rs2014 houstonnov hortonworksdevonenergy kellykohlleffelrickhowell...

34
© Copyright 2014-15 OSIsoft, LLC. Presented by How to achieve Operational Intelligence by becoming a Data- Driven Organization Frank Besch, Director of Business Integration, Noble Energy Rick Howell, Real-time Information Systems Supervisor, Devon Energy Kelly Kohlleffel, Industry Executive, Hortonworks Matt Ziegler, Product Manager, OSIsoft

Upload: john-archer

Post on 29-Sep-2015

5 views

Category:

Documents


1 download

DESCRIPTION

Big Data and operational intelligence in energy

TRANSCRIPT

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    Presented by

    How to achieve

    Operational

    Intelligence by

    becoming a Data-

    Driven Organization

    Frank Besch, Director of Business Integration, Noble Energy

    Rick Howell, Real-time Information Systems Supervisor, Devon Energy

    Kelly Kohlleffel, Industry Executive, Hortonworks

    Matt Ziegler, Product Manager, OSIsoft

  • Copyr i gh t 2014-15 OSIso f t , LLC. 2

    83% improved process cycle times

    12% less operating expense

    6% more profitable

    49% had payback in one year or less

    54% report ROIs >100%

    Sources: Harvard Business Review, Forbes, IDB

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    Modern Information Architecture

    3

    3

  • Copyr i gh t 2014-15 OSIso f t , LLC. 4

    Real-time Data isnt perfect

    Naturally incomplete (delays, shutdowns)

    Not evenly spaced

    Doesnt look and behave like SQL (RDBMS)

    Subject to errors in measurement

    Varies in fidelity

    Needs Context (Assets, Events)

    Hard to Collect effectively

    The Truth about Real-time Data

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    Decision-Ready Data

    5

    PI AF

    PI Server

    FinanceTrading ERP

    Central OpsR&D Facilities

    Central ITData Science

    Single

    Source

    Other

    Data 1

    Other

    Data 2

    Other

    Data 3Model

    A

    Model

    B

    Raw Data

    Conditioned,

    Trustworthy, Targeted

    Data

  • Copyr i gh t 2014-15 OSIso f t , LLC. 6

    Process Optimization

    Quality Improvement

    Asset Health & Uptime

    Energy Efficiencies

    Regulatory Requirements

    Safety

    Assets Assets Assets Assets Assets Assets

    Wherever your data

    starts

    to Wherever your data

    needs to go

    Enterprise

    Infrastructure

    Real-Time Data Infrastructure

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    Evolving Data Goals

    Real-time visibility

    Real-time and historical view across all assets

    Fleet-wide performance comparison

    Prediction and Prevention

    HMIPI System via

    PI ProcessBook

    Big Data?

    Monitoring Process Optimization Benchmarking System Optimization

    7

  • Copyr i gh t 2014-15 OSIso f t , LLC. 8

    64% of large enterprises plan to implement a big data project in 2014, but 85% of

    the Fortune 500 will be unsuccessful in doing so. These time-consuming data

    preparation tasks are largely to blame.

    Gartner

    Data cleansing and preparation tasks can take 50-80% of the development time

    and cost in data warehousing and analytics projects.

    poor data quality is the primary reason for 40% of all business initiatives failing to

    achieve their targeted benefits.

    Harvard Business Review

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    Project CAST

    9

    All PI System data delivered on your terms, in your language, to

    the tools you use, and to the people that can make a difference.

  • Copyr i gh t 2014-15 OSIso f t , LLC. 10

    Guaranteed Delivery & Storage Full Fidelity of Sensor Optimized for Real-Time Backup/Restore HA Security

    System of Record

    Statistical Analytics

    Visual Analytics

    Designed to Analyze Large Sets Expects that the Data Exists Problem Defines Data Shape Typically Evenly Spaced in Time

    Needs:

    Analytics Packages

    From Raw Data to Decision Ready Data

    Project CAST

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    Anatomy of a Data Publication

    11

    What our early adopters are saying

    input

    output

    Describes Assets

    Time Ranges,

    Events, Filters,

    Output Style

    Many styles, columnar, json, file

    Maintain one version of the truth no

    matter where the data is used

    Captures business rules to make

    the data ready for broad

    consumption

    PI manages it. Great!

  • Copyr i gh t 2014-15 OSIso f t , LLC. 12

    Project CAST Components

    Business Intelligence Accelerator

    Publication Buffer

    PI Server

    PI System

    SAP HANA

    PI Integrators for

    Oracle

    Hadoop

    Trustworthy Data

    Publications

    Coming to the PI System

    Managed

    by the PI

    System

    AF

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    The Minimum Viable Product Process

    13

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    Presented by

    Business Driven Data

    Rick Howell, Devon Energy

    14

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    About Devon Energy

    One of North Americas leading independent producers of natural gas and oil

    Engaged in exploration and production

    Corporate headquarters in Oklahoma City

    More than 5,000 employees

    Member of the S&P 500

    On Fortune magazines 100 Best Companies to Work For list each year since 2008.

