Where The Information Is

By John C. Kapron & Sumanth K. Makunur

A well-run business needs to know and control its process costs, optimize its performance, monitor its asset health and enhance its market value. This is a challenge for an electric utility like ours that must deal with assets and discrete technologies that span decades across multiple locations.

Detroit Edison is the largest operating subsidiary of DTE Energy. It generates and distributes electricity to 2.3 million customers in southeastern Michigan. Detroit Edison has an 11,080 megawatt system capacity and uses coal, nuclear fuel, natural gas and hydroelectric pumped storage to generate its electrical output.

In February 2004, the Fossil Generation Organization of DTE Energy’s Detroit Edison company was given the charge by senior management to be the “Best Process Control Company in the World” and bring the best value to it customers and share holders. Using existing technology assets, Fossil Generation developed a multi-layered strategy and developed a totally integrated solution across our 19-unit fossil fleet.

The Technology Framework

The framework illustrates the processing of data through technology. Development continues in each layer with horizontal and vertical integration to support ther objective of achieving performance excellence.

Our strategy displays actionable information that enables best decisions to be made with respect to operations, maintenance, budget and marketing strategies. This multi-layered approach incorporates technology solutions provided by industry leaders. Each process technology solution is totally integrated across our fleet with a focused objective. 

Process Implementation

Our technical solutions work in conjunction with our EPRI Plant Reliability Optimization business model and are coupled with our enterprise business systems (SAP/Maximo). SAP is our enterprise resource planning (ERP) system and Maximo is our work management system. All of these solutions are organized in a hierarchical system index (HSI) for reporting consistency.

All of these technologies are implemented consistently throughout our fossil fleet with the same information available to all sites, disciplines and organizations. Although there is still work to be completed, at this time each layer of the organization has at minimum a working pilot.

To optimize our fleet operations and its assets, a fleet-wide Performance Center in Ann Arbor, Mich., was put into 24x7 operation in 2005. The Performance Center analysts work closely with the power plants to help identify developing problems. The Performance Center analysts and the Merchant Operations Center analysts work side by side to make accurate capacity and cost assessment of our units when offering them to the Midwest Independent Transmission System Operator.

We started by defining a multi-layered strategy that would build on our present investments in technology and would also allow us to expand and drive performance excellence. This multi-layered integrated infrastructure would measure and optimize the performance of 19 fossil generating units across seven power plants. Our strategy requires us to use solutions from technology leaders and adherence to standards. The final goal is total integration among applications, the process and end users.

In implementing this strategy, it became evident that no single vendor was an industry technology leader across multiple layers. Our challenge was to select the best vendors, minimize the total number of vendors and install the solution fleet wide.

Historian Foundation Layer

DTE-Detroit Edison standardized on an ABB distributed control system (DCS) across our fleet, and this standardization is nearly complete. The DCS provides the source for a large set of real-time and historical process data.

OSIsoft’s PI data historians were installed at each plant to collect DCS data and also to collect data from our many non-DCS systems. The data historians are our primary data source of real-time and historical process data and provide a common user interface to access data from multiple systems.

PI is not only a valuable technology to store data, it is also an integral part of many process and business applications. PI also serves as an effective data communication conduit for many applications through its PI-to-PI interface. How does one get the most value out of 300,000 data points? By turning discrete data into actionable information. 

Engineering Applications Layer

This layer brings value to the discrete data with powerful applications.  These fleet-wide applications standardize analysis and reduce support costs. Applications include:

• Digital Fuel Tracking Systems (DFTS) developed by Engineering Consultants Group (ECG) were installed on major units with online fuel analyzers.  The Monroe Power Plant is a 3,100MW four-unit coal fired plant with a modern coal-blending system and is an example of a plant where all coal passes through an online analyzer. The coal characteristics of each sample are tracked by the DFTS through a series of cascaded belts that feed coal into one of 28 silos supplying fuel to four boilers. The DFTS lets plant operators and combustion engineers know the characteristics of the fuel entering the boiler.  Although designed as base-load units, the generation output is normally reduced during off-peak times and fuel blends are adjusted for economics.
• Included in the application suite developed by EGC is eNotification, which provides pager and email notification alerts for any user-selected and configured PI data points. But more than that, users can easily configure reports and trends that can then be emailed at a scheduled time or by a triggered event.
• Matrikon’s ProcessGuard collects all plant alarms from multiple sources into a common plant database. This application includes statistical tools for alarm analysis and management. User selectable subsets of these alarms are sent to the fleet-wide Performance Center.
• Scientech’s PMAX system prepares real-time thermal unit performance calculations  across our 19 units. Individual system and equipment performance data is continuously written to the PI historian.