Establishing a statewide electric vehicle charging station network in Maryland: A corridor-based station location problem
Introduction
The transportation sector has been going through transformational changes that are widely recognized as three revolutions: electrification, automation, and shared mobility [1]. Among them, the electrification of the transportation systems has been evolving rapidly in tandem with the technological advances in vehicle and fuel and energy supply/storage technologies. Policies and regulations that enable adoption of these technologies, primarily targeted towards promoting electric vehicles (EVs), have been introduced both in the US and around the world at various degrees.
According to Global EV Outlook [2], there has been rapid growth in EV deployment in the past ten years. Globally, the number of EV passenger cars passed 5 million in 2018, increasing 63% from 2017. The US has 22% of these EVs, behind Europe with 24% and China 45%. Other modes of transportation have also started to appear in the EV market. For example, electric two-wheelers for the shared micro-mobility market, electric buses, and trucks mostly deployed as light-commercial vehicles (LCVs) reached 260 million, 250,000, and 540,000 globally in 2018, respectively. On the charger side, there are 5.2 million chargers, of which 540,000 of them are publicly accessible. While these are encouraging numbers, the continuous adoption of EVs requires policies, e.g., greenhouse gas emission (GHG) reduction targets, programs that incentivize and promote EVs, and deployment of charging stations. Globally, and in the US, the availability of the charging infrastructure continues to be one of the significant factors influencing EV adoption, as well as the higher vehicle costs and the limited driving ranges.
Many local and federal government agencies around the world have been providing funding for infrastructure development to boost EV adoption. For instance, the US Federal Highway Administration (FHWA) has been working to establish a national network of alternative fueling and charging infrastructure along national highway system corridors. This program, introduced through the Fixing America's Surface Transportation Act (FAST Act) [3], intends to promote clean and domestically produced alternative fuels, address range anxiety, accelerate public adoption of alternative fuels and vehicles to reduce the dependence on foreign oil. The FAST Act has provided an opportunity for formal corridor designations on an annual basis, given that proposed corridors support forming a national network. Since rounds 1–3 (2016, 2017, and 2018) of FHWA's Alternative Fuel Corridor (AFC) designations, over 135,000 miles of the National Highway System are designated as alternative fuel corridors covering 46 states and District of Colombia.
The state of Maryland, the focus of the case study in this paper, is working to develop statewide EV corridors as the numbers of electric vehicles continue to increase. Utilizing the FHWA's AFC solicitation, Maryland has designated EV corridors including I-95, US 50, I-270, I-70, I-83, I-81, I-695, I-495, part of US 301, I-97, MD 32 and MD 100 (see Fig. 1) as EV-ready and others as EV-Pending, bringing the number of EV corridors up to 20 in the state. There are currently over 628 charging stations, 99 of which are DC fast charging stations, in Maryland. The state plans to build 24,000 residential, workplace, and public charging stations with the help of environmental groups, government agencies, and other stakeholders. Maryland needs to ensure that the charging stations are in locations that maximize travel and support future infrastructure development while enabling these corridors.
Northeast and Mid-Atlantic states, including Maryland, currently use an interactive GIS-based tool developed through Transportation and Climate Initiative (TCI), as well as input from the community for charging station location decisions. These states use the tool to identify the location of chargers along EV corridors at state and regional levels. The GIS-based tools address complexities introduced by many stakeholders involved in the process and the need for satisfying many criteria that a mathematical modeling approach alone would not be able to provide. However, an optimization-based approach can provide complementary information to this complex decision problem and help guide the existing approach by quantifying the impact of various alternatives. This paper describes an optimization-based approach to inform EV charging station location decisions, and evaluate various objectives considering the designated EV corridors. We formulate the problem adopting the Flow Refueling Location Model (FRLM) [4]. We introduce a new metric to the FRLM where we maximize Corridor-Miles Traveled (CMT) to prioritize EV corridors. The CMT is defined by the distance traveled on corridors multiplied by the flow using that corridor. We also introduce two new objective functions to expedite the deployment of EV corridors. We then explain how the proposed methodology can enhance current planning and policy development efforts in Maryland and elsewhere.
