve to serve as both an EDP and a letdown-to-flare should follow GP 30-35. This common usage also needs review during the design hazard review to ensure this design does not result in a common mode failure that could contribute to the cause and thus defeat a protective system.

d.

The scope and requirement to provide online testing of the depressuring valves shall be developed during design to meet required integrity levels.

Application of depressurisation systems General Plants and process systems covered by this GP require some form of operational depressurisation. This section summarizes where emergency depressurisation systems are required. The use of a depressurising system maintains the integrity of equipment by reducing the possibility of vessel rupture. a.

If a depressurising system is not required by regulation, the risk and effectiveness of the system should be considered. This entails developing a fire damage assessment, assessing the viability of depressurising to mitigate undesirable leaks and spills, then fully evaluating personnel risks and plant replacement costs with and without a depressurising system installed. During this assessment the following points should be considered: 1.

The probability of a fire and/or significant leakage of flammable material.

2.

Stress rupture may not occur as the fire may die out or be extinguished. Note the inventory of a gas plant is generally significantly less than an oil installation and therefore oil fires could last appreciably longer.

3.

Water deluge, or insulation, or both, may provide sufficient protection to minimise wall temperatures over the duration of the fire.

4.

Vessel design pressure may be substantially higher than the maximum operating pressure; reducing the possibility of rupture.

5.

Upgrading material selections for equipment subjected to low temperatures as a result of depressurisation and auto-refrigeration can substantially increase investment costs.

6.

Inadvertent activation of depressuring system can cause significant production loss and possible downtime.

7.

Heavy flaring caused by activation of depressuring system (refer to BP gHSEr).

8.

The potential risk of personnel injury shall be adequately assessed.


Page 16 of 38

DRAFT 3 January 200

9.2.

9.3.


Manned production platforms and floating production facilities a.

Manned platforms shall have emergency depressurisation systems to safeguard personnel and protect the investment. On platforms where process plant is totally enclosed or enclosed by louvers, gas detection should initiate yellow shutdown followed by automatic depressurisation.

b.

To minimise spurious trips, the shutdown and depressurisation should be the result of gas detection by two detectors in a voting system. These detectors should produce a pre-alarm signal to allow possible operator intervention before the plant trips and depressurises.

c.

On open deck platforms, gas detectors are unlikely to detect anything but very large leaks. However if they are used they should initiate yellow shutdown and depressurisation. On open deck platforms a vapour cloud is much less likely to form than in an enclosed platform and if the leak is large, depressurisation is unlikely to be necessary after an ESD has taken place.

d.

Floating production facilities are invariably manned and are therefore covered by the same, above criteria.

Unmanned production platforms a.

Unmanned platforms should be fitted with emergency depressurisation systems if the production facilities contain more than 30 tonnes (30 tons) of hydrocarbons either stored at or capable of exceeding API 521 criteria (50% of vessel design pressure) during normal operations, a fire case, or a process excursion.

b.

Gas or fire detection on unmanned platforms shall result in a process shutdown. The depressurisation should be manual; either initiated by operators visiting the platform or if there are no personnel on the platform, by operators at the controlling platform or terminal. Initiation of depressurisation should consider helicopter operation in the platform vicinity.

c.

Use should be made of fire protection for systems most at risk from a delayed depressurisation.

9.4.

Onshore gas/condensate plants

9.4.1.

Gas terminals

If a terminal contains pipework and tankage but no process plant, depressurisation is unlikely to be appropriate in emergency situations. Fire protection with water deluge and/or ive protection should be used with drain down if appropriate. 9.4.2.

Stabilisation or fractionation plants and gas terminals with processing plant

Emergency depressurisation should be considered for high pressure plants (significant hydrocarbon stream(s) above 70 barg or 1 000 psig) and those with a large inventory (greater than 10 tonnes [10 tons]) of pressured hydrocarbon. Fire protection in association with paving and drainage systems design to limit the spread of fire should be considered. Refer to API RP 521. 9.4.3.

Pressured LPG storage

The protection and safety of pressurised LPG storage systems is covered by IP 9. Fire protection, both ive and with water deluge, shall be used for fire protection, as appropriate, and no vapour depressurisation is required.


Page 17 of 38

DRAFT 3 January 200


9.5.

Hydrotreating/hydrocracking reactors

9.5.1.

General

9.5.2.

9.5.3.

a.

Hydroprocessing reactors above 34 barg (500 psig) separator operating pressure shall be provided with a fail safe depressuring system that can be activated from the control room. The main purpose of this depressuring system is for high pressure inventory disposal under unit emergency conditions such as fire. Fail safe means that the valve(s) used for depressuring the unit fails open on loss of signal or de-energisation. The depressuring valve can also serve as the normal pressure control valve.

b.

The protection and safety of the pressurized reactor section of a Hydroprocessing unit shall comply with BP Refining PSS 10.

New build hydroprocessing reactors with cracking catalyst

a.

New build hydroprocessing reactors (post January 2000) containing catalyst with cracking function and above 70 barg (1 000 psig) separator operating pressure shall be provided with a fail safe depressuring system that can be activated from the control room and in the field local to the plant. Fail safe means that the valve(s) used for depressuring the unit fail open on loss of signal or de-energisation. The depressuring valve/s shall be dedicated shutdown valve/s separate from normal unit controls.

b.

The depressuring system shall have automatic initiation by high reactor temperature. High reactor temperature is defined as 30°C (54°F) above normal operating temperature or reactor mechanical design temperature, whichever occurs first. Depressuring shall not be able to be interrupted until all temperatures are 25°C (45°F) below trip setting. Initial depressuring rate shall be maximized up to the industry practice of 21 bar/min (300 psi/min) subject to reactor bed and flare system constraints, but in no cases shall be less than 14 bar/min (200 psi/min).

