Ruben Schulkes Self-similarity and the need for subsea technology
rms Date: Page: 2 Outline of presentation Trends & challenges Technological needs Research needs
rms Date: Page: 3 Trends Since the early 80’s the global production rate has exceeded discoveries Number of new discoveries of Giant and Super Giant fields (Ultimate Recoverable Resource (URR) > 500Mbbl) decreases Is global oil production about to peak? ?
rms Date: Page: 4 Self-similarity? Production histories from selected oil provinces look very similar This self-similarity may suggest an eminent peak in global oil production Increasing focus on - Increased Oil Recovery - production from small fields (URR<50Mbbl) - Heavy (ultra-heavy) crudes US Norway North Sea
rms Date: Page: 5 Potential IOR in large fields Average recovery factor in NCS is less than 50% Total volume of unrecoverable reserves > 3000 MSm 3 oil (~5 x GDP)
rms Date: Page: 6 Small fields are challenging % more fields 5% more oil Small fields do not contain much oil (by definition!) Development of small fields is only possible when - Field development costs are low - Operating costs are low
rms Date: Page: 7 Subsea field developments Trends - Development of smaller fields - Development of large fields subsea-to-beach Being able to predict what happens in flow lines and processing units is key to success - Further development of accurate simulation tools is crucial
rms Date: Page: 8 Current modelling approaches OLGA - 1D frame work, pre-integrated 2D model for stratified flow LEDA - Coupled 1D-3D simulator, 1D/2D for pipe flow, 3D for processing units Fluent, StarCD, CFX - 3D codes
rms Date: Page: 9 CFD - Reasonable on “short” length scales - Difficulties with interacting dispersed phases and free boundaries - Difficult closure relations (turbulence/dispersed phase interactions) 1D models - Difficulties on short length scales which influences long-scale phenomena - Difficult to get closure relations which are universally valid L/D CFD 10 1 Physical complexity Hierarchy of models 1D models Simplified 1D models
rms Date: Page: 10 Increased processor speed and algorithm improvements lead to speed-up of factor 1000 per decade in CFD codes Demand for more accurate physical modelling leads to slow down of current 1D and future 2D simulators - Pre-integrated models (use of turbulent velocity field information) Simulation time CFD 1D codes Trends
rms Date: Page: 11 CFD simulation of 2-phase system - 30m pipe, 2x10 4 cells requires 5 days CPU time (4 processors) to simulate 30 second real time Field case - 5km pipe, 1hr transport time - 3x10 6 cells, days of simulation time With speed-up factor of 1000/decade it becomes possible to perform real- time 3D transient simulation of 5km pipe within 20 years Naive application of CFD
rms Date: Page: L/D FACE 10 1 Horizon/LEDA Physical complexity Multiphase research in Norway Research is driven by industry and institutes Academic activity is scattered - Who is doing the really difficult ground work? - Where are the future researchers being educated? Industry
rms Date: Page: 13 hydro.com
rms Date: Page: 14 Massively parallel simulations Use information from 1D/quasi-1D/2D codes as pre-conditioners for 3D solution Multi-scale simulator - Decouple high-frequency/short-length scale effects from low-frequency/long- length scale phenomena Intelligent CFD
rms Date: Page: 15 Potential IOR in large fields (2)