Combined Hydrogen and Power (CHyP) from Hybrid Wind–PV Plants: A Techno-Economic Pathway Toward Cost-Competitive Green Hydrogen in Spain

Abstract

Green hydrogen produced via water electrolysis using renewable energy is central to the energy transition, but high production costs remain a major barrier. Solar photovoltaic (PV) and wind farms are dominant renewable sources, yet each has drawbacks: PV is inexpensive but limited in operating hours, while wind is costlier but offers steadier output. This paper introduces the Combined Hydrogen and Power (CHyP) concept, a co-production configuration in which hydrogen and electricity are treated as primary products for an industrial off-taker within a hydrogen hub. The electrolyzer is sized from the generation duration curve to optimize annual hydrogen output and operating hours, operated at nominal load to limit failures and maintenance. The resulting surplus electricity is assumed to be sold to nearby consumers to offset hydrogen production costs. Simulations of 500 MW hybrid wind–photovoltaic plants at selected locations in Spain identify an average optimal wind share of 27%. Assuming a surplus electricity price of 60 €/MWh, minimum LCOH values range from 1.68 to 4.40 €/kg. Sensitivity analysis shows that average LCOH decreases from 5.30 to 0.73 €/kg as the selling price increases from 40 to 80 €/MWh, highlighting the importance of electricity sales conditions (industrial demand, temporal marching and distribution capacity in the hydrogen hub) for the economic viability of the proposed concept.

1. Introduction

1.1. Green Hydrogen and Techno-Economic Challenges

Achieving climate neutrality by 2050 requires deep decarbonization of emission-intensive sectors where direct electrification is technically or economically challenging. Global hydrogen demand reached almost 100 Mt in 2024, with production still dominated by unabated fossil fuels and low-carbon hydrogen accounting for less than 1% of the total [1,2]. Green hydrogen produced via water electrolysis powered by renewable energy sources (RES) is therefore increasingly viewed as a key energy carrier for decarbonizing hard-to-abate sectors such as heavy transport, steelmaking, aviation, and chemicals owing to its versatility and relatively high specific energy [3,4,5,6].

The European Union has positioned itself as a frontrunner in this transition. The 2020 Hydrogen Strategy targets 40 GW of renewable electrolysis capacity by 2030 and promotes “hydrogen valleys”, integrated regional ecosystems for hydrogen production, storage, transport, and use, as flagship instruments to develop hydrogen hubs around industrial clusters [4,7]. In this paper, the term “hydrogen hub” is used in a broad sense to denote spatial clusters where hydrogen production, infrastructure, and demand (often co-located with power and heat networks) are integrated; these may include projects formally labeled as “hydrogen valleys” as well as other industrial clusters with substantial hydrogen and electricity demand.

Despite this momentum, green hydrogen remains considerably more expensive than fossil-based alternatives under current market conditions. The International Energy Agency (IEA) [2] reports that the cost gap between low-carbon hydrogen and unabated fossil-based production remains a key barrier, currently exacerbated by lower natural gas prices and increased electrolyzer costs outside China. However, this gap is expected to narrow by 2030; in the IEA’s Stated Policies Scenario (STEPS), renewable hydrogen production costs could fall to under 2 USD/kg in China and to a range of 3–4 USD/kg in other regions with abundant renewable resources and low cost of capital [2]. Detailed case studies of optimally sited hybrid photovoltaic (PV)–wind farm (WF)–electrolyzer plants in California and northern Germany, Brazil, Egypt, and Australia report levelized costs of hydrogen (LCOH) between roughly 2 and 6 €/kg depending on resource quality, financing conditions, and plant configuration [8,9,10,11,12,13]. These values vary widely with site-specific and market parameters and should therefore be interpreted as location- and scenario-specific rather than as globally representative cost benchmarks.

Techno-economic assessments consistently highlight electricity price, electrolyzer investment cost, and annual operating hours as dominant drivers of LCOH, together with site-specific factors such as renewable resource quality, grid and infrastructure availability, and proximity to large off-takers [1,6,14,15]. According to [16], the weighted-average levelized cost of electricity (LCOE) from newly commissioned utility-scale solar PV and onshore wind projects in 2023 lies in the range of a few tens of USD/MWh, with best-in-class projects in resource-rich regions achieving values below about 30 USD/MWh. Such low renewable generation costs are a necessary but not sufficient condition for competitive green hydrogen: even with cheap electricity, high electrolyzer capital expenditures (CAPEX), limited utilization, and additional balance-of-plant costs can keep LCOH well above fossil benchmarks [2,14,17,18]. For the Iberian region, the Iberian Renewable Hydrogen Index (IBHYX) [19] provides a benchmark for the estimated LCOH of a representative electrolysis facility under current technological and economic conditions, rather than an actual hydrogen spot market price. Overall, these considerations motivate plant concepts that simultaneously increase electrolyzer utilization and unlock additional value streams in order to close the remaining cost gap.

1.2. Electrolysis Technologies and Operational Constraints

Alkaline water electrolysis (AEL) and proton exchange membrane (PEM) electrolysis have reached a technology readiness level (TRL) of 9 [20,21]. AEL has lower specific investment costs but limited part-load flexibility, whereas PEM electrolyzers respond rapidly to variable power inputs at higher CAPEX, which makes them well suited for direct coupling with PV plants and wind farms [20,21]. Solid-oxide electrolysis (SOEC) can achieve high efficiency when waste heat is available but is less mature [21].

Beyond technology choice, operating strategies have a material impact on performance and lifecycle cost. Recent degradation studies on PEM water electrolysis show that frequent load cycling and operation at very high current densities can noticeably increase voltage decay rates and shorten stack lifetimes compared with smoother operating profiles [22,23]. Complementary experimental work on AEL reveals that strongly variable operation accelerates catalyst degradation, whereas steady-state conditions preserve catalyst integrity over longer periods [24]. Optimization-based analyses that couple degradation models with replacement strategies further demonstrate that aggressive flexible operation can increase operating expenditures through more frequent stack refurbishments [25]. Recent reviews of PEM electrolyzer degradation mechanisms and modeling frameworks likewise stress that operating profiles, start-stop cycles, and power allocation strategies should be explicitly represented when assessing lifetime costs and optimal plant operation [23,24,25,26]. Operation-oriented studies address these issues through optimal scheduling and energy management, for example, by accounting for electrolyzer start-up characteristics [27,28], by scheduling complementary wind–PV plants on the basis of generation forecasts [29], or by coordinating hydrogen production with power and hydrogen networks [30].

Taken together, this evidence suggests that operational strategies ensuring high but relatively stable loading are desirable from a techno-economic standpoint, particularly for large plants directly coupled to variable RES. This consideration motivates the Combined Hydrogen and Power (CHyP) framework investigated in this study, in which the electrolyzer is intentionally operated at (or near) its nominal load during production periods, while renewable generation exceeding its electricity demand is allocated to industrial consumers.

