Multi-Criteria Decision-Making of Hybrid Energy Infrastructure for Fuel Cell and Battery Electric Buses

Abstract

This study evaluates four hybrid infrastructure scenarios for supporting battery electric buses (BEBs) and fuel cell electric buses (FCEBs), analyzing different combinations of grid power, solar energy, battery storage, and fuel cell systems. A multi-stage framework—comprising energy demand forecasting, infrastructure capacity planning, and multi-criteria decision-making (MCDM) evaluation incorporating total cost of ownership (TCO), carbon emissions, and energy resilience—was developed and applied to a real-world transit depot. The results highlight critical trade-offs between financial, environmental, and operational objectives. The limited rooftop solar configuration, integrating solar energy through a Solar Power Purchase Agreement (SPPA), emerges as the most cost-effective near-term solution. Offsite solar with onsite large-scale battery storage and offsite solar with fuel cell integration achieve greater sustainability and resilience, but they face substantial cost barriers. The analysis underscores the importance of balancing investment, emissions reduction, and resilience in planning zero-emission bus fleets.

1. Introduction

Hybrid charging stations represent a transformative advancement in public transportation infrastructure, integrating renewable energy sources to support the diverse energy needs of hydrogen fuel cell buses (FCEBs) and battery electric buses (BEBs). These stations utilize clean energy technologies, such as solar and wind power, for hydrogen production and battery charging, reducing dependency on fossil fuels and significantly lowering carbon emissions [1,2,3]. As transit agencies transition toward zero-emission bus fleets, significant attention is being directed toward the infrastructure required to support these technologies. FCEBs and BEBs each demand distinct energy systems for daily operation—hydrogen production and storage in the case of FCEBs, and electric charging infrastructure supplemented by battery storage for BEBs. Planning such infrastructure is especially challenging at the depot level.

Infrastructure planning for mixed FCEB and BEB fleets must account for each technology’s operational requirements and constraints [4,5,6,7,8,9]. FCEBs, with their extended range and rapid refueling capabilities, are ideal for longer routes or areas with limited charging infrastructure, while BEBs excel in energy efficiency for shorter, predictable routes. By supporting a heterogeneous fleet, hybrid stations empower transit agencies to meet diverse energy demands, adapt to specific route characteristics, and enhance service reliability [10,11,12,13]. Battery storage systems can buffer renewable generation and reduce peak electricity demand, while onsite hydrogen production enables more flexible fueling logistics and reduces reliance on delivered hydrogen. Meanwhile, fuel cells provide a continuous, on-demand energy source by converting stored hydrogen into electricity, ensuring uninterrupted service even during extended grid outages. Dynamic grid interaction enhances resilience by enabling hybrid charging stations to maintain reliable operations during renewable fluctuations while simultaneously supporting grid stability through load balancing, peak demand reduction, and renewable energy integration.

Despite the potential benefits of hybrid charging stations, several challenges must be addressed to ensure their feasibility. Recent studies, such as that of Bansal et al. [14], have shown that the dual-functionality design—combining electric charging and hydrogen refueling—leads to significantly higher capital investment and infrastructure complexity compared to single-technology stations. In addition, operational complexity increases when managing two distinct energy systems, as demonstrated by Erdemir and Dincer [15] and Eltoumi et al. [16], who emphasize the need for effective system integration and coordinated energy management to ensure reliable and efficient performance of hybrid charging stations. Moreover, while fuel cells are often cited for their potential to enhance system resilience, prior work by Maestre et al. [17] and Castillo et al. [18] has shown that efficiency limitations, system complexity, and hydrogen supply constraints can increase operational costs and reduce the overall energy efficiency of hydrogen-based or hybrid systems. These studies collectively underscore the need for comprehensive infrastructure planning methods that evaluate not just economic performance but also emissions and resilience trade-offs. In response to this gap, our work integrates a TCO framework with a multiple-criteria decision-making (MCDM) approach to support informed investment planning for hybrid energy infrastructure.

This paper incorporates an MCDM approach to evaluate hybrid charging stations, considering not only financial metrics such as the TCO but also critical sustainability and resilience aspects, leveraging data from real-world transit operation parameters. The key contributions of this study are as follows:

• Development of a hybrid energy infrastructure planning methodology tailored to transit depots supporting BEB and FCEB fleets.
• Integration of TCO modeling with broader sustainability, emission, and resilience criteria through an MCDM framework.
• Scenario-based evaluation to identify cost-effective and operationally feasible infrastructure configurations.
• Application of real-world demand and fleet data to ensure practical relevance for transit agency decision-making.

The remainder of this paper is organized as follows: Section 2 outlines the four-stage methodology—covering energy demand forecasting, infrastructure capacity planning, TCO analysis, and the MCDM approach—used to evaluate hybrid energy infrastructure scenarios for BEB and FCEB transit operations. Section 3 applies this methodology to a real transit depot, comparing four infrastructure scenarios based on cost, emissions, and resilience. Section 5 discusses the scenario results, highlights key trade-offs, and addresses limitations to support infrastructure planning for zero-emission fleets. Section 6 concludes the paper by summarizing the main findings and providing recommendations for future research and decision-making in sustainable transit development.

