Upstream Engineering Centre Ocean predictions and the oil

  • Slides: 13
Download presentation
Upstream Engineering Centre Ocean predictions and the oil and gas industry - room for

Upstream Engineering Centre Ocean predictions and the oil and gas industry - room for improvement? Colin Grant Metocean Technical Authority

Ocean predictions and the oil & gas industry • Ocean predictions have uses in

Ocean predictions and the oil & gas industry • Ocean predictions have uses in 3 main areas in the industry − Forecasts – real time operations and short term planning (1 to 5 day) − Reanalysis products / hindcasts – operational planning − “weather windows” − Databases to establish design criteria such as 100 year and 10, 000 year values OP Net Seminar Oslo May 11 2011 2

How useful are present ocean predictions? • Waves − Generally acceptable quality for both

How useful are present ocean predictions? • Waves − Generally acceptable quality for both short term forecasts and as hindcast databases. Suitable calibration with measured data permits use for design studies and operability analyses. • Currents − Lack of accuracy in temporal comparisons – frequently miss peak events when compared to measured data. Issues of sampling. Some skill in certain regions on a climatological basis. Used with care by the industry, for design studies, often after calibration & adjustment. • Water level − Useful products for design when combined with extreme wave predictions to establish total water levels. Setting of platform decks. • Temperatures & Salinity − Industry not a major user. Hindcast archives used for design (flow assurance issues – hydrate formation etc). Oil spill weathering issues • Sea ice conditions − Increasingly important with the move towards Arctic environments OP Net Seminar Oslo May 11 2011 3

Wave and current comparisons with data OP Net Seminar Oslo May 11 2011 4

Wave and current comparisons with data OP Net Seminar Oslo May 11 2011 4

Major challenges & requirements • Industry moving to deeper water and arctic areas. •

Major challenges & requirements • Industry moving to deeper water and arctic areas. • Deep water operations require knowledge of currents through the water column, both as forecasts and as hindcast databases. − Riser and mooring design and control − Combinations of parameters for response-based analyses using joint probabilities of key parameters − Winds, waves & currents − Wave crests & water levels − Waves, currents and sea ice • Example − Gulf of Mexico OP Net Seminar Oslo May 11 2011 5

Gulf of Mexico - Loop & Eddy Forecasts OP Net Seminar Oslo May 11

Gulf of Mexico - Loop & Eddy Forecasts OP Net Seminar Oslo May 11 2011 6

Macondo Oil Spill OP Net Seminar Oslo May 11 2011 7

Macondo Oil Spill OP Net Seminar Oslo May 11 2011 7

Macondo oil spill modeling • Much modeling activity both by BP and the US

Macondo oil spill modeling • Much modeling activity both by BP and the US Govt • BP’s in-house oil spill model is being standardised on SINTEF’s “Marine Environmental Modelling Workbench” − OSCAR – Oil Spill Contingency & Response − DREAM – Dose-related Risk & Effects Assessment Model − DEEPBLOW • Require initialisation and boundary conditions from atmospheric and ocean models in order to run effectively • Several other oil spill modelling providers e. g. ASA, BMT etc http: //www. sintef. no/static/ch/environment/numerical_modelling. htm OP Net Seminar Oslo May 11 2011 8

Scenario – West of Shetland spill • “My Ocean” Resource • Time res –

Scenario – West of Shetland spill • “My Ocean” Resource • Time res – Hourly (+ daily mean) • Spatial res - 0. 1 deg − (approx 6 km) • Depth − Numerous levels to 600 m • Variables − N & E Velocity − Salinity − Temperature − Sea surface height above geoid OP Net Seminar Oslo May 11 2011 9

West of Shetland complexity OP Net Seminar Oslo May 11 2011 10

West of Shetland complexity OP Net Seminar Oslo May 11 2011 10

Issues for oil spill modelling • Availability of input data varies regionally • Prediction

Issues for oil spill modelling • Availability of input data varies regionally • Prediction of deep water oil spill trajectories is now a 3 D dispersion problem • Global models e. g. US Navy HYCOM – daily mean values − Nest finer models from global models • Use the ocean model that the regulator uses to enhance acceptability? − Australia - Blue Link CSIRO / Bo. M / RAN • BP long term collaboration with Imperial College, London − Re. EMS - Regional Environmental Monitoring System − Open source codes − WRF + ROMS (including sea ice) + SWAN + NOAH − Main focus is climate change but also being used to assist in oil spill modeling − Caspian used as a test bed (4 km resolution) − Work on Mediterranean, West Africa, Brazil etc OP Net Seminar Oslo May 11 2011 11

The future of ocean prediction? • Linking modelling with observations − In-situ, remotely sensed

The future of ocean prediction? • Linking modelling with observations − In-situ, remotely sensed (satellite, aerial, HF radar etc) − Formatting and geo-referencing issues − Data assimilation • Product development and dissemination − Web / GIS based − User friendly – aimed at decision makers, not specialists • Integrated approaches to observation, forecasting and ultimately end user problem solving. OP Net Seminar Oslo May 11 2011 12

Questions OP Net Seminar Oslo May 11 2011 13

Questions OP Net Seminar Oslo May 11 2011 13