Sustainable Transportation,
Sustainable Research:
Consumer Adoption and Grid Peak-Shaving
of Battery Electric Vehicle Smart Charging,
Inspiring the surveydown Survey Platform

Pingfan Hu
George Washington University

About Me

• Pingfan Hu, PhD Candidate at George Washington University, advised by Dr. John Helveston.

• Research interests:

  1. Data Science in sustainable transportation and human behavior
  2. Open-Source Software in tools for survey research
  3. Agentic AI Engineering around Claude Code

• For more information, visit pingfan.org

EV sales in US reaching ~10% of sales

Source: Argonne National Lab, www.anl.gov/ev-facts/model-sales

Introduction

• Unmanaged BEV charging is becoming a problem to the grid.
• Managed charging is cheaper and smooths out the grid load.
• Smart charging: Supplier-Managed Charging (SMC) and Vehicle-to-Grid (V2G).

Unmanaged BEV charging is becoming a problem to the grid due to the spike. That’s why we need managed charging to smooth out the grid load as well as keep the price low.

Smart charging is a type of managed charging, which includes Supplier-Managed Charging, shortened as SMC, and Vehicle-to-Grid, shortened as V2G.

SMC - Supplier Managed Charging

• SMC smooths out the BEV charging spike (called “Peak-shaving”).
• Electricity demand is controlled below the capacity threshold.
• It saves money and reduces pollution.

For SMC, Supplier-Managed Charging, it smooths out the overnight BEV charging demand.

As you can see, with unmanaged charging, the BEV load easily exceeds the capacity threshold.

SMC - Supplier Managed Charging

• SMC smooths out the BEV charging spike (called “Peak-shaving”).
• Electricity demand is controlled below the capacity threshold.
• It saves money and reduces pollution.

But with SMC, the electricity demand is controlled below the threshold. It reduces the peak demand and saves money.

V2G - Vehicle-to-Grid

V2G is short for Vehicle-to-Grid. It is another type of smart charging that allows BEV owners to sell electricity back to the grid. With V2G, BEVs become potential sources of electricity.

Smart charging effectiveness depends on
enrollment and peak-shaving.

Three Inter-connected Studies

Survey Research & Modeling

Battery Electric Vehicle Smart Charging Adoption

Simulation Research

Grid Peak-Shaving Quantification

Software Development

The surveydown Survey Platform

My dissertation consists of three inter-connected studies: a survey study on smart charging adoption, a grid simulation quantifying peak-shaving, and the surveydown survey platform. After the three studies, I’ll also briefly summarize my teaching experience.

Study 1
Battery Electric Vehicle Smart Charging Adoption

Published in Environmental Research Letters:

Hu, Pingfan, Tarroja, B., Dean, M., Forrest, K., Hittinger, E., Jenn, A. & Helveston, J.P. (2025) “Measuring electric vehicle owners’ willingness to participate in smart charging programs” Environmental Research Letters. https://doi.org/10.1088/1748-9326/ae2597

Prior studies have few EV owners in sample

1. A survey-based study by Wong et al. (2023) examined incentives’ effect on EV owners’ acceptance, but EV ownership in sample was only 19%.
2. A survey-based study by Philip & Whitehead (2024) found range anxiety matters, but EV ownership in sample was only 1.28%.
3. A survey-based study by Huang et al. (2021) indicates the importance of fast charging, but the sample size was only 157.

We need high BEV ownership & a large sample size.

There have been a few studies that looked at people’s willingness to participate in smart charging programs, but they all have small sample sizes and low EV ownership. Therefore, we aimed to conduct a large-scale survey with high BEV ownership to study BEV owners’ preferences.

Research Questions

1. Sensitivity: How do changes in smart charging program features influence BEV owners’ willingness to opt in?

2. Enrollment Rate: Under what combinations of features will BEV owners be more willing to opt in to smart charging programs?

Approach

---

1. Conjoint survey to collect BEV owners’ willingness.
2. Mixed logit model for utility simulations.

Here we have 2 research questions. Firstly, we want to know the sensitivity of each smart charging program feature; secondly, we want to know under what conditions BEV owners will be more willing to opt in to the programs.

In order to answer these questions, we designed a conjoint survey to collect BEV owners’ willingness to participate. Then we used mixed logit models to simulate the utility of each program feature.

Survey Design

Conjoint Questions

1. Monetary Incentives
2. Charging Limitations
3. Flexibility

Demographic Questions

1. BEV Ownership
2. Personal Info
3. Household Info

Conjoint Question Explained

A Sample Conjoint Question

1. Provide respondents with different sets of attributes.
2. Observe choices across random sets.
3. Estimate utility of each attribute.