2. Methodology

2.1. Analysis Framework

The methodology for analyzing infrastructure for hybrid charging stations involves three main stages: energy demand forecasting, scenario-based infrastructure capacity planning, and MCDM analysis. Figure 1 shows an overview of the analysis methodology, outlining key tasks and considerations within each stage. The proposed analysis framework helps transit authorities balance financial, environmental, and operational objectives when selecting the optimal infrastructure configuration.

• Energy Demand Forecasting: The demand forecast stage involves estimating the energy requirements of the transit system. Utilizing General Transit Feed Specification (GTFS) data, the analysis captures detailed information about bus schedules and routes to understand energy demands by different bus types: BEBs and FCEBs. Additionally, this forecast considers the daily hydrogen and electricity demand based on anticipated operational needs, helping to ensure that infrastructure is properly aligned with energy requirements.
• Scenario-Based Infrastructure Design: Based on the demand forecast, the scenario-based infrastructure capacity planning stage involves selecting the necessary infrastructure type, determining specific capacity requirements, and establishing the quantity needed for each scenario. This approach ensures that each scenario is matched with an optimized infrastructure setup tailored to projected energy demands. Four scenarios are considered in this paper to evaluate different approaches to powering the hybrid charging stations.
• MCDM Analysis: In the final stage, a comprehensive evaluation is performed across multiple critical criteria: (1) A TCO analysis assesses long-term financial performance by considering capital investment, operational, and maintenance costs over a 12-year period. (2) A carbon emissions analysis estimates annual greenhouse gas emissions associated with each scenario, reflecting the environmental footprint. (3) An energy resilience analysis evaluates each system’s ability to maintain operations during grid disruptions or solar shortages.

2.2. Energy Demand Forecast

This section outlines the methodology used to forecast the charging demand for BEBs and FCEBs by incorporating real-world data and key operational assumptions. Figure 2 shows the workflow for energy demand forecast.

The General Transit Feed Specification (GTFS) data forms the foundation of this analysis, providing detailed information on transit schedules, routes, stop locations, and service frequencies [19]. While GTFS data enables precise modeling of each bus’s energy requirements by capturing variations in route patterns, stop intervals, and service timings, it does not include information on actual bus schedules, which is essential for calculating hourly charging demands. Additionally, GTFS lacks data on deadhead distances—distances buses travel without passengers between routes or to/from depots—which are instead obtained from external sources such as Google Maps. These limitations in GTFS data introduce some uncertainties in forecasting hourly demands and accurately estimating total travel distances. To address these gaps, manual adjustments to bus schedules and supplementary deadhead distance data are incorporated to improve forecasting accuracy and better reflect the real-world operational context. Despite these limitations, GTFS data remains a critical resource for aligning charging schedules with bus operations and informing overall infrastructure planning.

The BEB Power Consumption Model, based on data from NJ Transit, utilizes information collected from three BEBs deployed in Camden in 2022. Each bus was equipped with a sensor that tracked energy consumption (kWh/mile), power consumption (kW), ambient temperature (°F), and average speed (mph) from November 2022 to August 2023. A regression model is developed with a high R-squared value, as shown in Equation (1), indicating a strong fit with the observed data. Predicted energy consumption (kWh) for each block is calculated by multiplying power consumption by trip duration, providing a comprehensive assessment of energy requirements across service blocks. This calculation informs total projected energy demands.

$$Power Consumption=54.1-1.5\times T+0.01\times T^2+1.16v$$

(1)

Here, T represents temperature, and v represents speed.

Based on the projected energy demands for each block, it is initially assumed that all service blocks will be assigned to BEBs. If the total energy consumption for a service block exceeds the BEB’s battery capacity, it is assumed that the bus would be unable to complete the route, indicating the necessity of using an FCEB instead. This approach allows for a practical assessment of BEB limitations and highlights blocks that require FCEB deployment to ensure reliable operation. Operational scheduling assumptions are then applied to ensure feasibility:

• Each bus is assigned a maximum of two service blocks per day, balancing operational demands with adequate downtime for charging.
• A minimum two-hour interval between consecutive blocks is required if blocks share the same stops, while a three-hour interval is assumed for blocks without common stops, allowing for travel and charging as needed.

Finally, the forecasted hourly charging demand for BEBs and FCEBs is calculated by aggregating projected demands across all blocks for each hour. This breakdown supports determining the configuration and capacity of charging stations, ensuring they can meet the fleet’s operational daily energy needs efficiently. By calculating total energy requirements based on these assumptions, the methodology informs the design of a charging infrastructure aligned with operational schedules and minimizes potential service interruptions.

2.3. Hybrid Charging Station Infrastructure Design

2.3.1. Infrastructure for Battery Electric Buses Integrated with PV Arrays

Battery electric bus infrastructure with PV arrays combines renewable energy with sustainable transportation. Solar panels generate electricity from sunlight to charge bus batteries, while energy storage systems ensure consistent power. This reduces reliance on fossil fuels, lowers carbon emissions, and provides a green, cost-effective solution for public transit.

(1) PV Arrays

PV arrays, a cornerstone of a hybrid charging station, are designed to capture sunlight and convert it into electricity. Their efficiency and capacity depend on factors such as the type of solar cells used, the array’s orientation, and the amount of sunlight received. As of 2022, the average purchase cost of solar panels is around USD 0.50 to USD 0.70 per watt, with costs between USD 2.40 and USD 3.60 per watt including installation [20,21]. Most commercial panels have efficiencies from 17% to 20% [22]. In our analysis, solar power is procured through a Solar Power Purchase Agreement (SPPA), where the capital and maintenance costs of the PV panels are covered by the provider, and the transit agency only pays a predetermined rate for the generated solar power [23].

