background-image: url("images/blue.jpg") background-size: cover class: inverse <br><br><br><br> ## Quantifying Plug-in Electric Vehicle Mileage **.white[John Paul Helveston]**, George Washington University<br> Eliese Ottinger, George Washington University<br> Lujin Zhao, George Washington University<br> January 11, 2024 --- class: middle background-color: #fff ### .center[.font90[Electrifying the passenger vehicle fleet is a critical climate goal]] <center> <img src="images/ghg_emissions.png" width=850> </center> --- # .center[Two Studies, One Dataset] .leftcol[ ### Measuring Electric Vehicle<br>**Mileage** in the United States Lujin Zhao (Ph.D. Student)<br> Eliese Ottinger (Undergraduate RA)<br> **Status**: Published in _Joule_ ] .rightcol[ ### Measuring Electric Vehicle<br>**Resale Value** in the United States Laura Roberson (Ph.D. Student)<br> Saurav Pantha (Former Graduate RA)<br> **Status**: Submitting soon ] --- background-color: #fff ### .center[**Data**: ~13M used vehicle listings from 60k dealerships (2016 - 2022)] <center> <img src="images/table-data-summary.png" width=850> </center> --- class: middle, inverse ## Going the Distance:<br>Quantifying Electric Vehicle Mileage in the United States Lujin Zhao (Ph.D. Student)<br> Eliese Ottinger (Undergraduate RA)<br> John Paul Helveston, Ph.D. The George Washington University --- # .center[We really need to understand PEV usage] <br> -- ### - PEV emissions reduction benefit **depends on vehicle usage**<br>.font80[[Jenn (2020)](https://www.nature.com/articles/s41560-020-0632-7)] -- ### - Modelers typically assume **BEV miles = CV miles** -- ### - Revenue from proposed mileage tax **depends on vehicle usage**<br>.font80[[Metcalf et al. (2022)](https://doi.org/10.1086/722672); [Zhao and Mattauch (2022)](https://doi.org/10.1016/j.jeem.2022.102747); [Davis and Sallee (2020)](https://doi.org/10.1086/706793)] -- ### - PEV adoption depends on **how well PEVs substitute for CVs**<br>.font80[[Xing et al. (2021)](https://doi.org/10.1016/j.jeem.2021.102432)] --- background-image: url("images/table-lit1.png") background-size: cover ### .center[Conflicting prior results on BEV mileage] --- background-image: url("images/table-lit2.png") background-size: cover ### .center[.blue[Conflicting prior results on BEV mileage]] --- background-image: url("images/table-lit3.png") background-size: cover ### .center[.red[Inconsistent data quality in prior studies]] --- class: center background-color: #fff ## BEVs are driven significantly less than other powertrains <center> <img src="images/mileage-all.png" width=100%> </center> --- background-color: #fff class: center ### Teslas driven more than non-Tesla BEVs (but not as much as CVs) <center> <img src="images/mileage-bev.png" width=800> </center> --- class: center background-color: #fff ## BEVs are driven significantly less than other powertrains <center> <img src="images/mileage-all.png" width=100%> </center> `\(mileage = \beta_0 + \beta_1age+\beta_2{age*powertrain}+\beta_3{age}*cents\_p\_mile+\epsilon_i\)` --- background-color: #fff .leftcol80[ <center> <img src="images/table-reg-pooled.png" width=100%> </center> ] --- background-color: #fff .leftcol80[ <center> <img src="images/table-reg-pooled1.png" width=100%> </center> ] .rightcol20[ ### .red[BEVs driven 4,500 miles less than CVs on average] ] --- background-color: #fff .leftcol80[ <center> <img src="images/table-reg-pooled2.png" width=100%> </center> ] .rightcol20[ ### BEVs