Evaluating the Potential and Limits of Green Electrolysis in Future Energy Scenarios with High Renewable Share

Abstract

Water electrolysis is a potential contributor to global decarbonization, enhancing the flexibility and resilience of the electricity system and enabling integration with different sectors, such as industry and transportation, by acting as an energy vector and storage, as well as chemical feedstock. This study investigates the potential of hydrogen production by electrolysis in future national electric grid scenarios for Italy as a case study. It examines the impact of increasing photovoltaic and wind capacities up to five times the 2019 levels, considering an electricity storage capacity of up to 200 GWh. The feasibility of fully meeting current national hydrogen consumption through electrolysis in these scenarios is assessed by considering different overall electrolysis capacities. Specific CO₂ emissions associated with hydrogen production are evaluated as an indicator of environmental feasibility and compared with the conventional steam methane reforming. In addition, the levelized cost of hydrogen production is evaluated as an indicator of economic feasibility. Some limitations of electrolysis emerge when it is considered the sole way to decarbonize hydrogen production. Very high renewable shares are required to make electrolysis alone a feasible solution. Aiming to maximize the use of renewable curtailment for electrolysis conflicts with maximizing the electrolyzers’ utilization factor, thus, negatively affecting hydrogen production costs. Furthermore, since priority is given to the use of renewable and stored electricity to meet electricity demand, the remaining electricity is insufficient to produce the entire hydrogen demand in most of the considered scenarios, particularly when substantial storage supports the grid, as this reduces the curtailment available for electrolysis.

1. Introduction

The increasing concern about climate change and efforts for decarbonization have brought the potential of hydrogen as an energy carrier to the forefront. Hydrogen is not directly available in its molecular form in nature, apart from rare exceptions such as the Bourakebougou hydrogen field in Mali. Instead, it is usually found combined with other elements in compounds such as water or hydrocarbons. Its production is mainly based on fossil fuels, in particular, steam methane reforming (SMR), and results in more than 900 Mt of CO₂ emissions [1]. Almost all current global hydrogen demand comes from refining and industry, mainly as a feedstock [2]. However, the increase in renewable energy (RE) capacities in electricity generation causes increasing amounts of power curtailment in some hours of the day and periods of the year. Furthermore, the variability and unpredictability of photovoltaic (PV) and wind sources, the two fastest growing renewable sources, call for more flexible and highly integrated energy systems. In this context, hydrogen and oxygen as a by-product, can be produced from renewable electricity through electrolysis (green hydrogen) without associated CO₂ emissions. In addition, electrolysis allows for the use of otherwise curtailed electrical overproduction. Hydrogen can then be further converted into other chemicals or fuels (ammonia, methane, or fuels for aviation). Moreover, electrolytic hydrogen could substitute fossil-based hydrogen as feedstock in industry [3], especially in the hard-to-abate sectors, and as a heat source for processes requiring high temperatures. In this way, it contributes to decarbonizing industry and constitutes an alternative to the coupling of traditional hydrogen production processes with carbon capture and storage systems (blue hydrogen). Hydrogen can also serve as storage for balancing inter-seasonal mismatching between electricity production and consumption and be converted back into electricity through fuel cells or combustion engines when the electric demand exceeds production. It can also be injected into existing natural gas networks [4], with specific hydrogen blending limits that currently generally range from 5% to 20% in volume.

Several studies have focused on water electrolysis and power-to-gas systems in scenarios with high renewable energy penetration. Among them, Caumon et al. [5] investigated the impact of hydrogen production flexibility on French power systems. They found that the production of significant amounts of competitive low-carbon hydrogen was possible, with costs ranging from EUR 3 to 5/kg, only in scenarios with very high RE penetration and high interconnection capacities. Grube et al. [6] conducted an analysis on the quantitative potential, cost, and environmental impact of electrolytic hydrogen production, transmission, and storage in Germany. They considered two cases: one where electricity curtailment was utilized for hydrogen production and the other where the electrolyzer operated based on market prices. The second case yielded the lowest levelized cost of hydrogen (LCOH), amounting to EUR 3.63/kg.

National strategies for hydrogen development have been published by several countries. In particular, the European Union (EU) [7] set the goals of installing 6 GW of electrolyzers and producing up to 1 million tons of renewable hydrogen by 2024. These goals increase to 40 GW of electrolyzers and 10 million tons of renewable hydrogen by 2030. In June 2023, the EU adopted a strict definition of renewable hydrogen and its derivatives, named ‘renewable liquid and gaseous transport fuels of non-biological origin’. With the Delegated Act 2023/1184 [8], the EU defined rules about the additionality of renewable generation and temporal/geographical correlation between that generation and electrolysis. Kakoulaki et al. [9] evaluated the capacity of replacing carbon-intensive hydrogen production with RE-based water electrolysis in the EU. They concluded that the available technical potential for green electricity is enough to provide all current electricity demand as well as reach the current annual EU hydrogen production through RE-based electrolysis. Lagioia et al. [10] reviewed production and handling methods of green and blue hydrogen exploitation in the EU to meet the 2030 targets. They pointed out that the large use of green hydrogen will be feasible only when RE sources are able to decarbonize the entire electricity demand, or rather when additional renewable energy surplus is available for green hydrogen production. In addition, hydrogen strategies should focus primarily on its use in sectors that cannot take advantage of electrification, such as heavy industry and heavy-duty transport [10].

