Likelihood of Adoption, Technology Preferences and Charging Flexibility of Future EV Adopters R. Kenneth Skinner, PhD Integral Analytics, Inc. Spatial Electric Expansion & Risk
If we could fast forward to 2020, what would the typical EV owner look like? What technologies would exist to encourage EV adoption? How would the local utility and governments influence our choices? Although we can know with certainty, we can speculate The U.S. electric vehicle owners will number will between 2 million and 14 million - or roughly between 1 and 5% of the national fleet
Half are plug-in hybrid (PHEV), more popular in suburbs and rural areas, and half are EV s more popular in cities and warmer climates Smart technologies within the vehicles will help manage energy consumption and risk Smart-grid technologies will help the electric utilities will manage the potential negative impact to the electric grid Several charging options will be available to the EV adopters, slow and quick charging, time-of-day, utility managed, V2G, V2H
Although EV adopters will retain the flexibility to charge when and where they want, the cost to charge will vary with the marginal cost of electricity. Electric rates will encourage adopters to slow charge at night and provide discounts for allowing the electric utility to manage or actively control when and how the batteries are charged. Fast charging during system peak-load hours will be very expensive. Utilities may prevent charging during these times.
But what do we know today about typical EV adopters of 2020? Who is that person or group? Is there a consistent demographic profile that we can use to encourage adoption and better manage electric utility capital investment dollars, keeping electric costs down for everyone?
Recent EV Studies Integral Analytics has participated in several recent EV adoption studies. Copies of the following studies are available in the back or at our booth #415. Phillips, E.G., Quanta Technology, S. Smith, Integral Analytics, T. Weaver, Northern Virginia Electric Cooperative, H.L. Willis, Quanta Technology, (2009), LoadSEER Case Study: A Land-Use Simulation Based Spatial Electric Load Forecast. Wu, M., Integral Analytics, S. Smith, Integral Analytics, T. Osterhus, Sageview Associates, (2010), Integral Analytics Planning Tips for Modeling Electric Vehicle Adoption. Stevie, R.G., Duke Energy Corp, P. Mohseni, Duke Energy Corp, (2009), Electric Vehicles: Holy Grail or Fool s Gold.
Assessing Consumer Preferences A series of consumer research methods were employed to determine the relative appeal of PEVs, hybrids and alternative transportation modes to area consumers. Differences and key drivers were assessed for both existing purchases (e.g., existing hybrids) as well as future PHEV and EVs described as a combination of textual attributes or functionality (e.g., all electric, 90 MPG, 80mph top speed, charging available only at night).
Assessing Consumer Preferences Demographic and attitudinal information were also collected, with which demographic segments were developed consistent with each group s tendency to prefer certain types of PEV functionality, charging rates, convenience, MPG, MPH and other vehicle characteristics.
Assessing Consumer Preferences Although forecasters may not know exactly which home or customer might eventually adopt a fast charging, performance oriented Tesla or Fiskar type vehicle, it is possible to discern from customer studies the likely customer segments, areas or even streets where such adoption may be more likely.
Assessing Consumer Preferences Two key aspects are emphasized in the customer research and analytics, namely predicting which customers are likely to adopt which types of vehicles, and which customers are likely to prefer rapid and/or unrestricted charging. These factors appear to be the most consequential among the possible set of EV key drivers, given their significant consequence on long term integrative utility plans.
Adopters Tend to Cluster One such example is depicted below where, for the particular EV vehicle in question (not named, here), the forecasted groups of customers that appear to be more interested in purchasing this type of EV tend to be co-located or clustered near one another, reflecting either demographic or regionally similar propensities for adopting this type of EV.
Adopters Tend to Cluster EV adopters tend to cluster by demographic concentrations Birds of a feather flock together
The observed spatial clustering of the predicted probability scores is too concentrated to be of the result of chance, as shown. Statistical Proof Of Clustering
Statistical Proof Of Clustering Visual inspection suggests that there is clustering, and the Z Score analysis confirms this statistically. We can reject a null hypothesis that the predicted probability score of electric vehicle adoption is spatially random and conclude that the predicted probability scores occur in statistically significant neighborhood clusters. This implies that several transformers in these neighborhood clusters may be at eventual risk or may present an opportunity for community charging.
Clustering and Distribution Planning As such, distribution planners working in conjunction with customer forecasting analysts may jointly determine that additional capacity may be required on one more selected sections of circuit. Customer Forecasting 30 Mile Range Forecasted Loads In 15 Years With SmartCharging Example: 11pm KVA Forecast Increase City Center Higher EV Densities Suburbs Highways KVA PER ACRE - 11:00PM High : 0.166708 Low : 0
Other Key Attributes of EV Owners Sequential analysis enables us to identify key drivers of early adoption, and to further discern which types of customer segments appear to have similar preferences, and hence, similar future EV adoption and charging behaviors. In some cases, certain customer segments preferred unrestricted charging (e.g., minivan owners). In other cases, some segments tended to embrace the concept of night time charging, early adoption of EVs, or EVs that would require larger batteries.
