Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

EP Info Data Mgement 3-4 Feb 2015

582 views

Published on

  • Be the first to comment

  • Be the first to like this

EP Info Data Mgement 3-4 Feb 2015

  1. 1. Analytics Predict Profits – Is business leadership digitally aware ? Andrew Moore Exploration Data Systems Consultant
  2. 2. © Andrew Moore 2015 Outline  Reduced margins globally and especially in Australia are driving cost cutting  The innovation response is limited to the engineering comfort zone  Advanced analytics, especially in Exploration, can both save and make money  How to persuade a reluctant management to apply 21st Century thinking  Visualisation can help get the message across and raise thinking horizons
  3. 3. $200Bn is being invested in Australian projects to find and produce “clean, safe energy” But cost over-runs and delays have deterred further investment. Meanwhile the oil price has crashed and current margins are uneconomic. Sophisticated financial analytics are directing investments and raising profitability in other industries. So why is the upstream O&G business slow to adopt E&P data analytics? & How can we turn this around?© Andrew Moore 2015
  4. 4. Costs have crippled the oil and gas business, especially in Australia  An Aug 2014 report from Ernst & Young showed that, on average, 64% of O&G “megaprojects” had exceeded budgets and 73% missed deadlines.  On average, cost forecasts for 205 projects surveyed were 59% above initial estimates, that’s US$500Bn over an initial cost estimate of US$1.2t  Chevron estimate that Australian costs are 40% higher than US © Andrew Moore 2015
  5. 5. © Reelwell AS What techniques is the industry using to reduce these huge costs and delays ?  Improved planning, and cutting budgets  Pad drilling & SIMOPS for onshore field development, un-manned off-shore platforms  Extended horizontal drilling and geosteering  Improved stimulation and recovery for tight gas These are mainly engineering responses, they do not exploit insights gained from operational data. Industry leaders proclaim ours is a technology industry, but most E&P engineers think of “technology” as hardware. Source: Ernst & Young © Andrew Moore 2015
  6. 6. Where can information and data analytics have the most impact ?  Process improvements from better IM & KM  Reducing unplanned downtime of rotating equipment - “Traditional” predictive analytics  Exploiting new data relationships as well as the obvious E.g. frac design, microseismic, fracture propagation and resulting production  Real-time drilling analytics and synthetic logs in HPHT zones – could save $Ms per well. © Sekal AS The prize ? In 2012, 1GJ of contracted sales gas was worth around A$4*. A 2% improvement in Santos Cooper Basin production of 66PJ would have delivered ~$5M p.a. * Source Credit Suisse E. Australia Gas Prices 04/14, Santos annual report 2012 © 2014 MicroSeismic, Inc. © Andrew Moore 2015
  7. 7. What is data ?: Even with no margins, this makes no material difference. Santos was investing over $400M/yr in the Cooper*. Does analytics in Exploration offer more real value ? So, if oilfield analytics can reduce costs and, more importantly, increase revenue: Why are we slow to adopt and innovate?  The focus is on improved cycle times or reduced down time based on existing disciplines, processes and technology – 20th Century thinking, closed mind-set. Let’s look at the big picture - $5M ?  Exploration data is already Big, but adoption of analytics within the established scientific and engineering disciplines is slow. Perversely, science is holding back the data scientist. What is required to bridge the culture gap between the data scientist and the geoscientist ? :  Mutual comprehension, foresight, and the ability to persuade leadership. * Source: Santos Cooper Gas Growth Program © Andrew Moore 2015
  8. 8. Oilfield Analytics can be learned. But do we have the Foresight ? – Is the promise of data-driven E&P compromised by a 20th century view of its potential ? Let’s look at some analogues :-  "I bought a JEEP” – brilliant ad – but the futile act of a dying breed – 20th Century thinking  Tesla – 21st Century – A sexy solution to win hearts and open minds.  Outlander PHEV – The perfect everyday car?  Also BMW i8, i3, Nissan Leaf etc. etc. etc.  Spend some time to think just a little outside the box. © Andrew Moore 2015
  9. 9. © Andrew Moore 2015 Find (or be) a champion with foresight, exchange knowledge, make it stick  Through stories – win hearts and open minds  Through reason – demonstrate in the real world – why wouldn’t you use it?  Through example – find a “killer app” and exploit the app delivery culture ! Ability to Persuade How can “IT” convince a sceptical subsurface leadership that really using information as an asset and relying on predictive data models can revolutionise our industry as it has for banking, retail and manufacturing ?
