The Electric Commute

Over one quarter of the US greenhouse gas emissions can be traced back to transportation. Decarbonizing this sector through electrification will impact private households most directly. At the least, substituting a gas-fueled car with an electric vehicle requires drivers to adapt their fueling/charging routines. Some citizens may instead opt for established electrified modes of transportation, such as buses or trains. Others may prefer emerging electric micro-mobility such as electric bikes or scooters. The exact transportation mode-mixture in a future electrified transportation sector will depend on numerous factors such as personal cost, convenience, capacity of infrastructure, and safety of infrastructure.

The New York State legislature and Governor Kathy Hochul recently passed legislation banning the sale of many gasoline-powered passenger cars and light duty trucks by 2035. With this upcoming ban and ambitious decarbonization goals, New York policy makers are confronted with a significant challenge of understanding the (power) infrastructure needs of this transition and creating an affordable and equitable pathway towards meeting these needs.

This work will be an interactive guide for policy- and decision-makers to understand NYC transportation needs and the impact of its electrification.

The New York State Climate Leadership and Community Protection Act (CLCPA) mandates the achievement of carbon-free electricity by 2040 and a net-zero carbon economy by 2050. Without the abatement of fossil-fuels transportation modes, this mandate will not be met.

Around 2 million people commute in and out of New York City by car every day; only 1% of these cars are electric. Hence, with the goal of achieving an electrified transportation sector within the next one or two decades, there is significant potential for policy makers to influence how households that are currently dependent on gas-fueled cars will transition towards electrified mobility. Besides socio-economic considerations, e.g., equitable access to EV charging, electrified mass transit, and secure infrastructure for micro-mobility, the ability of the electric power system to support comprehensive transport electrification must be considered.

Most of the Manhattan workforce is comprised of commuters with very different travel patterns and distances.

Who are these almost 3 million people?

If commuters keep their current mode of transportation but gradually make the switch from gas-powered vehicles to electric vehicles, the electrical grid will experience a significant increase in demand from EV charging. Without expensive infrastructure upgrades to the grid and additional charging stations, costs will skyrocket whenever the electric transmission and distribution system operates at its technical limits. This impact can be softened if some commuters chose different modes of transportation and EVs are charged on a flexible schedule.

Available data on infrastructure, travel distance, and household demographics informs an estimation of how likely a commuter is to chose a certain mode of transportation. This serves two purposes. Firstly, it allows us to identify a relationship between certain demographic and external factors and the choice of transportation modes. Second, we can use these estimates to create conditional probability distributions of a certain household choosing a certain mode of transportation. Applying these distributions to the entire population obtain a proxy for the spatial distribution of mode choices. The underlying parameters can then be adapted to model the impact of policy decisions, e.g., incentivizing the use of electric bikes or expanding infrastructure, on the overall mode choice.

With the commuter model we can create scenarios of how certain policies may affect transportation mode choices. We focus on four central commuting scenarios with various work-from-home (WFH) assumptions to capture new hybrid working patterns. These scenarios include a public transit-focused scenario, a car-focused scenario, a micro-mobility focused scenario, and a "mix" scenario that captures the most diverse choice of all modes. A baseline 2019 scenario that captures current modes of transportation taken from the American Community Survey (ACS) data is provided for comparison. Each of these scenarios puts a heavy priority on a chosen mode of transportation, meaning if a commuter has more than one option of commuting methods, the public transit focused scenario would more likely assign that commuter to take public transit instead of their other options (give the method a higher weight in the weighted random assignment).

Through a better understanding of the spatial and temporal distribution of commuting routes we can assess energy requirements and efficiency of various mode mixes. Using actual data on state-of-the-art battery capacity and efficiency, as well as the efficiency of various electric drives, we identify the electric power needs for various scenarios. This information can provide useful insights for policy makers on which modes and / or infrastructure development plans should be prioritized from the perspective of energy efficiency.

We created a computational model that combined the hourly travel distance for each mode of transportation with the parameters of our electric vehicle efficiency and battery capacity dataset to obtain the per-mode hourly energy demand. Some complexities had to be considered. First, some commuting methods require electricity at the time of use, like subways or commuter trains, while others, like cars or e-bikes, require electricity from the grid only afterwards, upon post-use charging. Therefore, an understanding of the delays of each mode’s use on its resulting charging demand was needed.

An important techno-economic consideration is flexible/smart charging, which we also included in our model. This captures a future where some commuters (driven by price incentives or automated smart-charging) charge their vehicles immediately upon arrival to Manhattan, while others charge at lunch, and some others even charge towards the end of their work day. Building this parameter into the model provides even more insight into how a given commute can affect the existing grid.

Finally, given all of these calculations, we combined these results with data on existing electricity demand.

Each mode of transportation produces a unique energy demand. Some modes are linearly correlated to the number of people using them, like EVs and micromobility, but some modes are less impacted by the number of riders, like subways and ferries. (Of course, fuller trains require more power but this change is minor relative to the base load of running the train.)

There are large differences in charging demand based on people’s behaviors. The purple dashed line reflects charging demand when charging starts immediately after plugging-in (the current state-of-the-technology), which creates a morning peak after commuters arrive at work. The blue dashed line represents what the energy profile would look like if everybody charged their car at the end of their work day, i.e., if charging was pushed as far into the afternoon as possible, creating an afternoon peak. The red solid line, which is much flatter, shows the charging demand if charging starts randomly between the first and the last possible hour over the duration of each commuter's workday, which we use to approximate an optimally distributed charing pattern. In the 2019 baseline scenario, there is an over 200MW difference in demand between the scenarios of drivers choosing to charge at the start of work vs randomly distributed. The largest difference is created when driving commuters decide to charge, so the difference in demand is most significant in the Car Heavy Scenario.

Comparing the electricity demand identified in our commuter model with the current state of the NYISO power system paints a dire picture. For most scenarios, charging peaks will be very close to the theoretical system limit, some will even surpass it. On the one hand, this indicates the necessity for massive infrastructure investments, as the computed limit is only theoretical and practical limits are likely much lower. On the other hand, these results highlight the value of incentivizing smart charging infrastructure that allows to distribute charging demands over the day.

Transparent data-driven insights are fundamental for democratic policy-making. The TEC-NYC project provides such insights for policy makers working towards the future electrified transport sector and the interested public. Based on real-world data and interpretable parameters that are fed into our model, we can answer import questions that can inform policy designs and priorities.

What if ... Commuters prefer micromobility? What if ... Even more commuters start using subsidized individual electric vehicles? What if ... Public transport access is expanded? What if ... Biking infrastructure is improved? What if ... ?

Read some exemplary "What if ... ?" stories below.

The TEC NYC project offers the opportunity to analyze these and many more questions. As all data and model implementations are released open-source, it will allow users to create infinite customized scenarios to explore all policy planning.

The TEC NYC project is published as an interactive dashboard. It allows users to explore, analyze, and understand the models and their resulting data in greater detail.

Read the publication

Dashboard link:  https://tecnyc.herokuapp.com/  GitHub repo:  https://github.com/BNewborn/mobility-electrification 

This dashboard provides both an overview and a detailed breakdown to further help users identify patterns and gain insights. It can be utilized by a wide range of users, ranging from the everyday interested citizen to policy makers who need to see the larger impacts of potential policy.