Скачать презентацию THE BIG DEAL ABOUT BIG DATA IN OIL Скачать презентацию THE BIG DEAL ABOUT BIG DATA IN OIL

Hitachi_Bert-Beals_BAH2013.pptx

  • Количество слайдов: 15

THE BIG DEAL ABOUT BIG DATA IN OIL & GAS BERT BEALS CHIEF TECHNOLOGIST THE BIG DEAL ABOUT BIG DATA IN OIL & GAS BERT BEALS CHIEF TECHNOLOGIST – ENERGY INDUSTRIES 1

Big Data Potential Defining Big Data and Analytics Big Data and analytics technologies describe Big Data Potential Defining Big Data and Analytics Big Data and analytics technologies describe a new generation of technologies and architectures designed to economically extract value from very large volumes of a wide variety of data (structured and unstructured) by enabling high-velocity capture, discovery, and/or analysis. © IDC Energy Insights Visit us at IDC-ei. com 2

SOURCES OF BIG DATA GROWTH THE DATA MULTIPLIER EFFECT MACHINE SENSOR DATA, COMPLEX DATA SOURCES OF BIG DATA GROWTH THE DATA MULTIPLIER EFFECT MACHINE SENSOR DATA, COMPLEX DATA VIDEO RECORDING HUMAN SENSORS DOCUMENTS EMAIL SATELLITE IMAGING VARIETY VOLUME VELOCITY M 2 M LOG FILES ENTERPRISE CONTENT, EXTERNAL SOURCES BUSINESS PROCESS WEB LOGS VARIETY VOLUME SOCIAL DATABASE DATA VOLUME OLTP 1 X 10 X BIOINFORMATICS 100 X More Data with More Complex Relationships…in Real Time and At Scale To manage, govern and analyze 3

BIG DATA – THE PROMISE SORTING THE REALITY FROM THE HYPE § § Big BIG DATA – THE PROMISE SORTING THE REALITY FROM THE HYPE § § Big Data application stack is complex Data is highly fragmented New skill sets Difficult to know where to start MPP Hadoop Search Model Text Database Stream Processing Dark Data NAS ETL Connectors Clickstream Machine Data Audio Video Business Process Web Logs The Reality Structured. HTML Semi-structured Unstructured § Turning data flood into transformative insight § Exploration without preconceived notions In Memory DB Meta Data The Hype Social Media Federation Virtualization No. SQL Analytics Visualization Machine Learning Real Time How does Big Data help my Business? How much will it cost? What’s the ROI? How risky is it? When will I see results? 4

BIG DATA DRIVING BIG INNOVATION TODAY Hitachi Transportation: Bullet Trains • Track/train sensors • BIG DATA DRIVING BIG INNOVATION TODAY Hitachi Transportation: Bullet Trains • Track/train sensors • Keep trains on schedule • Proactive maintenance Hitachi Power: Power Stations • Telemetry from seismic sensors • Time series data Hitachi Construction: Excavators Hitachi, Ltd. : Tokyo Stock Exchange 5 • Operational data from sensors • Insight for fleet managers • High-speed index service • 1/100 th faster

#1 INDUSTRY MEGA DRIVER – SIMPLY STATED: WE NEED MORE ENERGY INCREASE PRODUCTION AND #1 INDUSTRY MEGA DRIVER – SIMPLY STATED: WE NEED MORE ENERGY INCREASE PRODUCTION AND KNOWN ENERGY RESERVES Dramatic increase in demand from emerging economies (non-OECD) Discover new reserves now: 75% of today’s oil was discovered before 1980 © Exxon. Mobil 2012 © BP 2011 6 © Exxon. Mobil 2012 Sources: “ 2012 The Outlook for Energy: A View to 2040”, Exxon. Mobil 2012; and “BP Energy Outlook 2030”, BP 2011

OIL&GAS REPRESENT THE MAJORITY OF THE ENERGY INDUSTRY § 60% of Global Energy demand OIL&GAS REPRESENT THE MAJORITY OF THE ENERGY INDUSTRY § 60% of Global Energy demand fulfilled by Oil&Gas ‒ Natural Gas is the fastest growing fuel source © BP 2011 © Exxon. Mobil 2012 7 Sources: “ 2012 The Outlook for Energy: A View to 2040”, Exxon. Mobil 2012; and “BP Energy Outlook 2030”, BP 2011