  10. 10. A Good Story Repsol started the Kaleidoscope Project in 2007. The aim: compete with and out perform the majors in deep water Gulf of Mexico plays.  They partnered with IBM to build a high performance computing platform  They partnered with Stanford University to develop new seismic reverse time migration algorithms.  They processed seismic 6 times faster and improved imaging beneath salt domes, raising exploration success for Repsol GOM JV projects from the industry norm of 20% to 50%. http://www.repsol.com/es_en/corporacion/prensa/galeriamultimedia/transcripcion_video_francisco_ortigosa.aspx http://www-07.ibm.com/innovation/au/shapingourfuture/downloads/repsol_case_study_2010.pdf “We realized that for a project like Kaleidoscope, which was aiming for a clear shift in our exploration model, we needed out-of-the-box thinking in every dimension.” Francisco Ortigosa © Andrew Moore 2015
  11. 11. What is data ?: The basis of reasoning ? This innovative thinking lead to outstanding success in Repsol deep water projects in GOM (Buckskin 2009), Brazil and West Africa with various partners.  By 2013 Repsol was able to report: “The proven reserve replacement ratio was 275%, one of the highest in the industry worldwide and Repsol's all-time high. During 2013, Repsol continued its track record of success, with nine finds in Brazil, Alaska, Algeria, Russia, Colombia and Libya.” The world’s highest reserve replacement ratio in 2013 Source: Repsol Annual Report 2013 And now Repsol has bought Talisman. Big thinking at board level has transformed Exploration – and the company – by changing the mind-set. Result :– © Andrew Moore 2015
  12. 12. Necessity is the mother of invention - and NOW is the time !  A combination of solvents and microwaves “melt” the bitumen in oil sands. Tests suggest ESEIEH could slash SAGD energy costs by 80%.  If there was ever a time for innovation it is now.  Alberta companies are testing enhanced solvent extraction incorporating electromagnetic heating (ESEIEH) - an example of innovative thinking (now 5 years old) but more important than ever at $50 oil.. * Source: Calgary Globe and Mail Jun 2012 © Andrew Moore 2015
  13. 13.  Definitions of Data – Google: “About 263,000,000 results (0.30 seconds)” “the quantities, characters, or symbols on which operations are performed by computer” ? “Things known or assumed as facts, making the basis of reasoning or calculation” ? “Factual information, especially information organized for analysis or used to reason or make decisions.” ?  Exploration data has already cost $billions, but it is only valuable if used to make decisions with a commercial outcome. Could our decisions be better ?  Think of data management and analytics as decision support and then ask:  Why does Exploration data sit idle whilst drilling data is left with the contractor ? Why do old trends and new unexploited data relationships not inform decisions ?  “The data is talking to us but we are not listening !” - Why not ? Is this is the biggest waste of money in history ? Data: The Basis of Reasoning © Andrew Moore 2015
  14. 14.  The industry is cutting costs, but is this the right time to cut IT ?  We can see, and learn about, how analytics can both save money and make money.  We should be investing in monetizing our data, not reducing our innovative capability  Google “Analytics in Oil and Gas Upstream”: About 372,000 results – this isn’t vapourware!  Reason this: IT solutions are repeatable for a fraction of the initial cost.  I said $5M p.a. was not “material”. But that’s just for Santos in the Cooper Basin. If these techniques were applied to all fields we could easily be talking $50M p.a.  1 saved stuck pipe and 1 HPHT incident avoided could easily add another $50M p.a..  $100M p.a. is very material. This could turn IT into a profit centre.  Reason this: Google “Analytics in Accountancy”: About 922,000 results !  Accounts should be correct, yes ? So if data = money - Why not databases ? Focus on the business, “monetize” the data © Andrew Moore 2015
  15. 15.  Traditionally, data management has been a service assisting “the business”  DM is not recognised as a profession or given credence, but it is now critical.  Today, the service customer makes all the decisions – DM has little authority.  Meanwhile the Digital Oilfield is driving more and more new data delivery  Some form of data scientist must now exist to filter, interpret and analyse – gaining credence & trust, influencing decisions. This role is critical and requires formal recognition.  Is this a problem ? What is a geophysicist anyway, if not a data scientist ?  Geoscience must connect to data science and adapt to new data processes  This means collaborative workflows and standards are imperative to integrate data, and  Data workers must adopt new techniques and automate to cope with data volumes Support the Evolution of a New Breed © Andrew Moore 2015
