Designing a Sustainable Off-Grid EV Charging Station: Analysis Across Urban and Remote Canadian Regions

Abstract

Electric vehicles are becoming more commonplace as we shift towards cleaner transportation. However, current charging infrastructure is immature, especially in remote and off-grid regions, making electric vehicle adoption challenging. This study presents an architecture for a standalone renewable energy-based electric vehicle charging station. The proposed renewable energy system comprises wind turbines, solar photovoltaic panels, fuel cells, and a hydrogen tank. As an energy storage system, second-life electric vehicle batteries are considered. This study investigates the feasibility and performance of the charging station with respect to two vastly different Canadian regions, Windsor, Ontario (urban), and Eagle Plains, Yukon (remote). In modeling these two regions using HOMER Pro software, this study concludes that due to its higher renewable energy availability, Windsor shows a net-present cost of $2.80 million and cost of energy of $0.201/kWh as compared to the severe climate of Eagle Plains, with a net-present cost of $3.61 million and cost of energy of $0.259/kWh. In both cases, we see zero emissions in off-grid configurations. A sensitivity analysis shows that system performance can be improved by increasing wind turbine hub heights and solar photovoltaic panel lifespans. With Canada’s goal of transitioning towards 100% zero-emission vehicle sales by 2035, this study provides practical insights regarding site-specific resource optimization for electric vehicle infrastructure that does not rely on grid energy. Furthermore, this study highlights a means to progress the sustainable development goals, namely goals 7, 9, and 13, through the development of more accessible electric vehicle charging stations.

1. Introduction

Global concerns over climate change and environmental degradation have accelerated the adoption of electric vehicles (EVs) as a sustainable alternative to traditional combustion engine vehicles. The transportation sector alone consumes nearly 50% of global petroleum-derived fuels, contributing significantly to greenhouse gas (GHG) emissions [1]. EVs, particularly when integrated with renewable energy sources (RESs) via charging stations, offer a viable path to reduce fossil fuel dependence and enhance energy sustainability [2].

Renewable energy (RE) systems, while environmentally friendly, are inherently intermittent: solar photovoltaic (PV) generation is unavailable at night or under overcast conditions, and wind energy output fluctuates throughout the day. Integrating multiple RES with an energy storage system (ESS) improves performance and ensures a reliable energy supply [3]. Second-life batteries (SLBs) provide an additional advantage by lowering capital costs by 20–80% compared to new lithium-ion batteries (LIBs) [4], reducing Net Present Cost (NPC) and Levelized Cost of Energy (LCOE). Retired EV batteries typically retain 70–80% of their capacity, making them suitable for off-grid EV charging, microgrids, and stationary applications [5]. Proper battery management and hybrid backup (e.g., fuel cells) ensure system reliability, extend battery lifecycle, reduce e-waste, and support circular economy principles, aligning with sustainable development goals [6].

Recent studies highlight the role of BESSs in balancing supply and demand, optimizing site selection, and improving operational efficiency for EV charging infrastructure [7,8,9,10,11]. Figure 1 illustrates the levels of EV charging infrastructure, including charging speeds, voltages, connector types, and typical applications, integrated into the discussion for clarity.

Remote and northern regions of Canada pose unique challenges for off-grid EV charging due to lack of grid access, harsh sub-zero climates, limited winter solar resources, and high logistical costs [12]. The Dempster Highway, stretching 734 km from near Dawson City, Yukon, to Inuvik, Northwest Territories, exemplifies the need for autonomous, renewable powered charging stations. A standalone charging station at Eagle Plains could provide emissions-free EV charging while serving as a real-world test case for system performance under extreme climatic and operational conditions.

This study develops a comprehensive, site-specific framework for off-grid EV charging by combining several complementary approaches:

• Empirical traffic modeling to quantify EV charging demand.
• Stochastic queuing theory (M/M/C: FCFS/∞/∞) to determine the optimal number of fast charging ports.
• Techno-economic and lifecycle assessment of a hybrid renewable energy system incorporating SLBs, hydrogen storage, and fuel cells.
• Time resolved optimization using HOMER Pro (v3.18.3), accounting for regional RES variability and dynamic cost modeling.

By integrating these components, the framework enables realistic, data driven, and economically optimized system design for carbon-neutral EV fast-charging stations.

The remainder of this paper is structured as follows: Section 2 reviews the literature; Section 3 outlines the system architecture; Section 4 presents the mathematical model; Section 5 describes meteorological data; Section 6 presents EV arrival rates and load estimation; Section 7 presents the results and discussion; and Section 8 concludes this study.

2. Literature Review

Solar PV is increasingly a preferred RES, and integrating wind energy with hydrogen energy storage can significantly reduce GHG emissions. Due to the intermittent nature of renewable sources, energy storage systems (ESSs), particularly LIB, are widely used to enhance reliability and stability [5]. However, LIBs are cost effective only for short-term storage and face material constraints due to lithium and cobalt demand [13]. Alternative ESSs include green hydrogen storage, which offers eco-friendly, transportable solutions for long-term storage. Combining ESSs with fluctuating solar and wind resources improves grid flexibility, reduces curtailment, and enables low GHG electricity generation [14].

Recent studies have explored integrating RE into EV charging stations with various optimization strategies. Ref. [15] highlight smart charging, vehicle-to-grid, and advanced storage solutions, noting ongoing challenges in economic feasibility, technical integration, and regulatory barriers. Ref. [16] optimize a solar PV and grid-powered station using the Spider Wasp Optimizer and a Multi-Scale Hypergraph Alignment Network, achieving 1.04% total harmonic distortion, 0.986 power factor, and $124,000 annual operating cost. Ref. [17] apply Moth-Flame Optimizer for a PV-powered station in Crete, Greece, yielding EUR 856,477 upfront cost and EUR 6,426,355 profit over 20 years. Ref. [18] uses Grey Wolf Optimization and Dynamic Fitness-Distance Balance to reduce power losses and voltage deviations, demonstrating that the optimal placement of charging stations and distributed generation improves grid stability and system efficiency.

Ref. [19] used Dollmaker Optimization and spatial Bayesian neural networks to improve efficiency, reliability, and power quality in PV, wind, supercapacitor, and battery-integrated microgrid stations. Ref. [20] examined grid-connected solar–wind electric vehicle charging stations (EVCSs) in three coastal Bangladeshi cities, supplying excess energy back to the grid. Ref. [21] proposed a bi-level model integrating a hydrogen ESS with RE to maximize social welfare. Ref. [22] optimized a solar-powered grid-connected charging station via scheduling for efficient PV use. Ref. [23] analyzed RE fractions (0–100%) for EVCS in South Korea, assessing GHG emissions and economics using HOMER Pro. A summary of these studies is presented in

Table 1. Summary of literature review.

Author	Year	Objective	Main Components	Software Used	Location	Results	Identified Gaps
[24]	2025	This study simulates a hybrid PV/WT/BG/FC/HT/Elz energy system using a genetic algorithm based on the NSGA-II non-dominated sorting technique.	PV/Wind/Electrolyzer/ Biogas/Fuel cell	MATLAB R2025a, v25.1	Bangladesh	For the PV/WT/FC/BG configuration, the COE is $0.1634/kWh, ecosystem damage 0.00098, human health impact 0.1732 DALYs, and development index 0.696 DALYs. Lifecycle GHG emissions are 123,730 kg CO2-eq/year, carbon penalties $1856/year, with 30 jobs/MW over a 25-year project life.	The work focused on a single Bangladeshi site without analyzing different climates, or second-life battery use, and lacked SDG context.
[25]	2025	In this work, the authors explored the optimal hybrid system in Adana, Türkiye. The focus was to ensure EVCS continuous supply.	PV/Wind/Electrolyzer/ Biogas/Fuel cell.	HOMER version 3.14.2	Türkiye	The optimization results show that the system comprising solar PV, biogas, electrolyzer, hydrogen tank, fuel cell, inverters, and grid is the most viable, with a total NPC of $611,283 and an LCOE of $0.0215/kWh.	Focused only on one Turkish Adana city; lacked demand variation, second-life batteries.
[26]	2024	A comprehensive study evaluated the technical and economic performance of an off-grid system in Pakistan.	PV/Wind/ Hydrogen	HOMER version 3.18.1	Paki-stan	Compare the NPC and COE for various cities of Pakistan for the off-grid charging station	Limited to Pakistani sites without analyzing remote vs. urban contexts; did not assess grid-connected scenario. Detailed feasibility analysis was not presented.
[27]	2024	The study analyzed a hybrid energy system in Australia, comprising solar PV, WT, and BESS, across five regions to identify the optimal location for an EV charging station.	PV/Wind/ Hydrogen/ Battery	HOMER version 3.18.1	Australia	HOMER Pro results indicate that Broken Hill and Murray River achieve optimal costs of $0.32/kWh (off-grid) and $0.030/kWh (on-grid), whereas Tasmania has higher costs of $0.38/kWh and $0.034/kWh for the respective modes.	Compared Australian sites but omitted EV-specific load profiles and policy guidance for remote areas.
[28]	2024	This research evaluates the technical and economic impacts of a renewable energy-based BSS integrated into an EV charging station in Malaysia.	PV/Wind/ Electrolyzer/ Battery/Hydrogen/Fuel cell	HOMER version 3.18.2	Malaysia	The findings revealed positive results with the overall NPC varying between $1.4 M and $3.4 M at all sites, and the COE was observed between the value of $0.03/kWh up to $0.16/kWh.	Explored economics but did not evaluate detailed feasibility analysis and lacked SDG-focused evaluation.
[23]	2023	In this system, the authors presented EV charging stations that are based on the various RESs.	PV/Wind/ Battery/Grid	HOMER version 3.16.2	South Korea	Results indicate that the PV/ESS-based configuration is the most optimal scenario, particularly regarding RE fractions involving PV, WT, and ESS.	Concentrated on Korean scenarios only; missing discussion on multi-region diversity, hydrogen systems, grid-connected systems, and second-life battery impacts.
[29]	2022	This study assessed the techno-economic and environmental performance of a solar PV plant producing electricity and hydrogen across five Indian cities: Chennai, Indore, Kolkata, Ludhiana, and Mumbai.	PV/Hydrogen	HOMER version 3.14.2	India	The highest hydrogen production was observed in Kolkata (82,054 kg/year), followed by Chennai (79,030 kg/year), Ludhiana (78,524 kg/year), Indore (76,935 kg/year), and Mumbai (74,510 kg/year).	Studied hydrogen production but did not include the grid-connected scenario and multi-energy integration strategies.
[30]	2022	This study focuses on a standalone, renewable energy-based EVCS.	PV/Wind/ Electrolyzer/ Battery/Bio	HOMER version 3.14.2	Qatar	The NPS is estimated to be between $2.53 M and $2.92 M, while the COE is estimated to vary between $0.285 and $0.329/kWh	The authors focused on cost metrics for Qatar but missed second-life battery analysis and SDG linkage.

