Ignition modelling Are our approaches aligned?

Similar documents
CFD Based Approach for VCE Risk Assessment

Determination of the Design Load for Structural Safety Assessment against Gas Explosion in Offshore Topside

Practical Modelling & Hazard Assessment of LPG & LNG Spills

CHEMICAL ENGINEERING TRANSACTIONS

I.CHEM.E. SYMPOSIUM SERIES NO. 97 BUOYANCY-DRIVEN NATURAL VENTILATION OP ENCLOSED SPACES

Fire and Safety for Offshore drilling and production Ajey Walavalkar ANSYS Inc.

DIGITAL SOLUTIONS TRAINING CATALOGUE. QRA and CFD simulation. Phast, Safeti and KFX SAFER, SMARTER, GREENER


REQUIREMENTS FOR VALIDATION OF MATHEMATICAL MODELS IN SAFETY CASES

Identification and Screening of Scenarios for LOPA. Ken First Dow Chemical Company Midland, MI

Part 2.5 Dispersion Modeling Using ALOHA

OIL & GAS OPTIMIZATION OF AN OFFSHORE PLATFORM ORIENTATION

Benefits of Detailed CFD Ventilation Analysis During Early Design Phases. Josué Quilliou Gexcon Consulting

Gas Accumulation Potential & Leak Detection when Converting to Gas

Well Control Modeling Software Comparisons with Single Bubble Techniques in a Vertical Well

Comparison of Large-Scale Vented Deflagration Tests to CFD Simulations for Partially Congested Enclosures

WIND LOADS / MOORING & FISH TAILING. Arjen Koop, Senior Project Manager Offshore Rogier Eggers, Project Manager Ships

Analysis of Variance. Copyright 2014 Pearson Education, Inc.

Abstract. Geometry Obstructions Ventilation Wind conditions

Software Reliability 1

Introduction to consequence modelling

EFFECTIVE DESIGN OF CONVERTER HOODS. 111 Ferguson Ct. Suite 103 Irving, Texas U.S.A. 400 Carlingview Dr. Toronto, ON M9W 5X9 Canada.

FIRE PROTECTION. In fact, hydraulic modeling allows for infinite what if scenarios including:

ACCIDENTAL HYDROGEN RELEASE IN GC-LABORATORY; A CASE STUDY

At each type of conflict location, the risk is affected by certain parameters:

Quantitative risk assessment and risk-based decision making

MODELING OF HYDROGEN EXPLOSION ON A PRESSURE SWING ADSORPTION FACILITY

QUANTIFYING THE TOLERABILITY OF POTENTIAL IGNITION SOURCES FROM UNCERTIFIED MECHANICAL EQUIPMENT INSTALLED IN HAZARDOUS AREAS

Overview of Recovery and Rescue Validation, Verification and Extrapolation

Finite Element Analysis of an Aluminium Bike Frame

Experimental Study of Hydrogen Release Accidents in a Vehicle Garage

Aerodynamic study of a cyclist s moving legs using an innovative approach

LEAK FREQUENCY MODELLING FOR OFFSHORE QRA BASED ON THE HYDROCARBON RELEASE DATABASE

MAHB. INSPECTION Process Hazard Analysis

Misuse of Combustible Gas Meters

Improvement in Release Frequencies for Quantitative Risk Assessment

Internal Explosion Methodologies

1200 New. Jersey Ave., SE. producers. and potassium. A. Summary. the

Impact of the tides, wind and shelf circulation on the Gironde river plume dynamics

Application of fuzzy logic to explosion risk assessment

Full Scale Experimental Testing of Aerosol Filling Facilities Experiments and Validation

Hydrogen Bubble Dispersion and Surface Bursting Behaviour

Designing a Traffic Circle By David Bosworth For MATH 714

Analysis and Comparison of Calculation Methods for Physical Explosions of Compressed Gases

A NEW PROCESS FOR IMPROVED LIQUEFACTION EFFICIENCY

Engineering Models for Vented Lean Hydrogen Deflagrations

Deep water plume models - What s special about deep water

CORRELATION EFFECTS IN THE FIELD CLASSIFICATION OF GROUND BASED REMOTE WIND SENSORS

Industrial Explosion Protection How Safe is your Process?

CONCLUSIVE SUMMARY OF CARBON DIOXIDE

Calculation of Trail Usage from Counter Data

Determination of characteristic accidental actions. Outline

DEVELOPMENT AND IMPLEMENTATION OF RISK ASSESSMENT METHODS FOR ONSHORE NATURAL GAS TERMINALS, STORAGE SITES AND PIPELINES

USE OF THE EXCEEDANCE CURVE APPROACH IN OCCUPIED BUILDING RISK ASSESSMENT

CFD BENCHMARKING: HAMER HALL AUDITORIUM CASE STUDY

HAP e-help. Obtaining Consistent Results Using HAP and the ASHRAE 62MZ Ventilation Rate Procedure Spreadsheet. Introduction

