Grid Modernization

Emerald PUD uses DEED grant for data analytics project with TEA

Emerald People’s Utility District in Oregon has used a Demonstration of Energy & Efficiency Developments (DEED) grant from the American Public Power Association to launch an effort to make better use of its data.

The Oregon public power utility began switching over to Advanced Metering Infrastructure (AMI) in early 2018, providing it with a wealth of customer data, and was beginning to reach the end of the installations as the DEED project was getting under way. The challenge was how to make the best use of that data.

When it applied for the DEED grant, the primary driver was proof of concept, “proving that we could derive value from all the new AMI data,” Kyle Roadman, power resources manager at EPUD, said.

“We had a couple of options,” Roadman explained. One was to invest significant funds into an off-the-shelf analytics product. Another was to try out a new analytics-as-a-service model with The Energy Authority (TEA).

Because the as-a-service option was still “somewhat unproven, we were hesitant to put a lot of ratepayer dollars into it,” Roadman said. “The DEED grant allowed us to get the service off the ground and determine if it would be a good long-term investment for our customers.”

In addition to AMI, EPUD was looking to find the best use for the data generated by all the technology upgrades on its roadmap, including GIS, outage management, staking, work order management, and enterprise software.

Going into the DEED project, EPUD’s goals were twofold: assist TEA in developing a flexible advanced analytics platform that can accept data inputs from any source regardless of vendor and to run a series of real world studies using advanced analytics to determine if the benefits are suitable for a smaller public power utility.

EPUD provides electric service to about 22,000 customers in the Eugene-Springfield metropolitan area of Lane County and has a peak demand of 130 MW in the winter.

“The DEED grant gave us initial confidence to try out something new and different, something we would not have done on our own,” Roadman said.

“A lot of utilities invest in technology upgrades” with the understanding “that a bunch of data” will certainly be valuable, Todd O’Keefe, client services manager at The Energy Authority, said. The attitude, he said, seems to be, “We’ll gather the data and figure out what to do with it later.” In fact, “what you do with the data is the tricky part,” O’Keefe said.

The first step in the DEED project was to build out a platform to collect, store, access and analyze the utility’s data. The next step was to perform actual advanced analytics. The goal, as stated in EPUD’s DEED report, was to show that “data analysis can help a utility like Emerald solve real world issues.”

The original list of potential studies was quite broad. It included everything from transformer load management and meter tampering detection to system connectivity mapping and energy efficiency marketing.

EPUD quickly realized, however, that there were data quality issues that made the mapping project unfeasible. Specifically, EPUD’s AMI vendor supplies voltage data in whole numbers. EPUD realized that the advanced analytics it was hoping to see is not possible without decimal point precision.

EPUD is working with its vendor to add voltage precision, but that was not possible within the DEED timeframe. EPUD and TEA realized that energy efficiency and rate design analysis could be performed using primarily usage interval data, rather than voltage data, and so they pivoted.

EPUD has seen rapid retail load growth since 2015 and is working under a directive from its board of directors to look at new ways of marketing energy efficiency without incurring big cost increases. In addition, a resource capacity shortage has surfaced in the Northwest and conventional products for dealing with short-term power supply needs have become scarce. So, EPUD wanted to study peak demand reduction through both energy efficiency and time of use rate design.

The first question EPUD sought to answer was if it could improve the cost effectiveness of its energy efficiency programs by targeting customers who could potentially save the most energy. Analyzing its user data with TEA, EPUD was able to develop a marketing campaign aimed at 458 high potential customers it identified.

The marketing campaign had a response rate of 39.5%, with responses from 181 customers. Of those, 38 had moved forward with energy efficiency improvements by April 2020, 29 customers were determined to have additional buildings on their account, and 18 were determined to likely be running an indoor agricultural operation.

EPUD estimates the pilot marketing campaign has reaped energy savings of between 20,000 kilowatt hours and 140,000 kWh, and it intends to continue its marketing and outreach program throughout 2020.

EPUD also wanted to test if it could use AMI data to determine if a customer has completed an energy efficiency project. Unexpectedly, the program uncovered increases in some customers’ electricity usage.

In most cases, it turned out the increase was the result of customers who previously burned wood for heat, switching to an efficient heat pump system. So, even though the result was higher load, it also will provide data that will help with the utility’s program planning moving forward.

EPUD also wanted to see if it could more accurately measure energy efficiency savings. In most residential projects, EPUD, like other utilities, tended to rely on “deemed” savings estimates established by standard setting bodies, which are generally seen as accurate but never fully vetted at the utility level.

Comparing year-over-year changes in electricity consumption among participants in a heat pump water heater program with a control group of non-participants, EPUD and TEA found the median energy savings between a participating customer and a non-participating customer came in at 10%, or about 130 kWh per month.

EPUD was also looking to understand how to best design rates to reduce peak demand. Although the rate design was ongoing as of April, EPUD reported “significant benefits” from the process. The analysis was able to quickly isolate the customer-level impacts of potential rates based on varying design parameters. Using the utility’s AMI data, TEA provided EPUD with a variety of rate scenarios and their impact on customers that could be compared with current residential rates.

Based on those capabilities, EPUD plans to launch a voluntary pilot time-of-use program later this year and to further utilize data analytics to determine both savings potential and true customer bill impacts. “This analysis will provide crucial insight as the utility decides whether to move forward with a broader time-of-use effort for its customers,” EPUD said in the DEED report.

Overall, EPUD concluded the DEED project “has proven that a utility its size can benefit from advanced analytics work. Although not every method resulted in an expected result, each study provided valuable insight and allowed the utility to approach its operations in a different way.” The utility’s staff was also “encouraged by the flexibility and efficiency of this ‘Analytics as a Service’ approach and sees great potential for continuing with it in the future.”

The DEED project began in July 2018 and was substantially complete by Oct. 31, 2019. EPUD staff continues to perform additional marketing and outreach based on the original analytic studies, with results of these efforts compiled through the end of April 2020.

On June 25, APPA will host a webinar related to the DEED project. Additional information on the webinar is available here.