.

2.1. Research Gap

Existing studies on RE-based EV charging systems primarily focus on techno-economic optimization, often overlooking lifecycle emissions, SLB integration, material recovery, carbon policy impacts, and realistic load modeling based on actual traffic and charging data. They also rarely consider regional climatic variations affecting system performance in urban and remote areas.

This research addresses these gaps by integrating techno-economic, environmental, circular economy, and real-world load modeling aspects, including SLB reuse, material recovery, CO₂ emission reduction, and load profiles derived from real traffic data across diverse Canadian climates, providing a comprehensive framework for sustainable EV charging infrastructure.

2.2. Primary Contribution

• This study applies location-specific optimization across two diverse Canadian regions to evaluate how regional renewable resource variability and climatic conditions affect system design and economics under a uniform load demand.
• This study develops a site-specific, data-driven approach that integrates traffic flow analysis, EV capture probability, daily load estimation, and queuing theory to enable optimal sizing of charging station, minimize waiting times, and accurately represent urban charging demand patterns.
• This work integrates hydrogen and SLB technologies with performance degradation modeling accounting for temperature effects, derating, and capacity fades to enhance system autonomy, resilience, and long-term sustainability while addressing RE intermittency under real-world cold climate conditions.
• This study performs detailed economic scaling analysis to evaluate the impact of equipment capacity expansion on NPC, COE, CAPEX, and OPEX, demonstrating a cost-effective, fully renewable off-grid system capable of supporting EV charging infrastructure in both urban and remote Canadian regions.
• This research provides a multi-dimensional assessment of renewable EV charging by integrating lifecycle emissions, SLB reuse, recoverable materials, and operational CO₂ reduction, while linking carbon tax savings and resource recovery to offer a holistic sustainability framework aligned with SDGs 12 and 13.
• This study provides a region-specific assessment of renewable energy that powers EV fast charging stations in Canada, integrating real networks, energy-based pricing, and estimating annual revenues by offering insights for sustainable and strategically located infrastructure.

3. System Architecture

Figure 2 presents the system architecture for a remote EV charging station based on renewable energy (RE). In particular, the model uses a combination of a solar PV, wind turbine, converter, and second-life battery-based energy storage system (SLB-based ESS). This set up directs energy to an electrolyzer and hydrogen tank along with fuel cells (FCs). Moreover, Figure 3 presents the component-based architecture of the charging system.

The details of the model are explained below:

• Solar PV: generates DC electricity, converted to AC for system compatibility.
• Wind Turbine: produces AC electricity directly to supply charging stations.
• SLB Storage: stores excess RE for low generation periods or peak loads, ensuring continuous supply.
• Electrolyzer: converts surplus RE into hydrogen, stored in tanks.
• Fuel Cell: utilizes stored hydrogen to provide clean energy during peak demand, enhancing system reliability beyond typical renewable generation.
• Energy Management System: optimizes energy distribution and battery cycles based on real-time demand and generation.
• EV Charging Stations: deliver hybrid energy from RE, batteries, and fuel cells, even off grid, using demand response and pricing optimization.
• Hydrogen Load: considered in analysis to utilize electrolyzer and storage effectively.
• IoT: provides real-time data on generation, consumption, storage, and charging; red arrows indicate power flow, blue arrows data flow, enabling optimized, reliable, and low-cost off-grid EV charging with reduced environmental impact.

4. Mathematical Modeling and System Formulation

This section presents the mathematical framework for modeling, sizing, and optimizing the hybrid renewable energy system (HRES) for standalone EV charging stations, integrating wind, solar PV, battery storage, electrolyzers, hydrogen storage, and fuel cells. All equations are drawn from peer-reviewed literature, with HOMER Pro used exclusively for simulation and optimization. A methodological flow chart for this process is shown in Figure 4 below.

• Objective Function: the NPC of the HRES is minimized using HOMER Pro, subject to all operational and physical constraints.
• Constraints: operational constraints, technical limits, and energy balance equations are formulated to ensure that the system reliably meets load demand; constraints include power balance, PV installation area, battery state of charge limits, hydrogen inventory, and energy production requirements.
• Decision Variables: sizing and operational parameters including PV and wind capacities, wind turbine hub height, battery and hydrogen storage capacities, converter and electrolyzer capacities, and dispatch strategies are selected as decision variables for optimization.
• Component Modeling: each system component (PV, wind turbine, BESS, electrolyzer, fuel cell, hydrogen storage, and converter) is mathematically modeled using validated equations from the literature to accurately capture performance, operational behavior, and physical constraints.
• Simulation Tool Selection: HOMER Pro is used to perform hourly simulations, capturing stochastic variability, battery cycling, hydrogen production, and renewable generation intermittency.
• Economic and Technical Analysis: system performance is evaluated through NPC, COE, and Total Cost of Ownership (TCO). Component degradation, climate adjustments, and operational limitations are incorporated to ensure accurate and reliable results.
• Meteorological Data: hourly solar irradiation, wind speed, and temperature data are collected from the NREL database to provide site-specific inputs for accurate RE modeling.
• Traffic Data: annual average daily traffic data are obtained from the Office of the City Engineer, Windsor, to estimate EV charging demand and inform system sizing.
• Results and Discussion: the analysis interprets numerical outputs, compares site performance, and examines sensitivities to key variables including solar irradiance, wind availability, ambient temperature, battery degradation, and hydrogen subsystem efficiency.

A methodological flowchart is provided in Figure 4.

As shown in Figure 4, every task is assigned to a section or subsection in the manuscript.

4.1. Optimization Framework and Objective Function

The objective of the optimization is to identify the cost of optimal configuration of the hybrid energy system while satisfying all operational, technical, and reliability constraints.

$$\mathrm{Minimize}: \mathrm{f}\left(\mathrm{x}\right)=\mathrm{NPC}\left(\mathrm{x}\right)$$

(1)

where NPC includes capital, replacement, O&M, fuel, and salvage costs over the project lifetime.

4.1.1. Operational Constraints

The system is subject to several constraints derived from the literature:

• Power Balance Constraint: total generation from PV, wind, fuel cell, and battery must meet load demand practices [3].

$${\mathrm{P}}_{\mathrm{PV}}+{\mathrm{P}}_{\mathrm{WT}}+{\mathrm{P}}_{\mathrm{FC}}+{\mathrm{P}}_{\mathrm{BES}}={\mathrm{P}}_{\mathrm{Load}}$$

(2)

• Solar Installation Area Constraint: PV capacity is constrained by site-specific physical limits [30].

$${\mathrm{A}}_{\mathrm{PV}}\le 3000 {\mathrm{m}}^2$$

(3)

• Daily Electricity Production Constraint: ensures sufficient energy to meet EV charging demand (Section 6).

$${\mathrm{E}}_{\mathrm{prod},\mathrm{daily}}\ge 3591.126 \mathrm{kWh}/\mathrm{day}$$

(4)

• Power Shortage Constraint: the off-grid system must have zero annual unmet demand [30].

$${\mathrm{E}}_{\mathrm{shortage},\mathrm{annual}}=0 \mathrm{kWh}$$

(5)

• Battery SOC Constraint: maintains safe battery operation to prevent degradation [3].

$$20\%\le \mathrm{S}\mathrm{O}\mathrm{C}\le 75\%$$

(6)

• Minimum Hydrogen Inventory Constraint: ensures safe operation of the fuel cell [31].

$${\mathrm{S}}_{{\mathrm{H}}_2}\left(\mathrm{t}\right)\ge {\mathrm{S}}_{{\mathrm{H}}_2,\min }$$

(7)

4.1.2. Decision Variables

Decision variables define the sizing and operational parameters optimized in HOMER:

$$\mathrm{x}=\{{\mathrm{P}}_{\mathrm{PV}},{\mathrm{P}}_{\mathrm{WT}},{\mathrm{H}}_{\mathrm{WT}},{\mathrm{C}}_{\mathrm{bat}},{\mathrm{C}}_{\mathrm{ely}},{\mathrm{C}}_{\mathrm{FC}},{\mathrm{C}}_{\mathrm{conv}},{\mathrm{S}}_{{\mathrm{H}}_2},{\mathrm{D}}_{\mathrm{strat}}\}$$

(8)

Rationale for selection:

• PV and Wind Capacities (${\mathrm{P}}_{\mathrm{PV}},{\mathrm{P}}_{\mathrm{WT}}$): directly determine RE generation and ability to meet load.
• Wind Turbine Hub Height (${\mathrm{H}}_{\mathrm{WT}}$): optimizes wind energy capture via the power law.
• Battery Capacity (${\mathrm{C}}_{\mathrm{bat}}$): provides energy reliability and reduces dependence on backup.
• Electrolyzer and Fuel Cell Capacity (${\mathrm{C}}_{\mathrm{ely}},{\mathrm{C}}_{\mathrm{FC}}$): controls hydrogen storage for long-term energy supply.
• Converter Capacity (${\mathrm{C}}_{\mathrm{conv}}$): ensures efficient AC/DC conversion between sources and loads.
• Hydrogen Storage (${\mathrm{S}}_{{\mathrm{H}}_2}$): maintains minimum hydrogen for continuous fuel cell operation.
• Dispatch Strategy (${\mathrm{D}}_{\mathrm{strat}}$): determines system operation between load monitoring and cycle charging strategies.