But first, I want to explain what a conjoint question is. If I care about apples, I would want to know how people value type vs price vs freshness. In this apple example, you can only choose one set of the attributes, so if you pick one option, you pick all the attributes in it. In short, conjoint is about forcing people to make trade-offs.

SMC Programs

Attributes

No.	Attributes	Range
1	Enrollment Cash	$50 to $300
2	Monthly Cash	$2 to $20
3	Monthly Override	0 to 5
4	Min Battery	20% to 40%
5	Guaranteed Battery	60% to 80%

Sample Program

Attributes	Values
Enrollment Cash	$300
Monthly Cash	$20
Override Allowance	5

(Range determined by stated vehicle they own)

Let’s move on to the conjoint question for SMC programs. Again SMC is short for Supplier-Managed Charging. These are the 5 attributes of SMC. You can see there are cash, override, and battery thresholds, and they all have reasonable ranges. For example, for enrollment cash we have $50 to $300.

Here is a sample SMC program. You can see the one-time enrollment cash, recurring monthly cash, and how many times per month the charging can be shifted back to normal.

The battery thresholds here are well-designed to provide an immediate visual aid of the minimum battery and guaranteed battery. We also used mileage instead of percentage to make it more intuitive. The mileage will be shown according to the real values of each participant.

Sample SMC Question

There are 6 SMC conjoint questions. In each question, participants are provided with 2 valid options and a “Not Interested” option. It’s important to have this “Not Interested” option, because it is a validation of willingness. When participants choose from these 2 valid options, they are sure to be interested in what they choose. Otherwise, they will choose “Not Interested”.

The V2G programs follow the same design, with occurrence-based attributes instead of override-based ones, and 6 V2G conjoint questions in parallel.

Survey Fielding - 1,356 in Total

Meta Ads: Voluntary participants

• 803 responses
• March to July in 2024

Dynata Recruitment: Paid survey

• 553 responses
• September to November in 2024

We collected a total of 1,356 complete, filtered responses from Meta and Dynata throughout 2024, and we made sure that all participants are real BEV owners.

Survey Question - Car Ownership

Survey Results - Top 10 BEV

This is a showcase of the top 10 BEV models from our survey results. You can see that Tesla is the most popular brand, followed by Ford, Chevrolet, and some other popular brands. These results are also representative of the BEV market in the United States.

Survey Results - Demographics

Other than the BEV models, there are also some important demographic information collected in the survey. You can see most families have 2 cars; most owners charge at home for more than 8 hours; and most of them care about climate very much. And more importantly, we have 100% BEV ownership.

Survey Results - Willingness to Participate

Mixed Logit Models

\[ \small \begin{align*} u_j = v_j + \epsilon_j = \beta' x + \epsilon_j \qquad P_j = \frac{e^{v_j}}{\sum_{k=1}^{J} e^{v_k}} \end{align*} \]

Utility estimated using maximum likelihood estimation (MLE).

SMC Estimates

V2G Estimates

Without compensation, users will not participate.

The willingness to participate is quantified with the mixed logit model, shortened as MXL. The utility of each option is a linear combination of the attributes, which is the \(v_j\) here, plus an error term, shown as \(\epsilon_j\). The probability of choosing an option is the exponential of its utility divided by the sum of exponentials of all utilities.

Here are the coefficient estimate plots of both programs. As you can see, the “No Choice” has the highest coefficient, which simply means that without compensation, users will not participate in the programs.

However, coefficients alone are not enough to interpret the willingness to participate. A more intuitive approach is the sensitivity, which means by how much the enrollment rate will change when a feature value changes.

Enrollment Sensitivity

Baseline Simulation

Choice between “None” and this program:

Attributes	Values
Enrollment Cash	$0
Monthly Cash	$0 - $20
Monthly Override	0

Sensitivity Plot

This is one example of the sensitivity plots. As the enrollment cash increases, the user enrollment increases as well. The solid part is from the results, and the dotted part is the extrapolation with the model.

Enrollment Sensitivity

1. Steeper slope indicates higher sensitivity.
2. Diminishing returns exist.

And here we have a full demonstration of all sensitivity plots. Steeper slope indicates higher sensitivity. Also, as the feature values become higher, we see diminishing returns. But the solid parts are basically linear.

SMC Scenario Analysis

1. Flexibility is highly valued.
2. Recurring incentives are more important than one-time ones.
3. Payment alone is not enough.

So, after analysis of all these individual features, we can combine them into programs. We call them “scenarios”. I want to emphasize that the purpose of doing these evaluations is to wisely allocate the resources of the grid. Therefore, we need to maximize the enrollment of each program with the least effort.