(2) Chargers

Charging infrastructure for BEBs encompasses a variety of technologies tailored to meet different operational needs and charging speeds [24,25]. Plug-in charging is the most common method, involving cables to charge buses overnight or during long stops, ensuring they are ready for daily routes but requiring sufficient downtime. Inductive charging (ground-based wireless charging) uses electromagnetic pads installed on the ground for convenient top-ups at stops or terminals without physical connections. Pantograph charging uses a retractable arm to quickly charge buses during brief stops, minimizing downtime and supporting continuous operations. Each charging method offers unique advantages, contributing to the flexibility and efficiency of BEB infrastructure for public transportation. In our analysis, overnight plug-in charging is considered to determine the number of chargers, which can be calculated using Equation (2).

$$No. of Chargers={\displaystyle \frac{Daily BEBs energy demand \left(kWh\right)}{Charging power of charger \left(kW\right)\times Operating time of charger \left(hours\right)}}$$

(2)

(3) Battery storage system

In infrastructure for BEBs integrated with PV arrays, the battery storage system plays a critical role as a bridge between solar energy generation and electric bus charging. Battery capacity can be optimized to balance costs and performance, as extensively discussed in numerous studies on energy storage optimization for bus charging stations [26,27]. In our analysis, which is simplified, the battery capacity is determined based on the one-day maximum surplus solar energy generated between 7 a.m. and 3 p.m., ensuring sufficient storage to meet charging demands when solar generation is low, such as at night or under cloudy conditions. By capturing surplus solar power and utilizing it during peak demand, the system maximizes renewable energy usage, reduces reliance on fossil fuels, and contributes to a more sustainable and efficient transit network.

2.3.2. Infrastructure for Fuel Cell Electric Buses with an Electrolyzer

Hydrogen bus infrastructure with an electrolyzer offers a sustainable solution for public transportation. The electrolyzer uses electricity, often from renewable sources like solar, to produce hydrogen through electrolysis. The hydrogen is compressed, stored, and used to refuel FCEBs at hybrid charging stations, and fuel cells can also convert hydrogen into electricity to power BEBs when needed. This system enables zero-emission transportation while leveraging renewable energy for a sustainable energy cycle.

(1) Electrolyzer

Alkaline and PEM electrolyzers are the most affordable options among electrolyzers nowadays. Apart from the electrolyzer stack, the electrolyzer plant also includes the cost of the balance of plant (BOP). The BOP is a range of system elements such as cooling, purifiers, thermal management, water treatment, etc. [28,29].

Many studies have looked into the potential cost decrease for increasing the module size and reaping the economies of scale [30,31,32]. Therefore, many cost estimation models, including technology development and electrolyzer plant size, have been developed based on the collected cost data. In [30], the cost of electrolyzer plants, including the cost of the BOP, based on the plant capacity and a learning curve/technology development rate, was developed and used in this study.

The cost of an electrolyzer plant in USD/kW is shown in Equation (3).

$$C=(k_0+{\displaystyle \frac{k}{Q}}{Q}^{ \propto })({{\displaystyle \frac{V}{V_0}}) }^{\beta }$$

(3)

Here, C is the electrolyzer plant cost per kW, k₀ and k are fitting constants, Q is the electrolyzer plant capacity (decided based on daily hydrogen demand), and V and V₀ are the plant installation year and reference year, respectively. α and β are fitting constants and are usually referred to as a scaling factor and a learning factor, respectively. In this study, we use the following values based on reference [30] for an alkaline electrolyzer: k₀ = 301.04; k = 11,603; α = 0.649; and β = –27.33.

The formula to calculate the required electrolyzer capacity Q in kW, considering the daily hydrogen production demand and electrolyzer efficiency, is shown in Equation (4).

$$Q={\displaystyle \frac{Daily hydrogen demand \left(kg\right)\times HHV of hydrogen (kWh/kg)}{Operation hours per day\times Electrolyzer efficiency}}$$

(4)

(2) Compressor, hydrogen tank, and dispenser

A compressor is a mechanical device that increases the pressure of a gas by reducing its volume [33,34]. This process is achieved through the compression of gas particles, which results in an increase in gas temperature and pressure. In the context of hydrogen fuel systems, compressors are used to compress hydrogen to high pressures, making it suitable for efficient storage in tanks or for fueling hydrogen-powered vehicles, thus enabling the practical use of hydrogen as a clean energy carrier. Equation (5) calculates the number of compressors needed for the station.

$$No. of compressors={\displaystyle \frac{Daily hydrogen demand \left(kg\right)}{Flow rate (kg/h)\times Operational hours per day\times Compressor efficiency}}$$

(5)

A hydrogen tank is a high-pressure vessel designed for storing hydrogen gas safely and efficiently. Constructed using advanced materials capable of withstanding the pressures required to store hydrogen at high densities (often up to 700 bar (10,000 psi) or more), hydrogen tanks are engineered to ensure safety, durability, and performance [35]. In our analysis, the hydrogen storage capacity is determined based on three days of daily hydrogen demand to ensure uninterrupted supply during periods of high usage or limited hydrogen production.