All decision variables directly influence NPC by determining component sizing, energy production, storage, and dispatch strategy. Optimizing them identifies the configuration that minimizes NPC while satisfying operational constraints, including power balance, PV area, energy shortages, battery SOC, and hydrogen inventory.

4.2. Component Modeling

This section presents the detailed modeling approach for each component of the hybrid RE system. The objective of these component models is to accurately represent the operational behavior, performance characteristics, and physical constraints of the solar PV array, wind turbine, battery energy storage system, electrolyzer, hydrogen storage, hydrogen vs. SLB storage system, fuel cell, and converter. These component models serve as the foundation for the optimization framework described in later sections. All equations used in this study are taken directly from the established literature and manufacturer specifications, not from HOMER’s internal algorithms, ensuring transparency and reproducibility.

4.2.1. Wind Turbine

In the proposed system, the wind turbine serves to complement the solar PV system, particularly during nighttime and in conditions of overcast or rainy weather. Wind speed at hub height can be calculated from Equation (9), see [32,33].

$${\mathrm{W}}_{\mathrm{h}\mathrm{u}\mathrm{b}}={\mathrm{W}}_{\mathrm{a}\mathrm{n}\mathrm{e}\mathrm{m}}\ast {\left({\displaystyle \frac{{\mathrm{H}}_{\mathrm{h}\mathrm{u}\mathrm{b}}}{{\mathrm{H}}_{\mathrm{a}\mathrm{n}\mathrm{e}\mathrm{m}}}}\right)}^{\mathrm{\alpha }}$$

(9)

where:

${\mathrm{W}}_{\mathrm{h}\mathrm{u}\mathrm{b}}$ = wind velocity at the hub elevation (m/s)

${\mathrm{W}}_{\mathrm{a}\mathrm{n}\mathrm{e}\mathrm{m}}$ = wind velocity at the elevation of anemometer (m/s)

${\mathrm{H}}_{\mathrm{h}\mathrm{u}\mathrm{b}}$ = hub height (m)

${\mathrm{H}}_{\mathrm{a}\mathrm{n}\mathrm{e}\mathrm{m}}$ = anemometer height (m)

$\mathrm{\alpha }$ = exponent of power law

Power obtained from wind turbines is expressed in Equation (10), see [30,32,34].

$${\mathrm{P}}_{\mathrm{w}\mathrm{i}\mathrm{n}\mathrm{d}}=\left({\displaystyle \frac{\mathrm{\rho }}{{\mathrm{\rho }}_0}}\right){\mathrm{P}}_{\mathrm{w}\mathrm{i}\mathrm{n}\mathrm{d},\mathrm{S}\mathrm{T}\mathrm{P}}$$

(10)

where:

${\mathrm{P}}_{\mathrm{w}\mathrm{i}\mathrm{n}\mathrm{d}}$ = output power (kW)

${\mathrm{P}}_{\mathrm{w}\mathrm{i}\mathrm{n}\mathrm{d},\mathrm{S}\mathrm{T}\mathrm{P}}$ = output power at standard temperature and pressure (kW)

$\mathrm{\rho }$ = air density (kg/m³)

${\mathrm{\rho }}_0$ = air density at STP (1.225 kg/m³)

4.2.2. Solar PV

PV systems convert solar radiation into electricity. The output power of the PV system can be ascertained through the application of Equation (11) below [14,30,32].

$${\mathrm{P}}_{\mathrm{P}\mathrm{V}}={\mathrm{Y}}_{\mathrm{P}\mathrm{V}}{\mathrm{f}}_{\mathrm{P}\mathrm{V}}\left({\displaystyle \frac{{\mathrm{G}}_{\mathrm{T}}}{{\mathrm{G}}_{\mathrm{T},\mathrm{S}\mathrm{T}\mathrm{C}}}}\right)\left(1+{\mathrm{\delta }}_{\mathrm{P}}\left({\mathrm{T}}_{\mathrm{C}}-{\mathrm{T}}_{\mathrm{C},\mathrm{S}\mathrm{T}\mathrm{P}}\right)\right)$$

(11)

where:

${\mathrm{Y}}_{\mathrm{P}\mathrm{V}}$ = rated capacity of PV array (kW)

${\mathrm{f}}_{\mathrm{P}\mathrm{V}}$ = PV derating %

${\mathrm{G}}_{\mathrm{T}}$ = solar radiation (kW/m2)

${\mathrm{G}}_{\mathrm{T},\mathrm{S}\mathrm{T}\mathrm{C}}$ = the incident radiation at standard test condition, STC, (kW/m²)

${\mathrm{\delta }}_{\mathrm{P}}$ = the power coefficient at temperature (%/°C)

${\mathrm{T}}_{\mathrm{C}}$ = cell temperature (°C)

${\mathrm{T}}_{\mathrm{C},\mathrm{S}\mathrm{T}\mathrm{P}}$ = cell temperature under standard temperature and pressure, STP, (25 °C).

4.2.3. Energy Storage System (ESS)

If resources are not available, such as solar PV at night and low wind speed, an ESS can help maintain power supply. So, in this study, an energy storage system (ESS) can be interchangeably labeled as a battery energy storage system or BESS. Equation (12) below calculates the charging process of the battery [25].

$$\mathrm{E}\left(\mathrm{t}\right)=\left(1-\mathrm{\phi }\right)\ast \mathrm{E}\left(\mathrm{t}-1\right)+\left({\mathrm{E}}_{\mathrm{g}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{d}}\left(\mathrm{t}\right)-{\displaystyle \frac{{\mathrm{E}}_{\mathrm{d}\mathrm{e}\mathrm{m}\mathrm{a}\mathrm{n}\mathrm{d}}\left(\mathrm{t}\right)}{{\mathrm{\eta }}_{\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{v}}}}\right)\ast {\mathrm{\eta }}_{\mathrm{C}\mathrm{C}}{\mathrm{\eta }}_{\mathrm{r}\mathrm{b}\mathrm{a}\mathrm{t}}$$

(12)

where:

$\mathrm{E}\left(\mathrm{t}\right)$ = energy storage at time t

$\mathrm{E}\left(\mathrm{t}-1\right)$ = energy storage at time $\left(\mathrm{t}-1\right)$

$\mathrm{\phi }$ = battery self-discharge rate

${\mathrm{E}}_{\mathrm{g}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{d}},\mathrm{t}$= energy generated at time t

${\mathrm{E}}_{\mathrm{d}\mathrm{e}\mathrm{m}\mathrm{a}\mathrm{n}\mathrm{d}},\mathrm{t}$ = energy demand at time t

${\mathrm{\eta }}_{\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{v}}$ = converter efficiency

${\mathrm{\eta }}_{\mathrm{C}\mathrm{C}}$ = charge controller efficiency

${\mathrm{\eta }}_{\mathrm{r}\mathrm{b}\mathrm{a}\mathrm{t}}$ = battery round-trip efficiency

Equation (13) below represents the discharging process of the battery [25].

$$\mathrm{E}\left(\mathrm{t}\right)=\left(1-\mathrm{\phi }\right)\ast \mathrm{E}\left(\mathrm{t}-1\right)+\left({\mathrm{E}}_{\mathrm{g}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{d}}\left(\mathrm{t}\right)-{\displaystyle \frac{{\mathrm{E}}_{\mathrm{d}\mathrm{e}\mathrm{m}\mathrm{a}\mathrm{n}\mathrm{d}}\left(\mathrm{t}\right)}{{\mathrm{\eta }}_{\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{v}}}}\right)-{\displaystyle \frac{{\mathrm{E}}_{\mathrm{g}\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{a}\mathrm{t}\mathrm{e}\mathrm{d}}\left(\mathrm{t}\right)}{{\mathrm{\eta }}_{\mathrm{r}\mathrm{b}\mathrm{a}\mathrm{t}}}}$$

(13)

4.2.4. Modeling of SLB Degradation

A linear degradation model was adopted to represent SLB capacity fade because it offers a practical, transparent, and computationally efficient framework for system-level simulations. While real SLB aging is nonlinear and sensitive to temperature, DOD, and cycling behavior, the linear approximation reflects average long-term trends reported in the literature [5,35,36,37,38]. Its limitations are acknowledged, and future work will incorporate more detailed temperature and cycle-dependent aging curves for improved fidelity. Based on observed annual degradation rates, the refined model quantifies reductions in capacity, round-trip efficiency, and remaining cycle life using the specifications listed below.

• Initial Capacity/SOH (State of Health): 80% [5]
• Capacity Fade per Year: 1.8% [35]
• Round-Trip Efficiency (RTE): 80% [36]
• Round-Trip Efficiency Loss per Year: 0.86% [37]
• Initial Number of Cycles: 2000 cycles [38]
• Cycle Loss per Year: 1.5% [38]
• Depth of Discharge (DOD): 80% [37]

Capacity loss, remaining number of cycles, and round-trip efficiency loss can be calculated by using Equations (14), (15), and (16), respectively.

$$\mathrm{C}\left(\mathrm{t}\right)={\mathrm{C}}_{\left(\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\right)}-(1.8\times \mathrm{t})$$

(14)

where:

C(t): is the capacity of the SLB at time t (in years).

${\mathrm{C}}_{\left(\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\right)}$: is the initial state of health (SOH).

1.8: represents the annual degradation rate of capacity

t: is the time in years.

$$\mathrm{N}\left(\mathrm{t}\right)={\mathrm{N}}_{\left(\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\right)}-(0.02\times \mathrm{t})$$

(15)

where:

N(t): is the remaining number of cycles of the SLB at time t (in years).

${\mathrm{N}}_{\left(\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\right)}$: is the initial number of cycles at t = 0.

0.02: represents the annual loss of cycles (in cycles per year).

t: is the time in years.

$$\mathrm{\eta }\left(\mathrm{t}\right)={\mathrm{\eta }}_{\left(\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\right)}-(0.86\times \mathrm{t})$$

(16)

where:

η(t): is the round-trip efficiency of the SLB at time t (in percentage).

${\mathrm{\eta }}_{\left(\mathrm{i}\mathrm{n}\mathrm{i}\mathrm{t}\mathrm{i}\mathrm{a}\mathrm{l}\right)}$: is the initial round-trip efficiency of the SLB at t = 0.

