Duke Energy EIM IMA Overview Charlie Ward Chief

Duke Energy EIM (IMA) Overview Charlie Ward Chief Architect - Smart Energy Systems

Summary – Why EIM? PRODUCTS & SERVICES OPERATIONS WORKFORCE ASSETS & FINANCIALS CUSTOMERS ONE COMPANY: Single version of the truth, shared management of information assets YESTERDAY • Inefficient access to company-wide information • Multiple companies and multiple functional information silos • Various definitions and rules for information • Limited oversight of information quality and usage • Uncontrolled data proliferation through “spreadmarts” TODAY AND INTO THE FUTURE • Information strategies & stewardship implemented • “One company” approach; functional silos broken for a process view • Highly leverage analytics for Business decisions • Acquisitions expedited • Cost and efficiency gains through improved information access and quality 2

BACKGROUND § In 2005 Duke Energy adopted a strategy of growth through acquisition § Target was 4 X current size over 7 years = 14 states, 12 - 14 M customers, 70, 000 employees § IT was commissioned to build a “scalable platform” to support future M&A § Standardized core systems § Managed business data and information § 1 system / 1 process § “Plug & Play” for future acquisition § 2005 – 2006 acquired/merged with Cinergy = 5 states, 4. 5 M customers, ~19000 employees § Designed and delivered the Information Management Architecture (IMA) for the scalable platform § Future M&A transactions would be executed as acquisitions – target company’s data converted and Duke Energy systems and processes adopted

EIM/IMA Benefits § Reduces Application Dependency § When applications reference each other directly, changing one means all who reference must change also. Hubs eliminate direct reference and an application may change without impacting others. Lower cost of application life cycles. § Provides Single Version of the Truth § When applications maintain their own version of enterprise data, inevitably, that data over time will differ from the other versions found in other applications. Greater confidence in the data. § Provides Integrated Reporting/Analytics § Reports can be constructed from the “single version of the truth” directly rather than replicating data from various sources into numerous report-specific copies. Fewer data and reporting errors. § Multiple systems performing the same function—whether region-specific customer information, or systems of newly acquired companies—contribute to singular, consolidated reporting. Lower cost to prepare consolidated reports. Total cost is lowered while information becomes more reliable

IMA 2006 Security Financial Systems Enterprise Application Integration Data Hubs M&A Materials Products & Services Messaging Customer System Smart Grid other Systems Catalog Layer Processes Work Mgmt Systems Extract, Transform & Load Reporting Standard Reports Finances Supply Chain Systems EAM Analytics Ad Hoc Reports Managed Reports Customer Dashboard Web Services etc. Infrastructure Business Processes Req to Pay, Order to cash, etc. Transactional Systems Integration Data Hubs Meta Data Analytics Reporting Applications and source data, Application Integration, Data models independent of transactional systems, What’s available, Standard Reports, Transaction processing, Data Move & Transform What’s certified, Collection of Cubes & Marts , Operational reporting Data Quality Data Move & Transform Analytic Capabilities Ad hoc capabilities, Dashboards

What is a Data Hub? § A Data Hub is a collection of data for a specific business subject area and is the authoritative source of data for application integration, and the authoritative source of data for reporting and analytics. § Each data hub is comprised of an operational data store (ODS) for system integration, and a data warehouse (DW) for reporting/business intelligence/analytics §The collection of subject area data warehouses in the IMA comprise Duke Energy’s enterprise data warehoues Based on this concept we manage information independent of the applications supporting business processes.

Duke Energy Enterprise Data Model T&D Facilities Customer Facilities Products & Services transport products to are de livered Customers are sold to to prod rk uce at Workforce n reve ue ed at maintained at co mp per fo rms Nuclear Facilities wo Materials ms uce maintain Duke Energy Facilities for per d pro wor k at en fund Equipment Finances sa te F/H Facilities X Y means X “is a type of” Y. Sub-types inherit relationships of the parent.

Business Strategy Changes § Late 2007 - Duke Energy scales back M&A plans to focus on streamlining operations to strengthen the balance sheet § Late 2008 - Strategy changes to emphasize revenue growth via smart grid, investment shifts to: § § § Green generation technologies – wind, solar, clean coal, CO 2 scrubbers, etc. . Distribution grid modernization AMI meter and system deployment/upgrade New customer products and services Energy management and conservation

New Business Strategy Drives Change § Evolving integration requirements § § § Future merger and acquisition (M&A) activities More external service providers and partners Data integration with exponentially larger volumes of data Cross functional data analytics Customer facing services § Reduce solution development, enhancement, and maintenance costs § Faster time to market for new or enhanced solutions § Decrease development time via standard, reusable *services § Shorten solution test cycles, particularly regression testing § Better visibility into system operational and performance metrics to preclude failure and speed problem troubleshooting and resolution § Business process and supporting system flexibility