1.3. Hybrid Renewable Supply and Renewable-to-Hydrogen System Design

Dedicated RES-to-hydrogen plants face a well-known design trade-off: electrolyzers are capital-intensive assets that require high annual operating hours, whereas stand-alone PV plants typically provide about 1500 equivalent full-load hours per year [16,31], well below the several thousand hours sought to amortize electrolyzer CAPEX [32,33,34]. Oversizing renewable capacity relative to the electrolyzer increases its operating hours but generates substantial electricity surpluses during peak production [2,32]; if these surpluses are curtailed or remunerated at low tariffs, the effective cost of the electricity used for hydrogen production rises, partly offsetting the expected cost reduction [12].

Hybrid PV–wind configurations are increasingly recognized as a baseline strategy to mitigate these issues [8,35]: owing to the partial temporal complementarity of solar and wind resources, they can smooth the net power profile, enhance electrolyzer utilization, and reduce the need for short-term battery storage [8,9,10,11,12,27,29,36,37,38,39,40]. For hybrid PV–WF–electrolyzer systems, techno-economic assessments identify the ratio between renewable capacity and electrolyzer power as an additional key driver of LCOH, alongside the renewable LCOE and the electrolyzer full-load hours already mentioned in Section 1.1 [2,8,9,10,11,12,36,37,40,41] while oversized battery storage tends not to be cost-effective unless hydrogen prices are very high or solar resources are extremely intermittent [38,39].

1.4. Co-Production, Polygeneration and Multi-Product Configurations

Beyond pure RES-to-hydrogen schemes, various co-production and polygeneration architectures seek to improve project economics by valorizing multiple energy outputs. In this work, the term “co-production” is used when electricity and hydrogen are the two main products of a system, and “polygeneration” when additional products such as heat, water, or hydrogen-derived chemicals are also supplied. Some studies use the term cogeneration for electricity and hydrogen production [42,43,44,45]; however, this term is avoided here because it conventionally denotes the combined production of power and useful heat, whereas no heat is recovered in the system studied in this work.

At the plant scale, techno-economic analyses of wind–PV–electrolysis–battery hybrid systems have examined the joint generation of power and hydrogen [40]. Scenario-based assessments of wind-based co-production report levelized costs, payback periods, and rates of return for both electricity and hydrogen under alternative degradation and discount-rate assumptions [46]. At the distribution and micro-grid level, hybrid systems combining PV, WF, diesel backup, batteries, and electrolysis co-produce electricity and green hydrogen, usually with hydrogen acting as a storage medium or secondary product that supports off-grid or weak-grid supply [27,38,39,42].

Heat and by-product utilization further widen the design space. Combined hydrogen and heat schemes recover low-grade electrolyzer heat for district heating, raising overall energy efficiency where a suitable heat demand exists [47]. Polygeneration systems in wastewater-treatment and steel plants co-produce hydrogen together with ammonia, electricity, or process heat [48,49], and solar-thermal and biomass-based concepts co-produce electricity, water, and hydrogen under favorable resource and cost assumptions [43,44,50,51,52]. Power-to-X routes that convert hydrogen from hybrid PV–wind plants into ammonia or methanol have also been assessed [13], and valorizing the electrolysis oxygen by-product can reduce LCOH by about 10–20% where high-purity oxygen can be marketed [53].

1.5. Hydrogen Hubs, Valleys and Regional Integration

At the regional scale, hydrogen hubs and valleys introduced in Section 1.1 provide the framework for integrating production, infrastructure, and demand [4,7]. Spatial-optimization methods such as H2Locate [54] and hydrogen-valley modeling approaches [55,56] show that resource availability, existing grids and pipelines, and proximity to large industrial consumers jointly determine the most suitable locations for large-scale electrolysis, and that purely resource-based siting can lead to sub-optimal outcomes even in areas with excellent renewable potential.

Value-chain analyses further show that hydrogen demand is concentrated in industrial clusters where integration with power networks is essential [15]. On the operational side, recent frameworks link electrolyzer dispatch to electricity-market conditions. Offshore wind-to-hydrogen systems participating in multi-level electricity markets can improve profitability and reduce curtailment [57], and market-optimization frameworks have been developed for solid-oxide systems that co-produce electricity and hydrogen [58].

Mixed-integer linear programming has also been used to compute the LCOH from the off-taker’s standpoint, showing for a small, demand-driven case that idle electrolyzer capacity raises costs and that additional production with surplus or low-priced electricity can lower them [59].

1.6. Research Gap and CHyP Concept

Taken together, the literature can be broadly classified into three overlapping categories. The first group comprises surplus-to-hydrogen schemes that produce hydrogen from curtailed or very low-priced electricity [9,10,11,12,13,14,35,36,37,39,60]. A second group encompasses multi-product energy systems, including studies described in the literature as cogeneration or polygeneration in which electricity, hydrogen, heat, and sometimes water or chemicals are co-produced. In many of these studies, market interactions and revenues from multiple products are represented using simplified assumptions, such as fixed selling prices or predefined economic scenarios [13,43,44,45,47,48,49,50,51,52,53,61]. A third group focuses on the decentralized hybrid systems and micro-grids described in Section 1.4, in which hydrogen is mainly an energy-storage medium or secondary by-product [38,39,42]. Although these categories are not mutually exclusive, they highlight differences in the technical objectives and valuation approaches of existing studies.

More recent works explicitly address the interaction between hydrogen and electricity products and markets, for instance, by analyzing the economics of converting renewable power to hydrogen under fluctuating electricity prices and efficient sizing considerations [32], or through off-taker-oriented and market-based frameworks discussed in Section 1.5 [57,58,59]. These frameworks clearly point to the importance of product flexibility and market-based operation, but they address different designs or evaluation settings. Earlier Spanish grid-connected PV–wind studies had already examined hydrogen production alongside surplus-electricity sales [62,63]. A subsequent study addresses plant sizing and control at an individual site [28], whereas broader assessments include Spain among the countries analyzed [64,65].

What remains less directly addressed by these precedents is a structured framework in which a hybrid PV–WF plant is assessed across selected Spanish locations to identify the preferred wind–PV mix and resulting LCOH under a common electrolyzer-sizing procedure and an integrated economic framework that accounts for revenues from surplus electricity sales. In particular, three aspects are considered together. First, the electrolyzer is sized based on the generation duration curve of a hybrid RES plant, selecting an operating point that substantially increases electrolyzer operating hours while limiting the reduction in annual hydrogen production, which remains between 85% and 95% of its maximum value. Second, surplus electricity is valued through an assumed PPA-like tariff for industrial offtake within hubs, with the corresponding revenues explicitly deducted in the calculation of the revenue-adjusted LCOH. Third, a spatially explicit assessment of hybrid PV–WF–electrolyzer configurations is carried out at one selected site in each of Spain’s 17 autonomous communities, comparing regional renewable resources and their impact on LCOH under common technical and economic assumptions.

CHyP is defined as a hybrid PV–wind power plant whose output is allocated between hydrogen production and direct electricity supply to an industrial off-taker, with both hydrogen and electricity treated as main products.