0.86: represents the annual degradation rate of efficiency (in percentage points per year).

t: is the time in years.

If we consider a default end-of-life battery to have 80% capacity, 2000 cycles, and 80% round-trip efficiency, our findings show that in 15 years of use, the battery will degrade to 53% capacity, 1550 cycles, and 67.1% round-trip efficiency, as shown in Figure 5.

4.2.5. Electrolyzer

The energy conveyed from the electrolyzer to the hydrogen storage tank is denoted as per Equation (17) [39].

$${\mathrm{P}}_{\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{y}\mathrm{z}\mathrm{e}\mathrm{r}/\mathrm{H}\mathrm{T}}=\mathrm{\eta }\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{y}\mathrm{z}\mathrm{e}\mathrm{r}\times {\mathrm{P}}_{\mathrm{r}\mathrm{e}\mathrm{n}/\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{y}\mathrm{z}\mathrm{e}\mathrm{r}}$$

(17)

where:

${\mathrm{P}}_{\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{y}\mathrm{z}\mathrm{e}\mathrm{r}/\mathrm{H}\mathrm{T}}$ = output of the electrolyzer

$\mathrm{\eta }\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{y}\mathrm{z}\mathrm{e}\mathrm{r}$= efficiency of electrolyzer

${\mathrm{P}}_{\mathrm{r}\mathrm{e}\mathrm{n}/\mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{y}\mathrm{z}\mathrm{e}\mathrm{r}}$ = renewable electrical energy supplied to electrolyzer

4.2.6. Hydrogen Storage Tank

The quantity of hydrogen present in HT is governed by specified minimum and maximum thresholds, as given in Equation (18) [39].

$${\mathrm{H}\mathrm{T}}_{\mathrm{m}\mathrm{a}\mathrm{s}\mathrm{s},\mathrm{m}\mathrm{i}\mathrm{n}}\le {\mathrm{H}\mathrm{T}}_{\mathrm{m}\mathrm{a}\mathrm{s}\mathrm{s}}\left(\mathrm{t}\right)\le {\mathrm{H}\mathrm{T}}_{\mathrm{m}\mathrm{a}\mathrm{s}\mathrm{s},\mathrm{m}\mathrm{a}\mathrm{x}}$$

(18)

4.2.7. Hydrogen vs. Battery Storage: Pros and Cons

In off-grid remote regions of Northern Canada, both hydrogen systems and SLB can provide energy storage for EV charging stations. A critical comparison is summarized below:

• Energy Storage Duration

• Hydrogen: suitable for long-duration storage, capable of covering energy needs for days or weeks, particularly useful when renewable generation is seasonal.
• SLB: effective for short to medium-term storage, but performance declines with frequent deep discharges, limiting reliability during prolonged low-generation periods.

2.

Scalability and Energy Independence

• Hydrogen: highly scalable and independent, ensuring uninterrupted power in off-grid locations.
• SLB: can be scaled, but increasing storage requires more battery units, leading to higher replacement frequency due to cell degradation.

3.

Capital and Operational Costs

• Hydrogen: higher upfront costs but longer lifespan (20–25 years) and low degradation; lower long-term operational costs for continuous energy availability despite lower round-trip efficiency (40–60%) [39].
• SLB: lower initial costs but may require replacement every 10–15 years; operational costs depend on charge/discharge cycles and efficiency [36]

4.

Safety and Handling

• Hydrogen: requires high-pressure storage (350–700 bar) with strict safety standards (ISO 16111, ASME, ISO 14687-2) and redundant safety mechanisms [31].
• SLB: safer in terms of storage pressure but can be susceptible to thermal runaway if not properly managed.

5.

Infrastructure and Integration

• Hydrogen: on-site electrolysis eliminates supply chain dependency; infrastructure allows for future hydrogen refueling.
• SLB: simple integration with PV and wind systems; requires careful thermal management and monitoring.

Hydrogen-only systems provide long-term, scalable, and reliable off-grid energy but have higher costs and lower efficiency. Battery-only systems are suitable for short-term storage with lower initial cost but limited lifespan and reduced reliability under deep cycling. A hybrid system leveraging both technologies ensures efficient, reliable, and sustainable energy for off-grid EV charging stations in remote northern regions.

Impact of Extreme Winter Temperatures

Extreme cold significantly affects SLBs, hydrogen systems, and PV production in off-grid EV charging stations. Low temperatures reduce battery efficiency, usable capacity, and lifespan due to higher internal resistance and slower electrochemical reactions [40]. Hydrogen storage and electrolyzer performance decline in cold conditions, requiring insulation or heating to maintain operability [39]. PV modules benefit from lower temperatures, but reduced sunlight, snow cover, and ice accumulation limit energy generation These effects necessitate thermal management, system insulation, and optimized PV mounting to ensure reliable, efficient off-grid operation in northern regions [41,42].

4.2.8. Fuel Cell (FC)

A proton exchange membrane, PEM, and fuel cells are used in this research work. The output power of the FC is contingent upon its overall efficiency and is determined by Equation (19) [39,43].

$${\mathrm{F}\mathrm{C}}_{\mathrm{p}\mathrm{o}\mathrm{w}\mathrm{e}\mathrm{r}}=\mathrm{\eta }\mathrm{F}\mathrm{C}\ast {\mathrm{P}}_{\mathrm{H}\mathrm{T}/\mathrm{F}\mathrm{C}}$$

(19)

where:

${\mathrm{F}\mathrm{C}}_{\mathrm{p}\mathrm{o}\mathrm{w}\mathrm{e}\mathrm{r}}$ = fuel cell power

$\mathrm{\eta }\mathrm{F}\mathrm{C}$ = FC efficiency

${\mathrm{P}}_{\mathrm{H}\mathrm{T}/\mathrm{F}\mathrm{C}}$ = HT power supplied to FC

4.2.9. Converter

The converter is employed within the system to convert electrical energy from AC current to DC current. The power produced from RE source is represented by Equation (20) [44].

$${\mathrm{P}}_{\mathrm{o}\mathrm{u}\mathrm{t}}={\mathrm{P}}_{\mathrm{i}\mathrm{n}}\times {\mathrm{\eta }}_{\mathrm{i}\mathrm{n}\mathrm{v}}$$

(20)

4.3. Optimization Tool Selection

HOMER Pro (v3.18.3) was used to assess the technical, economic, and environmental feasibility of the hybrid RE system. Developed by NREL and enhanced by HOMER Energy, it is widely applied in rural electrification, island microgrids, military installations, and grid-connected systems. HOMER Pro enables hourly time series simulations, capturing stochastic variations in renewable resource availability, EV charging demand, battery cycling, hydrogen production, and component dispatch. This dynamic modeling provides a realistic representation compared to monthly or annual average-based methods.

The software incorporates adaptive control strategies, component degradation, and replacement effects while performing integrated techno-economic optimization. Multiple system configurations can be compared automatically, and sensitivity and scenario analyses are supported, ensuring robust decision making and reproducibility. HOMER Pro simulation workflow for the hybrid RE system is presented in Figure 6 below.

4.4. Economic Analysis

The economic performance of the proposed hybrid RE system is evaluated using multiple financial indicators, including NPC, LCOE, Total Cost of Ownership (TCO), inflation rate, discount rate, and operational expenses. These metrics, combined with detailed capital, replacement, and O&M cost specifications, provide a comprehensive assessment of system affordability, long-term viability, and cost effectiveness. All economic equations are drawn from the widely recognized energy systems literature, and cost parameters were sourced from peer-reviewed studies and manufacturer data to ensure realistic estimates.

4.4.1. Net Present Cost (NPC)

The NPC represents the present value of all capital and lifetime costs associated with establishing the EV charging station, as shown in Equation (21) [26].

$$\mathrm{N}\mathrm{P}\mathrm{C}={\displaystyle \frac{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{a}\mathrm{n}\mathrm{n}\mathrm{u}\mathrm{a}\mathrm{l} \mathrm{c}\mathrm{o}\mathrm{s}\mathrm{t}}{\mathrm{C}\mathrm{R}\mathrm{F}(\mathrm{i},{\mathrm{R}}_{\mathrm{p}\mathrm{r}\mathrm{o}\mathrm{j}\mathrm{e}\mathrm{c}\mathrm{t}})}}$$

(21)

where:

CRF = capital recovery factor

i = interest rate

${\mathrm{R}}_{\mathrm{p}\mathrm{r}\mathrm{o}\mathrm{j}\mathrm{e}\mathrm{c}\mathrm{t}}$ = project total life (year)

4.4.2. Levelized Cost of Energy (LCOE)

The LCOE (or COE) is the ratio of the system’s total annual cost to the total annual energy delivered, representing the unit cost of electricity [25,26], as shown in Equation (22).

$$\mathrm{L}\mathrm{C}\mathrm{O}\mathrm{E}={\displaystyle \frac{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{A}\mathrm{n}\mathrm{n}\mathrm{u}\mathrm{a}\mathrm{l} \mathrm{C}\mathrm{o}\mathrm{s}\mathrm{t}}{\mathrm{T}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{A}\mathrm{n}\mathrm{n}\mathrm{u}\mathrm{a}\mathrm{l} \mathrm{E}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y} \mathrm{S}\mathrm{e}\mathrm{r}\mathrm{v}\mathrm{e}\mathrm{d}}}$$

(22)

4.4.3. Inflation, Discount Rate

Incorporating economic realism, this study assumes an inflation rate of 2% [45] and a discount rate of 6% (6% in this study, based on Canadian mortgage capital investment trends) [46].

4.4.4. Total Cost of Ownership (TCO)

TCO is calculated as the sum of capital, replacement, O&M, and climate-related operational costs over the project lifetime. This allows direct comparison of alternative configurations, highlighting trade offs between high-efficiency but expensive hydrogen storage and lower-cost second-life batteries.

4.4.5. Economic Analysis of System Components

Table 2. Cost of system components.

