Energy Conversion and Management 311 (2024) 118514

Contents lists available at ScienceDirect

Energy Conversion and Management

journal homepage: www.elsevier.com/locate/enconman

Research Paper

Integrated assessment of green hydrogen production in California: Life

cycle Greenhouse gas Emissions, Techno-Economic Feasibility, and
resource variability
Loiy Al-Ghussain a, *, Mohammad Alrbai b, Sameer Al-Dahidi c, Zifeng Lu a
a
Energy Systems and Infrastructure Analysis Division, Argonne National Laboratory, Lemont, IL 60439, USA
b
Department of Mechanical Engineering, School of Engineering, University of Jordan, Amman 11942, Jordan
c
Department of Mechanical and Maintenance Engineering, School of Applied Technical Sciences, German Jordanian University, Amman 11180, Jordan

A R T I C L E I N F O A B S T R A C T

Keywords: This study explores green hydrogen production, focusing on life cycle assessment (LCA) and techno-economic
Green hydrogen feasibility; specifically, life cycle greenhouse gas emissions and hydrogen production cost. Existing literature
Solar energy lacks a comprehensive exploration of the impact of solar and wind resource variations, ambient conditions, and
Wind energy
proton exchange membrane (PEM) electrolyzer sizing on the carbon intensity (CI) and hydrogen production cost.
Techno-economic feasibility
Life cycle greenhouse gas emissions
This work explores the relationship between the performance of photovoltaic (PV) and wind systems and the CI
of renewable electricity in the U.S. Secondly, this study incorporates the hourly variations in solar and wind
resources to assess the variation in hydrogen CI and production costs across California. The study examines how
reductions in PEM and Renewable Energy Systems (RES) costs, technological advancements, and the inclusion of
co-products’ carbon dioxide equivalent (CO2eq) credits and revenues impact hydrogen CI and production costs.
Wind-based systems consistently demonstrate lower CI (0.35 − 1.9 kg CO2eq/kg H2) except for the central regions
(CI can reach up to 4.34 kg CO2eq/kg H2), with geographical variations larger than solar-based systems (1.58 −
2.95 kg CO2eq/kg H2). Similar trends are observed for the hydrogen production cost which varies between 1.5
and 15 USD/kg H2 for wind-based systems and between 3.0 and 5.2 USD/kg H2 for the PV-based systems. This
study highlights the importance of optimal sizing of PEM electrolyzer to maximize its utilization for hydrogen
production and the use of hourly-based models to assess the CI and production costs of hydrogen. Additionally,
the geographical variation in the renewable energy resources and the ambient conditions are important factors
that affect the hydrogen cost and the CI of both renewable electricity and green hydrogen.

revolutionize sectors such as transportation, industry, and energy pro­

1. Introduction duction. It can be used in fuel cell electric vehicles, reducing fossil fuel
use and air pollutants emissions from the transportation sector [3]. In
1.1. Background and literature review industry, hydrogen can serve as a feedstock for chemical processes and
an energy source for high-temperature applications, contributing to
Hydrogen has garnered significant attention due to its pivotal role in decarbonization efforts [4]. Additionally, hydrogen facilitates energy
transforming the energy sector and addressing pressing environmental storage and grid balancing, enhancing the integration of renewable
concerns, primarily the imperative to combat climate change by tran­ energy sources by storing excess energy for later use [5]. The growing
sitioning away from fossil fuels and reducing carbon emissions [1]. Its interest in hydrogen reflects a global commitment to a cleaner, more
unique characteristics, including efficiency as an energy carrier and resilient energy future, making it a crucial subject of scientific study
diverse applicability in various sectors, position it as a key player in [3,6]. Keep in mind that the carbon footprint of hydrogen highly de­
shifting towards a low-carbon, renewable energy future. The global pends on its production pathways [7,8].
commitment to ambitious decarbonization goals underscores hydro­ Several distinct pathways for hydrogen production have emerged,
gen’s significance in the pursuit of sustainable and environmentally each with its unique advantages and challenges. The most traditional
friendly energy solutions [2]. Hydrogen has the potential to method, known as “grey hydrogen,” involves the high-temperature

* Corresponding author.
E-mail address: [email protected] (L. Al-Ghussain).

https://doi.org/10.1016/j.enconman.2024.118514
Received 26 February 2024; Received in revised form 30 April 2024; Accepted 2 May 2024
Available online 9 May 2024
0196-8904/© 2024 Argonne National Laboratory and The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license
(http://creativecommons.org/licenses/by-nc-nd/4.0/).
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

Nomenclature IRef Reference irradiance at reference conditions [W/m2]

z1 , z An arbitrary and hub heights [m]
CI Carbon Intensity u1 , uh Wind speeds collected at z1 and z [m/s]
CCS Carbon Capture and Storage ζ Shear coefficient
CCUS Carbon Capture, Utilization, and Storage EWT,el Hourly electric energy produced from the wind farm
GHG Greenhouse Gas [kWh]
GREET Greenhouse Gases, Regulated Emissions and Energy use in NWT Number of WTs
Technologies fd,WT Average annual degradation factor over the lifespan of the
H2 Hydrogen WT system [%/year]
IRA Inflation Reduction Act uC , uF , uR Cut-in, cut-off, and rated wind speeds [m/s]
LCA Life Cycle Assessment Pe,R Rated WT power [kW]
MOO Multi-Objective Optimization ω Weibull distribution coefficient
PEM Proton Exchange Membrane Std, u Standard deviation and mean of hourly wind data [m/s]
PV Photovoltaic AMtH2 The amount of H2 produced at each hour [kg/h]
RE Renewable Energy EtRES Hourly energy generated by the RES [kWh]
RES Renewable Energy Systems (PV or Wind) SECPEM Specific electricity consumption [kWh/kg H2]
SMR Steam Methane Reforming CostH2 Unit cost of H2 [USD/kg H2]
SOEC Solid Oxide Electrolysis Cells CPEM Capital cost of PEM system [USD/kWe]
TMY Typical Meteorological Year CRES Capital costs of RES system [USD/kWe]
TOPSIS Technique for the Order of Preference by Similarity to O&MPEM Operation and maintenance cost of PEM [USD/kWe]
Ideal Solution O&MRES Operation and maintenance cost of RES [USD/kWe]
WTs Wind Turbines H2,PEM PEM system’s annual H2 production [kg/year]
LCOH Levelized Cost of Hydrogen [USD/kg H2] LT Lifespan of the green H2 system [Years]
IT Total solar radiation incident on the PV modules [W/m2] CBoP Balance of plant cost [USD/kWe]
EPV,el Hourly energy production from a PV system [kWh] Cstack Stack capital cost [USD/kWe]
ηPV,Ref Reference efficiency of the PV modules [%] RPstack Number of stacks’ replacements
βRef Temperature coefficient of PV cell [%/◦C] UF Utilization Factor [%]
STC Standard Test Conditions FUR Full Utilization Ratio [%]
TRef,STC Reference cell temperature at STC [◦C] SYel Annual Specific Yield of electricity [kWh/kW]
Am Area of a single PV panel [m2] SYH2 Annual Specific Yield of H2 [kg H2/kW]
Nm Number of PV panels EE Embodied Emissions [kg CO2eq/kW]
Ls Performance ratio [%] EEstacks EE by stacks manufacturing [kg CO2eq/kW]
TPV PV panel temperature [◦C] EEBoP EE by BoP [kg CO2eq/kW]
fd,PV Average annual degradation factor over the lifespan of the CFRES Capacity Factor of the RES [%]
PV system [%/year] CIRES CI of produced electricity by RES [g CO2eq/kWh]
Tamb Ambient temperature [◦C] CAP System’s capacity [kW]
NOCT Nominal operating conditions CIH2 RES CI of green H2 [kg CO2eq/kg H2]
TCell,NOCT Cell temperature at NOCT [◦C] CRPEM PEM size ratio [%]
TRef,NOCT Reference cell temperature at NOCT [◦C]

steam methane reforming (SMR) of natural gas, which releases carbon costs [7,11–14]. The choice among the mentioned pathways ultimately
dioxide from both the chemical conversion process and the combustion depends on a region’s specific energy mix, policy priorities, and eco­
of natural gas for heat requirements [9]. In contrast, “blue hydrogen” nomic considerations, and the balance between these factors will shape
employs carbon capture and storage (CCS) technology to capture and the future landscape of hydrogen production for a sustainable and
store the CO2 emissions produced during SMR [7]. “Green hydrogen” is decarbonized future [15].
generated through the electrolysis of water using electricity from Among different hydrogen production pathways, green hydrogen
renewable energy systems (RES), such as wind or solar power, providing offers a clean and sustainable energy option and has more grid inte­
a potential cleaner and environmentally friendly option [10]. Lastly, the gration potential, aligning with the global transition towards a low-
relatively novel “purple hydrogen” pathway leverages nuclear power as carbon and environmentally responsible energy future. For instance,
a primary energy source for the electrolysis of water, offering a low- Gallardo et al. [27] introduced an innovative approach for the optimal
emission and consistent energy supply [11]. sizing of solar photovoltaic (PV) − proton exchange membrane (PEM)
Table 1 briefly compares the carbon intensities (CI) and the costs of systems. The key contribution lies in proposing the solar plant to elec­
different hydrogen production pathways in different countries from trolyzer capacity ratio as the focal optimization variable. The study in­
literature. Grey hydrogen, while economically viable, results in sub­ vestigates how this ratio influences the Levelized Cost of Hydrogen
stantial carbon emissions and is less aligned with carbon reduction goals (LCOH). The research encompasses case studies across a global hori­
[7]. Blue hydrogen mitigates the carbon footprint but still relies on zontal irradiation range of 1400–2600 kWh/m2-year. The resulting
natural gas as a feedstock and faces challenges related to CCS infra­ LCOHs vary from 5.9 to 11.3 USD/kg H2, depending on sizing and
structure [7]. Purple hydrogen, with its nuclear power source, offers a location. Significantly, the solar/PEM ratio emerges as a pivotal factor
stable energy supply and low emissions but faces public perception influencing cost, production, and LCOH optimality. Moreover, Fabianek
challenges and concerns regarding nuclear safety and waste manage­ & Madlener [28] conducted a study to assess the economic viability of
ment [12]. Green hydrogen generally has low carbon footprint, but is power and hydrogen generation using solar and wind energy resources
often hampered by intermittent energy sources and higher production in Northern Germany and California, both recognized for their

