Fuel Cell Electric Vehicle as a Power Plant: Techno-Economic Scenario Analysis of a Renewable Integrated Transportation and Energy System for Smart Cities in Two Climates

Renewable, reliable, and affordable future power, heat, and transportation systems require efficient and versatile energy storage and distribution systems. If solar and wind electricity are the only renewable energy sources, what role can hydrogen and fuel cell electric vehicles (FCEVs) have in providing year-round 100% renewable, reliable, and affordable energy for power, heat, and transportation for smart urban areas in European climates? The designed system for smart urban areas uses hydrogen production and FCEVs through vehicle-to-grid (FCEV2G) for balancing electricity demand and supply. A techno-economic analysis was done for two technology development scenarios and two different European climates. Electricity and hydrogen supply is fully renewable and guaranteed at all times. Combining the output of thousands of grid-connected FCEVs results in large overcapacities being able to balance large deficits. Self-driving, connecting, and free-floating car-sharing fleets could facilitate vehicle scheduling. Extreme peaks in balancing never exceed more than 50% of the available FCEV2G capacity. A simple comparison shows that the cost of energy for an average household in the Mid Century scenario is affordable: 520–770 €/year (without taxes and levies), which is 65% less compared to the present fossil situation. The system levelized costs in the Mid Century scenario are 71–104 €/MWh for electricity and 2.6–3.0 €/kg for hydrogen—and we expect that further cost reductions are possible.


Introduction
The Paris Agreement, which pledges to keep global warming well below 2 degrees Celsius above pre-industrial levels and to limit the increase to 1.5 degrees Celsius, needs a boost [1]. The highest emitting 100 cities, or so-called urban areas, account for 18% of the global carbon footprint [2,3]. Therefore, cities are increasingly focusing on and shaping the trajectory and impacts of climate change and air quality [4][5][6][7][8][9]. The C40 Cities Climate Leadership Group connects more than 90 of the performance are compared for two different technology development scenarios and two European climate zones for five years (2012-2016). Analyzing the system over five years will give insight into the inter-annual variability of the cost and energy performance. To our knowledge, no such comprehensive study has been performed up to now. Many studies and pilot projects investigate stand-alone and national grid-connected renewable energy systems using hydrogen as energy storage and stationary fuel cells for the reconversion of the stored hydrogen [74][75][76][77][78]. Some studies use the produced hydrogen for transportation [79][80][81][82][83][84] or solely use the fuel cell in the vehicle as an electric generator [85][86][87][88][89] without considering hydrogen production. Integration of FCEVs through V2G into a local electricity network for operating in island mode, emergency power, or balancing local renewables has been done mostly on a smaller or a very large scale [90][91][92][93][94]. Some studies include a cost analysis [95][96][97], do not compare with a future scenario with improved cost and efficiency (scenario and trend analysis) [98], are dependent on the grid electricity, do not compare different climate zones nor include inter-annual variability [99], or do not include seasonal hydrogen storage [98]. The authors of [100] focus on a small-scale system in a specific region without considering hydrogen transportation, although includes a future cost scenario. The authors of [101] look into urban areas and road transportation in different regions in different Japanese climate zones, but the described system is not 100% renewable and does not include economics or consider V2G electricity services with FCEVs. A study [102] performs a future techno-economic 100% renewable energy analysis, including multi-annual variability for multiple large national and trans-national regions. Various energy sectors are coupled, where hydrogen is used as energy storage and road transportation fuel along with several other energy carriers and storage techniques. However, here too, but also here V2G electricity services with FCEVs are not considered.

Approach
The techno-economic scenario analysis of a fully autonomous renewable and reliable integrated transportation and energy system for a smart city area is performed in four steps: 1. Location selection, system design and dimensioning, technological and economic characterization for the system components in two technology development scenarios (Section 2.2). 2. Developing a calculation model for hourly simulation of all energy flows for multiple years and sizing of system components, for two different European climates zones in two technology development scenarios (Section 2.3). 3. Calculating the cost of energy for the two technology development scenarios in two climate zones based on the sizing and economic characterization of the system components (Section 2.4). 4. Inter-annual variability analysis of wind and solar energy production on the cost of energy (Section 2.5).