Here for SMC we have 3 scenarios. They are flexibility, recurring cash, and one-time cash.

As we can see, for SMC, flexibility is highly valued, and recurring incentives are more important than one-time. So due to the range anxiety, payment alone is not enough. There should be a combination of flexibility and monetary incentives.

Smart Charging Enrollment Simulator

Here we’ve made a smart charging enrollment simulator with Shiny. We can toggle the feature values of each attribute, and the app shows the anticipated enrollment of each program.

1. More enrollment increase in SMC: Guaranteed threshold from 60% to 100%.
2. More enrollment increase in V2G: Recurring cash from $2 to $20.

Study 2
Quantifying the Peak-Shaving Potential and Cost Efficiency of BEV Supplier-Managed Charging

Submitted to Environmental Research: Infrastructure and Sustainability

Manuscript ERIS-100972, awaiting review

Recall the SMC Peak-shaving

• SMC smooths out the BEV charging spike (called “Peak-shaving”).
• Electricity demand is controlled below the capacity threshold.
• It saves money and reduces pollution.

Quick recall from Study 1: SMC shifts BEV charging away from the evening peak, so the total demand stays below the capacity threshold.

Prior studies assume enrollment is high and free

1. Most grid studies assume optimal or near-100% participation (Tarroja et al., 2015), but real pilots enroll far fewer.
2. Enrollment is not free: utilities must pay incentives (Bailey et al., 2023), yet costs are rarely counted.
3. Few studies compare peak-shaving across grid regions with different load shapes.

We connect consumer enrollment with grid simulation.

The technical potential of managed charging is well documented, but most studies assume everyone participates for free. In reality, participation depends on incentives, and incentives cost money. That’s the gap this study fills: we bring the empirical enrollment model from Study 1 into a grid-scale simulation.

Research Questions

1. Peak-shaving: How much can SMC reduce peak demand in contrasting grid regions?

2. Cost efficiency: How much does peak shaving cost per unit, and is more enrollment always better?

Approach

---

1. Scenario-based simulation for CAISO and NYISO.
2. Enrollment model: incentives vs participation.

Two research questions. First, the physical side: how much peak demand can SMC actually shave, and does the answer depend on the grid region? Second, the economic side: what does that peak shaving cost, and is pushing enrollment higher always worth it?

To answer them, we run a scenario-based simulation of CAISO and NYISO, and we join it with the enrollment model from Study 1, which links incentive levels to participation rates.

18 GEA regions in the USA

The U.S. grid is divided into 18 Generation and Emissions Assessment regions, as defined by NREL’s Cambium dataset. For the grid simulation, we zoom into two contrasting regions: CAISO in California and NYISO in New York.

	CAISO	NYISO
Single Families	7.69 million	3.16 million
Different Energy Sources	Solar	Hydro
Net load shape	“Duck curve”	Flat profile
Seasonal effect	Minimal	Strong in winter (heating)

Both lead in BEV adoption, with contrasting grid structures.

We picked California and New York because they’re both BEV leaders but structurally opposite grids. CAISO has the solar duck curve with a steep evening ramp, and it behaves essentially the same across all four seasons. NYISO is flatter, relies more on dispatchable hydro, and is season-sensitive: winter overnight heating keeps the nighttime load elevated, which matters later in the results. This contrast lets us see how grid context shapes SMC performance.

Data Sources

Source	Provides	Role in Simulation
2022 NHTS	Real trip records nationwide	Charge time & energy needs
2025 GridStatus.io	Net load time series	Regional grid baselines
2025 U.S. Census	Single-family counts	Scale to total fleet energy
Study 1	Enrollment curve	Cost efficiency analysis

Four data sources: NHTS trips define when vehicles travel and how much energy they need; GridStatus.io gives us actual 2025 net load for both regions; the Census scales our 10,000 simulated vehicles up to regional fleet sizes; and crucially, the discrete choice experiment from Study 1 translates incentive dollars into enrollment rates. Net load means total consumption minus variable renewables, so it’s the demand that dispatchable generation must serve.

Simulation Overview

Stage 1: Peak-shaving model sweeps thousands of scenarios.
Stage 2: Cost efficiency joins results with Study 1 enrollment curves.

The framework has two stages. Stage one is the peak-shaving model: three data sources feed a simulation that sweeps TOU window configurations, adoption rates, date ranges, and enrollment levels. Stage two takes the best-performing outcomes and joins them with the consumer enrollment curve from Study 1 to compute cost efficiency frontiers.