2
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

Table 1 in five Moroccan cities. They found that this hybrid approach can pro­
Brief environmental and economic comparison of hydrogen production path­ duce hydrogen at low costs, particularly in Dakhla ($2.54/kg), the
ways in different countries. lowest in Morocco. The research emphasizes the importance of energy
Ref. Country H2 type CI (kg CO2eq/ H2 cost (USD/ storage and shows that water desalination costs are minimal, impacting
kg H2) kg) the overall costs insignificantly. In another study, Nasser and Hassan
Koroneos et al. Germany Grey 10.66 2.08 [35] conducted a sensitivity analysis on green hydrogen production
[16] from solar PV panels and wind turbines under varying climate condi­
Aydin and Dincer Canada Grey + 6.8 2.13–3.64 tions. The study compared system efficiency, hydrogen output, and
[17] Green
economic viability across different setups, including solo and hybrid
Ozawa et al. [18] Australia Blue 4.16 2.79
Vilbergsson et al. Belgium Blue 3.02 2.49 configurations of PV and wind, under specific solar and wind conditions.
[19] Key findings indicated that PV systems performed better at high solar
Mohideen et al. − Purple 0.3–0.6 4.2–7.0 intensity but were costlier and had longer payback periods than wind
[20] systems, which excelled at higher wind speeds. The study provided
Granovskii et al. U.S. Green/PV 2.1 5.79–10.49
[21]
detailed comparisons to guide the optimal setup for green hydrogen
Simons & Bauer Spain Green/PV 4.4 7.54 production based on environmental conditions. Garcia and Oliva [36]
[22] performed similar work, including an emissions analysis, particularly of
Cetinkaya et al. Canada Green/PV 2.41 5.8 CO2 emissions. They developed a model to evaluate the costs and
[23]
emissions associated with producing green hydrogen using hybrid solar
Rivera et al. [24] China Green/PV 3.07 6.03
+ Wind and wind energy systems in four locations across Chile. By analyzing
Al-Qahtani et al. U.S. Green/PV 3.1 9.5 local solar and wind data, the model determined the optimal setup for
[25] solar PV, wind subsystems, and electrolyzers. The findings showed that
Kolb et al. [15] Germany Green/ 5.08 3.77 green hydrogen could be produced at competitive prices, ranging from
Hybrid
$2.09 to $3.28 per kg, and resulting in low CO2 emissions (1.06–1.57 kg
Freire Ordóñez UK Green/PV 5.65 8.55
et al. [26] CO2e/kg H2). The study highlighted the environmental advantages of
Freire Ordóñez Italy Green/PV 4.1 6.91 this hybrid approach over the traditional steam methane reforming
et al. [26] process, suggesting its potential as a benchmark for future emission
regulations and green hydrogen pricing strategies.
Other studies such as [37,38] investigated the techno-economic and
commitment to sustainable energy transition. The research models
carbon footprint of the complementary use of grid electricity in
green hydrogen generation plants, incorporating inputs and outputs for
hydrogen production systems. For instance, Hurtubia and Sauma [37]
wind power plants, photovoltaic power plants, and PEM electrolyzers.
investigated the economic feasibility and the environmental footprint of
They reported that the LCOH ranges from 4.5 €/kg to 5.2 €/kg in
hydrogen produced using renewable and grid electricity in Chile. They
Northern Germany and from 4.6 USD/kg to 5.3 USD/kg in California.
concluded that if 10 % of the electricity used for hydrogen production
Henry et al. [29] conducted a techno-economic analysis investigating
comes from grid electricity, the hydrogen production cost will drop by
hydrogen (H2) production from wind. The study compares PEM and
almost 26 % without significant increase in the CI. Similarly, Bhandari
alkaline electrolyzer units, considering the intermittency of wind and
and Rakesh Shah [38] showed that off-grid green hydrogen system is not
comparing it with grid-purchased electricity. In the first scenario, H2 is
feasible in Germany.
produced using grid energy, with electricity purchased when below a
For hydrogen production from grid-electricity only, Panah et al. [39]
specified price point or during specified hours. In the second scenario,
compared Alkaline, PEM, and solid oxide electrolysis cells (SOEC),
the user owns a wind turbine, and the electricity price is not a factor, but
emphasizing economic differences. PEM and alkaline electrolyzers can
the capital expenditure for the turbine is considered. The study finds that
produce hydrogen for less than 3 €/kg, potentially dropping below 2
the price of H2 production from wind is comparable to natural gas-
€/kg with mass adoption. If electricity is half-priced, the levelized cost of
derived H2 at a larger scale, with a minimum selling price of £4.85/kg
hydrogen can reach around 1 €/kg. SOEC can be as economical as other
at a setpoint of 500 kg of H2/hour.
technologies, but the study notes the persistence of grey hydrogen
Several researchers have explored the potential of off-grid electrol­
dominance without subsidized electricity or carbon tax adjustments.
ysis plants for green hydrogen production using renewable energy
Similarly, Ramadan et al. [40] conducted a comprehensive study on
sources. For instance, Kalinci et al. [30] focused on a solar/wind hybrid
regional hydrogen production costs in the entire U.S., focusing on the
system equipped with hydrogen storage on a Turkish island, showcasing
LCOH and CI. The investigation specifically delved into electrolysis
the advantages of hydrogen as a storage solution in isolated areas. Onar
technologies, including Alkaline, PEM, and SOEC, comparing them to
et al. [31] investigated a system combining wind, solar, fuel cells, and
the widely used SMR. In the year 2020, the study found that SMR with
ultracapacitors aimed at powering off-grid loads during daylight hours.
90 percent CCS exhibited a lower average LCOH and CI for hydrogen
In contrast, Fernandes et al. [32] examined the financial feasibility of
production compared to SOEC electrolysis. However, in states with
wind-solar hybrid systems for green hydrogen production and storage in
cleaner grids, hydrogen produced through SOEC demonstrated a
Brazil, concluding that these systems were not economically viable there
reduced carbon intensity compared to SMR with 90 percent Carbon
yet. Al-Buraiki et al. [33] looked into the use of renewable energy re­
Capture, Utilization, and Storage (CCUS). Notably, Washington emerged
sources in Saudi Arabia for generating electricity and hydrogen. They
as a leader with one of the lowest carbon footprints and the most cost-
employed computer simulations to fine-tune a hybrid renewable energy
effective LCOH for hydrogen production through alkaline electrolysis.
setup comprising solar panels, wind turbines, batteries, an electrolyzer,
Projections for the year 2050 indicate an estimated LCOH of 3.2 USD per
and a hydrogen storage tank for a residential home in Dhahran. Their
kg for Alkaline, 3.1 USD per kg for PEM, and 2.6 USD per kg for SOEC,
goal was to cut costs and meet reliability standards. This study proposed
assuming constant electricity prices.
new metrics for evaluating hydrogen production and storage effective­
Debates have been raised in recent years regarding how much green
ness, including hydrogen supply deficiency and the likelihood of
is green hydrogen and the need to include embodied emissions in related
hydrogen supply loss, and found a potential reduction in CO2 emissions
hydrogen production facilities when comparing the carbon footprint of
of 9.66 tons annually.
different hydrogen production pathways. In a recent study by Bracci
Recently, Ourya et al. [34] investigated the cost and technical
et al. [41], they carried out an intercomparing of net-zero hydrogen
feasibility of producing hydrogen using combined solar and wind power
production cost in the U.S. from different production pathways with and

