Climate Change Impacts on Gaseous Hydrogen (H2) Potential Produced by Photovoltaic Electrolysis for Stand-Alone or Grid Applications in Europe

Abstract

The EU’s hydrogen strategy consists of studying the potential for renewable hydrogen to help decarbonize the EU in a cost-effective way. Today, hydrogen accounts for less than 2% of Europe’s energy consumption. It is primarily used to produce chemical products. However, 96% of this hydrogen production is through natural gas, leading to significant amounts of CO₂ emissions. In this paper, we investigated PV electrolysis H₂ gas (noted H_2(g)) production for mapping this resource at Europe’s scale. The Cordex/Copernicus RCPs scenarios allow for evaluating the impact of climate changes on the H₂-produced mass and the equivalent energy, according to both extreme RCPs scenarios. New linear regressions are investigated to study the great dependence in H_2(g) produced masses (kg·yr⁻¹) and equivalent energies (MWh·yr⁻¹) for European countries. Computational scenarios are investigated from a reference year (2005) to the end of the century (2100) by steps of 5 years. According to RCPs 2.6 (favorable)/8.5 (extreme), 31.7% and 77.4% of Europe’s area presents a decrease of H_2(g)-produced masses between 2005 and 2100. For the unfavorable scenario (8.5), only a few regions located in the northeast of France, Germany, Austria, Romania, Bulgaria and Greece present a positive balance in H_2(g) production for supplying remote houses or smart grids in electricity and heat energy.

1. Introduction

Hydrogen appears to be an energy carrier of the future. According to Hunt et al. [1], oil and gas companies are beginning since the COVID crisis to invest heavily in sustainable hydrogen-based energy production technologies. A large part of the activities of these industries are in the transport, mobility and building sectors. Hydrogen production for the construction of a H₂ economy appears as a model of future development for these industries. As suggested by Salehi et al. [2] and Scheller et al. [3], companies whose activities or products are related to high-level greenhouse gas emissions will be affected by the need for significant restrictions on their CO₂ production. Several interesting works exist in the literature to assess the impact of carbon taxes on energy demand, CO₂ emissions or on the economy [4,5].

Green Hydrogen, produced by electrolysis from renewable energies, can also play a major role in reducing CO₂ emissions. According to Lagioia et al. [6], hydrogen energy can contribute to the decarbonization of the energy system in Europe for the actual century, for example, reducing emissions by 55% in 2030 compared to 1990 [7]. The notion of environmental cost is linked to environmental damage or degradation representing any action that deteriorates, damages or durably alters the quality or functioning of the environment, health and ecosystems, and/or the quantities of available natural resources. Thus, with an environmental cost that is too high for oil, coal and natural gas, and a climate change and global warming context, hydrogen produced by renewable energy sources can contribute to reduce the air pollution [8,9] and thus limit human health damages [10]. Moreover, hydrogen can help the world to obtain a carbon-neutral future [11].

The war situation in Ukraine since February 2022 complicates the energy supply of many countries with a significant increase in costs and fears of blackout for the winter of 2023 (especially in France). In this respect, we can see that some countries are highly dependent on electricity and gas imports. The future development of a mix of energy sources and vectors will contribute to a better response capacity to uncertain geopolitical contexts and will give countries greater sovereignty in terms of energy.

However, it is now established that the uses of hydrogen are currently limited by two main criteria: the lack of infrastructure and the high production costs [12]. Additionally, with a world population expected to reach nearly 10 billion in 2050 [13], energy consumption for transport and buildings will continue to grow up to the end of the century.

The high costs of deploying hydrogen technologies are due to both technical and economic parameters. Technically, the resource itself is rare in its pure state. Atomic hydrogen is everywhere—92% of the atoms in the universe are hydrogen atoms. However, H₂ is naturally very rare. It is necessary to produce it as an energy carrier. Even if it can be easily stored, hydrogen remains expensive to produce because of the energy cost of its production and its restitution, presenting a maximum efficiency of 30% (electrolysis-storage-fuel cell). In addition, platinum, an essential element used in fuel cells, remains a rare and therefore expensive raw material to move quickly to a generalization of this solution on a large scale. Economically, the cost of green hydrogen is above all that of electricity. With electrolysers installed at USD 1000/kW and a 50% load factor, electricity at USD 70/MWh represents 85% of the cost of hydrogen. Thus, the determining cost factor is the price of the electricity consumed since it makes up 60 to 80% of the complete costs of hydrogen production by electrolysis depending on the operating time [14].

The main energetic interest for hydrogen is its high energy density of 121 MJ·kg⁻¹ in comparison with gasoline or diesel fuels, respectively 47 and 44 MJ·kg⁻¹ [15]. Hydrogen can release about 4 to 5 times more energy in its combustion than biofuels currently used in France (only 25 MJ·kg⁻¹ [16]). Another interest of this gas lies in its capacity to be compressed for storage over long periods. Thus, for clean mobility, the main challenge is to manufacture tanks that can maintain high-pressure hydrogen (350 to 700 bars) without the hydrogen escaping.

Today, hydrogen is used in numerous applications for electrical and thermal energy production. Several papers reviewed the different hydrogen applications for massive or local energy production. Yue et al. [17] presented hydrogen applications in energy storage, power-to-gas systems, and fuel cell co- and tri-generation or mobility applications. Hydrogen makes it possible to increase medium- and long-term storage capacities and can also be used to shift the renewable resources across the seasons due to the seasonal difference in energy production. Moreover, hydrogen can be introduced in an existing natural gas network using a photovoltaic or wind power plant [18]. Using a Matlab software called ORIENTE (Optimization of Renewable Intermittent Energies with HydrogeN for auTonomous Electrification), the MYRTE (Mission hYdrogène Renouvelable pour l’inTégration au réseau Electrique) platform in France (Corsica island) has demonstrated the use of hydrogen to smooth the photovoltaic intermittent production [19,20] or to supply the grid during power peaks in the electricity consumption [21]. Today, hydrogen (liquid or gas) is considered for numerous applications [22], as a fuel for the future mobility [23], under water situations [24] or for power-to-gas strategies [25].

Taking into account that 770 million of people do not have access to electricity in 2019 vs. 860 million in 2018 [26], it is therefore logical to ask the question of how hydrogen could provide energy solutions to people considered in remote sites (single or grouped houses).

There are few articles in the literature on Europe-wide mapping of green hydrogen potential via photovoltaic electrolysis as an energy carrier for heat and power generation. A very recent review explores the production of green hydrogen at the scale of several regions of the world to show that this energy carrier can contribute to a transition towards a more environmentally friendly future energy but also improve energy independence in some countries [27]. Moreover, a few localized studies exist at the country level. Mouli-Castillo et al. [28] presented a recent paper on geological storage of hydrogen to meet regional heat production needs in the UK. In Niger (Africa), Bhandari [29] is working on the potential of green hydrogen obtained from solar energy and concludes that it would be easy to produce this energy carrier for the needs of the transportation and electricity sectors. Other studies concern Vietnam and Ecuador [30,31]. They all point to a growing interest in this energy vector in the needs of a country. However, these studies do not present perspectives of this potential by the end of the current century under the influence of global warming.

More researchconcerns hydrogen use on remote sites, i.e., not connected to an electrical network due to the distance from the power grid or the non-possible installation of renewable energy systems on protected ecological areas. In Africa, based on a multi-criteria analysis, a hybrid off-grid Wind/PV (photovoltaic)/FC (Fuel Cells)/electrochemical batteries system is preferred to produce energy in a small household in Nigeria [32]. Hydrogen coupled with wind turbine has demonstrated its interest to ensure uninterrupted power supply to an isolated zero-energy-house located at Catalca/Turkey [33]. Projects such as PEPITE in France (Projet d’Etudes et d’expérimentation de Puissance pour la gestion des énergies Intermittentes par les Technologies électrochimiques [34]) with H₂ under gas-state or for mobile network operators in Sub-Saharan Africa with H₂ under solid-state [35], supply off-grid telecom towers for mobile or TV telecommunications. Using temporal simulations, several papers have investigated the hybridization PV/Electrolyser with or without fuel cells. As an example, for the electricity and thermal supplies of a passive house, Motalleb et al. [36] simulated the behavior of the hybrid system for a yearly total heating load of 4395.7 kWh·yr⁻¹ and a daily maximum electrical DC peak of 300 W in the evening (21 h). Pal and Mukherjee [37] proposed an investigation based on the techno-economic feasibility assessment to choose the best configuration corresponding to off-grid PV/hydrogen fuel cell system applications in northeast India. Using HOMER-based optimization software, Mohammed et al. [38] determined the optimal design of a PV/FC hybrid system for the city of Brest in France. Typically, with a PV plant and FC of respectively 4200 kW and 2000 kW, corresponding to an electrolyser power of 3400 kW with 955 tons of hydrogen tank, a competitive COE (Cost Of Energy) is estimated at USD 0.120/ kWh⁻¹.

Taking into account this state of the art, our paper proposes a first mapping of H_2(g) gas potential produced by a PV (100 kW_p)/Electrolyser/H₂-H₂O storage tanks for Europe’s scale (12.5 $\times$ 12.5 km² grid) using climate change data by the end of the century. The equivalent H₂ energy will be considered for supplying electricity or heat energy for one or a few houses isolated from the grid or integrated into a smart grid.

Working from the latitude/longitude coordinates, the surface solar radiation downwards and the ambient temperature at 2 m extracted from the Coordinated Regional Climate Downscaling Experiment (CORDEX) database, this paper aims to simulate the behavior of the hybrid system leading to map the yearly electrolyzed H_2(g) mass (kg·yr⁻¹) and its equivalent energy (MWh·yr⁻¹) on the domain. Thus, mappings of hydrogen capabilities in the past and actual conditions at two spatial scales (Europe and France) are proposed. Using two extreme RCPs (Representative Concentration Pathways) scenarios (2.6, favorable and 8.5, unfavorable), climate change impact on solar irradiation and ambient temperature are investigated to determine the evolution of gaseous H_2(g) production and equivalent energy until the end of the 21st century.