Peak Shaving: Annual Average

CAISO		NYISO
Peak shave	3.7 GW (30.5 → 26.8)	Peak shave	1.6 GW (21.2 → 19.6)
Percentage	12.1%	Percentage	7.5%

Now we move from the method to the results. On an annual average day with full enrollment, SMC shaves 3.7 gigawatts off CAISO’s evening peak, about 12%, and 1.6 gigawatts in NYISO, about 7.5%. CAISO does better both because it has more households and because its deep overnight valley absorbs the deferred charging.

Enrollment Curve from Study 1

Mixed logit prediction based on ~1300 real BEV owners.

This is the bridge between the two studies. The mixed logit model from Study 1, holding non-monetary attributes fixed, gives the enrollment rate as a function of the monthly incentive. About 30% would participate with no payment at all, $20 a month gets you to about 60%, but full enrollment would take $85 per household per month. The curve is concave, meaning every additional dollar buys less enrollment, and since utilities pay a uniform incentive to everyone, total program cost grows non-linearly.

Peak from BEV charging decreases with higher enrollment

The final step joins the peak-shaving results with the enrollment curve: as enrollment rises, the peak from BEV charging falls in both regions. Each dot along a curve is a 10% enrollment step, and program cost is the incentive times the number of enrolled households. The dots spread out along the cost axis at higher enrollment: attracting marginal participants requires raising the payment for all existing enrollees, so marginal costs rise sharply.

Cost Efficiency Frontier

Cost efficiency is highest at moderate enrollment (30–50%).

Here’s the central result. Each dot is a 10% enrollment step, and the dots stretch out along the cost axis: the last increments of enrollment are extremely expensive. Moderate enrollment, 30 to 50 percent, captures most of the achievable peak reduction at a fraction of the cost. And on about 15% of CAISO days, pushing enrollment too high actually reconstitutes a new overnight peak.

Conclusions

1. The cost-effectiveness of peak shaving requires: moderate enrollment (30-50%), and seasonal effect consideration.
2. Peak-shaving potential is: 14.4% for CAISO, and 7.9% for NYISO.
3. SMC has co-benefits beyond monetary value: fewer emissions, longer battery life, and lower storage capacity requirements.
Limitations	a. Stated preference being optimistic. b. Only considered single-family homes. c. Grid conditions based on today’s data.

I close with three takeaways and the limitations.

First, the conclusion: SMC delivers meaningful evening peak reductions in both regions, but returns to enrollment diminish sharply: 30 to 50% participation captures most of the achievable reduction at a fraction of the cost, and pushing too high can even create a new overnight peak. So utilities should design SMC programs around moderate enrollment targets, not maximum participation.

Second, the magnitude: at best, SMC shaves 14.4% off CAISO’s evening peak and 7.9% off NYISO’s. CAISO does better because its deep overnight valley absorbs the deferred charging.

Third, co-benefits. Shifting charging into the overnight valley reduces marginal emissions, the 20-to-80% SOC window extends battery life, and an enrolled fleet acts as dispatchable capacity that defers new transmission and generation investment. None of these are monetized in our efficiency frontiers, so the frontiers are a lower bound on the total value of SMC.

Finally, the limitations. The enrollment curve comes from stated preference, so it may overstate real uptake and our cost estimates may be a lower bound. We model residential Level 2 charging in single-family homes only, which is less relevant for NYISO, where New York City is largely multi-family. And results are calibrated to 2025 net load, so they shouldn’t be extrapolated to future high-solar, high-storage grids.

Study 3
The surveydown Survey Platform

Published in PLoS ONE:

Hu, Pingfan, Bunea, Bogdan, & Helveston, J. P. (2025). “surveydown: An open-source, markdown-based platform for programmable and reproducible surveys” PLOS ONE, 20(8), e0331002. https://doi.org/10.1371/journal.pone.0331002

Typical Web Survey

Google Form - GUI Interface

Limitations

❌ Reproducibility

❌ Version control

❌ Limited features

❌ Open source

Qualtrics - Commercial Platform

Expensive!

Why not make surveys from code?

Limitations
Features

✅ Reproducibility

✅ Version control

✅ Lots of features

✅ Open source

✅ Free!

Introducing surveydown!

