Center for e-field and Integrated Operations for Upstream Petroleum Activities

Transcription

1 Center for e-field and Integrated Operations for Upstream Petroleum Activities Program 2 - Real-time Reservoir Management and Production Optimization Workshop

2 Program 12:00-12:15 Introduction Bjarne Foss 12:15-12:45 Each participant gives a 2-3 min presentatoin All participants :15 Program 2 - expectations as seen from the IO-center Jon Lippe 13:15-13:45 Program 2 - goals and activities Bjarne Foss 13:45-14:00 Break 14:00-16:00 Reservoir management Norne case study (60 min) Jørg Aarnes History matching challenges (30min) Jon Kleppe/Martin Landrø History matching EnKF (15min) Bjarne Foss Discussion 16:00-16:15 Break 16:15-17:45 Production optimization Important challenges in production optimization and system integration methods Petrostreamz platform (45min) Curtis Whitson FMC study (15min) Ingrid Schjølberg Production optimization and CMU cooperation (30min) Vidar Gunnerud 17:45-18:00 Closing discussion Bjarne Foss 19:00 Dinner at Kvilhaugen Gård All participants 2

3 Scope Present goals and on-going activities in Program 2 Introduce new members to the Program 2 research environment Identify and explore links between activities as a basis for collaboration between researchers in different disciplines Plan for 2008 is in the making. This workshop will provide important input for prioritizing next year s activities 3

4 People and activities 2.1 Development and testing of methods for reservoir model updating Bjarne A. Foss Professor, NTNU Dept. of Engineering Cybernetics Program leader 2 + supervisor/researcher Systems and control Jørg Aarnes Researcher, SINTEF Applied mathematics Project leader 2.1 (pilot study only) Reservoir simulation Knut Andreas Lie Senior researcher, SINTEF Applied mathematics Supervisor/researcher Reservoir simulation Petter Almklov NTNU Social Research Ltd. Researcher Work processes Jon Kleppe Martin Landrø Professor, NTNU Dept. of Petroleum Engineering and Applied Geophysics Supervisor/researcher Reservoir simulation Professor, NTNU Dept. of Petroleum Engineering and Applied Geophysics Supervisor/researcher Reservoir seismics Khalid Aziz, Jan D Jansen Stanford University, TU Delft Supervisor/researcher Reservoir simulation Petroleum eng Eka Suwartadi NTNU Dept. of Engineering Cybernetics PhD researcher Systems and control Foss supervisor 4 Agus I. Hasan Pamela Tempone Mosen Dadashpor Amir Ghaderi Alexey Stovas NN NN NTNU Dept. of Engineering Cybernetics PhD researcher Systems and control NTNU Dept. of Petroleum Engineering and Applied Geophysics PhD researcher Reservoir NTNU Dept. of Petroleum Engineering and Applied Geophysics PhD researcher Reservoir NTNU Dept. of Petroleum Engineering and Applied Geophysics NTNU Dept. of Petroleum Engineering and Applied Geophysics Foss supervisor Kleppe supervisor, Landrø co-supervisor (partial funding) Kleppe supervisor, Landrø co-supervisor (partial funding) PhD researcher (postdoc aug 2007-aug 2009) Reservoir seismic Landrø supervisor Senior postdoc reseacher Reservoir seismic NTNU Dept. of Petroleum Engineering and Applied Geophysics PhD researcher Reservoir NTNU Dept. of Petroleum Engineering and Applied Geophysics PhD researcher Reservoir Kleppe, Landrø supervisor Kleppe, Landrø supervisor

