Surprising Data on When People are Charging
by Brewster McCracken, President and CEO, Pecan Street Research Institute
What if electric vehicle drivers don’t charge at the same time each afternoon after all?
That’s one of the tantalizing possibilities raised by the most recent data from the Pecan Street Research (PSR) Institute’s electric vehicle (EV) research trial.
The most recent data is for 30 homes during the June 1 – August 31, 2013 quarter. (PSR’s EV research trial consists of 74 EV drivers in Austin who own a mix of Leafs and Volts, along with two Tesla Model S owners). Half of the sample consists of EV owners participating in a time-of-use pricing trial. All participants have Level 2 home chargers, and their electricity use and charging are measured at one-minute intervals.
The data from this 30-home sample of EV drivers suggests that weekday EV home charging may be much more distributed in the real world than has been projected in behavioral models. In evaluating electric system preparedness for electric vehicles, a critical question is: are residential EV charging patterns more similar to residential air conditioner loads (which come on en masse at roughly the same time on summer afternoons), or are they more similar to electric clothes dryer use patterns (another very large electric load whose use is more randomly distributed than air conditioner use)?
Figure 1: Weekday distribution of all EV charging
To the extent the measured patterns in the studied sample are confirmed by other research results, these findings suggest EV home charging patterns could more closely resemble electric clothes dryers. If so, that raises the possibility that utilities may be able to serve comfortably much higher levels of EV adoption than has been forecast by studies that relied on modeled charging behavior.
As shown in Figure 1, key findings include:
- For weekday charging, the portion of home charging that occurred during peak demand hours was 17%. (PSR used ERCOT’s definition of summer peak hours: 3-7 pm CDT). This is a far lower percentage than what is used in most EV charging behavioral models.
- For the 15 participants participating in the time-of-use pricing trial, the portion of weekday home charging during peak was 12%. For the other 15 participants (who are not participating in any pricing or EV behavioral trial), the portion of charging occurring during peak was 22%. Even this value is still far less charging during peak than has been projected in most models.
Figure 2: Weekday distribution of all EV charging for off-peak and peak times
- For homes in the pricing trial, the highest levels of weekday charging occurred roughly between 11:45 pm and 2:30 am. For homes not in the pricing trial, the highest levels of charging occurred between 9 pm and 11 pm (see Figure 2).
- For both groups, the lowest period of weekday charging occurred roughly between 6 am and 7 am.
- As shown in Figure 3, the average duration of weekday charge events was 121 minutes. This means the EV participants were, on average, returning home from work with more than one-third of their battery life remaining. (With a Level 2 (240 V) charger, a Volt with a fully discharged battery takes about 210 minutes to charge fully, while a Leaf takes about 420 minutes.)
EV’s represent the largest new electric load to appear in homes in a generation – even at Level 1 charge rates (120 V). On top of that, EV early adopters are frequently clustered in a small number of neighborhoods. This creates the risk that disruptive impacts to utility system operations could occur even at relatively low levels of consumer adoption.
Whether utilities will find serving these loads problematic (even with distributed charging patterns) will vary for utilities. Key determinants for whether a given utility will find clusters of EV’s problematic likely include:
- Geographical location of adoption clusters. For example, if a cluster of EV adopters emerges in an area where people are more likely to have long commutes, this would clearly influence what time people tend to arrive back home, whether midday trips home are common, and how depleted the EV battery was on returning.
- Distribution system sizing. The extent to which the utility’s distribution systems is sized for air conditioning and/or all-electric homes can have a big impact on a utility’s preparedness for EV’s. Sunbelt utilities that have high seasonal cooling loads frequently are sized to accommodate these wide load swings, meaning they typically have ample available capacity. Utilities with low air conditioning loads, such as along the Pacific coast and in the Mountain West, have historically had little business purpose for sizing distribution systems to accommodate very wide load swings.
There are intuitive reasons to suspect that no matter where EV drivers live, charging patterns will prove much more variable than the research models have largely predicted. In all of our lives, stuff comes up. We have a work dinner. Sometimes we need to stop at the grocery store on the way home. Our kids need to be transported to and from after-school activities. Sometimes our hands are full when we get out of the car after arriving home.
This fluidity in daily life patterns is good news for utilities. In fact, the electric system’s stability is predicated in part on such fluidity. The key for utility planners is to understand early on how many EV’s are in the service territory, where they are located and the charging patterns of these customers.
Brewster McCracken is President and CEO of the Pecan Street Research Institute and Pike Powers Laboratory and Center for Commercialization.
Headquartered at The University of Texas, the institute’s research focuses on electric and gas reliability and environmental and behavioral economics aspects of energy use. Mr. McCracken was one of three global smart grid project leaders invited by the government of Japan to present at the one-year anniversary conference for the reconstruction of Fukushima in March 2012. He is lead author of the institute’s research analysis comparing customer electricity use in green-built and non-green older homes, Data-driven Insights from the Nation’s Deepest Ever Research on Customer Energy Use, and he is the lead author of the institute’s forthcoming whitepaper characterizing diffusion of innovation categorization of electric vehicle owners participating in Pecan Street’s electric vehicle research (which includes the nation’s highest concentration of electric vehicles).
He was elected to two terms on the Austin City Council, serving in a city-wide at large position. Through his elected position, he founded and chaired the city council’s Emerging Technologies Committee, led the city’s collaboration with The University of Texas to establish technology incubators in bioscience and wireless technologies and served for six years as a board member of Austin Energy, the nation’s fourth largest municipally-owned utility.
Prior to holding elected office, he practiced commercial litigation for nearly a decade with two large international law firms. He is an honors graduate of Princeton University and The University of Texas School of Law, and he also holds a Masters in Public Affairs from UT’s Lyndon B. Johnson School of Public Affairs.