R package that renders Quarto files into surveys

Quarto is a publishing system

Original qmd file

Markdown + R code chunks

---
format: html
title: "HTML Page with R Code"
---

# Hello, World!

This is a simple **HTML** page with *R* code.

```{r}
library(ggplot2)

df <- data.frame(x = rnorm(100))
ggplot(df, aes(x = x)) +
  geom_histogram()
```

Rendered HTML

Original qmd file

Markdown + Python code chunks

---
format: pdf
title: "PDF File with Python Code"
---

# Hello, World!

This is a simple **PDF** file with *Python* code.

```{python}
import matplotlib.pyplot as plt
import numpy as np

data = np.random.normal(0, 1, 100)
plt.hist(data)
plt.show()
```

Rendered PDF

survey.qmd

---
format: html
echo: false
warning: false
---

```{r}
library(surveydown)
```

--- welcome

# Welcome to `surveydown`!

```{r}
sd_question(
  type  = "mc",
  id    = "has_fav_hero",
  label = "Do you have a favorite super hero?",
  option = c(
    "Yes" = "yes",
    "No"  = "no"
  )
)

sd_next()
```

--- some_other_page

Other content...

Rendered Survey

survey.qmd

---
format: html
echo: false
warning: false
---

```{r}
library(surveydown)
```

--- welcome

# Welcome to `surveydown`!

```{r}
sd_question(
  type  = "mc",
  id    = "has_fav_hero",
  label = "Do you have a favorite super hero?",
  option = c(
    "Yes" = "yes",
    "No"  = "no"
  )
)

sd_next()
```

--- some_other_page

Other content...

YAML header for a “clean” output

survey.qmd

---
format: html
echo: false
warning: false
---

```{r}
library(surveydown)
```

--- welcome

# Welcome to `surveydown`!

```{r}
sd_question(
  type  = "mc",
  id    = "has_fav_hero",
  label = "Do you have a favorite super hero?",
  option = c(
    "Yes" = "yes",
    "No"  = "no"
  )
)

sd_next()
```

--- some_other_page

Other content...

Load the surveydown Package

survey.qmd

---
format: html
echo: false
warning: false
---

```{r}
library(surveydown)
```

--- welcome

# Welcome to `surveydown`!

```{r}
sd_question(
  type  = "mc",
  id    = "has_fav_hero",
  label = "Do you have a favorite super hero?",
  option = c(
    "Yes" = "yes",
    "No"  = "no"
  )
)

sd_next()
```

--- some_other_page

Other content...

Use triple-dash marker (---)

to define survey pages

Quarto fences (:::) also OK:

::: {#welcome .sd-page}

Page content

:::

survey.qmd

---
format: html
echo: false
warning: false
---

```{r}
library(surveydown)
```

--- welcome

# Welcome to `surveydown`!

```{r}
sd_question(
  type  = "mc",
  id    = "has_fav_hero",
  label = "Do you have a favorite super hero?",
  option = c(
    "Yes" = "yes",
    "No"  = "no"
  )
)

sd_next()
```

--- some_other_page

Other content...

Page content

• Markdown for text, images, etc.
• sd_question() for survey questions
• sd_next() for page navigation

survey.qmd

---
format: html
echo: false
warning: false
---

```{r}
library(surveydown)
```

--- welcome

# Welcome to `surveydown`!

```{r}
sd_question(
  type  = "mc",
  id    = "has_fav_hero",
  label = "Do you have a favorite super hero?",
  option = c(
    "Yes" = "yes",
    "No"  = "no"
  )
)

sd_next()
```

--- some_other_page

Other content...

Rendered Survey

survey.qmd

---
format: html
echo: false
warning: false
---

```{r}
library(surveydown)
```

--- welcome

# Welcome to `surveydown`!

```{r}
sd_question(
  type  = "mc",
  id    = "has_fav_hero",
  label = "Do you have a favorite super hero?",
  option = c(
    "Yes" = "yes",
    "No"  = "no"
  )
)

sd_next()
```

--- some_other_page

Other content...

Rendered Survey

Quarto renders to static html pages.
Shiny turns them interactive.

A complete surveydown survey

survey.qmd

A Quarto doc defining the survey content (pages, text, images, questions, etc.).

app.R

An R script defining the survey Shiny app.

PostgreSQL for response data storage.

supabase.com

Technologies of surveydown

surveydown is feature-packed!

surveydown is highly flexible and customizable to user needs

Randomization: Conjoint question

Shiny compatibility: leaflet map

surveydown vs other platforms

Star Count on GitHub

The companion sdstudio package

The companion sdstudio package

Research Contributions

• Study 1: First large N survey study of BEV owner preference sensitivities to smart charging.
• Study 2: Quantifies peak-shaving potential and cost efficiency of SMC by integrating empirical enrollment with grid simulation.
• Study 3: surveydown, a free open-source survey platform for programmable, reproducible survey designs.

I start with survey research on smart charging, followed by a nation-wide grid simulation, and a survey software.

I hope these studies could benefit sustainable transportation, and make lasting contributions to future research.