Component	Capital Cost ($/kW)	Replacement Cost ($/KW)	Operation and Maintenance Cost $/kW Per Year	Lifetime (Years)	Source
Solar PV	640	640	10	25	[47]
Wind Turbine	1000	1000	4	25	[14,30]
Electrolyzer	1500	1200	30	15	[26]
Hydrogen Tank	700/KG	700/kg	0	25	[30]
Fuel Cell	600	500	0.08/h	40,000 h	[39,43]
Converter	300	300	0	15	[29]
Li-Ion Battery	550	550	10	15	[30]

summarizes the capital, replacement, and O&M costs used in the assessment. All values are justified based on the current literature and market data, and critical assumptions such as component lifetimes, maintenance requirements, and usage patterns are explicitly considered.

4.4.6. Performance Adjustments and Degradation Analysis

To enhance accuracy beyond default HOMER Pro settings, key performance adjustments were made. These modifications offer a climate and degradation-aware model, which is more suitable for the Canadian conditions and presented in

Table 3. Derating and degradation parameters for system simulation.

Component	Adjustment Type	Value	Source
Wind Turbine	Winter Icing and Mechanical Losses	4% loss	[48]
	Low Temperature Shutdown Loss	5% loss	[48]
Solar PV	Global Solar Derating Factor Temperature Coefficient Adjustment	9% loss −0.4%/°C	[41] [42]
Battery (SLB)	Annual Capacity Fade Minimum State of Charge (SOC)	1.8% per year 20%	[35] [3]

.

4.4.7. Technical Specifications of System Components

The technical specifications of the solar PV, wind turbine, FC, hydrogen storage tank, electrolyzer, converter, and battery energy storage system are summarized in

Table 4. Components technical description.

Component	Model	Unit Capacity	Efficiency (%)	Lifetime	Source
Wind Turbine	Lagerwey LW30/250	250 kW		25 years	[30]
Solar PV	Longi Solar LR6-72	350 W	18.10	25 years	[47]
Electrolyzer	Proton Exchange Membrane	1 kW	85	15 years	[26]
Fuel Cell	Proton Exchange Membrane	1 kW	45	40,000 h	[43]
Converter	Generic System Converter	1 kW	95	15 years	[29]
Hydrogen Tank	Hydrogen Storage Tank	1 kg		25 years	[26]
Energy Storage System	Power Wall	1 kWh		15 years	[30]

.

5. Meteorological Data

Meteorological data from the two chosen locations have been examined, owing to their varying geographical and climatic conditions. The meteorological data encompass Global Horizontal Irradiance (GHI), wind speed, and temperature.

5.1. Selection of Site

This study focuses on two regions: Windsor, Ontario, and Eagle Plains along the Dempster Highway between the Yukon and Northwest Territories. Windsor, located on the southern bank of the Detroit River opposite Detroit, is Canada’s southernmost city and features strong solar and wind potential with the country’s highest temperatures. The second site, Eagle Plains, lies near the midpoint of the 734 km Dempster Highway from Dawson (YT) to Inuvik (NWT) [12]. This remote location is off-grid and relies on diesel generators for electricity. The site coordinates are listed in

Table 5. Geographic coordinates and classification of selected sites [49].

Site	Province/Territory	Latitude (°N)	Longitude (°W)	Type of Region
Windsor	Ontario	42.17°	82.53°	Urban (Base Case)
Eagle Plains	Yukon Territories	66°22	136°43	Remote (Off Grid)

, with locations shown in Figure 7 and Figure 8.

5.2. Annual Average Global Horizontal Irradiance (GHI)

Utilizing information from the NREL database, a bar graph has been generated and is presented in Figure 9. The annual average solar GHI for Windsor is 3.82 kWh/m²/day, and Eagle Plains is 2.58 kWh/m²/day [49].

5.3. Wind Speed Data

From the NREL, data on wind speed, at a height of 50 m above the earth’s surface, are collected. As illustrated in the graph presented in Figure 9, the annual average wind speed for Windsor is 7 m/s, and Eagle Plains is 5.31 m/s [49].

5.4. Temperature Data

By using the dataset from the NREL, accessed through HOMER Pro, the annual average air temperature is reported as follows: Windsor, 9.1 °C; and Eagle Plains is −8.02 °C [49], as shown in Figure 9.

5.5. Climatic Data Integration in HOMER

All climatic data were incorporated into HOMER using its built-in integration with the National Renewable Energy Laboratory (NREL) databases, including the National Solar Radiation Database (NSRDB) and the Wind Toolkit. After entering the precise geographic coordinates for Windsor and Eagle Plains, HOMER automatically downloaded the required hourly (8760-point) datasets for GHI, wind speed, and temperature in the appropriate file format. The imported datasets were then checked within HOMER’s resource interface to ensure completeness and internal consistency, ensuring an accurate representation of local climatic conditions. This automated workflow eliminated the need for manual preprocessing and guaranteed reliable integration of meteorological data into all hybrid system simulations.

6. EV Arrival Rate and Load Estimation

Searching for data about traffic flow, we were unsuccessful in securing statistically significant data. The datapoints obtained from the city of Windsor are limited to only five days, covering twenty-four hours per day. No data were available for the Eagle Plains location. This type of data, if available in a significant number of datapoints, would help determine the number of charging ports and consequently the load demand. Thus, due to this limitation, we consider this part of the charging station study as conceptual rather than complete. The arrival rate to the charging station is assumed Poisson distributed and roughly estimated from the datapoints obtained from the city of Windsor; this rate is equal to 5.990 EV per hour, or the highest value shown in

Table 6. EV arrival rate and load profile per day.

Hour	Traffic Flow (Vehicles) per Hour	EV Arrival Rate ${\mathsf{\lambda }}_{\mathbf{EV}}\left(\mathbf{t}\right)={\mathbf{p}}_{\mathbf{EV}} \mathbf{Flow}\left(\mathbf{t}\right)$ ${\mathbf{p}}_{\mathbf{EV}}=\frac{72}{\mathrm{62,897}}$ $=0.0011447$	Hourly Load (EV/day) × 50 kWh
1.0	190	0.217	10.896
2.0	152	0.174	8.717
3.0	284	0.325	16.287
4.0	322	0.369	18.467
5.0	1314	1.504	75.357
6.0	2830	3.240	156.621
7.0	4656	5.330	267.019
8.0	4874	5.579	279.521
9.0	3564	4.080	204.394
10.0	3074	3.519	176.292
11.0	3404	3.897	195.218
12.0	3380	3.869	193.841
13.0	3396	3.887	194.759
14.0	4210	4.819	241.441
15.0	5231	5.988	300.500
16.0	5221	5.977	294.900
17.0	5233	5.990	295.300
18.0	3460	3.961	198.429
19.0	2466	2.823	141.424
20.0	1846	2.113	104.433
21.0	1441	1.650	82.641
22.0	1398	1.600	80.175
23.0	489	0.560	28.044
24.0	462	0.529	26.45
Sum	62,897/day	72 EVs/day	3591.126/day

below. This assumption/estimation of highest (peak) arrival rate can be explained in comparison to current gasoline stations, which have unused gas pumps during low traffic volume or at night and in early morning hours. Also, we assume that the rate of EV capture will increase as more electric vehicles will be driven in the future. The assumptions of Poisson distribution of arrival rate and exponential service (charging) time are well reported in the literature [50,51,52,53,54]. The charging time or service time is assumed to be exponentially distributed with rate = 20–30 min per vehicle, which translates to approximately 2.5 vehicles [55].

The hourly traffic flow at the proposed site is shown in Table 6 [56]. Considering a 3.4% EV penetration rate in Canada [57] and applying a 4% market capture rate, the proposed charging station is expected to serve approximately 72 EVs per day. The hourly arrival rate depends on EV probability of capture and the total vehicle flow, which is described in Equation (23):

$${\mathrm{\lambda }}_{\mathrm{EV}}\left(\mathrm{t}\right)={\mathrm{p}}_{\mathrm{EV}} \mathrm{Flow}\left(\mathrm{t}\right)$$

(23)

where:

${\mathrm{\lambda }}_{\mathrm{EV}}\left(\mathrm{t}\right)$ = EV arrival rate at time $\mathrm{t}$,

${\mathrm{p}}_{\mathrm{EV}}$ = probability of a vehicle being an EV

$\mathrm{Flow}\left(\mathrm{t}\right)$ = total vehicle flow at time $\mathrm{t}$.

The probability of an individual EV requiring charging is determined using the ratio of the total number of EVs charged in a day to the total daily traffic flow, expressed in Equation (24):

$$\mathrm{p}={\displaystyle \frac{{\mathrm{N}}_{\mathrm{E}\mathrm{V}}}{\sum _{\mathrm{i}=1}^{24}{\mathrm{V}}_{\mathrm{i}}}}$$

(24)

where:

${\mathrm{N}}_{\mathrm{E}\mathrm{V}}=$ total number of EVs charged per day

${\mathrm{V}}_{\mathrm{i}}$ = total number of vehicles during hour $\mathrm{i}$.

The probability $“\mathrm{p}”$ represents the fraction of vehicles within the hourly traffic stream that are expected to be EVs requiring fast charging at the proposed station. Using this probability, the hourly charging load, and the corresponding number of EVs charged, was computed directly from the observed hourly traffic flow and the resulting arrival rates, as presented in Table 6.

Modeling of Charging Port (Estimation)

The number of charging ports (servers) required at the proposed fast charging station was modeled using queuing theory to ensure efficient service and minimal waiting times. The following parameters were adopted based on observed traffic data and the modeled charging demand, which was derived from the hourly vehicle arrival patterns over a 24-h period.

• Arrival rate (λ): 5.990 vehicles per hour, representing the peak observed flow during busy periods (Table 1).
• Battery capacity: 64 kWh, consistent with mid-size EV models such as Hyundai Kona, Kia Niro, and Kia Soul [58].
• Desired average waiting time (Wq): ≤10 min.
• Service rate per charging port (μ): corresponding to 20–30 min per vehicle or approximately 2.5 vehicles/h [55].
• The charging system is modeled as an M/M/C:FCFS/∞/∞ queue system.

