It’s been a long time since many electric utilities have had to ask their rate commissions for the amounts of money they’re asking for today. States with deregulation programs either have frozen rates or reduced them over the last decade, in the hopes that competition would naturally lower prices to consumers. Now those programs are ending and their success is questionable. Utilities in more regulated states haven’t faced since the 1970s new build programs like the ones currently contemplated. And, everyone is paying far more for fuel today than they expected to only a few years ago, so fuel adjustment clauses or pass-through are also on the docket.
As a result, utilities are preparing new rate-case strategies, i.e., plans for how to communicate with the regulator and the public about why higher rates are necessary. One question often asked by regulators is “what are you doing to get the most out of the assets you have?” Testimony from recent hearings reveals the application of real-time performance monitoring and diagnostic (PM&D) technologies can bolster the case.
These new PM&D technologies provide plant operators and engineers one of the following: early detection and warning of impending problems to prevent unexpected component failures; continuous tracking and comparing of critical performance parameters, such as heat rate, so that operators can make adjustments to controllable process variables; collection and archiving of on-line data for tracking of long-term trends in component and system health; and validation of sensors and process measurements. These technologies transform reams of plant data into actionable information, so staff can focus on solving problems, not just looking for them.
A number of companies provide PM&D software products, which fall into several categories. The first type of software is the data historian. A data historian is a time-series database that interfaces with a plant’s distributed control system. The historian stores the time history of a plant’s instrumentation, and typically also provides a common infrastructure to access and view plant data across a plant or corporation. The second type of software analyzes thermal performance. This software analyzes the factors that affect the heat-rate performance of a plant against design conditions, helping to indentify areas of thermal loss. The third type of software is predictive-analytic software, which compares the current performance of a piece of equipment to its past performance and is used to detect the early signs of problems, so maintenance can be more proactive instead of reactive.
Several U.S. and Canadian electric utilities have included their use of PM&D software in their rate-case hearings and commission testimony. Reasons for including these software technologies in testimony are diverse, but mostly stem from today’s challenges such as recruiting and training a new generation of workers to deal with a loss of expertise as experienced workers retire; supporting rate increases at a time when capital costs for new capacity have skyrocketed; and ensuring that everything is being done to extract productivity from existing assets before requesting money for new build.
For example, in the case of Kansas City Power & Light, predictive analytics was referenced directly by F. Dana Crawford in his January 2007 testimony. The testimony involves a line of questioning related to “supply plan maintenance expense normalization.” To address the question, “What is KCPL doing to address performance improvements needed to maintain high levels of output from its existing generating assets?” Mr. Crawford responded, in part:
“For a number of years, KCP&L has utilized dedicated predictive maintenance teams at each plant site to gather data (vibration, oil sampling, thermography, sonic testing, etc.) to proactively look for early ‘warning’ signs of possible equipment failures. These efforts have been successful and are a key component of the PRO [plant reliability optimization] process.”
Crawford also said the company uses real-time predictive analytics and data-historian technology to support maintenance and troubleshooting activities.
Several other utilities are including, have included, or plan to include PM&D software in their rate-case strategies. However, rate cases are sensitive exchanges of information and, for various competitive and other reasons, utilities understandably aren’t always willing to disclose their evidence to a wider audience. Nevertheless, at a recent predictive analytics user group meeting, one utility executive observed, “One of the values we’re seeing in our analytics is we’re able to take it in front of our regulator. We’re proposing significant rates increases over the next couple of years to deal with our capital program, and by sharing this tool with the regulator we’re able to say this is what we’re doing to minimize those capital increases and really stretch the life of the assets that we’re operating.”
In a later session at the same meeting, the utility representative continued, “One of the tasks we face is being transparent in how we make our investment decisions and having to justify those in front of our regulator. [Predictive analytics] can be a valuable tool for us to do that because it allows us to start to show trends in the equipment and very easily come up with that information when we’re asked.”
A representative of a large investor owned utility added that they’ve “already presented [predictive analytics] to the regulators as being proactive, preventing cost and rate increases, and [got] a very favorable reception from the public service commissions.” At this utility, PM&D technology is a key component of centralized performance monitoring and diagnostics at afacility serving multiple sites.
In April of this year, Mark C. Birk, vice president of power operations at AmerenUE, testified on heat rate and efficiency testing to the Missouri Public Service Commission. “By monitoring heat rates, the company can track the efficiency of its units and address observed reductions in a unit’s efficiency appropriately,” he said. “This in turn allows the company to make efficient use of the fuel it buys by getting as much electric generation as it reasonably can from each unit of fuel burned. Performance monitoring systems allow AmerenUE to continuously track and record generator outputs, heat rates, and controllable parameters. Plant operators use this real-time performance information to continuously optimize the heat rates of the AmerenUE fossil units by making operational adjustments.”
In yet another example, PM&D technology is mentioned in BC Hydro’s 2007 Annual Report, along with smart grid, smart metering technology, and spatial asset management, as examples of new technology initiatives to improve system performance. “The advanced equipment monitoring and diagnostic system tracks changes in equipment condition to help maintenance engineers at our generating facilities optimize and plan preventive maintenance.”
BC Hydro monitors 14 of its 30 hydroelectric facilities using predictive analytic technology, although those units cover more than 90 percent of the utility’s hydro capacity. Gas turbine and other hydro assets also will be added, according to Mr. David Lebeter, director of generation operations.
Many owners and operators with large and geographically dispersed generating assets deploy real-time heat rate and predictive analytic software through centralized maintenance and diagnostic centers. One leading predictive analytic software product now is being used to avoid failures and improve reliability at hundreds of power stations, either at individual plants or through a centralized M&D center. Such software implementations have been progressing over the last decade or more.
Beyond bolstering rate cases, owners and operators also are reporting performance improvements by deploying these technologies. Entergy, for example, avoided a catastrophic failure of a 411- MW steam turbine-generator at its gas-fired Waterford station. Analytics software began to show anomalies in the operating data and vibration specialists were dispatched to investigate, ultimately prompting a unit shutdown. Subsequent inspections revealed a 1.5-inch deep crack in a generator rotor shaft. While repair of the unit was estimated at around $5 million, Entergy estimated it would have incurred an over $30 million financial hit if the unit failed catastrophically, plus the potential for loss of life and limb. The Waterford units are 33 years old.
In addition to enabling plant owners and operators to optimize heat rates and use exorbitantly priced fuel as wisely as possible, these technologies can increase uptime by avoiding failures and converting unplanned equipment events into predictive maintenance activities, and generally raise capacity factors by reducing forced outages. From the perspective of utilities reporting to their commissions, they are relatively low-cost approaches to reigning in operating costs and supporting necessary capital expenditures.
The nation’s flagship fossil, hydro, and nuclear generating fleets continue to age, and the reality is that new replacement capacity isn’t coming to the rescue anytime soon. The earliest a nuclear unit could be on line by all accounts is 2017. Most coal-fired projects have been postponed, delayed, or bottled up over global warming issues, as three major Wall Street investment firms have issued their “carbon principles,” stating that they will not finance new coal until carbon capture and sequestration is commercially available. That’s not expected until 2020. Some gas-fired assets are getting permitted and built, but at today’s natural gas prices, they’ll face difficulty competing against the aging, but long since paid off, existing fleet.
Thus, the industry is working to extract additional kilowatt-hours from these aging assets, by avoiding forced outages, shrinking planned outage schedules, and raising capacity factors. Software and digital technologies can help to extract this stealth capacity, while also ensuring the nation’s aging electricity system operates reliably. In time, power-generating utilities increasingly will rely on data from such technologies to support their positions in rate-case proceedings.