The contribution of this work does not lie in proposing a fundamentally new system architecture or introducing surplus electricity sales into hydrogen production for the first time, as both approaches have precedents [28,62,63]. Instead, it lies in consistently combining electrolyzer sizing, renewable generation hybridization, and revenue-adjusted hydrogen cost assessment across selected Spanish locations. Although degradation and electricity markets are represented through average rates and a single PPA-like tariff rather than fully dynamic models, the resulting framework links plant sizing, product allocation, and regional resource variability within a common assessment.

The remainder of this paper is organized as follows. Section 2 describes the CHyP plant configuration, the electrolyzer sizing procedure based on the generation duration curve, and the LCOH model and economic assumptions. Section 3 presents the LCOH as a function of the wind share for a selected site in each Spanish autonomous community, together with the resulting optimal configurations. Section 4 discusses the LCOH breakdown, its relationship with LCOE and electrolyzer operating hours, and the operating profiles and sensitivity analyses for two selected locations. Section 5 summarizes the main conclusions.

2. Methods

The production unit analyzed in this study consists of a PEM electrolyzer directly powered by a hybrid renewable energy plant, combining PV and WF sources. The electrolyzer always operates at its nominal design point during hydrogen production periods to limit degradation associated with variable-load operation [22]. To determine the annual electrolyzer operating hours, the generation duration curve of the hybrid plant is obtained, by sorting hourly net electricity production from the highest to the lowest value. The product of the coordinates at each point on this curve (power multiplied by operating hours) determines the amount of electricity generated at the corresponding power level. Assuming an average specific electricity consumption ($\overline{g}$), the corresponding annual hydrogen production curve can be obtained. This curve generally reaches its maximum at approximately 2000 operating hours per year. However, such a low utilization level results in high hydrogen production costs. For this reason, the electrolyzer is sized by increasing its operating hours beyond the maximum-production point, seeking a compromise between longer operating hours and the increasingly steep decline in annual hydrogen production. To establish a consistent and reproducible selection criterion, a reference level corresponding to 90% of the maximum annual electricity available to the electrolyzer, calculated from the generation duration curve, is adopted. The algorithm selects the operating point with the highest number of operating hours among those whose annual electricity availability corresponds to the established 90%. The same criterion is applied to all 51 wind–PV configurations at each of the 17 selected locations. For the configurations analyzed, the resulting hydrogen production levels are approximately 85–95% of the maximum annual production, depending on the generation profile and the average specific electricity consumption of the electrolyzer.

The annual electricity supplied to the electrolyzer ($E_{ELY}$) represents therefore only a fraction of the total electricity production of the hybrid plant, which includes contributions from both PV (E_PV) and WF (E_WF) sources. The remaining electricity constitutes a surplus that may be difficult to export to the grid. This provides the basis for the co-production concept: industrial hydrogen consumers may also require electricity, and part of this demand could be supplied using the surplus electricity generated by the hybrid plant. In order to facilitate this integrated energy supply, the facility should ideally be located within a hydrogen hub, as promoted by the European Union [7]. An internal microgrid could be implemented within the hub to distribute surplus electricity to several industrial consumers. Figure 1 illustrates the CHyP concept, described in detail in the introduction section.

Figure 1. CHyP concept. Renewable electricity generation supplies the electrolyzer, while surplus electricity is delivered to a nearby hydrogen hub, whose remaining electricity demand is met by the grid.

Once the annual electricity supplied to the electrolyzer is determined, the surplus electricity is assumed to be sold to the microgrid of the industrial hub at a predefined tariff ($T_{se}$). Given that the surplus generation duration curve closely resembles that of a WF, a selling tariff of 60 €/MWh has been assumed, consistent with recent wind power purchase agreement signed in Spain [66]. The base-case analysis assumes that all surplus electricity can be sold at this tariff, without explicitly accounting for the hourly electricity demand of industrial consumers or the capacity constraints of the internal microgrid.

While solar PV exhibits low costs [16], it typically produces electricity for approximately 1500 equivalent full-load hours per year, with time-varying output [31,33,34]. In contrast, wind farms, although generally more expensive [16], can produce electricity over a larger fraction of the year [67], while its output remains variable. To balance these complementary characteristics, both energy sources have been considered in this study at different shares.

Since the LCOH is evaluated within the Spanish context, the economic assumptions are aligned with those adopted in the IBHYX [19] and are summarized in Table 1, reflecting the current technological and economic conditions considered for Spain. According to [19], the investment of the electrolyzer includes water treatment equipment, stack, balance of plant (components, auxiliary systems and equipment, instrumentation and control, etc.), power supply (substation and connection), engineering, procurement and construction. Land acquisition costs, fees and licenses are not included. The stack degradation is assumed to be 0.12% per 1000 operating hours, with a nominal electricity consumption ($g$) of 55.5 kWh/kg [19]. The stack is assumed to be replaced after 80,000 h of operation [19]. Following the methodology presented in [68], the average specific consumption over the project‘s lifetime ($\overline{g}$) is determined in each case. The LCOH is obtained from Equation (1), whose components are detailed in Equations (2)–(9).

$$LCOH={CAPEX}_{ELY}+{CAPOPEX}_{feedstock}+{OPEX}_{omELY}+{OPEX}_{repstack}+{REVENUES}_{surpluses}$$

(1)

Table 1. Economic assumptions used to calculate the LCOH (Equation (1)), based on the Spanish reference conditions reported in the IBHYX [19].

Variable	Value	Units
${INV}_{ELYe}$	1600	€/kW
fom	4	%
$g$	55.5	kWh/kg
${INV}_{PVe}$	500	€/kW
${OM}_{PV}$	5	€/kW-year
${INV}_{WFe}$	1100	€/kW
${OM}_{WF}$	50	€/kW-year
N	25	years
$H_{stack}$	80,000	h
$f_{rs}$	15	%
wacc	7.74	%

Equation (2) defines the investment cost of the electrolyzer (CAPEX_ELY) as a function of its specific investment cost (INV_ELY,e) and the annual operating hours (H). The operational costs associated with electricity supply to the electrolyzer (CAPOPEX_feedstock) are calculated using Equation (3), where the LCOE is derived in Equations (4)–(6). These equations consider the specific investment costs of PV and WF plants (INV_PVe and INV_WFe), their respective operation and maintenance costs (OM_PV and OM_WF), and their installed peak capacities (P_PV and P_WF). Electrolyzer operation and maintenance costs are calculated using Equation (7), based on the capital investment and a maintenance factor (f_om). Stack replacement costs are modeled by Equation (8) as a function of the total operating hours before replacement (H_stack), and the replacement factor (f_rs). Finally, revenues from surplus electricity sales (REVENUES_surpluses) are quantified in Equation (9).

$${CAPEX}_{ELY}={\displaystyle \frac{{INV}_{ELYe}·\overline{g}·CRF}{H}}$$

(2)

$${CAPOPEX}_{feedstock}={\displaystyle \frac{LCOE·\overline{g}·\left(E_{PV}+E_{WF}\right)}{1000·E_{ELY}}}$$

(3)

$$LCOE={CAPEX}_E+{OPEX}_E$$

(4)

$${CAPEX}_E=\left({\displaystyle \frac{{INV}_{PVe}·P_{PV}+{INV}_{WFe}·P_{WF}}{E_{PV}+E_{WF}}}\right)·CRF·1000$$