Hydrogen fuel cell buses are refueled at these stations using high-pressure hydrogen pumps that can deliver hydrogen gas at pressures up to 700 bar [36]. This technology allows for the rapid refilling of a bus’s hydrogen tanks, usually within 10 to 20 min, enabling long-range operation without the long recharge times associated with batteries [37]. Equation (6) calculates the number of dispensers needed for the station.

$$No. of Dispensers={\displaystyle \frac{Daily hydrogen demand \left(kg\right)}{Dispensing rate (kg/min)\times Operational time per day\times Number of hoses per dispenser}}$$

(6)

(3) Fuel cell

Fuel cells play a crucial role in hybrid charging stations, offering a sustainable and efficient solution to supplement the power supply at the charging station. In these stations, fuel cells convert hydrogen into electricity through an electrochemical process, providing a reliable energy source with zero emissions. The capital cost of fuel cell systems is a significant consideration, as fuel cell stacks typically involve a high upfront investment compared to other energy technologies [38,39]. The variability in cost is influenced by factors such as the scale of deployment, technological advancements, and manufacturing processes. Despite the initial investment, fuel cells offer long-term benefits by reducing greenhouse gas emissions and enhancing energy resilience, making them a valuable component in the transition to cleaner transportation solutions.

2.4. Total Cost of Ownership Analysis

In this study, we embark on an in-depth TCO analysis to scrutinize the cumulative costs entailed in the life cycle of a hybrid charging station designed for BEBs and FCEBs [40,41,42,43,44,45]. This analysis meticulously considers the initial expenses involved in establishing the required infrastructure and the ongoing operational and maintenance expenditures that will accrue throughout the infrastructure’s operational lifespan. Our TCO draws upon a wealth of data extracted from the existing literature, providing a solid foundation for our estimations and conclusions. Additionally, we incorporate GTFS data to accurately reflect bus schedules and energy demands. To ensure our analysis is grounded in practicality and relevance, we specifically tailor our investigation to reflect the operational dynamics and bus demand of the Wayne Garage operated by NJ Transit. This focused approach allows us to model our analysis on a real-world scenario, enhancing the applicability and accuracy of our findings.

Equation (7) provides a mathematical method for calculating the NPV of the TCO. This equation incorporates the time value of money, indicating that future costs are discounted back to their present value to account for inflation and other factors that influence the value of money over time.

$$TCO={\displaystyle \frac{Initial capital costs}{{(1+r)}^n}}+\sum {\displaystyle \frac{Annual O\&M costs}{{(1+r)}^t}}$$

(7)

Here, annual O&M costs are operating and maintenance costs incurred over the life cycle of the system; r is the discount rate, which is the rate used to adjust future costs to their present value; t is the number of years from the present to each future year when the costs are incurred; and n is the number of years over which the initial costs are depreciated or amortized.

For the capital expenditure component of the analysis, the scope encompasses a comprehensive suite of infrastructure and equipment investments, should the transit authority opt for direct ownership of the system. This includes overnight plug-in chargers for BEBs, a battery storage system, an onsite electrolyzer for hydrogen production, hydrogen compressors, hydrogen storage tanks for gaseous hydrogen, dual-hose dispensers for hydrogen fueling, a precooling unit, and fuel cells. The analysis assumes a life cycle of 12 years. The BEBs are equipped with a battery capacity of 520 kWh (usable: 390 kWh). The operating and maintenance (O&M) costs for the hybrid charging station cover several essential components to ensure seamless operation for both BEBs and FCEBs. These costs are categorized as follows: electricity costs for BEB charging, electricity costs for hydrogen production, and infrastructure maintenance costs. Additionally, annual maintenance costs for battery storage and hydrogen fueling infrastructure are assumed to be 5% of their respective capital costs.

Table 1. Input parameter values for TCO calculation.

Parameter	Value	Reference
General
Discount rate	3.90%	[46]
Retail electricity price/peak	USD 133.6/MWh	[47]
Retail electricity price/off-peak (9 pm to 6 am)	USD 60/MWh	[47,48,49,50]
SPPA solar price	USD 49.09/MWh	[51,52,53,54]
BEBs
Capital cost of overnight plug-in chargers	USD 106,000	[55,56]
Annual maintenance costs for overnight plug-in chargers	USD 12,000	[55,56]
Charging power of overnight plug-in chargers	40 kW	[57,58,59]
Capital cost of battery storage	USD 1325/kW	[55,56,60]
Fixed annual O&M cost of battery storage	USD 25.96/kW	[60]
Battery efficiency	94%	[61]
FCEBs
Capital cost of electrolyzer	USD 1125/kW	[30]
Electrolyzer efficiency	0.73	[28]
Capital cost of compressor	USD 297,185	[62,63]
Flow rate of compressor	63 kg/h	[62,63]
Compressor efficiency	65%	[62,63]
Capital cost of hydrogen storage tank/gas	USD 700/kg	[64,65]
Capital cost of dispenser/dual hose	USD 140,000	[62]
Capital cost of precooling unit	USD 227,000	[62]
Capital cost of fuel cell	USD 7748/kW	[60,66,67]
Fixed annual O&M cost of fuel cell	USD 32.23/kW	[60]
Fuel cell efficiency	50%	[60,66,67]

displays the cost parameters established for the TCO, providing a comprehensive overview of the financial considerations involved.