The assumption of Poisson arrivals and exponential service times aligns with the aforementioned EV charging queuing studies, which commonly employ M/M/C models to evaluate waiting times, charger utilization, and optimal station sizing. Using this framework, the required number of charging ports or chargers was determined for a peak arrival rate of λ = 5.990 vehicles/hour and average service rate of μ = 2.5 vehicles/h; the system stability condition ($\mathrm{\rho }=\mathrm{\lambda }/\left(\mathrm{C}\mathrm{\mu }\right)<1$) is satisfied with C = 4 chargers, resulting in an average waiting time of approximately 4.27 min, well below the target threshold of 10 min. This provides a thorough analytical basis for sizing fast charging infrastructure under stochastic arrival conditions while ensuring high service quality and minimal queue delays. To summarize, if statistically representative datapoints become available, the same framework remains applicable whether the model is an M/M/C or another queuing model, and the number of charging ports and load profile can be established accordingly for the selected sites.

7. Results and Discussion

This section presents a detailed performance assessment of the proposed off-grid EV fast charging systems in Windsor (base case) and Eagle Plains (remote northern site). Using HOMER Pro’s optimization engine, each location was evaluated under its unique resource, climatic, and load conditions. The analysis is structured to (i) interpret numerical outputs, (ii) compare performance drivers between the two sites, and (iii) discuss sensitivities to key variables such as solar irradiance, wind availability, temperature, battery degradation, and hydrogen subsystem efficiency.

7.1. System Optimization and Component Sizing

The optimized system configurations generated by HOMER Pro illustrate how each site’s resource profile influences the techno-economic design of the off-grid EV charging system.

7.1.1. Windsor (Base Case)

Windsor was selected as the base case to determine the optimal sizing of the off-grid EV charging system components. The corresponding design variables and model input parameters used in the HOMER Pro optimization are summarized in

Table 7. Model input parameters and designs variable values.

Component/Parameter	Input Values
Solar PV	350 kW
Wind Turbine	100 kW
Fuel Cell	100 kW
Electrolyzer	100 kW
Converter	319 kW
Hydrogen Tank	100 kg
Wind Turbine Hub Height	50 m
SLB Capacity	64 kWh
SLB State of Health	80%
State of Charge (min)	20%
State of Charge (max)	75%
Energy Storage System	400 kWh
PV Derating Factor	9%
PV Temperature Coefficient	−0.4%/°C
Wind Turbine Icing + Shutdown Losses	4% + 5%
Battery Capacity Fade	1.8%/year
Annual Load Demand	1,310,761 kWh/year
Estimated Number of Charging Ports	4
Estimated Number of Cars Charged/Day	72
Dispatch Strategy	LF or CC
Inflation Rate	2%
Discount/Interest Rate	6%
Project Lifetime	25 years

.

7.1.2. Eagle Plains (Remote Northern Site)

Eagle Plains exhibits harsher climatic constraints: GHI of 2.58 kWh/m²/day, lower wind speeds (5.31 m/s), and extreme cold conditions with an average −8.02 °C. These conditions lead to:

• reduced PV efficiency due to low winter irradiance.
• increased wind turbine losses due to icing.
• higher battery degradation stress.
• greater reliance on fuel cells and hydrogen storage.

As a result, HOMER Pro shifts the system toward greater fuel cell utilization and more frequent operation of the electrolyzer during periods of excess generation. Battery behavior is strongly influenced by climate and SOC limits (20–75%), making the hydrogen subsystem essential for energy shifting.

7.2. Comparative Performance Analysis

This section compares the hybrid energy systems at Windsor and Eagle Plains, with detailed site-specific performance discussed in the following subsections.

7.2.1. Renewable Penetration and Resource Utilization

• Windsor: achieves higher renewable contribution due to abundant solar resources. PV accounts for a substantial fraction of annual energy generation, reducing reliance on hydrogen conversion pathways.
• Eagle Plains: shows lower renewable penetration because both solar and wind resources are weaker, and winters impose long periods of low irradiance.

7.2.2. Hydrogen System Behavior

Hydrogen storage is used differently across sites:

• Windsor: Hydrogen functions primarily as a seasonal buffer and supplementary storage. Fuel cell operation is used for peak shaving rather than continuous supply.
• Eagle Plains: The fuel cell becomes a major energy supplier during prolonged low renewable periods. Hydrogen demand increases due to reduced PV and wind availability.

This difference significantly affects COE and NPC, with the hydrogen subsystem contributing a larger share to total system costs in Eagle Plains.

7.2.3. Battery Degradation and Temperature Effects

In both sites, second-life batteries degrade at 1.8% annually; however:

• Windsor maintains mild operating temperatures, limiting performance derating and extending usable capacity.
• Eagle Plains requires more conservative SOC operation and experiences greater efficiency losses in extreme cold, amplifying degradation related impacts on economic performance.

7.3. Economic Results

Using Windsor as the base case, the techno-economic performance of the hybrid system was evaluated for Eagle Plains under a uniform annual load of 1,310,761 kWh/year. Windsor recorded the lowest costs, with an NPC of $2.80 million, COE of $0.201/kWh, OPEX of $56,067/year, CAPEX of $1.92 million, and annual generation of 1,315,038 kWh. In contrast, Eagle Plains showed higher costs NPC of $3.61 million, COE of $0.259/kWh, OPEX of $77,714/year, and CAPEX of $2.39 million despite comparable energy output (1,329,709 kWh). These differences are driven by variations in renewable resource potential and climatic conditions across the regions.

Table 8. Cost metrics and energy production results for Windsor and Eagle Plains.

Location	NPC (M$)	COE ($/kWh)	CAPEX (M$)	OPEX ($/year)	Energy Production (kWh/year)
Windsor	2.80	0.201	1.92	56,067	1,315,038
Eagle Plains	3.61	0.259	2.39	77,714	1,329,709

summarizes the techno-economic results.

Figure 10 presents a comparative analysis of the NPC, COE, and CAPEX across the two locations.

7.4. Economic Impact of Equipment Capacity Expansion

To evaluate the economic implications of scaling system components, the hybrid EV charging system was analyzed under a uniform annual load of 1,310,761 kWh/year, with Windsor serving as the base case and Eagle Plains representing a remote, resource constrained location. In Windsor, higher solar irradiance (3.82 kWh/m²/day) and wind speed (7.0 m/s) enable solar PV and wind turbines to dominate energy production, reducing reliance on backup systems such as fuel cells, electrolyzers, and energy storage. Conversely, Eagle Plains’ lower solar irradiance (2.58 kWh/m²/day), lower wind speeds (5.31 m/s), and colder temperatures (−8.0 °C average) necessitate larger capacities of hydrogen storage, battery energy storage, and auxiliary generation to maintain supply reliability.

Table 9. Economic assessment of system component sizing.

Location	NPC (M$)	COE ($/kWh)	CAPEX (M$)	OPEX ($)	Energy Production (kWh/year)
Eagle Plains	0.81	0.058	0.47	21,647	1,329,709

summarizes the resulting techno-economic metrics for Eagle Plains under these conditions.

By maintaining a constant load across proposed locations, the optimal component capacities were determined for Eagle Plains to meet the demand while accounting for variations in renewable resource availability and climatic conditions, as presented in

Table 10. Optimized component sizes for Windsor and Eagle Plains.

Component	Unit	Windsor	Eagle Plains
Solar PV	kW	350	430
Wind Turbine	kW	100	200
Fuel Cell	kW	100	250
Electrolyzer	kW	100	350
Converter	kW	319	363
Hydrogen Tank	kg	100	350
Energy Storage System	kWh	400	1000

.

The annual energy contribution of each source for the selected locations is summarized in

Table 11. Annual energy production of each source.

Component	Unit (kWh/yr)	Windsor	Eagle Plains
Solar PV	(kWh/yr)	771,436	318,515
Wind Turbine	(kWh/yr)	412,240	769,913
Fuel Cell	(kWh/yr)	131,362	241,281
Total	(kWh/yr)	1,315,038	1,329,709

.

Regional resources and climate strongly influence hybrid EV charging system sizing and performance. Windsor’s favorable conditions enable smaller, efficient configurations, while Eagle Plains requires larger capacities for reliability. Sensitivity analyses highlight the effects of PV lifetime, wind turbine height, battery SOC limits, hydrogen costs, discount rate, and EV load on system outcomes. Scenario comparisons provide critical insights for designing resilient, site specific off-grid EV charging systems across diverse Canadian regions.

7.5. Revenue Analysis of EV Charging Station

To charge an EV, a DC fast charger is highly effective, capable of recharging the battery from 10% to 80% in approximately 20–30 min [55]. In Canada, two pricing models are commonly used: time-based pricing ($0.27–$0.60 per minute) and energy-based pricing ($0.60–$0.70 per kWh) [59]. The major DC fast charging networks operating across Canada include Petro Canada EV Fast Charging Network, FLO Ultra-Fast Chargers, Electrify Canada, and the Tesla Supercharger Network [60].

The estimated annual revenue generated by the proposed EV charging station for each provider and province (considering 72 EVs charged per day, 50 kWh per EV, and the applicable energy-based rates) is presented in Figure 11.

7.6. Environmental and Lifecycle Impact Analysis

To assess the full environmental implications of the proposed standalone EV charging station, this study expands the analysis beyond direct CO₂ avoidance to include lifecycle emissions, critical material recovery, operational emissions, and carbon pricing impacts. These considerations offer a more comprehensive evaluation of the system’s sustainability performance and alignment with global climate and resource efficiency goals.

7.6.1. Battery Reuse and Emissions Reduction

Lithium-ion EV batteries have a high manufacturing footprint, with a cradle-to-gate global warming potential, GWP, of ~13 t CO₂-eq per unit [61]. This study uses 64 kWh second-life batteries [58] to form the energy storage systems, requiring 400 kWh at Windsor and 1000 kWh at Eagle Plains, yielding significant emission savings.

The number of SLBs required for each system is derived using Equation (25):

$$\mathrm{N}\mathrm{u}\mathrm{m}\mathrm{b}\mathrm{e}\mathrm{r} \mathrm{o}\mathrm{f} \mathrm{S}\mathrm{L}\mathrm{B}(\mathrm{N})={\displaystyle \frac{{\mathrm{E}}_{\mathrm{s}\mathrm{t}\mathrm{o}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}}{{\mathrm{E}}_{\mathrm{b}\mathrm{a}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{y}}}}$$

(25)

where:

${\mathrm{E}}_{\mathrm{s}\mathrm{t}\mathrm{o}\mathrm{r}\mathrm{a}\mathrm{g}\mathrm{e}}=$ the total energy storage capacity (kWh)

${\mathrm{E}}_{\mathrm{b}\mathrm{a}\mathrm{t}\mathrm{t}\mathrm{e}\mathrm{r}\mathrm{y}}$= SLB capacity (64 kWh).