In Italy, the updated Integrated National Energy and Climate Plan [11] set the 2030 national renewable share targets towards decarbonization: 63.4% in the electricity sector, 35.9% in the thermal sector, and 34.2% in the transport sector. To accomplish these established goals, significant increases in PV and wind capacities are necessary, while other renewable sources are expected to largely maintain their current levels. In 2030, the electricity generated from solar energy needs to be more than three times the amount generated in 2019, and that from wind should be double the 2019 level [12]. As outlined in the National Hydrogen Strategy Preliminary Guidelines [13], Italy aims to fulfill 2% of its energy demand by producing green hydrogen by 2030. This will involve the installation of about 5 GW of electrolyzers, leading to an associated reduction of 8 million tons of CO₂ emissions. By 2050, hydrogen is projected to cover up to 20% of the energy consumption. According to Crespi et al. [14], a target of 250–300 thousand tons per year of renewable hydrogen by 2030 (480,000 tons was the 2019 hydrogen consumption [15]) should be considered realistic, but certainly not ambitious, for Italy.

In a previous study [16], the authors investigated the quantitative potential of electrolytic hydrogen production in future scenarios in Italy with increasing PV and wind capacities by considering several installed electrolysis capacities. The results highlighted some critical aspects of hydrogen production from power curtailment. Indeed, producing hydrogen only from power curtailment generally leads to low electrolysis utilization factors (UF) and, consequently, high costs. Alternatively, by allowing the use of additional non-curtailed fossil fuel-based grid electricity at the same overall installed electrolysis capacity, the electrolyzers’ UF can be increased at the cost of increasing specific CO₂ emissions.

Short-term storage will be essential for future grid needs to help smooth electricity production and increase the share of RE in the total supply of electricity. According to the Italian transmission system operator (Terna) [17], in the scenario “Fit-For-55”, which assumes an increase in PV and wind capacities to, respectively, about 3.6, and 2.5 times their 2019 levels, the total Italian need for additional storage capacity will be about 95 GWh by 2030, excluding the existing pumped storage. In a previous study [16], the presence of additional electric storage (ES) compared to the year 2019 was not considered. This was a limitation of the analysis since, by considering additional electric storage capacity, the amount of curtailed electricity available for electrolysis is further reduced.

In this study, the impact of the presence of additional ES capacity on the quantitative potential of hydrogen production from electrolysis has been evaluated. Future Italian electric grid scenarios with overall PV and wind capacities of up to five times the 2019 levels are analyzed by including additional ES capacity of up to 200 GWh.

Then, the feasibility and techno-economic implications of producing all the current national hydrogen consumption through electrolysis in the analyzed scenarios are discussed by considering different overall electrolysis installed capacities and evaluating the levelized cost of hydrogen production. The impact on specific CO₂ emissions associated with electrolytic hydrogen produced is also evaluated and compared with the conventional SMR method.

Although the quantitative results are specific to the Italian context, the main findings can be generalized. A key strength of this study is having developed and applied a relatively simple method to the Italian context. Although the quantitative results are country-specific, this method could also be applied to other countries with similar characteristics, such as other European countries. Its relative simplicity does not compromise the key general insights obtained; rather, it helps highlight them more effectively.

2. Future Scenarios of the Italian Electric System with Increased PV and Wind Penetration

The year chosen as the reference for the analyzed scenarios of the electricity generation mix in Italy is 2019, i.e., the last year before the COVID-19 pandemic began. Hourly national electricity load and generation profiles by source for 2019 were taken from the ENTSO-E platform [18]. The annual renewable share of the total electricity load, annual RE production, and installed capacity in 2019, divided by source [18,19], are shown in Figure 1.

The investigated scenarios of the electricity generation mix are the combinations of multiples of PV and wind overall installed power in 2019, ranging from 1 up to five times (as in [16]). Hence, a total of 25 (5 × 5) scenarios are considered. Each scenario is denoted in the format “PV×m W×n”, where “m” and “n” are the multiplication factors of installed PV power and wind power, respectively. To provide a point of reference for what these installed capacity values correspond to, according to Kakoulaki et al. [9], the annual Italian PV technical potential is about 247 TWh of ground PV plus about 94 TWh (88.7 TWh according to Bódis et al. [20]) of rooftop PV. Instead, wind potential is 282.2 TWh of offshore wind and 12.6 TWh of onshore wind [9]. Therefore, both PV and wind technical potentials are more than 10 times the corresponding 2019 generation and much greater than the annual generation at the maximum RE-penetration taken into account in this study.

The electricity demand was assumed to be the same in all the scenarios, and hydrogen logistics/distribution were not taken into account. These assumptions are simplified but represent a more optimistic case for electrolytic hydrogen production. Indeed, the national electric load is projected to increase because of the progressive electrification of sectors like transport and residential heating. As a consequence, with the same renewable generation, the electricity available for electrolysis would be less.

For each scenario, the presence of additional electric storage (ES) was included by investigating several energy and power capacities of the ES. The energy capacities (${\mathrm{C}}_{\mathrm{E}\mathrm{S}}$) considered are: 0 (baseline case without ES), 50, 100, and 200 GWh. For reference, 95 GWh is the total Italian need for additional storage capacity by 2030 in the scenario “Fit-For-55”, net of the existing pumped storage [17]. As for the power capacities, the energy-to-power (E/P) ratios considered for each energy capacity are 4 h, 6 h, and 8 h. E/P ratios indicate the minimum time for the complete charge (or discharge) of the overall ES. Charging and discharging power capacities are assumed to be the same. A round-trip efficiency (RTE) of 90% is assumed for the ES.