(5)

$${OPEX}_E=\left({\displaystyle \frac{{OM}_{PV}·P_{PV}+{OM}_{WF}·P_{WF}}{E_{PV}+E_{WF}}}\right)·1000$$

(6)

$${OPEX}_{omELY}={\displaystyle \frac{{INV}_{ELYe}·\overline{g}·f_{om}}{H}}$$

(7)

$${OPEX}_{repstack}={\displaystyle \frac{floor\left(\frac{H·N}{H_{stack}}\right)·{INV}_{ELYe}·\overline{g}·f_{rs}}{H·N}}$$

(8)

$${REVENUES}_{surpluses}=-T_{se}·\left({\displaystyle \frac{\overline{g}}{1000}}\right)·\left({\displaystyle \frac{E_{PV}+E_{WF}-E_{ELY}}{E_{ELY}}}\right)$$

(9)

The nominal escalation rate of the operational costs is assumed to be zero and the capital recovery factor (CRF) is specified in Equation (10), where wacc stands for the weighted average capital cost and N is the lifespan of the project.

$$CRF={\displaystyle \frac{wacc·{\left(1+wacc\right)}^N}{{\left(1+wacc\right)}^N-1}}$$

(10)

For comparison purposes, a conventional LCOH reference (LCOH_conv) is also calculated for the same plant configuration. This indicator includes electrolyzer investment, operation and maintenance, stack replacement, and the cost of electricity actually consumed for hydrogen production, valued at the LCOE of the hybrid renewable plant, without accounting for revenues from surplus electricity sales. This conventional LCOH is calculated using Equations (11) and (12).

$$LCO{H}_{conv}={CAPEX}_{ELY}+{CAPOPEX}_{feedstock,conv}+{OPEX}_{omELY}+{OPEX}_{repstack}$$

(11)

$${CAPOPEX}_{feedstock,conv}={\displaystyle \frac{LCOE·\overline{g}}{1000}}$$

(12)

The proposed methodology was applied to 17 selected locations in Spain, one in each autonomous community. One representative site was selected in each autonomous community, prioritizing areas with the highest wind energy potential based on data from the Global Wind Atlas [69]. This selection criterion was adopted considering the geographical characteristics of solar and wind resources in Spain. Solar irradiation exhibits a marked regional distribution, generally increasing from northern to southern Spain, as documented by the Spanish Solar Radiation Atlas published by AEMET [70]. In contrast, wind resources are strongly influenced by local topography and exhibit substantial spatial variability within the Iberian Peninsula [71]. Consequently, site selection focused on identifying locations with favorable wind resources within each autonomous community, while the solar resource was evaluated at the same selected locations. The geographical coordinates of the 17 selected locations are reported in Table 2. These coordinates were used to obtain the hourly wind and solar generation profiles considered in the techno-economic simulations.

Table 2. Geographical coordinates of the selected modeling locations in the 17 Spanish autonomous communities.

Autonomous Community	Latitude (°N)	Longitude (°E)
Andalucía	36.11	−5.67
Aragon	41.73	−1.75
Canarias	28.10	−17.17
Cantabria	43.25	−3.58
Castilla-La Mancha	38.43	−2.40
Castilla y León	42.96	−4.65
Cataluña	40.89	0.36
Comunidad de Madrid	40.79	−3.94
Comunidad Valenciana	40.37	−0.03
Extremadura	40.19	−5.66
Galicia	42.44	−8.41
Islas Baleares	39.90	2.96
La Rioja	42.23	−2.93
Navarra	42.88	−1.94
País Vasco	42.99	−2.39
Principado de Asturias	43.12	−6.03
Región de Murcia	38.27	−1.19

For each selected location, hourly wind power generation profiles were obtained from [72], based on the methodology described in [33] and [67], while hourly solar PV generation profiles were obtained from the European Commission’s PVGIS tool [34].

At each site, a 500 MW hybrid renewable energy system was modeled where the contribution of wind power was varied from 0% to 100% in increments of 2%. Each increment corresponds to replacing 10 MW of photovoltaic capacity with 10 MW of wind capacity, while maintaining a constant total installed capacity of 500 MW. Thus, 51 different configurations were evaluated at each location to identify the renewable energy mix that minimizes the LCOH. The selected capacity of the hybrid renewable field is representative of the scale of large renewable energy projects in Spain, where four of the largest wind farms range from 130 MW to 800 MW [73], whereas there are 11 PV plants with more than 300 MW, being the maximum 430 MW [74]. Nevertheless, under the linear scaling assumptions adopted in the model, the main results are expected to be independent of the total installed capacity. The corresponding LCOH was calculated for each configuration to identify the optimal wind–PV capacity mix.

To assess the influence of revenues from surplus electricity sales on the economic performance of the CHyP system, a sensitivity analysis was performed considering three electricity selling prices: 40, 60, and 80 €/MWh. These values were selected considering recent Spanish PPA market assessments [66,75], with the upper value representing a favorable electricity selling scenario. For each selling price, the optimization procedure was repeated for all 17 Spanish autonomous communities to determine the corresponding minimum LCOH.

Additionally, another sensitivity analysis was performed to evaluate the effect of partial surplus electricity sales on the economic performance of the CHyP system. For each optimal configuration, the minimum fraction of surplus electricity that must be sold for the revenue-adjusted LCOH to equal the conventional LCOH was determined, assuming the reference selling tariff of 60 €/MWh.

3. Results

Figure 2 illustrates the calculated LCOH and the corresponding LCOE as a function of the wind power share for a representative location in the autonomous community of Aragón (Spain). The wind power share ranges from 0% (i.e., a 500 MW plant fully based on solar PV) to 100% (i.e., 500 MW entirely from wind power). When relying exclusively on wind power, the LCOH reaches 3.5 €/kg H₂, whereas using only solar power results in a cost of 5.75 €/kg H₂. All hybrid wind–solar configurations considered result in a lower LCOH than the solar-only configuration, while the optimum hybrid configuration also outperforms the wind-only case. For this specific location—which exhibits the lowest LCOH among the 17 selected sites—the minimum value is 1.68 €/kg H₂, corresponding to a wind power share of 26% and an LCOE of 37.34 €/MWh.

Figure 2. LCOH and LCOE as a function of the wind power share for the autonomous community of Aragón in Spain.

The abrupt decrease in LCOH observed between wind power shares of 22% and 24% is associated with a change in the electrolyzer sizing determined from the hybrid generation duration curve. As the wind–PV capacity mix changes, the operating point selected under the adopted sizing criterion shifts towards a substantially longer operating period and a lower nominal electrolyzer capacity. Specifically, electrolyzer operating hours increase from 2520 to 4945 h/year, while its capacity decreases from 135.0 to 68.7 MW. Annual hydrogen production remains nearly constant (approximately 5900 t/year), whereas the lower electrolyzer capacity significantly reduces the investment and maintenance costs per unit of hydrogen produced. Consequently, the LCOH decreases from 4.41 to 2.12 €/kg. This sharp variation reflects the sensitivity of the electrolyzer sizing procedure to changes in the shape of the hybrid generation duration curve.