The paper is organized as follows. After having reported on measurements and methods in Section 2, mainly concerning physical models used to characterize the behavior of each energy subsystem, Section 3 is devoted to the main results with (i) hydrogen and equivalent energy potential for a reference year 2005 and their evolution up to 2020, (ii) the determination of simple regressions to estimate yearly H_2(g) and energy potential in relation to latitude, (iii) the future tendencies of H_2(g) mass and equivalent energy for future decades up to 2100.

2. Material and Methods

2.1. Data Description

Regional climate models (RCM) data are provided by the CORDEX/COPERNICUS database [39]. Meteorological data are computed from a single level (Earth’s surface). High resolution CNRM-ALADIN 63 (CNRM: Centre Régional de Recherches Météorologiques; ALADIN: Aire Limitée Adaptation Dynamique Développement InterNational) Regional Climate Model from Météo-France coupled with the CNRM-CERFACS-CM5 Global Climate Model provide climate change information on regional and local scales in relatively fine detail. Gridded data present a 0.11° × 0.11° horizontal resolution (12.5 × 12.5 km²). Net CDF4 files (Network Common Data Form) are offered to download with air temperature (T_a, K) at 2 m from the ground, orography (m), latitude and longitude coordinates (°), surface solar radiation downwards (I, W.m⁻²) and total cloud cover (N, % of cloudy sky). Each proposed experiment is a simulation of the regional climate either for the past or future. Calendar data are Gregorian and grid mapping is Lambert conformal. Solar irradiances I are only considered for horizontal surfaces. Finally, one notes that we have worked with bias-adjusted CORDEX simulations downloaded on the similar platform with the same search options (see Section 2.2).

Historical data refer to experiments for the past using Global Climate Models’ (GCMs) boundary conditions [40]. These experiments cover a period for which modern climate observations exist. These experiments show how the RCMs performed for the past climate when forced by GCMs and can be used as a reference period for comparison with scenario runs for the future. In our case, the year N_ref = 2005 (called “historical” mode in this paper) will be considered as a reference for future experiments.

RCPs data refer to the experiments running for the future using both Representative Concentration Pathways (RCPs) forcing scenarios. They are registered for N = 2006 (first year with RCPs data) and all 5 years (N = 2010, 2015, …, 2100) with a 3 h time-step.

The mapping covers Europe from Norway to South of Morocco and from the Atlantic Ocean to Turkey (from 27° N to 72° N and from 22° W to 45° E). The domain is discretized in 453 × 453 pixels. RCPs’ hypotheses correspond to a variation in the radiative balance (difference between incoming and outgoing radiation) at the top of the troposphere (located between 10 and 16 km elevation), due to a change in one of the factors of climate change such as the concentration of greenhouse gases. In this study are considered the favorable RCP 2.6 (~3 W.m⁻², 490 ppm eq. CO₂) and unfavorable RCP 8.5 (>8.5 W.m⁻², >1370 ppm eq. CO₂) scenarios [41,42], thus providing two extreme pathways of the future climate forcing.

2.2. Data Quality and Bias Correction

CORDEX simulations are chains of climate models: socio-economic scenario, global climate models, regional climate models. They are not perfect because uncertainties creep in at each step. These uncertainties are mainly of two kinds: some intrinsic to the climate system and not reducible (chaotic variability) and others related to modeling errors that can be characterized and corrected (models are simplified representations of reality).

A recent article presents the issue of sources of uncertainty in projection climate data [43]. According to the authors, main uncertainties come from Global and Regional Climate Models (GCM/RCM) and forcing RCPs scenarios. Due to many missing combinations of emission scenarios and climate models leading to sparse scenario–GCM–RCM matrices, these large ensembles, however, are very unbalanced, which makes uncertainty analyses impossible with standard approaches.

According to [44], the simulated temporal evolution of future climate is subject to uncertainties that are tackled by different ensemble simulation strategies: (i) scenario uncertainty, (ii) internal climate variability and (iii) model uncertainty [45]. To evaluate the scenario uncertainty, a test with extreme multi-scenarios allows for covering a large bandwidth of future climate evolution, related to a systematic comparison with the historical climate simulations in order to derive projected climate change signals. Internal climate variability is simulated by models of the climate system [46]. Its temporal evolution strongly depends on the initialization of each model component. Multi-model ensemble simulations are based on a certain scenario, sample modelling uncertainties, but also different initial conditions of the climate system.

A complete bias correction study was applied to EURO-CORDEX data [47]. This work characterizes the quality of RCMs on a number of classical climate variables oriented towards their use in impact models. The results show that the simulations are globally too cold, too wet and too windy compared to the available observations and re-analyses for the period 1981–2010. Some simulations show strong systematic biases on temperature, others on precipitation or dynamic variables, but none of the models/simulations can be defined as best or worst on all criteria.

The CERFACS-CNRM-CM5-ALADIN63 coupling, chosen in this study, has shown its interest for the determination of wind power potential and intermittency issues in the context of climate change [48] or to stabilize energy supply with wind and solar photovoltaic energy under climate change scenarios in the Iberian Peninsula [49]. Another work chooses this hybrid model to study the climate projection in Brussels (Belgium), especially for air temperature variations [50]. Finally, the CNRM-ALADIN63/CM5 model was investigated for water scarcity studies in the Mediterranean climate [51]. All these applications conclude that this model compared to other approaches leads to comparable tendencies.

2.3. Specific Studied Sites

We consider two spatial scales for Europe (453 $\times$ 453 pixels) and France (~90 $\times$ 90 pixels). On each scale, five specific studied sites are chosen to their local specificities (topography, meteorology) and their interest in population density (Table 1). The chosen sites present a variability relative to climate classification (Köppen–Geiger; Cfb, Csa, Bsh, Dfb), GPS coordinates (Lat: 31.6° to 55.7° and Long: −8.0° to 21°) and altitude above sea level (between 1 and 647 m asl).

2.4. PV Model

At the scale of a PV plant with several tens of kW_p, PV production depends on solar irradiance (surface-down welling shortwave in the range 0.2–4 μm) but also on ambient air temperature and wind speed, which affect PV cell efficiency [52]. According to Wild et al. [53], a very strong correlation is mainly observed between irradiation and ambient temperature. Therefore, the PV production model implemented in this paper only includes these two significant meteorological parameters.

A literature survey presents different models for photovoltaic modules and plants. We consider here a PV plant with its DC-DC converter. Our approach integrates ambient temperature T_a (°C) to determine the cell temperature T_c (°C) using:

(1)$T_c=T_a+\frac{NOCT-20}{800}\times I$

where NOCT is the Normal Operating Cell Temperature according to the testing standard operational conditions of solar cells, defined as the temperature reached by open circuited cells in a module assuming 800 W.m⁻² irradiance, 20 °C ambient temperature and wind speed of 1 m·s⁻¹.

Taking into account losses due to connection or module design and layout, the mismatch efficiency can be calculated with:

(2)${\eta }_{loss}={\eta }_{cable}\times {\eta }_{conex }\times {\eta }_{ser}\times {\eta }_{par}$

Thus, the resultant PV MPPT power (W) is given by:

(3)$P_{mppt}={\eta }_{loss}\times N_{pv}\times \left(\frac{I}{I_{ref}}\right)\times \left(P_{nom\_mod}+\mu \times \left(T_c-T_{ref}\right)\right)$

where N_pv represents the number of modules of PV array. Taking into account the DC-DC converter efficiency (see Section 2.6), the PV power available for electrolysis is given by:

(4)$P_{mppt\left(DC/DC\right)}={\eta }_{DC/DC }\times P_{mppt}$

In this work, we consider a PV array of a 100 kW_p peak power operating on each time-step (3 h) of the CORDEX database. In terms of reference PV module, we chose the VertexS TSM-390 DE09.08 PV module distributed by TrinaSolar corresponding to a nominal PV power of P_{nom_mod} = 0.39 kW_p [54] (PV plant constituted with 100/0.39 $\approx$ 257 modules for an active area of about 494 m²). All technical parameters of the PV module and array are summarized in Table 2.

2.5. Electrolyser Model

For this study, we consider the technical characteristics of a Proton Exchange Membrane (PEM) electrolyser (EL) such as implemented on the Myrte’s R&D platform in Ajaccio, France [57] (Table 3). Produced by Helion hydrogen Power (Alstom Hydrogen SAS today), these PEM electrolyser systems are regulated to obtain a constant trend for the stack temperature (60 °C) and pressure (35 bars). Taking into account the wide time-step data from the Copernicus Cordex database (= 3 h), we neglect the phenomena or rise in temperature/pressure because they take place only during a few minutes of stack operation [58].

Classically, the electrical power output is approached with the I-V curve supplied by the company and presented in previous works [34,58]. Using the number of electrolyser cells N_EL (series association), its voltage V_EL (V), the active area of a single cell S_EL (cm²) and its current density J_EL (A.cm⁻²), the EL power P_EL (W) is given by:

(5)$P_{EL}=N_{EL}\times V_{EL}\times J_{EL}\times S_{EL}$

Firstly, we consider an EL operating range between its nominal power and a low threshold whose value corresponds to 10% [57] of its nominal value. Secondly, one notes that the simulation code integrates the auxiliaries’ power for the EL subsystem such as cooling, regulation, purification using the following equation:

(6)$P_{EL/Aux}=P_{EL/0}+C.P_{EL}$

where P_EL/0 (W) and C (%) respectively represent the no-load and proportional auxiliaries’ consumptions according to the EL production.