The emission savings for each site are calculated by using Equation (26) and presented in

Table 12. Emission savings from SLB integration.

Sites	ESS (kWh)	Number of SLB	GHG Emissions Saving (Tons)
Windsor	400	6	78
Eagle Plains	1000	16	208

.

$$\mathrm{Emission} \mathrm{Savings} =\mathrm{N}\times 13 \mathrm{t}\mathrm{o}\mathrm{n}\mathrm{s}(\mathrm{c}\mathrm{r}\mathrm{a}\mathrm{d}\mathrm{l}\mathrm{e}-\mathrm{t}\mathrm{o}-\mathrm{g}\mathrm{a}\mathrm{t}\mathrm{e} \mathrm{G}\mathrm{W}\mathrm{P})$$

(26)

Repurposing second-life EV batteries instead of manufacturing new ones yields emission savings ranging from 78 tons CO₂-eq (Windsor) to 208 tons CO₂-eq (Eagle Plains).

This represents a measurable reduction in embedded emissions within the energy storage subsystem. Such reuse supports the principles of circular economy by extending the lifecycle of existing batteries and reducing the demand for energy and resource incentives for new batteries production.

7.6.2. Material Recovery and Resource Circularity

LIBs also contain strategically critical and resource-intensive materials. Based on the LIB chemistry, the materials Cobalt, Nickel, and Lithium can be recovered from end-of-life EV batteries through recycling processes. The quantity of material that can be recovered is 116 g/kWh Cobalt, 400 g /kWh Nickel, and 73 g /kWh Lithium [61]. The recoverable mass of each material can be calculated by using Equation (27) and is presented in

Table 13. Recoverable materials from SLBs in selected locations.

Locations	ESS (kWh)	Cobalt (kg)	Nickel (kg)	Lithium (kg)
Windsor	400	46.4	160	29.2
Eagle Plains	1000	116	400	73

.

$$\mathrm{Mass} \mathrm{of} \mathrm{Material} \left(\mathrm{kg}\right)={\displaystyle \frac{ \mathrm{ESS} \mathrm{Capacity} \left(\mathrm{kWh}\right)\times \mathrm{Material} \mathrm{Factor} (\mathrm{g}/\mathrm{kWh})}{1000}}$$

(27)

Recovering these materials through recycling reduces environmental harm from mining, supports responsible material sourcing, and aligns with SDG 12 (Responsible Consumption and Production) by extending the useful life of embedded resources.

7.6.3. Operational Emissions and Carbon Tax Implications

In addition to lifecycle benefits, the proposed RE system substantially reduces annual operational CO₂ emission and associated carbon tax costs. Regional grid emission factors of 35 g CO₂/kWh (Ontario), and 70 g CO₂/kWh (Yukon Territories) were adopted based on data from the Canada Energy Regulator (2024) [62].

The annual electricity demand of the EV charging station is 1,310,761 kWh/year. The corresponding annual CO₂ emissions were calculated using Equation (28), while the economic implications were estimated by applying the 2025 Canadian carbon tax rate of $0.095 per kg CO₂ [63]. The findings are summarized in

Table 14. Estimated annual CO₂ emissions and carbon tax by region.

Locations	Provence/ Territories	Grid Emission Factor (g CO2/kWh)	Annual Energy Demand (kWh)	Annual Emission (Tons CO2)	Annual Carbon Tax ($)
Windsor	Ontario	35	1,310,761	45.99	4369
Eagle Plains	Yukon Territories	70	1,310,761	91.99	8739

.

$$\mathrm{Annual} {\mathrm{CO}}_2 \mathrm{Emission} (\mathrm{t}/\mathrm{year})={\displaystyle \frac{\mathrm{Annual} \mathrm{Energy} \mathrm{Demand} \left(\mathrm{kWh}\right)\times \mathrm{Emission} \mathrm{Factor} (\mathrm{g} {\mathrm{CO}}_2/\mathrm{kWh})}{\mathrm{1,000,000}}}$$

(28)

The results summarized in Table 14 show the estimated annual CO₂ emissions and potential carbon tax costs for Windsor and Eagle Plains. Adopting a 100% renewable energy-based charging system can completely eliminate these emissions and related costs, demonstrating strong contributions toward SDG 13 (Climate Action).

7.7. Policy Framework and Sustainable Development Goals (SDGs)

This section places this study’s findings within Canada’s clean energy transition, highlighting only the policy implications directly supported by system performance results.

• Regulatory Alignment

Zero-emission operation in Windsor and Eagle Plains supports federal commitments under the 2030 Emissions Reduction Plan and RE programs for remote regions, showing that off-grid EV charging can meet demand without grid or diesel-related emissions and costs.

2.

Infrastructure and Technology Development

The proven feasibility of integrating solar, wind, hydrogen, and second-life EV batteries aligns with Canada’s Hydrogen Strategy and circular economy policies. Sensitivity results showing reduced NPC and COE with higher turbine hub heights and longer PV lifetimes underscore the value of technology standards and incentive structures that enhance system durability.

3.

Environmental and Community Considerations

The proposed system fully eliminates operational emissions, avoiding 45.99 to 91.99 tons of CO₂ per year, compared to grid-based charging supporting national climate action policies and sustainable land-use planning for remote and Indigenous communities.

4.

Stakeholder Engagement

Higher storage requirements and cost premiums in northern climates highlight the importance of co-development with Indigenous and local communities, supported by government funding and technical partners. This ensures appropriate siting, long-term operation, and community benefit, in line with federal engagement frameworks [64].

SDG Contributions

• SDG 7: (Affordable and Clean Energy): zero-emission charging and avoidance of carbon tax costs provide clean and economically competitive energy access.
• SDG 9: (Industry, Innovation, and Infrastructure): the system offers a scalable off-grid model for regions where conventional grid expansion is impractical.
• SDG 13: (Climate Action): complete elimination of operational CO₂ emissions directly supports national decarbonization targets.

7.8. Contributions to Stakeholders

This research presents significant insights for stakeholders, including energy planners, utilities, technology developers, and communities with concrete data on costs, energy output, and GHG reductions to support informed investment, planning, and policy decisions. By analyzing both remote northern and urban southern regions of Canada, it highlights region-specific opportunities and constraints, enabling strategies tailored to NPC, COE, OPEX, CAPEX, and emissions, thereby enhancing the feasibility of a nationwide energy transition.

7.9. Contributions to Policymakers

Policymakers shape regulations and incentives that guide energy and infrastructure development. In remote regions, uncertainty about feasibility can hinder support for EV adoption. Demonstrating the self-sufficiency and reliability of standalone EV charging systems can increase policymakers’ confidence in promoting sustainable initiatives. Evidence of system performance under extreme conditions also provides insights that can be applied to other regions and policy contexts.

Recommendations for Policymakers and Stakeholders

Based on the findings, this study proposes the following policy actions:

• Incentivize Modular Off-Grid Charging: provide subsidies, grants, or low-interest loans for modular EV charging systems in remote areas, and streamline permitting for standardized designs.
• Support Second-Life Batteries: offer tax credits or reduced import duties for using SLBs in energy storage, lowering capital costs and promoting circular economy practices.
• Fund Climate-Resilient Deployment: establish funding to offset costs of cold climate technologies such as battery thermal management, insulated enclosures, and hybrid system setups.
• Invest in Local Workforce Development: collaborate with Indigenous communities and rural institutes to train skilled personnel for installation, monitoring, and maintenance, enhancing reliability and local ownership.
• Integrate into Clean Energy Strategy: use techno-economic and environmental data (NPC, COE, emission reductions) to guide clean mobility policies, especially in areas where grid extension is unfeasible.

Equipping policymakers with evidence-based insights on standalone EV charging systems supports informed decision-making, advancing Canada’s Net Zero Emissions goals and UN SDGs 7, 9, and 13.

7.10. Challenges in Remote Areas

Deploying off-grid EV charging stations in northern Canada faces significant technical, environmental, and logistical challenges. Isolation from the national grid, extreme cold, heavy snowfall, and limited winter sunlight reduce solar and battery efficiency. Rugged terrain and limited infrastructure increase transportation, installation, and maintenance costs, while access to skilled labor and replacement parts is constrained, further complicating system reliability.

Mitigation Strategies

To address these challenges, several mitigation strategies are proposed:

• Modular, Prefabricated Systems: containerized, plug and play designs allow preassembly, testing, and rapid deployment, reducing logistical complexity.
• Cold Climate Technology Adaptation: use insulated enclosures, battery heaters, cold-optimized PV modules, and hybrid PV/Wind/Fuel Cell systems to ensure reliable energy supply in harsh conditions.
• Local Workforce Training: partner with community colleges or Indigenous organizations to train technicians for diagnostics, maintenance, and repairs, fostering sustainability and community ownership.
• Remote Monitoring and Predictive Maintenance: IoT-based systems enable real-time tracking of energy production, battery status, and fault detection, minimizing site visits and operational costs.
• Strategic Partnerships and Incentives: collaborate with governments, clean energy agencies, and Indigenous communities to share investment costs, provide subsidies, and ensure culturally and economically viable solutions.

These strategies collectively improve resilience, cost-effectiveness, and scalability of off-grid EV charging systems in remote regions.

7.11. Sensitivity Analysis

A sensitivity analysis using Windsor, Ontario as the base case evaluates the impact of variations in solar PV lifetime, wind turbine hub height, EV load demand, hydrogen fuel cost, discount rate, and battery SOC limits on system performance and is shown in

Table 15. Sensitivity analysis.

