risk user group meeting 2005 presentation - offshore consequence modelling
TRANSCRIPT
1
DNV Software
Offshore Consequence ModellingHow you can use DNV software to easily model offshore scenarios
Slide 2
Contents
BackgroundOnshore v. OffshoreOffshore Consequence Modelling
ExamplesSmoke ModellingTurbine Exhaust Modelling
2
Slide 3
Offshore versus Onshore
Onshore considerations
Relatively open
Off-site Impacts
Immediate impacts to the public and/or
workers
Slide 4
Offshore Risk Considerations
Enclosed/confined spaceImpingementEscalationConfined explosionsFlaresGas turbine exhaust
EscapeImpairment of Temporary Refuge (TR)Impairment of escape, evacuation and rescue (EER) facilities
Subsea Releases
3
Slide 5
Escalation
Flares
Impingement
Smoke
Subsea releases
Gas turbine exhaust
What special consequences may be modelled?
Offshore Consequence Modelling
Slide 6
Offshore Consequence Modelling
What information do we need?
What are the physical constraints of an offshore platform?
How will this affect my consequence modelling?
What assumptions do I need to make?
Which results are most useful?
4
Slide 7
Software Tools
PHASTConsequence modelling toolIncludes models for all aspects of a release from discharge and dispersion to end fire, explosion and toxic effects
NEPTUNEAdvanced risk analysis toolIncludes models for consequence modelling, leak frequency generation, event tree modelling and offshore specific requirements (ship collision, EER, etc.)
Model SpreadsheetsExpert consequence modelling toolIndividual spreadsheets for each consequence model Offers more modelling flexibility than is available through the standard software
DNV Software
Practical Examples
5
Slide 9
Smoke Dispersion
Smoke generation is caused byFires from initiating releaseLarge fires due to escalation
Smoke dispersion can lead toImpairment of EER facilitiesImpairment of TR
We want to look at distance to concentration results, specifically:Does the cloud reach the TR?How long is it over the TR?Will it ingress into the TR?
Slide 10
Smoke Dispersion Assumptions
Combustion products are primarily a mixture of nitrogen, steam, carbon dioxide and carbon monoxide.
High proportion of nitrogen in airNitrogen tends to dominate in hot combustion products⇒ Nitrogen properties representative of mixture
For hydrocarbon fuels the stoichiometric ration is 15 kg air/kg fuel
Well ventilated areasSufficient air for complete combustion⇒ Smoke production rate = 16 x fuel burning rate
6
Slide 11
Smoke Dispersion Assumptions
Real smoke plume tends to be longer than predicted dispersions of nitrogen gas
⇒ Artificially extend the plume using calm weather conditions, eg F stability
High temperature However if too hot, combustion products would tend to rise clear and give false impression that TR is not impaired⇒ Temperature ≤ 600 ºC
Concentration of interest 1e4 ppm = 1%Indicative of edge of smoke plume that would lead to impairment of visibility
Slide 12
Smoke Dispersion Example
The Unified Dispersion Model (UDM) can be used to model smoke dispersion using the previous assumptions.
The UDM is part of PHAST/SAFETI, NEPTUNE and Model Spreadsheets.
For this example we will be using PHAST.
7
Slide 13
Smoke Dispersion in PHAST
Smoke generation from a pool fire
User Defined Source Model
Material = NitrogenDirection = Vertical
NB: Pool fire results will include burn rate from which release rate can be calculated.
Slide 14
Smoke Dispersion in PHAST
8
Slide 15
Turbine Exhaust DispersionArea of concern for the helideck
Offshore plant and equipment can cause localised environmental anomalies near the helideck and on available flight pathsCan compromise helicopter safety (lift)Must be taken into account during installation design and modifications
We want to look at temperature profiles results, specifically:
Does the cloud interfere with the helideck?Does the cloud interfere with any flight paths?
Slide 16
Turbine Exhaust Dispersion Assumptions
Exhaust is primarily a mixture of nitrogen, oxygen, carbon dioxide and steam.
Nitrogen composes approximately 75% of the mixture⇒ Nitrogen properties representative of mixture
Real exhaust plume tends to be longer than predicted dispersions of nitrogen gas
⇒ Artificially extend the plume using calm weather conditions, eg F stability
High temperature, approximately 500 ºC (optional)
9
Slide 17
Turbine Exhaust Dispersion Example
The Unified Dispersion Model (UDM) can be used to model turbine exhaust dispersion using the previous assumptions.
The UDM is part of PHAST/SAFETI, NEPTUNE and Model Spreadsheets.
For this example we will be using PHAST.
Slide 18
Turbine Exhaust Dispersion in PHAST
User Defined Source Model
Material = NitrogenDirection = Horizontal
NB: Release rate and temperature are part of turbine design specifications.
NB: Discharge velocity can be calculated based on release rate and exhaust exit area.
10
Slide 19
Turbine Exhaust Dispersion in PHASTFind detailed concentration and temperature information in the Detailed Dispersion report.
Can extrapolate the concentration equivalent for temperatures of interest
Slide 20
Concentration equivalentsExtrapolated by plotting a graph and generating trend line in Excel®
Temperature profile graphsCan plot the concentration equivalents to get a view of the temperature profiles
Turbine Exhaust Dispersion in PHAST
Temperature Concentration(º C) (ppm)
50 64,000100 148,000200 317,000300 486,000
11
Slide 21
Turbine Exhaust Dispersion in PHAST
50º C100º C200º C300º C
Slide 22
Summary
Offshore consequence modelling presents very different challenges due to
Physical constraints of the platformEnclosed/confined spaceEscalation impactsEscape impacts
Additional upstream aspectsRisers, wellheadsSubsea releases
However, these challenges can be met through the correct application of your existing software.
12
2005 DNV Software