driven 4,500 miles less than CVs on average ### .red[Non-Tesla BEVs:<br>-5,400 miles] ### .blue[Tesla:<br>-2,800 miles] ] --- background-color: #fff <center> Cars <img src="images/mileage-scatter-cars.png" width=100%> </center> <br> <center> SUVs <img src="images/mileage-scatter-suvs.png" width=100%> </center> ??? BEVs:<br>~7k mi/yr All others:<br>~11k - 12k mi/yr Far less variability in BEV mileage than CV mileage --- background-color: #fff .leftcol65[ <center> <img src="images/table-reg-powertrain-zoom.png" width=100%> </center> ] --- background-color: #fff .leftcol65[ <center> <img src="images/table-reg-powertrain-zoom1.png" width=100%> </center> ] .rightcol35[ **Non-linear range effect**: +10 mi range: .red[Low range (<100 mi):<br>+640 mi/yr] .red[Mid range (100-200 mi):<br>+420 mi/yr] .red[High range (\>200 mi):<br>+90 mi/yr] ] --- background-color: #fff .leftcol65[ <center> <img src="images/table-reg-powertrain-zoom2.png" width=100%> </center> ] .rightcol35[ **Non-linear range effect**: +10 mi range: .red[Low range (<100 mi):<br>+640 mi/yr] .red[Mid range (100-200 mi):<br>+420 mi/yr] .red[High range (\>200 mi):<br>+90 mi/yr] <br> .blue[Tesla effect isn't just from range] ] --- background-color: #fff ### .center[BEV mileage less sensitive to operating cost than CV mileage] <center> <img src="images/table-reg-powertrain-cpm.png" width=1000> </center> 1 cent increase in operating cost: BEV: -69 mi/yr<br> CV: -136 mi/yr<br> --- background-color: #fff ### .center[BEV mileage less sensitive to operating cost than CV mileage] <center> <img src="images/table-reg-powertrain-cpm.png" width=1000> </center> .leftcol70[ <center> <img src="images/cost-per-mile.png" width=600> </center> ] .rightcol30[ 1 cent increase in operating cost: BEV: -69 mi/yr<br> CV: -136 mi/yr<br> .blue[BEVs have much lower operating costs] ] --- name: why-vmt class: center, middle, inverse # Why low BEV mileage? --- # Why low BEV mileage? .leftcol40[ ### **Intra-household substitution?** Maybe current adopters have multiple cars? Perhaps, but NHTS data suggests **secondary cars are only driven 1,000 - 2,000 miles less per year**. ] .rightcol60[ <center> <img src="images/table-nhts.png" width=100%> </center> ] --- # Why low BEV mileage? .leftcol40[ ### **Maybe newer models are driven more?** Some (limited) evidence this may be the case (MY 2019: only 10,484 listings, max age of 3.2 years old) ] .rightcol60[ <center> <img src="images/table-my-zoom.png" width=100%> </center> ] --- # Why low BEV mileage? .leftcol40[ ### **Selection bias?** Maybe current adopters just have lower driving needs? No way for us to measure this, but it seems very plausible ] .rightcol60[ <center> <img src="images/plane.jpg" width=100%> </center> ] --- ## Key takeaways -- ### - BEVs are driven significantly less than other powertrains:<br>.red[Non-Tesla BEVs: -5,400 miles]; .blue[Tesla: -2,800 miles] -- ### - Far less variability in BEV mileage than CV mileage<br>(BEVs only substituting for lower-mileage CV usage) -- ### - BEV mileage less sensitive to operating cost than CV mileage -- ### - Can't say **why** low BEV mileage, but still relevant for policy --- class: inverse background-image: url("images/blue.jpg") background-size: cover <br><br><br><br><br><br><br><br><br><br> # Thanks! ### Slides: ### https://slides.jhelvy.com/2024-ev-policy-council-vmt/ .footer-large[.white[.right[ @jhelvy@fediscience.org
<br> @JohnHelveston
<br> @jhelvy
<br> jhelvy.com
<br> jph@gwu.edu
]]]