Existing hydroprocessing reactors with cracking catalyst

a.

Existing hydroprocessing reactors containing catalyst with cracking function and above 70 barg (1000 psig) separator operating pressure shall be provided with a fail safe depressuring system that can be activated from the control room and in the field local to the plant. Fail safe means that the valve/s used for depressuring the unit fail open on loss of signal or de-energisation. The depressuring valve/s shall be dedicated shutdown valve/s separate from normal unit controls.

b.

The depressuring system shall have manual or automatic initiation by high reactor temperature. High reactor temperature is defined as 30°C (54°F) above normal operating temperature or reactor mechanical design temperature, whichever occurs first. Depressuring shall not be able to be interrupted until temperatures are 25°C (45°F) below trip setting. Initial depressuring rate shall be at least 14 bar/min (200 psi/min).

10. Time for depressurisation 10.1.

Fire case a.

If depressurisation is not the controlling rate for design of the flare, vessels designed to BSI PD 5500 should be depressurised to 6,9 barg (100 psig) or 50% of the design pressure, whichever is the lower pressure, in 15 minutes. Vessels designed to ASME VIII, BSI PD 5500 or equivalent which are 25 mm (1 in) wall thickness or greater, should be depressurised to 50% of their design pressure in 15 minutes if the vessel is uninsulated carbon steel construction (see Annex A).

b.

Vessels designed to ASME VIII and pipework of wall thickness less than 25 mm (1 in) should be depressurised proportionately faster (see Annex A). ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 18 of 38

DRAFT 3 January 200


c.

If depressurisation governs flare design capacity, the rate of vessel(s) and pipework depressurisation should be related to the reducing strength of the steel exposed to fire. The depressurisation rate selected and method of calculation shall be subject to BP approval.

d.

If necessary to reduce the flow rate of gas to flare or vent, longer depressurisation times may be proposed for BP approval for vessels containing liquid, having external insulation, or having wall thickness greater than 25 mm (1 in). Vessels containing liquid or having external insulation take longer to heat up than non-wetted or bare vessels. Also vessels of wall thickness greater than 25 mm (1 in) take longer to heat up than the 25 mm (1 in) thick vessels used as a basis in API RP 521. For equivalent wall thicknesses, stainless steel vessels generally allow lower depressurisation rates than carbon steel (see Annex A) because stainless steel begins to lose strength at a higher temperatures than carbon steel.

10.2.

Non-fire case The 15 minute depressurisation guideline does not apply to non-fire case. Refer to formal risk analysis, Annex A, and API RP 521 guidelines.

10.3.

Stopping depressurisation Depressurisation should not be stopped on reaching the target pressure (See Annex A). Pressure normally continues to reduce until atmospheric pressure or the flare or vent backpressure is reached. Stopping the depressurising process should be evaluated on a case-by-case basis if it is deemed discontinuing depressurisation can be conducted safely. See clause 13.

11. Methods of depressurisation Initial depressurisation flow rates may be substantial and create difficulty in the design of the flare or vent systems. There are ways to reduce this high initial flow as discussed below. 11.1.

Uncontrolled depressurisation In a normal uncontrolled depressurisation the total system is shutdown and isolated and the pressure in the process equipment is discharged to flare or vent through the emergency blowdown valves, restriction orifices, and headers. This creates an initial peak flow rate that decays over the major portion of the depressurising period. The depressuring time and the end pressure (clause 10.3) are of great importance when determining the load to flare or vent. In addition to the load from the system being depressurised there is a possibility of continued in-flow into the system being depressured, e.g. from wells that have not shut-in, and consequently more gas is routed to flare increasing the total gas depressuring rate.

11.2.

Depressurisation by zone Offshore, process plant often occupies a number of fire zones, i.e. areas separated by firewalls within which a fire can be contained. Onshore, fire in one area can be prevented from reaching other areas by distance and bunding or kerbing. This division into fire zones is an important means to reduce the maximum flare load in cases of emergency depressurisation. It is more applicable to onshore plants, but can be applied offshore if good fire zone segregation is achieved. a.

Consideration should be given to depressurising only those vessels and equipment within the fire zone in which fire or gas is detected. Depressurisation of vessels and equipment in ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 19 of 38

DRAFT 3 January 200


adjacent fire zones should be timed to follow, or be at the discretion of the operator, but in any case should be delayed to minimise the size of the flare or vent system. A reduction in flare size by zoned depressurisation shall only be accepted if total depressurisation cannot occur from a common cause (e.g. simultaneous loss of instrument air to all depressurisation valves).

11.3.

b.

In exceptional circumstances, BP may approve a reduction in the peak discharge rate due to the distributive nature of the relieving flow and the requirement for pressure build-up to discharge the system inventory. The use of any such reduction shall require specific, written approval of the BP responsible Engineering Authority (EA).

c.

The designer should be aware of and prevent common cause failures, e.g. breakdown of a distributed control system in one area shall not affect any other process areas.

Controlled depressurisation In a controlled depressurisation, the peak flow rate can be maintained for a substantially longer period before the decaying phenomenon sets in. Since peak flow is maintained over a longer time, the depressurisation flow rate is reduced in magnitude compared to the peak flow rate of an uncontrolled depressurisation. Flow from each section being depressurised is normally initiated by the opening of a valve with the flow controlled by a restriction orifice. It is possible to limit the total flow rate, but still depressurise in the designated time, by either controlling the flow rate bying the orifice with a control valve or snap open valve(s) initiated by a timing sequence. This increases the flow to flare after the peak flow has ed and the system pressure has reduced. The use of controlled depressurisation is subject to BP approval, since controlled depressurisation instrumentation can be less reliable than that used for uncontrolled depressurisation. Systems containing fail open or fail shut valves rather than control valves are generally favoured, but in all cases reliability analysis shall be required for a controlled depressurisation system.