3
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

without embodied emissions. To achieve net-zero hydrogen, they and environmental impact associated with green hydrogen production,
incorporate the cost of CO2eq removal in both fossil fuel and electricity crucial for making sustainable decisions. It is important to clarify that
pathways (PV-based and grid-based systems). Additionally, they pre­ the methodology and techniques employed are versatile and can be
sented a comparison between the carbon footprint of each technology applied to any geographical location. California serves as a representa­
with and without embodied emissions and investigated the eligibility of tive example, emphasizing that the findings extend beyond its specific
each pathway for the U.S. Inflation Reduction Act under the 45 V Section context.
(IRA 45 V) incentives. They highlighted the importance of hourly-based
analysis when investigating the cost of green hydrogen due to the 2. Theory and methodology
variability in solar resources. Additionally, they concluded that
electricity-based systems have lower net-zero hydrogen production cost This section illustrates the numerical framework devised for
(between 2 and 2.9 USD/kgH2) compared to the fossil-based systems modeling the proposed RES-based hydrogen production system while
(between 2.7 and 5.9 USD/kgH2). This is only applicable if electricity- optimizing its economics and environmental footprint. Specifically,
based systems were eligible for the IRA45V incentives, i.e., without Sections 2.1 and 2.2 detail the modeling of the PV and wind systems,
counting for the embodied emissions. respectively. Whereas Section 2.3 presents the modeling of the hydrogen
The carbon footprint of green hydrogen varies due to climate and production system. Sections 2.4 and 2.5 define the parameters used to
RES supply chain variations [11,42,43]. Kolahchian et al. [42] carried assess the techno-economic and environmental analysis of the proposed
out a comprehensive literature review regarding the carbon footprint of system, respectively. Section 2.6 demonstrates the process of optimizing
PV-based hydrogen around the world and they concluded that the the system’s performance and economics. Lastly, Section 2.7 presents
embodied emissions in PV systems are the major contributors in the the parameters used to evaluate the sensitivity of the system and the
green hydrogen CI. Additionally, they re-estimated the CI of PV-based potential reduction of H2 CI and production cost. As aforementioned,
hydrogen by updating several parameters such as the PV efficiency, California has been used in this work as a showcase to prove the validity
supply chain and material inventory. They concluded that incorporating and the importance of the developed model. By utilizing inputs such as
the updated numbers decreases the CI by almost 60 % in Europe. the Typical Meteorological Year (TMY), embodied emissions in PEM
electrolyzer and RES technologies, and electricity carbon intensity data
1.2. Research gap and contribution specific to California, the analysis effectively demonstrates the validity
and significance of the developed model, demonstrating its applicability
Despite the numerous studies that investigated the life cycle assess­ beyond the showcased region.
ment and the techno-economic feasibility of green hydrogen. These Finally, this study aims to investigate the CI and cost of hydrogen
studies have established the feasibility of using solar and wind energy production across CA and does not delve into the sizing of the hydrogen
systems for green hydrogen production but have not examined the real- system to cover certain demand. Hence, we adopted a methodology
time dynamics of resource availability, the effect of geographic varia­ where the size of the hydrogen production unit was not predetermined.
tions in RE resources, and PEM sizing on the cost and carbon intensity of Instead, we generalized the results using normalized output metrics,
green hydrogen. Understanding these variations is crucial for optimizing specifically kilograms of hydrogen produced (kgH2) or kilowatts (kW).
production systems to minimize environmental impact and operational This approach was deemed appropriate as we assumed that capital costs
costs, informing policy decisions, and advancing scientific knowledge in (USD/kW) and embodied emissions (kgCO2eq/kW) do not significantly
the field. In summary, to the best of our knowledge none of the studies in vary with the system size. Consequently, our analysis did not rely on
the literature: (1) investigated the effect of solar and wind resources specific energy demands tied to a fixed unit size. For the PV-based sys­
variation as well as the ambient conditions on the carbon intensity of tem, we assumed a fixed PV capacity of 1 MW whereas, for the WT-based
renewable electricity, (2) investigated the effect of solar and wind re­ system, we assumed a fixed capacity of 2 MW WT. Additionally, the
sources hourly and geographic variation on the H2 carbon intensity and system’s boundaries are assumed to be from well-to-production and
production cost, (3) investigated the effect of PEM electrolyzer sizing in hence the costs and the CI of downstream processes are not included in
PV-based and wind-based on the H2 carbon intensity and production the analysis.
cost, or (4) considered multi-objective optimization for sizing the PEM
electrolyzer in green hydrogen systems to minimize both the H2 pro­ 2.1. Solar PV system
duction cost and minimize the carbon intensity.
Hence, this study aims to: The typical meteorological year (TMY) data files for California,
United States of America, have been generated by using the PVGIS
• Explore the relation between the performance of PV and wind system software [44]. The software has been used to estimate the total solar
and their carbon intensity of electricity in the U.S. considering the radiation incident on the PV modules (IT ) as detailed in [45,46]. For
supply chain of PV and wind systems. conciseness, the standard calculations used for the estimation are not
• Perform a multi-objective optimization to estimate the optimal PEM repeated here and interested readers can refer to [45,46] for more de­
electrolyzer to RES capacity ratio across CA that minimizes the tails. Consequently, the hourly energy production from a PV system
hydrogen production cost and CI. (EPV,el ) (in kWh) of CS3L-365 monocrystalline PV modules [47], man­
• Analyse how hourly and geographic fluctuations in solar and wind ufactured by CanadianSolar (specifications are shown in Table A1 in the
resources throughout the year affect the CI and production cost of Appendix), employed in this work, fluctuates considerably and has been
green hydrogen, crucial for optimizing production systems and estimated by using Eq. (1):
promoting widespread adoption of clean energy. ( ( ))
• Investigate the effect of PEM and RES costs reduction as well as the EPV,el = ηPV,Ref 1 − βRef TPV − TRef ,STC IT Am Nm Ls fd,PV (1)
technological advancements on the H2 carbon intensity and pro­
duction cost in CA. where ηPV,Ref represents the reference efficiency of the PV modules (in
• Assess the effect of co-products (O2 and excess electricity) credits and %), βRef represents the temperature coefficient of the PV cell (in %/◦C),
revenues on the H2 carbon intensity and production cost in CA. TRef,STC represents the reference cell temperature at standard test con­
ditions (STC) (in ◦C), Am represents the area of a single PV panel (in m2),
California was employed as a case study in this analysis to showcase Nm represents the number of PV panels, Ls represents the performance
the functionality and the importance of the developed model. The focus ratio (in %) (approximated as 0.85 [48]), which comprises the losses in
of the study is to provide insights into the techno-economic feasibility wiring, shading, and inverter, TPV represents the PV panel temperature

4
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

(in ◦C), and fd,PV represents the average annual degradation factor over ambient temperature with a purity ≥ 99.9 % [14,54].
the lifespan of the PV system (in %/year), assuming annual degradation ⎧ t
⎪ ERES
of 0.7 % [43]. ⎪ t
⎨ SEC , if ERES < PEMCAP
⎪
The estimation of the TPV (Eq. (2)) relies on the Tamb (the operating t
AMH2 =
PEM
(8)
⎪
ambient temperature, in ◦C), ⎪
⎪
⎩
PEMCAP
, else
TCell,NOCT (the cell temperature at the nominal operating conditions SECPEM
(NOCT), in ◦C), TRef,NOCT (the reference cell temperature at NOCT in ◦C),
where PEMCAP is the optimal PEM capacity (in kWe).
IT (the total solar radiation incident on the PV modules, in W/m2) and
IRef (the reference irradiance at reference conditions, in W/m2):
( ) IT 2.4. Techno-Economic parameters
TPV = Tamb + TCell,NOCT − TRef ,NOCT (2)
IRef
To evaluate the economic feasibility of the green hydrogen produc­
2.2. Wind system tion, the unit cost of hydrogen (denoted as CostH2 , in USD/kg H2) is used
which is estimated using Eq. (9) [1,14,37,55].
TMY wind data for California has been generated using the PVGIS CPEM, total × CRPEM + CRES + (O&MPEM × CRPEM + O&MRES ) × LT
software [44]. The G114-2.0 wind turbines (WTs) have been used in this CostH2 =
H2, PEM × LT
work [49], the specifications of the WT are shown in Table A1 in the
(9)
Appendix. The extracted wind speed data (u1 , in m/s) collected at the
ground level (z1 = 10 m) given the shear coefficient (ζ) have been used where CPEM,total and CRES represent the lifespan capital costs of the PEM
to compute the wind speed (uh , in m/s) at hub height (z, in m) by system and the RES (in USD/kWe), respectively; O&MPEM and O&MRES
resorting to Eq. (3). The average wind shear coefficient in the U.S. have represent the operation and maintenance costs of the PEM system and
been employed in this study which is around 0.225 [50]. the RES (in USD/kWe), respectively; H2,PEM represents the PEM system’s
( )ζ
z annual H2 production (in kg), the annual sum of the hourly H2 pro­
uh = u1 (3) duction in Eq. (8); LT is the lifespan of the green hydrogen system (in
z1
years), which is assumed to be 30 years; and CRPEM is the optimal PEM
It is worth mentioning that the wind speed is variable, thus the gener­ capacity ratio (CRPEM = PEM
CAPRES ) which is found through the optimization
CAP

ated electrical energy from WTs fluctuates to a great extent compared to procedure described in Section 2.6. The capital cost of the PEM system is
those obtained by the PV system. Consequently, the hourly electric en­ estimated using Eq. (10) taking into account the balance of plant cost
ergy produced from the wind farm (EWT,el , in kWh) is determined as per (CBoP , in USD/kWe), the stack capital cost (Cstack , in USD/kWe) and the
Eq. (4): number of stacks’ replacements (RPstack ). The economic parameters
⎧ utilized in this work are listed in Table 2. It is important to note that the
⎨ 0, uh < uC oruh > uF hydrogen compression, storage and transportation costs are not
(4)
ω
EWT,el = NWT × fd,WT × a + b(uh ) , uC ≤ uh ≤ uR
⎩ included in this work.
Pe,R , uR < uh ≤ uF
CPEM, total = CBoP + Cstack × RPstack (10)
where NWT represents the number of WTs, fd,WT is the average annual
degradation factor over the lifespan of the WT system (in %/year), where CBoP equals to 320.65 USD/kWe (0.55 × CPEM ), Cstack is 262.35
assuming annual degradation of 1.19 % [51], uC , uF , and uR represent USD/kWe (0.45 × CPEM ), and RPstack equals to three. With CPEM being the
the cut-in, cut-off, and rated wind speeds (in m/s); and Pe,R represents cost of the PEM unit (583 USD/kWe) without additional cost associated
the rated WT power (in kW). Equations (5) and (6) have been used to with stack replacement.
estimate the coefficients a and b, respectively. On the other hand, Eq. (7) We evaluate the performance of the PEM electrolyzer using two
has been used to estimate the Weibull distribution coefficient (ω), parameters, the utilization factor (UF, also known as the capacity factor
following Justus theory [52]: of PEM, in %) and the full utilization ratio (FUR, in %) which are esti­
mated using Eq. (11) and Eq. (12), respectively. The UF represents the
(uC )ω
a = Pe,R (5) ratio between the total annual H2 production and the amount that can be
(uC ) − (uR )ω
ω
generated if the PEM is run at full (rated) capacity all the year. Whereas
FUR represents the frequency of running the PEM at full capacity
Pe,R
b= (6) throughout the year. 8760 is the number of hours in a year.
(uR )ω − (uC )ω
∑8760 t
{ t=1 AMH2
UF = PEMCAP (11)
ω = (std/u)− 1.086
, 1 ≤ ω ≤ 10 } (7) × 8760
SECPEM

where std and u are the standard deviation and the mean of the hourly
wind data (in m/s), respectively.
Table 2
The capital and operation and maintenance expenses of the RES-based hydrogen
2.3. Hydrogen production system
production system, along with their expected operational lifespans.

In this study it is assumed that the PEM is only run by electricity from System Capital Cost O&M Annual Lifespan
Cost (years)
the RES and due to the variation in wind and solar resources, the PEM
electrolyzer is sized to ensure the maximum utilization of it. The amount PV system 876 USD/kWe [56] 13 USD/kWe [56] 30 [57]
Wind system 1285 USD/kWe [56] 28 USD/kWe [56] 30 [58]
of hydrogen produced at each hour (denoted as AMtH2 , in kg/h) can be
PEM unit 583 USD/kWe 4 % of CPEM [61] 30 [61,62]
estimated as per Eq. (8) using the hourly energy generated by the RES [59,60]
(EtRES , in kWh) and the specific electricity consumption (denoted as PEM stack 45 % [60] − 10 [63]
SECPEM , in kWh/kg H2) which is assumed to be 56 kWh/kg H2 [51,53]. share
PEM BoP share 55 % [60] 30 [61,62]
In this study, it is assumed that the hydrogen is produced at 80 bar and at
−

5
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

∑8760
Nt
FUR = t=1
(12) CIH 2RES =
EERES × CAPRES + (EEBoP + EEstack × RPstack ) × PEMCAP
(18)
8760 H2, PEM × LT
{
t
1, PEMCAP ≤ ERES
Nt = (13) CIH 2RES =
EERES + (EEBoP + EEstack × RPstack ) × CRPEM
(19)
0, else H2, PEM × LT