Location Selection
The following criteria apply to the selection of two locations in different European climate zones. They are listed in order of significance ( Figure 1): 1. Close to a large European functional urban area [12] or city with at least 50,000 inhabitants, preferably in one of Europe's five most populous countries [103]. 2. Located in different European climate zones, as defined by the Köppen-Geiger climate classification [44] and supplement of [44]. 3. Located in a region with underground salt formations suitable for underground gas storage [104]. 4. One location should have a relatively high, and one location should have a relatively low solar global irradiation compared to European measurements [59,60,62]. 5. One location should have a relatively low annual precipitation compared to European measurements [105]. 6. All required statistical and hourly modeling data should be available for the selected locations (wind velocity, solar irradiation, precipitation, building energy consumption, etc.).
The urban area of Hamburg in Germany and Murcia in Spain were selected, see Figure 1. Hamburg is the cooler, windier, and rainier area; Murcia is the warmer, sunnier, and dryer area. In Appendix A.1, Table A1 shows key figures characterizing Hamburg in Germany and Murcia in Spain and their climates.

System Design and Dimensioning
The smart city area energy and transportation system is designed in such a way that it fulfills the following design requirements:  uses only electricity and hydrogen as energy carriers and is all-electric in end-use  uses only hydrogen as seasonal energy storage and fuel to power all road vehicles  can be applied to an average European city area and is a scalable design  can be operated in a network of multiple smart city areas and renewable hydrogen and electric energy hubs or centers [32,[106][107][108][109][110]]  can be integrated into existing infrastructure and buildings  is not dependent on an in-urban area underground hydrogen pipeline transportation network  uses abundant renewable energy sources in Europe: local solar and large-scale wind only  is independent of high and medium voltage electricity grids, natural gas, and district heating grids or the expansion of these.
By applying the design requirements, the integrated system design of the smart city area has the following seven major elements and functional energy performance and conversion steps ( Figure 2 and Table 1): 1. Local solar electricity and hydrogen production (orange): Local rooftop solar electricity and rainwater collection, purification, and storage systems (S1-S3) produce solar electricity (ES) and pure water (H2OS). A part of the solar electricity is directly consumed (EDC) in buildings and other subsystems. The remaining surplus solar electricity (ES) is used with purified water (H2OS) in the hydrogen production, purification, and compression system (S4-S6) for filling tube trailers (TT1) with hydrogen (HS). 2. Fuel cell electric vehicle-to-grid, building electricity consumption, and smart grid control (yellow): The smart electric grid is managed by a controller, which connects all buildings, grid-connected FCEVs (FCEV1and2), the hydrogen fueling station (HFS1-HFS4), solar electricity and hydrogen production (S1-S6), and the tube trailer filling station (SHS2) at the seasonal hydrogen storage (SHS1). The directly consumed solar electricity (EDC) is divided amongst the all-electric residential and services sector buildings (EB), HFS (EHFS), and SHS (ESHS) electricity consumption. Any shortage of electricity is met by the electricity produced from hydrogen (EV2G) through parked (at home or in public or commercial spaces) and V2G connected FCEVs (FCEV1and2). 3. Hydrogen tube trailer transportation (grey): Tube trailers (TT1) towed by tube trailer tractors (TT2) transport hydrogen from either the local solar hydrogen production or the SHS to the HFS, or from the local solar hydrogen production to the SHS. 4. Hydrogen fueling station (blue): Hydrogen from tube trailers is further compressed (HFS1) and stored at high pressure (HFS2). A chiller (HFS3) cools the dispensed hydrogen (HHFS), including sufficient dispensers (HFS4) to provide hydrogen for both road transportation (Hroad) and V2G (HV2G) use. 5. Road transportation (purple): A fleet of road transportation FCEVs, namely passenger cars, vans, buses, trucks, and tractor-trailers. 6. Large-scale and shared wind hydrogen production (green): A large-scale wind turbine park (W1) that is not located near or in smart city areas is shared with other smart city areas and renewable hydrogen hubs and consumers. All wind electricity (EW) is used with purified water (H2OW) from local surface water or seawater in hydrogen production (W4), purification (W5), and compression system (W6), which includes a water collection and purification system (W2 and W3). The hydrogen produced (HW) is stored in a large-scale underground seasonal hydrogen storage (SHS1). 7. Large-scale and shared seasonal hydrogen storage (red): Large-scale underground seasonal hydrogen storage (SHS1), including a tube trailer filling and emptying station (SHS2).
The system design configuration is sufficiently flexible to allow other renewable energy sources, if present, to be used (e.g., offshore wind, biomass, or hydropower). However, this was not analyzed in this study. The smart urban area operates in a network of multiple smart urban areas, hydrogen fueling stations, other renewable hydrogen and electric energy hubs, and other hydrogen and electricity consumers (not part of this study). Hydrogen is produced within the smart urban areas from local surplus solar electricity and at large-scale wind parks. These large-scale wind parks, as well as the large-scale seasonal underground hydrogen storage, are jointly owned by the smart urban areas and other hydrogen consumers. Hydrogen is transported via tube trailers from the smart urban areas to hydrogen fueling stations, or the large-scale and shared underground seasonal hydrogen storage [104,111].  Water for hydrogen production via wind H2OS Water for hydrogen production via solar The size of a Hamburg-or Murcia-based illustrative smart city area for this study was determined using the dispersion of supermarkets and gas stations in Europe, Germany, and Spain. In the EU 28 countries, for every 2000 households, there is one medium-sized supermarket and one gas station [55,[112][113][114]. In Germany and Spain, there is one gas station per 2600 and 1700 households, respectively [55,113,114]. Thus, 2000 households are a good indicator for dimensioning the smart integrated city area; see Table 2 (common parameters). This hydrogen fueling station will serve a similar vehicle population as current gasoline stations [115,116]. Total capital cost per capacity for large HFS (≥1500 kg/day) is lower than for smaller HFS [117], thus also defining the minimum size of this scalable and illustrative smart city area.
The floor area of residential and services buildings is derived from national statistical data and scaled to 2000 households: German and Spanish average household floor area Shh is, respectively, 91.60 and 91.78 m 2 [54,55]. Residential and service sector roofs will be used for solar electricity systems and rainwater collection [121][122][123][124]. Solar electricity systems are installed on all technically suitable roof areas: 9 m 2 per person on residential buildings and 4 m 2 per person on service sector buildings area [125,126]. Facçades are not considered.
For ease of comparison between Hamburg and Murcia, the roof area available for solar electric modules and rainwater collection in Murcia is based on the Hamburg parameters.  [125,126] 56,000 56,000 7 1 Assuming that only one household lives in a dwelling. 2 German data [119,120] defines a van as a vehicle with a weight of less than 3.5 tons; the Odyssee database [57] used for Spain defines a van as a vehicle with a weight of less than 3 tons. 3 Including commercial vehicles of 3.5-6.0 tons. 4 No distinction is made between trucks and tractor-trailers in [57]; therefore, the same relation between the number of trucks and tractor-trailers as in Germany is used. 5 Based on the surface area of permanently occupied dwellings [54]. 6 The floor area represents the floor space that needs to be heated, cooled, or illuminated [127]. 7 For ease of comparison, the value is kept equal to the Hamburg case.