5 People and activities 2.2 Production optimization Ingrid Schjølberg Curtis Whitson Michael Golan SINTEF Applied Cybernetics Professor, NTNU Dept. of Petroleum Engineering and Applied Geophysics Professor, NTNU Dept. of Petroleum Engineering and Applied Geophysics Senior reseacher Supervisor/researcher Supervisor/researcher Systems and control Petroleum engineering Petroleum engineering Erik Ydstie Professor, Carnegie Mellon University Supervisor/researcher Systems and control Michael Wartman Silvya Dewi Rahmawati Carnegie Mellon University NTNU Dept. of Engineering Cybernetics PhD researcher PhD researcher Systems and control Systems and petroleum engineering Ydstie supervisor - Foss co-supervisor Foss supervisor - Whitson co-supervisor 2.3 Courses applying IO technoogy Vidar Gunnerud NTNU Dept. of Engineering Cybernetics PhD researcher Optimization Foss supervisor NTNU Dept. of Petroleum Engineering and Applied Snjezana SunjergaGeophysics PhD researcher Michael Golan NTNU Dept. of Petroleum Engineering and Applied Geophysics Professor, NTNU Dept. of Petroleum Engineering and Applied Geophysics PhD researcher Supervisor/researcher Petroleum engineering Petroleum engineering Petroleum engineering Whitson supervisor Whitson, Golan supervisor 2.4. New pilot studies Morten Hovd Robert Drysdale Professor, NTNU Dept. of Engineering Cybernetics Project leader 2.3 Researcher, SINTEF Petroleum Research Project leader 2.4 5

6 Program 12:00-12:15 Introduction Bjarne Foss 12:15-12:45 Each participant gives a 2-3 min presentatoin All participants :15 Program 2 - expectations as seen from the IO-center Jon Lippe 13:15-13:45 Program 2 - goals and activities Bjarne Foss 13:45-14:00 Break 14:00-16:00 Reservoir management Norne case study (60 min) Jørg Aarnes History matching challenges (30min) Jon Kleppe/Martin Landrø History matching EnKF (15min) Bjarne Foss Discussion 16:00-16:15 Break 16:15-17:45 Production optimization Important challenges in production optimization and system integration methods Petrostreamz platform (45min) Curtis Whitson FMC study (15min) Ingrid Schjølberg Production optimization and CMU cooperation (30min) Vidar Gunnerud 17:45-18:00 Closing discussion Bjarne Foss 19:00 Dinner at Kvilhaugen Gård All participants 6

7 Program 12:00-12:15 Introduction Bjarne Foss 12:15-12:45 Each participant gives a 2-3 min presentatoin All participants :15 Program 2 - expectations as seen from the IO-center Jon Lippe 13:15-13:45 Program 2 - goals and activities Bjarne Foss 13:45-14:00 Break 14:00-16:00 Reservoir management Norne case study (60 min) Jørg Aarnes History matching challenges (30min) Jon Kleppe/Martin Landrø History matching EnKF (15min) Bjarne Foss Discussion 16:00-16:15 Break 16:15-17:45 Production optimization Important challenges in production optimization and system integration methods Petrostreamz platform (45min) Curtis Whitson FMC study (15min) Ingrid Schjølberg Production optimization and CMU cooperation (30min) Vidar Gunnerud 17:45-18:00 Closing discussion Bjarne Foss 19:00 Dinner at Kvilhaugen Gård All participants 7

8 Program 2 Research structure Tools Continuity Applications Technology transfer Basic research New, daring directions Deliverables: PhD degrees Publications (level 2) Case studies Prototype methods and algorithms Publication is a mandatory deliverable for all Program 2 projects 8

9 Level 2 publications - examples Petroleum engineering SPE Drilling & Completion SPE Journal SPE Production & Facilities SPE Reservoir Evaluation and Engineering Journal of Canadian Petroleum Technology System, optimization, control Automatica Journal of Process Control IBM SystemsJournal SIAM Journal on Optimization 9

10 Level 2 publications - examples! #$ " %& #$ "' Special issues on Business Collaboration Compliance Management Information-Based Medicine Model-Driven Software Development Online Game Technology e-business Management Information Integration... 10! +, ( -./ )* " " %$ + 0.