We further use another technical parameter to evaluate and compare the

2.6. Optimization procedure
potential for hydrogen production across CA where we employ the
annual specific yield of electricity (SYel , in kWh/kW) and hydrogen
Two objective functions have been considered to optimally define
(SYH2 , in kgH2/kW) which can be estimated using Eqs. (14) and (15),
the parameters of the proposed RES-based green H2 system, namely H2
respectively.
production cost (CostH2 , in USD/kg H2) and the carbon intensity of H2
∑8760 t
E (CIH2 RES , in kgCO2eq/kg H2). Thus, a multi-objective optimization
SYel = t=1 RES (14) (MOO) problem is formulated and solved by using the ParetoSearch
CAPRES
technique [64] to optimally identify the optimal PEM size ratio (CRPEM ,
∑8760
AMHt 2 in %). The optimal values are those at which the CostH2 and CIH2 RES are
(15)
t=1
SYH2 =
CAPRES minimized. The detailed analysis of the MOO is formulated in Eq. (20):
Minimimze : CostH2 = f1 (CRPEM )
2.5. Environmental assessment
Minimimze : CIH 2RES = f2 (CRPEM ) (20)
Greenhouse Gases, Regulated Emissions and Energy use in Tech­
nologies (R&D GREET) model [51] has been employed in this study to The ParetoSearch toolbox in MATLAB was used [65,66]. The region
estimate the embodied emissions of RES-based hydrogen system in under investigation, California, has been divided into a grid consisting of
California. These emissions are produced during the extraction of ma­ 1052 points. These points cover sub-areas with evenly spaced intervals
terials, manufacturing processes and transportation involved in RES of latitude and longitude (0.1◦ ) in order to visualize the spatial variation
manufacturing. In this study, the estimation of CI for hydrogen pro­ in hydrogen CI and production cost. The MOO problem at hand produces
duction involves the utilization of specific values pertinent to PV and multiple solutions that are considered optimal and are presented in the
wind systems as deployed in California (USA). These values were form of a Pareto frontier. Technique for the Order of Preference by
derived from the R&D GREET model, specifically tailored to the Cali­ Similarity to Ideal Solution (TOPSIS) algorithm (explained in reference
fornia context. The R&D GREET model provides a robust framework for [67]) was employed to identify the desired optimal solution. Random
assessing well-to-gate/well-to-wheel life cycle emissions, encompassing samples of Pareto lines and point selection for PV and Wind-based sys­
various facets such as the supply chain dynamics (including different tems are shown in Fig. A2 in the Appendix.
manufacturing countries) and production pathways [51]. By leveraging
the GREET model tailored to California, we ensure a nuanced under­ 2.7. Sensitivity analysis
standing of the environmental implications associated with hydrogen
production, accounting for region-specific factors and variations in In order to accurately assess the economic aspects of the proposed
technology deployment. hybrid system, it is important to take into account the variability
The lifespan embodied emissions (EE, in kgCO2eq/kW) obtained from observed in the data of the main techno-economic parameters reported
R&D GREET 2023 for the year of 2022 for commercial monocrystalline in various literature sources. For example, the capital costs of PV, wind,
PV, onshore wind and PEM systems are 1335.48, 766.03, and 86.28 and PEM systems as well as PV and PEM specific energy in addition to
kgCO2eq/kW, respectively for CA [51]. Note that the PEM embodied wind turbine hub height and the manufacturing country (which changes
emissions are divided into two parts; emissions associated with stacks embodied emissions of the RES). Thus, a sensitivity analysis is con­
manufacturing (EEstacks , in kgCO2eq/kW) which is equal to 14.55 ducted for both the H2 production cost and CI considering the range (i.e.,
kgCO2eq/kW and emissions associated with BoP (EEBoP , in kgCO2eq/kW) optimistic/minimum and pessimistic/maximum values) of data reported
which is equal to 71.73 kgCO2eq/kW [51]. First, we aim to create a in the literature as shown in Table 3.
relation between the capacity factor of the RES (CFRES , in %) and the CI In addition to the sensitivity analysis, the potential reduction in H2
of produced electricity (CIRES , in gCO2eq/kWh) to better understand the production cost and CI in 2030 is investigated using the 2030 predicted
effect of resources variation on the CI of electricity as shown in Eq. (16). capital costs, technical specifications, and the change in the
manufacturing country taking into account the future change in the PV
EERES
CIRES = (16) and WT supply chains. Moreover, the effect of selling the RES excess
8760 × CFRES × LT
electricity and O2 co-product and the allocation of CO2eq credit for these
with, co-products in H2 production cost and CI are also investigated. Table 4
shows the parameters used to estimate the H2 production cost and CI in
∑8760 t
t=1 ERES California in 2030.
CFRES = (17)
8760 × CAPRES
3. Results and discussion
where EtRES is the hourly electricity generated by the RES in kWh, CAP is
the system’s capacity in % and the subscript RES is either the PV or the 3.1. Variation of carbon intensity of renewable energy systems
wind system.
Afterward, the carbon intensity of green hydrogen (CIH2 RES , in The manufacturing and installation as well as the recycling of
kgCO2eq/kg H2) is estimated using Eq. (18) (or Eq. (19) [37] using the renewable energy systems such as PV and wind systems are associated
optimal PEM size ratio (CRPEM )). It should be noted this study adopts the with Greenhouse Gas (GHG) emissions. These embodied emissions vary
PV supply chain reported in [43] whereas default R&D GREET values for significantly for the same renewable energy (RE) technology depending
wind and PEM systems is used when estimating the embodied emissions. on the manufacturing country [43]. Additionally, the carbon intensity of
electricity generated from RES, when embodied emissions are included,
varies significantly even if the manufacturing country of RE technology

6
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

Table 3
Optimistic and pessimistic values of the main techno-economic parameters used
in this work for the sensitivity analysis.
Parameter Symbol Unit Optimistic Pessimistic Ref.

PV capital cost (CPV ) USD/kW 569 1878

[56]
PV O&M (O&MPV ) USD/kW 11 20
[56]
PV efficiency (ηPV,Ref ) % 22 15
[68]
PV (EEPV ) (kgCO2eq/ USA APAC1
[43]
manufacturer kW) (1249.6) (1369.4)
(emissions)
Wind capital (CWT ) USD/kW 987 1907
[56]
cost
Wind O&M (O&MWT ) USD/kW 25 81
[56]
WT hub height (z) m 160 70 −
WT (EEWT ) (kgCO2eq/ EU2 China
[51]
manufacturer kW) (762.2) (774.7)
(emissions)
PEM capital (CPEM ) USD/kW 480 1400
[69]
cost
PEM Specific (SECPEM ) kWh/ 50 83
[69]
Energy kgH2 Fig. 1. Variation of the carbon intensity of electricity generated from com­
1
APAC refers to Asia-Pacific countries and 2EU refers to the European Union. mercial scale monocrystalline PV systems and onshore wind turbines with the
capacity factor in the U.S.

Table 4 Green Hydrogen production via PEM electrolyzer using renewable

2030 predicted technical, economic and environmental parameters. electricity from PV and wind has gained significant interest in the last
Parameter Symbol Unit Value Ref. few years. Numerous studies in the literature have introduced green
hydrogen as clean one; however, including embodied emissions in­
PV capital cost (CPV ) USD/kW 587
[70] creases the carbon intensity of green hydrogen. The carbon intensity of
PV O&M (O&MPV ) USD/kW 9.11
[71] renewable electricity is one of the main factors that affect the carbon
PV efficiency (ηPV,Ref ) % 30 footprint of green hydrogen. Additionally, the variation in the RE re­
[72] sources does not only affect the carbon intensity of renewable electricity
PV manufacturer (EEPV ) (kgCO2eq/ USA
[51] but also affects the amount of green hydrogen that can be produced. The
(emissions) kW) (1116.04)
Wind capital cost (CWT ) USD/kW 1075
last factor can be overcome by oversizing the PEM to utilize all the
[73] renewable electricity; nevertheless, such methodology will increase the
Wind O&M (O&MWT ) USD/kW 15.73 H2 production cost due to the inefficient utilization of PEM electrolyzer
[71]
WT hub height (z) m 150 due to the fluctuations of RE resources. Hence, there is a tradeoff be­
[74]
tween the H2 production cost and its carbon intensity when sizing the
WT manufacturer (EEWT ) (kgCO2eq/ USA (694.88)
[51] PEM electrolyzer as shown in Fig. 2.
(emissions) kW)
PEM capital cost (CPEM ) USD/kW 250
[75] 3.2. Optimal sizing of PEM electrolyzer
PEM Specific Energy (SECPEM ) kWh/kgH2 45
[69]
O2 Price − USD/kg 0.14 [76] Due to the variation in the RE resources, the optimal PEM size ratio
Electricity Price − USD/kWh 0.0268 − varies depending on the geographical location hence it is vital to map
the PEM size ratio to ensure the optimal utilization of RE resources for
H2 production. In this study MOO algorithm was employed to find the
is the same. This is because the performance of RES varies significantly
optimal PEM size ratio that minimizes both the H2 CI and production
due to the variation in the RE resources as well as the ambient condi­
cost across CA. Fig. 3 shows the optimal PEM size ratios across CA for
tions. Hence, it is important to consider the effect of the RES perfor­
both PV-based (Fig. 3(a)) and WT-based (Fig. 3(b)) systems. The spatial
mance variation on the carbon intensity of renewable electricity. By
variation in the optimal PEM size ratio is clear in both systems and is
adopting the PV supply chain information for PV systems used in the U.
highly related to the geographical locations of CA and the corresponding
S. reported in [43] and by assuming that the wind turbines in the U.S.
variations in the magnitude and frequency of the RE resources as shown
are manufactured in the U.S., the lifecycle GHG emissions embodied in
in Fig. A1 in the appendix. For instance, the high variability in wind
PV and wind turbines are obtained from R&D GREET 2023. Using Eq.
resource is impacted by factors, such as the Sierra Nevada and coastal
(16) we analyze the variation in the carbon intensity of renewable
mountain ranges [78,79] leading to substantial fluctuations in WT-based
electricity, where embodied emissions are normalized by the lifespan
system’s energy production and the corresponding PEM size ratio. On
electricity that can be produced at different capacity factors of solar PV
the other hand, the consistent solar PV yields in the southern regions of
systems or wind turbines [70,77] to simulate the variation in the re­
CA reflect its high solar radiation, e.g., Central Valley and southern
sources and ambient conditions as shown in Fig. 1.
deserts, which is conducive to a more uniform PEM size ratio for solar PV
Fig. 1 confirms the importance of analyzing the performance of the
systems. The MOO algorithm, in this regard, aims to increase the utili­
RES before assuming certain intensity for the renewable electricity
zation of the PEM electrolyzer by optimizing the PEM size that best
where the CI varies almost by four folds for PV and three folds for wind
matches the variable RE resources. Specifically, notice that in Fig. 3,
systems over the capacity factor values bounds. Additionally, it should
there is larger range of variability in the PEM size ratio (CRPEM ) for the
be noted that Eq. (16) can be used to estimate the carbon intensity of
WT-based system (between 30 and 80 %) compared to the PV-based
electricity produced by PV and wind in any location in the US if the
system (between 57 and 66 %) due to the increased variability in the
capacity factor (estimated using Eq. (17)) is known.
wind resources (both frequency and magnitude) compared to the solar

7
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

Fig. 2. Effect of PEM size ratio on green H2 carbon footprint and production cost in CA using average solar and wind resources: a) PV-based system and b) WT-
based system.