  16. 16. Traditionally, the management of (mainly static) data has been a service. Today, the service customer (decision maker) is always right – Education is required. DM is rarely recognised as a profession or given credence, its importance is down-played. Meanwhile the “Digital Oilfield” is driving more and more new data delivery Some form of data scientist is required to filter, interpret and analyse – gaining credence & trust, influencing decisions. This role is critical and requires formal recognition. Is this a problem ? What is a Geophysicist anyway, if not a data scientist ? Data science must constantly adapt to monitor new data streams. This means collaborative workflows and standards are imperative to integrate data And data workers must adopt new techniques and automate to cope with data volumes  We don’t have enough eye balls ! How long before we evolve ? Automate or grow more eyes © Andrew Moore 2015
  17. 17. A Science ?!Can’t wait that long ? – Get Collaborating ! © Andrew Moore 2015
  18. 18.  Find the right sponsor and establish an agile development project  Look for a seasoned risk taker who can balance the investment with potential benefit  Engage with up-and-coming professionals recently trained in probability theory  Look for “Explorationists” who can see beyond their own silo, who support collaboration  Recruit good statisticians, preferably from within the company or industry  Deep understanding of stochastic methodologies, i.e. a masters in statistics, is required  But more so, industry knowledge, to bridge the gap with geoscientists and engineers.  Your database platform and data integration may be an issue  In-memory database platforms are better suited to analytics, is your platform suitable ?  Look for discrete data-sets to assemble into a pilot analytics data mart. e.g. seismic attributes to identify HPHT zones not visible in seismic sections A Science ?!Can’t wait that long ? – Get Collaborating ! © Andrew Moore 2015
  19. 19. There’s an app for that … © Andrew Moore 2015
  20. 20. There’s an app for that …  Find a simple data relationship with obvious impact and create demand  Safety improvements are always supportable – cite the Shell example post Macondo  Don’t ignore mobile field-based technology – LDAR for example, and visualise in the office  Exploit business intelligence tools like Spotfire to visualise data relationships  Target specific user groups, control access through conditions of use  Create demand through word of mouth, support utilisation, plan for it to go viral ! Examples:  Self Organising Maps for multivariate analyses – e.g. reservoir characterisation  Mapping seismic attributes to reservoir properties e.g. wide / multi azimuth & fractures  Estimated Ultimate Recovery – Scenarios with well spacing, type curves & fluid dynamics © Andrew Moore 2015
  21. 21. Visualisation is key to innovative thinking  In the Quality vs. Quantity debate, both are right  Visualisation is critical to expose both the correctness and completeness of data  Any lack of quality or quantity may be embarrassing but will drive rapid improvements  Improved access to data of a known quality – good or bad – informs better decisions  Revealing “expert” data to “non-experts” encourages new insights  Visualisation is a keystone of discipline integration, collaboration and innovation  The juxtaposition of “expert” data from different disciplines encourages new thinking  Expect entrenched views to present impediments to progress The following slides are courtesy of Marathon and iStore, whose PetroTrek portal is being released in Santos after many political hurdles were crossed © Andrew Moore 2015
  22. 22.  Connection to multiple data stores  Great QC / QA tool to raise quality  No additional database  Quick + easy access to data
  23. 23.  Focus on Safety to sell your case
  24. 24.  Within a few years a geoscientist with data training will be promoted to a position where this can happen – or maybe an accountant will !  Until then, the old crew will ignore the evidence, saying IT costs the Earth. 21st Century companies know that IT is 2.5 times cheaper, in $50 boe terms, than it was in 1990* plus it has the potential to significantly enhance ROI.  Or possibly, the incumbent software providers will get there without you.  More likely, major financial systems providers (IBM, Oracle, SAP), currently spending $Ms to enter the market, will recoup their costs via your CEO.  Explorationists are smart risk takers. Sooner or later someone will risk it and use analytics to manage data volumes and reduce uncertainty. ”Professional curiosity will become an industry imperative” Keith Holdaway Or Wait for The Big Crew Change * Source Paradigm::-1990 industry IT cost $0.25 per a boe @ $20. Today it’s still $0.25 @ $50 boe. © Andrew Moore 2015
  25. 25. © Andy Moore, Exploration Data Systems Consultant andrew.moore@dataco.co.uk “Big Data” is a 21st Century issue. 21st Century thinking and volition is required to apply new scientific methods to realise orders of magnitude more benefit. Thank You Smart E&P requires smart thinking

×