Workshops & TA Experience

	Topic	Venue & Time
Workshops	Agentic Workflows with Claude Code	GWTAI, 2026
	surveydown: Open-source survey platform	Quant UX Con, 2025
Teaching Assistant	EMSE 4572/6572: Exploratory Data Analysis	GWU, Fall 2025
	EMSE 4571: Programming for Analytics	GWU, Spring 2025

Mentored students through substantial R programming and data analysis projects.

Beyond the research, a quick summary of my teaching: workshops all co-taught with my advisor — the agentic workflows workshop, plus the surveydown workshop at Quant UX Con.

And two semesters as a teaching assistant in GW’s data science courses: Exploratory Data Analysis and Programming for Analytics, where I mentored students through substantial R programming and data analysis projects.

Thanks for listening!

Pingfan Hu

PhD Candidate, Systems Engineering

pingfan.org

About

Blog

Researches

Works

Appendix - Study 1Smart Charging Adoption

Equivalencies of 5% Enrollment Increase

SMC

Attribute	Equivalence Value	Unit
Enrollment Cash	77.7	$
Monthly Cash	4.0	$
Override Days	2.5	Days
Minimum Threshold	65.5	%
Guaranteed Threshold	6.3	%

V2G

Attribute	Equivalence Value	Unit
Enrollment Cash	55.7	$
Occurrence Cash	2.9	$
Monthly Occurrence	1.9	Times
Lower Bound	11.7	%
Guaranteed Threshold	9.1	%

1. Smaller value indicates higher efficiency.
2. Monetary incentives are valued more in V2G than SMC.
3. Guaranteed thresholds are highly efficient, indicating range anxiety.
4. Attribute equivalencies can be used to inform incentive design.

V2G Scenario Analysis

1. Still, recurring incentives are more important than one-time ones.
2. But flexibility is not as important as in SMC.
3. Owners are willing to leverage BEV as a source of income.

For V2G programs, recurring incentives are also more important than one-time. But flexibility is compromised since owners are willing to use their BEVs as a source of income. So the V2G programs are more incentive-powered than SMC programs.

SMC Logit Model

\[ \begin{align*} u_j = \beta_1 x_j^{\text{enroll_cash}} + \beta_2 x_j^{\text{monthly_cash}} + \beta_3 \delta_j^{\text{override_allowed}} + \beta_4 x_j^{\text{num_overrides}} \notag \\ + \beta_5 x_j^{\text{min_threshold}} + \beta_6 x_j^{\text{guaranteed_threshold}} + \beta_7 \delta_j^{\text{no_choice}} + \epsilon_j \end{align*} \]

Attribute	Coef.	Est.	SE	Level	Unit
Enrollment Cash	β₁	0.0037	0.0002	50, 100, 200, 300	USD
Monthly Cash	β₂	0.0728	0.0031	2, 5, 10, 15, 20	USD
Override Days	β₃	0.1191	0.0140	0, 1, 3, 5	Days
Override Flag	β₄	0.4357	0.0654	Yes, No	-
Minimum Threshold	β₅	0.0044	0.0023	20, 30, 40	%
Guaranteed Threshold	β₆	0.0490	0.0028	60, 70, 80	%
No Choice	β₇	2.8984	0.2215	-	-

V2G Logit Model

\[ \begin{align*} u_j = \beta_1 x_j^{\text{enroll_cash}} + \beta_2 x_j^{\text{occur_cash}} + \beta_3 x_j^{\text{num_occurrences}} + \beta_4 x_j^{\text{lower_threshold}} \notag \\ + \beta_5 x_j^{\text{guaranteed_threshold}} + \beta_6 \delta_j^{\text{no_choice}} + \epsilon_j \end{align*} \]

Attribute	Coef.	Est.	SE	Level	Unit
Enrollment Cash	β₁	0.0051	0.0003	50, 100, 200, 300	USD
Occurrence Cash	β₂	0.0972	0.0045	2, 5, 10, 15, 20	USD
Monthly Occurrence	β₃	0.1595	0.0262	1, 2, 3, 4	Times
Lower Threshold	β₄	0.0263	0.0036	20, 30, 40	%
Guaranteed Threshold	β₅	0.0368	0.0037	60, 70, 80	%
No Choice	β₆	2.4283	0.1964	-	-

Appendix - Study 2Grid Peak-Shaving Quantification

Simulation Assumptions

Vehicle & Charging
BEV range	200 miles
Charging	Level 2 (6.48 kW)
Battery SOC	20% – 80%
Travel needs	Guaranteed
Overrides	2 – 3 per month

Program & Scenarios
BEV : Household	1 : 1
Enrollment	0% – 100%
Valley start	8 PM – 12 AM
Best valley	11 PM
Time spans	Year / day / season

Key assumptions: 200-mile BEVs on Level 2 home charging, batteries kept between 20 and 80 percent, and travel is never compromised: sessions that can’t be shifted stay in place as override events, which stay within 2 to 3 per vehicle per month. We simulate the full range of enrollment levels and five time-of-use window configurations; the 11 PM valley start wins in both regions, so all reported results use it.