Technological and Economic Characterization of System Components in Two Scenarios
The technological and economic characteristics of the selected components will be listed according to the latest available figures in two technology development scenarios. The two scenarios, in different time frames, can be characterized as follows:


The Near Future scenario uses current state-of-the-art renewable and hydrogen technology and current energy demand for buildings and transportation. It is an all-electric energy system, which means space heating is done using heat pumps, meeting the present heat demand for houses and buildings. Only commercially available hydrogen technologies are used. For all systems, including hydrogen technologies, current technology characteristics and cost figures are used. The Near Future scenario presents a system that could be implemented in 2020-2025.  In the Mid Century scenario, a significant reduction in end-use energy consumption is assumed. Hydrogen and fuel cell technologies have become mature with mass production and performing on the cost and efficiency targets projected for 2050. Also, for all the other technologies, such as solar, wind, and electrolyzers, the learning curves are taken into account.
The detailed technical and cost-related parameters of the system components are presented in Appendix A.2 Table A2 and Table A3. The technology selection for the system components and sizing methods is based on the component description in [36]. Figure 3 shows the simplified simulation scheme of the calculation model, consisting of five major steps that are executed hourly for an entire year. A detailed description and input data are described in Appendix B, Table A4, Table A5 and Table A6.  Two sets of energy balances are calculated, on both an hourly and an annual basis ( Figure 3 in red and green) for both hydrogen and electricity energy carriers. Energy consumption takes place in buildings and for mobility. Energy production is by roof-top solar and wind turbines and covers all energy consumption needs, taking into account all efficiencies of the different energy conversion and storage processes.