According to Faraday’s law [34], gaseous hydrogen H_2(g)/Oxygen O_2(g) production (mol·s⁻¹) and H₂O_(l) volume consumption (mol·s⁻¹) are determined using Faraday efficiency from [59]:

(7)$n_{H_{2}O}^C={\eta }_{F/EL}\times S_{H_{2}O}\times \frac{N_{EL}{I}_{EL}}{2F}$

(8)$n_{H_2}^P={\eta }_{F/EL}\frac{N_{EL}{I}_{EL}}{2F}$

(9)$n_{H_2}^P=\frac{n_{H_2}^P}{2}$

To characterize the operation of the electrolyser τ_i (%), we calculated the ratio between the operating time of the electrolyser $t_{EL}$ normalized with (i) the corresponding sunshine hours t_sun (i.e., only when I > 0) or (ii) the whole number of hours on the considered period t_tot (i.e., for one year, 8760 h = 3 $\times$ 2920 time-steps):

(10)${\tau }_{sun}(\%)=\frac{t_{EL}}{t_{sun}} \mathrm{and} {\tau }_{tot}(\%)=\frac{t_{EL}}{t_{tot}}$

2.6. H₂O_(l)/H_2(g) Storage Tanks

Electrolyzed water is extracted from the classical water network with a filtration and purification process. Considering that obtained O_2(g) flow is transferred to the ambient air (not valorized here), produced H_2(g) is sent to a storage tank (35 bars, no need for gas compression because the EL and the storage are at the same pressure). For each time-step, hydrogen storage ($n_{H_2}^P$, mol) and pumped H₂O_(l) quantity for electrolysis ($n_{H_{2}O}^C$, mol) are calculated from (dt = 3 h):

(11)${\dot{Q}}_{H_2}=\frac{d{Q}_{H_2}}{dt}=n_{H_2}^P-L_{H_2}$

(12)${\dot{Q}}_{H_{2}O}=\frac{d{Q}_{H_{2}O}}{dt}=n_{H_{2}O}^C$

where L_H2 represents the H_2(g) storage losses in the tank. In accordance with [34], we consider monthly losses equal to 0.01% of the maximal quantity that can be contained in the H_2(g) tank. No losses are considered in the liquid water tank.

If the hydrogen produced is not directly used on-site or is not intended to be consumed locally (for export or mobility use, for example), this energy carrier must be compressed to a higher level than its outlet pressure from the electrolyser (typically 30–35 bars for PEM technologies). The energy consumption of such a compressor is estimated at 2.7 kWh·kg⁻¹ H₂ [60] corresponding to about 8.2% of the H₂ lower heating (LHV) value (LHV_H2 = 33 kWh·kg⁻¹). It is not the case in our study.

2.7. DC-DC Converter

DC-DC converters are electronic devices used to efficiently convert direct current from one voltage to another. In a simplified approach, the DC-DC converter efficiency can be modelled by a constant value, generally taken as equal or greater than 90% [61]. In our work, the efficiency curve is modelled by classical equations [34,62], using η₁₀ et η₁₀₀ supplied by the manufacturer, respectively the converter efficiencies at 10% (η₁₀ = 97.1%) and 100% (η₁₀₀ = 97.8%) of the nominal power P_{DC-DC_nom}:

(13)${\eta }_0=\left(\frac{10}{{\eta }_{10}}-\frac{1}{{\eta }_{100}}-9\right)/99$

(14)${\eta }_1=\frac{1}{{\eta }_{100}}-{\eta }_0-1$

Thus,

(15)${\eta }_{DC-DC}\left(t\right)=\frac{P_{out}\left(t\right)}{P_{DC-DC\_nom}}\times \frac{1}{\left(\frac{P_{out}\left(t\right)}{P_{DC-DC\_nom}}+{\eta }_0+{\eta }_1\times {\left(\frac{P_{out}\left(t\right)}{P_{DC-DC\_nom}}\right)}^2\right)}$

2.8. Energy Flow

The simulated energy flow is resumed on Figure 1 for a given time-step. For a specific location defined by the Lat./Long. pixel coordinates, the available PV production is computed from the irradiance I and the ambient temperature T_a. This energy is transferred to the electrolyser in an operating range between its nominal power and its low operating threshold. Beyond the thresholds of the EL, the energy is lost. The molar quantities of hydrogen produced and liquid water consumed are calculated and impacted on the corresponding gas storage and the overall volume of water required for electrolysis. The simulation is repeated for each time-step on the same pixel and at the scale of all the map pixels over a period of 365 days (i.e., 365 days $\times$ 8 time-steps/day = 2920 time-steps).

3. Results—Discussion

3.1. H_2(g) Resource Mapping (Historical Data)

First, we worked with the year 2005 (historical data) for simulation inputs in order to describe the reference year. The objective is to obtain a first map of hydrogen resources for Europe corresponding to a 100 kW_p PV plant. For each pixel, the yearly output parameters consist of a triplet composed by the PV production (MWh·yr⁻¹), the gaseous H_2(g) produced volume (sum (Q_H2), kg·yr⁻¹) and the liquid H₂O_(l) needed for electrolysis (L.yr⁻¹). Figure 2 exhibits the first yearly reference mapping for H_2(g) mass (kg) in (a) Europe and (b) France (N_ref = 2005). Figure 2c,d presents the cumulated H_2(g) production mass profiles sum (Q_H2) according to the studied sites. Table 4 summarizes the main results at a local scale.

Concerning the historical data (year 2005), H_2(g) production in Europe shows a great dependency with latitude in relation to solar irradiation potential (weak dependence with longitude). A fairly clear boundary can be seen around latitude 45° N (Figure 2a) with H_2(g) mass exceeding 1 ton by year for latitudes lower than this threshold. Visually, a stratified repartition with latitude can be distinguished. For the most northern countries (latitude > 60° N), the estimated quantity of green H_2(g) remains very favorable with 700 to 800 kg H_2(g).yr⁻¹. The Sahara Desert clearly appears as a great location for H_2(g) production from the Atlantic coast to the Nile Delta in Egypt. However, the problem of access to water resources will arise in these regions. In the best conditions, some territories such as the southern regions of Morocco, Algeria, Libya or Egypt can exhibit H_2(g) productions exceeding 1.5 t H_2(g).yr⁻¹. Concerning Warsaw in Poland in Eastern Europe, the H₂ production corresponds to the quantities obtained at London i.e., 922.3 vs. 937.7 kg·yr⁻¹ (N_ref = 2005).

France presents a tempered climate with H_2(g) masses mainly in the range of 800 to 900 kg·yr⁻¹ (Figure 2b). The Southeast of the country can reach 1.1 to 1.2 t per year and the high mountainous areas, (alt. > 3000 m asl) such as the Alps and Pyrenees chains, exhibit best yields with 1.3 to 1.5 t H_2(g).yr⁻¹. There is undoubtedly an interesting perspective for hydrogen energy for the electric and heat supply in remote sites (mountain huts or pastoral shelters). Finally, on Corsica island, the most mountainous island of the Mediterranean Sea (max alt. = 2706 m, Monte Cinto) clearly shows a decrease in the production of H_2(g) from the seaside to the rural and mountainous areas (about 1.2 t for coastal areas to 0.9 t for mountainous locations relative to the presence of degraded sky conditions due to nebulosity and/or fog).

Figure 2c,d presents the yearly H_2(g) produced evolution for all studied sites. Clearly, we confirm both influences of seasonality and altitude on H_2(g) mass capacities. With the exception of Marrakech (Figure 2c) where the trend is practically linear throughout the year with sum (Q_H2(t)) = 0.508 $\times$ t (R² = 0.992), the other H_2(g) electrolyzed evolutions for all sites suffer on season influences (see rounded curves in winter periods, Figure 2c,d). For Copenhagen, the winter period is marked by a very small increase in the amount of H_2(g) produced (${\dot{Q}}_{H_2}\cong 0)$ due to the weak amount of sunlight available horizontally at these latitudes.

H_2(g) production capacities (kg·yr⁻¹) depend on several parameters as mainly the PV peak power (kW_p), latitude/longitude (°) and altitude (m) of the site, and the irradiation potential (W.m⁻²). For the 10 studied sites (Table 4), sum (Q_H2) has values in the range of 0.882 t and 1.4 t respectively obtained for Copenhagen and Marrakech. This highly energetic gas is the result of the electrolysis of liquid water for volume in the range of 8.9 to 14.2 m³.yr⁻¹. At a regional scale, one notes in France that the following repartition is observed in terms of H₂ produced: Paris $\approx$ Strasbourg < Toulouse < Marseille $\approx$ Ajaccio. Between Marseille and Paris, a substantial sum (Q_H2) difference of +32% is observed in one year.

The annual equivalent energy corresponds to 26.7 MWh·yr⁻¹ (minimum) and 42.4 MWh·yr⁻¹ (maximum). According to local specificities (mainly heating and domestic hot water needs), these equivalent energies could be corresponding to the supply of some remote grouped houses connected to a specific microgrid. For example, the average British household has 3 people living in it and uses 13.8 MWh of electricity [63]. In the same conditions (3 people by house), means of 20 MWh·yr⁻¹ and 2.3 MWh·yr⁻¹ are respectively observed in France and Morocco.

Finally, whatever the measurement site, the EL works practically 2/3 of the time of sunshine (i.e., when I > 0). The remaining third corresponds to periods when the horizontal irradiation is not sufficient to exceed the low threshold of EL (10% of its nominal power). During the sunny hours, the average operating time can reach about 3 h out of 4 in the best conditions (Marrakech, more than 74%). On the contrary, if the EL operating time is compared to the annual simulation time, the electrolysis time phenomenon only represents a minimum of 13.3% in London and a maximum of 15.5% in Marrakech.