Fig. 3. Optimal PEM size ratio across CA for green hydrogen production: a) PV-based system and b) WT-based system.

resources across CA. The WT-based system has a higher FUR (up to 25 %) compared to PV-
Moreover, it is clear in Fig. 3 that the distribution of the optimal PEM based system (up to ~ 12 %) where the magnitude and the frequency
size ratio matches the specific electricity yield of PV-based and WT- of the RE resources, justified by the geographic and topographical in­
based systems shown in Fig. 4(a) and Fig. 4(b), respectively. In fact, fluences on wind/solar patterns, are the key factors that affect the FUR.
most of the southern parts of CA are characterized by higher solar re­ Moreover, the nature of solar energy, being available only during day
sources compared to the north as shown in Fig. 4(a) which corresponds hours, compared with the wind energy which can be generated during
to higher H2 annual specific yield (defined in Eq. (15)) as shown in Fig. 4 the day and night hours, thus affecting the FUR and underscoring the
(c). On the other hand, the central, north-west and south-west regions of necessity to align system utilization with the inherent variability of each
CA are characterized by low wind resources, where topographical fea­ renewable resource. Interestingly, notice in Fig. 5(d) that the point near
tures reduce wind speeds in contrast to the other regions, as shown in coordinates (− 118◦ , 35◦ ) has the low FUR despite the evident abun­
Fig. 4(b) which is similar to what was reported by [80]. This justifies the dance of wind resources (as shown in Fig. 4(b)). This indicates that the
smaller H2 annual specific yield for WT-based system in such regions, as PEM runs at rated capacity only a few times throughout the year.
shown in Fig. 4(d) and the small optimal PEM size ratio (Fig. 3(b)). Considering the high availability of wind resources at that point, the
The frequency and magnitude of wind resources exceed those of solar optimization algorithm maximizes the size of the PEM electrolyzer
resources, leading to increased utilization of the PEM electrolyzer as (CRPEM is around 0.82 as shown in Fig. 3(b)) to increase the harnessing
shown in Fig. 5(a &b), respectively and hence, increased H2 production. of high wind energy instances however this causes a drop in the in­
The average annual PEM utilization factor (UF defined in Eq. (11)) for stances during which the PEM runs at rated capacity.
PV-based system ranges between 20 % and 32 % compared to the After determining the optimal PEM size ratio, it is now possible to
broader range of 10 % and 55 % for the WT-based system. This indicates estimate the CI and production cost of green hydrogen across CA as
that the PEM electrolyzer is not consistently utilized (at rated capacity), shown in Fig. 6. The heat maps of CI and production cost share similar
as shown in Fig. 5(c &d), where the FUR (as defined in Eq. (12)), features that regions with high CI coincide with high production cost
quantifies the proportion of time the PEM is utilized at its full capacity. regions which directly related to the specific H2 yield and specific

8
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

Fig. 4. Specific energy (first row, defined in Eq. (14)) and H2 (second row, defined in Eq. (14)) yields of PV (first column) and wind (second column) systems
across CA.

electricity yields shown in Fig. 4. These regions of high CI and produc­ (Fig. 6(a)) and wind (Fig. 6(b)), respectively. Whereas, the cost color
tion cost for WT-based systems are typically those with lower wind scale was set to a max of 4.4 and 7 USD/kg H2 for PV (Fig. 6(c)) and wind
speeds, as the geographical diversity of CA leads to variability in wind (Fig. 6(d)), respectively to enhance the readability of the figures.
frequency and magnitude, particularly when compared to the more
stable solar radiation patterns. It is clear in Fig. 6 (a& b) that wind-based
systems have lower CI almost in all regions across CA with the lowest 3.3. Comparative analysis
being in the south-east and northeast regions (less than 1 kg CO2eq/kg
H2) compared to less than 2 kg CO2eq/kg H2 in the south regions for PV- The reported CI of wind-based and PV-based systems are lower than
based system. Similar trends are observed in the H2 production cost the reported emissions for fossil fuel-based H2 technologies for CA (re­
Fig. 6(c& d) with the costs between 3 and 5.2 USD/kg H2 for PV-based ported to be 3.8 − 12 kg CO2eq/kg H2 [41]) whereas the H2 production
systems whereas for WT-based systems the costs are between 1.5 and 15 costs in RE-based systems are comparable with fossil fuel-based systems
USD/kg H2. Wind resources across CA have larger geographical varia­ when GHG mitigation costs are included (reported to be 2.73 − 5.94
tion in magnitude and frequency compared to solar resources which USD/kg H2 [41]). The H2 production costs in RE-based systems become
results in larger variation in the H2 CI and production cost for WT-based more competitive with fossil fuel-based considering the incentives
H2 systems compared to PV-based systems. These results emphasize that offered by the U.S. Inflation Reduction Act (IRA) under the 45 V Section.
CI and H2 production cost are location-specific and not equally created. The IRA 45 V offers 3 USD/kgH2 for H2 facilities with CI less than 0.45
As aforediscussed, the central region of CA has low wind resources kgCO2eq/kgH2 for 10 years which reduces the H2 production costs in CA
which results in high CI (up to 4.34 kg CO2eq/kg H2) and production cost by almost 1 USD/kg H2. Although embodied emissions in solar PV and
(up to 15 USD/kg H2) as shown in Fig. 6(b& d). It should be noted that wind turbines are included in this study, it should be noted that they are
the CI color scale was set to a max of 2.4 and 1.9 kg CO2eq/kg H2 for PV not included in IRA 45 V to determine CIs of H2 produced through water
electrolysis by solar and wind power (i.e., green H2 has a CI of 0 in IRA

9
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

Fig. 5. Average annual utilization factor of the PEM electrolyzer (first row) and the average annual full utilization ratio of PEM electrolyzer (second row). The first
column presents the results for PV-based hydrogen production system whereas the second column presents the results for the WT-based system.

45 V) [81]. considering different input parameters.

Moreover, we compare recent research in green hydrogen generation
with the current study in terms of CostH2 and CI for specific locations
3.4. Sensitivity analysis of H2 production cost and carbon intensity
around the world. In this study and depending on the location, the CostH2
ranges between 3 and 4.4 USD/kg H2 for PV-based and 1.5–15 USD/kg
Here, we investigate the sensitivity of the average H2 production cost
H2 for wind-based systems, respectively. These values align with the
and CI in CA (excluding the extreme results) considering different
value reported by the U.S. Department of Energy for green hydrogen
realistic minimum and maximum input techno-economic parameters
production in the U.S. (~5 USD/kg H2) [82]. Moreover, the obtained
reported in the literature, providing valuable insights into the factors
costs are comparable with the values reported in the literature as shown
driving cost variations as shown in Fig. 7. It is evident from Fig. 7 that
in Table 5. Additionally, the current study indicates relatively low CI
fluctuations in the reported PEM capital cost and PEM specific energy
values, with PV systems ranging from 1.58-2.95 kg CO2eq/kg H2 and
exert the most pronounced influence on the average H2 production cost.
wind systems ranging from 0.35-4.34 kg CO2eq/kg H2. Compared to the
PEM electrolyzer technology is not commercialized yet, and they have
referenced studies, the presented findings generally align with the
been used in pilot projects so far to investigate their feasibility. Addi­
existing research. It should be noted that the variation in the values
tionally, there are significant ongoing technological and manufacturing
reported in this study and the values in the literature is due to the
enhancements that vary among the research facilities causing a source of
variation in geographic location and the variation in the input param­
uncertainty in the PEM capital cost and PEM specific energy. Hence,
eters (capital costs and RES supply chain). Hence, in the following sec­
PEM technology possesses significant uncertainty in the average H2
tion we present the variation in the average CostH2 and CIH2 RES
production cost. Furthermore, while RE-specific techno-economic

10
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

Fig. 6. Carbon intensity (first row) and H2 cost without incentives (second row) of the RE-based systems with optimized PEM size ratio across CA. The first column
presents the results for PV-based hydrogen production system whereas the second column presents results for the WT-based system.

parameters also contribute to cost variations, their relative influence enhancing the efficiency of renewable energy systems to minimize the
appears to be secondary to that of PEM-related parameters. This suggests environmental footprint of green hydrogen production. Addressing un­
that advancements in RE technologies, while crucial for achieving sus­ certainties associated with reported PEM and RE parameters through
tainable hydrogen production, may have a comparatively lesser impact continued research and development efforts will be essential for real­
on mitigating cost fluctuations arising from variations in PEM izing the full environmental benefits of hydrogen as a clean energy
parameters. carrier. Moreover, fostering innovation and deployment of high-
Similarly, the sensitivity analysis reveals that fluctuations in the re­ efficiency renewable energy technologies will be instrumental in
ported PEM specific energy and the RE efficiency values exert a signif­ achieving sustainable hydrogen production pathways with minimal
icant influence on the CI of hydrogen production. It is important to note carbon emissions.
that while other parameters contribute to CI variations, the relative in­ Significant efforts have been invested in enhancing efficiency and
fluence of reported PEM specific energy and RE efficiency values ap­ reducing the costs of RES and H2 production systems which contribute
pears to be particularly pronounced. For instance, if 100 % of PV toward reducing the production cost of H2 and its CI. Hence, we look
modules came from APAC countries the CI would increase by ~ 2 %. into the potential reduction in H2 production cost (excluding the
Keeping in mind that currently almost 87 % of the PV modules used in extreme results) considering the increase in the RES performance
the U.S. are manufactured in APAC countries [43]. On the other hand, if (mainly PV efficiency, PEM specific energy, and WT height), the
all the PV modules used were manufactured in the U.S., the CI would reduction in RES capital costs, and operation and maintenance costs in
decrease by ~ 6 %. In contrast, less sensitivity of CI is observed for wind 2030. Moreover, we investigate the effect of exporting excess electricity
turbine manufacturing country (less than 1 %). This underscores the to the grid at fixed tariff (0.0268 USD/kWh) and selling O2 as co-product
critical importance of advancing PEM electrolyzer technology and on the H2 production cost (without incentives) as shown in Fig. 8. It can

