Knowledge may be power, but it can also take the form of speed.
Implementing any improvement process is a journey, for example, so it can be very useful to find someone that can help you along the path. In the case of Meridian Energy, this meant using Matrikon's tools to better access its own information, and accelerate its trip to beneficial asset management.
New Zealand's largest electricity generator, Meridian, has eight power stations with an installed capacity of 2,460 Megawatts, which it uses to generate approximately 30% of the 12,400 Gigawatt hours that the nation consumes per year. However, the utility recently found itself in a position familiar to many industries"”having to monitor and maintain dispersed and increasingly fragile plants ever more closely with fewer and less experienced staff.
Consequently, beginning in 2000, Meridian implemented a plant asset management (PAM) system to care for its assets, and manage the health of those assets throughout their lifecycles. PAM integrates sophisticated data gathering, plant modeling, analysis and presentation software with legacy systems to keep the lights on for Meridian's 200,000 energy customers.
Garth Dibley, B.E., M.B.A., Meridian's asset manager, reported May 8 at Matrikon Summit 2007 that the company's asset management plan includes initiation and delivery of major rehabilitation projects that will keep plants operating effectively and efficiently in the future. A long term focus on preventive maintenance has seen the recent development of a PAM expert system that predicts mean time to failure for critical assets.
To develop its PAM plan, Dibley explains that Meridian first listed its main asset management aims. These included maintaining asset capability and reliable revenue capacity, enhancing existing asset capacity, extracting latent value, developing efficiencies from existing assets and processes, optimizing its business processes, and creating a unified "One Team" of staffers to accomplish all these goals.
Before starting on its new asset management journey, Dibley reports that Merdian's asset management efforts were mostly time based and involved little predictive maintenance. After redefining its asset management aims in 1999 and 2005, Meridian began to focus on reliability engineering methods using block diagrams, RCIM defect management, and performance monitoring scorecards, and eventually drafted Matrikon into its initiaive. "The most important aspect of all of this is doing asset performance in a feedback loop," says Dibley.Between 1999 and 2006, Dibley reports that Meridian's performance improved dramatically. "We went from 204 forced outages per year to 80-90, and this represented about $1 million in savings," he says. "We also achieved a 30% savings in our unit cost of production, and developed a long-term asset management plan. This is important because the costs of compliance, consultants, and original resources are always increasing, while skills and competent resources are decreasing."
Despite the improvements it gained, Dibley adds that Meridian's journey wasn't always easy. "We found collected data was still questioned, timely data wasn't always appropriate, and that plant data analysis wasn't always consistent," he says. "We found that we still had to optimize the effectiveness of our maintenance. If you've got a big question mark here, then you're running blind. This is why PAM looks at what your upstream processes are doing, and feeds that information forward. The objective is to have a predictive maintenance tool, enable plant condition management, detect any impending faults early, and be proactive based on the health of you assets.
"PAM turns masses of data into useful information that can be acted upon."
To monitor its transformers, turbines, governors and generators, Meridian's PAM generally standardized data entry, prioritizes diagnostics to deliver consistent and objective results, and uses visualization and notification functions to also display those results consistently. The utility's PAM plan developed asset-based software with help from Matrikon's models, condition monitoring, web-based dashboard, and then integrated these overview screens that can drill down from plant to individual unit/transformer overviews. This allows user to check maintenance and alert summaries, and generate specific work orders.
"You can then go all the way down to power factor predictions, and interpolate this information to predict days to failure for that unit," says Dibley. "PAM and Matrikon enabled us to implement our asset performance feedback and achieve a single version of the truth. They really speed us up, and accelerated our journey to asset management excellence."