3.2. Influence of Latitude Coordinates

Figure 3 presents (a) the correlation between annual H_2(g) produced quantities sum (Q_H2) (kg·yr⁻¹) and (b) the yearly cumulative PV energy sum (E_Pv-DC) (MWh·yr⁻¹), according to latitudes of the 10 studied locations. Between 31° N and 56° N, we investigate a linear regression between latitude and sum (Q_H2) or sum (E_H2), corresponding to the following equations (for a 100 kW_p PV plant):

(16)$sum(Q_{H2})=-25.5\times lat(°)+2257 ({\mathrm{R}}^2=0.897)$

(17)$sum(E_{H2})=-0.77\times lat(°)+68.41 ({\mathrm{R}}^2=0.898)$

These simple expressions quickly represent estimations of yearly H_2(g) potential mass and available energy under Cfb, Csa, Dfb and BSh Köppen–Geiger classifications in Europe. These results can be compared to those obtained by Dumas et al. [64] for an altitude of 0 m. Our study leads to H_2(g) capacities of 3.03, 2.25 and 1.48 kg·m⁻².yr⁻¹ for latitudes of 30°, 45° and 60°, respectively, while Dumas et al. [64] obtained 3.21, 2.17 and 1.68 kg·m⁻².yr⁻¹ of H₂ production by water electrolysis. This difference can be explained by the different choices of EL or PV modules characteristics.

One notes for Lat. = 0° (equator), the estimated sum (Q_H2) (~ 2.25 t·yr⁻¹) is not realistic (from Equation (16)). In accordance with Figure 2a, the maximum quantity observed does not exceed 1.8 t·yr⁻¹. According to [64], the predicted H₂(g) mass for latitude 0° is estimated at 1.8 kg·m⁻².yr⁻¹ in accordance with our simulation mapping for the southern regions of the map (Figure 2a).

3.3. Climate Change Scenarios (RCP)

3.3.1. Past Evolution (2005–2020)

The PV/EL hybrid system was simulated using predicted data for year 2020 according to both emission scenarios RCPs 2.6 and 8.5. Results were compared to reference values computed with historical data (Table 5).

Over the past fifteen years, the general trend shows a stabilization of the studied parameters: mass of produced H_2(g), quantity of consumed water, equivalent H_2(g) energy, operating time of the EL. For the 10 sites and all variables combined, both scenarios (RCPs 2.6/8.5) show for 42% of cases, a decrease of the parameters. Undoubtedly, these are the first results of a climate change already in dynamics as suggested by [65].

When reasoning at the site-specific level, some European regions are more affected by the hydrogen mass decrease or equivalent H_2(g) energy. This reduction is more pronounced when greenhouse gas emissions are higher (RCP 8.5). These areas are the cities of London (LD) and several sites in the south of France or in the western Mediterranean, particularly in Marseille and Ajaccio where 80% of statistics coefficients are decreasing. These reductions are not observed at other latitudes such as Strasbourg, Marrakesh, Copenhagen and Paris, where for the French capital, a significant increase in H_2(g) is observed (+0.9% and +7.2% for RCP2.6/8.5). To a lesser extent, for Marrakech is affected, +0.5% and +2.7% for RCP2.6/8.5.

Considered as a great hotspot for biodiversity and to confirm the previous conclusion, the Sixth Assessment Report of the Intergovernmental Panel on Climate Change dedicates a cross-chapter to Mediterranean regions, considering for the first time the Mediterranean Basin as an entity [66]. Experts have shown that virtually all parts of this area are vulnerable and face significant risks due to climate change (water scarcity, droughts, human health, ecosystem losses, resource decrease, etc.).

3.3.2. Typical Tendencies 2005–2100

• Cartography results

The tendencies of the predicted variables are computed from the time derivative each dt = 5 years (with one exception, between 2006 and 2009, only dt = 4 years). For H_2(g)-produced mass (kg·yr⁻¹) and equivalent energy (MWh·yr⁻¹), one thus considers, with $Y_y=Q_{H_2,y}$ the yearly H_2(g) quantities or $Y_y=E_{H_2,y}$, the yearly H_2(g) equivalent produced energy for the year y:

(18)${\dot{Y}}_{dt}=\frac{d{Y}_y}{dt}$

The cumulative function of ${\dot{Y}}_{dt}$ starting in 2005 (historical data, reference) and ending in 2100 gives the global variation of $Y_y$:

(19)$\Delta Y_y={\sum }_{2005}^{2100}{\dot{Y}}_{dt}=Y_y\left(2100\right)-Y_y\left(2005\right)$

This absolute function of $\Delta Y_y$ was computed in relative value (%) by dividing the absolute variation $\Delta Y_y$ with the result of the reference year 2005.

For Europe, respectively 68.3% (Figure 4a) and only 22.6% (Figure 4b) of the pixels show an increase in H_2(g) production according to RCPs 2.6/8.5 between 2005 and 2100. The effects of climate change due to the increased presence of greenhouse gases and higher soil evapotranspiration are undeniable. Respectively, 31.7%/77.4% (Figure 4a,b) of the mapping presents a decrease in H_2(g) produced quantities. In scenario 2.6, the most affected regions are limited to a geographical band in Western Europe covering a large part of Norway, Sweden and Spain but also a small circular area in the Mediterranean Sea covering Corsica and North Sardinia. The African continent is mainly impacted in Morocco, Algeria, Tunisia and Mauritania. Scenario 8.5 is much bleaker as the decrease in H_2(g) mass impacts the entire North African continent, the western part of the Mediterranean, Spain, Portugal and Northern Europe (UK, Norway, Sweden, Russia).

Statistical parameters for the mapping are summarized in Table 6. As described, the yearly averaged H_2(g) mass is estimated at 16.8 $\pm$ 9.2 kg·yr⁻¹ (i.e., 1.8%, min/max: −142.0/223.3 kg·yr⁻¹) and −31.3 $\pm$ 16.2 kg·yr⁻¹ (i.e., −2.7%, min/max: −251.5/176.8 kg·yr⁻¹) for RCPs 2.6/8.5. Considering that respectively the Standard Deviation (SD) and the Double Standard Deviation (DSD) represent 68.3% and 95.4% of the values around the mean, with SD = 9.2/16.2 for RCPs 2.6/8.5, we conclude that the changes for 2100 can be considered small considering the long period of experiments and the high values of standard deviations. When the DSD is bigger than the mean, predicting a number for changes is still very difficult. At most, we can predict an area to which the number may move towards. The observed minimum and maximum values represent only a few pixels of the map. Considering that the observed mean values will be lower than the DSD, we can conclude that this would represent 95% of the statistically possible results. Thus, in the majority of cases, small changes will be expected in the coming decades. This also indicates a form of inertia of the climate system where it is known that the current trend will continue until the middle of the century (up to 2050/2060) and will experience changes in the distant future. The favorable situation (RCP2.6) leads to 50% of the map surface (median value) presenting a mass balance greater than 17.3 kg·yr⁻¹. For the extreme scenario (RCP8.5), 50% of pixels (median value) present a mass balance lower than −27.4 kg·yr⁻¹. These results can be explained, among other things, by the increase in ambient temperatures that will necessarily influence the photovoltaic production and the performance of the cells as still demonstrated in a recent paper [67].

The most affected regions (ΔQ_H2,y > −100 kg·yr⁻¹) would be western Norway, Spain and Portugal, the French and Italian Alps, North Africa and the islands of the western Mediterranean (Corsica, Sardinia, Balearics islands). Only the eastern Mediterranean, part of the French and German regions and its eastern countries up to Greece and Turkey show an increase in this parameter by 2100.

On a global scale, the main results of our study are not consistent with the findings of the Crook et al. paper [68] showing a global increase in PV production in Europe of up to 8% between 1990 and 2080 (climate models HadGEM1 and HadCM3 under the IPCC SRES A1B scenario). However, our results could be expected following recent work [52] demonstrating the alteration of solar PV by the end of the century estimated in the range [−14%; +2%] compared with current climate conditions. Similar conclusions for PV-energy potential have been observed on other longitudes in China with a slight decrease of up to 6% in most of the study regions under RCP4.5 and RCP8.5 [69].

At a local scale, Burnett et al. [70] have demonstrated the impact of climate change on annual solar resources in the UK with a slight increase in southern UK and a marginal decrease in the north that correspond to our study for both RCPs 2.6 and 8.5 scenarios by the end of the century. Another study by Perez et al. [71] for Canary Islands (Spain) shows the influence of seasonality in PV potential under climate changes with a decrease in PV production during summer due to the rise in temperature (RCPs 4.5 and 8.5 for simulations covering the period 2090–2099). A good correlation of our study is obtained with the work of Panagea et al. [72], who demonstrate an increase in photovoltaic production of 4% in Greece despite an increase in ambient temperature of up to 3.5 °C and mean total radiation (up to 5 W.m⁻²). We obtain equally a good correlation with the study of Feron et al. [73] based on the RCP 4.5 scenario. The authors conclude slight changes in solar PV output in Central Europe (+5%) and the Arabian Peninsula (−4%).

More locally (Figure 4c,d), in mainland France (continental and Corsica island), the unfavorable scenario 8.5 leads to strong decreases in the west and in Corsica (between 0 and −70 kg·yr⁻¹) but especially in the southeast of the country (between −50 and −100 kg·yr⁻¹).

• results for studied cities

The yearly Q_H2 and E_H2 values were computed from N = 2006, 2010 to 2100 by a step of 5 years (20 values for a site and for an RCP scenario). Taking N_ref = 2005 (historical), the difference $\overline{Q_{H_2}}=\overline{Q_{H_2, N}-Q_{H_2, N_{ref}}}$ and $\overline{E_{H_2}}=\overline{E_{H_2, N}-E_{H_2, N_{ref}}}$ were calculated to obtain minimum and maximum values for all cities or for each RCP scenario. The results are summarized in Table 7.