2.1. Management of the Electric Storage

The hourly generated electricity exceeding the demand is, as a priority, stored as much as possible in the electric storage (ES). Secondly, it is used by the electrolyzers if the ES is completely charged or if the surplus power exceeds the ES power capacity. As soon as the electric load exceeds RE production, the ES provides all the amount it can to cover the additional required power, compatibly with the amount of energy stored and the ES power capacity. The ES serves only the electrical demand while the electrolyzers are not operated with electricity from the ES.

The ES state of charge (SOC) is determined at each time step k as follows:

$$\mathrm{S}\mathrm{O}\mathrm{C}\left(\mathrm{k}\right)=\mathrm{S}\mathrm{O}\mathrm{C}\left(\mathrm{k}-1\right)+{\displaystyle \frac{{\mathrm{E}}_{\mathrm{R}\mathrm{E}\to \mathrm{E}\mathrm{S}}\left(\mathrm{k}\right)-\frac{1}{\mathrm{R}\mathrm{T}\mathrm{E}}\cdot {\mathrm{E}}_{\mathrm{E}\mathrm{S}\to \mathrm{l}\mathrm{o}\mathrm{a}\mathrm{d}}\left(\mathrm{k}\right)}{{\mathrm{C}}_{\mathrm{E}\mathrm{S}}}}$$

(1)

where ${\mathrm{E}}_{\mathrm{R}\mathrm{E}\to \mathrm{E}\mathrm{S}}$ is the amount of RE that is stored in the ES and ${\mathrm{E}}_{\mathrm{E}\mathrm{S}\to \mathrm{l}\mathrm{o}\mathrm{a}\mathrm{d}}$ is the electricity from the ES used to cover the load, already net of the losses due to the round-trip efficiency (RTE). Therefore, in Equation (1), ${\mathrm{E}}_{\mathrm{E}\mathrm{S}\to \mathrm{l}\mathrm{o}\mathrm{a}\mathrm{d}}$ is divided by RTE to include these losses.

At the beginning of the simulation, the ES is assumed to be empty (SOC(1) = 0).

Such a management strategy of the ES maximizes the RE share in the electric load coverage. In addition, it minimizes the annual RE curtailment in the case of no electrolysis capacity.

The annual RE share in electric load coverage is defined as the sum of the annual RE electricity supplying the load (${\mathrm{E}}_{\mathrm{R}\mathrm{E}\to \mathrm{l}\mathrm{o}\mathrm{a}\mathrm{d}}$) and the possible electricity from the ES (${\mathrm{E}}_{\mathrm{E}\mathrm{S}\to \mathrm{l}\mathrm{o}\mathrm{a}\mathrm{d}}$), excluding losses due to RTE, divided by the annual electric load:

$$\mathrm{R}\mathrm{E} \mathrm{s}\mathrm{h}\mathrm{a}\mathrm{r}\mathrm{e}\left[\mathrm{\%}\right]=100\cdot {\displaystyle \frac{{\sum }_{\mathrm{k}=1}^{8760} \left[{\mathrm{E}}_{\mathrm{R}\mathrm{E}\to \mathrm{l}\mathrm{o}\mathrm{a}\mathrm{d}}\left(\mathrm{k}\right)+{\mathrm{E}}_{\mathrm{E}\mathrm{S}\to \mathrm{l}\mathrm{o}\mathrm{a}\mathrm{d}}\left(\mathrm{k}\right)\right]}{{\sum }_{\mathrm{k}}^{8760}{\mathrm{E}}_{\mathrm{l}\mathrm{o}\mathrm{a}\mathrm{d}}\left(\mathrm{k}\right)}}$$

(2)

The annual RE curtailment before the possible production of electrolytic hydrogen is calculated as the sum of the hourly difference between RE production (${\mathrm{E}}_{\mathrm{R}\mathrm{E}}$) and the amount of RE that either supplies the load (${\mathrm{E}}_{\mathrm{R}\mathrm{E}\to \mathrm{l}\mathrm{o}\mathrm{a}\mathrm{d}}$) or is stored in the ES (${\mathrm{E}}_{\mathrm{R}\mathrm{E}\to \mathrm{E}\mathrm{S}}$):

$$\mathrm{R}\mathrm{E} \mathrm{c}\mathrm{u}\mathrm{r}\mathrm{t}\mathrm{a}\mathrm{i}\mathrm{l}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t}=\sum _{\mathrm{k}=1}^{8760}\left\{{\mathrm{E}}_{\mathrm{R}\mathrm{E}}\left(\mathrm{k}\right)-\left[{\mathrm{E}}_{\mathrm{R}\mathrm{E}\to \mathrm{l}\mathrm{o}\mathrm{a}\mathrm{d}}\left(\mathrm{k}\right)+{\mathrm{E}}_{\mathrm{R}\mathrm{E}\to \mathrm{E}\mathrm{S}}\left(\mathrm{k}\right)\right]\right\}$$

(3)

2.2. Preliminary Results

As an example of the power management strategy, power flows and the ES state of charge over a 2-week period are shown in Figure 2 for the scenario PV×4W×3 with an ES energy capacity of 50 GWh and an E/P ratio of 4 h. As can be noticed, there are some hours on some days in which the RE power generation (green line) exceeds the electric load (black line). In these cases, the excess power is used to charge the ES (red line), compatibly with limitations in power capacity and energy capacity. The storage charges during diurnal hours (red line indicates charging power) and discharges when PV generation decreases. On some days, the storage reaches its maximum capacity, and the additional generation (black dotted line) is curtailed. Charge and discharge power are limited by the maximum ES power capacity. If RE power cannot be stored in the ES because the maximum energy capacity has already been reached or the power exceeds ES power capacity, the excess RE power is curtailed (dotted line).

