How to Optimize the Value of Information Through Data Analytics


by Brian Crow, Sensus

Data and data analytics are critical for electric utilities. Outage management is one example of the power of data analytics with which most utilities are familiar.

When the lights are out and the only sound is ringing phone lines, just knowing which territories no longer have service can be the difference between a customer service nightmare and a model success story for power restoration.

If not properly handled, power outages can cause system damage and significant costs. They also mean utility customers might be stuck without electricity for hours, days or even weeks.

Consider what happened to one electric utility in the middle of the infamous Tornado Alley. The area experienced a series of devastating storms and tornados, resulting in serious damage to the utility’s transmission and distribution systems. By leveraging meter data with its outage management system, the utility identified outages faster and completed all repairs to its transmission and distribution systems within a month. This system integration also enhanced the utility’s relationship with its customers by restoring power and resuming operations in record time.

This is one example of the many opportunities utilities have to maximize customer service and operational efficiency through data analytics. With mountains of data piling up and seemingly endless opportunities, all utilities should be asking, “How do I get started?”

Getting Started

As utilities implement communication systems to better their operations, they receive endless amounts of data from their infrastructure and external sources; however, the data alone is just that: data. What matters is what utilities do with this data.

There are three steps to turning data into information:

  1. Collect the data.
  2. Analyze the data.
  3. Convert the analysis into actionable insights.

For each step, there are key takeaways to successfully implementing data analytics and using the information.

Collecting the data. Communication networks can provide utilities with data about power usage, the utilities’ infrastructure and outages. This information is useful, but utilities are asking, “What else does this information tell me? Are there other opportunities to leverage this information to improve operations?”

To answer these questions, utilities must consider the other sources of data that they could be tapping to gain a more comprehensive view of their systems. They also must avoid getting ahead of themselves.

How do you determine what data to collect and integrate? First, prioritize what issues can be solved with data and begin collecting the data accordingly. For example, the combination of advanced metering infrastructure (AMI) data and supervisory control and data acquisition (SCADA) data can help a utility conduct a distribution loss analysis. If power losses in the network are an issue, this data collection can help solve it. More specifically, AMI data, which provides a bottom-up review, in conjunction with SCADA data, which is top-down, can help utilities determine the difference between power distributed and power sold to their customers. This loss information is helpful, but utilities can take this analysis further by using analytics to profile and trend these losses. By monitoring loss patterns, a utility can understand better when and where losses are occur. This analysis might point to underlying issues such as a faulty wire connection, or even larger issues like a failed distribution asset.

Sometimes the actions of one department can affect the entire utility. Data collection improves coordination within the utility, helping tear down walls among departments. Through data collection and analysis, every department can see the big picture and work together to improve operations for the utility and for the benefit of its customers.

Analyzing the data. Data analytics can help utilities receive the maximum value out of the information collected.

One example is using the data to identify significant changes in usage patterns. Often these usage pattern variations point to a change in customer behavior that could lead to power outages or failed transformers. How can that be? With the rise in electric vehicles, for example, customers who live in older homes might not have the appropriate transformers to accommodate this new surge in energy use. By analyzing customer usage for unusual energy spikes, a utility can contact those customers to determine the reasons behind the flux in energy use and evaluate the transformers before outages occur. Utilities also can enhance their customer relationships by proactively offering customers different rate plans or suggesting customers charge their electric vehicles during off-peak times, should that be the reason for the spike in power use.

Data analytics also can help utilities develop the best rate structures for their customers while incentivizing customers to use energy when it is most beneficial to the utilities. As tiered rate structures become more complex, utilities will need to understand how current pricing structures affect when customers use electricity or curtail total use. Through data analytics, utilities can review customer profiles to ensure customers have the proper rate structures for their needs.

Finally, this intelligence allows utilities to automatically adjust to any perceived discrepancies in the data the moment such discrepancies appear. Utilities can preselect responses and organizational tactics for different types of incoming information. In the event that an issue or discrepancy occurs with a customer’s power consumption, such as an outage or drastic increase in power usage, a utility could receive an immediate alert. Alerts improve utility response rates and enhance customer service. With alerts, utilities often are the ones letting customers know of issues and that solutions are on the way.

Converting analysis into actionable insights. With the massive amount of data available to utilities, a key part of data management is not just sorting through the data but being able to pull in actionable insights. One actionable insight enabled by data analytics is revenue forecasting.

With on-demand meter-reading capability, utilities can gather and immediately analyze meter data using predetermined checks and reports. This information and analysis helps track revenue in real time and enables utilities to make revenue forecast adjustments based on issues or discrepancies. By pulling customer information, utilities also can manage their business better by segmenting sales data via customer classes and estimating budgets to conserve costs and improve operations.

Data analytics also can provide environmental and societal benefits. Utilities can monitor customer usage and educate customers on their consumption by providing them with regular alerts on energy usage. With improved communication to customers about their energy consumption, self-initiated conservation can occur.

Bringing it all Together

Given the ever-changing nature of the electric industry, data analytics provides flexibility through vast customization options to address the varying skill sets and needs within a particular utility. This agility allows for enhanced integration of complex networks. By collecting the data, analyzing its information and pulling actionable insights, utilities can improve operations, reduce cost and inefficiencies and enhance customer service.

Brian Crow is vice president of data analytics at Sensus. Prior to founding Verdeeco, a Sensus company, he worked for the SAS National Utility Practice, focusing on providing utilities with analytic products such as load forecasting and energy trading risk measurement. He is a licensed professional engineer in Georgia and has a Bachelor of Science from the University of Georgia.

More POWERGRID International Issue Articles
POWERGRID International Articles Archives
View Power Generation Articles on
Previous articleImproving Grid Reliability Through Machine Learning
Next articleOne Source of Truth- The Modern Era of Network Model Management in Power Transmission Companies

No posts to display