The same analysis was performed for the 17 Spanish autonomous communities. Figure 3 shows the geographical distribution of the minimum LCOH and the corresponding optimal wind power shares, while Table 3 summarizes these results together with the associated LCOE and electrolyzer operating hours. All results correspond to the specific modeling locations identified in Table 2, with the autonomous communities ordered by increasing LCOH in Table 3 and also the LCOE corresponding to the wind share that minimizes the LCOH and the electrolyzer operation hours. As expected, a similar trend is observed between LCOH and LCOE, with lower LCOH values generally associated with lower LCOEs. The lowest LCOH values are observed in locations with higher wind energy potential, with the optimal wind power share being 27% on average. The LCOH ranges from 1.68 to 4.4 €/kg, meaning that even in the least favorable autonomous communities, the hybridization of solar and wind yields a competitive cost when compared to the official Iberian renewable hydrogen price index (6.28 €/kg as of 20/10/2026) [19]. It is also worth noting that the assumed electricity sale tariff of 60 €/MWh is significantly higher than the resulting LCOE in the best-performing locations—by up to 60%. Table 3 also includes the conventional LCOH reference (LCOH_conv), calculated without considering revenues from electricity sales. Its average value across the 17 locations is 5.11 €/kg, compared with 3.06 €/kg for the LCOH obtained under the CHyP configuration, highlighting the economic contribution of surplus electricity valorization.

Figure 3. Geographic distribution of (a) minimum LCOH and (b) optimal wind power share across the Spanish autonomous communities. Results correspond to the selected modeling locations whose coordinates are provided in Table 2. [Maps created with Datawrapper (https://www.datawrapper.de/ (accessed on 15 October 2026))].

Table 3. Results for the optimal CHyP configurations in each Spanish autonomous community, including LCOH accounting for surplus electricity revenues, conventional LCOH reference (LCOH_conv), optimal wind power share, LCOE and electrolyzer operating hours.

Autonomous Community	Min. LCOH [€/kg H2]	LCOHconv [€/kg H2]	Wind Share [%]	LCOE [€/MWh]	H [h/Year]
Aragón	1.68	4.35	26	37.34	5829
Canarias	1.86	4.37	30	37.03	5727
Cataluña	1.98	4.63	24	37.18	5121
Comunidad Valenciana	2.08	4.63	24	36.43	5032
Andalucía	2.1	4.70	28	36.11	4861
La Rioja	2.78	5.05	26	41.36	4780
Comunidad de Madrid	3.01	5.03	32	42.27	4917
Islas Baleares	3.02	5.16	24	39.44	4412
Castilla-La Mancha	3.17	5.27	32	40.43	4330
Galicia	3.24	4.92	30	45.62	5386
Cantabria	3.4	5.24	18	46.36	4957
Región de Murcia	3.45	5.27	28	43.52	4610
Castilla y León	3.73	5.68	26	42.80	3961
Navarra	3.84	5.58	26	45.86	4337
Principado de Asturias	4.05	5.78	24	46.21	4082
País Vasco	4.23	5.58	22	49.85	4695
Extremadura	4.4	5.61	40	48.04	4472

The sale of surplus electricity plays a key role in the economic performance of the proposed CHyP model. Equation (1) considers the hybrid renewable power plant as an asset integrated into the hydrogen production facility. Accordingly, its electricity generation can either be used to supply the electrolyzer or sold to other industrial consumers through the hub microgrid. In contrast, Equation (11) (LCOH_conv) represents a conventional configuration in which the electrolyzer purchases electricity from an external supplier at the corresponding LCOE. The lower LCOH achieved by the integrated configuration relies on revenues from surplus electricity sales. However, the assumption that all surplus generation can be sold may not always be realistic, as the actual electricity absorbed by the hub depends on the demand profiles of its industrial consumers and the available distribution capacity. This limitation might reduce the competitiveness of the proposed idealized model.

To assess the influence of this assumption, Figure 4 presents the minimum percentage of surplus electricity that must be sold for the integrated CHyP configuration to achieve the same LCOH as the conventional configuration. Locations with lower LCOH generally require a smaller percentage of surplus electricity sales, although some variations arise from differences in LCOE. Selling a higher percentage of the surplus electricity than the threshold shown in Figure 4 results in a lower LCOH for the integrated configuration, whereas selling a smaller percentage makes the conventional configuration more economically attractive. Therefore, depending on the magnitude and temporal distribution of electricity demand within the industrial hub, these results can help assess the suitability of integrating the renewable power plant into the hydrogen production facility rather than purchasing electricity from an external supplier.

Figure 4. Percentage of surplus electricity that must be sold for the integrated CHyP configuration to achieve the same LCOH as the conventional configuration (LCOH_conv). Higher percentages of surplus electricity sales result in lower LCOH values for the integrated configuration.

Table 4 summarizes the main technical parameters of the optimal CHyP configurations across the 17 Spanish autonomous communities. The electrolyzer capacity ranges from approximately 45 to 83 MW, considerably below the total installed renewable capacity of 500 MW. On average, approximately two-thirds of the renewable electricity generation is available as surplus electricity, which is assumed to be sold to industrial consumers under the reference scenario, while the remaining third is used for hydrogen production. The sold electricity requires that the industrial demand matches with the surplus electricity profile and the distribution capacity of the microgrid is enough. The electrolyzer capacity factor varies between approximately 45% and 67%, reflecting the differences in renewable resource availability and the resulting operating hours across the selected locations. Electrolyzer downsizing was defined as the percentage reduction in nominal electrolyzer capacity relative to the total installed renewable capacity of 500 MW. The electrolyzer capacity factor was calculated as the ratio of its annual full-load operating hours to the total hours in a year (8760 h), consistent with the nominal-load operating strategy adopted in this study.

Table 4. Main technical parameters of the optimal CHyP configurations in each Spanish autonomous community, considering a surplus electricity selling price of 60 €/MWh.

Autonomous Community	ELY (MW)	H2 (t/yr)	RES (GWh/yr)	ELY cons. (GWh/yr)	Surplus (GWh/yr)	Down. (%)	CF (%)
Aragón	58.2	5849	1028.1	339.0	689.1	88.4	66.5
Canarias	65.9	6514	1090.7	377.3	713.4	86.8	65.4
Cataluña	65.3	5788	1005.7	334.4	671.3	86.9	58.5
Comunidad Valenciana	71.0	6186	1026.5	357.2	669.3	85.8	57.4
Andalucía	77.7	6547	1090.8	377.9	712.9	84.5	55.5
La Rioja	62.5	5172	928.3	298.5	629.8	87.5	54.6
Comunidad de Madrid	67.0	5709	979.1	329.6	649.5	86.6	56.1
Islas Baleares	76.7	5868	948.2	338.6	609.5	84.7	50.4
Castilla-La Mancha	82.7	6203	1023.6	358.0	665.7	83.5	49.4
Galicia	52.4	5086	902.2	294.4	607.7	89.5	64.2
Cantabria	45.0	3863	741.9	223.0	518.9	91.0	56.6
Región de Murcia	67.5	5394	905.1	311.3	593.9	86.5	52.6
Castilla y León	76.4	5236	897.1	302.5	594.6	84.7	45.2
Navarra	61.8	4646	837.2	268.1	569.1	87.6	49.5
Principado de Asturias	62.5	4417	809.2	255.1	554.2	87.5	46.6
País Vasco	47.0	3820	730.0	220.4	509.6	90.6	53.6
Extremadura	76.5	5930	944.9	342.2	602.7	84.7	51.1

ELY: electrolyzer capacity; H₂: hydrogen production; RES: renewable electricity generation (wind + solar); ELY cons.: electricity consumed by the electrolyzer; Surplus: electricity sold; Down.: electrolyzer downsizing; CF: electrolyzer capacity factor.