2.5. Analysis of Carbon Emissions and Energy Resilience

This section outlines the methodology used to assess the carbon emissions and energy resilience performance of various hybrid charging station configurations. Carbon emissions in this analysis are estimated based on the life-cycle emission intensity of electricity sources used for battery charging and hydrogen production. Two types of electricity are considered: grid electricity and onsite solar PV.

The grid electricity emission factor is assumed to be 0.26 kg CO₂/kWh, based on the 2023 electricity generation mix in New Jersey [68]. The 2023 grid mix is primarily composed of natural gas (50.5%), nuclear (45.8%), and renewables (3.7%). The solar electricity emission factor is 0.048 kg CO₂/kWh, reflecting the life-cycle emissions associated with PV system manufacturing, installation, and maintenance [69]. Although solar energy is commonly regarded as producing zero emissions at the point of use, it still carries a non-negligible carbon footprint when life-cycle impacts are considered. Therefore, total carbon emissions for each scenario are calculated as shown in Equation (8).

$$Total {CO}_2=Total grid electricity consumed\times 0.26+Total solar electricity used\times 0.048$$

(8)

Energy resilience is evaluated based on the system’s ability to supply the required daily BEB energy demand during disruptions to all external energy inputs, including grid electricity and solar generation. The analysis considers the maximum available capacity of local resources (i.e., stored battery energy, hydrogen, and fuel cell) in a simplified way, rather than real-time operational states. Since all scenarios are configured with at least one day’s worth of hydrogen storage, the energy demand of FCEBs is assumed to be fully met for one day during disruptions. Therefore, the resilience assessment focuses solely on the system’s ability to meet BEB charging needs. The primary metric used is the Energy Resilience Ratio (ERR), as shown in Equation (9).

$$ERR={\displaystyle \frac{Total Backup Energy Available}{Total Daily BEBs Energy Demand}}$$

(9)

Total backup energy available is calculated as the sum of usable battery energy and the electricity that can be generated from stored hydrogen via the fuel cell.

3. Scenario-Based Infrastructure Design and Analysis

3.1. Energy Demand Analysis

The demand analysis for this case study is based on publicly available GTFS data from the Wayne Garage of NJ Transit. This data includes detailed block schedules and stop information, serving as the foundation for the bus scheduling and charging demand analysis for BEBs and FCEBs.

Using the BEB Power Consumption Model described in Equation (1), the predicted energy consumption for all service blocks related to the Wayne Garage was calculated. This model provides a detailed estimation of energy usage for each block, accounting for the block’s own energy requirements as well as the energy needed for deadhead trips to and from the block. Figure 3 illustrates the distribution of predicted energy consumption across all blocks, offering insights into the variability of energy demand. This distribution serves as a critical input for determining the energy requirements and operational feasibility of deploying BEBs and FCEBs in the Wayne Garage network.

Based on the assumptions outlined in Section 2 and the energy consumption calculations derived from the BEB Power Consumption Model, manual bus schedules were developed for both BEBs and FCEBs to meet service requirements. Given the large number of buses operating from the Wayne Garage, a representative schedule is presented in Figure 4 for BEBs, illustrating typical scheduling patterns and how energy demands and operational constraints are addressed. Each horizontal line corresponds to a service block assigned to a specific bus, with the block ID displayed above the line. The total energy consumption for each bus, calculated based on the block’s own energy demand and deadhead energy, is listed on the right side of the figure in kWh.

Based on the full schedule developed for the Wayne Garage, the required fleet size is determined to be 192 BEBs and 41 FCEBs to meet the service-level objectives. The bus numbers are assigned according to the complete scheduling plan, ensuring that the energy demands and operational constraints of each block are met efficiently. Figure 5 and Figure 6 illustrate the hourly charging profiles for FCEBs and BEBs based on the charging rates of the hydrogen dispenser and battery charger. The hydrogen charging profile shows a significant peak at midnight, with nearly 200 kg of hydrogen dispensed, and additional smaller peaks in the early morning and late afternoon to evening hours, indicating refueling primarily during off-peak hours. Similarly, the battery charging profile exhibits a notable peak in the evening, reaching over 8 MWh, with additional charging activity in the early morning and at midnight, highlighting the strategic charging of BEBs during off-peak times to optimize electricity rates and minimize operational downtime. These profiles emphasize the need for a robust charging infrastructure to handle peak loads and ensure efficient refueling and recharging to meet the transit agency’s service-level objectives. The daily electricity demand for the BEB fleet is 49.74 MWh, while the daily hydrogen demand for the FCEB fleet is 832.64 kg, further underscoring the importance of efficient energy management and infrastructure planning.

3.2. Scenario-Based Infrastructure Capacity Planning

Four energy supply scenarios are considered for infrastructure capacity planning, each analyzing different combinations of energy sources, storage systems, and backup solutions to meet the charging demands of BEBs and FCEBs. As illustrated in Figure 7, Scenario #1 relies solely on grid power, without solar energy, battery storage, or fuel cells. Scenarios #2 and #3 incorporate solar power and onsite battery storage while excluding fuel cells, prioritizing solar energy for BEB charging and onsite hydrogen production via electrolyzers. Scenario #4 explores a potential hybrid setup that combines solar power, battery storage, and a fuel cell. Hydrogen is produced onsite using electrolysis and later converted back into electricity by the onsite fuel cell system to help power BEB charging when solar and battery resources are insufficient, enhancing system resilience. In all scenarios with a hydrogen supply, electrolyzers are used onsite to produce hydrogen from solar, battery, or grid power inputs.

