Is Predictive Analytics in Your Future?
- Apr 14, 2016 8:20 pm GMT
DATA AND ITS ANALYSIS are all the rage these days. Pick up any business magazine or newspaper and you will see numerous articles about companies using new and old sources of data to try to improve operations, increase efficiencies, retain customers and grow revenue.
Utilities, grid operators, transmission companies and other electric industry entities naturally want in on the action. And like their counterparts in other fields, they are adopting increasingly sophisticated analytics methods. The transition that is occurring reflects a shift from using basic reporting and dynamic dashboards that present business intelligence information, telling companies what has happened or is happening, to using predictive and prescriptive analytics that do much more.
Understanding what these less familiar technologies can do for energy industry organizations will tell you why there is growing interest in using them. Most applications of analytics are descriptive. For example, a transmission company develops maintenance schedules by analyzing historical records of transformer outages or a utility plans power purchases by deriving electricity usage trends from past meter readings.
In contrast, predictive analytics uses regression analysis, modeling, neural network and simulation algorithms taking into account multiple independent variables to determine the likelihood of an event happening in the future. For example, a utility planning restoration services before a major storm might factor in historical information about past outages from similar storms, detailed weather information about the expected path of the current storm, and information about recent vegetation-clearing efforts to determine where damage is most likely to take place. The information can then be used to take preventive actions, such as clearing branches or staging repair crews before the storm.
Essentially, companies using predictive analytics can transition from being reactive (taking some action after an event occurs) to being proactive (taking actions to prevent or minimize the effect of an event). For example, predictive analytics can be used to help utilities shift from traditional time-based asset management - where network repairs are done on schedule regardless of how much useful life is left in an asset - to a more informed reliability-based approach of making repairs when they are actually needed. There are many applications of predictive analytics in other operational areas.
Prescriptive analytics takes things to a higher level. It builds on the benefits of predictive analytics by recommending one or more courses of action - and showing the likely outcome of each decision. For example, a utility might use predictive analytics to identify what percentage of customers is most likely to fall behind on paying bills, whereas a prescriptive analytics approach would provide guidance as to which debt relief approach to use on a specific customer or set of customers.
Tools to conduct the more sophisticated predictive and prescriptive analytics work are coming from traditional analytics software companies such as IBM and SAS. There also are some newer companies such as Opower and C3 Energy that offer industry-specific solutions that are increasingly being used by utilities.
For example, earlier this year, C3 Energy announced a large new project with Baltimore Gas & Electric, a project that the utility hopes will save it hundreds of millions of dollars by using analytics to manage operations of the 2 million smart meters it's deploying, and by tapping its data to discover and prevent energy theft and revenue losses.
Although there have been instances where these less familiar technologies have been applied in the past, the general consensus is that this is the year when the industry will see a much broader embracement of predictive analytics, with adoption of prescriptive analytics to follow close behind.
At the start of 2015, the Utility Analytics Institute noted that analytics use by most of its members fell somewhere between basic reporting and predictive. However, the institute pointed out that no utility had adopted analytics adoption across all of its business divisions. The primary reason that most utilities have staggered their programs is a strategic decision to tackle analytics in key areas such as advanced meter infrastructure, asset management or customer service.
One factor leading companies to the use of predictive analytics is the abundance of data. "Last year, there was much greater understanding of how important data is to solving multiple problems," said Scott Stallard, vice president of asset management and a data analytics subject matter expert at the consultancy Black & Veatch. At the same time, there has been an explosion in the amount of data available for analysis, thanks to the increased deployment of smart devices, whether it involves meters at customer sites or monitoring sensors on critical infrastructure equipment. Stallard noted that these smart devices create new opportunities for using predictive analytics.
Critically, many different business and operational areas within all energy industry organizations now have more data available. This in turn means that many aspects of running a company can benefit from predictive analytics. "With the availability of data, there are many predictive analytics applications including demand response, preventive maintenance [of transformers, for example], energy trading and many more," said Ken Budka, chief technology officer for strategic industries at Alcatel-Lucent.
"Predictive analytics is being applied in the asset space, energy forecasting, and customer area and on the operational side in many utilities," said Alyssa Farrell, product marketing manager for energy and sustainability at SAS. She noted an effort that involves SAS working with the electric company Ameren Missouri. The companies developed a predictive model for storm restoration using historical outage information, near real-time National Weather Service data and other third-party information to predict a storm's effect and to guide restoration efforts. The models helped Ameren smartly stage materials, equipment and crews, leading to a 22 percent improvement in restoration time.
Use of predictive analytics has already delivered similarly impressive results for one company in an entirely different area. In one case, WNS Limited, a global business process management company working with a large utility, was able to increase the utility's debt collection by 50 percent using a propensity-to-pay predictive analytics model.
These are just the tip of the iceberg when it comes to the ways predictive analytics can help. As more companies realize the potential applications and benefits of predictive analytics, 2015 is expected to see companies embrace the technology on a much wider scale in diverse areas including demand response, revenue collection, capacity planning, energy purchasing, asset management, maintenance, marketing and operations.
No discussions yet. Start a discussion below.
Get Published - Build a Following
The Energy Central Power Industry Network is based on one core idea - power industry professionals helping each other and advancing the industry by sharing and learning from each other.
If you have an experience or insight to share or have learned something from a conference or seminar, your peers and colleagues on Energy Central want to hear about it. It's also easy to share a link to an article you've liked or an industry resource that you think would be helpful.