Capabilities Business Drivers 1 § The smart energy leadership team defined a set of business expectations IT needed to address to support program initiatives. 2 3 4 § The Core Issues are the obstacles that need to be overcome. § The Architecture Capabilities are required specifically to: § § § Meet the Smart Energy and Smart Grid business needs Overcome the Core Issues Deliver and support the Business Drivers 5 6 7 8 9 Application Integration A B Data Integration Time to Market Cheaper Builds Lower O&M SG Analytics Biz Agility Biz Process Flexibility Reliable HA C D E Core Issues Architecture Capabilities Unproven scalability 12 and reliability 9 Data services Long project timelines 3 Difficult to coordinate projects 1 Inconsistent or inaccurate data 4 8 3 A G Application and A G integration services H Integrated test methodologies A Ability to integrate capabilities and requirements C Standards based architectures C Services governance B E 7 2 6 7 7 Scope and budget 8 force “get by” design B I D E B I D B G F Poor visibility into technologies & use 2 6 4 G Complexity of 5 point-to-point designs H I J Building “one offs”, poor reuse 1 Testing takes too long 3 No SG Business Context 3 7 8 8 4 9 H Master data governance D Consistent and measurable data quality D SLA definition and fulfillment A Business Architecture F H H I E F B E C H G J

The Enhanced IMA End to End Testing Tools Data Hubs SOA Financial Systems • Information Integrity Services • ETL Services • EII Services Standard Reports Facilities Master Data Management • • • Lifecycle Management Hierarchy & Relationship Mgt Data Quality Management Authoring Base Services MMD Interface Services Materials Equipment. Integration Middleware • • Enterprise Service Bus Queue Registry & Repository Quality of Service Customers Catalog Layer other Other Systems Finances Managed Master Data Registry Business Processes Customer System Reporting Information Integration Services Supply Chain Systems Work Mgmt Systems Analytics Ad Hoc Reports Managed Reports Dashboard Products & Services GIS Enterprise Application Integration HRMS ETL Messaging Web Services Workforce Management and Monitoring (initially Registry & Repository, later BAM) Security Infrastructure 11

Business Strategy Changes Again § 2011 – Duke Energy merges with Progress Energy to create the largest IOU in US. Announces it will probably execute another acquisition once Progress deal closes. Agility and Flexibility are keys to success. EIM should be able to support any business strategy with little to no change. Duke Energy’s IMA currently supports 3 major strategic initiatives… 12

(EIM) IMA supports integration of smart grid systems

Enhanced IMA supports a digital customer interaction strategy

Enhanced IMA supports M&A system integration

EIM Governance § IT Governance Committee commissioned a working team to : § § § Evolve a definition of data stewardship Perform the role on an on-going basis Define a structure for Duke Energy data Make policy recommendations Data Stewardship will evolve over time § A cross functional technical governance team was formed to ensure the proper technologies are deployed, maintained, and managed § An Integration Center of Excellence was created to develop and provide expertise on the technologies and processes of EIM

Successes To-Date § People. Soft Financials 8. 9 avoided interface costs § EAM able to design its interfaces to Finances and Customers data hubs rather than being dependent on projects in those areas. § Customers data hub allows a single interface in spite of their being four customer information systems. OLS (~$200 K /yr avoidance) , e. Bill, IVR (~$750 K/yr avoidance) all benefit through interface consolidation § Enterprise Customer System architecture - ~$3 M/2009 – 10 avoidance § Customer data hub allows consolidated customer business intelligence (retail marketing, call center agent operations). § Financial analysts spend less time on data verification and more time on data analysis due to improved data quality § Financial report run times greatly reduced. Some taking hours before now run in minutes. § Supply chain and work management report designers need not understand complex Maximo (e. Max) data structures, only the hub structures which are designed specifically for business intelligence and analytics

Challenges Remain Category Challenge Response Call to Action Change Management Pushback from projects: “it costs more” Building hubs is a lower cost approach, although projects that build them are often not the beneficiary. Optimizing costs in total across the Company and over time sometimes means incurring higher costs in certain areas. Promote the use of data hubs for data sharing: – Find out what major projects in respective areas are and are not using hubs – Demonstrate your support for utilization given the benefits – Ensure projects include resource scope for data hubs Managing data independent of applications The tendency is to think in terms of applications. But application are narrow in scope whereas hubs are intended to meet broad requirements across the Company. Build out and populate the hubs: – Sponsor projects to identify enterprise requirements (external to “owning” area) – Sponsor and sequence projects to close gaps left by application projects Prioritization and funding when dependency is on others to provision information Data hub owners serve the needs of Duke Energy, but those needing the data usually don’t control the resources that can populate hubs. Ensure areas supporting a data hub budget for continued build out: – Review with each hub owner their plans and budgets – Anticipate additional data needs Ramifications of an “enterprise data platform”

Data Hub Maturity Customer Facilities T&D Facilities are ork at Customers co en mp d te ntai n fun sa e om ec rk at F/H Facilities t produce revenue sw nb Nuclear Facilities s wo mai da are sold to Workforce ca orm d to ate m for maintained at Materials* stored at perf deli vere oc per uce Duke Energy Facilities el Products & Services transport products to d pro ar ed a t Equipment* Finances Opportunity Built thru 2010 * Materials and Equipment includes Nuclear

QUESTIONS?