    15

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    Devon TodayDevons Core & Emerging Assets

    Core

    Emerging

    Heavy Oil

    Rockies Oil

    Mississippian-Woodford

    Barnett Shale

    Permian Basin

    Anadarko Basin

    Eagle Ford

    Q3 2014 net production: 640 MBOED(1)

    Deep inventory of oil opportunitiesTop-tier Eagle Ford developmentStrong Permian Basin positionWorld-class heavy oil projectsUpside potential in emerging plays

    Strong liquids-rich gas optionality

    EnLink ownership valued at $8 billionAdditional midstream value in Access

    and Victoria Express pipelines

    16

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    Business Driven Data

    17

    Executive Buy In

    Devon Enterprise US and Canada Drilling Completions Production Facilities Midstream

    Shape our data culture with tools Spotfire SAS Excel

    Rick HowellDevon Energy

    Supervisor Real-Time Data

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    Making Drilling Repeatable

    18

    WellCon Drilling Dashboard

    PI is source of real-time data

    Identify and characterize top performers

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    Spotfire, OFM and SAS

    19

    Spotfire Screenshot

    Reservoir characterization ESP performance Comparison across wells Performance metrics

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    New Tools

    20

    NetworkNOT a traditional Hadoop Top of Rack (ToR) Configuration. Solution leverages

    redundant collapsed core model that delivers 40 gigabit aggregate.

    Physical Infrastructure

    40ea HP DL380 Servers

    Each has

    16 CPU Cores 96 Gigs Ram 30 TB Local Storage

    (1.2 Petabytes Raw)

    10 Gigabit Ethernet

    Services Include MapReduce Data Nodes Hbase

    (Every Server)

    Apache Solr Search(4 Server Instances)

    Virtual Infrastructure

    3ea HP DL580 VMWare

    Host. Each has

    32 CPU Cores 512 Gigs Ram 6 TB Local Storage 10 Gigabit Ethernet

    Services Include Name Nodes

    (Primary & Backup)

    Hive Catalog PostGreSQL DB Zookeeper DataMeer Informatica Data

    Virtualization

    Other Ancillary Services

  • Page 21 Hortonworks Inc. 2011 2014. All Rights Reserved

    Become a Data Driven Organization with

    OSIsoft and Hadoop

    Kelly Kohlleffel Hortonworks Industry Executive13 November 2014 - OSIsoft Houston Regional Seminar

    Hortonworks. We do Hadoop.

  • Page 22 Hortonworks Inc. 2011 2014. All Rights Reserved

    Our Mission:Power your Modern Data Architecture

    with HDP and Enterprise Apache Hadoop

    Who we are

    June 2011: Original 24 architects, developers, operators of Hadoop from Yahoo!

    June 2014: An enterprise software company with 520+ Employees

    Key Partners

    Our model

    Innovate and deliver Apache Hadoop as a complete enterprise data platform

    completely in the open, backed by a world class support organization

    22

  • Page 23 Hortonworks Inc. 2011 2014. All Rights Reserved

    Hadoop in Oil and Gas

    Real Time Operations

    Join disparate sources of data together presenting real time and historical

    combinations of E&P data at each stage

    of the oil and gas production process.

    Production Optimization Production parameter optimization is

    intelligent management of the parameters that maximize a wells useful life, such as pressures, flow rates, and thermal characteristics of injected fluid mixtures.

    Seismic Analytics/Management

    Storing seismic data from multiple experiences permits learning in the

    aggregate across all of those

    experiences.

    LAS Predictive Analytics

    Leverage the shovel-ready nature of LAS files for predictive analytics across

    multiple datasets and the power of

    Hadoop for normalization,

    transformation and economical storage

    Other Preventative Maintenance

    Condition Monitoring

    Supply Chain and Manufacturing

    Asset Optimization

    Lease Bidding

    QHSE

    Enterprise Archive (Unstructured)

    Process unstructured data into an enterprise archive and blend search with

    machine-learning algorithms to discover

    value and automatically categorize the

    data for eDiscovery and other applications

    23

  • Page 24 Hortonworks Inc. 2011 2014. All Rights Reserved

    OSIsoft and Hadoop for Oil and GasS

    OU

    RC

    ES

    Sensor &

    Machine

    Logs

    SO

    UR

    CE

    S

    Unstructured Existing

    Systems

    Web &

    Social

    Geolocation

    Weather

    AN

    ALY

    ZE

    OPERATIONAL USER

    AN

    ALY

    ZE

    BUSINESS USER

    DATA SCIENTIST

    1

    Script

    Pig

    SQL

    Hive

    Java

    Scala

    Cascading

    Stream

    Storm

    Search

    Solr

    NoSQL

    HBase

    Accumulo

    HADOOP : HORTONWORKS DATA PLATFORM (HDP)