Calculation Model and Hourly Simulation
The amount of rooftop area available for solar electricity systems is fixed in both scenarios and locations for ease of comparison of the system performance between the two climates. The amount of installed wind capacity is the degree of freedom in the calculation model and completes the annual energy balance.
The system is simulated for five years using weather data from 2012 to 2016, which results in varying hourly electricity production consumption profiles, as well as electricity production per installed capacity. For ease of comparison between the years, the annual building electricity demand is kept constant.

Calculating the Cost of Energy
Three components of the cost of energy (CoE) will be calculated for each location in both scenarios.
1. Smart city area total system cost of energy (TSCoESCA) in euros per year (Appendix C.1). 2. System levelized cost of energy for electricity (SLCoEe) in euros per kWh and for hydrogen (SLCoEH) in euros per kg of hydrogen (Appendix C2). 3. Cost of energy for households (CoEhh) in euros per household per year (Appendix C.3).

Smart City Area Total System Cost of Energy
The TSCoESCA in euros per year is the sum of the total annual capital and operation and maintenance costs TCi (€/year) of the total number of components (n) in the smart city area. The TCi of an individual component is calculated using the annual capital cost CCi (€/year) and operation and maintenance cost OMCi (€/year); cost formulas used are listed in Appendix C.1.
The cost analyses are in constant 2015 euros. An exchange rate of 0.88 USD to 1 EUR is used as in [36]. The website [128] is used to convert all USD values to USD2015 values. A weighted average cost of capital WACC of 3% is used from Appendix A of [102].

System Levelized Cost of Energy
The system levelized cost of energy, for either electricity SLCoEe (€/kWh) or hydrogen SLCoEH (€/kg H2), is calculated by allocating a share of the TSCoESCA (€/year) related to either electricity TSCoESCA,e (€/year) or hydrogen consumption TSCoESCA,H (€/year). These shares are then divided by either the annual electricity consumption ECe (kWh/year) or the annual hydrogen consumption ECH (kg H2/year), resulting in, respectively, the SLCoEe (€/kWh) or the SLCoEH (€/kg H2). The cost formulas used are listed in Appendix C.2. The key energy balance parameters and annual energy balances of the years 2012-2015 show similar outcomes. Several major trends can be seen when looking at the FCEV2G, wind and solar electricity production, direct consumption of solar electricity, and seasonal hydrogen storage. The 87% higher solar electricity output in the Mid Century scenario in both locations results in less required external wind-to-hydrogen production to close the energy balance. This, together with more than a 30% reduction in building and road transportation energy consumption, and improvements in energy conversion processes, results in reductions of 70% and 90% of wind electricity production for, respectively, Hamburg and Murcia.  The 490% higher solar hydrogen production in the Mid Century scenario in both locations compared to the Near Future scenario. Due to lower building electricity consumption and higher solar electricity production, there is more solar surplus electricity for hydrogen production. In Hamburg, solar electrolyzer power consumption always peaks in the summer's time, whereas, in Murcia, solar electrolyzer power consumption peaks in winter ( Figure A2).  The 40% and 56% higher coverage of electricity consumption with direct solar electricity production in the Mid Century scenario in, respectively, Hamburg and Murcia compared to the Near Future scenario. Due to higher solar radiation and lower building and system electricity consumption, a higher percentage can be met directly with solar electricity. Nighttime electricity consumption has to be met with FCEV2G electricity production.  The 15%-25% lower seasonal hydrogen storage requirements in the Mid Century scenario due to a better match of higher solar electricity production and lower building electricity demand compared to the Near Future scenario. For Hamburg, the maximum storage content of hydrogen occurs in the fall for both scenarios, whereas, in Murcia, this period shifts from spring to fall. The minimum storage content occurs in winter for both locations and scenarios. In the Mid Century scenario, a typical salt cavern [104] (Table A3) could serve approximately 23 similarly operating smart city areas in Hamburg and 40 Murcia smart city areas.  The 40% lower seasonal hydrogen storage and FCEV2G requirements in Murcia compared to Hamburg, in all scenarios. In the Mid Century scenario, solar electricity alone is almost able to supply all of Murcia's energy needs for buildings and road transportation (despite its 21% higher consumption of road transportation hydrogen; Appendix B.2). If approximately 15% more solar panels were to be installed, either on facades, in public spaces, or nearby solar fields, the entire energy demand could be met with solar energy. The reason for the lower SHS and FCEV2G requirements in Murcia compared to Hamburg is the better match in time (daily and seasonal) between solar electricity production and building electricity consumption. In addition, Murcia also has a relatively higher solar electricity output and lower building demand compared to Hamburg. In the Mid Century scenario in Murcia, the same solar system produces 73% more electricity than in Hamburg.  Relatively, 70% and 30% more seasonal hydrogen storage is needed in the Mid Century scenario for, respectively, Hamburg and Murcia. Even though absolute hydrogen and electricity production, energy consumption, and seasonal hydrogen storage decrease in the Mid Century scenario, the higher dependency on solar electricity production increases the seasonal effect. Hence, there is an increase in relative seasonal hydrogen storage compared to the annual hydrogen and electricity production in the Mid Century scenario.