The sensitivity analysis described in Section 2 was conducted to evaluate the effect of surplus electricity selling prices on the economic performance of the CHyP system. Figure 5 compares the minimum LCOH values obtained in each autonomous community for the three electricity selling prices. As expected, increasing the selling price reduces the LCOH in all locations, although the magnitude of this reduction varies among regions due to differences in renewable electricity generation, electrolyzer operation, and surplus electricity availability. At the highest selling price, negative LCOH values are obtained in certain locations. These negative values indicate that revenues from surplus electricity sales exceed the annualized costs considered in the CHyP economic framework. They should therefore be interpreted as the result of the adopted revenue-adjusted cost calculation, rather than as negative physical hydrogen production costs.

Figure 5. Sensitivity of the minimum LCOH to surplus electricity selling prices of 40, 60, and 80 €/MWh across the 17 Spanish autonomous communities.

Figure 6 summarizes the sensitivity results in terms of the average LCOH across the 17 locations and the corresponding minimum–maximum range. The average LCOH decreases from 5.30 €/kg at a selling price of 40 €/MWh to 3.06 €/kg under the reference scenario of 60 €/MWh. A further increase to 80 €/MWh reduces the average LCOH to 0.73 €/kg. The corresponding regional ranges are 4.03–6.67 €/kg, 1.68–4.40 €/kg, and −0.67–2.35 €/kg, respectively. At the highest selling price, five of the 17 locations exhibit negative LCOH values, as it has been explained.

Figure 6. Sensitivity of the average LCOH to the surplus electricity selling price, showing the minimum–maximum range across the 17 Spanish autonomous communities. The dashed line indicates the reference scenario of 60 €/MWh.

In addition to its impact on the LCOH, the surplus electricity selling price also affects the optimal wind–PV capacity mix. Figure 7 summarizes the variation in the average optimal wind share across the 17 Spanish autonomous communities, together with the corresponding minimum–maximum range. The average optimal wind share decreases from 36.5% at 40 €/MWh to 27.2% at 60 €/MWh and 26.5% at 80 €/MWh. Therefore, the average optimal wind share of approximately 27% obtained under the reference scenario remains relatively stable when the electricity selling price increases to 80 €/MWh, whereas a lower selling price of 40 €/MWh favors configurations with a higher wind contribution.

Figure 7. Variation in the optimal wind share with the surplus electricity selling price. The figure shows the average value and the minimum–maximum range obtained across the 17 Spanish autonomous communities. Dashed line shows the selling price used in this study.

4. Discussion

To interpret the resulting LCOH, its components have been grouped into two terms: ${CAPEX}_{ELY}$, ${OPEX}_{omELY}$ and ${OPEX}_{repstack}$ have been combined in the electrolyzer term (ELY), whereas ${CAPOPEX}_{feedstock}$ and ${REVENUES}_{surpluses}$ have been combined in the energy term (ENERGY). Figure 8 shows these two terms for each Spanish autonomous community. In the nine communities with the lowest LCOH, the ENERGY term is negative or very small. This occurs when revenues from surplus electricity sales exceed the renewable electricity costs allocated to hydrogen production under the adopted accounting framework. In the remaining communities, the LCOE is higher, so hybrid-plant costs are not fully offset by revenues and the ENERGY term becomes substantially larger, although the ELY term remains the dominant contribution.

Figure 8. Main terms of the LCOH in each autonomous community.

The ENERGY term mainly depends on the LCOE, because including revenues ensures that all electricity generation is used, either in the electrolyzer or supplied to an industrial consumer. The ELY term is largely driven by CAPEX and therefore correlates with the inverse of the electrolyzer operating hours (η). Figure 9 shows the strong correlation obtained with these independent variables. Higher dispersion in ENERGY curve is due to the variability of electricity surplus due to the different share of wind energy in each optimum case, affecting the revenues.

Figure 9. Linear regression analysis of LCOH terms as a function of its corresponding explanatory variables.

Taking this into account, a multivariable linear regression is derived, yielding Equation (13), which attains an R² value of 98.31%. The strong correlation is partly attributable to the mathematical structure of the LCOH formulation, which explicitly depends on the LCOE and electrolyzer operating hours. Therefore, the regression should be interpreted as a descriptive relationship rather than as an independent validation of the model.

$$LCOH=-6.367+0.149·LCOE+{\displaystyle \frac{\mathrm{14,974}}{H}}$$

(13)

Figure 10 shows the scatter of minimum LCOH values (red points), together with iso-LCOH lines derived from Equation (13) used to classify them. The electrolyzer operating hours vary markedly across locations, reflecting differences in the local renewable resource, whereas LCOH is strongly influenced by the LCOE. This is evident in the five locations with the lowest LCOH, which all cluster around 2 €/kg because their LCOE is nearly the same (about 37 €/MWh), even though electrolyzer operating hours range from 4800 to 5800. The iso-LCOH lines should be taken as a mere clustering tool, as the strong correlation shown in Figure 9 is due to structural factors.

Figure 10. Coordinates of the minimum LCOH for each autonomous community (red points) on the LCOE-H plane. Black lines show the iso-LCOH contours derived from Equation (13).

Figure 8 identifies two groups of autonomous communities according to the sign of the ENERGY contribution to the LCOH: those with negative values and those with positive values. To further investigate this behavior, two representative locations with intermediate ENERGY contributions within each group were selected: Comunidad Valenciana, with a negative ENERGY term, and Región de Murcia, with a positive ENERGY term. Figure 11 and Figure 12 show the electrolyzer electricity consumption and surplus electricity generation profiles for a representative day in the middle of each season at both locations. The electrolyzer operates continuously once started, except during autumn, when interruptions lasting between 3 and 5 h occur. This relatively steady operating pattern is expected to mitigate accelerated stack degradation.

Figure 11. Electrolyzer consumption and electricity surplus profiles for each season in the location of Comunidad Valenciana.

Figure 12. Electrolyzer consumption and electricity surplus profiles for each season in the location of Región de Murcia.