11.4.

11.5.

Depressurisation flow rates a.

Flow restriction orifices are used in conjunction with snap open valves to set the maximum rate of depressurisation to flare or vent. Alternatively, depressurisation valves may be sized to ensure the disposal system capacity is not exceeded when all valves go to the full open position. Depressurisation valves may also be designed to fail closed or in the last position to eliminate common mode failures provided that reliable valve actuation to the appropriate SIL is assured.

b.

Depressurisation flow rates and the resulting system temperatures for gas and gas/liquid systems shall be calculated in accordance with one of the methods given in Annex B. Maximum depressurisation flow rates tied to a flare system shall not exceed the flare system capacity.

c.

For controlled depressurisation, the flow may not be totally through single fixed orifice(s). Alternative calculation methods that take this into shall be considered for this case.

Draindown a.

In some instances it may be inappropriate to depressurise to flare. An example of this would be if a vessel has a considerable inventory of highly volatile hydrocarbon liquid such as LPG.

b.

In a fire, vessels containing volatile hydrocarbon should be protected by ive fire protection and/or a water deluge system. To limit vapour generation and possible spread of fire, facilities may allow for removal of liquid from the system. This may be possible using the normal product withdrawal system to remove liquid from the vessel being protected ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 20 of 38

DRAFT 3 January 200


during fire exposure, but could require special consideration of the electrical power and utility system status. c.

In many cases the advantages of retaining the liquid in the vessel to minimise the rate at which the vessel heats up outweighs the benefits from removing the flammable liquid from the fire zone. The reliability of the deluge system, the availability of separate liquid storage and the fire resistance of the system shall be considered for each case, since this will probably only be used when a fire has reached a high severity.

d.

Draindown can also be used to avoid low temperatures resulting from boil-off of highly volatile liquids and hence reduce the requirement for expensive alloy steel.

12. Effects of depressurisation 12.1.

Auto-refrigeration

12.1.1.

General

If the contents of a vessel, pipework, or a pipeline are depressurised from a high pressure, gas expansion and liquid evaporation normally cause cooling. 12.1.2.

Vessels and pipework

In general, if the vessel fluid is all vapour, the rate of heat transfer from the vessel to vapour is negligible. The heat capacity of the vessel is also high compared to that of the vapour and the reduction in wall temperature is low. An exception to this can be exit nozzles where significant forced convection takes place during depressurisation requiring the use of stainless steel nozzles, internal sleeves, or other mitigations to accommodate these low temperature blowdown effects. Available information should be carefully considered before selecting the thermodynamic process for vessel depressurisation. Downstream of the orifice where fluids are usually at their coolest and the flow very turbulent, the rate of heat transfer is high, resulting in the pipework generally approaching isenthalpic flash temperature. Note that flow up to the orifice throat is an isentropic process while flow in the piping downstream of the orifice is generally considered to be an isenthalpic process. Technical studies have yet to firmly establish the type of process within the vessel being depressured (isentropic, isothermal or somewhere in between). The isentropic assumption neglects heat gain from the surrounding environment and can predict low temperatures upon depressurisation, potentially suggesting a need for low temperature materials. In contrast, an isothermal process may often suggest that plain carbon steel construction would be appropriate, even though fluid kinetic energy gains downstream of an orifice can limit available fluid thermal energy and push fluid temperatures below those predicted by an isenthalpic flash calculation. In a vessel containing liquid and vapour, vapour expansion and evaporation from the liquid cools the liquid. The rate of heat transfer from the container to the liquid is significant. Transfer of heat in from the surroundings often is not negligible in the timescale of the depressurisation except for fire, sweating (re-condense water from humidity), and sweating up to ice layer. Annex B gives four methods of increasing complexity and accuracy when calculating minimum wall temperatures resulting from depressurisation. The time and effort expended on this calculation could depend on the likely equipment cost savings. For example, if a simple flash (Method 1) shows stainless steel is required for a large vessel by a small margin, it would be reasonable to go to a more complex ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 21 of 38

DRAFT 3 January 200


analysis to establish if the minimum wall temperature, so calculated, would allow use of a less expensive material. 12.2.

Hydrates and ice a.

Under certain conditions it is possible for hydrates or ice to form in a hydrocarbon system containing free water. In such systems, if the temperatures calculated on emergency depressurisations using a simple adiabatic flash are low enough to hydrates or ice; consideration should be given to using Method 2 in Annex B to establish if a ing temperature and pressure can exist at the same time.

b.

If a severe hydrate or ice formation problem either upstream or downstream of the depressurisation valve is possible, provision should be made for methanol injection, other BP approved chemical injection, or other methods used to inhibit hydrate formation. Consideration should be given to providing a nitrogen pressurised methanol injection system upstream of the valve. Less severe problems may occur when depressurising downstream of dehydration plants. In this case although hydrate or ice formation may be predicted, turbulence around the valve coupled with the relatively short time needed for plant depressurisation will probably prevent formation of a mass large enough to cause a blockage or other damage.

13. Repressurisation a.

If depressurisation results in temperatures lower than the design minimum temperature of a system, the system shall not be repressured until all temperatures throughout the depressurisation system have risen to a calculated safe value above the design minimum temperature. This may require a process system to be substantially above the minimum design temperature to allow for temperature reduction from gas expansion or high rates of liquid evaporation during the process restart. Repressuring a process at temperatures below a known, safe restart temperature may result in brittle fracture.

b.