FCEV2G Usage and Electricity Balance Discussion and Results
Figure 6 provides further insight into seasonal and hourly FCEV2G usage. The FCEVs needed for producing V2G electricity (# cars left y-axis, % of car fleet right y-axis) are shown by means of boxplots for every hour of the day. For both locations and scenarios, usage is shown separately for both the colder winter period (in blue, left, 1 October-31 March) and the warmer summer period (in orange, right, 1 April-30 September).


Reliable electricity supply can be realized at all hours of the day, as extreme FCEV2G peaks never exceed 50% of the total car fleet. The number of cars needed to balance the system peaks in the morning (06:00-09:00) and the late afternoon/early evening (16:00-20:00) and correspond to driving rush hours. These peaks are extreme outliers, and values close to these occur for only a small number of hours per year ( Figure A1).  In Murcia, virtually no cars are required during daylight hours. This is valid in all scenarios and seasons, except for some moments. In Hamburg, this is only the case in the summer period, for both scenarios.  Hamburg faces a greater seasonal, and Murcia a greater day-night storage challenge, particularly in the Mid Century scenario. In Hamburg, peak FCEV2G electricity production occurs in the winter period, whereas, in Murcia, the production is highest in both the summer and the winter period (see also Figure A2).

Total System Cost of Energy
The total system cost of energy per year TSCoE (k€/year) in the Near Future and Mid Century scenarios for Hamburg and Murcia is shown in Figure 7. The subsystems are grouped into hydrogen and electricity. The average component installed capacities and their total annual costs (TCi) are listed in Appendix E Table A7 and serve as input for Figure 7. The following major trends can be observed when comparing both locations and scenarios.


The  Table  A7 in Appendix E). This can be explained by the variations in daily and annual wind and solar electricity production, as well as the varying mismatch between solar electricity production and consumption. Seasonal hydrogen storage has relatively higher cost variations (8-12%) in comparison to other components, as the SHS is responsible for coping with all the abovementioned variations.  The cost of hydrogen components in the Mid Century scenario drops up to 75%. For both locations, in the Near Future scenario, the hydrogen components represent about 70% of the TSCoE; this reduces to 63% on average. As hydrogen technology is relatively new, economies of scale have a bigger impact on future cost reductions than on solar and wind electricity technology. In addition, the increase in solar output reduces storage requirements.  Hydrogen transportation, seasonal hydrogen storage, and the solar system are the only components that share in the total costs' relative increase compared to all other components. This is because the cost reductions for these components are relatively lower compared to the other components. The relatively higher use of seasonal hydrogen storage in the Mid Century scenario compared to the annual hydrogen production (see Section 3.1) is another contributing factor.

System Levelized Cost of Energy
The levelized and system levelized cost of electricity and hydrogen for Hamburg and Murcia in the Near Future and the Mid Century scenario are listed in Table 4. The values represent the average of the five simulated years. The levelized cost of energy (LCoE) and SLCoE parameters are calculated using the total costs (TCi, Appendix E) of the various components and the corresponding energy flows (Figures 4 and 5). Detailed calculation methods can be found in Appendix C and [36].


The system levelized cost of energy of electricity (SLCoEe) is 239  The levelized cost of energy of hydrogen (LCoEH,W&S) before transportation and storage is based on hydrogen from both wind and solar. Even in Murcia, in the Mid Century scenario, the cost of solar electricity (LCoEe,S) is lower than the cost of wind electricity LCoEe,W. The reason for this is that a significantly higher capacity factor is achieved when the electrolyzer is connected to the wind turbine than to the solar electricity system, which only uses surplus solar electricity peaks.