Likelihood of Early PHEV Adoption To determine the effect of added load in electric transmission and distribution systems, we identified which households on the circuits are more likely to adopt plug-in electric vehicles using a scoring system related to their revealed choices from the consumer research. Note that the given score is not a probability per se, but an index of likelihood similar to the way in which one is provided a score for credit risk, upon requested a loan.
Likelihood of Early PHEV Adoption Households more likely to adopt plug-in electric vehicles
Preferred Vehicle Types Additionally, we would like to project which vehicle types are likely to possibly be selected by those early adopting households. Here, we combine our adoption predictions and our segmentation analysis to project which types of vehicles we might see early on in the adoption process.
Preferred Vehicle Types Once we scale up to 20% EV penetrations, the Volt Segment clusters become more pronounced, and clustered, whereas the other vehicle adopters appear to be more dispersed. 20% Electric Vehicle Penetration All Electric (4-wheel) Plug-in Hybrid Hybrid (Non plug-in) Segway/Scooter Small All Electric Hybrid or All Electric Motorcycle
Preferred Vehicle Segmentation Chevy Volt Age between 56 to 65 (82.61% prefer) Age <=56 and house built after 1990 (60% prefer) Primary sedan driver, between 35 and 56, house built before 1990 (63.64% prefer) Small Electric Vehicles (Segway) Students: Age <24 (83.3% prefer) Retired: Age > 65 ( prefer)
Propensity to Engage in Flex Charging There are two attributes which relate to charging, namely time of day and duration, broadly reflecting a tolerance or flexibility on the part of customers to engage in some type of flexible charging options. First, is customer preference to fuel the vehicle during the day. Second is preference regarding the duration of that charge, or speed of charge. Both attributes have consequences on the load and voltage required by the distribution and supply systems.
Propensity to Engage in Flex Charging The results indicate a spatial clustering of adopters with the propensity to charge at night. The results are statistically significant and indicating that customer propensity to charge at night occurs in statistically significant spatial clusters.
Propensity to Engage in Flex Charging A large section of the clustering tends to appear within multi family space, apartment settings and townhomes, potentially reflecting both a) the possibility of increased charging flexibility, given one or two people in the dwelling vs. single family home, and/or b) the slight possibility that closely situated living arrangements (e.g., apartments) might suggest increased flexibility as a character trait.
High Propensity to Flex Charge Segments with the highest propensity to engage in flex charging include Annual income <$29,000 (81.82% prefer) Mostly condo/duplex owners, but where they are not retired or older, age <= 65 (76% prefer) Single Family, with owner aged 45 to 65, and annual income > $30,000 (64.44% prefer) Current hybrid owners, income < $35,000 NOT convertible owners, nor pick up truck owners, unless coupled with the above.
Long-Duration Charging Segments Luxury vehicle drivers are slightly more flexible on long-duration charging (59%). Moreover, the slightly less rich, living in higher income areas, tend to be more receptive to early adoption (consistent with 2008 actual hybrid purchase). Household size (>2) significantly undermines tolerance for long duration charging (36%). Small house hold size (<=2) and younger demographics, age <=35, help (65% prefer)
Time-of-Day Restrictions Overall, all customer groups would prefer to fuel their vehicle at any time of day Customers between the age of 46 and 65 are relatively flexible to time-of-day restrictions. Customers who make less than $29,000 are the only income group totally okay with noafternoon fueling Males are indifferent generally to fueling at night only, while females do not prefer it. Target men for smart charging approach. Hybrid owners are flexible regarding charging.
Flex Charging and Rates Sadly, many customers prefer convenience over cost. So, at least some on peak charging is likely to occur. 80% will pay the $3 to get home at 4pm (~$300/mwh) Most customers would rather drive home in the afternoon, versus saving $3 to $10. Willingness to Pay Time of Day Rates
Further Customer Insights Several customer insights were generated in these studies including: Hybrid owners tend to be more flexible regarding charging. 10 miles is the upper limit on fueling station availability. 18-24 year olds strongly prefer fueling stations every 1 mile. Customers 36-45 and 56-65 do not readily accept fueling at home only, and will be looking for options, presumably due to work, school, kids, etc.
Further Customer Insights Low income customers strongly prefer fueling stations every 1 mile, along with women and younger customers. All customers would prefer to plug their vehicle into any outlet, regardless of current vehicle type. Few customers want to plug their vehicle into only one dedicated outlet. All customers, irrespective of current vehicle type owned, would prefer to have automated charging, or have the vehicle plug itself in, except for minivan owners.
Further Customer Insights Reduced insurance rates and free battery charging at night are acceptable incentives regardless of age. Customers who make less than $29,000 prefer free recharging in public lots, while all other income brackets prefer reduced insurance rates. Minivan owners prefer to be able to use backup power from the vehicles, for their house more than other groups, but it is not their strongest preference. Really, they don t care about any of these incentives, since they don t want EVs to begin with.
Spatial Forecast Methodology The spatial forecasting information mentioned in these studies is based on the Integral Analytics LoadSEER software. We invite you to our booth #415 for more information about LoadSEER and to meet with our LoadSEER product development director Scott Smith. Spatial Electric Expansion & Risk
For more information contact Integral Analytics, Inc. 513-762-7621 Or email kenneth.skinner@integralanalytics.com www.integralanalytics.com