Whether immediate repressurisation after EDP activation is required shall be determined by the project in conjunction with operating staff. If immediate repressurisation is not possible due to low metal temperatures, temperature monitoring and warm up facilities shall be provided in order to permit restart/repressurisation within the time period defined by the project and by operational requirements.

c.

Depressurisation systems immediately downstream of a re-pressurisation valve may be much colder than expected during the early re-pressurisation stages due to JouleThompson expansion across this valve coupled with the thermal energy transferred to the downstream fluid from its kinetic energy gain.

d.

If below minimum design temperatures can occur, re-pressuring shall be adequately controlled such that an adequate warm-up period exists. If not, a lower design minimum temperature should be employed.

e.

If depressurisation is interrupted, process consideration should be taken for units with a cracking catalyst function to minimize the possibility of reaction runaway.


Page 22 of 38

DRAFT 3 January 200


Annex A

(Normative)

Background to the selected depressurisation time A.1.

Depressurisation purpose The two primary purposes for an emergency depressurisation system are:

A.2.

a.

Remove internal pressure from a gas-filled vessel(s) to reduce the applied stress on the pressure boundary material, assumed to be a metal, exposed to a pool fire in order to extend the time before failure occurs. This is needed because the strength or load carrying ability of a metal decreases with temperature.

b.

Remove inventory from a vessel(s) to minimize the amount of material that would be released through a leak or rupture. This would reduce or mitigate the consequences from the leak or rupture.

Depressurisation systems designed for pool fire exposure a.

API RP 521 suggests depressurisation of equipment to 50% of the design pressure in 15 minutes. Although the methodology can be applied to any metal vessel, the API RP 521 criteria is specifically based on a relatively large and/or high pressure carbon steel vessel (i.e., wall thickness of 25,4 mm (1 in)) that is gas-filled (i.e., no liquid so all non-wetted surfaces) and exposed to a pool fire. The desired depressurisation rate can vary significantly based on vessel wall thickness, material of construction, initial temperature, type of fire (i.e., pool fire or jet fire), type of failure to be mitigated (vessel yield or vessel rupture), and presence of fireproofing or other fire protection means (e.g., water spray systems).

b.

Note that the following approach is applicable to standard hydrocarbon pool fires. Exposure to jet fires can result in intense, localized heating rates that significantly exceed that of a standard hydrocarbon pool fire. Higher fire heat fluxes require higher depressurisation rates than predicted below. If available, jet fire heat flux data should be utilized that is representative of jet fires one might expect in a plant as opposed to theoretical analyses or data obtained in strictly controlled tests.

c.

The normal starting point in specifying depressurisation criteria for a specific vessel would be determining the vessel wall heat up time from pool fire exposure. Figure A1 of API RP 521 provides data on average heat-up rates of steel plates exposed on one side to an open gasoline pool fire. This Figure indicates it takes about 15 minutes for a 25,4 mm (1 in) thick steel plate to heat from ambient temperature to 650°C (1200°F) in the pool fire. The plate would reach an “equilibrium” temperature of about 760°C (1400°F) in 20 minutes. The equilibrium temperature is the temperature at which heat gain from a typical pool fire is balanced by heat loss to the environment and assumes a bare, unwetted plate that is exposed to fire on one side. There can be significant variability in the heat up rate and ultimate temperature attained in a pool fire due to the influence of total heat capacity of the vapour within the vessel, fire intensity, fire duration, presence of external insulation, etc.

d.

A typical carbon steel pressure vessel is designed on a conservative basis using a recognized “code”, such as ASME Section VIII and Section IID. For example, a vessel fabricated with SA-516 Grade 70, has a maximum allowable design stress of 138 MPa (20 ksi or 20 000 psi) per ASME Section IID at up to 260°C (500°F) temperature. Note While fabrication stress concentration factors consume a portion of this design stress differential, allowable stress ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 23 of 38

DRAFT 3 January 200


values from ASME Section VIII provide a significant safety factor well below the ultimate tensile strength of the material. Typically the allowable stress is the lower of 2/3 the yield strength or 1/3.5 of tensile strength at the design temperature. The allowable stress value is based on time independent properties. This means that the material properties do not change with time up to a “code” allowable temperature that is related to a specific material. Above this temperature, the material undergoes small metallurgical changes with time. The “code” refers to these as time-dependent properties. These are typically reported as creep or stress rupture strengths. e.

For a pressure vessel exposed to a pool fire, an increase in the unwetted wall temperature above the “code” allowable temperature causes a decrease in material strength. Immediate failure may not be predicted because of conservative “code” design. However, as the metal wall temperature gets quite high, short term failures can occur as predicted based on creep or stress rupture data. Figure A2 (taken from API RP 521) shows the effect of temperature increase on rupture strength of typical carbon steel pressure vessel plate (i.e., grade SA-515). This data was obtained from bibliographical reference [1] (see Figure A3 for the original source data) and is characteristic of all grades of carbon steel. These figures plot rupture stress (i.e., load or stress on the pressure boundary material required to rupture a vessel) versus the time to rupture for several elevated temperatures. For a carbon steel vessel with a 138 MPa (20 ksi) design stress, Figure A2 (or Figure A3), predicts the vessel would rupture in about 1 hour at 593°C (1 100°F) and 0,1 hour (6 minutes) at 650°C (1 200°F). Hence, vessel rupture can occur in a relatively short time period once the metal reaches a high temperature as long as the internal pressure remains high (e.g., near design conditions).

f.