The system levelized cost of energy of hydrogen (SLCoEH) is 70-80% higher than the combined levelized cost of energy of hydrogen from solar and wind (LCoEH,W&S). The SLCoEH includes the costs of hydrogen transportation by tube trailers, seasonal and fueling station storage, and dispensing on top of the solar and wind electricity costs, and the electrolyzers and low-pressure compressors, which is only the case for the LCoEH,W&S.

LCoE and SLCoE Comparison with Other Studies
Studying "100% renewable energy systems" is relatively new [130], and no integrated transportation and energy systems are the same. Comparing the SLCOEe with other 100% renewable energy systems should be taken as a general indication since there are many differences; for example, differences in geographical locations, renewable energy sources, energy carriers, storage technologies, and simulation criteria, such as energy self-sufficiency ratios or cost input parameters. Despite such differences, we can, to a certain extent, compare some subsystem costs, onshore wind and solar electricity, stored and dispensed hydrogen, and all-time available system electricity costs, including daily and seasonal storage.


Onshore wind electricity costs (LCoEe,W) are relatively low in comparison with other studies. Near Future scenario 24-27 €/MWh compared to 30-50 €/MWh for 2025 [131], and Mid Century scenario 16-18 €/MWh with 20-35 €/MWh for 2050 [131]. There are three reasons for this. First, the exclusion of grid connection costs of 11.5% [132,133] in this study, because of the direct coupling between the wind turbine and the electrolyzer. Second, the use of a lower WACC (3%) compared to other studies (3.5-10%) [131]. Third, the placement of wind turbines on sites with good wind conditions, resulting in good onshore wind capacity factors (33-38%  [36]. Several hydrogen electricity reconversion pathways in the north of Germany have been designed and evaluated for the year 2050, including underground seasonal hydrogen storage [141]. The study reports higher values of 176-247 €/MWh, although it confirms that the costs are dominated in all pathways by the costs of purchasing electricity [141]. The authors of [102] and [142] report similar values of 75-85 €/MWh and 100 €/MWh for 100% renewable and self-sufficient energy systems in 2050. Although they have similar system electricity costs, there are several differences: [102] and [142] use different storage technologies simultaneously, include more sectors (industry, agriculture, fishing, and forestry) and renewable energy sources, and either simulate for entire countries (Germany and Spain) [102] or cities in a different continent (North America) [142].

Cost of Energy for Households (Without Taxes and Levies)
Total system costs or system levelized energy costs do not represent the combined effect of energy-saving measures, higher efficiencies, and decreased costs. Therefore, the cost of energy for an average household CoEhh (€/hh/year) is introduced as an example. To put the designed system into perspective, a comparison with today's household energy costs would be interesting to make. This, however, is not as straightforward as it seems.
The developed system and the technologies used are very different from today's fossil-based energy and transportation system. Cities today are not self-sufficient: They import energy from both the national and the international power and fuel network. These national and international electricity and fuel supply chain networks also come at a cost. This, however, falls outside the scope of this study.
The analyzed size of this system is very small; one could compare it to a neighborhood within these big urban areas or a very small village. In addition, only the building and the road transportation sector are analyzed and integrated here. Increasing the system size and combining several different sectors would create more integration opportunities and reduce costs. For example, the equipment could be shared to avoid underutilization.
Environmental and health savings and welfare creation (e.g., jobs) [143] compared to the present fossil system are difficult to express in costs for this specific and small-scale system. In the present situation, taxes and levies on energy can represent a great part of the energy costs for household consumers, but future estimates of taxes and levies are not within the scope of this study.
Summarizing, it is very difficult to make a fair cost comparison. Nevertheless, a very simple energy cost comparison for an average household is shown below, without any taxes or levies. The present fossil situation is compared with the designed 100% renewable system in the Near Future and the Mid Century scenarios. Additional background data for the present situation can be found in Appendix F.
The cost of energy for a single household CoEhh (€/hh/year) consists of the cost of energy for the building energy CoEhh,B (€/hh/year) and the transportation energy CoEhh,T (€/hh/year); see Table 5.
The Near Future scenario CoEhh shows an increase compared to the present situation, although not by several magnitudes. For Murcia, the increase is only 30% in the Near Future scenario. This shows that even though new hydrogen technologies are used, Near Future scenario costs can come close to the present situation costs and thus give reason to explore further. We should bear in mind that the Near Future scenario only changes technologies (e.g., electric water heating and heat pumps for heating) and has no significant energy savings as in the Mid Century scenario. However, in reality, the installation of a heat pump often goes hand in hand with energy-saving measures like insulation. What's more, any further integration with other sectors and increasing the system size could also further reduce costs.
The cost of energy for households (without taxes and levies) in the Mid Century scenario is significantly lower (up to 65%) compared to the present situation-namely 770 and 520 €/year per household for Hamburg and Murcia, respectively. Therefore, the designed system is not only renewable and reliable but also affordable.