11
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

Table 5 observed for the exclusion of low solar resources regions on the H2
A comparison between the current study and the other research. production cost in CA due to the smaller variation in the solar resources
Ref. RES Location CostH2 ($/kg CIH 2 RES (kg compared to wind-based systems.
H2) CO2eq/kg H2) Similarly, we investigate the reduction in the average H2 CI in CA
Current Study PV USA 3.0–5.2 1.58–2.95 (excluding the extreme results) considering the increase in the RES and
Wind USA 1.5–15 0.35–4.38 PEM performance in 2030, manufacturing PV and wind turbines in the
Al-Orabi et al. PV/ Egypt 3.94 1.74–1.83 U.S., and accounting for co-products credits as shown in Fig. 9. Inter­
[83] Wind estingly, accounting for excess electricity credits has the largest effect on
Jahangiri et al. Grid/ Chad 4.695 2.15
[84] PV/
the CI where it decreases the CI of H2 by almost 1.65 and 0.6 kg CO2eq/
Wind kg H2 for PV-based and wind-based systems, respectively. As afore dis­
Salameh et al. PV/ KSA 21 3.57–5.23 cussed, the PV system has lower utilization of PEM as shown in Fig. 5 (a
[85] Wind &c) since solar energy is only available in the day and due to the limited
Bareiβ et al. [55] PV/ Germany 3.3
−
PEM capacity (which was sized to ensure the maximum economic
Wind
Weidner et al. PV Switzerland 3–7 2.4 feasibility and minimum carbon footprint); hence, significant excess is
[86] available to be exported to the grid. Unlike wind energy that could be
Garcia and Oliva PV/ Chile 2.09–3.28 1.06–1.57 generated during day and night which increases the PEM utilization and
[36] Wind hence less excess available to be exported. Despite CA’s plans to reduce
Valente et al. Wind Spain 0.36–0.43
their CI of electricity, the credits given to the clean excess electricity are
−
[87]
BloombergNEF − − 4.5–12 − significant (0.1624 kg CO2eq/kWh in 2030) and contribute towards
[88] decreasing the carbon footprint of green H2 in CA. Moreover,
Clark et al. [80] Wind USA 2.46–10 − manufacturing the PV modules in the U.S. would decrease the H2 CI by
Al-Qahtani et al. Wind USA ~5.5 0.86
almost 0.3 kg CO2eq/kg H2 whereas manufacturing the wind turbines in
[25]
Al-Qahtani et al. PV USA ~9.5 3.1 the U.S. would decrease it by almost 0.08 kg CO2eq/kg H2. In R&D
[25] GREET 2023, wind turbines are already assumed to be manufactured in
the U.S. where the only change in this case would be the change in the CI
of electricity in 2030 (from 0.469 to 0.228 kg CO2eq/kWh, which is the
be deduced from Fig. 8 that the option of selling O2 could significantly average CI in the U.S.). Whereas for the PV modules, the change is more
reduce the average H2 production cost by almost 1 USD/kg H2 in both significant since R&D GREET uses the current supply chains of PV used
PV-based and wind-based systems, whereas the reduction in the PEM in the U.S. which is mostly manufactured in APAC countries with 0.677
capital cost reduces the average H2 production cost by 0.8 USD/kg H2. kg CO2eq/kWh CI of electricity [43]. Overall, the continuous advances in
Overall, the reduction in the RES and PEM capital and O&M and the the RES and PEM technologies could decrease the H2 CI on average to
increase in their efficiencies as well as the ability to sell the co-products − 0.7 and − 0.02 kgCO2eq/kgH2 for PV-based and wind-based systems,
could decrease the average H2 production cost in CA to − 0.06 and 0.67 respectively.
USD/kgH2 for PV-based and wind-based systems, respectively. It should
be noted that, if the analysis shown in Fig. 8 was applied only to the
4. Conclusions
regions with high wind potential (excluding low wind energy regions) in
CA, the H2 production cost for the wind-based systems could be negative
This study aims to investigate the impact of solar and wind resource
and smaller than the PV-based ones. On the contrary, smaller effect is
variations, ambient conditions, and PEM electrolyzer sizing on the CI of

Fig. 7. Sensitivity of H2 production cost (first row) and CI (second row) to the variation in the main technical and economic parameters: a& c) PV-based system and
b& d) WT-based system.

12
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

Fig. 8. Potential reduction in the average H2 production cost (without incentives) in CA considering predicted enhancement in performance and reduction in capital
cost as well as the ability to sell O2 and electricity as co-products: a) PV-based system and b) WT-based system.

Fig. 9. Potential reduction in the average H2 CI in CA considering predicted enhancement in performance as well as the credit for O2 and excess electricity: a) PV-
based system and b) WT-based system.

renewable electricity and CI and production cost of green hydrogen in PV-based and wind-based systems, respectively. In summary, the study
CA considering the supply chain of RES. The study achieves its primary underscores the significance of appropriately sizing PEM electrolyzers to
objectives by developing models correlating the performance of PV and maximize their utilization in RES-based hydrogen production systems.
wind systems with the CI of electricity in California. These models Employing hourly-based models is crucial for the sizing of the PEM
illuminate the intricate relationship between the capacity factors of electrolyzer and for evaluating the CI and production costs of green
these systems and the CI of electricity. Hourly variations in solar and hydrogen. Moreover, geographical variations in renewable energy re­
wind resources are taken into account in the analysis, revealing their sources and ambient conditions play pivotal roles in influencing the cost
significant influence on PEM optimal sizing as well as the hydrogen CI of hydrogen, as well as the CI of both renewable electricity and
and production costs. Various factors, including reductions in PEM and hydrogen.
RES costs, technological advancements, and the inclusion of co-prod­
ucts’ O2 and excess electricity credits and revenues, are examined to CRediT authorship contribution statement
understand their impact on hydrogen carbon intensity and production
costs. Geographical variations in CI and production costs are identified, Loiy Al-Ghussain: Writing – review & editing, Writing – original
highlighting consistently lower CI for wind-based systems with larger draft, Visualization, Validation, Supervision, Software, Project admin­
variations compared to solar-based systems. Notably, the results indi­ istration, Methodology, Investigation, Formal analysis, Data curation,
cate decreasing hydrogen production costs, reaching as low as − 0.06 Conceptualization. Mohammad Alrbai: Writing – review & editing,
and 0.67 USD/kg H2 for PV-based and wind-based systems, respectively. Writing – original draft, Conceptualization. Sameer Al-Dahidi: Writing
Continuous technological advancements are projected to further – review & editing, Writing – original draft, Conceptualization. Zifeng
decrease hydrogen CI on average to − 0.7 and − 0.02 kg CO2eq/kg H2 for Lu: Writing – review & editing, Resources, Conceptualization.

13
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

Declaration of competing interest Technologies Office under the Office of Energy Efficiency and Renew­
able Energy of the U.S. Department of Energy. The views and opinions of
The authors declare that they have no known competing financial the authors expressed herein do not necessarily state or reflect those of
interests or personal relationships that could have appeared to influence the U.S. Government or any agency thereof. Neither the U.S. Govern­
the work reported in this paper. ment nor any agency thereof, nor any of their employees, makes any
warranty, expressed or implied, or assumes any legal liability or re­
Data availability sponsibility for the accuracy, completeness, or usefulness of any infor­
mation, apparatus, product, or process disclosed, or represents that its
Data will be made available on request. use would not infringe privately owned rights. Argonne National Lab­
oratory is a U.S. Department of Energy laboratory operated by UChicago
Acknowledgement Argonne, LLC under Contract No. DE-AC02-06CH11357.

This work was supported by the Hydrogen and Fuel Cell

Appendix A

Fig. A1. Average hourly renewable energy resources and ambient temperature across CA: a) global horizontal radiation (GHI), b) wind speed at ground level (10 m),
and c) ambient temperature.

14
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

Fig. A2. Random sample of Pareto lines and the selection of the optimal point using TOPSIS: a) PV-based, and b) WT-based hydrogen systems.

Table A1
Specifications of the PV module (HiKu7_CS7N-MS) [47] and the wind turbine (Gamesa G114-2.0) used in
this study [49].

System Parameter Value Unit

PV Module rated capacity 670 W

ηPV,Ref 21.6 %
βRef − 0.0034 1/◦ C
TRef,STC 25 ◦
C
NOCT 41 ◦
C
TRef,NOCT 20 ◦
C
IRef 800 W/m2
Am 3.11 m2
Wind Turbine Pe,R 2000 kW
uC 2 m/s
uF 21 m/s
uR 9 m/s
z 140 m

References [11] Incer-Valverde J, Korayem A, Tsatsaronis G, Morosuk T. “Colors” of hydrogen:

