You can use this DOI to cite this repository:

(zenodo link here…)

This repository contains the data and code to reproduce results from our study titled “Grid-Integration of Electric Vehicles: Consumer Preferences for Supplier Managed Charging and Vehicle-to-Grid Programs”. All code is written using the R programming language.

It also contains the source code of a Smart Charging Enrollment Simulator.

For a brief introduction of this study, proceed to this link. It also contains the survey manuscript, enrollment simulator, and related poster and slides.

Authors:

Name	ORCID
Pingfan Hu	0009-0001-4877-4844
Brian Tarroja, Ph.D., P.E.	0000-0002-7736-8642
Matthew Dean, Ph.D.	0000-0002-0346-4316
Kate Forrest, Ph.D.	0000-0001-6375-1299
Eric Hittinger, Ph.D.	0000-0001-8439-4016
Alan Jenn, Ph.D.	0000-0003-4232-0697
John Paul Helveston, Ph.D.	0000-0002-2657-9191

DOI: DOI here…

Abstract: As power systems transition toward renewable energy sources, integrating Battery Electric Vehicles (BEVs) into grid operations becomes crucial for optimizing energy resources and enabling multi-energy sector coupling. Smart charging programs can help utilities balance electricity supply and demand while facilitating renewable energy integration, but their success depends on BEV owner participation. This study examines two grid-integration strategies: Supplier-Managed Charging (SMC), which enables utilities to optimize charging timing and duration, and Vehicle-to-Grid (V2G), which transforms BEVs into distributed energy storage resources that can support grid stability. Using a discrete choice experiment with 1,356 current BEV owners, recruited through social media advertisements and survey panels, we quantify how different program attributes influence enrollment decisions. Our multinomial logit models reveal distinct preference patterns that inform program design: SMC participants predominantly value operational flexibility and recurring payments, while V2G participants show stronger preferences for monetary incentives, indicating willingness to provide grid services for compensation. Through simulation analysis, we identify program ``attribute equivalencies’’ that quantify the changes needed in program attributes to achieve equivalent enrollment levels, offering utilities guidance for designing grid-integration programs that balance system needs with consumer preferences. These findings offer insights for developing market mechanisms and policy frameworks that can accelerate the integration of BEVs into future energy systems while supporting power system decarbonization.

Replication Steps

1. Install R and RStudio.
2. Double click on the smart-charging.Rproj file to launch RStudio.
3. Open codes/run.R, select all scripts using Ctrl/cmd + A.
4. Run the code scripts using Ctrl/cmd + Enter, or by pushing the “Run” button on the top right corner.

File Organization

codes

Contains the source codes of survey design and analysis. To reperform the analysis, simply run the run.R file.

file	description
run.R	A single file to reproduce all analysis.
analysis/0_source_scripts.R	Triggers all analysis codes, called by run.R.
analysis/1_clean_data.R	Cleans the raw survey data and saves into the results/ directory.
analysis/2_summary_statistics_plots.R	Generates the summarized plots of the raw survey data and saves into the results/1_summary/ directory.
analysis/3_mnl_pref_model.R	Constructs the multinomial logit model with user preference space and saves into the results/2_models/ directory.
analysis/4_mnl_enrollment_sensitivity.R	Simulates the user enrollment sensitivity of smart charging programs and saves into the results/3_enrollment_sensitivity/ directory.
analysis/5_mnl_scenario_analysis.R	Simulates the scenario analysis and saves into the results/4_scenario_analysis/ directory.
design/1_power_analysis_smc.R	Performs power analysis conjoint survey questions to help determine the expected number of participants.
design/2_make_choice_questions.R	Generates conjoint choice questions and save csv files into the data/ directory.
design/3_make_smc_battery_figs.R	Generates battery condition figures for the SMC smart charging programs and saves into the figs/ directory.
design/4_make_v2g_battery_figs.R	Generates battery condition figures for the V2G smart charging programs and saves into the figs/ directory.
design/5_make_cars_df.R	Generates car makes and models and save csv files into the data/ directory.

Comments:

1. Simply run the run.R file will generate all analysis codes and save the results in the results/ directory.
2. The run.R file triggers the analysis/0_source_scripts.R file, and then further triggers the five other R files under analysis/.
3. The five R files under analysis/ cannot be executed alone unless the run.R file is executed.
4. The files in the design/ directory are used for designing the survey and don’t need to be executed anymore.
5. The design/2_make_choice_questions.R file contains randomization. If you run this file again, the resulted smc_questions.csv and v2g_questions.csv files will be different, which will not match with the questions used in our survey recruitment. If you want to see the performance of this file, better save the original question csv files somewhere else.

data

Contains miscellaneous files that are used in survey design and analysis.

file	description
bev_makes.csv	Brands that sell BEVs in the U.S. market.
bevs.csv	BEV models with makes, models, and ranges.
car_makes.csv	All car makes in the U.S. market.
car_models.csv	All car models in the U.S. market.
cars.csv	Raw car data scrapped from cars.com. All other car data files are generated from this file, with further manual modification.
smc_questions.csv	SMC program conjoint question file generated from codes/design/2_make_choice_questions.R.
smc_scenarios.csv	SMC program scenarios used for scenario analysis, hand generated.
v2g_questions.csv	V2G program conjoint question file generated from codes/design/2_make_choice_questions.R.
v2g_scenarios.csv	V2g program scenarios used for scenario analysis, hand generated.
Six Image Files	Images used as banners and educational visual aids in survey. The banner images are generated using DALL-E, and educational images are made using Affinity Designer 2.

figs

Contains all battery conditional figures of SMC and V2G smart charging programs, generated by codes/design/3_make_smc_battery_figs.R and codes/design/4_make_v2g_battery_figs.R.

results

Contains all data files generated by codes/run.R, which triggers the R code scripts in the codes/analysis/ directory.

1. The csv files and .RData files are the processed survey results. We generated these 2 file types to comply with the needs of data analysis in R codes.
2. The four folders contain the results of different code script files in code/analysis/.

simulator

This is a standalone folder that contains the Shiny App source codes of the Smart Charging Enrollment Simulator.

survey

Contains the raw survey responses. We recruited our survey on these two sources:

1. Meta - the well-know social media, including Facebook, Instagram and Messenger.
2. Dynata - a largely used survey panel.

Survey files are in 7 pieces (for both sources) due to the limitation of formr, the survey platform we used for this study.

tables

The .tex formatted tables used in our paper.

file	description
table_demographics.tex	Demographic results table, generated from codes/analysis/2_summary_statistics_plots.R.
table_sample_size.tex	Sample size of the survey, containing both Meta and Dynata.
table_smc_attr.tex	SMC program attributes, helps to design the survey.
table_smc_equiv.tex	SMC program equivalency table, generated from the “Equivalency Table” section in codes/analysis/4_mnl_enrollment_sensitivity.R.
table_v2g_attr.tex	V2G program attributes, helps to design the survey.
table_v2g_equiv.tex	V2G program equivalency table, generated from the “Equivalency Table” section in codes/analysis/4_mnl_enrollment_sensitivity.R.
table_vehicles.tex	Vehicle ownership results table, generated from codes/analysis/2_summary_statistics_plots.R.