Whatever the RCP scenarios, 5 cities out of 10 (AJ, LD, MD, MK, MS) of the studied sites will know a reduction in H_2(g) production mass. Except for LD, they are all located in southern Europe. The two scenarios lead to an increase in production only for the cities of Copenhagen, Paris, Strasbourg, Toulouse and Warsaw (50% of studied sites). However, one notes that for these cities, only Copenhagen and Warsaw are strongly influenced by the climate forcing scenario: the transition between the RCPs scenario (2.6 vs. 8.5) leads to a reduction in the H₂-produced increase during the 21st century (2.7% vs. 0.8% for Copenhagen and 2.4% vs. 0.4% for Warsaw). In contrast, the city of Strasbourg has the highest surplus with +4.3% (40.5 kg·yr⁻¹/N_ref) and +3.8% (35.6 kg·yr⁻¹/N_ref).

The following ranking is observed, AJ (−1.4% vs. −2.4% for RCPs 2.6/8.5) ~ LD ~ MS ~ MD > MK > TL > WS > CH > PR ~ SB (+4.3% vs. +3.8%) for the favorable scenario and already exhibits the critical specificity of western Mediterranean islands. For RCP 8.5, the Mediterranean regions and Spain (i.e., Madrid, MD) clearly appear as extreme regions that are the most impacted by climate changes. The recent literature has already identified this part of the world as the future most impacted region of the world concerning aridity and water scarcity [74], rainfall [75] and consequences on streamflow regime [76] or droughts [77].

4. Conclusions

Hydrogen appears to be an energy carrier of the future to limit greenhouse gas emissions and thus contain the impact of climate change. With its ecological and economic advantages, green hydrogen produced by water electrolysis using renewable energy sources will contribute to the deployment of gaseous hydrogen throughout Europe.

The current conflict in Ukraine creates a complex energy situation in Europe. It sheds light on the lack of energy sovereignty of certain countries and demonstrates the fragility of several European electricity production systems. However, this geopolitical context did not influence the research conducted in our article because of the easy access to online European climate databases such as ECMWF.

By using historical meteorological data from the ECMWF Copernicus/Cordex database (reference year 2005), we have obtained a reference mapping of the H₂ gas potential at the European scale by considering its production from a 100 kW_p PV array hybridized with a PEM electrolyser/storage tank (35 bars). Yearly H_2(g) production profiles have been established for 10 European cities thought Europe using two spatial scales (Europe and France). In these conditions, simulations have exhibited that the yearly electrolyser operating time can reach 74% (mean/min values: 68.1%/63.9%) when calculated in relation to the site’s sunlight hours. Brought back to the real time of the simulation, the ratios collapse to reach at best 15.5% (mean/min values: 14.2%/13.3%).

Under European latitudes between 31° N and 56° N, linear regressions were established to easily estimate the actual yearly H₂ gas mass production (kg·yr⁻¹) and its equivalent energy (MWh·yr⁻¹) relative to latitude coordinates (°) with the objective of valorization into electricity or domestic heat for remote areas or renewable energy smartgrids.

The 2005 reference mapping was used to experiment with the impacts of climate change on the H_2(g) mass production and equivalent energy over the period 2005–2100 according to two extreme scenarios, RCPs 2.6 and 8.5. For the least likely scenario (2.6), 2/3 of Europe’s surface will see its hydrogen production capacity increase up to 24.7% (mean/min values: 1.8%/−13.2%). However, for the most probable scenario (8.5), almost 3/4 of the European surface will experience a decrease in H_2(g) production averaging −2.7% +/− 26% and up to −28.5% at the end of the century. The western countries in Europe will be the most strongly impacted (Norway, northern UK, western France, Spain, Portugal, western Mediterranean and the whole of northern Africa). The regions located in East Europe such as Germany, Romania, Bulgaria, Croatia, Greece will know an increase in H_2(g) capacities than can reach 27.4% (RCP8.5). Considering that the observed mean values will be lower than the DSD, we can conclude that this would represent 95% of the statistically possible results. Thus, in the majority of cases, small changes will be expected in the coming decades. This also indicates a form of inertia of the climate system where it is known that the current trend will continue until the middle of the century (up to 2050/2060) and will experience changes in the distant future.

For decentralized applications (smartgrids, remote houses), power-to-gas technologies based on hydrogen (or synthetic methane) remain competitive in certain aspects compared to electrochemical batteries. Indeed, hydrogen offers storage capacities over several months up to 10 GWh, with a low electrical efficiency (30%) but a competitive global efficiency (heat/electricity, 80%). Electrochemical batteries offer daily or even weekly storage capacities of up to several tens of MWh. For identical storage capacities, hydrogen offers response times of about 100 ms compared to 1 min with compressed air. Thus, through innovative projects, the decentralized approach to energy production will aim to test the hydrogen injection at progressive rates (less than 20% in volume to start with) into a distribution network to supply a new district, or to develop clean mobility. This is the challenge to develop this technology for the future.

Author Contributions

Conceptualization, J.-N.A. and M.M.; Methodology, P.-A.M., J.-N.A. and M.M.; Software, P.-A.M. and M.M.; Investigation, P.-A.M.; Writing—original draft, P.-A.M.; Writing—review & editing, P.-A.M. and M.M. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

Data can be found on request at [email protected].

Acknowledgments

The authors acknowledge the innovative company Demeures Corses and its President/CEO (J.N.A.), to have proposed a student internship to P.A.M. author, student at the Paolitech engineering school of the University of Corsica.

Conflicts of Interest

The authors declare no conflict of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Tables

Figure 1. Synoptic of the green H2(g) production from PV/EL system.

Figure 2. H2(g) resource maps (kg·yr−1) for a 100 kWp PV plant (year 2005, historical data). (a): Europe; (b): France. H2(g) mass production profiles (kg·yr−1) versus time for year Nref = 2005: (c): Europe; (d): France. Red letters: city abbreviations (see Table 1).

Figure 2. H2(g) resource maps (kg·yr−1) for a 100 kWp PV plant (year 2005, historical data). (a): Europe; (b): France. H2(g) mass production profiles (kg·yr−1) versus time for year Nref = 2005: (c): Europe; (d): France. Red letters: city abbreviations (see Table 1).

Figure 3. (a) Latitude impacts on the H2(g) capacities (kg·yr−1). (b) Yearly PV cumulative energy (MWh·yr−1) for a 100 kWp PV plant. Histograms and Marks: blue square: year 2005 (historical); “+” (red): year 2020 (RCP 2.6, favorable, see Section 3.3.2); “[Forumla omitted. See PDF.] ” (black): year 2020 (RCP 8.5, unfavorable, see Section 3.3.2). Blue line: linear regression on year 2005 for historical data.

Figure 4. Future evolution (2100 compared to 2005) of relative H2(g) produced masses (%) according to both emission RCPs scenarios (left: low emission and right: high emission), respectively, for Europe (a,b) and France (c,d). Red letters: city abbreviations (see Table 1).

Table 1

Studied sites on each spatial scale for the determination of H₂ potential.

City	Country	Abbreviation	Lat. (°)	Long. (°)	Alt. (m asl.)	Köppen–Geiger Classification
Copenhagen	Denmark	CH	55.676	12.566	13	Cfb
London	UK	LD	51.507	−0.128	17	Cfb
Madrid	Spain	MD	40.417	−3.704	647	Csa
Marrakesh	Morocco	MK	31.630	−8.009	459	BSh
Warsaw	Poland	WS	52.229	21.012	100	Dfb
Paris	France	PR	48.857	2.352	41	Cfb
Strasbourg	France	SB	48.573	7.752	139	Cfb
Toulouse	France	TL	43.605	1.444	153	Cfb
Marseille	France	MS	43.296	5.369	1	Csa
Ajaccio	France	AJ	41.919	8.739	5	Csa
MIN	-	-	31.630	−8.009	1	-
MAX	-	-	55.676	21.012	647	-
MEAN	-	-	45.771	4.739	158	-
SD	-	-	6.666	7.909	209	-
MEDIAN	-	-	46.089	3.861	71	-

Table 2

Technical characteristics of the PV module (TrinaSolar, module VertexS TSM-390 DE09.08, Mono-Si) and the PV power model.

Variable	Parameter	Value	Unity
Pnom_mod	PV module nominal power (STC a)	$390 \pm 0/5$ d	W
-	Maximum PV module efficiency (STC)	20.3 d	%
-	PV module surface	1.922 d	m²
${\eta }_{cable}$	Cables efficiency (for line losses)	0.99	-
${\eta }_{conex }$	Connection efficiency	0.996 b	-
${\eta }_{ser}$	Series association efficiency	0.99 c	-
${\eta }_{par}$	Parallel association efficiency	0.98 c	-
μ	Temperature coefficient for Pnom_mod	−0.0034 d	%.°C−1
NOCT	Normal operating cell temperature	$43\pm {2 }^{\mathrm{d}}$	°C

^a: STC: I_ref: reference solar irradiation 1 000 W.m⁻², T_ref: temperature cell 25 °C, AM 1.5; ^b: [55]; ^c: [56]; ^d: [54].

Table 3

Technical parameters for the electrolyser subsystem [20,57,58].

Variable	Parameter	Value	Unity
$N_{EL}$	Electrolyser cells (series association)	60	-
$S_{EL}$	Active area of a single cell	290	cm2
$P_{EL/0}$	Auxiliary power with no-load	5	%
C	Auxiliaries’ consumption coefficient	17	%
$S_{H_{2}O}$	H2O stoichiometry	1.05	-
F	Faraday’s constant	96,485	C.mol−1
-	H2 production flow at nominal power	10	Nm3.h−1
-	Electrolyser operating temperature	20–60	°C
-	Electrolyser operating pressure	35	Bar

Table 4

Simulation results for yearly H_2(g) local capacities [kg·yr⁻¹], equivalent energy [MWh·yr⁻¹] and others physical parameters for year 2005 (historical data).