The presence of the ES allows a great reduction in RE curtailment but is not able to eliminate it in all the scenarios (Figure 3). In the scenarios with the highest RE penetration, RE curtailment is consistently reduced by an increase in ES capacity for all the energy capacities considered. However, the reduction with the first 100 GWh of ES is much greater, even more than twice the additional reduction obtained by adding a further 100 GWh. Instead, in the scenarios with limited RE penetration, such as PV×2-3 W×1-3, energy capacities up to 100 GWh are enough to drastically reduce or eliminate RE curtailment.

It can be noticed that, in the scenarios with the highest PV penetration (PV×5), doubling the ES energy capacity from 50 GWh to 100 GWh allows a great reduction in RE curtailment for all the installed wind capacities considered. Indeed, a similar further reduction, just slightly less than the one obtained with a 50 GWh ES, is obtained by adding a further 50 GWh (i.e., moving from 50 GWh ES to 100 GWh ES). Instead, when doubling again the ES energy capacity from 100 GWh to 200 GWh, the additional reduction in RE curtailment is slightly less than half of the one obtained by going from 0 GWh ES to 100 GWh ES. Nevertheless, RE curtailment is still significantly further reduced by moving from 100 GWh ES to 200 GWh ES in the highest RE penetration scenarios.

At the same time, the presence of the ES allows for an increase in the RE share in covering the annual electric load (Figure 4) by storing electricity overproduction and providing it back when RE production is low. In this way, it contributes to the decarbonization of the electric system. This contribution is more significant as RE penetration increases and depends on the ES capacity. The highest ES capacity considered (200 GWh) allows for a maximum RE share increase of about 10 percentage points compared to the case without ES in the scenario with the highest RE penetration (PV×5 W×5) with an E/P ratio of 4 h (i.e., 50 GW of ES power capacity).

If, on one hand, at increasing ES capacities, the electric system becomes greener, on the other, the amount of RE curtailment available for the electrolysis decreases. The hydrogen production potential from RE curtailment is analyzed in the following section.

3. Hydrogen Production Through Electrolysis

Firstly, the amount of green electrolytic hydrogen that could be potentially produced from the sole RE curtailment and the required electrolysis capacity were evaluated in all scenarios. This potential for green hydrogen production was also compared with the current annual national demand for hydrogen by expressing it as a fraction of the annual hydrogen demand.

The annual Italian consumption of hydrogen in 2019, equal to 480⋅10⁶ kg, was assumed as a reference. By assuming an average specific electricity consumption of 50 kWh/$\mathrm{k}{\mathrm{g}}_{{\mathrm{H}}_2}$ for the electrolyzers, this corresponds to an equivalent consumption of 24 TWh of electrical energy if the entire hydrogen demand were produced through electrolysis. As a consequence, a minimum overall capacity of 2.74 GW is necessary to produce the annual hydrogen demand if the electrolyzers operate at all times at nominal conditions. This electrolysis power capacity was taken as the reference capacity (${\mathrm{P}}_{\mathrm{e}\mathrm{l}\mathrm{z}}^{\mathrm{*}}$) in the following.

Afterwards, the entire annual hydrogen consumption was supposed to be produced via electrolysis by allowing the use of non-RE electricity. Hydrogen produced in this way is no longer green hydrogen, but hydrogen produced with associated CO₂ emissions (yellow hydrogen). The hourly remaining required electricity that was not available from RE curtailment (after the ES), if any, was supposed to be provided by NG-based power plants with associated specific emissions ($\mathrm{e}{\mathrm{m}}_{\mathrm{N}\mathrm{G}\mathrm{p}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{t}\mathrm{s}}$) of 368.7 $\mathrm{k}{\mathrm{g}}_{\mathrm{C}{\mathrm{O}}_2}$/MWh_el [21]. These emissions correspond to 18.435 $\mathrm{k}{\mathrm{g}}_{\mathrm{C}{\mathrm{O}}_2}$/$\mathrm{k}{\mathrm{g}}_{{\mathrm{H}}_2}$, given the assumed specific electrical consumption of the electrolyzers of 50 kWh/$\mathrm{k}{\mathrm{g}}_{{\mathrm{H}}_2}$.

In order to meet the entire hydrogen demand, the minimum installed electrolysis capacity required is the reference one (${\mathrm{P}}_{\mathrm{e}\mathrm{l}\mathrm{z}}^{\mathrm{*}}$= 2.74 GW). However, greater electrolysis capacities can be employed as well. Therefore, several electrolysis capacities were considered for this analysis: they were equal to or greater than the reference one. In the case of an electrolysis capacity equal to the reference one, the electrolyzers must operate at all times at nominal conditions and, consequently, specific CO₂ emissions per kilogram of hydrogen are the highest possible. By increasing the overall electrolysis capacity, the utilization of electrolyzers decreases while CO₂ emissions can be reduced, since a higher amount of RE curtailment during power generation peaks can be used by the electrolyzers.

The annual RE curtailment after hydrogen production via electrolysis was calculated as follows:

$$<{\mathrm{R}\mathrm{E} \mathrm{c}\mathrm{u}\mathrm{r}\mathrm{t}\mathrm{a}\mathrm{i}\mathrm{l}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t}}_{\mathrm{a}\mathrm{f}\mathrm{t}\mathrm{e}\mathrm{r}{ \mathrm{H}}_2 \mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}}>=\sum _{\mathrm{k}}^{8760}{\mathrm{E}}_{\mathrm{R}\mathrm{E}}\left(\mathrm{k}\right)-[{\mathrm{E}}_{\mathrm{R}\mathrm{E}\to \mathrm{l}\mathrm{o}\mathrm{a}\mathrm{d}}\left(\mathrm{k}\right)+{\mathrm{E}}_{\mathrm{R}\mathrm{E}\to \mathrm{E}\mathrm{S}}\left(\mathrm{k}\right)+{\mathrm{E}}_{\mathrm{R}\mathrm{E}\to \mathrm{e}\mathrm{l}\mathrm{z}}\left(\mathrm{k}\right)]$$

(4)

The results were compared in terms of the electrolysis utilization factor, the percentage of green hydrogen over the total hydrogen produced, and specific CO₂ emissions.

