The 5 best places to find value in your Big Data
By: SGN Staff
It was a couple years back, at an invitation-only utility summit. As soon as the first Q&A session began, the CIO of a Midwest utility started waving his hand. "Can somebody please, please, please tell me what I should be doing with all my data?" he pleaded.
I encounter similar attitudes wherever I go. When it comes to Big Data, most utilities don't have a clue where to start. Where to find the low-hanging fruit. Where to find those first starter applications that can show results quickly.
And I had no answers for them... until recently. I sit on the Advisory Board of a utility analytics company called TROVE. As TROVE began to work with multiple utilities, they started to notice patterns. Places that were most likely to yield an early quick win.
So I asked if I could share their findings with you. Herewith, the top five spots where TROVE customers are finding value in their Big Data. Your mileage may vary, of course, but I predict that one or more of these five will apply to your situation as well. - Jesse Berst
By Isaias Sudit
Energy Efficiency Targeting: Your consumption data can help you identify the best candidates for enrollment. When it treats all customers the same, a utility is not properly marketing its energy efficiency products to the customers who need it most. For example, TROVE is working with a large Midwestern utility, where we were able to double the total energy reduction for roughly half the campaign costs. The company had undiscovered value in its data. By taking advantage of that data, it was able to increase its response rates by 2.3 times and its energy reduction per customer by 1.5 times. By fusing multiple data sets and applying our algorithms, we were able to identify which customers were most likely to take action, and how much they could reduce energy consumption.
Demand Response Targeting: Finding the right DR candidates is difficult. Fortunately, interval data can give you valuable insights into who has the greatest opportunity to shed load. TROVE worked with a large Midwestern utility to help it optimize its DR enrollment, leading to $7 million in deployment cost savings while increasing the maximum load reduction by 22%. The utility's marketing consultants had identified segments they thought would be the best customers. Yet TROVE used interval data to prove that those segments were not even close to the best ones for demand response. That project served to emphasize that human bias often interferes with analysis. If you let the data tell the story - if you use predictive data science - you can extract accurate insights not evident to human intuition.