Appendix D: Long-Distance Travel Frequency from the SHRP2 NDS
FHWA’s second Strategic Highway Research Program’s Naturalistic Driving Study (SHRP2 NDS) data set is used to evaluate LDT frequency. As a longitudinal safety study, SHRP2 NDS tracked 3,710 vehicles (exclusively non-PEVs, consisting mostly of conventional gasoline vehicles and a small number of HEVs) across six cities for up to 3 years to record crash, near-crash, and baseline driving events, resulting in data on approximately 5.4 million driving trips (Campbell 2012, Hallmark et al. 2013, Dingus et al. 2015). To estimate LDT frequency, the 3,352 vehicles that were tracked for at least one year or more were considered. Trip distances were aggregated on a daily basis using the unique vehicle ID provided in the data set. The resulting SHRP2 NDS sample was compared to the 2009 NHTS (FHWA 2017a) to assess its overall representativeness of driving in the United States. SHRP2 NDS shows a higher share of vehicles classified as “car” compared to the 2009 NHTS (71% in SHRP2 NDS versus 49% in the 2009 NHTS), and the average vehicle age across all classifications in the SHRP2 NDS data set is 10.84 years, whereas an average age of 9.33 years is reported in the 2009 NHTS. Figure D-1 and Figure D-2 compare daily and annual VMT distributions between the 2009 NHTS and the SHRP2 NDS. Daily VMT and long-distance driving patterns are mostly consistent between the two data sets; nearly 96% of SHRP2 NDS vehicle-days have VMT less than 100 miles, compared with 94% for the 2009 NHTS. However, a higher share of SHRP2 NDS vehicles had annualized VMT of 4,000 to 10,000 miles compared with vehicles from the 2009 NHTS.
Figure D-1. Cumulative distribution of daily VMT in 2009 NHTS and SHRP2 NDS.
Figure D-2. Annual VMT distribution in 2009 NHTS and SHRP2 NDS.
Of the 3,352 vehicles in the SHRP2 NDS, 1,669 show a daily VMT of above 100 miles for at least 1 day during the tracked period. The average SHRP2 NDS vehicle traveled 100 miles or more 6 days/year (2 days/year for daily VMT greater than 200 miles and 1 day/year with daily VMT greater than 300 miles). These results imply that a large degree of heterogeneity exists between LDT frequencies in the SHRP2 NDS data set. On the one hand, 50% of the sample never drove more than 100 miles on any travel day, implying a potentially good match with short-range BEVs with approximately 100 miles of single-charge driving range. On the other hand, a large segment of the SHRP2 NDS vehicles exhibits long-distance travel days multiple times per month: travels that could only be accomplished in a BEV with a large single-charge driving range and DCFC support on corridors.
Longitudinal surveys, such as the SHRP2 NDS, are time consuming and expensive to administer. Instead, annual VMT is much easier to estimate using odometer readings from traditional single-day travel surveys (such as the 2009 NHTS), and can serve as a reasonable proxy for LDT frequency. Figure D-3 shows the correlation between annual VMT and LDT frequency. Vehicles with higher annual VMT have generally higher LDT frequency (days/year) for all daily distance thresholds selected (100, 200, and 300 miles).
Figure D-3. Impact of annualized VMT (AVMT) on average long-distance travel frequency from SHRP2 NDS.
As mentioned previously, a significant degree of heterogeneity exists in driving data sets such as SHRP2 NDS. Figure D-4 reflects this fact by displaying percentile curves for frequency of LDT days with at least 200 miles of driving versus annual VMT (note that percentile curves below 60 are not shown as 50% of the sample exhibited no LDT). Taking the 70th percentile curve as an example, the data show that 70% of the SHRP2 NDS vehicles drove more than 200 miles on 3 or fewer days per year.
Figure D-4. Impact of annualized VMT (AVMT) on long-distance travel frequency from SHRP2 NDS (percentile bins, frequency of 200+ mile days).
Evidence from the SHRP2 NDS data set suggests that a large segment of drivers routinely use their existing conventional vehicles for long-distance travel (e.g., 50% of SHRP2 NDS vehicles drove more than 100 miles one day per month on average) and would presumably require long-distance single-charge ranges and DCFC support along corridors to consider adopting a BEV as a fully capable replacement for their existing vehicle. And even the remaining class of drivers that never makes a LDT over the course of a year might exhibit similar adoption requirements based on their perceived need for long-distance driving.