The electrolysis utilization factor ($\mathrm{U}{\mathrm{F}}_{\mathrm{e}\mathrm{l}\mathrm{z}}$) was defined as follows:

$${\mathrm{U}\mathrm{F}}_{\mathrm{e}\mathrm{l}\mathrm{z}}={\displaystyle \frac{{\sum }_{\mathrm{k}=1}^{8760}{\mathrm{E}}_{\mathrm{R}\mathrm{E}\to \mathrm{e}\mathrm{l}\mathrm{z}}\left(\mathrm{k}\right)}{{8760\mathrm{h}\cdot \mathrm{P}}_{\mathrm{e}\mathrm{l}\mathrm{z},\mathrm{n}\mathrm{o}\mathrm{m}}}}$$

(5)

where ${\mathrm{P}}_{\mathrm{e}\mathrm{l}\mathrm{z},\mathrm{n}\mathrm{o}\mathrm{m}}$ is the overall electrolysis power capacity.

The percentage of green hydrogen over the total produced hydrogen (green %) was calculated as follows:

$$\mathrm{g}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\mathrm{\%}=100\cdot {\displaystyle \frac{<\mathrm{a}\mathrm{n}\mathrm{n}\mathrm{u}\mathrm{a}\mathrm{l}{ \mathrm{H}}_2 \mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{e}\mathrm{d} \mathrm{f}\mathrm{r}\mathrm{o}\mathrm{m} \mathrm{R}\mathrm{E} \mathrm{c}\mathrm{u}\mathrm{r}\mathrm{t}\mathrm{a}\mathrm{i}\mathrm{l}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t}>}{<\mathrm{t}\mathrm{o}\mathrm{t}\mathrm{a}\mathrm{l} \mathrm{a}\mathrm{n}\mathrm{n}\mathrm{u}\mathrm{a}\mathrm{l} {\mathrm{H}}_2 \mathrm{p}\mathrm{r}\mathrm{o}\mathrm{d}\mathrm{u}\mathrm{c}\mathrm{e}\mathrm{d}>}}$$

(6)

Specific CO₂ emissions associated with hydrogen production were calculated as follows:

$$\mathrm{e}{\mathrm{m}}_{\mathrm{y}\mathrm{e}\mathrm{l}\mathrm{l}\mathrm{o}\mathrm{w} {\mathrm{H}}_2}=\left(1-{\displaystyle \frac{\mathrm{g}\mathrm{r}\mathrm{e}\mathrm{e}\mathrm{n}\mathrm{\%}}{100\mathrm{\%}}}\right)·{\mathrm{e}\mathrm{m}}_{\mathrm{N}\mathrm{G} \mathrm{p}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{t}\mathrm{s}}$$

(7)

Economic Evaluation

For the economic evaluation, the levelized cost of electrolytic hydrogen (LCOH), both green and yellow hydrogen, was calculated as follows:

$$\mathrm{L}\mathrm{C}\mathrm{O}\mathrm{H}=\left[\left(\mathrm{C}\mathrm{R}\mathrm{F}+{\mathrm{f}}_{\mathrm{O}\&\mathrm{M}}\right)·\mathrm{C}\mathrm{A}\mathrm{P}\mathrm{E}{\mathrm{X}}_{\mathrm{e}\mathrm{l}\mathrm{z}}·{\mathrm{P}}_{\mathrm{e}\mathrm{l}\mathrm{z},\mathrm{n}\mathrm{o}\mathrm{m}}+{\mathrm{E}}_{\mathrm{a}\mathrm{d}\mathrm{d},\mathrm{N}\mathrm{G} \mathrm{p}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{t}\mathrm{s}}·{\mathrm{c}}_{\mathrm{e}\mathrm{l},\mathrm{N}\mathrm{G} \mathrm{p}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{t}\mathrm{s}}+{\mathrm{c}}_{{\mathrm{H}}_2\mathrm{O}}·{\mathrm{M}}_{{\mathrm{H}}_2}\right]/{\mathrm{M}}_{{\mathrm{H}}_2}$$

(8)

where the capital recovery factor, CRF, is calculated as follows:

$$\mathrm{C}\mathrm{R}\mathrm{F}={\displaystyle \frac{\mathrm{I}\mathrm{R}}{1-{\left(1+\mathrm{I}\mathrm{R}\right)}^{-\mathrm{L}\mathrm{t}}}}$$

(9)

where ${\mathrm{f}}_{\mathrm{O}\&\mathrm{M}}$ is the operation and maintenance factor, ${\mathrm{c}}_{\mathrm{e}\mathrm{l},\mathrm{N}\mathrm{G} \mathrm{p}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{t}\mathrm{s}}$ is the cost of grid electricity from NG-based power plants, ${\mathrm{E}}_{\mathrm{a}\mathrm{d}\mathrm{d},\mathrm{N}\mathrm{G} \mathrm{p}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{t}\mathrm{s}}$ is the additional electricity provided by NG-based power plants, ${\mathrm{M}}_{{\mathrm{H}}_2}$ is the annual hydrogen production in units of mass, ${\mathrm{c}}_{{\mathrm{H}}_2\mathrm{O}}$ is the cost of water per unit of hydrogen mass, IR is the interest rate, and $\mathrm{L}\mathrm{t}$ is the electrolyzers’ lifetime.

