SPE logo

Technical Sessions

Session 1: Scalable Solvers for Linear and Nonlinear Systems

Session Chairs: Arthur Moncorge, Total; Kees Vuik, Delft University of Technology; Klaus Stueben, SCAI; Knut-Andreas Lie, SINTEF

A primary challenge for reservoir simulation is the accurate description of multiphase flow in porous, highly heterogeneous and fractured media with very complex geometries. The lack of robust and efficient solvers for fully implicit formulations is still one of the main bottlenecks that most simulator developers face in the oil and gas industry. Generally, the underlying systems of partial differential equations are nonlinear, highly non-symmetric and indefinite. The condition number and degree of coupling of these systems may be subject to dramatic changes due to abrupt flow variations induced by the high-heterogeneity and complex well operations during the simulation process.

In addition, reservoir models have been growing in complexity regarding geometry, discretisation and physical models including, for instance, thermal and chemical influences, geomechanics, heterogeneity, and size, causing these systems to get increasingly large and difficult to solve. In fact, the computational time required to solve these systems of equations is still a major bottleneck in the practicability of numerical simulation. Hence, advanced reservoir simulators need to combine the numerical scalability (optimality) of efficient solvers with the parallel scalability of modern hardware.

This session addresses all relevant numerical aspects such as the efficient and robust treatment of linear and nonlinear problems, the treatment of fully coupled systems (either the Jacobian directly or by CPR-type approaches), efficient preconditioning techniques, fast linear solvers (hierarchical solvers such as multigrid or multilevel ILU as well as parallel performance aspects).

0930

 

A Real-Field Multiscale Reservoir Simulator
Jostein Natvig, Kyrre Bratvedt, Seong Lee, Antonina Kozlova, Zhuoyi Li, Carlos Boneti, Shingo Watanabe, Yifan Zhou, Schlumberger

1000

 

A Constrained Pressure Residual Multiscale (CPR-MS) Compositional Solver
H. Hajibeygi, M. Cusini, Delft University of Technology; A.A. Lukyanov, Schlumberger

1030

 

Tuning Systems Algebraic Multigrid in Reservoir Simulation
Sebastian Gries, SCAI

1130

 

Asynchronous Multirate Newton—A Class Of Nonlinear Solver That Adaptively Localises Computation
Soham Sheth, Rami M Younis, The University of Tulsa

1200

 

Scalable Hybrid Preconditioners
P. Jolivet, F. Nataf, N. Spillane, University Paris

1230

 

Comparison of Multiscale Finite Volume Solvers on Stratigraphic Grids
Knut-Andreas Lie, Olav Moyner, Stein Krogstad, SINTEF

Alternates/Knowledge Sharing ePosters

1030
ePoster
Station 1

 

A Parallel Algebraic Multigrid Method For Elliptic Problems With Strongly Discontinuous Coe_Cients
Markus Blatt, HPC-Simulation—Software & Services

1045
ePoster
Station 1

 

Fully Implicit Numerical Modelling Of Natural/Hydraulic Fracturing Using XFEM
K.C. Das, Ankur Narang, IBM; Harshita Dude, IIT

 

Session 2: Big Data Discovery and Its Impact on Decision Making

Session Chairs: Hector Klie, ConocoPhillips; Sylvain Desroziers, IFP Energies Nouvelles; Ahmet Duran, Istanbul Technical University

Data discovery or mining of large volumes of data is a challenging but, without a question, a revolutionising topic in several branches of science and engineering. Major challenges include acquisition, cleaning, storage, transformation, and visualisation of these data to be able to extract relevant knowledge that may lead to improved predictive models, reliable decisions, and ultimately, unlock new business opportunities. In oil and gas industry there has been a notable spread in the use of sensors, mobile technology, and in general digital equipment to be able to track the complex physics and somehow mitigate the uncertainty associated with large scale reservoirs.

The increasing adoption of digital technology is shifting the business towards the processing of unprecedented volumes of data in a timely and reliable fashion. Although there is a general awareness about data mining and analytics in the oil and gas industry, most challenges and business potentials associated with the processing of big data are practically unknown. In this session, we will pay particular attention to state-of-the-art big data discovery algorithms and practices that may be differential for improving the decision making process. Presentations and discussions would be oriented to identify challenges associated with the generation of fast, integrated, automated, and self-learning decisions workflows in oil and gas.