Abbrev. (see Table 1)	sum (QH2(g)) [kg·yr−1]	sum (QH2O(l)) [L.yr−1]	Equiv. En. [MWh·yr−1]	τsun [%]	τtot [%]
CH	882.4	8929.4	26.7	64.7	13.4
LD	937.7	9489.3	28.4	63.9	13.3
MD	1302.5	13,181.1	39.5	71.9	14.9
MK	1400.7	14,174.8	42.4	74.1	15.5
WS	922.3	9333.6	27.9	66.5	13.8
PR	943.7	9550.7	28.6	64.5	13.4
SB	936.5	9477.4	28.4	66.1	13.8
TL	1097.4	11,105.2	33.2	69.0	14.3
MS	1246.8	12,617.4	37.8	70.0	14.5
AJ	1265.2	12,804.3	38.3	70.5	14.7
MIN	882.4	8929.4	26.7	63.9	13.3
MAX	1400.7	14,174.8	42.4	74.1	15.5

Table 5

Past evolution (historical data, 2005) to actual situation (RCPs data, 2020). For year 2020, both extreme scenarios RCP 2.6 and RCP 8.5 are simulated. Red numbers correspond to a negative balance of the studied quantity in the period 2005–2020.

Abbrev. (see Table 1)	H2(g)		H2O(l)		Equiv. En.		τ sun		τtot
Scenarios	RCP 2.6	RCP 8.5	RCP 2.6	RCP 8.5	RCP 2.6	RCP 8.5	RCP 2.6	RCP 8.5	RCP 2.6	RCP 8.5
CH	3.6%	2.0%	3.6%	2.0%	3.7%	2.2%	0.9%	−0.2%	1.1%	0.0%
LD	−5.1%	−2.4%	−5.1%	−2.4%	−4.9%	−2.5%	−0.1%	1.0%	0.2%	1.3%
MD	0.1%	−0.9%	0.1%	−0.9%	0.0%	−1.0%	−0.8%	−1.8%	−0.6%	−1.6%
MK	0.5%	2.7%	0.5%	2.7%	0.7%	2.8%	1.6%	1.6%	1.7%	1.7%
WS	−0.1%	2.5%	−0.1%	2.5%	0.0%	2.5%	1.4%	1.8%	2.2%	2.2%
PR	0.9%	7.2%	0.9%	7.2%	1.0%	7.0%	2.4%	1.5%	2.7%	1.7%
SB	2.9%	5.2%	2.9%	5.2%	2.8%	4.9%	−0.7%	1.3%	−0.6%	1.5%
TL	−0.3%	1.8%	−0.3%	1.8%	−0.3%	2.1%	−1.0%	−1.6%	−0.8%	−1.5%
MS	−2.5%	−2.6%	−2.5%	−2.6%	−2.6%	−2.6%	0.1%	−1.4%	0.4%	−1.2%
AJ	−1.5%	−2.5%	−1.5%	−2.5%	−1.3%	−2.3%	0.0%	−1.2%	0.3%	−1.0%
MIN	−5.1%	−2.6%	−5.1%	−2.6%	−4.9%	−2.6%	−1.0%	−1.8%	−0.8%	−1.6%
MAX	3.6%	7.2%	3.6%	7.2%	3.7%	7.0%	2.4%	1.8%	2.7%	2.2%

Table 6

Climate change tendencies 2005–2100 (statistical parameters for Europe).

		$\mathit{\Delta }{\mathit{Q}}_{{\mathit{H}}_2,\mathit{y}}$		$\mathit{\Delta }{\mathit{E}}_{{\mathit{H}}_2,\mathit{y}}$
RCP	Statistics	kg·yr−1	%	MWh·yr−1
2.6	Min	−142.0	−13.2	−4.3
	Max	223.3	24.7	6.8
	Mean	16.8	1.8	0.5
	SD	9.2	1.7	0.3
	Median	17.3	1.7	0.5
8.5	Min	−251.5	−28.5	−7.6
	Max	176.8	27.4	5.4
	Mean	−31.3	−2.7	−1.0
	SD	16.2	2.6	0.5
	Median	−27.4	−2.9	−0.8

Table 7

Climate change tendencies for only the 10 studied sites between 2005 and 2100. Data are provided in absolute (kg·yr⁻¹ or MWh·yr⁻¹) and relative (%, parentheses) values. Italic and underlined characters respectively represent minimum and maximum values.

Cities	RCP	$\overline{{\mathit{Q}}_{{\mathit{H}}_2}}$	$\mathbf{SD}({\mathit{Q}}_{{\mathit{H}}_2, \mathit{N}}-{\mathit{Q}}_{{\mathit{H}}_2, {\mathit{N}}_{\mathit{r}\mathit{e}\mathit{f}})}$	$\overline{{\mathit{E}}_{{\mathit{H}}_2}}$	$\mathbf{SD}({\mathit{E}}_{{\mathit{H}}_2, \mathit{N}}-{\mathit{E}}_{{\mathit{H}}_2, {\mathit{N}}_{\mathit{r}\mathit{e}\mathit{f}}}$
	-	kg·yr−1	kg·yr−1	MWh·yr−1	MWh·yr−1
AJ	2.6	−17.5 (−1.4%)	14.1	−0.5 (−1.3%)	0.4
	8.5	−30.4 (−2.4%)	16.3	−0.9 (−2.3%)	0.5
PR	2.6	39.8 (4.2%)	23.5	1.2 (4.2%)	0.7
	8.5	36.5 (3.9%)	25.6	1.1 (3.8%)	0.8
MS	2.6	−13.2 (−1.1%)	15.3	−0.4 (−1.1%)	0.5
	8.5	−29.2 (−2.3%)	19.2	−0.9 (−2.4%)	0.6
TL	2.6	18.5 (1.7%)	20.1	0.6 (1.9%)	0.6
	8.5	13.1 (1.2%)	16.6	0.4 (1.3%)	0.5
SB	2.6	40.5 (4.3%)	31.5	1.2 (4.3%)	1.0
	8.5	35.6 (3.8%)	26.2	1.0 (−3.7%)	0.8
MK	2.6	−3.9 (−0.3%)	23.0	−0.1 (−0.2%)	0.7
	8.5	−26.4 (−1.9%)	28.4	−0.7 (−1.8%)	0.9
LD	2.6	−15.5 (−1.7%)	22.0	−0.5 (−1.6%)	0.7
	8.5	−14.2 (−1.5%)	20.1	−0.4 (−1.5%)	0.6
CH	2.6	23.9 (2.7%)	21.4	0.8 (2.9%)	0.7
	8.5	6.8 (0.8%)	29.5	0.3 (0.9%)	0.9
WS	2.6	22.1 (2.4%)	21.8	0.7 (2.6%)	0.7
	8.5	3.4 (0.4%)	25.2	0.1 (0.5%)	0.8
MD	2.6	−11.2 (−0.9%)	19.8	−0.4 (−0.9%)	0.6
	8.5	−43.5 (−3.3%)	24.6	−1.4 (−3.4%)	0.7
MIN (2.6)	-	−17.5 (−1.4%)	-	−0.5 (−1.6%)	-
MAX (2.6)	-	40.5 (4.3%)	-	1.2 (4.3%)	-
MIN (8.5)	-	−43.5 (−3.3%)	-	−1.4 (−3.4%)	-
MAX (8.5)	-	36.5 (3.9%)	-	1.1 (3.8%)	-

References

1. Hunt, J.D.; Nascimento, A.; Nascimento, N.; Werncke Vieira, L.; Romero, O.J. Possible pathways for oil and gas companies in a sustainable future: From the perspective of a hydrogen economy. Renew. Sustain. Energy Rev.; 2022; 160, 112291. [DOI: https://dx.doi.org/10.1016/j.rser.2022.112291]

2. Salehi, M.; Fahimifard, S.H.; Zimon, G.; Bujak, A.; Sadowski, A. The Effect of CO₂ Gas Emissions on the Market Value, Price and Shares Returns. Energies; 2022; 15, 9221. [DOI: https://dx.doi.org/10.3390/en15239221]

3. Scheller, C.; Schmidt, K.; Spengler, T. Effects of CO₂-Penalty costs on the production and recycling planning of lithium-ion batteries. Procedia CIRP; 2021; 98, pp. 643-647. [DOI: https://dx.doi.org/10.1016/j.procir.2021.01.168]

4. Runst, P.; Höhle, D. The German eco tax and its impact on CO₂ emissions. Energy Policy; 2022; 160, 112655. [DOI: https://dx.doi.org/10.1016/j.enpol.2021.112655]

5. Lin, B.; Jia, Z. How does tax system on energy industries affect energy demand, CO₂ emissions, and economy in China?. Energy Econ.; 2019; 84, 104496. [DOI: https://dx.doi.org/10.1016/j.eneco.2019.104496]

6. Lagioia, G.; Spinelli, M.P.; Amicarelli, V. Blue and green hydrogen energy to meet European Union decarbonisation objectives. An overview of perspectives and the current state of affairs. Int. J. Hydrogen Energy; 2022; in press [DOI: https://dx.doi.org/10.1016/j.ijhydene.2022.10.044]

7. Seck, G.S.; Hache, E.; Sabathier, J.; Guedes, F.; Reigstad, G.A.; Strauss, J.; Wolfgang, O.; Ouassou, J.A.; Askeland, M.; Hjorth, I. et al. Hydrogen and the decarbonization of the energy system in Europe in 2050: A detailed model-based analysis. Renew. Sustain. Energy Rev.; 2022; 167, 112779. [DOI: https://dx.doi.org/10.1016/j.rser.2022.112779]

8. Van Ruijven, B.; Lamarque, J.F.; Van Vuuren, D.P.; Kram, T.; Errens, H. Emission scenarios for a global hydrogen economy and the consequences for global air pollution. Glob. Environ. Change; 2011; 21, pp. 983-994. [DOI: https://dx.doi.org/10.1016/j.gloenvcha.2011.03.013]

