Explicit consideration of human exposure to minimize freight routing impacts

As freight demand grows, so do its negative externalities, such as health and environmental impacts. This study integrates a reduced-complexity air quality model (RCM) with a multimodal traffic assignment model to analyze the scenario in which trucks minimize their public health costs. The study compares the least health-cost path (LHCP) with the shortest, fastest, and least-emissions paths. Two LHCP scenarios with geofences around sensitive populations differed in travel time (TT), health costs, and community impacts. LHCP had 10.9% higher TT and 29% lower costs than the next lowest scenarios. Under the LHCP, within-scenario burdens increased for some communities (e.g., PM2.5 burden) but decreased for others (e.g., ozone burden). The other routing scenarios similarly produced differential effects across communities, highlighting tradeoffs. Geofences increased TT and health costs relative to the LHCP and produced different community impacts. This study provides new, actionable insights into methods to mitigate transportation externalities, particularly in communities.

Introduction
As manufacturing and shopping patterns change, the US relies more heavily on multiple sources and freight modes to move massive quantities of goods across the country and worldwide (USEPA & OAR, 2016a). The US Department of Transportation (2022) reports that in 2023, the daily average tonnage was 55.5 million, valued at approximately $51 billion. The Freight Analysis Framework (2024) estimates that tonnage will increase by 1.2 % annually.

As demand for freight transportation grows, so do emissions and the negative externalities borne by nearby communities. Further, heavy-duty trucks (HDTs) have the fastest-growing share of emissions within the transportation sector. Among the most harmful transportation pollutants to air quality and public health are criteria pollutants, including nitrogen oxides (NOX), volatile organic compounds (VOCs), and particulate matter (PM) (USEPA, 2022). The US Environmental Protection Agency (USEPA) estimates that the transportation sector accounts for over 50 % of total NOx emissions, over 30 % of VOCs, and over 20 % of PM in the US (USEPA & OAR, 2016a). The health impacts of transportation-attributable emissions are well-documented and include elevated asthma risks, acute lower respiratory infections, lung cancer, irritation and inflammation, and premature mortality (Gauderman et al., 2005, McConnell et al., 2006, McConnell et al., 2010, Zhu et al., 2002, USEPA., 2022, CARB, xxxx, Brugge et al., 2007; Cesaroni et al., 2013; HEI, 2022).

Emissions of criteria pollutants from the transportation sector disproportionately impact disadvantaged communities in the region, as identified by California State Bill 535 designations (SB 535 Disadvantaged Communities, 2022). Thus, California has historically prioritized policy and planning measures to improve air quality (e.g., the Clean Air Act of 2006), which have progressively set targets to reduce emissions significantly. Several rules have focused on the transportation sector (e.g., Advanced Clean Cars, Advanced Clean Fleets, 2045 decarbonization plan).

Most current state policies and strategies in California promote air quality improvements by adopting and deploying zero- or near-zero-emission heavy-duty trucks (ZEHDTs), which will take time to become fully market-ready. Identifying complementary and supplementary strategies to improve local air quality and achieve the State’s targets on a shorter timeline is critical. To this end, this study analyzes a quick-to-implement strategy to address vehicle operations that disproportionately burden communities (Samet and White, 2004, Schlossberg, 2021, Tessum et al., 2021). Specifically, the work focuses on routing strategies that consider the local health impacts of freight flows, coupled with spatially constrained strategies such as geofencing.

In doing so, this study builds upon eco-routing tools in the literature, which generate unique routing options based on network conditions and private objectives, such as minimizing transportation costs, fuel consumption, or emissions. Some of these studies have developed origin-based multi-class traffic assignment models, solved using the paired alternative segments (mTAPAS) algorithm, to estimate the network-wide effects of eco-routing (Pahwa and Jaller, 2024, Jaller et al., 2021). This type of eco-routing model dynamically estimates vehicle emissions based on traffic assignment-induced congestion, and its network-based nature provides high-resolution, link-level emissions estimates. Empirical evidence demonstrates that eco-routing and geofencing can lead to significant local emissions reductions, depending on the characteristics of the geofenced strategy, without negative consequences for the rest of the region. However, a critical limitation of existing models is that they consider only changes in emissions and do not account for air pollution-related health impacts on populations affected by the routing scenarios. Therefore, this study integrates health exposure into eco-routing models to explore the opportunities and challenges of freight eco-routing for minimizing public health costs.

Specifically, this study performs traffic assignment to minimize the public health costs imposed on communities by HDTs, then analyzes community impacts between different routing scenarios. The primary contribution of this study to the existing field of knowledge is the novel integration of a reduced-complexity air quality model (RCM) that quantifies exposure-related health impacts by accounting for the complex transition from vehicle emissions to estimated pollutant concentrations, with a multi-class traffic assignment model. This integrated model seeks an impact-based equilibrium on health. This study develops and implements an algorithm that accounts for the complexity of the multi-class public health assignment. Based on the least health-cost path (LHCP) assignment, the community impacts are quantified and discussed in relation to the other scenarios under study, namely, the least emissions path (LEP), the shortest path (SP), the fastest path (FP), and geofenced scenarios. Furthermore, this study employs different objectives for passenger cars (minimizing travel time) and HDTs (minimizing health costs) to ensure that the outcomes provide practical insights for mitigating health effects from the freight sector.

This paper first discusses the literature focusing on the health impacts of truck travel and traffic routing methods. Then, it describes the methodology that integrates an RCM into the multi-class traffic assignment and the proposed approach for evaluating community impacts. The remainder of the paper presents the empirical results and their implications, concluding with policy, operational, and planning recommendations.