The paper continues in Section 2 with a discussion of literature related to refueling station location problem in the context of AFVs and AFCs. Next, in Section 3, we present the Corridor-Based Station Location Problem. In Section 4, we provide additional information regarding the state of Maryland's EV corridor initiative and present experimental design parameters. In Section 5, we present the numerical experiment results and discuss policy implications. Finally, we discuss conclusions and future directions in Section 6.
Section snippets
Literature review
There are two main optimization-based approaches used in locating refueling stations for alternative-fuel vehicles, including fast-charging stations for electric vehicles: node-based demand models and flow-based demand models. There is also emerging literature on location of destination chargers. These, Level 2 slow chargers, are less commonly used to recharge on longer distance trips due to their longer charge times and associated wait periods for travelers. We do not consider their placement
Problem definition
Consider a decision-maker who plans to create an EV corridor for a geographical area by locating m charging stations given a set of candidate stations, M. The traffic flow is between Origin-Destination (O-D) pairs, and we assume EVs follow the shortest path between O-D pairs. We also assume N (N ⊆ M) is the set of O-D nodes, Q is the set of all O-D pairs, q is the index for O-D pairs, and fq is the traffic flow that travels between O-D pair q, QC is a set of O-D pairs that flow between them
Maryland case study
The Maryland Department of Transportation (MDOT) oversees the EV-related policies and implementations in the state while collaborating with relevant stakeholders through its Zero Emission Electric Vehicle Infrastructure Council [25] founded by legislation in 2011. The Council's major functions are to evaluate incentives for the EV purchase and ownership, develop recommendations for a statewide infrastructure plan and other policies, oversee EV infrastructure deployment through the Electric
Experiment results and policy implications
The experiment results show that the majority of the original problem instances (52 out of 90) have alternate optimal solutions. Consequently, prioritization plays a crucial role in terms of putting more stations on corridors. On average, the prioritization increased the number of stations on corridors by 7.1 charging stations compared to no prioritization (neutral policy) and by 8.4 stations compared to deprioritized corridor nodes when there are alternate optimal solutions. Fig. 6 shows that
Conclusions
This research is the first study that considers the mathematical modeling approach for enabling EV corridors on a real-world transportation network. The methodology can be used in conjunction with the GIS-based approaches to help guide complex station location decision problems, that require multiple stakeholder input and expert knowledge, by adding quantitative analysis capability of various alternatives. Our research shows that while maximizing captured flow and vehicle-miles traveled are
Author statement
This research is not published elsewhere and it is not under review elsewhere.
Sevgi Erdoğan, is an associate research professor and director of the Transportation Policy Research Group at the National Center for Smart Growth at the University of Maryland. She will join Syracuse University as an associate professor in the School of Information Studies in spring 2022. Her research includes transportation systems analysis, sustainable transportation and logistics, and modeling transportation systems in relation to other interconnected systems. She has served on the
References (27)
- et al.
Automation, electrification, and shared mobility in urban freight: opportunities and challenges
Transport. Res. Procedia
(2020) - et al.
An arc cover-path cover formulation and strategic analysis of alternative-fuel station locations
Eur J Oper Res
(2013) - et al.
The flow-refueling location problem for alternative-fuel vehicles
Soc Econ Plann Sci
(2005) - et al.(2013)
- et al.
An efficient heuristic algorithm for the alternative-fuel station location problem
Eur J Oper Res
(2018) - et al.
A corridor-centric approach to planning electric vehicle charging infrastructure
Transport Res Ser B: Methodological
(2013) - et al.
A general corridor model for designing plug-in electric vehicle charging infrastructure to support intercity travel
Transport Res C Emerg Technol
(2016) - et al.
Efficient and equitable transnational infrastructure planning for natural gas trucking in the European Union
Eur J Oper Res
(2017) - et al.
An approach for the optimal planning of electric vehicle infrastructure for highway corridors
Transport Res E Logist Transport Rev
(2013) - et al.
