by Alyssa Farrell, SAS
Utilities are under pressure to optimize asset replacement funds, improve operating efficiencies and meet regulatory requirements for adopting smart grid technologies. Grid reliability also is threatened by forces of nature outside the utility’s control (Figure 1) and more distributed energy resources. Large capital investments are needed, but rate increases are difficult to secure. Managing assets in the usual reactive fashion will not allow utilities to meet growing business demands.
Reliability Paradigm CHANGES
Much has been written about the convergence of information technology (IT) and operations technology (OT) in the utility industry. Few areas of the business have as much to gain through this intersection as reliability and maintenance. By analyzing asset data with prevention in mind, utilities can re-engineer processes to do more with less.
For example, a U.S.-based transmission and distribution company wanted to use analytics to improve reliability and operational performance for its 2.5 million metered customers. It had expended significant resources for transformer repairs and wondered why failures occurred and how to improve. The transformers were not equiped with intelligent sensors, but the company had secondary data from which to build predictive models, including load profile, physical characteristics, associated meter and weather data, as well as geospatial locations and outage history. Even with little data, the company believed it could build a predictive model based on data related to dumb transformers.
Smart grid analytics allow the company to identify critical factors driving transformer failures. These models enable proactive prediction of pending transformer failures, allowing the utility to streamline planning and coordination required for transformer placement and maintenance.
In the process, the utility can:
- Improve methods to anticipate transformer failures;
- Enhance regular equipment maintenance schedules;
- Reduce repair truck rolls for unscheduled maintenance except during unplanned incidents.
Turning Data into Insight
Information available to asset managers is ripe for analytics that turn data into insights. If IT and OT continue to merge, utilities soon could see social media, predictive analytics and enhanced data visualization impact reliability.
Reliability professionals must build partnerships with IT to successfully deploy and maintain these applications and the data that support them.
Griping, boasting, informing; tweets of all kinds abound on Twitter. Twitter is the social media application that allows users to share snippets of text. The text of the tweet typically contains a statement of opinion or fact and can also link to news stories, photos or other tweets. Users search for and follow other tweets by tagging the comments using a hashtag (#).
For example, ComEdison has the hashtag #ComEd. A recent survey of tweets (Figure 2) showed relatively high activity during a recent power outage. All Twitter information, Facebook pages and LinkedIn groups are public.
|2 Twitter.com #ComEd|
Transient social media content can be a valuable information asset if the utility can continuously monitor, capture and integrate online and social conversation data. For reliability professionals, having tools to identify important topics and content categories, and determine their relevance to customers and operations, can help pinpoint areas that need attention.
In the future, customers might turn to social media outlets as a way to vocalize community needs to the utility. Analytics brings insight to the rapidly changing world of social media by interpreting sentiment and data mining the unstructured text for clues about ongoing operations.
The use of basic monitoring consoles provided by asset manufacturers are limited in scope and provide isolated views that are prevalent in capital-intensive industries. The processes are driven mostly by domain knowledge and biased judgment, and are resource-intensive, time consuming and create many false alerts. This is reactive maintenance, which means fixing a problem as quickly as possible after it occurrs.
Utilities are investing in sensor technologies that add a layer of intelligence to networked assets. In addition to managing big data, utilities need to move from a reactive or condition-monitoring mode to a predictive planning mode, especially for critical assets. Addressing service needs as they occur rather than proactively anticipating them costs more in the long run–both in terms of personnel, equipment and customer satisfaction.
Using predictive analytics to alert a utility to upcoming maintenance needs can improve asset performance, optimize outages and reduce overall maintenance costs. This approach also results in:
- Reduced capacity losses due to unexpected shutdowns.
- Avoided health, safety or environmental violations and resulting financial penalties.
- Reduced prolonged outages, which affect the utility’s reputation.
- Increased ability to meet service level agreements.
- Reduced workforce overtime.
To achieve these results, utility reliability professionals must make the best use of data integration, automation, analysis and predictive analytics to boost uptime, performance and productivity while lowering maintenance costs and the risk of revenue loss. Capital-intensive companies, including utilities, have used software to enable predictive and preventive maintenance of their assets with minimal disruption to production (Figure 3).
|3 SAS Predictive Asset Maintenance|
Enhanced data visualization
Maintenance professionals have to master the science of physics and the art of interpretation. The true meaning of information locked up in numerical data can be understood only after years of experience. With an aging workforce and increasing variety of the utility assets, experience can be a guide for only so long. New tools must be brought in to help speed the delivery of new insights.
Visual data exploration lets users correlate events, identify key relationships and make more precise decisions, faster than before. Analytically savvy users can quickly identify opportunities or concerns, so further investigation can be quick.
Enhanced data visualization is a perfect marriage to the high volume, low latency data of most complex maintenance applications. The ability to parse the data, quickly search for outliers and display results on a map can shorten outages, inlcuding those caused by storms. With higher visibility for service time restoration, utilities need tools to communicate with key stakeholders who are not familiar with the underlying data sources.
Making (Dollars and) Sense of Asset Analytics
While it’s impossible to prevent every outage, the analytics available to reliability and maintenance engineers enable proactive resource management, including assets, workforce and supply chain. Priorities will be set based on business cases. In the example mentioned earlier in this article, reduced contractor hours justified the decision. For other utilities, decisions will be determined by the capital investments that can be deferred by extending the lifetime of aging assets.
As IT and OT converge and assets become digitized, maintenance and reliability engineers are applying analytics to derive value from big data.
Alyssa Farrell leads global industry marketing for SAS’ business within the energy sector, including utilities and oil and gas. Farrell is a member of the North Carolina Technology Association Green Technologies Council and contributes to the product strategy for SAS’ sustainability software portfolio. Prior to joining SAS, she was a senior consultant in the Deloitte Public Sector practice. She earned her MBA degree with a concentration in management information systems from the University of Arizona. She also holds a bachelor of arts degree from Duke University.More PowerGrid International Issue Articles
PowerGrid International Articles Archives
View Power Generation Articles on PennEnergy.com