11 Decisioin horizons vary a lot There is a need for models of wells, near well regions and production system. There is a need for system integration methods There is a need for detailed dynamic drainage models. This is critical for optimal asset management Oilfield Review

12 Program 2 Closing the Decision Loop Reservoir Wells, Transport and Production System Data from Field Operation /Oilfield Review 2005/ Many data sources Production data Process data Maintenance data Test/campaign data Seismic data Well logs Drilling data Data rates and transmission rates are increasing (Much) more data available Data integration platforms Easier to integrate various data Data quality and data reliability is improving new sensors and transmission More trustworthy data 12

13 Program 2 Closing the Decision Loop Reservoir Wells, Transport and Production System Data from Field Operation Automated real-time decision making Control systems Emergency/process shutdown systems Human-in-the-loop realtime decision making Visualization of current state Assess current situation /Oilfield Review 2005/ Decide course of action 13 Implement decisions Control and Operational Operational Decision Making Decision Making /Advantech/ /CO-OP, Canada/ Real-time feedback loop Manual Operational Decision Making Automated Operational Decision Making

14 Program 2 Improving the Decision Loop New methods for assessing the state of the system Reservoir Wells, Transport and Production System Allows us to continuously monitor key variables and key indicators Data from Field Operation Research areas New Data Assimilation Methods New Reservoir Simulation Methods New Reduced Order (Proxy) Models System Integration Methods Control and Operational Decision Making Data Filtering, Data Assimilation, Model Updating Manual Operational Decision Making Automated Operational Decision Making 14

15 Data filtering - data assimilation - model updating - history matching The reservoir models are needed for estimating current states predicting the consequence of alternative strategies Norsk Hydro P. Sarma et al,

16 Program 2 Improving the Decision Loop New methods for realtime decision making Higher quality decisions making based on experience and optimization tools Reservoir Wells, Transport and Production System Data from Field Operation Research areas New Data Assimilation Methods New Reservoir Simulation Methods New Optimization Techniques New Reduced order (proxy) models System Integration methods Control and Operational Data Filtering, Model-based Decision Making Data Assimilation, Optimization, Model Updating Predictive Control Manual Operational Decision Making Automated Operational Decision Making 16

17 Optimized operational decision making - example Snøhvit 17 Challenge: Optimize the supply chain (well => storage tanks) when an LNG-ship arrives late

18 Optimized operational decision making - example Snøhvit This (red) LNG-ship is delayed Use heuristics - reduce production Optimized solution - reduce and redistribute production A system integration platform is important 18 for these problems

19 Optimized decision making - general concept 19 Brouwer et al, 2004

20 Program 2 One view Physical system Reservoir Wells, Transport and Production System Data from Field Operation Foundation: The world is represented in two alternative ways Control and Operational Data Filtering, Model-based Decision Making Data Assimilation, Optimization, Model Updating Predictive Control Mathematical representation/ Abstract models Manual Operational Decision Making Automated Operational Decision Making 20

21 Program 2 Research structure 2.2 Production optimization System Integration methods New Reduced order (proxy) models New Optimization methods 2.1 Reservoir management System Integration methods New Reservoir Simulation Methods New Reduced order (proxy) models New Data Assimilation methods Oilfield Review

22 Program 2 Link to Program 4 Integration across disciplines To capitalize on New methods for assessing the state of the system and New methods for real-time decision making close cooperation with Program 4 is necessary. Innovative visualization and interpretation technologies is a must. Tools, incentives and work processes to facilitate a common situational awareness and mutual trust across disciplines and geographical locations is a must. /answers.com/ 22

23 Program 2 Link to other programs To capitalize on New methods for assessing the state of the system and New methods for real-time decision making close cooperation with Program 4 is necessary. Innovative visualization and interpretation technologies is a must. Tools, incentives and work processes to facilitate a common situational awareness and mutual trust across disciplines and geographical locations is a must. Close links to Program 3 Operation and maintenance Predictive Maintenance Control Condition Monitoring of Facilities Integrated planning 23