Long-term strategic planning of inter-city fast charging infrastructure for battery electric vehicles
Transport Res E Logist Transport Rev
(2018)
Heuristic algorithms for siting alternative-fuel stations using the flow-refueling location model
Eur J Oper Res
Global EV Outlook 2019
Fixing America's Surface transportation Act, enacted by the 114 United States congress
Cited by (20)
Decarbonization in Mexico by extending the charging stations network for electric vehicles
2023, Results in EngineeringSustainable construction project of electric vehicle charging stations: A risk-based hybrid decision-making approach
2023, Journal of Cleaner ProductionEvacuation route planning for alternative fuel vehicles
2022, Transportation Research Part C: Emerging TechnologiesCitation Excerpt :However, emergency planners do not often make these decisions since refueling infrastructure networks are installed and operated by industry. Critical Stakeholder: Studies in the literature have contributed to planning refueling stations deployment in cities (Ghamami et al., 2015; He et al., 2015) and corridors (Erdoğan et al., 2022; Ghamami et al., 2020; Kim and Kuby, 2012; Nie and Ghamami, 2013). In practice, there are ambitious support packages from federal and state governments, investing in alternative refueling stations.
The spatial planning of public electric vehicle charging infrastructure in a high-density city using a contextualised location-allocation model
2022, Transportation Research Part A: Policy and PracticeCitation Excerpt :Elsewhere, Jia et al. (2019) devised a two-step optimisation model to minimise total travel time from demand clusters to charging stations, and Vazifeh et al. (2019) expanded the set covering objective to minimise charging station volume, excess driving distance from trip end-points to stations and energy overheads. Meanwhile, extending the flow refuelling model, Erdoğan et al. (2022) tailored a corridor-based problem and demonstrated that it outperformed the traditional objectives in a case considering candidate locations along and outside the corridor. The LA analysis in the studies mentioned above were typically solved under a series of constraints.
Sevgi Erdoğan, is an associate research professor and director of the Transportation Policy Research Group at the National Center for Smart Growth at the University of Maryland. She will join Syracuse University as an associate professor in the School of Information Studies in spring 2022. Her research includes transportation systems analysis, sustainable transportation and logistics, and modeling transportation systems in relation to other interconnected systems. She has served on the editorial boards of Transportation Research Part-E (2015–2020), Urban Planning, Energies, and TRB committees of Transportation Network Modeling, and Transportation Demand Forecasting. She earned a Ph.D. in Civil Engineering from the University of Maryland.
İsmail Çapar, is an associate professor at the Industrial Distribution Program at Texas A&M University. His research interests include inventory and transportation coordination, distribution and logistics management, warehouse design and optimization, supply chain network optimization, transportation issues during incidents of national significance, location/pricing decisions for alternative fuel vehicles. He serves as associate editor of Networks and Spatial Economics journal. He earned his Ph.D. in Industrial and Systems Engineering from Mississippi State University, and M.S. in Industrial Engineering from Sabanci University, Turkey.
İbrahim Çapar, is an assistant professor at the Applied Statistics and Operations Research Department at Bowling Green State University. He earned his Ph.D. in the Operations Management (OM) program at the University of Alabama (UA) and his MBA degree from the University of Alabama in 2011 with dual concentrations in Business Analytics and Supply Chain & Operations Management. His research interests mainly focus on applied, large-scale optimization problems with applications in routing. He has been an active member of several professional organizations, including INFORMS and POMS.
Mohammad Motalleb Nejad, is a Ph.D. candidate in the Civil and Environmental Engineering Department and a graduate research assistant at the National Center for Smart Growth at the University of Maryland. His research areas include Data Analysis, Optimization, and Machine Learning and its application in civil engineering. He applied optimization methods to structural designs and transportation operations in extreme events. His current research focuses on integrating machine learning approaches such as Neural Networks, Copulas, etc., into generative models. His expertise also includes transportation demand modeling and forecasting. He received his M.S. degree from the University of Delaware in Civil Engineering.