Discussion
The designed and analyzed integrated transportation and energy system is an extreme hypothetical scenario, because: 1. The city area is not connected to any national electricity or natural gas grid or a transportation fuel network. It is self-sufficient and stand-alone. 2. Only the residential, services, and road transportation sectors have been taken into account as energy consumers (e.g., not industry, agriculture, rail, or air transportation sectors).
3. Space heating and hot water production are all-electric. 4. It uses a single set of technologies for road transportation, transportation fuel, energy storage, and balancing, namely hydrogen, hydrogen production, and fuel cells (FCEVs), (no batteries or Battery Electric Vehicles, BEVs). 5. The city area is relatively small, based on approximately 5000 people.
In the future, a mix of multiple energy carriers, storage methods, and energy technologies could all work together. Cities in Europe already have connections to national electricity and sometimes natural gas grids. In addition, all sectors should be considered, not only the residential, services, and road transportation sectors. Increasing the system size and combining several different sectors would create more integration opportunities and could reduce costs.
However, the calculated energy costs of the designed system are affordable and in line with other studies. This gives reason to explore whether variations in system designs and balancing methods can reduce total system costs even further. The system designs and balancing methods discussed below are a non-exhaustive selection of possible options.


A national electricity grid connection would make it possible to import electricity or export peaks of solar electricity to other cities or electricity consumers in different sectors, such as industry, for example, by importing lower-cost onshore or offshore wind electricity during periods of insufficient solar electricity production (e.g., at night). This would reduce the need for hydrogen storage and FCEV2G electricity. High solar output at midday in the Mid Century scenario results in high surplus peaks to be absorbed by the solar electrolyzer. Exporting these high peaks of solar electricity to, for example, industrial cooling warehouses would reduce solar electrolyzer installed capacity and costs. Using only one electrolyzer connected to the national grid and placed next to the hydrogen station could reduce hydrogen transportation. Smart placement of electrolyzers in the electricity grid could obviate electricity grid congestion and reduce or avoid the need for expensive capacity expansion [144].  A hydrogen pipeline network [32,[145][146][147][148][149] could reduce hydrogen transportation via tube trailers and fueling station capacity. Multiple electrolyzers and hydrogen fueling stations could be interconnected via a pipeline network [150]. In this way, tube trailer hydrogen transportation could be replaced, and hydrogen transportation costs reduced. Furthermore, the partial recompression of hydrogen when emptying a tube trailer could also be reduced or avoided altogether. The compressor could even be omitted, provided the electrolyzer hydrogen output pressure is higher than the pipeline pressure. In the case of parked FCEVs delivering V2G electricity, the fuel cell could be connected directly to the hydrogen distribution pipeline network, instead of using hydrogen from the on-board hydrogen tank [151]. Not using hydrogen from the 700 bar tank eliminates the need for refueling for V2G purposes, which in turn reduces the required capacity of hydrogen fueling stations.  Import of low-cost renewable hydrogen could partially replace, possibly costlier, local hydrogen production and seasonal hydrogen storage, and thus total system costs. Locally and at certain times of the year, there could be insufficient solar and onshore wind sources available to produce hydrogen. Regions with abundant and low-cost hydro, solar, or wind power [152][153][154][155][156][157][158] could produce lowcost hydrogen for export. This hydrogen could be imported at demand centers instead of being produced and stored on-site. Several ideas already exist, for example, producing hydrogen (far) offshore [159] from fixed or floating wind [32,[160][161][162] and solar structures [163,164], or wave energy [165] and bringing the hydrogen onshore via existing natural gas or newly built pipelines [32] or ships [166,167]. The onshore pipeline network would then distribute the hydrogen to the consumers.  Using a lower-cost mix of renewable energy sources. In this study, the rooftop solar surface area was kept equal in both locations, even though solar electricity is more expensive in Hamburg than in Murcia. Therefore, using the lowest cost renewable energy source locally available could reduce total system costs even further. For example, hydropower, offshore wind, biomass, concentrated solar power, by-product hydrogen, or tidal or wave energy could result in lowercost electricity than onshore wind or solar Photovoltaic (PV).  Tailor electricity mix and its supply pattern to local demand. In Murcia, solar electricity production has a better time match with electricity consumption on both a daily and a seasonal basis. During the day, solar electricity production in summer aligns well with electricity demand in buildings for space cooling. Therefore, a lower total system cost can be achieved by tailoring the renewable energy mix to allow for a better match between the production pattern and the demand pattern [61,63,65,102,142,[168][169][170][171][172][173][174][175][176][177]. This would result in lower hydrogen production, storage, transportation, fueling, and FCEV electricity production costs.