In the baseline scenario of grid only, the transit agency relies exclusively on grid power to meet the charging demands of both BEBs and FCEBs. All electricity is sourced from the grid, covering both the charging needs of BEBs and the operation of electrolyzers that produce hydrogen for FCEBs. This scenario excludes renewable energy sources, additional power agreements, advanced storage systems, and fuel cell integration, involving only the standard capital costs associated with charging infrastructure and electrolyzers. The configuration provides a foundational model for analyzing alternative energy integration in other scenarios.

In the scenario #2, rooftop PV arrays installed on the Wayne Garage through a Solar Power Purchase Agreement (SPPA) supply renewable energy for BEB charging and hydrogen production via electrolyzers. The system generates approximately 4.67 MWh of electricity per day, with surplus energy stored in a small-scale (4.21 MWh) battery for use during non-solar hours. Grid electricity serves as a backup when both solar and battery power are insufficient. Energy is dispatched in a prioritized sequence—solar, battery, and then grid—to reduce reliance on non-renewable sources. Off-peak grid electricity is utilized for battery charging and hydrogen production when storage levels fall below a defined threshold, supporting cost efficiency and operational resilience. This scenario includes 833 kg of hydrogen storage, which is sized to cover approximately one day of FCEB hydrogen demand, but does not incorporate a fuel cell.

The scenario #3 features utility-scale solar power procured under an SPPA from offsite installations, supplying approximately 94.67 MWh per day—enough to meet the full daily demand for BEBs and electrolyzer operations. Excess generation is stored in a large-scale (85.27 MWh) battery system supplemented by an 833 kg hydrogen storage tank sized to provide one day of hydrogen supply for FCEBs. This ensures continued service during periods of low solar output or high demand. When needed, the grid provides backup energy. Dispatch logic remains consistent with Scenario #2, prioritizing renewable energy and supplementing it with grid power only as necessary. Strategic use of off-peak electricity enables cost-effective battery charging and hydrogen production, maintaining system reliability without the need for a fuel cell.

The scenario #4 represents a hybrid energy system that combines offsite solar generation, small-scale battery storage, an onsite small-scale fuel cell, and grid backup to enhance energy reliability. The offsite PV arrays, provided under an SPPA, generate approximately 94.67 MWh of electricity per day. Solar energy is prioritized for meeting daily operational demands, with surplus stored in a 4.21 MWh battery system. An integrated 3.38 MW fuel cell converts stored hydrogen into electricity, offering a resilient backup during periods of low solar generation or peak demand. To support the continuous operation of the fuel cell, a large 2498 kg hydrogen storage system is included, sized to provide approximately three days of hydrogen supply for FCEBs. Grid power acts as a final safety net, ensuring uninterrupted service. The system prioritizes energy flow from solar first, followed by battery discharge, fuel cell generation, and finally grid electricity. Off-peak grid power is strategically used to recharge the battery and produce hydrogen when storage levels are low, enhancing both cost-efficiency and operational resilience.

Table 2. Technical specifications of TCO.

Parameter	Value
Number of overnight plug-in chargers (all scenarios)	108
Number of dispensers (all scenarios)	3
Number of compressors (all scenarios)	2
Number of precooling units (all scenarios)	3
Electrolyzer capacity (all scenarios)	1872 kW
Hydrogen storage tank capacity (Scenario #2 and #3)	833 kg
Hydrogen storage tank capacity (Scenario #4)	2498 kg
Battery storage energy capacity (Scenario #2 and #4)	4.21 MWh
Battery storage energy capacity (Scenario #3)	85.27 MWh
Fuel cell capacity (Scenario #4)	3.38 MW

outlines various technical specifications and capacities of a solar-powered hybrid charging infrastructure.

3.3. Comparison of Total Cost of Ownership

The TCO analysis evaluates the long-term financial performance of various energy strategies for hybrid charging stations over a 12-year period. It includes two major cost components: capital investment and total operation and maintenance (O&M) cost. The latter encompasses all recurring expenses such as electricity for BEB charging and hydrogen production for FCEBs.

As shown in Figure 8 and Figure 9, the TCO and cost component breakdown across scenarios highlight key financial trade-offs over the 12-year analysis period. Scenarios #1 and #2 exhibit the lowest total costs, with Scenario #2 being slightly more cost-effective due to reduced energy purchase costs for BEB charging and hydrogen production, enabled by rooftop solar integration under an SPPA. Scenarios #3 and #4 exhibit significantly higher TCOs due to large capital investments. Scenario #3’s high cost stems from the deployment of a large-scale battery storage system to support offsite solar generation, leading to increased capital and O&M expenses. In contrast, Scenario #4 incorporates a fuel cell system and small-scale battery storage but results in a higher overall cost due to the fuel cell’s lower efficiency and higher capital cost per unit capacity (USD/MW).

3.4. Comparison of Carbon Emissions and Energy Resilience

Table 3. Annual carbon emissions, grid power costs, and energy resilience for each scenario.