Note that the API curve shown in Figure A2 is extrapolated from the source data shown in Figure A3 for exposure times less than 0,1 hours. Figure A3 also plots the Short Time Tensile Strengths at or near the 0,1 hour data. The Short Time Tensile Strengths should be considered the limiting stresses for a given temperature (i.e., exceeding these cause failure). Hence, the extrapolations in Figure A1 at exposure times less than 0,1 hours should not be used because the vessel integrity cannot be maintained with an internal pressure that results in total stresses, including residual and applied, that exceed the vessel’s tensile strength. It is also important to note that the vessel can be damaged (e.g., yielded) at stresses significantly lower than the rupture stress. Hence, any equipment exposed to fire would need a follow-up inspection and assessment to determine fitness-forservice even if out-right failure did not occur.

g.

One way to extend the time before failure occurs is to reduce the internal pressure within a vessel (i.e., depressure it). As long as the internal pressure results in stresses below the rupture stress, the vessel should not rupture (unless there are material defects/ flaws/ previous damage or if one of the components such as a nozzle or flange is a weaker link). The data in Figure A2 or A3 can be used to determine the maximum stresses to avoid failure, but this requires a fairly comprehensive analysis. However, during the initial heatup phase, it can be easier to use the short-time tensile strength as a limiting stress that, if exceeded even briefly, would cause vessel failure. Tensile and yield strength data for typical carbon steel are given in Table A1 and illustrated in Figure A4. This data can be used to determine depressurisation versus time for vessels using carbon steel with about 483 MPa (70 ksi) tensile strength when coupled with heat-up rates.

h.

For example, the effect of depressurisation on the time to rupture is required for a large (say 25,4 mm (1 in) wall thickness) vessel fabricated from SA-516 Grade 70 carbon steel with a design allowable stress of 138MPa (20 ksi). The scenario involves pool fire exposure whereby the internal pressure increases to that equivalent to the design allowable stress in the pressure boundary wall and then stays constant as the vessel continues to be heated. Maintaining a constant pressure near the design allowable stress would be comparable to having the vessel protected by a pressure relief valve. If the vessel was ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 24 of 38

DRAFT 3 January 200


initially at ambient temperature, Figure A1 indicates it will take slightly over 10 minutes for the average vessel wall temperature to reach 538°C (1 000°F). Based on the data in Table A1 and Figure A4, the vessel starts to yield at this temperature because the applied stress (138 MPa (20 ksi)) would become equal to the yield stress. If the applied stress stays constant at 138 MPa (20 ksi) and further wall heating occurs, then the vessel will rupture when the wall temperature reaches about 650°C (1 200°F), about 14 minutes into the fire (see Figure A1). Rupture can occur at lower temperatures if there are material defects or a weaker component of the vessel construction than the base metal. i.

API RP 521 suggests the internal pressure be reduced by 50% within 15 minutes of pool fire exposure. Figure A1 indicates the average wall temperature 15 minutes into the pool fire would be about 670°C (1 240°F). For our vessel with the 138 MPa (20 ksi) allowable design stress (thus applied stress of only 69 MPa (10 ksi) once depressured to 50% internal pressure), Figure A4 indicates that at 670°C (1 240°F) the 69 MPa (10 ksi) stress on the carbon steel pressure boundary would exceed the yield point (about 62 MPa (9 ksi)), but not the rupture stress (about 90 MPa (13 ksi)). Thus, the vessel will be damaged, typically in the form of bulging. Note that this rupture stress is based on the vessel seeing a wall temperature of 670°C (1 240°F) for the full 15 minutes of pool fire exposure (see Figure A2 or A3). In other words, it ignores heat-up time and assumes the wall reaches 670°C (1 240°F) at the onset of fire exposure. At the other extreme would be the case where the vessel wall temperature does not appreciably increase until the end of the 15 minute exposure in which case the metal wall sees 670°C (1 240°F) but only for a very short time. In this case, the tensile strength would be the limiting factor (about 117 MPa (17 ksi) at 670°C (1 240°F) from Figure A4). The actual internal pressure where rupture occurs will likely be in between the rupture stress and tensile strength given the pool fire heat-up rate is not instantaneous.

j.

It is important to note that the vessel wall temperature can continue to increase per Figure A1. Based on a maximum wall temperature of 760°C (1 400°F), further depressurisation would be required to avoid rupture. The bottom curve in Figure A2 (or Figure A3) indicates the internal stress would need to be reduced to about 34,5 MPa (5 ksi) or 34,5/138 (5/20) = 25% of the MAWP to extend the time to rupture to about one hour and reduced to about 27,6 MPa (4 ksi) or 27,6/138 (4/20) = 20% of the MAWP to extend the time to rupture to about 4 hours.

k.

The depressurisation path as a function of time for carbon steel of several thicknesses is illustrated in Figure A5. This figure illustrates the depressurisation path to minimize failure potential due to fire exposure as a function of percent of MAWP. Because of the faster heat-up rate (see Figure A1), thinner walled vessels require faster depressurisation. These varying rates of depressurisation become important when deg a common depressurisation system for interconnected vessels of different sizes and/or design pressures. In such cases, a fire could simultaneously expose multiple vessels of varying wall thicknesses. Judgement needs to be taken as to which vessels are included in the depressurisation system design. Rapid depressurisation required for small vessels must be balanced not only with the available disposal system capacity but also with the failure consequences of small versus large vessels.

l.

In addition to reducing the likelihood for rupture, depressurisation would reduce the consequences in the event of rupture due to fire exposure. Upon rupture, the mechanical energy (gas stored under pressure) is converted to a blast wave that can damage adjacent equipment due to overpressure. In addition, fragments can be produced upon failure. Possible impacts of the failure should be analysed on a case-by-case basis. The general effect of depressurisation on distance to overpressure effects (e.g., side-on overpressures) due to vessel failure is illustrated in Figure A9.

m.