1400

 

Machine Learning Analytics For Big Data From Reservoir Simulations
Jochen Garcke, Rodrigo Iza-Teran, University of Bonnand Fraunhofer SCAI

1430

172982

Principles Of Big Data Algorithms And Application For The Unconventional Oil And Gas Resources
Avi Lin, Halliburton

1500

172983

Cloud-Based Remote Visualisation Of Big Data To Subsurface Exploration
Nicole Sultanum, Alécio Binotto, Renato Cerqueira, IBM

1530

172984

Spectral Effects Of Large Matrices From Oil Reservoir Simulators On Performance Of Scalable Direct Solvers
Ahmet Duran, Mehmet Tuncel, Istanbul Technical University

1600

172985

State Of Art Emerging Technology For Technical Data Mining And Analysis
Kaushal Kishore, Puneet Sharma, Pooja Khanapurkar, Halliburton

Alternates/Knowledge Sharing ePosters

1500
ePoster
Station 1

172994

An Advanced Solution For Unit Data System Management
Rajeev Panda, Vishal Harne, Halliburton

1515
ePoster
Station 1

 

Systems Theory Approach For Heterogeneous Permeability Field Inversion
Sergiy Zhuk, Tigran Tchrakian, Sean McKenna, IBM

Session 3: Hardware Advances and Code Optimisation

Session Chairs: Jose Rodrigues, Petrobras; Bret Beckner, ExxonMobil; Ulisses Mello, IBM; Arthur Moncorge, Total

Hardware advances are a key driver in the development of effective reservoir simulators. Hardware options for high-performance computing are rapidly progressing in several directions. Current high performance computers have eclipsed one million cores with larger systems under development. Modern high performance computing systems typically include mixtures of CPUs, GPGPUs, and many integrated cores, active storage, and active networks yielding a variety of novel computational patterns which have to be considered if scalable parallel algorithms and efficiently optimised code are to be taken to the next level.

Development of efficient parallel algorithms on these systems must include an understanding of memory access and optimal data layout, message passing, connectivity topologies, and parallel I/O. Numerical methods to maximise locality or minimise node-to-node communication are needed to achieve good strong scalability performance.

These and other computational issues on high performance parallel computing systems present opportunities and challenges for reservoir simulation developers. This section will explore parallel code optimisation techniques under the current HPC landscape, as well as shed some light into what are the emerging HPC trends and perspectives for next few years in this area.

0900

172986

Challenges And Opportunities For Hpc Cloud In Natural Resources
Marco A. S. Netto, Alecio Binotto, IBM

0930

 

Parallel Efficiency And Algorithmic Optimality In Reservoir Simulation On Gpus
K. Esler, D. Dembeck, K. Mukundakrishnan, V. Natoli, J. Shumway, Y. Zhang , Stone Ridge Technology; J. Gilman, H-Z. Meng, iReservoir

1000

172987

Combining Aggregation And Classical Algebraic Multigrid In The CPR-AMG Linear Solver
Serge Gratton, Pascal Hénon, Pavel Jiranek, Xavier Vasseur, Total

1100

 

GPU-Accelerated Algebraic Multigrid For Commercial Applications
Joe Eaton, NVIDIA

1130

172988

A Multilevel Preconditioner And Its Shared Memory Implementation For New Generation Reservoir Simulator
Shuhong Wu, Baohua Wang, Qiaoyun Li, Xiaobo Li, Hua Li, PetroChina; Chensong Zhang, Chinese Academy of Sciences; Jinchao Xu, Penn State University

1200

172989

Reservoir Simulation Prototyping Platform For High Performance Computing
Arthur Moncorge, S. Jauré, R. de Loubens, Total

Alternates/Knowledge Sharing ePosters

1030
ePoster
Station 1

172995

Exploring Efficient Alternatives For High Performance Computing Requirements In Coupled Fluid-Flow And Stress Simulations For The Oil And Gas Industry
E. Rodrigues, J.M. Serra Segura, P. Vargas Mendoza, R. Ausas, K. Das, U. Mello, M. R. Lakshmikantha, IBM

Session 4: Novel Discretisations of PDEs and Impact on Solvers

Session Chairs: Mary Wheeler, University of Texas at Austin; David Keyes, KAUST; Knut-Andreas Lie, SINTEF; Arthur Moncorge, Total

Choice of discretisation scheme is pivotal in reservoir simulation since it is the bridge between the multi-scale, geometrically complex continuum and the finite-dimensional nonlinear and linear operators of the computational representation. Any discretisation is a balance between represent ability and solvability.  Recent years have witnessed great innovation in discretisation schemes for PDEs generally, some of which offer interesting practical opportunities for the reservoir simulation community, and there have been some advances specific to the complex physics of reservoirs. The talks of this session identify advantages of disadvantages of some new schemes in the context of practical reservoir modeling and discuss their implications for the algebraic solution process downstream.

1330

172990

A Multiscale Mortar Method And Two-Stage Preconditioner For Multiphase Flow Using A Global Jacobian Approach
Benjamin Ganis, Kundan Kumar, Gergina Pencheva, Mary F. Wheeler, University of Texas at Austin; Ivan Yotov, University of Pittsburgh

1400

 

High-Resolution Numerical Simulation Of Naturally Fractured Carbonate Reservoirs
Sebastian Geiger, Heriot-Watt University

1430

172991

Robust Fully-Coupled Multiphase Flow And Geomechanics Simulation

Faruk O. Alpak, Shell

1530

 

Uncertainty Quantification With Polynomial Chaos Expansions For Robust Static And Dynamic Modelling
Pallav Sarma, Chevron