    COMPLEX DATASETS - ENTERPRISE ANALYTICS

    In-Memory

    Spark

    Others

    ISV

    Engines

    YARN: Data Operating System(Cluster Resource Management)

    HDFS (Hadoop Distributed File System)

    Tez Slider SliderTez Tez

    OSIsoft PI SYSTEM

    SYSTEM OF RECORD REAL TIME ANALYTICS

    ASSETS / EVENTS ASSET BASED ANALYTICS

    PI Data Archive

    Data Publications

    24

  • Page 25 Hortonworks Inc. 2011 2014. All Rights Reserved

    Hadoop & OSISoft : Enabling Data Driven Innovation

    Hadoop &

    OSIsoft

    Joint Value

    Explore Complete Datasets Leverage Hadoop as a landing pad for all emerging data types and data silos

    Empower Operational Users, Business Users, and Data Scientists

    Enable Data Agility Schema on Demand Shorten development cycles

    Test 5x 10x more hypotheses

    Shorten innovation cycles

    Apply to any size dataset

    Deliver Data Scale and Variety Economically Enable exploration of larger datasets

    Preprocess raw data

    Expose unlimited data variety on premise or in the cloud ($250/TB)

    Create New Value and Business Innovation Unlock net new business value within and across emerging data types

    25

  • Page 26 Hortonworks Inc. 2011 2014. All Rights Reserved

    ..allows a shift from reactive to proactive interactions

    Hadoop and OSIsoft

    allow organizations to

    shift interactions from

    ReactivePost Transaction

    ProactivePre Decision

    to Real-time PersonalizationFrom static branding

    to repair before breakFrom break then fix

    to Dynamic AutomationFrom manual process

    to Real Time AutomationFrom gut feel

    to Accelerated InterventionFrom speed

    constraints

    A shift in Production

    A shift in Drilling

    A shift in GeoScience

    A shift in Retail

    A shift in Refining

    26

  • Page 27 Hortonworks Inc. 2011 2014. All Rights Reserved

    Next Steps...

    Download the Hortonworks Sandbox

    Read the datasheet: Oil & Gas and Hadoophttp://hortonworks.com/blog/modern-oil-gas-architectures-built-hadoop/

    Engage the Hortonworks/OSIsoft Joint Account Teamsfor a Business Use Case Workshop

    27

  • Page 28 Hortonworks Inc. 2011 2014. All Rights Reserved

    Frank BeschNoble Energy

    Director of Business Integration

    Increasing Production with Data

  • Page 29 Hortonworks Inc. 2011 2014. All Rights Reserved

    Noble Energy Company Highlights

    DJ Basin

    Marcellus Shale

    Deepwater Gulf of Mexico

    West Africa

    Israel and Cyprus

    Falkland Islands

    Northeast Nevada

    Levant Basin

    29

  • Page 30 Hortonworks Inc. 2011 2014. All Rights Reserved

    The ChallengeAn Innovative Approach to Unconventional Resources

    Production Growth Drives Margin Growth

    Production growth fuels cash margin growth which drives long term cash flow

    Speed Is Essential

    Working with disparate, complex datasets under a traditional analysis model limits innovation and does not allow the

    speed required for unconventional plays

    Data Volume Continues to Grow

    A single well has billions of time series data points and other key related data sources such as the production systems,

    subsurface information, and field information (many times in

    unstructured format) make it highly challenging to provide a

    consolidated view for analysis

    30

  • Page 31 Hortonworks Inc. 2011 2014. All Rights Reserved

    Our ApproachData as a Strategic Asset

    Gain New Insights Into Production Dynamics

    Across a wide variety of disparate data sources and variables such as well information systems, SCADA data, sensors, and unstructured data

    Build Executive Consensus and Business Sponsorship

    Start with a single business unit prove the value

    Recognition that data provides sustainable competitive advantage

    Widespread, systemic value creation because data is managed as professionally as capital or labor

    Rely on Trusted Partners to Assist

    Combination of Noble team members along with Hortonworks (Hadoop) and OSIsoft (PI / CAST)

    High performing operational and data science team (Python)

    31

  • Page 32 Hortonworks Inc. 2011 2014. All Rights Reserved

    Results and Next StepsJourney to a Data Driven Organization

    Operational Value Realized

    Proactive approach to identifying events causing production downtime resulting in significant savings per day

    Advanced analytics allowed us to move beyond the spreadsheet

    On a Path to Strategic and Transformational

    Fostering a data driven culture while realizing value across multiple areas

    Delivering advanced and transformational analytics to each business function and business unit

    Building out organizational design and capabilities, best practices, and a COE

    Enhanced Models and Net New Analytics

    Continuing to add additional datasets to the model for even greater enriched analysis

    Addressing other areas within the company

    Predictive analytics operational across business processes

    32

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    Get Involved

    Demo Pods

    See Project CAST and Hadoop in Action

    E-mail us [email protected] (2015 CTP)

    Download the Hortonworks Sandbox

    33

  • Copyr i gh t 2014-15 OSIso f t , LLC.

    Brought to you by