Scenarios #	Annual Carbon Emission (tons)	Annual Grid Power Cost (USD)	Energy Resilience Ratio (ERR)
1	8861	USD 2,830,175	0.00
2	8469	USD 2,292,416	0.08
3	3294	USD 407,620	1.61
4	5612	USD 1,048,257	0.74

provides a comprehensive comparison of the carbon emissions and energy resilience performance of the four scenarios over a 12-year period. Scenario #1, which relies solely on grid power without any renewable integration or storage, results in the highest annual carbon emissions (8861 tons) and offers no energy resilience (ERR = 0), highlighting its complete dependence on the grid. Scenario #2 introduces rooftop solar through an SPPA and limited battery storage, which reduces annual emissions by approximately 4% (to 8469 tons) and lowers annual grid power cost by 20% compared to Scenario #1. However, the system’s small-scale solar capacity limits its ability to reduce grid reliance further, resulting in only modest energy resilience (ERR = 0.08)—insufficient to cover even one full day of BEB charging demand. Scenario #3 achieves the most significant environmental and resilience benefits, with the lowest annual carbon emissions (3294 tons) and the highest energy resilience (ERR = 1.61). This performance stems from extensive offsite solar generation and large-scale battery storage, which allows most of the energy demand to be met with stored renewable power. Scenario #4 balances resilience and sustainability by integrating solar energy with both a smaller battery and a fuel cell. While it produces slightly more annual emissions (5612 tons) than Scenario #3, it still significantly outperforms Scenarios #1 and #2. It also achieves a moderate ERR of 0.74—higher than Scenarios #1 and #2—indicating improved resilience. In practice, the resilience of a hybrid charging station is inherently limited, as storing large quantities of electricity and hydrogen onsite is not typically feasible or economical for most transit agencies.

4. MCDM Analysis

The MCDM stage provides a structured framework to evaluate the hybrid charging station scenarios by incorporating financial and non-financial criteria. This approach enables decision-makers to assess trade-offs and select the optimal solution by balancing various objectives [70,71,72,73,74]. A spider chart is used to visually represent the performance of each scenario across key criteria, such as TCO, capital costs, O&M costs, carbon emissions, and resilience. The chart highlights areas where each scenario performs well or poorly, offering an intuitive and comprehensive view of trade-offs.

To ensure comparability across criteria with different units and scales, a normalization process is applied. The normalization approach depends on whether a higher or lower value indicates better performance for a given criterion:

• For criteria where a higher value is better (e.g., resilience), the normalized value is calculated using Equation (10).

$$Normalized Value={\displaystyle \frac{Criterion Value}{Maximum Criterion Value}}$$

(10)

For resilience, the value is determined by calculating the ratio of available backup energy (e.g., energy stored in batteries or provided by fuel cells) to the total daily energy demand. This metric quantifies the system’s ability to maintain operations during disruptions and reflects the capacity of the hybrid charging infrastructure to provide energy in case of grid outages.

• For criteria where a lower value is better (e.g., costs or carbon emissions), the normalized value is calculated using Equation (11).

$$Normalized Value={\displaystyle \frac{Minimum Criterion Value}{Criterion Value}}$$

(11)

This normalization process scales all criteria to a range between 0 and 1, enabling meaningful comparisons across scenarios. The normalized values are then used to generate the spider chart and facilitate the evaluation of trade-offs between scenarios, providing a clear basis for decision-making.

The MCDM framework is applied to evaluate the performance of four energy supply scenarios for hybrid charging stations over a 12-year period. This analysis considers five key metrics: resilience, capital cost, O&M cost, TCO, and annual carbon emissions. The results are illustrated in Figure 10, highlighting the trade-offs and advantages of each configuration.

Scenario #1, relying entirely on grid electricity, has the lowest capital cost but exhibits the highest carbon emissions and zero resilience. It demonstrates minimal infrastructure investment at the expense of sustainability and energy security.

Scenario #2 introduces limited rooftop solar and onsite battery storage, slightly decreasing carbon emissions and improving resilience. However, the system’s small-scale solar capacity restricts its ability to sustain even one full day of BEB and FCEB operation without grid power, resulting in minimal overall improvement.

Scenario #3, which includes large-scale offsite solar and extensive onsite battery storage, achieves the highest resilience and the lowest carbon emissions. However, these benefits come at a significant financial cost: a capital cost of USD 44.1 M and a TCO of USD 84.8 M—47% higher than Scenario #1—due to the expense and maintenance demands of battery infrastructure.

Scenario #4 incorporates offsite solar, a small onsite battery, and a fuel cell system. While its resilience is lower than Scenario #3, it still enables partial off-grid operation and reduces emissions to 5612 tons/year. However, it incurs the highest TCO (USD 88 M), driven by the combined costs and lower efficiency of fuel cell technology. Compared to Scenario #3, it shows similar capital and O&M costs but does not achieve proportional gains in resilience or emissions due to the high cost per unit of fuel cell capacity and hydrogen usage inefficiencies.

Overall, Scenario #3 demonstrates the strongest energy resilience and environmental benefits, but its practicality is constrained by the high cost of large-scale battery storage. Scenario #4 presents a promising alternative, though its viability depends on future advancements in fuel cell technology to improve efficiency and reduce costs. At present, Scenario #2 stands out as the most cost-effective option, although it provides only limited improvements in emissions reduction and energy resilience compared to the baseline grid-only configuration.

5. Discussions and Limitations

This study evaluated four infrastructure scenarios for supporting hybrid charging stations for BEBs and FCEBs, each representing a different configuration of energy supply technologies. The scenarios highlighted key trade-offs among capital costs, operational and maintenance expenses, carbon emissions, and energy resilience.

