Utility revenue protection: faster, easier, better with data analytics

Published: Fri 02 May 2014
A blog entry by Smart Grid Watch

Contributed by:

Smart Grid Watch
Smart Grid Blog
Siemens

Smart Grid Watch's Blog

Electricity theft is a significant and growing drain on utility revenues around the world. However, modern data analytics technologies now make it faster and easier for utilities to detect energy theft and other forms of non-technical loss anywhere on the network.

A breakout session at this year's Smart Grid Software Leadership Conference (May 5-7, Orlando, Fla.) will offer a preview of a new analytics product focused on revenue protection.

Don't miss: Revenue Protection Analytics

Sumeet Ganju, Product Manager for eMeter, a Siemens Business, explained what his session will cover, and what the new analytics offering will provide.

Most utilities already have revenue protection strategies. Why is it crucial to use advanced analytics to help solve this problem?

Ganju: Thanks to smart meters, utilities now have far more detailed data available on the energy usage patterns of individual customers. AMI meters have enhanced our ability not just to gather data, but to also generate deeper insights from it. Utilities no longer have to guess about consumption patterns based on one meter read a month. Interval data, combined with alarms and events from smart meters, can be very useful and effective tools for identifying a significant percentage of nontechnical losses.

In particular, interval data can be trended over time, allowing utilities to build detailed and useful consumption profiles. Utilities can also build consumption profiles for known cases of theft. By using advanced analytics techniques, customer usage patterns can be compared against fraudulent usage profiles, as well as to the profiles of similar customers. This can flag service points where energy theft may be occurring.

An analytics-based approach allows utilities to automatically review almost every account continuously. This greatly increases their ability to detect nontechnical losses. It not only allows utilities to recover more revenue (and prevent more losses), but it also greatly reduces the operational costs of a revenue protection program by increasing operational efficiency and accuracy.

What will eMeter's new revenue protection analytics system include?

Ganju: This system will provide a complete revenue protection workflow. It includes advanced theft detection algorithms and machine learning to identify potential cases of nontechnical loss. It allows utilities to perform online data review, recording the results of field investigations, and many other tasks that are critical to the revenue protection process.

Why is machine learning important for spotting electricity theft?

Ganju: Electricity thieves keep getting more sophisticated about concealing their activities. Machine learning allows our analytics tools to keep pace with shifting patterns of anomalous or potentially suspicious energy usage patterns.

The software can say, "Look, here's an unusual pattern, there may be something odd going on." Then, utility staff can use the analytics to drill down further into the data to make a determination. This way, advanced analytics can help you spot electricity theft -- even when you don't know exactly what that theft might look like.

----------------------------

More interviews and information about the Siemens Smart Grid Software Leadership Conference.

Register now for this conference.

Read more at Smart Grid Watch!