A surplus electricity selling price of 60 €/MWh was assumed, consistent with the average day-ahead market prices during periods of surplus electricity generation. Figure 13 presents the average hourly day-ahead electricity prices for each season in Spain over the period from 10 March 2024 to 9 March 2025. Weighting these hourly prices by the corresponding surplus electricity generation yields average electricity market values of 61.45 €/MWh for Comunidad Valenciana and 61.39 €/MWh for Región de Murcia. Therefore, the assumed selling price of 60 €/MWh is slightly below these surplus-weighted market values. Moreover, after the blackout in Spain on 28 April 2025, the electricity day-ahead market price has been rising, recently affected by Strait of Hormuz crisis. Thus, in 2024 the annual electricity day-ahead market price weighted by the surplus was 63.03 €/MWh, growing up to 65.73 €/MWh in 2025 and increasing up to 76.75 €/MWh in 2026 (until 30 September) [76]. These figures provide additional context for the assumed selling price, although future electricity market conditions remain uncertain. Moreover, direct electricity supply from the renewable power plant to the hub microgrid could reduce certain distribution grid-related charges, potentially improving the economic attractiveness of the proposed configuration. On the other hand, it is required that the surplus electricity profile matches with the industry demand in the hub, as well as the distribution capacity of the microgrid is enough.

Figure 13. Electricity day-ahead market price in Spain between 10 March 2024 to 9 March 2025 (average in each season) [76].

The ENERGY term can be expressed as shown in Equation (14), where op (overproduction) denotes the ratio of surplus electricity generation to electrolyzer electricity consumption, as defined in Equation (15). In this formulation, the LCOE is multiplied by a larger coefficient than the electricity selling price, and the sign of the ENERGY term depends on the balance between these contributions. The overproduction ratio is 2.84 p.u. for Comunidad Valenciana and 2.52 p.u. for Región de Murcia.

The break-even electricity selling price is 55.89 €/MWh for Comunidad Valenciana, below the assumed selling price of 60 €/MWh, and 68.12 €/MWh for Región de Murcia, above the assumed selling price. These results highlight the importance of the LCOE in determining the sign of the ENERGY contribution, despite the lower electricity overproduction ratio in Región de Murcia. Figure 14 presents the LCOH breakdown for both locations, illustrating the balance between electricity supply costs (CAPOPEX) and revenues from surplus electricity sales (REVENUES) within the ENERGY term.

$$ENERGY=\left({\displaystyle \frac{\overline{g}}{1000}}\right)·\left[LCOE·op-T_{se}·\left(op-1\right)\right]$$

(14)

$$op={\displaystyle \frac{E_{PV}+E_{WF}}{E_{ELY}}}$$

(15)

Figure 14. Breakdown of LCOH at the selected locations.

A sensitivity analysis of the main techno-economic parameters was conducted for the two selected locations, with the results presented in Figure 15 and Figure 16. Financial parameters (N and wacc) emerge as the main drivers of LCOH because they influence both the ENERGY (WF and PV CAPEX) and ELY (ELY CAPEX) cost components. The sensitivity to these parameters is greater in Región de Murcia, where revenues from surplus electricity sales provide a smaller offset to renewable generation costs. The influence of electrolyzer electricity consumption is similar in both locations, as its influence is limited to the ENERGY term, whose behavior is largely determined by renewable generation costs (WF and PV CAPEX) and electricity sales revenues.

Figure 15. Sensitivity analysis of the LCOH at selected locations.

Figure 16. Sensitivity analysis of the LCOH at selected locations against H_stack.

Stack lifetime is another relevant parameter. However, given the relatively limited annual operating hours of the electrolyzer, a wider variation range (±50%) was considered to assess its influence, compared with the ±20% applied to the other parameters. As shown in Figure 16, the resulting sensitivity of LCOH to stack lifetime is relatively small in both locations, although somewhat greater in Comunidad Valenciana, where the ELY component accounts for a larger proportion of the total LCOH than in Región de Murcia.

The electrolyzer operating hours were selected based on the normalized annual hydrogen production curves shown in Figure 17 for the two representative locations. These curves typically exhibit a region of maximum production, which may be relatively flat or contain local maxima, followed by a descending branch with an increasingly steep slope. The operating points obtained using the quantitative sizing criterion described in Section 2 lie near the onset of a steeper decline in annual hydrogen production, as illustrated by the two representative locations shown in Figure 17. Across the 17 modeled locations, the 90% electricity reference provides a practical compromise, with resulting hydrogen production levels of approximately 85–95% of the maximum. The shape of the descending branch also indicates that further increases in operating hours could lead to substantially greater hydrogen production losses. This approach increases electrolyzer operating hours while limiting the reduction in hydrogen production to approximately 5–15% of its maximum value.

Figure 17. Normalized hydrogen production at the selected locations. Dashed lines indicate the selected working point.

As an example, in Comunidad Valenciana (Figure 17a), selecting an operating point corresponding to 85% of the maximum annual hydrogen production allows a substantial increase in electrolyzer operating hours while avoiding the region of the curve where further increases would lead to a pronounced reduction in hydrogen production. In contrast, the hydrogen production curve for Región de Murcia (Figure 17b) exhibits a relatively flat region, allowing 95% of the maximum annual hydrogen production to be maintained before a marked decline occurs with increasing operating hours.

The resulting reduction in LCOH is mainly attributable to two effects: the decrease in the ELY component, associated with lower electrolyzer investment and maintenance costs per kilogram of hydrogen produced, and the reduction in the ENERGY component, resulting from increased revenues from surplus electricity sales. Table 5 illustrates these effects for the two representative locations. The ELY component decreases by more than 50% in Comunidad Valenciana and approximately 48% in Región de Murcia. Meanwhile, the ENERGY component decreases by more than 17% in Región de Murcia and changes from positive to negative in Comunidad Valenciana. Overall, the ELY component exhibits the greatest sensitivity to the increase in electrolyzer operating hours, although the additional electricity sales revenues also contribute to reducing the LCOH.

Table 5. Comparison of hydrogen production costs under different electrolyzer sizing criteria at the selected locations. Adopted criterion in this work consists of selecting the operating point with the highest number of operating hours among those whose annual electricity available to the electrolyzer corresponds to 90% of the maximum.

	Comunidad Valenciana		Región de Murcia
Normalized Cost	Maximum ELY Consumption	Adopted Criterion	Maximum ELY Consumption	Adopted Criterion
ELY CAPEX [€/kg]	3.54	1.68	3.68	1.83
om ELY OPEX [€/kg]	1.54	0.73	1.61	0.80
rep. stack OPEX [€/kg]	0.00	0.11	0.00	0.12
feedstock CAPOPEX [€/kg]	5.02	5.92	5.61	5.93
surpluses REVENUES [€/kg]	−4.87	−6.36	−4.75	−5.22
LCOH	5.23	2.08	6.14	3.45

An additional sensitivity analysis has been performed, as mentioned in the Methods section, that highlights the substantial influence of surplus electricity selling prices on the economic competitiveness of the CHyP system. In particular, at the highest selling price considered, five locations exhibit negative calculated LCOH values, indicating that revenues from surplus electricity sales exceed the total annualized costs considered within the CHyP economic framework. These results should not be interpreted as negative physical hydrogen production costs, but rather as a consequence of the adopted economic accounting approach, in which electricity sales revenues are deducted from the total system costs.