Cases	Wind Hub Height (m)	Solar PV Life Time (years)	Battery SOC Limits (%)	Discount Rate (%)	Hydrogen Fuel Cost ($/kg)	EV Load Demand (kWh)	SLB Deg Life Time (Years)	Scaled Temp °C	Inflow Rate ELz Efficiency (%)	Net Present Cost (NPC) (M$)	Cost of Energy (COE) ($/kWh)	Operating Cost ($/Year)
Case 1	31.8	20	10–100	6	0	3591.12	15	9.1	85	2.86	0.205	59,392
Case 2	50	20	10–100	6	0	3591.12	15	9.1	85	2.85	0.204	58,938
Case 3	50	22	10–100	6	0	3591.12	15	9.1	85	2.83	0.203	57,674
Case 4	31.8	22	20–100	6	0	3591.12	15	9.1	85	2.84	0.204	58,128
Case 5	31.8	25	10–100	6	0	3591.12	15	9.1	85	2.81	0.202	56,521
Case 6	50	25	10–100	6	0	3591.12	15	9.1	85	2.80	0.201	56,067
Case 7	31.8	25	10–80	6	0	3591.12	15	9.1	85	2.81	0.202	56,521
Case 8	50	25	20–80	6	0	3591.12	15	9.1	85	2.80	0.201	56,067
Case 9	50	25	20–80	5	0	3591.12	15	9.1	85	2.91	0.188	56,306
Case 10	50	25	20–80	7	0	3591.12	15	9.1	85	2.71	0.215	55,550
Case 11	50	25	20–80	6	0.1	3591.12	15	9.1	85	2.88	0.207	60,798
Case 12	50	25	20–80	6	0.5	3591.12	15	9.1	85	3.18	0.228	79,571
Case 13	50	25	20–80	6	0	3400.20	15	9.1	85	3.10	0.242	75,082
Case 14	50	25	20–80	6	0	800.20	15	9.1	85	3.26	0.216	84,319
Case 15	50	25	20–80	6	0	3591.12	20	9.1	85	2.29	0.164	39,766
Case 16	50	25	20–80	6	0	3591.12	10	9.1	85	3.07	0.22	89,044
Case 17	50	25	20–80	6	0	3591.12	15	−10	85	2.55	0.183	56,115
Case 18	50	25	20–80	6	0	3591.12	15	30	85	2.55	0.183	56,115
Case 19	50	25	20–80	6	0	3591.12	15	9.1	80	2.81	0.201	56,098
Case 20	50	25	20–80	6	0	3591.12	15	9.1	90	2.55	0.183	56,099

.

• Wind Turbine Hub Height: increasing hub height from 31.8 m (Case 1) to 50 m (Case 3) improves wind capture, lowering NPC from $2.86 million to $2.83 million, COE from $0.205 to $0.203/kWh, and OPEX from $59,392/year to $57,674/year.
• Solar PV Lifetime: extending PV lifetime from 20 years (Case 2) to 25 years (Case 6) reduces NPC from $2.85 million to $2.80 million, COE from $0.204 to $0.201/kWh, and OPEX from $58,938/year to $56,067/year.
• Battery SOC Limits: reducing upper SOC from 100% (Case 5) to 80% (Case 7) has negligible effect on economics (NPC $2.81 million, COE $0.202/kWh), primarily improving battery longevity.
• Discount Rate: lowering the rate from 6% (Case 8) to 5% (Case 9) increases NPC to $2.91 million; raising it to 7% (Case 10) reduces NPC to $2.71 million but increases COE to $0.215/kWh.
• Hydrogen Fuel Cost: introducing hydrogen costs raises NPC and COE. At $0.10/kg (Case 11), NPC = $2.88 million, COE = $0.207/kWh; at $0.50/kg (Case 12), NPC = $3.18 million, COE = $0.228/kWh, OPEX = $79,571/year.
• EV Load Demand: reducing annual demand from 3800 kWh (Case 14) to 3400 kWh (Case 13) lowers NPC from $3.26 million to $3.10 M, OPEX from $84,319/year to $75,082/year, while COE increases from $0.216 to $0.242/kWh.
• SLB Degradation / Useful Life: SLB lifetime is highly influential: extending SLB life (Case 15) delivers the largest reductions in NPC of $2.29 million and COE of $0.164/kWh, while a shorter life (Case 16) raises NPC to $3.07 million and OPEX to $89,044/year, confirming SLB aging as a critical sensitivity.
• Ambient temperature: temperature shifts (−10 °C or +30 °C in Cases 17 and 18) change NPC of $2.55 million and COE of $0.183/kWh relative to baseline, indicating that extreme climates meaningfully affect component performance and operating costs by $56,115/year.
• Electrolyzer efficiency: raising electrolyzer efficiency from 80% to 90% (Case 20) reduces NPC of $2.55 million and COE of 0.183/kWh; showing electrolyzer performance is an important lever for reducing hydrogen pathway costs.

7.12. Comparative Analysis

A comparison between this study and recent research is shown in

Table 16. Comparison with other research studies.

Sr. No	Author	Year	Country	Configuration	COE($/kWh)	NPC ($Million)
1	[24]	2025	Bangladesh	PV/WT/Elz/BG/FC	0.163	-
2	[65]	2025	Saudi Arabia	PV/Battery/Inv	-	0.823–1.05
3	[66]	2025	South Korea	PV/Battery/Inv	-	0.192
4	[25]	2025	Türkiye	PV/BG/Elz/HT/Inv/Grid	0.0215	0.6112
5	[28]	2024	Malaysia	PV/WT/Elz/Battery/FC/HT	0.03–0.16	1.4–3.4
6	[20]	2024	Bangladesh	Grid/PV/WT/Battery	0.11	0.4706
7	[26]	2024	Pakistan	PV/WT/FC	0.39–0.56	14.6–21
8	[31]	2023	Türkiye	PV/WT/Battery/DG	0.441–0.512	7.24–8.9
9	[67]	2023	Saudi Arabia	PV/Battery/Grid	0.064–0.403	10.33–35.7
10	[30]	2022	Qatar	PV/WT/Battery/FC/HT/BG/Elz	0.285–0.329	2.53–2.92
11	[68]	2023	Oman	PV/WT/DG/Grid	0.257–0.566	14.9–31.02
	This Study		Canada	PV/WT/Battery/FC/Elz/HT	0.201–0.259	2.80–3.61

.

Table 16 compares the proposed Canadian hybrid system (PV/WT/Battery/FC/Elz/HT) with international studies. The system achieves a COE of $0.201–0.259/kWh and NPC of $2.80–3.61 million, reflecting economic viability under Canadian conditions. While [25,28] report lower COE due to favorable resources and grid integration, Refs. [26,30] show higher COE and NPC due to fuel cell costs and limited renewable potential. The proposed system balances cost, reliability, and sustainability, with multi-source RE and hydrogen/battery storage enabling autonomous operation in both cold climate and remote off-grid EV charging contexts, supporting Canada’s clean transportation and emissions reduction goals.

8. Conclusions

This study presents a comprehensive techno-economic and environmental assessment of off-grid EV charging stations at two Canadian locations, Windsor and Eagle Plains. The proposed hybrid energy system integrates solar PV, wind turbines, fuel cells, electrolyzers, hydrogen storage, converters, and SLB to meet EV charging demand under diverse renewable resource and climatic conditions.

The results demonstrate the significant impact of site-specific resources on system design and performance. Windsor, with higher solar irradiance (3.82 kWh/m²/day) and wind speeds (7 m/s), required smaller storage and hydrogen capacities, achieving higher system efficiency, a lower COE of $0.201/kWh, NPC of $2.80 million, CAPEX of $1.92 million, OPEX of $56,067/year, and annual energy production of 1,315,038 kWh. In contrast, Eagle Plains, with lower solar input (2.58 kWh/m²/day), reduced wind speeds (5.31 m/s), and cold temperatures (−8.02 °C), required larger storage and hydrogen capacities to maintain reliability, resulting in lower operational efficiency, higher COE of $0.259/kWh, NPC of $3.61 million, CAPEX of $2.39 million, OPEX of $77,714/year, and annual energy production of 1,329,709 kWh.

The integration of second-life EV batteries demonstrated strong effectiveness as an energy storage solution, reducing lifecycle emissions by 78 tons CO₂-eq in Windsor and 208 tons CO₂-eq in Eagle Plains. In addition to emission reductions, these batteries facilitate the recovery of critical materials: 46.4 kg Co, 160 kg Ni, and 29.2 kg Li in Windsor, and 116 kg Co, 400 kg Ni, and 73 kg Li in Eagle Plains, thereby advancing circular economy principles. Furthermore, the adoption of a 100% RE mix substantially decreased operational CO₂ emissions and mitigated associated carbon tax liabilities. Sensitivity analyses confirmed that system economics and performance are most affected by wind turbine hub height, PV lifetime, hydrogen fuel cost, temperature, SLB degradation, electrolyzer efficiency, and EV load demand, while battery SOC limits and discount rates also have notable impacts.

The findings should be interpreted in light of several limitations. The analysis relies on HOMER Pro simulations that use simplified assumptions for component performance and load behavior. Meteorological and traffic data were site specific and may not capture full inter-annual variability or extreme conditions. Cost, efficiency, and degradation uncertainties for emerging technologies such as fuel cells, electrolyzers, and second-life batteries were not fully quantified. Moreover, although results were compared with standard design guidelines, direct validation against operational Canadian microgrids was limited. As a result, the feasibility and scalability of the proposed systems, especially in remote northern regions, should be viewed as indicative rather than definitive.

Finally, this study emphasizes policy and stakeholder implications, advocating for modular off-grid infrastructure, integration of SLB, climate resilient design, and local workforce development to enable scalable, sustainable deployment. The proposed systems contribute directly to SDG 7 (Affordable and Clean Energy), SDG 9 (Industry, Innovation, and Infrastructure), and SDG 13 (Climate Action), providing a viable model for renewable, off-grid EV charging infrastructure in both urban and remote Canadian regions.

Future research should incorporate more detailed SLB degradation models, extreme-temperature performance modeling, and probabilistic (e.g., Monte Carlo) sensitivity analyses of inflow rates and renewable availability. Field validation of hybrid hydrogen battery systems in northern climates is also needed, along with assessment of long-term maintenance strategies, reliability metrics, and socio-economic impacts on remote communities.