The Peak Rate Conundrum

When you’re a distribution utility, rate hikes are a fact of life. Routine base rate hikes, like TVA’s 1.5 percent increase in 2016 are generally treated as pass-through charges, with the utility simply raising customers’ base rates at the same percentage. Peak use rates create similar challenges, with additional complexities.

Lisa Greeson, supervisor of information systems for Alabama’s Sand Mountain Electric Cooperative, described the challenge that unpredictable weather can have on TVA’s peak rates and the co-op’s monthly bottom line. “Suppose you set rates in early February, hitting peak early due to a cold spell. Then the weather warms up, usage falls off, and you don’t sell enough to meet the high demand,'”said Greeson. “Without a steady flow of KW hours over the month, we lose money.”

While base rate increases may not be something a distribution utility can control, peak rate charges are. In fact, the more precisely a utility can understand the causes of its customers’ peak usage – and its impact on the peak pricing set by the generation utility – the better equipped it is to act to manage its bottom line by setting fair and optimal rates within its use classes.

At Sand Mountain, its TVA peak rates are based on usage measured at the substation level, while the co-op has additional visibility into usage at the customer level using meter data management (MDM) software installed in early 2014.

Sand Mountain’s MDM provides a detailed picture of not only each of its over 31,000 members’ total use, but of their individual peaks as well. They could, if desired, identify individual meter peak by the day of the month or hour of the day, then look for peak patterns over seasons or across years. This additional granularity can be leveraged in significant ways.

The Promise of Refined Rates

With the deployment of the MDM system in 2014, Sand Mountain General Manager Mike Simpson realized that MDM data at the individual meter level could be used to achieve two key goals: to refine rate classes more accurately around cost of service and to provide the basis for time-of-use (TOU) rate programs for the utility.

Primarily three TVA rate classes are currently applied to Sand Mountain customers: Rate Class 22 Residential (22kWh); Rate Class 40 Small Commercial (<50 kW & <15,000 kWh); Rate Class 50 (>50 kW & <1,000 kW; or <50 kW but >15,000 kWh). In the mostly rural region of Northeast Alabama the co-op serves, there are no current customers in the over 5000 KW category.

Simpson’s first objective was to examine data from coincident peaks, those times when TVA substation data and Sand Mountain MDM data both recorded peak. By comparing the utility’s system peak to the TVA peak, the co-op would begin to see what – and who – is causing it to happen. Armed with this information, they could potentially hone in the rates, determining the true cost of service and structuring rates to put the burden of peak rates on those creating the greatest peaks.

Here, Sand Mountain hit obstacles. While both Access and Excel databases feature data filtering, collation and querying, neither could access the data or provide the tools to analyze it in the way they envisioned simply due to the immense amount of data involved. Cloud-based solutions also fell short. Simpson explains, “At that point, it was simple math. We wanted to look at 5 pieces of data for each hour of every day for an entire year for each of our 31,000 customers. That comes out to nearly 1.4 billion pieces of data. If you look back at 2 years of history, you double that number. Excel and Access just can’t handle that many records.” Then at a CEO conference in 2015, Simpson discovered that SEDC (Sand Mountain’s provider for their billing software suite, MDM, and GIS solution), had launched an analytics solution, CatalystIQ. “I realized IQ could do what I wanted. Excel and Access need to physically ingest the data. IQ simply touches each database and gathers just what it needs – it doesn’t physically store the data separately like a spreadsheet does, so it can easily handle billions of pieces of data.”

While the software was designed to provide utilities with sophisticated analysis of any number of disparate data streams from across the utility enterprise, for Simpson, this was the answer to one core problem — the need to dig into MDM data to refine rates. As Greeson put it, “This was the whole reason we got it.”

Digging into the Data

Using the CatalystIQ core platform, Sand Mountain could at last join data from across distinct datasets and sources. The application blends, analyzes and displays it in seconds – not hours or weeks – enabling the co-op to see patterns, identify trends, and gain insights. It provided the power to join the multiple databases involved in answering the complex strategic questions involved in rate planning.

