Software Integration
Phil Flesch, manager, power team at SmartSignal Corp., also addressed the asset intelligence "islands" theme based on SmartSignal’s experience. SmartSignal provides the predictive analytics that power the 24/7 monitoring of over 18 GW of generation, and those systems have been implemented over the past six years. Flesch discussed how, together with partners OSIsoft and General Physics Corp., his firm is working to routinely integrate three of the asset intelligence islands: predictive analytics, thermal performance monitoring, and data infrastructure.
That integration is needed has been confirmed by data distilled from the Pearl Street surveys and interviews. Plant personnel are overwhelmed by data and alarms (see story p. 38). SmartSignal’s solution seeks to cut through the "data fog" to provide early detection of machine problems. In traditional condition-monitoring techniques, signals are monitored separately. SmartSignal’s software, in contrast, monitors all the signals simultaneously, correlates the data in real time, and provides alerts and warnings (Figure 3).
3. Early-warning system. Predictive analytics software monitors multiple signals and correlates them to provide early warning of problems and failures. Source: SmartSignal
An interesting point about digital intelligence developed through predictive analytics and confirmed by the many monitoring centers that employ SmartSignal is that early warning identifies machine-caused as well as human-caused problems. The agreed-upon ratio, from monitoring center managers, is that problems at a plant typically are equally distributed between human-caused and machinery-failure problems, which suggests an additional, interesting dimension to asset intelligence.
A new SmartSignal product, CycleWatch, will extend the company’s monitoring solution from "steady state" conditions to the dynamic start-up cycles for gas turbines (GT) and combined cycles. The rapidly changing conditions of GT-based plant start-up provide the earliest indications of failures, such as with sensors, combustors, nozzles, and valves. SmartSignal’s advanced (and patented) variable state similarity based modeling (VBM) technology is able to account for the chaos of start-up and point out just what is starting to deviate from expected behavior.
In addition, SmartSignal is working to close the gap among the different islands of analysis. Its software represents the front end of problem detection. Its customers also rely heavily on OSIsoft’s visualization tools, such as PI ProcessBook. Another new product, which was launched in 2009, SmartSignal xConnector, will replicate SmartSignal’s modeling information and incidents back into the PI environment from which the original data are read. This will allow customers to see the information within the graphical PI screens they prefer. Addressing the issue of intelligence islands eases pressures on power plant staff.
Functional Integration
Ross Convey, OSIsoft, showed how the application of his company’s plant-level data infrastructure software — so prevalent in power stations that it almost can be considered a de facto standard — is being expanded to an enterprise architecture platform "from fuel to meter." OSIsoft seeks to evolve into the common data infrastructure for generation, transmission, and distribution, and more recently, even advanced metering infrastructure. In recent years, OSIsoft has been supplying its solution suite to independent system operators around the U.S. and is demonstrating its software as a key smart grid component.
Many important power station applications from other software vendors, SmartSignal included, draw data continuously from OSIsoft’s PI system for individual plants and entire fleets. Many utilities today have taken a fleet-centric approach to managing their assets, facilitated by digital asset intelligence. Convey discussed how one leading utility, DTE Energy, uses OSIsoft PI to collect data and distribute actionable information through OSIsoft’s Process Book displays and a variety of third-party applications (Figure 4).
4. From data to actionable information. Performance excellence is driven through the organization via a hierarchy of knowledge management databases, software, and applications. (Diagram was simplified from original presentation version for clarity and general application.) Source: DTE Energy
Email
Comments
Print
Reuse