The main assumptions on the economic parameters adopted for evaluating the LCOH are provided in

Table 1. Main economic parameters.

Parameter	Value
Capex, $\mathrm{C}\mathrm{A}\mathrm{P}\mathrm{E}{\mathrm{X}}_{\mathrm{e}\mathrm{l}\mathrm{z}}$	EUR 750/kW·1.15·${\mathrm{P}}_{\mathrm{e}\mathrm{l}\mathrm{z},\mathrm{n}\mathrm{o}\mathrm{m}}$
Operation and maintenance factor, ${\mathrm{f}}_{\mathrm{O}\&\mathrm{M}}$	5%
Cost of grid electricity (from NG-based power plants), ${\mathrm{c}}_{\mathrm{e}\mathrm{l},\mathrm{N}\mathrm{G} \mathrm{p}\mathrm{l}\mathrm{a}\mathrm{n}\mathrm{t}\mathrm{s}}$	EUR 50/MWh
Specific cost of water, ${\mathrm{c}}_{{\mathrm{H}}_2\mathrm{O}}$*	EUR 0.057/$\mathrm{k}{\mathrm{g}}_{{\mathrm{H}}_2}$ [23]
Interest rate, IR	8% [24]
Lifetime, Lt	20 years [25]

* Assuming a specific water consumption of 15 dm³/$\mathrm{k}{\mathrm{g}}_{{\mathrm{H}}_2}$ and a water cost of EUR 3.8/m³.

. The curtailed electricity used for the electrolysis was assumed to be free of cost. A specific capital cost of EUR 750/kW [22], increased by 15% to take into account additional costs for the balance of plant, was adopted for the electrolyzers.

In this study, a sensitivity analysis of LCOH with respect to the specific capital cost of electrolyzers and the cost of grid electricity was not performed. However, a sensitivity analysis on that was conducted in a previous study [16], analyzing the same scenarios without any electric storage, considering specific capital costs of electrolyzers ranging from EUR 250 to 750/kW and costs of grid electricity ranging from EUR 25 to 75/MWh.

4. Results

4.1. Green Hydrogen Production Potential from Electrolysis

In the upper graphs in Figure 5a–d, hourly values of RE curtailment power are shown in ascending order for some selected scenarios. The abscissa of the intersection point with the x-axis represents the annual number of hours during which RE curtailment is zero. The maximum y-value (obtained at x = 8760 h) is the maximum RE curtailment power of the year. The black dash-dotted line indicates the overall electrolysis power required to produce the annual national hydrogen demand if all the electrolyzers are operating constantly at this power, i.e., the reference power (${\mathrm{P}}_{\mathrm{e}\mathrm{l}\mathrm{z}}^{\mathrm{*}}$).

In the lower graphs of Figure 5a–d, the y-coordinate of each point on the curves indicates the annual amount of RE curtailment that could be used for water electrolysis if the overall nominal electrolysis power is equal to the RE curtailment power reported in the corresponding upper graph at the same x-coordinate. The annual amount of RE curtailment that could be used for water electrolysis for a given overall nominal electrolysis power (${\tilde{\mathrm{P}}}_{\mathrm{e}\mathrm{l}\mathrm{z}}$) is equal to the following:

$$\begin{array}{c}<\mathrm{R}\mathrm{E} \mathrm{c}\mathrm{u}\mathrm{r}\mathrm{t}\mathrm{a}\mathrm{i}\mathrm{l}\mathrm{m}\mathrm{e}\mathrm{n}\mathrm{t} \mathrm{u}\mathrm{s}\mathrm{e}\mathrm{d} \mathrm{f}\mathrm{o}\mathrm{r} \mathrm{e}\mathrm{l}\mathrm{e}\mathrm{c}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}\mathrm{y}\mathrm{s}\mathrm{y}\mathrm{s} \mathrm{i}\mathrm{f} {\mathrm{P}}_{\mathrm{e}\mathrm{l}\mathrm{z},\mathrm{n}\mathrm{o}\mathrm{m}}={\tilde{\mathrm{P}}}_{\mathrm{e}\mathrm{l}\mathrm{z}}>=\\ =1\mathrm{h}·{\displaystyle \sum _{\mathrm{i}=1}^{8760}}\left\{\begin{array}{c} {\mathrm{P}}_{\mathrm{R}\mathrm{E} \mathrm{c}\mathrm{u}\mathrm{r}\mathrm{t}}\left(\mathrm{i}\right)\mathrm{if}{\mathrm{P}}_{\mathrm{R}\mathrm{E} \mathrm{c}\mathrm{u}\mathrm{r}\mathrm{t}}\left(\mathrm{i}\right)\le {\tilde{\mathrm{P}}}_{\mathrm{e}\mathrm{l}\mathrm{z}}\hfill \\ {\mathrm{P}}_{\mathrm{e}\mathrm{l}\mathrm{z}}^{\mathrm{*}}\mathrm{if}{\mathrm{P}}_{\mathrm{R}\mathrm{E} \mathrm{c}\mathrm{u}\mathrm{r}\mathrm{t}}\left(\mathrm{i}\right)>{\tilde{\mathrm{P}}}_{\mathrm{e}\mathrm{l}\mathrm{z}}\hfill \end{array}\right.\end{array}$$

(10)

The overall nominal electrolysis power required to produce a certain amount of green hydrogen (hydrogen obtained only from RE curtailment) can be read from Figure 5 for a given scenario and a given ES capacity. The maximum theoretical potential of green hydrogen production is slightly less than two times the annual hydrogen demand in the scenario PV×5W×5 without ES (as it can be observed from the ordinate of the intersection point between the purple line and the right y-axis in the bottom graph of Figure 5a). This value is reduced to slightly less than 1.5 times the annual hydrogen demand in the scenario PV×5W×3. In the scenarios with a lower RE penetration, such as PV×4W×3, it is not possible to produce the annual hydrogen demand by means of electrolysis from RE curtailment. In the scenario PV×3W×2, the annual green hydrogen production potential goes below one-quarter of the annual hydrogen demand. This means that in order to produce all the annual hydrogen demand, it becomes necessary to use further NG-based electricity.