SEDC worked with Sand Mountain, developing the tool to pull data from the MDM system and execute tasks like calculating percentages on the data. It might show, for example, that within a selected (peak) time interval, one rate class represents 60 percent of total use; then further consider this finding against the percentage of total customers the class represents.

This IQ dashboard shows intervals that are in the top 50 for Residential Rates where the monthly consumption is over 3,000 kWh, allowing Sand Mountain EC to identify the upper band of residential use.

Have there been any surprises in the data? “The biggest users have been residential,” said Greeson, who noted that there are agricultural and other commercial customers in the largely rural northeast Alabama service area. Other surprises emerged once the analytics drilled down to find who’s driving peak. To find the top 50 users within the customer base or a rate class, for example, Sand Mountain can flag customers whose use exceeds 3,000 kWh. The tool begins to “throw a mark” on the data any time a user hits that point. Tags begin to accumulate and reveal patterns. One customer showed up in top 50 results every day, with peak use occurring from 9 a.m. to noon. This triggered questions. Who was this customer and did he really belong in the residential rate class? That allowed the utility to calculate a true cost of service for this residential customer (who turned out to be a big church) with commercial level usage.

Once SEDC had customized the tool to achieve its objectives for usage analysis and provided essential training to utility users, Sand Mountain began to apply it not only to current data but also to the two years’ worth of data compiled since the MDM’s launch in 2014. “That data didn’t go away and now with the IQ tool, we could apply the analytics to it,” said Greeson. The historic data began to yield insights into longstanding (or evolving) use patterns that could be examined by rate class, by time of use, and over time.

The findings not only will help place customers accurately within rate classes, they advance fair and transparent rates tied closely to cost of use and, significantly, provide a foundation for proposed time of use rates.

Toward Time of Use Offerings

For Sand Mountain, time of use rates could empower customers with new ways to manage their energy costs, motivating customers to adjust their energy consumption behaviors to reduce peak use. Sand Mountain General Manager Mike Simpson also sees tremendous potential savings for customers and a potential reduction in the utility’s wholesale power bill. While one goal is to stabilize the utility’s margins, Simpson points out that “the customers are the utility, and the utility is the customer. If our customers save money, then the utility saves money, and we’ll pass that savings along to them.”

To secure approval for rate changes or new time of use offerings, Sand Mountain will need TVA approval. It is currently working with its rate consultants, providing them the data findings from CatalystIQ to substantiate its proposals. The aim is to use the advanced analytics to create a rate structure that is fair, transparent, and rooted in sound economics.

Together, the rate class refinements and time of use structures could help Sand Mountain to:

  • More fairly distribute the burden of peak rates on those whose use created the greatest peaks.
  • Motivate customers to adjust use behavior to reduce peak use.
  • Help control costs for all customers by avoiding or reducing extraneous peak usage.
  • Promote cleaner air and reduce environmental impact through fewer incidents requiring TVA to activate peaker plants.

Ultimately, Simpson sees CatalystIQ as a tool that will help the utility achieve an impact in the community that he has strived toward for years. He plans to use the insights gained through CatalystIQ to develop a new rate that rewards consumers who shift their usage patterns to take advantage of off peak savings. “The question has always been, how do you get to the point where customers can reduce their usage without just saying conserve, conserve, conserve,” he points out. “But if we create a rate that rewards them for shifting their patterns, they’ll have an incentive to make changes that can save them money on their bill. If each one of our customers can reduce their bill by $10 a month, keep that money in their pocket, spend it locally, over the course of a year that’s $3.7 million that will stay here and be spent in the local economy. It’s a win-win situation for them, for us, and for the community.”

Post Script

As of October 2016, Sand Mountain continues to utilize CatalystIQ for MDM analysis and to explore its future uses. The co-op continues to develop its staff’s skills in the solution, enrolling key employees in advanced Medallion training from SEDC. Because each user within the utility has the flexibility to create custom analytics and gather the data exactly the way they want to see it, the potential use cases to unlock additional value from utility business or operational data offers virtually unlimited options.