Definitions and carbon intensity. Energy Convers Manag 2023;291:117294.
https://doi.org/10.1016/j.enconman.2023.117294.
[1] Al-Ghussain L, Ahmad AD, Abubaker AM, Hassan MA. Exploring the feasibility of
[12] Arcos JMM, Santos DMF. The Hydrogen Color Spectrum: Techno-Economic
green hydrogen production using excess energy from a country-scale 100% solar-
Analysis of the Available Technologies for Hydrogen Production. Gases 2023;3:
wind renewable energy system. Int J Hydrogen Energy 2022;47:21613–33. https://
25–46. https://doi.org/10.3390/gases3010002.
doi.org/10.1016/j.ijhydene.2022.04.289.
[13] Palmer G, Roberts A, Hoadley A, Dargaville R, Honnery D. Life-cycle greenhouse
[2] Donald R, Boulaire F, Love JG. Contribution to net zero emissions of integrating
gas emissions and net energy assessment of large-scale hydrogen production via
hydrogen production in wastewater treatment plants. J Environ Manage 2023;344:
electrolysis and solar PV. Energ Environ Sci 2021;14:5113–31. https://doi.org/
118485. https://doi.org/10.1016/j.jenvman.2023.118485.
10.1039/D1EE01288F.
[3] Alrbai M, Qawasmeh BR, Al-Dahidi S, Ayadi O. Influence of hydrogen as a fuel
[14] Terlouw T, Bauer C, McKenna R, Mazzotti M. Large-scale hydrogen production via
additive on combustion and emissions characteristics of a free piston engine.
water electrolysis: a techno-economic and environmental assessment. Energ
Therm Sci 2020;24:87–99. https://doi.org/10.2298/TSCI181211071A.
Environ Sci 2022;15:3583–602. https://doi.org/10.1039/D2EE01023B.
[4] Islam AK, Dunlop PSM, Hewitt NJ, Lenihan R, Brandoni C. Bio-Hydrogen
[15] Kolb S, Müller J, Luna-Jaspe N, Karl J. Renewable hydrogen imports for the
Production from Wastewater: A Comparative Study of Low Energy Intensive
German energy transition – A comparative life cycle assessment. J Clean Prod
Production Processes. Clean Technol 2021;3:156–82. https://doi.org/10.3390/
2022;373:133289. https://doi.org/10.1016/J.JCLEPRO.2022.133289.
cleantechnol3010010.
[16] Koroneos C, Dompros A, Roumbas G, Moussiopoulos N. Life cycle assessment of
[5] Xiao B, Liu J. Effects of various pretreatments on biohydrogen production from
hydrogen fuel production processes. Int J Hydrogen Energy 2004;29:1443–50.
sewage sludge. Chinese Sci Bull 2009;54:2038–44. https://doi.org/10.1007/
https://doi.org/10.1016/J.IJHYDENE.2004.01.016.
s11434-009-0100-z.
[17] Aydin MI, Dincer I. A life cycle impact analysis of various hydrogen production
[6] Thungklin P, Reungsang A, Sittijunda S. Hydrogen production from sludge of
methods for public transportation sector. Int J Hydrogen Energy 2022;47:
poultry slaughterhouse wastewater treatment plant pretreated with microwave. Int
39666–77. https://doi.org/10.1016/J.IJHYDENE.2022.09.125.
J Hydrogen Energy 2011;36:8751–7. https://doi.org/10.1016/j.
[18] Ozawa A, Inoue M, Kitagawa N, Muramatsu R, Anzai Y, Genchi Y, et al. Assessing
ijhydene.2010.07.073.
Uncertainties of Well-To-Tank Greenhouse Gas Emissions from Hydrogen Supply
[7] Ajanovic A, Sayer M, Haas R. The economics and the environmental benignity of
Chains. Sustain 2017, Vol 9, Page 1101 2017;9:1101. doi: 10.3390/SU9071101.
different colors of hydrogen. Int J Hydrogen Energy 2022;47:24136–54. https://
[19] Vilbergsson KV, Dillman K, Emami N, Ásbjörnsson EJ, Heinonen J, Finger DC. Can
doi.org/10.1016/j.ijhydene.2022.02.094.
remote green hydrogen production play a key role in decarbonizing Europe in the
[8] Hermesmann M, Müller TE. Green, Turquoise, Blue, or Grey? Environmentally
future? A cradle-to-gate LCA of hydrogen production in Austria, Belgium, and
friendly Hydrogen Production in Transforming Energy Systems. Prog Energy
Iceland. Int J Hydrogen Energy 2023;48:17711–28. https://doi.org/10.1016/J.
Combust Sci 2022;90:100996. https://doi.org/10.1016/j.pecs.2022.100996.
IJHYDENE.2023.01.081.
[9] Kothari R, Buddhi D, Sawhney RLL. Comparison of environmental and economic
[20] Mohideen MM, Subramanian B, Sun J, Ge J, Guo H, Radhamani AV, et al. Techno-
aspects of various hydrogen production methods. Renew Sustain Energy Rev 2008;
economic analysis of different shades of renewable and non-renewable energy-
12:553–63. https://doi.org/10.1016/j.rser.2006.07.012.
based hydrogen for fuel cell electric vehicles. Renew Sustain Energy Rev 2023;174:
[10] Dincer I, Acar C. Review and evaluation of hydrogen production methods for better
113153. https://doi.org/10.1016/j.rser.2023.113153.
sustainability. Int J Hydrogen Energy 2015;40:11094–111. https://doi.org/
10.1016/j.ijhydene.2014.12.035.