Big Data Potential Volume, Velocity and Variety § More Volume § • Exploration: WAZ Big Data Potential Volume, Velocity and Variety § More Volume § • Exploration: WAZ and Iso. Metrix More Velocity • Drilling: Nuclear, electromagnetic, acoustic § • Stream vs. batch processing • Clearer view of potential resources • Unstructured: CAD drawings, specifications, seismic, well log or daily drilling reports • Semi-structured: Processed data • Complexity involving many engineering disciplines Potential Value § Potential Value • Help guide a fracking process • When consequences of failure are great • Show the way to enhancing production • When a delay in processing data may mean missing a bid for an oilfield © IDC Energy Insights Visit us at IDC-ei. com More Variety • Structured: Time series, relational • Real-time data from SCADA, drill heads, flow sensors, or condition sensors • Production: Optical sensors § § Potential Value • Access data previously inaccessible due to multiple access patterns or unstructured nature of data 8

BIG DATA IN ACTION IN OIL&GAS Big Data Analysis = Value Seismic data Pe BIG DATA IN ACTION IN OIL&GAS Big Data Analysis = Value Seismic data Pe tab Bore hole sensors Environmental sensors Weather data yte to Re Production utilization al Tim Storage capacities Exploration e. U ns tru ctu Spot pricing (trading) Geo-technical applications Transportation Production red Da ta Inventory levels Resource planning Demand & forecast Refining Business intelligence Distribution Databases Marketing and retail Office applications 9 Business Decisions E&P Investments Inventory locations Production planning Safety Location data

VELOCITY – INGEST MORE DATA FASTER • Growing data > 300 MB / Km VELOCITY – INGEST MORE DATA FASTER • Growing data > 300 MB / Km 2 early 90 s > 25 GB / km 2 in 2006 > Growing… to PBs / km 2 § Yesterday: 20 – 25, 000 sensors, 500 MB/s – 2 GB/s, 50 – 200, 000 shots, 50 – 200 TB data § Now: Full 3 D acquisition, 8 GB/s – 20 GB/s, 250 TB – 1 PB § Tomorrow: 50 GB/s Sources, Grid Computing Ahmar Abbas: 1 Luigi Salvador, High Performance Computing for the Oil and Gas Industry 2 ML Geovision www. alkorinternational. com 10

VOLUME - INSATIABLE APPETITE FOR INFORMATION § Technology as a competitive weapon ‒ Pushing VOLUME - INSATIABLE APPETITE FOR INFORMATION § Technology as a competitive weapon ‒ Pushing technology boundaries Source: Henri Calandra, John Etgen, Scott Morton – Rice University IT STARTS WITH THE DATA, THEN IT NEEDS TO BE ANALYZED TO EXTRACT INFORMATION ‒ Never resting, methods ready for anticipated technology advances § Big Data in use ‒ A different scale, PB file systems as caches ‒ Live in-feed from sensors from sites world wide 11 Realtime drilling operation center, a fully integrated part of the Valhall drilling operations in Norway.

VARIETY: FIND AND PRODUCE FASTER, SAFER, AND MORE EFFICIENTLY Increase Reserves - Capacity scaling, VARIETY: FIND AND PRODUCE FASTER, SAFER, AND MORE EFFICIENTLY Increase Reserves - Capacity scaling, changing workloads Data Acquisition Data Management Seismic Processing Visual Interpretation Modeling Petrophysical Automation Analysis Property Modeling Simulation Upstream Oil&Gas: § Many interconnected, increasingly complex and rapidly changing workflows § Exponential growth in the volume of data § Computational demands increasing by orders of magnitude § Manage complexity and increase efficiency § Facilitate collaborative computing and secure access to data Increase efficiency and accelerate the TOTAL workflow 12

Big Data Potential Data Access Challenges § Data locked in applications on the desktop Big Data Potential Data Access Challenges § Data locked in applications on the desktop and cannot be shared efficiently § Legacy data that is in proprietary formats § Storage I/O requirements can still be a challenge. There may not be enough capacity or upload bandwidth. § The workloads may be random or sequential or both in an unpredictable pattern. © IDC Energy Insights Visit us at IDC-ei. com 13

UNIFIED STORAGE FOR UPSTREAM OIL AND GAS ORGANIZATIONS Reduce Cost - Reduced complexity, higher UNIFIED STORAGE FOR UPSTREAM OIL AND GAS ORGANIZATIONS Reduce Cost - Reduced complexity, higher efficiency § Accelerated exploration with optimized file storage ‒ Highest per-node IOPs for unpredictable I/O ‒ Ideal for mixed workflows requiring performance for throughput-oriented and also for random data flows ‒ Easy scalability and high availability ‒ CIFS, NFS, i. SCSI, Lustre, Hadoop § One storage platform for E&P and the entire oil and gas enterprise Enterprise ‒ Open file system standards: Unified ‒ Single architecture for file and block ‒ Comprehensive data management across technologies ‒ No compromise of performance versus control 14 Structured Technical Unstructured

OBRIGADO 15 OBRIGADO 15