by Jim Chappell, InStep Software
It’s a common goal: Utilities across the globe strive to deliver reliable electricity as safely and efficiently as possible. To meet and surpass those goals, every critical equipment asset from generation to delivery should operate at optimal levels.
Transformers should be able to efficiently change transmission voltages down to lower distribution voltages, and circuit breakers should interrupt fault currents. Unfortunately, optimal operation is not always the case. Equipment becomes degraded and aged, environmental factors take their toll and assets become damaged. To counteract these issues and achieve ideal operating conditions, utilities implement equipment maintenance programs. Traditionally, these maintenance plans have been largely reactive, correcting issues as they occur; however, the exponential and continued growth of big data is creating opportunities for utilities to strengthen their maintenance plans by incorporating advanced predictive technology.
No matter how advanced technology becomes, reactive maintenance never will be completely eliminated because unplanned situations that require immediate action will continue to happen. Reactive maintenance, however, should be limited with proactive maintenance strategies’ largely deterring the need for reactive measures.
Instead, power utilities should implement comprehensive maintenance plans with more proactive and efficient strategies. A comprehensive plan incorporates several maintenance approaches, including a combination of condition-based, preventative, predictive and reliability-centered measures.
Components of a Comprehensive Approach
Preventative maintenance largely is a calendar-based approach that calls for equipment to be serviced or replaced at predetermined intervals or periods of time. This could include replacing a circuit breaker based on a specified time interval or number of operations.
Conversely, a condition-based maintenance program focuses on the condition of equipment and how it is operating rather than on a length of time or predetermined schedule. This approach is highly reliant on operational data collected and transmitted by equipment sensors. For example, engineers monitor the temperature of a transformer to ensure it doesn’t step outside a certain range, which could indicate a loading issue.
Online predictive maintenance is a key component of a comprehensive strategy that involves using software technology for real-time monitoring of equipment health and comparing its current operational state to a model that defines normal or ideal operating conditions. Predictive analytics software uses advanced algorithms to detect subtle operational variances for each piece of equipment, which often warn of impending problems that might have gone unnoticed otherwise. Utilities can create automated alarm notifications and use the software to diagnose the source of equipment and system anomalies, in addition to prioritizing issues based on severity.
Reliability-centered maintenance is a data-intensive strategy that involves performing a failure mode, effects and criticality analysis (FMECA) for assets and then implementing maintenance strategies based on the results. Using this strategy, equipment maintenance is altered and prioritized by its importance to the overall health of the plant, grid or facility; however, the value to be achieved through reliability-centered maintenance can’t be fully realized without incorporating preventative, condition-based and predictive techniques.
Implementing a Predictive Plan
Although predictive asset analytics in the power industry has been heavily concentrated in generation, the technology is becoming increasingly more prominent in transmission and distribution. Initially within the power industry, predictive maintenance and analytics were used to identify patterns and learn modes of failure in cyclical-operating mechanical equipment.
Predictive technology later expanded to provide similar insight into transmission and distribution assets.
A predictive maintenance strategy is most beneficial with the implementation of proper online condition monitoring and analytics software. Typically, predictive analytics software analyzes information from an enterprise historian, ensuring all historical and real-time sensor data is included in the analysis and model building.
Specifically within power delivery, the technology can be used to monitor and interpret the behavior of most systems and assets including transformers, breakers, capacitor banks and other substation equipment and devices. This increases situational awareness of geographic areas on the grid, making it easier to identify how assets affect one another.
An example of applying a predicitive maintenance strategy in transmission and distribution is using predictive software for transformer monitoring through dissolved gas analysis, thermal analysis and insulation breakdown. For example, preventative maintenance techniques and predictive analytics software are used together to evaluate gas levels and diagnose problems before a fault occurs. Improving the reliability of the transformer is of utmost importance because it is an important component of the transmission grid.
Using Predictive Analytics to Improve T&D
The benefits of including a predictive maintenance plan in a comprehensive strategy are immediate and long-term. The obvious benefits include increased reliability and efficiency, but how are those achieved?
Using predictive maintenance tactics, power delivery utilities can make smarter decisions about when and where maintenance should be performed. Utilities can reduce maintenance costs with better planning and gain the insight needed to delay maintenance that might not be immediately necessary. Instead, some suggested maintenance windows can be extended to a more convenient and cost-effective time.
Predictive analytics also can identify underperforming assets and help personnel understand what factors contribute to abnormal operation. In the same manner, predictive analytics technology can prevent equipment failures by providing early warning of subtle changes that otherwise might have gone unnoticed. The technology can identify problems days, weeks and months before a failure, which allows utilities to be more proactive. Not only do utilities reduce expenses by extending the life of their equipment, lengthening maintenance windows, increasing asset efficiency and increasing availability, other savings are realized when considering the costs that could have been, including loss of power, replacement equipment, lost productivity, additional man-hours, etc., when a major failure is avoided.
Implementing a software suite that uses predictive analytics enhances the ability of wide-area monitoring systems (WAMs) to monitor and analyze power use over large geographic areas. Predictive analytics also can be combined with WAMS to better monitor and analyze the power transmission system over large geographic areas for improved situational awareness. Through subsecond data monitoring, smart grid software can prevent the grid from collapsing by detecting small oscillations that can serve as early warning signs. By employing analytics to interpret data such as grid oscillations, utilities can launch proactive measures to prevent problems such as a grid collapse. The integration of real-time asset health information with WAMS can provide the additional insight necessary to identify risk points and analyze grid disturbances.
One of our large power delivery customers at InStep Software used predictive analytic software called PRiSM to identify a subtle equipment variation that would have gone unnoticed for some time.
Company personnel discovered that when a capacitor bank was energized, the neutral current was abnormally high. This condition did not trip the real-time monitoring alarm. Rather than waiting until more capacitors failed, which could have caused more problems, the utility received an early warning of the issue by using advance pattern recognition technology. This type of find or early warning is common when a predictive strategy is in place and spans from simple asset malfunctions to environmental effects to wide-scale efficiency loss on the grid.
Making Sense of Big Data
As new technology develops in the power sector, the use of predictive maintenance strategies has migrated from power generation plants to equipment assets deployed as part of the smart grid.
In particular, the use of predictive analytics has become prevalent in designing and monitoring infrastructure for transmitting and distributing electricity. After making significant investments in modern control, monitoring and smart devices, predictive monitoring techniques allow utilities to extend that investment by using and analyzing collected data to make more informed maintenance decisions.
Using advanced predictive analytics and diagnostic technology as part of a comprehensive maintenance program, utilities can monitor critical assets to predict, diagnose and prioritize impending equipment problems continuously and in real time.
As many utilities struggle to make sense of the massive amount of data available through smart devices, smart grids and machine sensors, predictive maintenance remains a practical application. Power delivery companies can transform their maintenance strategies by leveraging data and predictive asset analytics solutions to spend less time looking for potential issues and more time taking actions to get the most out of every asset.
Jim Chappell is an executive vice president at InStep Software, a leading global provider of real-time performance management and predictive asset analytics software products and solutions. He can be reached at email@example.com.
PowerGrid International Articles Archives
View Power Generation Articles on PennEnergy.com