15
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

[21] Granovskii M, Dincer I, Rosen MA. Exergetic life cycle assessment of hydrogen [46] Al-Ghussain L, Ahmed H, Haneef F. Optimization of hybrid PV-wind system: Case
production from renewables. J Power Sources 2007;167:461–71. https://doi.org/ study Al-Tafilah cement factory. Jordan Sustain Energy Technol Assessments 2018;
10.1016/J.JPOWSOUR.2007.02.031. 30:24–36. https://doi.org/10.1016/j.seta.2018.08.008.
[22] Simons A, Bauer C. Life cycle assessment of hydrogen production. Transit to [47] CanadianSolar. HiKu5 Mono PERC. 2021.
Hydrog Pathways Towar Clean Transp 2011:13–57. https://doi.org/10.1017/ [48] Sadati SMS, Jahani E, Taylan O. Technical and Economic Analyses for Sizing PV
CBO9781139018036.006. Power Plant With Storage System for METU NCC. Vol. 6B Energy, ASME; 2015, p.
[23] Cetinkaya E, Dincer I, Naterer GF. Life cycle assessment of various hydrogen V06BT07A044. doi: 10.1115/IMECE2015-50959.
production methods. Int J Hydrogen Energy 2012;37:2071–80. https://doi.org/ [49] Gamesa. Greater Energy Produced From Low and Medium Wind Sites. Zamudio,
10.1016/J.IJHYDENE.2011.10.064. Spain: 2014.
[24] Schmidt Rivera XC, Topriska E, Kolokotroni M, Azapagic A. Environmental [50] Kelley N. Evaluation of Wind Shear Patterns at Midwest Wind Energy Facilities
sustainability of renewable hydrogen in comparison with conventional cooking Wavelet Analysis of Turbulence View project Wind Turbine Control Technology
fuels. J Clean Prod 2018;196:863–79. https://doi.org/10.1016/J. View project 2002.
JCLEPRO.2018.06.033. [51] Wang M, Elgowainy A, Lee U, Baek KH, Balchandani S, Benavides PT, et al.
[25] Al-Qahtani A, Parkinson B, Hellgardt K, Shah N, Guillen-Gosalbez G. Uncovering Greenhouse gases, Regulated Emissions, and Energy use in Technologies Model ®
the true cost of hydrogen production routes using life cycle monetisation. Appl (2023 Excel) 2023. doi: 10.11578/GREET-Excel-2023/dc.20230907.1.
Energy 2021;281:115958. https://doi.org/10.1016/J.APENERGY.2020.115958. [52] Kumar KSP, Gaddada S. Statistical scrutiny of Weibull parameters for wind energy
[26] Freire Ordóñez D, Halfdanarson T, Ganzer C, Shah N, Mac DN, Guillén-Gosálbez G. potential appraisal in the area of northern Ethiopia. Renewables Wind Water, Sol
Evaluation of the potential use of e-fuels in the European aviation sector: a 2015;2:14. https://doi.org/10.1186/s40807-015-0014-0.
comprehensive economic and environmental assessment including externalities. [53] Fragiacomo P, Genovese M. Numerical simulations of the energy performance of a
Sustain. Energy Fuels 2022;6:4749–64. https://doi.org/10.1039/d2se00757f. PEM water electrolysis based high-pressure hydrogen refueling station. Int J
[27] Gallardo F, García J, Monforti Ferrario A, Comodi G, Chiu JN. Assessing sizing Hydrogen Energy 2020;45:27457–70. https://doi.org/10.1016/j.
optimality of OFF-GRID AC-linked solar PV-PEM systems for hydrogen production. ijhydene.2020.07.007.
Int J Hydrogen Energy 2022;47:27303–25. https://doi.org/10.1016/j. [54] Corti HR. Polymer electrolytes for low and high temperature PEM electrolyzers.
ijhydene.2022.06.098. Curr Opin Electrochem 2022;36:101109. https://doi.org/10.1016/j.
[28] Fabianek P, Madlener R. Techno-Economic Analysis and Optimal Sizing of Hybrid coelec.2022.101109.
PV-Wind Systems for Hydrogen Production by PEM Electrolysis in California and [55] Yukesh Kannah R, Kavitha S, Preethi, Parthiba Karthikeyan O, Kumar G, Dai-Viet
Germany. 2021. NV, et al. Techno-economic assessment of various hydrogen production methods –
[29] Henry A, McStay D, Rooney D, Robertson P, Foley A. Techno-economic analysis to A review. Bioresour Technol 2021;319:124175. doi: 10.1016/j.
identify the optimal conditions for green hydrogen production. Energy Convers biortech.2020.124175.
Manag 2023;291:117230. https://doi.org/10.1016/j.enconman.2023.117230. [56] Irena. Renewable Generation Costs in 2022. Abu Dhabi 2022.
[30] Kalinci Y, Hepbasli A, Dincer I. Techno-economic analysis of a stand-alone hybrid [57] Gaur MK, Shrivastava A, Pandit RK. Performance Analysis and Sustainability
renewable energy system with hydrogen production and storage options. Int J Assessment of a Building Integrated Solar PV System. Int J Ambient Energy 2023:
Hydrogen Energy 2015;40:7652–64. https://doi.org/10.1016/J. 1–22. https://doi.org/10.1080/01430750.2023.2283762.
IJHYDENE.2014.10.147. [58] Wiser RH, Bolinger M. Benchmarking Anticipated Wind Project Lifetimes: Results
[31] Onar OC, Uzunoglu M, Alam MS. Modeling, control and simulation of an from a Survey of U.S. Wind Industry Professionals. E-Scholarship Repository,
autonomous wind turbine/photovoltaic/fuel cell/ultra-capacitor hybrid power Berkeley, CA (United States) 2019. https://doi.org/10.2172/1564078.
system. J Power Sources 2008;185:1273–83. https://doi.org/10.1016/J. [59] Krishnan S, Koning V, Theodorus de Groot M, de Groot A, Mendoza PG,
JPOWSOUR.2008.08.083. Junginger M, et al. Present and future cost of alkaline and PEM electrolyser stacks.
[32] Macedo SF, Peyerl D. Prospects and economic feasibility analysis of wind and solar Int J Hydrogen Energy 2023;48:32313–30. https://doi.org/10.1016/j.
photovoltaic hybrid systems for hydrogen production and storage: A case study of ijhydene.2023.05.031.
the Brazilian electric power sector. Int J Hydrogen Energy 2022;47:10460–73. [60] Irena. Green Hydrogen Cost Reduction: Scaling up Electrolysers to Meet the 1.5◦ C
https://doi.org/10.1016/J.IJHYDENE.2022.01.133. Climate Goal. Abu Dhabi 2020.
[33] Al-Buraiki AS, Al-Sharafi A. Hydrogen production via using excess electric energy [61] James B, Colella W, Moton J, Saur G, Ramsden T. PEM Electrolysis H2A Production
of an off-grid hybrid solar/wind system based on a novel performance indicator. Case Study Documentation. PEM Electrolysis H2A Prod Case Study Doc 2013:1–27.
Energy Convers Manag 2022;254:115270. https://doi.org/10.1016/j. [62] Parra Ramirez MA, Fogel S, Reinecke SF, Hampel U. Techno-Economic Assessment
enconman.2022.115270. of PEM Electrolysis for O2 Supply in Activated Sludge Systems—A Simulation
[34] Ourya I, Nabil N, Abderafi S, Boutammachte N, Rachidi S. Assessment of green Study Based on the BSM2 Wastewater Treatment Plant. Processes 2023;11:1639.
hydrogen production in Morocco, using hybrid renewable sources (PV and wind). https://doi.org/10.3390/pr11061639.
Int J Hydrogen Energy 2023;48:37428–42. https://doi.org/10.1016/J. [63] Bareiß K, de la Rua C, Möckl M, Hamacher T. Life cycle assessment of hydrogen
IJHYDENE.2022.12.362. from proton exchange membrane water electrolysis in future energy systems. Appl
[35] Nasser M, Hassan H. Thermo-economic performance maps of green hydrogen Energy 2019;237:862–72. https://doi.org/10.1016/j.apenergy.2019.01.001.
production via water electrolysis powered by ranges of solar and wind energies. [64] Al-Ghussain L, Abubaker AM, Darwish AA. Superposition of Renewable-Energy
Sustain Energy Technol Assessments 2023;60:103424. https://doi.org/10.1016/J. Supply from Multiple Sites Maximizes Demand-Matching: Towards 100%
SETA.2023.103424. Renewable Grids in 2050. Appl Energy 2021;284:116402. https://doi.org/
[36] Garcia GM, Oliva HS. Technical, economic, and CO2 emissions assessment of green 10.1016/j.apenergy.2020.116402.
hydrogen production from solar/wind energy: The case of Chile. Energy 2023;278: [65] Wetter M, Wright JA. Comparison Of A Generalized Pattern Search And A. Genetic
127981. https://doi.org/10.1016/J.ENERGY.2023.127981. Algorithm Optimization Method 2003.
[37] Hurtubia B, Sauma E. Economic and environmental analysis of hydrogen [66] Madić M, Radovanović M. Optimization of machining processes using pattern
production when complementing renewable energy generation with grid search algorithm. Int J Ind Eng Comput 2014;5:223–34. https://doi.org/10.5267/j.
electricity. Appl Energy 2021;304:117739. https://doi.org/10.1016/j. ijiec.2014.1.002.
apenergy.2021.117739. [67] Luiz Junho Pereira J, Antônio Oliver G, Brendon Francisco M, Simões Cunha Jr S,
[38] Bhandari R, Shah RR. Hydrogen as energy carrier: Techno-economic assessment of Ferreira GG. Multi-objective lichtenberg algorithm: A hybrid physics-based meta-
decentralized hydrogen production in Germany. Renew Energy 2021;177:915–31. heuristic for solving engineering problems. Expert Syst Appl 2022;187:115939.
https://doi.org/10.1016/j.renene.2021.05.149. https://doi.org/10.1016/j.eswa.2021.115939.
[39] Ghaebi Panah P, Cui X, Bornapour M, Hooshmand RA, Guerrero JM. Marketability [68] Ameur A, Berrada A, Loudiyi K, Adomatis R. Performance and energetic modeling
analysis of green hydrogen production in Denmark: Scale-up effects on grid- of hybrid PV systems coupled with battery energy storage. Hybrid Energy Syst
connected electrolysis. Int J Hydrogen Energy 2022;47:12443–55. https://doi.org/ Model, Elsevier 2021:195–238. https://doi.org/10.1016/B978-0-12-821403-
10.1016/J.IJHYDENE.2022.01.254. 9.00008-1.
[40] Ramadan MM, Wang Y, Tooteja P. Analysis of Hydrogen Production Costs across [69] International Renewable Energy Agency - IRENA. Making the breakthrough: Green
the United States and over the next 30 years 2022. hydrogen policies and technology costs. 2021.
[41] Bracci JM, Sherwin ED, Boness NL, Brandt AR. A cost comparison of various [70] International Renewable Energy Agency - IRENA. Future Of Solar Photovoltaic:
hourly-reliable and net-zero hydrogen production pathways in the United States. Deployment, investment, technology, grid integration and socio-economic aspects.
Nat Commun 2023;14:1–13. https://doi.org/10.1038/s41467-023-43137-x. 2019.
[42] Kolahchian Tabrizi M, Famiglietti J, Bonalumi D, Campanari S. The Carbon [71] Al-Ghussain L, Darwish Ahmad A, Abubaker AM, Hassan MA. Techno-economic
Footprint of Hydrogen Produced with State-of-the-Art Photovoltaic Electricity feasibility of thermal storage systems for the transition to 100% renewable grids.
Using Life-Cycle Assessment Methodology. Energies 2023;16. https://doi.org/ Renew Energy 2022;189:800–12. https://doi.org/10.1016/j.renene.2022.03.054.
10.3390/en16135190. [72] Solar Energy Technologies Office UD of E (DOE). 2030 Solar Cost Targets 2021.
[43] Gan Y, Elgowainy A, Lu Z, Kelly JC, Wang M, Boardman RD, et al. Greenhouse gas [73] Executive Summary 2019.
emissions embodied in the U.S. solar photovoltaic supply chain. Environ Res Lett [74] DOE: Office of Energy Efficiency & Renewable Energy. Wind Turbines: the Bigger,
2023;18:104012. https://doi.org/10.1088/1748-9326/acf50d. the Better n.d. https://www.energy.gov/eere/articles/wind-turbines-bigger-better
[44] Huld T, Müller R, Gambardella A. A new solar radiation database for estimating PV (accessed April 12, 2023).
performance in Europe and Africa. Sol Energy 2012;86:1803–15. [75] DOE: Hydrogen and Fuel Cell Technologies Office. Technical Targets for Proton
[45] Duffie JA, Beckman WA, Blair N. Solar engineering of thermal processes, Exchange Membrane Electrolysis n.d. https://www.energy.gov/eere/fuelcells/
photovoltaics and wind. John Wiley & Sons; 2020. technical-targets-proton-exchange-membrane-electrolysis (accessed April 12,
2023).

16
L. Al-Ghussain et al. Energy Conversion and Management 311 (2024) 118514

[76] Li R, Jin X, Yang P, Sun X, Zhu G, Zheng Y, et al. Techno-economic analysis of a [83] Al-Orabi AM, Osman MG, Sedhom BE. Analysis of the economic and technological
wind-photovoltaic-electrolysis-battery hybrid energy system for power and viability of producing green hydrogen with renewable energy sources in a variety
hydrogen generation. Energy Convers Manag 2023;281:116854. https://doi.org/ of climates to reduce CO2 emissions: A case study in Egypt. Appl Energy 2023;338:
10.1016/j.enconman.2023.116854. 120958. https://doi.org/10.1016/J.APENERGY.2023.120958.
[77] Boretti A, Castelletto S. Trends in performance factors of large photovoltaic solar [84] Jahangiri M, Soulouknga MH, Bardei FK, Shamsabadi AA, Akinlabi ET,
plants. J Energy Storage 2020;30:101506. https://doi.org/10.1016/j. Sichilalu SM, et al. Techno-econo-environmental optimal operation of grid-wind-
est.2020.101506. solar electricity generation with hydrogen storage system for domestic scale, case
[78] Esteban-Parra MJ, García-Valdecasas Ojeda M, Peinó-Calero E, Romero-Jiménez E, study in Chad. Int J Hydrogen Energy 2019;44:28613–28. https://doi.org/
Yeste P, Rosa-Cánovas JJ, et al. Climate Variability and Trends. Landsc. Sierra 10.1016/J.IJHYDENE.2019.09.130.
Nevada, Cham: Springer International Publishing; 2022, p. 129–48. doi: 10.1007/ [85] Salameh T, Sayed ET, Abdelkareem MA, Olabi AG, Rezk H. Optimal selection and
978-3-030-94219-9_9. management of hybrid renewable energy System: Neom city as a case study.
[79] Lee J-C-Y, Fields MJ, Lundquist JK, Lunacek M. Determining variabilities of non- Energy Convers Manag 2021;244:114434. https://doi.org/10.1016/J.
Gaussian wind-speed distributions using different metrics and timescales. J Phys ENCONMAN.2021.114434.
Conf Ser 2018;1037:072038. https://doi.org/10.1088/1742-6596/1037/7/ [86] Weidner T, Tulus V, Guillén-Gosálbez G. Environmental sustainability assessment
072038. of large-scale hydrogen production using prospective life cycle analysis. Int J
[80] Clark CE, Barker A, Brunik K, Kotarbinski M, Grant E, Roberts O, et al. Hydrogen Energy 2023;48:8310–27. https://doi.org/10.1016/J.
Opportunities for green hydrogen production with land-based wind in the United IJHYDENE.2022.11.044.
States. Energy Convers Manag 2023;296:117595. https://doi.org/10.1016/j. [87] Valente A, Iribarren D, Dufour J. Prospective carbon footprint comparison of
enconman.2023.117595. hydrogen options. Sci Total Environ 2020;728:138212. https://doi.org/10.1016/J.
[81] Doe. Office of Energy Efficiency & Renewable Energy. GREET 2023. https://www. SCITOTENV.2020.138212.
energy.gov/eere/greet. [88] BloombergNEF. Green Hydrogen to Undercut Gray Sibling by End of Decade. 2023
[82] U.s.. Department of Eneergy H and FCTO. Hydrogen Shot 2021. https://www.ener n.d. https://about.bnef.com/blog/green-hydrogen-to-undercut-gray-sibling-by-
gy.gov/eere/fuelcells/hydrogen-shot. end-of-decade/.

17