24 Program 2 Scope of Work New Data Assimilation Methods New Reservoir Simulation Methods New Optimization Techniques New Reduced order (proxy) models System Integration methods /Oil and Gas Journal/ Program 2 - Value Increase in Reserves Increased Production Rates Reduced Operational Expenses Reduced Risk 24 /grinningplanet.com/

25 Program 12:00-12:15 Introduction Bjarne Foss 12:15-12:45 Each participant gives a 2-3 min presentatoin All participants :15 Program 2 - expectations as seen from the IO-center Jon Lippe 13:15-13:45 Program 2 - goals and activities Bjarne Foss 13:45-14:00 Break 14:00-16:00 Reservoir management Norne case study (60 min) Jørg Aarnes History matching challenges (30min) Jon Kleppe/Martin Landrø History matching EnKF (15min) Bjarne Foss Discussion 16:00-16:15 Break 16:15-17:45 Production optimization Important challenges in production optimization and system integration methods Petrostreamz platform (45min) Curtis Whitson FMC study (15min) Ingrid Schjølberg Production optimization and CMU cooperation (30min) Vidar Gunnerud 17:45-18:00 Closing discussion Bjarne Foss 19:00 Dinner at Kvilhaugen Gård All participants 25

26 26

27 Scope Develop and test new methods for automated history matching Why There is a lack of efficient methods for updating reservoir models Methodological foundation Fast simulation methods Data assimilation methods Work processes Deliverables Report real cases Prototype software code Dissemination 27 Program 2.1 Reservoir management Data assimilation - History matching Publications, workshops and talks 7 PhD candidates Active partners Statoil (Norne) Activities /streamsim.com/ NTNU (Petroleum, Cybernetics & Social science) SINTEF (Applied mathematics) Stanford University Master and PhD projects ( ) One case study project ( )

28 Scope Develop and test concepts and methods for integrating operation beyond traditional boundaries Why Program 2.2 Production optimization System integration methods The operation of the production chain is composed into parts suboptimal operation is the rule Methodological foundation Use state-of-the-art system integraton platforms Case studies Deliverables Report from real cases Prototype software code Dissemination 28 Publications, workshops and talks One PhD candidates Active partners FMC, Hydro Petrostreamz, Petrobras Activities /emersonprocess.com/ NTNU (Petroleum & Cybernetics) SINTEF (Cybernetics) Two case study projects ( ) One PhD project ( )

29 Scope Develop and test reduced order models for system integration, realtime use and optmization Why Full-fledged model are not suitable for real-time applications Methodological foundation Program 2.2 Production optimization Reduced order (proxy) models Model reductions techniques high fidelity => low cost models Physics based and emprical models Deliverables Testing on real cases New models and prototype software code Dissemination 29 Publications, workshops and talks Two PhD candidates Active partners FMC, Hydro Petrostreamz NTNU (Petroleum & Cybernetics) SINTEF (Cybernetics) Activities Two PhD projects ( )

30 Scope Develop and test new optimization method for full-field production optimization Why Program 2.2 Production optimization Decentralized optimization Present optimization technology is a constraining factor for real-time production optimization Methodological foundation Lagrangian decomposition and mixed integer optimization Active partners Hydro (Troll & Research) /mathworld.wolfram.com/ NTNU (Cybernetics & Industrial Economics) Carnegie Mellon University (CMU) Deliverables Report real cases Prototype software code Dissemination 30 Publications, workshops and talks Two PhD candidates Activities Master and PhD projects ( )

31 MORE Concrete examples (as we obtain such results) related to slide 2-8 Concretize research activities (case, methods, results, partners, 2.2: Presentation FMC March 2007 Concretize System Integration (Petrostreamz, FMC) Concretize people Concretize academic axis (including CMU, Stanford, TU Delft) Look into Stanford slides Where are we in 5 years (comments from last board meeting) What differentiates Prog.2 from other activities 31

32 32

33 Model Predictive Control (MPC) A successful concept from process control domain Now Future Model Reservoir, Well, Transport and Production System Alternative choices Consequences A Consequences B Consequences C Consequences D Continous updating with real time data 33