The material data presented above applies to typical carbon steel. Alloy materials, including low alloy steels, stainless steels, and nickel base alloys, show similar strength loss with temperature as carbon steel, but less dramatic: a sample of rupture strength versus ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 25 of 38

DRAFT 3 January 200


temperature for type 304 stainless steel (i.e., 18-8 grade) is shown in Table A2 and Figures A6 through A8. n.

A.3.

A.4.

It is important to recognize that the tensile and rupture strength data given in this Annex are used as an example of how depressurisation calculations for pool fire exposure can be made. The sources for this data are typical industry data (1, 2, 3, and 4). However, the designer should be aware that the values provided by all such references are not necessarily consistent with each other. Variability in the pool fire (heat-up rate, maximum equilibrium temperature, view factors, size, duration, extent) can further complicate depressurisation requirements. Given the variability in assumptions, the designer should be careful not to overanalyse the depressurisation system requirements but to select an appropriate criteria for their specific facility requirements.

Depressurisation systems designed to minimize leak size a.

The second use of a depressurisation system is to mitigate the effects of a leak. A reduction of vessel pressure through depressurisation to a “safe” location, such as a flare, reduces not only the leak rate but also the leak duration through reduction in vessel inventory. Depressurisation is one of the few mitigation measures for a jet fire whereby if the source of the jet fire were depressurised, the size and effects are attenuated. Once depressurised to near atmospheric pressure, the jet fire would in essence be “turned off” as the source of fuel would be depleted.

b.

The design of a depressurisation system to mitigate consequence of leaks is dependent upon the following: 1.

Nature and extent of consequence (toxicity hazard, flammability hazard (vapour cloud explosion and/or flash fire extent and potential), jet fires, environmental release);

2.

Location of potentially vulnerable public, plant personnel, environmental aspects, equipment;

3.

Location and configuration of adjacent equipment (particularly important regarding mitigation of jet fire impacts);

4.

Capacity of the disposal system where the depressurisation system discharges;

5.

Inventory of fluid requiring depressurisation including liquid that can flash into vapour upon depressurisation;

6.

Evacuation considerations.

c.

Generally, a starting point in the depressurisation system design would be using a total instantaneous depressurisation rate equal to no more than the capacity of the disposal system (e.g., flare). It is critical to stay within the disposal system capacity even during utility failures such as loss of instrument air that may cause all the depressurisation valves to open simultaneously (if so designed to fail in the open position upon air failure and without a backup air supply).

d.

Because of the variability in processes and reasons for depressurisation in non-fire situations, it is recommended the designs be handled on a case-by-case basis. Consequence modelling should be performed to determine specific depressurisation goals. In many circumstances, a depressuring level to no more than 0,7 MPa (100 psig) is commonly used as the design basis.

Calculation of depressurisation mass flow rates a.

With depressurisation calculations now embedded in process simulation software, manual calculation of blowdown times by step calculations at various reducing pressure increments is no longer adequate. Manual calculations are only approximate, always ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 26 of 38

DRAFT 3 January 200


require subsequent verification from a more rigorous analysis, and can provide very misleading results depending on the assumptions made and coefficients being used. b.

A.5.

Process simulation s should recognise there often can be limitations to these built-in depressurisation solutions. Some packaged software calculations fail to take proper of phase changes and heat transfer (or heat loss). As one example, high purity liquids stored near their bubble point are not handled well.

API 521 guidelines Refer also to API 521 Section 3.19 for further guidance on depressurisation flow rates.


Page 27 of 38

DRAFT 3 January 200


Figure A1 - API RP 521 figure on average rate of heating steel plates exposed to open gasoline fire on one side


Page 28 of 38

DRAFT 3 January 200


Figure A2 - API RP 521 figure on effect of overheating steel (ASTM A515 grade 70)

Figure A3 - Typical carbon steel (SA-515, grade 70) rupture stress versus time to rupture (bibliographical reference [1], Page 20)


Page 29 of 38

DRAFT 3 January 200


Table A1 - High temperature tensile properties for typical carbon steel (1) Temperature, °F

Temperature, °C

Tensile Strength, psi

Tensile Strength, MPa

Yield Stress 0,2% Set, psi

Yield Stress 0,2% Set, MPa

750

399

58 000

400

24 600

170

900

482

45 500

314

23 500

162

1 000

538

36 500

252

20 100

139

1 100

593

27 200

188

14 250

98

1 200

649

20 000

138

10 200

70

1 300

704

13 500

93

7 375

51

1 400

760

9 025

62

3 750

26

(1) Applies to SA-515 and SA-516 carbon steels, Grade 70. Reference: bibliographical reference [1], Page 16.

Figure A4 - Typical carbon steel (SA-515, grade 70) tensile strength and yield stress versus temperature (bibliographical reference [1], Page 16)

40000

Strength or Stress PSI

35000 30000 25000 20000 15000 10000 5000 0 1000

1100

1200

1300

1400

Degrees F Tensile Strength

Yield Stress


Page 30 of 38

DRAFT 3 January 200


Figure A5 - Typical carbon steel (SA-515, grade 70) internal pressure versus pool fire exposure time to minimize potential for vessel rupture

100% Maximum Pressure to Avoid Failure (% of MAWP)

90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 0

5

10

15

20

25

30

35

40

45

50

55

60

Time from Fire Start (minutes) 1/8 Inch Thick

1/4 Inch Thick

1/2 Inch Thick

1 Inch Thick

Figure A6 – 18-8 grade stainless steel (304, 304L) rupture stress versus time to rupture (bibliographical reference [1], Page 20)


Page 31 of 38

DRAFT 3 January 200


Table A2 - High temperature tensile properties for 18-8 stainless steel (1) Temperature, °F

Temperature, °C

Tensile Strength, psi

Tensile Strength, MPa

Yield Stress 0,2% Set, psi

Yield Stress 0,2% Set, MPa

1 000

538

53 000

365

14 000

97

1 100

593

48 500

334

12 000

83

1 200

649

43 000

296

11 000

76

1 300

704

35 000

241

11 000

76

1 400

760

27 000

186

10 500

72

1 500

816

20 500

141

10 000

69

1 600

871

17 650

122

---

---

1 800

982

9 600

66

---

---

2 000

1 093

4 900

34

---

---

(1) Applies to 304 and 304L stainless steels. Reference: bibliographical reference [1], Page 140.

