With all of this data, will we finally understand how our customers use electricity?
As we move into an era where electric utilities have fifteen-minute or hourly load information on most—if not all—of their customers, will we finally understand how our customers use electricity? I think the answer to this shockingly simple question is probably not. You see, it is entirely possible for two houses, constructed exactly the same, with identical appliances, same number and age of occupants, sitting side-by-side to have vastly different energy consumption profiles. It’s not simply the end-uses and customer demographics of each household that drives energy use but, perhaps more importantly, how the household members interact with the energy consuming appliances that matters. Individualized customer behavior creates customer-to-customer variability making all the difference in the world.
As an industry, we have been moving to a much more time differentiated planning horizon where loads and costs vary hourly. We are hearing the folks in charge of the electricity markets talk about truly integrating supply and demand including “prices to devices” (PTD) applications. Can we really understand the impact of “prices on devices” if we don’t understand the usage characteristics of the devices?
Only through end-use data and insight can the utility and market planners have the detail necessary to understand how the various end-uses contribute to the overall profile. Let’s be clear, end-use information refers to the energy and demand characteristics that break down the customers’ total household or facility energy use into the various end-use components like lighting, refrigeration, and HVAC. Figure 1 presents an end-use profile by displaying the annual energy shares of each end-use consuming appliance.
End-use information is available in various forms, such as appliance saturation surveys, laboratory ratings, and modeled energy consumption, including some limited, or dated, end-use metered data (i.e., load shapes). While modeled end-use data provides a useful starting point, we believe that only metered end-use consumption data captures the true variability inherent in actual customer consumption. Therein lays the rub because capturing actual in-field, end-use data is complex, often intrusive, and very expensive. While there are certainly numerous applications within the utility that benefit greatly from end-use information, developing a “slam dunk” business case for actual, end-use data collection has proven challenging. The primary reason is that assigning “value” to improving modeled data with actual information can be fairly subjective.
As we look into our crystal ball we see promises out on the horizon. The Electric Power Research Institute (EPRI) continues to work on advanced metering technologies like Non-Intrusive Load Monitoring (NILM). NILM and Single Point End-use Energy Disaggregation (SPEED) devices permit the collection of a multitude of appliance end-use loads without the intrusive wiring nightmares of past years. These hardware/software systems purport to allow for the collection of appliance-specific load data without entering customer premises and without installing metering devices on appliances. The analysis software seeks to recognize appliance signatures in the interval load data, for example, an electric water heater with a 3,500 Watt heating element looks significantly different than a 100 Watt light bulb.
DNV GL is collaborating with EPRI and Oak Ridge National Lab and hopes to release a white paper on this very topic in the coming weeks (stay tuned!). Figure 2 highlights the avenues we plan to explore in the white paper including statistical, modeling, direct metering, and advanced metering strategies. Our vision is that through a regional/national effort the industry will return to focusing on and understanding how our customers use and value electricity.
For more information, please contact me at email@example.com. I will also be moderating a Customer Engagement panel at DNV GL’s Utility of the Future Leadership Forum, June 2-4, 2014. Learn more about the panel and Forum here.
 End-use information is applicable load forecasting, integrated resource planning, generation resource acquisition, transmission and distribution planning, transformer load management, demand response and energy efficiency evaluation, rates and regulatory policy, marketing, and customer service operations,