9. Zhou, L.; Zhang, F.; Wang, L.; Zhang, Q. Flexible hydrogen production source for fuel cell vehicle to reduce emission pollution and costs under the multi-objective optimization framework. J. Clean. Product.; 2022; 337, 130284. [DOI: https://dx.doi.org/10.1016/j.jclepro.2021.130284]

10. Granovskii, M.; Dincer, I.; Rosen, M.A. Air pollution reduction via use of green energy sources for electricity and hydrogen production. Atmos. Environ.; 2007; 41, pp. 1777-1783. [DOI: https://dx.doi.org/10.1016/j.atmosenv.2006.10.023]

11. Yang Dong, Z.; Yang, J.; Yu, L.; Daiyan, R.; Amal, R. A green hydrogen credit framework for international green hydrogen trading towards a carbon neutral future. Int. J. Hydrogen Energy; 2022; 47, pp. 728-734. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2021.10.084]

12. Tarhan, C.; Cil, M.A. A study on hydrogen, the clean energy of the future: Hydrogen storage methods. J. Energy Storage; 2021; 40, 102676. [DOI: https://dx.doi.org/10.1016/j.est.2021.102676]

13. Word Population Data Sheet. Available online: https://interactives.prb.org/2020-wpds/ (accessed on 11 March 2022).

14. IFPEN, Institut Français du Pétrole Energies Nouvelles. Available online: https://polenergie.org/nos-actualites/lhydrogene-renouvelable-ou-decarbone-a-quel-cout/ (accessed on 13 December 2022). (In French)

15. Guibet, J.C. Carburants et Moteurs: Technologies, Énergie, Environnement; Publications de l’Institut Français du Pétrole: Technip, France, 1997; 830p.

16. Muselli, P.A.; Delmon, N. Hydrogène, Biocarburants: Carburants D’avenir. Technical project report (internship) Engineering and Energy Paolitech School, University of Corsica: Corsica, France, 24p.

17. Yue, M.; Lambert, H.; Pahon, E.; Roche, R.; Jemei, S. Hydrogen energy systems: A critical review of technologies, applications, trends and challenges. Renew. Sustain. Energy Rev.; 2021; 146, 111180. [DOI: https://dx.doi.org/10.1016/j.rser.2021.111180]

18. Melaina, M.W.; Antonia, O.; Penev, M. Blending Hydrogen into Natural Gas Pipeline Networks. A Review of Key Issues; Technical Report National Renewable Energy Laboratory: Golden, CO, USA, 2013; 131p.

19. Darras, C.; Muselli, M.; Poggi, P.; Voyant, C.; Hoguet, J.C.; Montignac, F. PV output power fluctuations smoothing: The MYRTE platform experience. Int. J. Hydrogen Energy; 2012; 37, pp. 14015-14025. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2012.07.083]

20. Poggi, P.; Darras, C.; Muselli, M.; Pigelet, G. The PV-Hydrogen MYRTE platform—PV output power fluctuations smoothing. Energy Procedia; 2013; 57, pp. 607-616. [DOI: https://dx.doi.org/10.1016/j.egypro.2014.10.215]

21. Darras, C.; Muselli, M.; Poggi, P.; Cristofari, C.; Le Pivert, X. Modeling and Simulating of A PV/H₂ Hybrid System for Reducing Load Peaks on An Electrical Grid?. ECS Trans.; 2007; 12, pp. 609-621. [DOI: https://dx.doi.org/10.1149/1.2921587]

22. Rasul, M.G.; Hazrat, M.A.; Sattar, M.A.; Jahirul, M.I.; Shearer, M.J. The future of hydrogen: Challenges on production, storage and applications. Energy Conver. Manag.; 2022; 272, 116326. [DOI: https://dx.doi.org/10.1016/j.enconman.2022.116326]

23. Ravi, S.S.; Aziz, M. Clean hydrogen for mobility—Quo vadis?. Int. J. Hydrogen Energy; 2022; 5, pp. 20632-20661. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2022.04.158]

24. Sezgin, B.; Devrim, Y.; Ozturk, T.; Eroglu, I. Hydrogen energy systems for underwater applications. Int. J. Hydrogen Energy; 2020; 47, pp. 19780-19796. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2022.01.192]

25. Fragiacomo, P.; Genovese, M. Technical-economic analysis of a hydrogen production facility for power-to-gas and hydrogen mobility under different renewable sources in Southern Italy. Energy Conver. Manag.; 2020; 223, 113332. [DOI: https://dx.doi.org/10.1016/j.enconman.2020.113332]

26. International Energy Agency. Available online: https://www.iea.org/reports/sdg7-data-and-projections/access-to-electricity (accessed on 11 March 2022).

27. Panchenko, V.A.; Daus, Y.V.; Kovalev, A.A.; Yudaev, I.V.; Litti, Y.V. Prospects for the production of green hydrogen: Review of countries with high potential. Int. J. Hydrogen Energy; 2022; in press [DOI: https://dx.doi.org/10.1016/j.ijhydene.2022.10.084]

28. Mouli-Castillo, J.; Heinemann, N.; Edlmann, K. Mapping geological hydrogen storage capacity and regional heating demands: An applied UK case study. Appl. Energy; 2021; 283, 116348. [DOI: https://dx.doi.org/10.1016/j.apenergy.2020.116348]

29. Bhandari, R. Green hydrogen production in West Africa—Case of Niger. Renew. Energy; 2022; 196, pp. 800-811. [DOI: https://dx.doi.org/10.1016/j.renene.2022.07.052]

30. Phap, V.M.; Sang, L.Q.; Ninh, N.Q.; Bonh, D.V.; Hung, B.B.; Huyen, C.T.; Tung, N.T. Feasibility analysis of hydrogen production potential from rooftop solar power plant for industrial zones in Vietnam. Energy Rep.; 2022; 8, pp. 14089-14101. [DOI: https://dx.doi.org/10.1016/j.egyr.2022.10.337]

31. Posso, F.; Pulido, A.; Acevedo-Pàez, J.C. Towards the hydrogen economy: Estimation of green hydrogen production potential and the impact of its uses in Ecuador as a case study. Int. J. Hydrogen Energy; 2022; in press [DOI: https://dx.doi.org/10.1016/j.ijhydene.2022.05.128]

32. Babatunde, O.M.; Munda, J.L.; Yamam, Y. Hybridized off-grid fuel cell/wind/solar PV /battery for energy generation in a small household: A multi-criteria perspective. Int. J. Hydrogen Energy; 2022; 47, pp. 6437-6452. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2021.12.018]

33. Firtina-Ertis, I.; Acar, C.; Erturk, E. Optimal sizing design of an isolated stand-alone hybrid wind-hydrogen system for a zero-energy house. Appl. Energy; 2020; 274, 115244. [DOI: https://dx.doi.org/10.1016/j.apenergy.2020.115244]

34. Darras, C.; Sailler, S.; Thibault, C.; Muselli, M.; Poggi, P.; Hoguet, J.C.; Melscoet, S.; Pinton, E.; Grehant, S.; Gailly, F. et al. Sizing of photovoltaic system coupled with hydrogen/oxygen storage based on the ORIENTE model. Int. J. Hydrogen Energy; 2010; 35, pp. 3322-3332. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2010.01.060]

35. Jansen, G.; Dehouche, Z.; Corrigan, H. Cost-effective sizing of a hybrid Regenerative Hydrogen Fuel Cell energy storage system for remote & off-grid telecom towers. Int. J. Hydrogen Energy; 2021; 46, pp. 18153-18166.

36. Motalleb, M.; Dukic, A.; Firak, M. Solar hydrogen power system for isolated passive house. Int. J. Hydrogen Energy; 2015; 40, pp. 16001-16009. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2015.09.134]

37. Pal, P.; Mukherjee, V. Off-grid solar photovoltaic/hydrogen fuel cell system for renewable energy generation: An investigation based on techno-economic feasibility assessment for the application of end-user load demand in North-East India. Renew. Sustain. Energy Rev.; 2021; 149, 111421. [DOI: https://dx.doi.org/10.1016/j.rser.2021.111421]

38. Mohammed, O.H.; Amirat, Y.; Benbouzid, M.; Elbast, A. Optimal Design of a PV/Fuel Cell Hybrid Power System for the City of Brest in France. Proceedings of the IEEE ICGE 2014; Sfax, Tunisia, 25–27 March 2014; pp. 119-123.

39. ECMWF. Available online: https://cds.climate.copernicus.eu/cdsapp#!/dataset/projections-cordex-domains-single-levels?tab=overview (accessed on 1 October 2022).

40. Euro-Cordex Community. Guidance for EURO-CORDEX Climate Projections Data Use. Available online: https://www.euro-cordex.net (accessed on 14 November 2022).

41. Moss, R.H.; Edmonds, J.A.; Hibbard, K.A.; Manning, M.R.; Rose, S.K.; Van Vuuren, D.P.; Kram, T. The next generation of scenarios for climate change research and assessment. Nature; 2010; 463, pp. 747-756. [DOI: https://dx.doi.org/10.1038/nature08823] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20148028]

42. O’Neill, B.C.; Carter, T.R.; Ebi, K.; Harrison, P.A.; Kemp-Benedict, E.; Kok, K.; Kriegler, E.; Preston, B.L.; Riahi, K.; Sillmann, J. et al. Achievements and needs for the climate change scenario framework. Nat. Clim. Change; 2020; 10, pp. 1074-1084. [DOI: https://dx.doi.org/10.1038/s41558-020-00952-0] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33262808]

43. Evin, G.; Somot, S.; Hingray, B. Balanced estimate and uncertainty assessment of European climate change using the large EURO-CORDEX regional climate model ensemble. Earth Syst. Dynam.; 2021; 12, pp. 1543-1569. [DOI: https://dx.doi.org/10.5194/esd-12-1543-2021]

44. EURO-CORDEX Guidelines, Version 1.1. Guidance for EURO-CORDEX Climate Projections Data Use. 2021; Available online: https://www.euro-cordex.net/imperia/md/content/csc/cordex/guidance_for_euro-cordex_climate_projections_data_use__2021-02_1_.pdf (accessed on 8 December 2022).