With the presence of additional ES capacity, RE curtailment is reduced. As a consequence, the green hydrogen production potential is reduced as well, as can be seen from Figure 5b–d for increasing ES capacities.

The cases shown in Figure 5b–d refer to an E/P ratio of 4 h. In the cases with lower power capacities (i.e., with higher E/P ratios), the RE curtailment and, consequently, the green hydrogen potential can be slightly higher.

In Figure 6, the (minimum) overall electrolysis capacities required to produce an amount of green hydrogen equal to 25%, 50%, and 100% of the annual hydrogen demand, respectively, are shown (in the cases in which this is possible) for the selected scenarios under different ES energy capacities (${\mathrm{C}}_{\mathrm{E}\mathrm{S}}$) and for two different values of the E/P ratio: 8 h (Figure 6a) and 4 h (Figure 6b).

At the same ES energy capacity, when the ES power capacity is higher (Figure 6b), the green hydrogen production potential is lower because of the lower RE curtailment. For example, with ${\mathrm{C}}_{\mathrm{E}\mathrm{S}}$ = 100 GWh, the whole annual national demand can be produced from RE curtailment in the scenario PV×5W×5 if E/P = 8 h (purple star mark circled in green in Figure 6a), i.e., if the ES power capacity is 12.5 GW, a minimum overall electrolysis capacity of about 32 GW is required. Conversely, it cannot be produced solely from the curtailment if E/P = 4 h, i.e., if the ES power capacity is 25 GW (indeed, there is no purple star mark in Figure 6b at ${\mathrm{C}}_{\mathrm{E}\mathrm{S}}$ = 100 GWh). Analogously, in the scenario PV×5W×3, 50% of the annual national demand can be produced from RE curtailment with ${\mathrm{C}}_{\mathrm{E}\mathrm{S}}$ = 100 GWh if E/P = 8 h, i.e., the ES power capacity is 12.5 GW (yellow circle mark circled in green in Figure 6a), but not if E/P = 4 h, i.e., the power capacity is 25 GW (there is no yellow circle mark in Figure 6b at ${\mathrm{C}}_{\mathrm{E}\mathrm{S}}$ = 100 GWh).

Furthermore, under the same scenario and ES energy capacity, a lower minimum electrolysis capacity is required to produce the same amount of green hydrogen when the ES power capacity is lower (i.e., when E/P is greater).

It can be noticed that the ES power capacity affects the minimum electrolysis power required to produce a certain amount of green hydrogen in a different way depending on the scenario. This depends, in particular, on the PV share. For example, at an ES energy capacity of 50 GWh and an E/P ratio of 8 h, the electrolysis capacity required to produce 25% of the annual hydrogen demand in the scenario PV×4W×3 is lower than that required to produce 50% of the annual hydrogen demand in the scenario PV×5W×3 (marks circled in blue in Figure 6a). Instead, at a greater power capacity (E/P = 4 h), the electrolysis capacity required to produce 25% of the annual hydrogen demand in the scenario PV×4W×3 is greater than that required to produce 50% of the annual hydrogen demand in the scenario PV×5W×3 (marks circled in blue in Figure 6b). Indeed, the increase in the minimum electrolysis capacity required to produce the same amount of green hydrogen with the increase in ES power capacity is more marked in the scenarios with higher PV penetration. The greater ES power capacity allows for reducing curtailment during PV power peaks.

4.2. Production of Current Italian Hydrogen Demand Through Electrolysis

The overall electrolysis capacities obtained and discussed above refer to the cases in which the production of a certain amount of green hydrogen is imposed and are generally much greater than the reference electrolysis capacity (${\mathrm{P}}_{\mathrm{e}\mathrm{l}\mathrm{z}}^{\mathrm{*}}$). Indeed, if the annual hydrogen demand has to be produced with those electrolysis capacities, the electrolyzers operate when there is RE curtailment and, therefore, for a limited number of hours per year, that is, with low utilization factors (UFs). By choosing lower electrolysis capacities (but higher than the reference one), it is possible to produce the same amount of hydrogen by using further NG-based electricity. This option allows for increasing the electrolyzers’ UF and reducing capital costs and, as a cascade effect, reducing the specific cost of hydrogen production. On the other side, hydrogen production is no longer 100% green, but there are associated CO₂ emissions. It is worth investigating intermediate solutions in which a lower LCOH is obtained with a slight increase in CO₂ emissions. The upper limit of CO₂ emissions that should not be exceeded is given by the specific CO₂ emissions related to the most widespread hydrogen production process of steam methane reforming (9 $\mathrm{k}{\mathrm{g}}_{\mathrm{C}{\mathrm{O}}_2}$/$\mathrm{k}{\mathrm{g}}_{{\mathrm{H}}_2}$ [2]). As a consequence of the assumption of 18.435 $\mathrm{k}{\mathrm{g}}_{\mathrm{C}{\mathrm{O}}_2}$/$\mathrm{k}{\mathrm{g}}_{{\mathrm{H}}_2}$ as specific emissions associated with electrolytic hydrogen production from NG-based electricity, a percentage of green hydrogen over the total electrolytic hydrogen production of 51.18% leads to the same specific emissions as SMR. At lower percentages of green hydrogen, specific CO₂ emissions associated with the electrolytic hydrogen production are, therefore, higher than those of SMR.