Step 1: BEV Charging Load Profile

Simulation results based on ~26,000 survey responses from 2022 NHTS.

Step one builds the unmanaged BEV charging load from about 26,000 travel records in the 2022 NHTS. People come home in the evening and plug in, so the load peaks at 7 PM, exactly on top of the existing grid evening peak.

Step 2: BEV Load Peak Shaving

Five configurations with different peak/valley windows.

Two messages in this plot. First, besides the baseline, there are five configurations that differ only in when the valley window opens, from 8 PM to 12 AM, so we can directly compare which peak/valley allocation shaves the most. Second, this is drawn at 50% enrollment, not 0 or 100, so you can see the interim shifting: half of the fleet’s charging has moved into the valley, and the other half still charges as usual. Mechanically, from 10 AM charging is suppressed unless the battery would fall below 20% or tomorrow’s travel would be at risk, and deferred demand is delivered at maximum rate once the valley opens.

Step 3: Grid Net Load Profile

Net Load = Grid Supply - Variable Sources

Step three is the grid baseline. CAISO’s 2025 average net load shows the classic duck curve: a midday dip from solar and a steep evening ramp. The evening peak and overnight valley together define the opportunity window for SMC. We plot days starting at 7 AM so the peak sits in the center and the valley forms one contiguous block on the right.

Step 4: Net Load + BEV & Peak Shaving

BEV load scaled to fleet size and stacked on net load.

Step four stacks the scaled BEV load on the grid baseline. Each simulated vehicle represents 769 vehicles in CAISO and 316 in NYISO. The black line is total load with no SMC; the green line is full enrollment: the evening peak drops and the load moves into the overnight valley.

Peak Shaving: CAISO Extreme Days

Best Day (Jun 1)		Worst Day (Dec 13)
Peak shave	5.4 GW (37.6 → 32.2)	Peak shave	1.9 GW (30.7 → 28.8)
Percentage	14.4%	Percentage	6.2%

Performance varies a lot day to day. CAISO’s best day shaves 5.4 gigawatts, its worst under 2. Important nuance: extreme days are ranked by how much peak can be shaved, not by how high the peak is: on heat-wave days, air conditioning dominates and BEV charging is a small share, so SMC has less leverage.

Peak Shaving: NYISO Extreme Days

Best Day (Jul 13)		Worst Day (Nov 8)
Peak shave	2.2 GW (27.8 → 25.6)	Peak shave	0.8 GW (17.8 → 17.0)
Percentage	7.9%	Percentage	4.5%

NYISO’s day-to-day range is narrower: the best day shaves 2.2 gigawatts, the worst 0.8. The same ranking nuance applies: days are ranked by achievable shaving, not by demand level.

Seasonal Patterns: CAISO

Spring	Summer	Fall	Winter
3.7 GW (27.5 → 23.8)	3.7 GW (34.3 → 30.6)	3.8 GW (32.3 → 28.5)	3.6 GW (29.1 → 25.5)
13.5%	10.8%	11.8%	12.4%
Peak shave for all seasons: 3.6–3.8 GW (10.8–13.5%)

Seasonally, CAISO is remarkably consistent: 3.6 to 3.8 gigawatts in every season, because the solar-shaped profile keeps the peak-to-valley gap wide year-round. Summer has the highest absolute peak from air conditioning, but the shaving potential barely moves.

Seasonal Patterns: NYISO

Spring	Summer	Fall	Winter
1.5 GW (18.4 → 16.9)	1.6 GW (24.6 → 23.0)	1.5 GW (19.8 → 18.3)	1.4 GW (22.1 → 20.7)
8.2%	6.5%	7.6%	6.3%
Peak shave for all seasons: 1.4–1.6 GW (6.3–8.2%)

NYISO’s winter is the exception: overnight space heating keeps the valley elevated, leaving less room for deferred charging and giving the smallest seasonal shaving at 1.4 gigawatts. The other three seasons cluster at 1.5 to 1.6.

Seasonal cost efficiency is
consistent in CAISO, but varies in NYISO.

The efficiency frontiers by season tell the same story: CAISO’s four curves nearly overlap, so a year-round program delivers consistent value. NYISO’s winter curve sits below the others at every enrollment level, so utilities there may need seasonal enrollment targets or incentive levels.

How Do Pilot Programs Compare?