1600

172988

Coupled Flow And Geomechanics For A Fractured Poreolastic Medium

V. Girault, University Paris; K. Kumar, G. Pencheva, G. Singh, M. F. Wheeler, University of Texas at Austin

1630

 

A Review Of Linear Solvers And Preconditioners In Reservoir Simulation

G.J. Shaw, T.B. Jonsthovel, Schlumberger; C. Vuik, Delft University

Alternates/Knowledge Sharing ePosters

1500
ePoster
Station 1

 

Innovative Accelerated Real Time Numerical Simulator For The Multi-Physics And Chemistry For Unconventional Oil And Gas Stimulation Treatments
Avi Lin, Dinesh Ananda Shetty, Joshua Camp, Halliburton

Session 5: Inverse Problems, Uncertainty Quantification and Optimisation

Session Chairs: Ruben Juanes, MIT; Khalid Aziz, Stanford University; Long Nhgeim, Computer Modelling Group

The emergence of the formal application of optimisation techniques for both development and operation of oil fields has introduced new challenges. How to develop multiple realisations of the reservoir to capture uncertainty? How to include geology in the history matching and uncertainty workflow? 

What are the best optimisation techniques for well scheduling and well control? When is it possible to find the global optimum? How can we minimise the number of reservoir simulations to make optimisation practical for large reservoirs? How to handle a large number of parameters? How to handle fields with a large number of wells? How can we efficiently assimilate new data in model updating and finding the optimum? What kinds of proxies are suitable for optimisation? 

0900

 

Metamodelling And Pareto Optimisation For History Matching Of Reservoir Simulation Models
T. Clees, N. Hornung, SCAI

0930

 

Massively Distributed Simulation And Optimisation On Commercial Compute Clouds
Pallav Sarma, Wen Chen, Jim Owens, Xian-Huan Wen, Chevron

1000

 

Application Of Gradual Deformation Method For History Matching Brugge Field Stud
Rubakumar Sankararaj, Andrew Wadsley, Nathan Reeves, Stochastic Simulation Limited

1100

 

Statistically Based Objective Functions Forassited History Matching And Model Selection
Gregory R. King, Baurzhan Kassenov, Aizada Abdrakhmanova, Moon Chaudhri, Elrad Iskakov, Tengizchevroil, Daniel H. Dye, ExxobMobil

1130

 

Use Of Flow Diagnostics In Waterflood Optimisation
Stein Krogstad, Knut-Andreas Lie, Olav Møyner, SINTEF

1200

172992

Automated History Matching Using Combination Of Adaptive Neuro Fuzzy System (ANFIS) And Differential Evolution Algorithm

Muzammil Hussain Rammay, A. Abdulraheem, King Fahd University of Petroleum and Minerals

Session 6: Case Studies/Large Scale Simulations

Session Chairs: Greg King, Chevron; Jeroen Vink, Shell; Turgay Ertekin, Pennsylvania State University; Wu Shuhong, PetroChina

Over the last several years many previously advanced or niche simulation techniques slowly are becoming more main stream: sssisted history matching and uncertainty quantification using design of experiment and other (stochastic or ensemble based) techniques, closed loop and life-cycle field development optimisation, integrated subsurface/facility flow modelling, coupled flow-geomechanics simulations, field-scale EOR development optimisations, and detailed high-resolution (conventional and unconventional) process simulations. Often such large scale simulation studies are pushing the limits on what can be achieved on current hardware and using present-day simulation techniques. In this session compelling examples of large scale or otherwise cutting-edge simulation studies will be presented.

1330

 

Ten Billion Cell Reservoir Simulations
P. Crumpton, U. Middya, L.Fung,  A.Dogru, T.Al-Shaalan, Saudi Aramco

1400

 

A Coupled Surface Facility For A Billion Cell Giant Oil Reservoir With Compositional Fluid Description
U.Middya,  T.Al-Shaalan, T.Al-Qasim, A.Dogru, Saudi Aramco

1430

 

Giga-Model Simulations In A Commercial Simulator—Challenges And Solutions
Dominic Walsh, Mark Wakefield, Paul Woodhams, Schlumberger; Matthieu Brucher, Eguono Obi, Total

1530

172993

 Induced Fracture Modelling and Its Integration with Pressure Transient Analysis—Study For A Shallow-Water Offshore Field, South-East Asia – Part 2
Adi Anand, Shell

1600

 

 Large-Scale Integrated Reservoir-Facility Simulations Of A Multi-Reservoir Miscible Gas Flood Of Sour Oil Reservoirs: The Importatnceof Multi-Fidelity Fluid Modelling
Jean van Berkel, Peter Cornelisse, Eric Hendriks, Toon Weisenborn, Chris Welch, Shell

1630

 

Full Field Reservoir Simulation Through Coupled Multi-Scale Models For Solvent Flooding Enhanced Oil Recovery
Abdullah Alkindi, Petroleum Development Oman; Paul te Riele, STO

Alternates/Knowledge Sharing ePosters

1500
ePoster
Station 1

 

“Hands free" Capacity Calibration of a Super Giant Gas Field
Javier Loaiza, Shell