Figure A7 – 18-8 stainless steel (304; 304L) tensile strength and yield stress versus temperature (bibliographical reference [1], Page 140)

60000

Strength or Stress PSI

50000 40000 30000 20000 10000 0 1000

1100

1200

1300

1400

Degrees F Tensile Strength

Yield Stress


Page 32 of 38

DRAFT 3 January 200


Figure A8 – 18-8 stainless steel (304; 304L) internal pressure versus pool fire exposure time to minimize potential for vessel rupture

100% 90% Maximum Pressure (% of MAWP)

80% 70% 60% 50% 40% 30% 20% 10% 0% 0

10

20

30

40

50

60

Time from Fire Start (minutes) 1/8 Inch Thick

1/4 Inch Thick

1/2 Inch Thick

1 Inch Thick

Failure Pressure (% of MAWP)

Figure A9 - Effect of depressurisation on reduction of distance to overpressure effects (e.g., side-on overpressure)

100 90 80 70 60 50 40 30 20 10 0 0

20

40

60

80

100

% Reduction in Distance to an Overpressure Effect Relative to Failure @ the MAWP ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 33 of 38

DRAFT 3 January 200


Annex B

(Normative)

Methods for estimating the minimum wall temperature of depressurised vessels and pipework B.1

Proposed methods a.

There are four methods of increasing complexity and accuracy for estimating minimum vessel and pipework temperatures during depressurisation: 1. 2. 3. 4.

B.2

A theoretical (reversible) adiabatic, instantaneous flash. Staged flash, infinite heat transfer between vessel and liquid. Staged flash, finite heat transfer between vessel and liquid. Staged flash, with vessel heat conduction calculation at each stage.

b.

The first method requires only process data; the second additionally requires vessel design data and the third and fourth additionally require details of the depressurising procedure.

c.

The first method is the most simple, least accurate, and most conservative, while the fourth is the most complex and time consuming but most accurate/ least conservative.

General assumptions a.

Process simulation programs generally assume the fluid is perfectly mixed in the vessel. Specialist software such as “BLOWDOWN™”0 shall be used to for different fluid strata within a vessel.

b.

Calculations are usually performed by depressurising from maximum design pressure and minimum ambient temperature to atmospheric pressure. More reasonable starting conditions might be maximum design pressure and minimum fluid flowing temperature, if the contents of the vessel would not be shut-in long enough for the temperature to drop to minimum ambient temperature.

c.

The vessel is warmed by the air surrounding it, but experience has shown that it may not have a significant effect during the rapid depressurising timescale in certain cases (particularly non-fire scenarios). However, it is possible for water in the air (i.e., humidity) to condense on the vessel wall maintaining the wall temperature no lower than the water dew point. It may be possible depending upon the rate of cooling for the vessel wall temperature to fall below 0°C (32°F) at which time the vessel could get an insulating covering of ‘snow’ or ice. Further cooling of the wall temperature could then occur to relatively low temperature, depending upon the fluid inside the vessel, depressurisation rate, and so on.

d.

The heat transfer between vapour and vessel walls is usually much poorer than between a liquid and the vessel walls. Additionally the thermal capacity of all but the highest pressure gas is much lower than that of the vessel walls and temperature depressurising effects on vessel walls containing only vapour are generally small and only of concern if the pressure is very high (greater than 100 bar (1,450 psi)), or the vessel is normally operated at subambient temperatures. Heat transfer downstream of the depressurisation valve can be sufficiently high that an internal sleeve, providing a convective separation gap between the depressurised fluid and outer pipe wall, may often be required.


Page 34 of 38

DRAFT 3 January 200

B.3

B.4

Vessel contents a.

Vessel fluid cooling comes from expansion of the contained gas and the liquid standard enthalpy change of vaporisation (or heat of vaporisation). In processes where the liquid phase is relatively non-volatile, the minimum temperature is experienced when vapour volume is greater. Therefore, when using Methods 2 to 4 below, the lowest reasonable liquid level (with maximum vapour volume) should be considered to accurately calculate the expected minimum temperature.

b.

If the vessel liquid boils (or vapourises) significantly as pressure is reduced, such as in the pressurised storage of pure gases as liquids, then the minimum possible temperature could be experienced at maximum liquid levels, or it could be insensitive to liquid level. If significant liquid vaporisation occurs, calculations should be done at both the reasonably highest and lowest liquid levels that may occur in the vessel.

Method 1 a.

b.

B.5

B.6


Using Method 1, several oversimplifying assumptions are made: 1.

Vessel contents are well mixed.

2.

The vessel depressurises instantaneously.

3.

A reversible, adiabatic depressurisation to the final system pressure is achieved.

To perform this type of calculation, vessel contents are depressurised isenthalpically using a process simulator such as HYSYS. The ensuing fluid temperature is taken to be the minimum wall temperature.