These findings emphasize that the LCOH values obtained under the reference scenario (60 €/MWh) are strongly dependent on the assumed electricity selling price. Consequently, the economic feasibility of future CHyP projects will depend not only on local renewable resources and electrolyzer operating conditions, but also on the prices and contractual arrangements under which surplus electricity can be sold. In fact, additional requirements must be met: profiles of industrial demand and surplus electricity should match and distribution capacity of the microgrid should be enough.

The sensitivity analysis also provides useful insight into the robustness of the optimal wind–PV mix. Under the reference scenario, the average optimal wind share is approximately 27%. This value remains relatively stable when the selling price increases from 60 to 80 €/MWh, with unchanged optimal configurations in 14 of the 17 locations. However, reducing the selling price to 40 €/MWh increases the average optimal wind share to 36.5%, suggesting that configurations with a higher wind contribution become more economically favorable when electricity export revenues decrease. Therefore, within the electricity selling price range analyzed, the optimal wind–PV mix is more sensitive to a reduction in the selling price below the reference value of 60 €/MWh than to an increase above it.

A limitation of the study concerns the geographical selection of the modeling locations. Only one site was considered within each autonomous community, with priority given to favorable wind resources. This selection criterion may influence the resulting optimal wind–PV configurations. Alternative locations with higher solar irradiation or lower wind availability could favor a greater photovoltaic contribution and lead to different LCOH values. Accordingly, the approximately 27% average optimal wind share should be regarded as a result specific to the selected wind-favorable sites and the adopted techno-economic assumptions, rather than as a universally applicable optimum for Spain. A broader spatial analysis considering multiple sites within each autonomous community, including solar-dominant locations, would be necessary to evaluate the geographical robustness of the optimal configurations. Moreover, interannual variability in wind and solar resources may affect the resulting LCOH, while the economic benefits of surplus electricity sales depend on actual industrial demand and export-capacity constraints, which have not been explicitly modeled. Therefore, the economic competitiveness of the proposed CHyP configuration should not be generalized beyond the conditions considered in this study.

5. Conclusions

Green hydrogen has emerged as a key enabler of the energy transition, especially in hard-to-abate sectors. However, its widespread adoption is currently limited by high production costs. To address this challenge, this paper proposes a co-production-based business model aimed at reducing the LCOH through two main strategies: (i) the hybridization of wind and solar power to optimize electricity supply, and (ii) the sale of electricity surpluses to industrial end-users. For this model to be economically attractive, the facility should be integrated into an industrial hub with sufficient demand well matched with the electricity surpluses and enough distribution capacity.

The main conclusions derived from this study are as follows:

• The hybridization of wind and solar power, combined with the valorization of surplus electricity, offers significant economic benefits for hydrogen production in Spain. For the 17 selected locations across all the autonomous communities, the combined use of wind and solar consistently lowers the LCOH compared to configurations relying on a single source, with an average optimal installed wind power share of approximately 27%. The lowest LCOH—1.68 €/kg H₂—was observed in Aragón, but even in less favorable regions, values remain below 4.4 €/kg, significantly under the official Iberian Renewable Hydrogen Index price. These values assume that all the electricity surplus can be absorbed by the nearby demand.
• The LCOE corresponding to the optimal wind share is substantially lower than the assumed electricity sale tariff (60 €/MWh), further reinforcing the economic viability of this co-production approach, if the electricity surplus can be absorbed by the nearby demand. In spite of this difference, the selling price is slightly lower than the market price weighted by the surplus in 2024 (63.03 €/MWh) and 2025 (65.73 €/MWh) and clearly lower than in 2026 (76.75 €/MWh). This reduction regarding the market is due to the wind contribution, which allows selling electricity in peak hours. This difference might be enlarged due to direct electricity supply through the hub microgrid could reduce certain grid-related charges, depending on the applicable regulatory framework.
• The surplus electricity selling price strongly influences the economic performance of the CHyP system. The average LCOH decreases from 5.30 to 0.73 €/kg H₂ as the selling price increases from 40 to 80 €/MWh. At the highest price considered, five locations exhibit negative calculated LCOH values, reflecting the adopted economic accounting approach rather than negative physical hydrogen production costs. These results emphasize the importance of electricity sales conditions in assessing project viability.
• The optimal wind–PV capacity mix is relatively robust to higher electricity selling prices but more sensitive to lower prices. The average optimal wind share is 27.2% and 26.5% at selling prices of 60 and 80 €/MWh, respectively, with unchanged optimal configurations in 14 of the 17 locations. However, reducing the selling price to 40 €/MWh increases the average optimal wind share to 36.5%, indicating that configurations with a greater wind contribution become more favorable when electricity export revenues decrease.
• Electrolyzer utilization is a key factor in the economic performance of the proposed configuration. The adopted duration-curve-based sizing approach selects operating points that increase electrolyzer operating hours while avoiding a pronounced reduction in annual hydrogen production. For the configurations analyzed, this procedure results in hydrogen production levels between 85% and 95% of the maximum annual production, depending on the generation profile. This approach enables the selection of electrolyzer capacities substantially below the total installed renewable capacity, reducing the investment cost per unit of hydrogen produced. The results also indicate that LCOH variations are strongly associated with LCOE and electrolyzer operating hours, although the regression describing this relationship should be interpreted as descriptive rather than predictive.
• The sensitivity analysis identifies financial parameters as important drivers of hydrogen production costs. In particular, the weighted average cost of capital and project lifetime have a substantial influence on LCOH, as they affect both renewable generation and electrolyzer investment costs. By comparison, stack lifetime exhibits a smaller influence within the operating conditions and parameter ranges examined.
• The findings are conditional on the adopted geographical and operational assumptions. The approximately 27% average optimal wind share corresponds to the selected wind-favorable locations and should not be regarded as a universal optimum for Spain. Furthermore, the economic competitiveness of the proposed CHyP configuration depends on the assumed renewable resource profiles, electricity selling prices, and surplus electricity sales. Although sensitivity analyses assess the influence of electricity selling prices and surplus utilization, uncertainties associated with interannual wind and solar resource variability and the temporal matching between surplus generation and industrial electricity demand have not been explicitly modeled. In addition, the reference scenario assumes that all surplus electricity can be sold at the specified tariff, without considering demand profiles or export-capacity constraints. Therefore, the favorable economic results obtained under this idealized assumption may not be fully achievable under practical operating conditions and should not be generalized beyond the scenarios analyzed. Future work should therefore consider multiple locations within each autonomous community, alternative renewable resource profiles, multi-year renewable generation data, and the integration of industrial demand and grid constraints to assess the practical feasibility and geographical robustness of the proposed CHyP configurations.

Overall, the results demonstrate the potential of integrating hydrogen production and surplus electricity supply within hybrid renewable energy systems to improve project economics. However, successful implementation of the CHyP concept will require suitable industrial off-takers, reliable electricity sales arrangements, and adequate grid and hydrogen infrastructure. Future developments should therefore combine renewable resource optimization with industrial demand assessment and infrastructure planning to identify the most promising deployment opportunities.