Juan de Fuca Recreational Chinook Fishery Proposal. Forum on Conservation & Harvest Planning for Fraser Salmon April 13, 2015

Experimental Characterization and Modeling of Helium Dispersion in a ¼-Scale Two-Car Residential Garage

Dynamic Modelling of Control Valves

Modelling Hazardous Consequences of a Shale Gas Well Blowout

Gravity Probe-B System Reliability Plan

Global Flow Solutions Mark Zagar, Cheng Hu-Hu, Yavor Hristov, Søren Holm Mogensen, Line Gulstad Vestas Wind & Site Competence Centre, Technology R&D

DETERMINATION OF HAZARDOUS ZONES FOR A GENERIC HYDROGEN STATION A CASE STUDY

Wind Farm Blockage: Searching for Suitable Validation Data

THREE DIMENSIONAL FLAME PROPAGATION ABOVE LIQUID FUEL POOLS

ERCBH2S Frequently Asked Questions.

THE SIGNIFICANCE OF HAZARD ENDPOINTS IN QUANTITATIVE RISK ANALYSIS

Goldfields Gas Pipeline Access Arrangement Comments on Pipeline Capacity Modeling and the Impact of Changing Gas Quality

Quantitative Risk Analysis (QRA)

DETAILS OF THE ACCIDENT PROGRESSION IN 1F1

On the Challenges of Analysis and Design of Turret-Moored FPSOs in Squalls

Sea-going vessel versus wind turbine

Safety assessments for Aerodromes (Chapter 3 of the PANS-Aerodromes, 1 st ed)

SUMMARY OF THE EXPERIMENTAL STUDIES OF COLD HELIUM PROPAGATION ALONG A SCALE MODEL OF THE LHC TUNNEL

Improving Accuracy of Frequency Estimation of Major Vapor Cloud Explosions for Evaluating Control Room Location through Quantitative Risk Assessment

ENHANCED PARKWAY STUDY: PHASE 2 CONTINUOUS FLOW INTERSECTIONS. Final Report

Verification and Validation Pathfinder

intended velocity ( u k arm movements

Plan B Dam Breach Assessment

Session 13: Fundamentals of Area Classification Studies

REPORT INVESTIGATION OF 3M CLEAN AND STRIP DISKS IN EXPLOSIVE ATMOSPHERES. Client: 3M Norge A/S. Authors: Bjørnar A. Johnsen Geir H.

Best Practice RBI Technology Process by SVT-PP SIMTECH

THESIS COMPLETION CERTIFICATE

Analysis of Pressure Rise During Internal Arc Faults in Switchgear

Recommendations for the Risk Assessment of Buffer Stops and End Impact Walls

Simulation and modelling of water spray in the 3D explosion simulation program FLACS. Elin Kristin Dale

Autodesk Moldflow Communicator Process settings

WindProspector TM Lockheed Martin Corporation

CFD Modelling of LPG dispersion

Tilt Detection Using Accelerometer and Barometric Measurements

DEFINING HAZARDOUS ZONES ELECTRICAL CLASSIFICATION DISTANCES

EXPLOSIVE ATMOSPHERES - CLASSIFICATION OF HAZARDOUS AREAS (ZONING) AND SELECTION OF EQUIPMENT

DAMAGE STABILITY TESTS OF MODELS REPRESENTING RO-RC) FERRIES PERFORMED AT DMI

QRA method for land-use planning around natural gas production, processing and transportation sites in the Netherlands

LES* IS MORE! * L ARGE E DDY S IMULATIONS BY VORTEX. WindEnergy Hamburg 2016

DETERMINATION OF HAZARDOUS ZONES FOR A GENERIC HYDROGEN STATION A CASE STUDY

Failure analysis of storage tank component in LNG regasification unit using fault tree analysis method (FTA)

Sesam HydroD Tutorial

ALOHA. Example Scenarios. September 2016

Golf Ball Impact: Material Characterization and Transient Simulation

Transcription:

Ignition modelling Are our approaches aligned? Lars Rogstadkjernet Gexcon

Outline Experience from 3 rd party reviews Areas of diverging approach Dispersion and ignition modelling

3 rd party reviews 3 rd party reviews of 5-6 prob analyses ala Z-013 GexCon perspective: as provider of the tool we like user to be successful Industry perspective: inconsistency does not benefit to anyone Accuracy vs consistency

Norsok Z-13 Annex F Detailed guidance 1.Model representation 2.Leak rates and transients 3.Leak directions 4.Winds conditions 5.Time dependent ignition model 6... 7... Widely adopted

ERA work tasks Assumption and input Geometry modeling Ventilation analysis Dispersion study Explosion study Risk calculation Reporting

ERA work tasks Assumption and input Geometry modeling Ventilation analysis Dispersion study Explosion study Risk calculation Reporting

Geometry modelling Early design stage: incomplete models Anticipated congestion modelling basis seem to vary (significantly) Congestion estimated based on experience from previous projects Not consistent measure for congestion quantification Congestion characterized as packing density: Meter of pipe/volume = OK Meter of box/volume???? Databases predictions relying on box length counts have fundamental flaws! Considerable spread in practice x3 x5 x9

Practicalities ACM A model that look reasonable can be assessed by the engineering party. Is this possible with extensive use of randomized congestion? How can correctness be challenged? Is this a reasonable congestion level around a compressor?