Program	Incentive	Enrollment
PCE Managed Charging (CAISO)	$0 → $40 per month	1.4% → 10.6%
MCE Sync (CAISO)	$50 bonus + $10 per month	10% of serviceable market
SmartCharge New York (NYISO)	Up to $400 per year	~7% of registered EVs (731 vehicles)

• Smart charging programs motivated EV owners to charge off-peak.
• More incentive results in more enrollment.
• Stated preference surveys are not calibrated against current pilot programs.

These are three deployed pilot programs: two in CAISO territory, one in NYISO. From them I draw three takeaways.

First, smart charging programs motivated EV owners to charge off-peak: MCE Sync cut charging in the 4-to-9 PM window by nearly 90%, with 96% of charging now off-peak. This matches our program setup, which shifts enrolled charging away from the evening peak.

Second, more incentive results in more enrollment. PCE’s randomized pilot shows this directly: 1.4% with no payment, rising monotonically to 10.6% at $40 per month. Our simulation reveals the same relationship from the stated preference side.

Third, stated preference surveys are not calibrated against these pilot programs: real enrollment sits around 4 to 10 percent, below our 30 to 50 percent target. That’s expected: stated preference always overstates real behavior, and these pilots are limited to small areas over short periods, so they carry their own biases. But this is exactly why stated preference is still valuable: it gives a theoretical benchmark, the optimistic upper bound of what these programs can achieve. And even though the numbers are optimistic, the trend is correct: the field data confirms the direction our model predicts.

If asked: the 7% SmartCharge New York enrollment is our calculation: 731 participants against roughly 9,900 EVs registered in Con Edison territory per NYSERDA data; the report itself doesn’t state a percentage.

Appendix - Study 3The surveydown Survey Platform

surveydown Feature Highlights

Question types

Conditional logic

12 Built-in Question Types

text

textarea

numeric

mc

mc_multiple

mc_buttons

mc_multiple_buttons

select

slider

slider_numeric

date

daterange

Question Type: text

sd_question(
  type  = "text",
  id    = "fav_hero_name",
  label = "Who is your favorite super hero?"
)

Question Type: mc_buttons

sd_question(
  type  = "mc_buttons",
  id    = "dream_power",
  label = "If you could have ONE superpower, which would you choose?",
  option = c(
    "🕸️ Web-slinging"   = "webslinging",
    "🛡️ Super Strength" = "strength",
    "✈️ Flight"         = "flight",
    "🧠 Telepathy"      = "telepathy",
    "⚡️ Super Speed"    = "speed"
  ),
  direction = "vertical"
)

Conditional logic

Conditional showing

Conditional skipping

Conditional stopping

Conditional showing - sd_show_if()

Conditional showing - sd_show_if()

survey.qmd

# Conditional Question
sd_question(
  type  = "mc",
  id    = "has_fav_hero",
  label = "Do you have a favorite hero?",
  option = c("Yes" = "yes", "No" = "no")
)

# Target Question
sd_question(
  type  = "text",
  id    = "fav_hero",
  label = "Who is your favorite super hero?"
)

app.R

# Inside server
sd_show_if(
  input$has_fav_hero == "yes" ~ "fav_hero"
)

Condition ~ Target

“If {condition}, then show {target}”

Generate surveys with LLMs!

The companion sdstudio package

Reference List

Bailey, M. R., Brown, D. P., Shaffer, B. C., & Wolak, F. A. (2023). Show Me the Money! Incentives and Nudges to Shift Electric Vehicle Charge Timing [Working {Paper}]. National Bureau of Economic Research. https://doi.org/10.3386/w31630

Huang, B., Meijssen, A. G., Annema, J. A., & Lukszo, Z. (2021). Are electric vehicle drivers willing to participate in vehicle-to-grid contracts? A context-dependent stated choice experiment. Energy Policy, 156, 112410. https://doi.org/10.1016/j.enpol.2021.112410

Philip, T., & Whitehead, J. (2024). Consumer Preferences Towards Electric Vehicle Smart Charging Program Attributes: A Stated Preference Study. https://doi.org/10.2139/ssrn.4812923

Tarroja, B., Shaffer, B., & Samuelsen, S. (2015). The importance of grid integration for achievable greenhouse gas emissions reductions from alternative vehicle technologies. Energy, 87, 504–519. https://doi.org/10.1016/j.energy.2015.05.012

Wong, S. D., Shaheen, S. A., Martin, E., & Uyeki, R. (2023). Do incentives make a difference? Understanding smart charging program adoption for electric vehicles. Transportation Research Part C: Emerging Technologies, 151, 104123. https://doi.org/10.1016/j.trc.2023.104123