Data analytics: how to, and why

Real-world applications and best practices

Phil Carson | Aug 29, 2012

Share/Save  

Every utility captures some forms of data and rolls that up into reports to understand their systems, gauge asset conditions and analyze business operations. But with new data sources and analytical capabilities available, where are the shortest paths to value?

In a webcast yesterday titled "Analytics Best Practices & Case Studies," hosted by Christine Richards, director of the Utility Analytics Institute (sister to Intelligent Utility under the Energy Central banner), our audience got an overview of North American utilities in this field. And they heard from one small utility (Fort Collins Utilities) and one very large utility (Sempra Utilities) in the same vein. 

You can listen to the presentation here and download the slide deck here.

The webcast served as a preview of the UAI's fourth report in its Executive Insight Series, titled "Case Studies and Best Practices," which looks at four key areas: asset and grid optimization, customer operations and engagement.

Richards offered insights on technology, people and business processes in the utility analytics field and how those factors can be aligned with the enterprise's strategy. She suggested seven phases of value, from the OMG! phase (see: Lucille Ball working on a candy conveyor belt) to the transformation of business processes based on data analysis. 

Most utilities today fall into the first four categories, the "foundational phase," in Richards' parlance, which culminates in dashboard presentations of historical data to answer "what happened?" Just ahead: the promise of modeling and planning based on that historic data. Then, in the next couple years, the big departure for the land of predictive analytics and the use of real-time data for real-time decision making.     

To reach those value-laden zones, the essential exercise is thinking backwards, Richards said in so many words. By using key utility goals as guidance, technology can be right-sized to produce the data and people can be trained to support data acquisition, analysis and the development of resulting business intelligence. The how-to? Leverage existing capabilities and systems to add value to existing business processes. 

Those ideas address both grid optimization and external customer engagement efforts, which teed up a presentation by John Phelan, energy services manager, Fort Collins Utilities, on how his organization has used analytics for customer engagement. 

Fort Collins Utilities is a municipal utility in a college town north of Denver with 65,000 customers of electricity, water and wastewater services. The city's energy policy goals, the result of its well-educated, environmentally aware citizens' outlooks, drive programs that this muni uses data to achieve, according to Phelan.

"Data analytics serve policy initiatives," Phelan said.

Thus Fort Collins is using data analytics to segment its customers and deliver segment-appropriate messages around, for instance, energy efficiency programs. Using a cloud-based database and combining four sources of data (electricity and water use, GIS and weather), the utility has identified 1,000 of its 5,000 commercial customers that are most likely to take advantage of energy efficiency programs. Then it ranked those 1,000 commercial customers by the size of their potential savings and began outreach efforts accordingly.  

How? Needing only quarterly data to achieve this end result, the utility used a third-party vendor for software licensing and consulting services to create its own system. Data categories include, for instance, which customers have availed themselves of an energy audit, the type/size of facility or residence, weather data and consumption ranges. By using filters to set ranges, the utility can quickly see, for instance, which commercial customers might achieve the biggest energy savings and be approached to consider, say, heating, ventilation and air conditioning (HVAC) upgrades.

Thus data analytics can serve the city's goals of capping growth in electric demand as well as shaving peak load and its costs in dollars and pollution. And in the specific case of commercial customers in Fort Collins, the utility also is working on those customers' behalf to help them manage pass-through demand charges from the city's wholesaler.

Lessons learned along the way included the fact that customer segmentation, for a municipal utility, remains a new concept. Follow-through to realize actual benefits is critical. Whether data analytics effectively informs program design and targeted marketing remains to be seen as this nascent effort begins to bear fruit, Phelan said. 

Lorie Mariano, business intelligence manager for Sempra Utilities, described a three-year analytics effort for San Diego Gas & Electric and Southern California Gas that, having achieved some early wins in one area is being expanded to other business units. 

The first step was developing a strategy and roadmap informed by understanding organizational pain points, Mariano said.

A review of existing capabilities asked whether pertinent data existed and whether it was in a useful form. Moving ahead meant beefing up the business intelligence-related IT infrastructure, implementing new capabilities and developing standards for uniform outputs. Success in these endeavors produced a six-fold expansion in the use of analytics in three years, according to Mariano. 

Best practices for business intelligence specialists meant leveraging and building upon their core competencies, deepening the bench of talent and cultivating evangelists who could communicate the value of data analytics. For internal clients, best practices included supporting "super-users," collaborating with executive sponsors and promoting BI self-service, Mariano said. 

In response to an audience question, both Phelan and Mariano agreed that practitioners should not underestimate the resources needed to support the ongoing effort to ensure data quality and the process of spinning data into business intelligence. Otherwise, the failure to realize swift and useful benefits can doom a program.

Many of these issues will be discussed in depth at the Utility Analytics Week event, Sept. 18-20, 2012, in Arlington, Texas

Phil Carson
Intelligent Utility Daily
pcarson@energycentral.com
303-228-4757   

 

Related Topics