In Figure 7, the percentage of green hydrogen over the total amount of produced hydrogen is shown for all the considered scenarios for a fixed overall electrolysis capacity of 5.48 GW, which is twice the reference capacity. The annual amount of hydrogen produced is set equal to the annual hydrogen consumption (480 × 10⁶ kg). As a consequence, the electrolysis UF is equal to 50% in all cases.

In all cases shown in Figure 7, in order to produce the annual hydrogen demand with a UF of 50%, the percentage of green hydrogen over the total produced hydrogen is low. Only in the scenario PV×5W×5 without ES is the percentage of green hydrogen slightly higher than 55%, although it decreases to about 13% with ${\mathrm{C}}_{\mathrm{E}\mathrm{S}}$ = 200 GWh and an E/P ratio of 4 h. Therefore, to reach greater percentages of green hydrogen, it is necessary either to have greater electrolysis capacities and, at the same time, reduce the UFs or forgo producing the whole national hydrogen demand through electrolysis.

Increasing the overall electrolysis capacity and, at the same time, reducing the overall electrolysis UF in order to increase the percentage of green hydrogen (hydrogen produced from curtailment) has an impact on the hydrogen production cost. Figure 8 shows the percentage of green hydrogen relative to the total electrolytic hydrogen production, with the latter set equal to the national hydrogen demand vs. the levelized cost of the hydrogen (LCOH) produced in the selected scenarios at several ES energy capacities and for an E/P ratio of 4 h. The overall electrolysis nominal power (${\mathrm{P}}_{\mathrm{e}\mathrm{l}\mathrm{z},\mathrm{n}\mathrm{o}\mathrm{m}}$) is increased from the lowest value of the reference electrolysis power (2.74 GW) in steps equal to the reference electrolysis power itself $({\mathrm{P}}_{\mathrm{e}\mathrm{l}\mathrm{z},\mathrm{n}\mathrm{o}\mathrm{m}}=\mathrm{n}\cdot {\mathrm{P}}_{\mathrm{e}\mathrm{l}\mathrm{z}}^{\mathrm{*}}$, with n ∈ N, n ≥ 0). Without ES, in the scenarios PV×5W×5 and PV×5W×3, it is possible to produce 100% green hydrogen at electrolysis power capacities of 11 GW and 15 GW, respectively. At an ES capacity of 50 GWh, it is possible to produce 100% green hydrogen only in the scenario PV×5W×5. To achieve this, the overall electrolysis capacity must be at least 17.8 GW (i.e., 6.5 times the reference electrolysis power) in the case of an E/P ratio of 4 h (i.e., an ES power capacity of 12.5 GW).

At the same green hydrogen percentage, the higher the RE penetration, the lower the LCOH. Furthermore, at the same LCOH, scenarios with the highest RE penetration allow for much higher percentages of green hydrogen to be reached. Scenario PV×3W×2 does not allow for both producing the annual hydrogen demand through electrolysis and obtaining associated specific CO₂ emissions lower than those of SMR in any case. Scenario PV×4W×3 allows for the production of the annual hydrogen demand through electrolysis with associated specific CO₂ emissions lower than those of SMR only in the case without ES. In the scenario PV×5W×5, it is possible to produce the annual hydrogen demand through electrolysis with associated specific CO₂ emissions slightly lower than those of SMR even in the case with the highest energy capacity (200 GWh) and the highest power capacity considered, at the price of a higher LCOH of at least EUR 8.5/kg and a minimum electrolysis capacity of at least 12 times the reference one.

5. Conclusions

This study highlights some major limitations in using electrolysis as the main process to meet current hydrogen demand. To be sustainable, electrolytic hydrogen must be produced from renewable electricity. Not only that, but the renewable electricity used should not be subtracted from that which could be used to cover the electric load; instead, it should come from otherwise curtailed electricity. This is an aspect that is rarely considered when addressing green hydrogen production, but it should be taken into account in future research to avoid partial and misleading perspectives.

In the investigated scenarios with increasing PV and wind penetration, the amount of RE curtailment is low in comparison to the current hydrogen demand. Only in the scenarios with the highest RE penetration would it be theoretically possible to satisfy the total annual hydrogen demand via electrolysis only based on RE curtailment, and only in cases of low electric storage (ES) capacities. The presence of ES increases the RE share in the electric grid but further reduces the electrical energy available for electrolysis. In addition, RE curtailment comes from power generation peaks, which are concentrated within some hours. Therefore, large electrolysis capacities are required in order to use all of it. As a consequence, the resulting electrolysis utilization factor (UF) is very low and leads to high specific hydrogen costs. By allowing the use of further grid electricity from NG-based plants, the electrolysis UF can be increased, but there are consequent, associated specific CO₂ emissions. In particular, in the theoretical scenario of producing the whole annual hydrogen demand through electrolysis, almost all the scenarios considered resulted in specific CO₂ emissions higher than those of SMR.

Priority of RE utilization should be given to the decarbonization of the electric system. Therefore, a substantial increase in RE capacities and the consequent large availability of excess power are fundamental to making electrolysis viable on a large scale and more feasible from an economic point of view.