World Wide Chemicals & Petroleum IBM - Fueling the Oil & Gas Industry Big Data in Oil & Gas - Improved Decision Making & Operational Efficiency Ole Evensen, IBM Chemical & Petroleum WW Upstream Business Development Executive Ole.Evensen@no.ibm.com
Oil & Gas Centers of Excellence - Addressing Industry Challenges People & Competence Industry Challenges CoE Enabling Technology 1990-2000 HW 2000-2010 2010 Analytics Cognitive 2
Big Data in E&P: Believe it Build for use Benefit from it Addressing industry challenges: Big Data in E&P Believe it! - Value chain of data sources - Growing Build for use Today and Tomorrow - Capture Manage Use Benefit from E&P Big Data - UC1 Cognitive Exploration Advisor - UC2 Drilling efficiency example - UC3 Operational Efficiency (Leading industry example) 3
Big Data in E&P Big Data in E&P Believe it!
IDC MarketScape and FutureScape - Oil and Gas
Some data about (Big) Data 80% of data is unstructured (logs, maintenance, production data, reports and studies) Variety Data in many forms Velocity Batch and streaming Up to 2 Tb of data produced from one rig / per day $150 million cost of incorrectly locating a new oil rig Data in doubt Veracity Big Data Terabytes and zettabytes of data Volume Up to 40,000 sensors per oil rig Cognitive? Analytical? 6
Low High Level of Insight Are we getting insight from data Or headache? How do we find, extract and contextualize data to Information Assumption? Modest volumes of data, more flexible time windows of decisions More realistic? Data types today 80% Unstructured It not managed? 20% Structured Low Volume of data (Increase / Timeline High
Big Data in E&P Big Data And an analytical evolution
Our ability to utilize data an evolution towards Cognitive Evolution towards Cognitive How do we optimize a dynamic, Big Data environment? Cognitive (Ask, Discover, Decide) E.g. Equipment & Design Advisor What should we do about it? Prescriptive (Direct) Proactive pump maintenance, to minimize downtime. 1 Understands natural language and human style communication 2 Generates and evaluates evidencebased hypothesis What will happen? Predictive Early indication of pump failure. What happened? Descriptive (Analyse) Pressure, temperature increase alerts on pumps. 3 Adapts and learns from training, interaction, and outcomes Menu
How to Support decisions - illustrated Different decisions require different support Different types of data require different tools... To secure Insight from available data... Request To Drill Cognitive Decisions: Do the right things Efficiency: Do the things right Assess Desirability Probability Hypotheses Analogies Insight Options Dialogue Decisions Infer Analytics Monitor Measure Compare Analyse Predict Prescribe Improve Insight Optimise Prevent Well Delivery: Well Project Initiation Secure Access to Rigs Develop project Plan Select Well Concept Conduct Detailed Design Pre Execution Activities Drilling Execution Well Close Out Menu
Introducing: Watson Cognitive Technology Transforming Oil & Gas Menu
12 2012 IBM Corporation
13 2012 IBM Corporation
The E&P Big Data and Analytics Value Chain UC2 UC1 UC3 Interpretation Analytics and Decision Support Explore & Appraise Analysis, cross discipline decisions Operations X-functional Planning and Assurance Well Productivity Securing the subsurface potential Seismic Big Data, HPC and Applied Analytics Drilling Data Mining, Analytics and Real Time Predictions Development Emerging analytics and Predictive (Time, Cost) Reservoir Analysis to determine optimal drainage Life of Field Planning Maximize Subsurface potential and secure Facility capacity Key messages: E&P functions depend on each other Interfaces are frequent fail-points! Interfaces depends on shared information Best-of-breed to be balanced against shared systems Technology must be matched by people and processes to realize Big insight from Big data
New Section Exploration/BD Cognitive Decision Advisors Menu
Repsol Cognitive Next wave is coming Menu
Identify: new exploration areas - Country, Region, Basin, Field considering the hydrocarbon potential, cost and risks. Evidence driven enabling comparative ranking Identify Explore Develop Operate Extend Decomission Assess Exploration Risk? Assess Country Risk? Business Development What exploration results are available or published for Uganda? Which services companies have been associated with exploration drilling in Uganda? Known geological characteristics? Do we have information of similar geological basins or fields? Do we have own operated fields of similar characteristics? Who is the exploration and appraisal SME for onshore exploration in environmentally exposed areas? How is oil, gas or mineral sector governed? How is the license (EPSA) regime governed, developed and perceived by stakeholders? Is the petroleum taxation clear / predictable? Any boarder disputes in Uganda, relevant for natural resource extraction? Any conflicts related to HSSE, indigenous citizens or NGO s? Key currency risk or regulations for Uganda? Assess Commercial Risk? Which Oil & Gas companies are currently associated with Uganda? Any conflicts or disputes between E&P companies and government? Do we have partnerships or network with incumbent companies? How is transparency of relationships and ownerships of suppliers, investors and Public Interest Companies. ( E.g.: government / employees ) What License or Asset transactions are known or can be identified? Any analogous fields in our portfolio? What are the refining and export facilities and capacity in Uganda? Data sources: Internal DB Internal EDRMS Internal Fileshare Private Studies External Subscribe External IHS External National External WWW Social Tweets
Pratt & Whitney broadens performance monitoring capabilities More than 4,000 operational commercial engines IBM will leverage world-class, military diagnostic, prognostic and health management capabilities to enable proactive and automated logistics to the rapidly expanding commercial fleet. This will provide Pratt & Whitney customers with - longer time on-wing, complement current asset maintenance alerts and - better insight into flight operational data. Enable accurate and proactive monitoring the health of customers' engines visibility to plan ahead for optimized fleet operations - reducing customers' costs." http://www.pw.utc.com/press/story/20140717-0300/2014/all%20categories#sthash.sjembrja.dpuf An Aircraft engine can generate up to a ½ TB data per flight. Coupled with predictive analytics (structured and un-structured data streams) early warning / fault detection Improve visibility of overall health of aircraft engines IBM Confidential Menu
E&P... Big Data Changing the game? Identify Explore Develop Operate Extend Decomission 1994 DISKOS Vision Geoscientists used more time to find/retrieve data than to analyze Company should compete based on understanding and interpretation of data, not access to it. 2015 Big Data Reality Companies unable to manage (acquire, clean, contextualize and share) data, are unable to utilize the potential of analytics Decision support democratizes insight Inhibitors: Technology People Process? Menu
Ole Evensen IBM Norway Business Development Executive Forusbeen 10 Upstream 4033 Stavanger WW Chemical & Petroleum Mobile: +47 9944 9010 ole.evensen@no.ibm.com no.linkedin.com/in/evensen Menu