Scenario #1 (grid only) required the lowest capital investment but resulted in the highest annual carbon emissions (8861 tons) and grid power costs (USD 27.76 million over 12 years), underscoring the long-term environmental and economic drawbacks of full grid reliance. Scenario #2 (rooftop solar with onsite small-scale battery storage) introduced solar generation via an SPPA with minimal infrastructure investment. This setup achieved a modest 4% reduction in carbon emissions, along with a 20% decrease in grid power costs compared to Scenario #1. However, the limited solar capacity restricted further gains in sustainability and resilience.

Scenario #3 (offsite solar with onsite large-scale battery storage) provided the strongest environmental benefits and the highest resilience (ERR = 1.61), which were enabled by its extensive battery capacity that maximized solar utilization and reduced grid dependence. However, it also incurred a significantly higher TCO (TCO = USD 84.78 million), driven by a capital cost nearly three times that of Scenario #1 and substantially higher O&M requirements. Scenario #4 (offsite solar with an onsite fuel cell and a small-scale battery) achieved improved energy resilience (ERR = 0.74) and lower emissions than the grid-dominant configurations. However, it did not deliver proportional benefits relative to Scenario #3, despite slightly higher capital and O&M costs. Its limited efficiency and the high unit cost of the fuel cell system contributed to elevated expenses, making it the most costly configuration overall (TCO = USD 88.02 million).

The MCDM analysis highlighted trade-offs across five key criteria: energy resilience, capital cost, O&M cost, carbon emissions, and the TCO. No single scenario performed best in all dimensions. Scenario #3 excelled in resilience and emissions reduction but incurred a high overall cost due to its reliance on large-scale battery storage. Scenario #4 provided a moderate balance of resilience and emissions performance but remained costly, driven by the low efficiency and high capital cost of the fuel cell system. Scenario #2 emerged as the most cost-effective near-term solution, offering modest sustainability gains through limited solar integration with minimal additional investment. Policy incentives, external funding, and continued advancements in battery and fuel cell technologies could make high-cost options like Scenarios #3 and #4 more feasible for widespread adoption.

Despite the practical insights offered by this analysis, several limitations must be acknowledged.

First, the current framework does not incorporate the degradation effects of batteries and fuel cells over time. In real-world applications, these components experience performance decay, which can impact both energy availability, bus schedule, and life-cycle costs. A sensitivity analysis that accounts for degradation rates would provide a more robust understanding of long-term operational reliability and replacement needs.

Second, the resilience metric used in this study is defined as the system’s short-term ability to maintain charging operations at the depot level. While appropriate for evaluating local infrastructure performance, this definition does not capture broader power system resilience, including grid recoverability, black-start capability, or regional interdependencies, which are critical during prolonged outages or system-wide failures.

Third, the model does not explicitly address spatial constraints, which are critical in transit depot planning. Infrastructure components such as the fuel cell system, battery storage, and hydrogen storage tanks require significant space, and siting limitations can substantially affect project feasibility and cost. Integrating spatial data—especially in urban or constrained sites—would improve the accuracy of infrastructure sizing and land-use economics.

Fourth, safety considerations are not included in the current analysis. Hydrogen systems, in particular, involve high-pressure storage and flammability risks that necessitate careful siting, design, and regulatory compliance. Including safety risk assessments in future work would strengthen the practical relevance of the decision-making framework for transit agencies and stakeholders.

Lastly, this study does not optimize infrastructure sizing or operational strategies. Instead, it evaluates predefined scenarios. Future work could incorporate optimization models to dynamically size and manage hybrid systems, aiming to minimize TCO while ensuring that operational requirements are met under variable solar generation, energy demand, and grid pricing conditions.

6. Conclusions and Recommendations

This study evaluated four infrastructure scenarios for supporting hybrid charging stations for BEBs and FCEBs, each representing a different configuration of energy supply technologies. The scenarios highlighted key trade-offs among capital costs, operational and maintenance expenses, carbon emissions, and energy resilience. A multi-stage framework—combining energy demand forecasting, infrastructure capacity planning, and an MCDM evaluation that includes TCO, carbon emissions analysis, and energy resilience assessment—was developed to systematically assess hybrid energy strategies for transit systems.

The results highlight critical trade-offs between financial, environmental, and operational objectives. The limited rooftop solar configuration, integrating solar energy through a Solar Power Purchase Agreement (SPPA), emerges as the most cost-effective near-term solution. Offsite solar with onsite large-scale battery storage and offsite solar with fuel cell integration achieve greater sustainability and resilience, but they face substantial cost barriers. The analysis underscores the importance of balancing investment, emissions reduction, and resilience in planning zero-emission bus fleets.

Future work should enhance this framework by

• Incorporating component degradation modeling to more accurately capture long-term life-cycle costs and the performance decay of batteries and fuel cells.
• Accounting for spatial and siting constraints, particularly for large infrastructure components such as fuel cell systems, battery storage, and hydrogen storage facilities.
• Developing optimization-based approaches to dynamically size and manage hybrid systems, minimizing TCO while meeting operational, environmental, and resilience goals under variable energy supply and pricing conditions.

By addressing these challenges, future research can further strengthen the realism and decision-making value of hybrid infrastructure planning models, supporting more sustainable and resilient transit system development.