45. Hawkins, E.; Sutton, R. The potential to narrow uncertainty in projections of regional precipitation change. Clim. Dynam.; 2011; 37, pp. 407-418. [DOI: https://dx.doi.org/10.1007/s00382-010-0810-6]

46. Deser, C.; Knutti, R.; Solomon, S.; Phillips, A.S. Communication of the Role of Natural Variability in Future North American Climate. Nat. Clim. Change; 2012; 2, pp. 775-779. [DOI: https://dx.doi.org/10.1038/nclimate1562]

47. Vautard, R.; Kadygrov, N.; Iles, C.; Boberg, F.; Buonomo, E.; Bülow, K.; Coppola, E.; Corre, L.; van Meijgaard, E.; Nogherotto, R. et al. Evaluation of the large EURO-CORDEX regional climate model ensemble. J. Geophys. Res.-Atmos.; 2021; 126, e2019JD032344. [DOI: https://dx.doi.org/10.1029/2019JD032344]

48. Cai, Y.; Bréon, F.M. Wind power potential and intermittency issues in the context of climate change. Energy Conv. Manag.; 2021; 240, 114276. [DOI: https://dx.doi.org/10.1016/j.enconman.2021.114276]

49. Costoya, X.; DeCastro, M.; Carvalho, D.; Arguillé-Pérez, B.; Gomez-Gesteira, M. Combining offshore wind and solar photovoltaic energy to stabilize energy supply under climate change scenarios: A case study on the western Iberian Peninsula. Renew. Sust. Energy Rev.; 2022; 157, 112037. [DOI: https://dx.doi.org/10.1016/j.rser.2021.112037]

50. Duchene, F.; Hamdi, R.; Schaeybroeck, B.V.; Caluwaerts, S.; De Troch, R.; De Cruz, L.; Termonia, P. Downscaling ensemble climate projections to urban scale: Brussels’s future climate at 1.5 °C, 2 °C, and 3 °C global warming. Urban Clim.; 2022; 46, 101319. [DOI: https://dx.doi.org/10.1016/j.uclim.2022.101319]

51. Soares, P.M.M.; Lima, D.C.A. Water scarcity down to earth surface in a Mediterranean climate: The extreme future of soil moisture in Portugal. J. Hydrol.; 2022; 615, 128731. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2022.128731]

52. Jerez, S.; Tobin, I.; Vautard, R.; Montavez, J.P.; Lopez-Romero, J.M.; Thais, F.; Bartok, B.; Christensen, O.B.; Colette, A.; Déqué, M. et al. The impact of climate change on photovoltaic power generation in Europe. Nat. Commun.; 2015; 6, 10014. [DOI: https://dx.doi.org/10.1038/ncomms10014] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26658608]

53. Wild, M.; Ohmura, A.; Makowski, K. Impact of global dimming and brightening on global warming. Geophys. Res. Lett.; 2007; 34, L04702. [DOI: https://dx.doi.org/10.1029/2006GL028031]

54. TrinaSolar. Technical Documentation for VertexS TSM-390 DE09.08 Module. 2022; Available online: www.trinasolar.com (accessed on 9 November 2022).

55. Shen, L.; Li, Z.; Ma, T. Analysis of the power loss and quantification of the energy distribution in PV module. Appl. Energy; 2020; 260, 114333. [DOI: https://dx.doi.org/10.1016/j.apenergy.2019.114333]

56. Picault, D. Reduction of Mismatch Losses in Grid-Connected Photovoltaic Systems Using Alternative Topologies. PhD Thesis; Institut National Polytechnique de Grenoble—INPG: Grenoble, France, 2010; Available online: https://tel.archives-ouvertes.fr (accessed on 15 September 2022).

57. Espinosa-Lopez, M.; Darras, C.; Poggi, P.; Glises, R.; Baucourb, P.; Rakotondrainibe, A.; Besse, S.; Serre-Combe, P. Modelling and experimental validation of a 46 kW PEM high pressure water electrolyser. Renew. Energy; 2018; 119, pp. 160-173. [DOI: https://dx.doi.org/10.1016/j.renene.2017.11.081]

58. HELION Hydrogen Power Company. Technical Documentation 2010; HELION Hydrogen Power Company: Aix-en-Provence, France, 2010.

59. Uzunoglu, M.; Onar, O.C.; Alam, M.S. Modeling, control and simulation of a PV/FC/UC based hybrid power generation system for stand-alone applications. Renew. Energy; 2009; 34, pp. 509-520. [DOI: https://dx.doi.org/10.1016/j.renene.2008.06.009]

60. FCH-JU. Study on Early Business Cases for H₂ in Energy Storage and More Broadly Power to H₂ Applications; Institution of Gas Engineers and Managers (IGEM): Brussels, Belgium, 2017.

61. Ursua, A.; Martin, I.S.; Barrios, E.L.; Sanchis, P. Stand-alone operation of an alkaline water electrolyser fed by wind and photovoltaic systems. Int. J. Hydrogen Energy; 2013; 38, pp. 14952-14967. [DOI: https://dx.doi.org/10.1016/j.ijhydene.2013.09.085]

62. Labbe, J. L’hydrogène Électrolytique Comme Moyen de Stockage D’électricité Pour Systèmes Photovoltaïques Isolés. PhD Thesis; Ecoles des Mines de Paris: Paris, France, 21 December 2006; Available online: https://pastel.archives-ouvertes.fr (accessed on 15 September 2022).

63. WorldData. Available online: https://www.worlddata.info/africa/morocco/energy-consumption.php (accessed on 15 November 2022).

64. Dumas, A.; Trancossi, M.; Anzillotti, S.; Madonia, M. Photovoltaic Production of Hydrogen at Stratospheric Altitude. J. Solar. Energy Eng.; 2013; 135, 011018. [DOI: https://dx.doi.org/10.1115/1.4007357]

65. WMO. Available online: https://public.wmo.int/en/media/press-release/climate-change-indicators-and-impacts-worsened-2020 (accessed on 17 November 2022).

66. IPCC. Climate Change 2022: Impacts, Adaptation and Vulnerability, Working Group II Contribution to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change; IPCC: Geneva, Switzerland, 3068p.

67. Djeziri, M.A.; Benmoussa, S.; Tapia Sanshez, R.; Palais, O.; Tina, G.M. Solar cell modeling in normal and degraded operations for simulation and monitoring. Sustain. Energy Technol. Assess.; 2022; 51, 101990. [DOI: https://dx.doi.org/10.1016/j.seta.2022.101990]

68. Crook, J.A.; Jones, L.A.; Forstera, P.M.; Crook, R. Climate change impacts on future photovoltaic and concentrated solar power energy output. Energy Environ. Sci.; 2011; 4, pp. 3101-3109. [DOI: https://dx.doi.org/10.1039/c1ee01495a]

69. Zhao, X.; Huang, G.; Lu, C.; Zhou, X.; Li, Y. Impacts of climate change on photovoltaic energy potential: A case study of China. Appl. Energy; 2020; 280, 115888. [DOI: https://dx.doi.org/10.1016/j.apenergy.2020.115888]

70. Burnett, D.; Barbour, E.; Harrison, G.P. The UK solar energy resource and the impact of climate change. Renew. Energy; 2014; 71, pp. 333-343. [DOI: https://dx.doi.org/10.1016/j.renene.2014.05.034]

71. Perez, J.C.; Gonzalez, A.; Díaz, J.P.; Exposito, F.J.; Felipe, J. Climate change impact on future photovoltaic resource potential in an orographically complex archipelago, the Canary Islands. Renew. Energy; 2019; 133, pp. 749-759. [DOI: https://dx.doi.org/10.1016/j.renene.2018.10.077]

72. Panagea, I.S.; Tsanis, I.K.; Koutroulis, A.G.; Grillakis, M.G. Climate Change Impact on Photovoltaic Energy Output: The case of Greece. Advanc. Meteorol.; 2014; 11, 264506. [DOI: https://dx.doi.org/10.1155/2014/264506]

73. Feron, S.; Cordero, R.R.; Damiani, A.; Jackson, R.B. Climate change extremes and photovoltaic power output. Nat. Sustain.; 2021; 4, pp. 270-276. [DOI: https://dx.doi.org/10.1038/s41893-020-00643-w]

74. Mendes, M.P.; Rofriguez-Galiano, V.; Aragones, D. Evaluating the BFAST method to detect and characterise changing trends in water time series: A case study on the impact of droughts on the Mediterranean climate. Sci. Total Environ.; 2022; 846, 157428. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2022.157428]

75. Marras, P.A.; Lima, D.C.A.; Soares, P.M.M.; Cardoso, R.M.; Medas, D.; Dore, E.; De Giudici, G. Future precipitation in a Mediterranean island and streamflow changes fora small basin using EURO-CORDEX regional climate simulations and the SWAT model. J. Hydrol.; 2021; 603, 127025. [DOI: https://dx.doi.org/10.1016/j.jhydrol.2021.127025]

76. De Girolamo, A.M.; Barca, E.; Leone, M.; Lo Porto, A. Impact of long-term climate change on flow regime in a Mediterranean basin. J. Hydrol. Region. Stud.; 2022; 41, 101061. [DOI: https://dx.doi.org/10.1016/j.ejrh.2022.101061]

77. Gomez-Gomez, J.D.; Pulido-Velazquez, D.P.; Collados-Lara, A.J.; Fernandez-Chacon, F. The impact of climate change scenarios on droughts and their propagation in an arid Mediterranean basin. A useful approach for planning adaptation strategies. Sci. Total Environ.; 2022; 820, 153128. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2022.153128]