Practicalities ACM Does it matter where the missing congestion are located? Should much of this be neatly arranged here? How much does it matter?

ERA work tasks Assumption and input Geometry modeling Ventilation analysis Dispersion study Explosion study Risk calculation Reporting

Dispersion modelling in ERA The normal approach: 1. Run X number of dispersion simulations 2. Derive cloud size and ignition probability 3. Not all scenarios can be simulated hence many are «extrapolated» X vary enormously! Consequently the degree of «extrapolation» varies greatly Little consistency with efforts on explosion simulations

Basic logic of TDIIM /OLF ignition models Ignition is more likely when the cloud covers a large area When explosive atmosphere persist for a long time, ignition would increase Likelihood of ignition should be high at first time exposure and then drop Gas alarm/activation of ignition control measures should have an effects All relate directly to gas dispersion

And it s complex. 6 kg/s 96 kg/s 24 kg/s 3 kg/s 1,5 kg/s Ignition probability drops as time passes

Case example 1 2 releases with same: Release rate Leak profile Release location Wind speed Wind direction Wind Only difference is leak direction

Case example 1

Case example 1 3,000 2,500 2,000 1,500 1,000 500 Cloudsize vs time 0 0 50 100 150 200 250

Case example 1 TDIIM ignition probability as function of time Sum ignition probability A: 3.65% Sum ignition probability B: 0.49%

Case example 1 2700 m 3 300 m 3

Case example 1 Scenario Ignition probability Pressure A 3.65% 0.40 B 0.49% 0.02 Can simplify or average? Can you get by with only doing one? Sensible approach seem to include both scenarios

Case example 1 What about the other scenarios? Scenario Ignition probability Pressure A 3.65% 0.40 B 0.49% 0.02 C D E An average rate specific ignition probability?

Case example 2 ERA for FLNG Generic TDIIM approach Ignition predicted from 1. Leak = source term 2. Quantity released 3. Ventilation = sink term 4. Module volume Ign prob calculated for each rate and wind condition

Case example 2 FLNG - buoyant and non buoyant gas Two identical release scenarios same vessel, same wind, same leak rate, same location Only difference is composition: A = Propane (refrigerant) B = Natural gas (LNG)

Case example 2: 8m/s wind

Cloud size [m3] Case example 2 Refrigerant and natural gas produce almost identical cloud sizes Ignition probability will be almost the same Wind trumps buoyancy effects 1,600 1,400 1,200 1,000 800 600 400 200 Flammable cloud vs time C3-8m/s C1-8m/s 0 0 50 100 150 200 250 300 350 400 450 500 Time [s] OK

Wind = 1m/s

Cloud size [m3] Case example 2 Flammable cloud vs time 6,000 5,000 4,000 3,000 2,000 1,000 0 C3-1m/s C1-1m/s 0 50 100 150 200 250 300 350 400 450 500 Time [s]

Ignition probability Ignition probability Case example 2 2.0E-05 Ignition probability: C3,1m/s wind 2.0E-05 Ignition probability: C1,1m/s wind 1.5E-05 Continuos Intermitent Sum 1.5E-05 Continuos Intermitent 1.0E-05 1.0E-05 Sum 5.0E-06 5.0E-06 0.0E+00 0 100 200 300 400 Time [s] 0.0E+00 0 100 200 300 400 Time [s] Sum ignition probability A: Sum ignition probability B: 1.67% 0.77%

Case example 2 Propane cloud: 4800m3: 2.4 barg Natural gas cloud: 2100m3 0.20barg

Case example 2 Scenario Ignition probability Pressure A 1.65% 2.40 B 0.77% 0.20 Not OK Because buoyancy trumps wind at low wind speed

Design changes Process deck (yellow) changed from grating to plating Reduce volume where gas can accumulate

Design change

Design change Design change affect non buoyant releases but not buoyant!

Summary Explosion risk is governed by many parameters that affect each other in a non linear manner We ve seen: 1. Leak orientation 2. How wind affect compositions differently 3. Effect of geometry vary with composition And there are more Scenari o Ign P Pressur e A 3.65% 0.40 B 0.49% 0.02 C........ AAA Is it possible to simplify dispersion and ignition modelling without loosing accuracy? Are we confident these simplifications predict the right answer? AAB AAC

Summary Industry practice for dispersion «Clever methods» and black box predictions (untraceable) Different approaches by different consultants Simplifications in dispersion and ignition modelling 1. The reason for diverging load prediction? 2. The impact may be substantial can affect design recommendations So: Scale back on black boxes and «clever» simplifications? CPU cost are not a valid argument Explicit ignition modelling from CFD results are easily available Increased transparency and trust in results A more explicit standard?

Does anyone benefit from this inconsistency?

Questions?