Method 2 a.

In this method, the thermal mass of the vessel is taken into . The heat transfer from the vessel wall in with the liquid is assumed to be infinite, so that the liquid contents of the vessel and the wetted wall are at the same temperature.

b.

The contents of the vessel are to be depressurised in steps corresponding to given small temperature drops. 1.

The temperature drop for the next stage is decided upon.

2.

The height of liquid before each depressurising stage is calculated from the liquid fraction.

3.

The thermal capacity of the wetted vessel walls is calculated, and the heat lost from the vessel walls over the fall in temperature is added to the internal energy of the vessel contents.

4.

The vessel is depressurised until the fluid temperature drop reaches the required figure.

5.

The amount of vapour considered in the model is reduced so that the volume of fluid corresponds to the vessel volume.

6.

The vapour and liquid streams are re-mixed.

7.

Return to stage 1 until the vessel pressure is down to the required figure.

Method 3 a.

This method is used for vapour filled vessels and also for vessels partially filled with liquid where the liquid is removed before significant depressurising occurs. The method is similar ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 35 of 38

DRAFT 3 January 200


in theory to Method 2, but heat transfer between the fluid and the vessel is finite and either calculated or estimated before depressurising starts. There are different heat transfer coefficients for thermal flux into the liquid and vapour, and it may prove necessary to recalculate the heat transfer coefficient as the depressurising continues. Warming of the fluid by the vessel walls is now dependent upon the time taken for depressurising. Therefore the time/ pressure relationship shall be calculated, usually by assuming choked flow through the venting orifice. The depressurising shall be done in small time steps.

b.

B.7

1.

The temperature drop for the next stage is decided upon.

2.

Liquid height before each depressurising stage is calculated from the liquid fraction.

3.

The vessel is depressurised until the fluid temperature drops to the required value.

4.

The time required for that pressure drop is calculated.

5.

Heat flows from the vessel to the liquid and vapour are calculated.

6.

The amount of vapour considered in the model is reduced so that the volume of fluid corresponds to the vessel volume.

7.

The vapour and liquid streams are re-mixed.

8.

Return to stage 1 until the vessel pressure is down to the required figure.

In practical , if liquid is still present in the vessel, the heat capacity of liquids are much higher than vapour and heat transfer into the gas is generally much less than the liquid temperature changes.

Method 4 a.

This method is based on the previous method, but entails a detailed analysis of the temperature profiles within the vessel walls at each de-pressuring stage. This allows the calculation of vessel wall stresses from the thermal gradients. This greater detail can allow a lower minimum temperature to be withstood by a metal than would otherwise be permissible.

b.

The stages are per Method 3 above, except for step 5 which is replaced as follows: 1.

The temperature profile throughout the vessel walls is calculated using a dynamic finite element or finite difference model of the system.

2.

In a finite element model, the vessel is divided up into sections small enough that the properties of the section can be assumed to be constant throughout the section. The most important properties in this case are obviously the heat flows and temperature gradients. Heat conduction between adjacent sections is calculated by a computer program, and from this, the new temperature profile is determined.

3.

The finite difference model differs from the finite element model in that it represents the vessel by a network of points, asg heat capacities to the points, and thermal conductivities to the connections between points. Each point corresponds to the centre of the sections in the finite element model. The differences between the two representations are more of mathematical and computing concern than of concern to the engineer. The choice of method probably depends upon the software available to the engineer.

4.

The number of sections or points depends upon the resolution of the temperature profile required by the mechanical engineer, and upon the needs of the mathematical method. If the model is too simple, the assumption that the properties can be assumed constant across the section is no longer valid. The usual way to determine whether the model is fine enough is to rerun the model with a finer grid, or smaller sections, and ed Date: 6/17/2008 11:12:31 PM The latest update of this document is located in the BP ETP and Projects Library

Page 36 of 38

DRAFT 3 January 200


see whether there is any significant change in the results. If there is, then the model shall be rerun with a yet smaller grid, until two grids give consistent results.


Page 37 of 38

DRAFT 3 January 200


Bibliography th

[1]

“Digest of Steels for High Temperature Service”, 5 Edition, 1946, published by the Timken Roller Bearing Company, Canton, Ohio.

[2]

“An Evaluation of the Elevated Tensile and Creep-Rupture Properties of Wrought Carbon Steel”, ASTM Data Series DS 11S1, Supplement to Publication DS 11, formerly ASTM STP 180, ISBN 8031-2004-4, January, 1970.

[3]

“Report on the Elevated-Temperature Properties of Chromium-Molybdenum Steels”, W.F. Simmons and H. C. Cross, ASTM Special Technical Publication 151, October 1953.

[4]

“Report on the Elevated-Temperature Properties of Stainless Steels”, W.F. Simmons and H. C. Cross, ASTM Special Technical Publication 124, January 1952.


Page 38 of 38

Gp 44-25 Guidance On Practice For Depressurisation 6w2gv

Overview 3e4r5l

More details w3441

Related Documents 3m3m1z

Gp 44-25 Guidance On Practice For Depressurisation 6w2gv

Guidance On Risk Assessment For Offshore Installations 5l6k2q

Coco Guidance On Control 4293

Depressurisation - A Practical Guide 5x70p

Gp 4ln6j

Career Guidance For Grade11 4v3z4n

More Documents from "YT" 235s6j

Drill Well Path Design d5m24

Kerangka Acuan Ispa Diare Ak 1-8 2017 6e506d

Olga7 Training Exercises 621yj

Lagu Bertangga Nada Mayor 6t2160

Gp 44-25 Guidance On Practice For Depressurisation 6w2gv

Inventory Management System Documentation 25w2g