Other Balancing Methods
 Using a mix of FCEVs, BEVs, or fuel cell plug-in hybrid electric vehicle (FCPHEV) and stationary batteries [84,87,[178][179][180]. Instead of only using hydrogen and FCEV2G for both daily and seasonal energy balancing, other technologies could be used in parallel. For example, batteries in BEVs or FCPHEV, as well as stationary batteries, could be used for storing or releasing peak surplus or shortage of electricity [181] for day-to-day storage. Especially in Murcia, this could result in lower total system costs, as the day-to-day storage is more prevalent in Murcia [182]. Capacity factors of electrolyzers could be improved, and so decrease costs. FCEVs and hydrogen production and storage could subsequently be used for energy balancing for longer periods, up to entire seasons [182].  Using other CO2-free hydrogen carriers for energy transportation, short and long-term energy storage.
There are several other proven and available carriers today, such as liquefied hydrogen [183][184][185][186][187], ammonia (NH3) [188], or liquid organic hydrogen carriers (LOHC) [189,190]. Transporting liquid hydrogen can be less costly compared to compressed hydrogen when volumes and distances are larger. Ammonia storage and LOHC storage are becoming commercial applications at scale, and both represent reasonable alternatives in the absence of salt caverns.  Increase passenger car FCEV2G power output, use other FCEVs and stationary fuel cells for combined heat and power. At the moment, only passenger cars with an output of 10 kW/car while having a 100 kW fuel cell system on-board are used for FCEV2G electricity. This limitation is mainly because of potential insufficient cooling radiator capacity when parked and providing FCEV2G electricity [38]. If V2G output could be increased by enhancing cooling capacity, then proportionally fewer passenger cars would be needed. Cooling capacity could be enhanced by installing, for example, a bigger radiator and cooling fans, or by using two-phase cooling fluids with a higher cooling capacity [191]. Commercial vehicles (vans, trucks, buses) are more widely used than passenger cars, although often not during the night. By also using commercial for V2G purposes [192], the number of passenger cars would be reduced. In the case of an underground hydrogen pipeline network, stationary fuel cells [193][194][195][196][197][198] could provide heat and power to buildings, and when necessary, FCEV2Gs could provide peak power.  Internet Technology (IT) usage for demand response forecasting, scheduling, virtual power plants, and autonomous driving. Weather and electricity demand forecasting [199][200][201][202][203][204][205][206][207][208] in combination with demand response [21,26,[209][210][211] could potentially avert peaks in temporal surplus or shortage of electricity. This would reduce installed capacity cost. Combining the output of thousands of grid-connected FCEVs would create so-called virtual power plants [212,213] with potentially large capacities. Similar to mobility as a service (MaaS) [214][215][216][217][218], power or electricity as a service (PaaS or EaaS) could be introduced. To create these markets, additional pricing structures, contract types, and aggregators scheduling and operating the cars will be required [219][220][221][222].
Upcoming technologies could facilitate the scheduling of cars, for example, self-driving, freefloating, cloud-connected car-sharing fleets [223][224][225], together with inductive (wireless) selfconnecting V2G infrastructure [226][227][228][229][230]. As mentioned earlier, most FCEV2G electricity is required at night, whereas most people travel and work during the day. So, even if car-sharing spreads widely and the total number of cars decreases, at night, car-sharing fleets will be used less and, therefore, will be available to provide power.

Conclusions
The designed and modeled system for smart urban areas is based on wind, solar, and hydrogen, where fuel cell electric vehicles provide year-round 100% renewable, reliable, and affordable energy for power, heat, and transportation in two different European climates.
The two locations in different climate zones-namely Murcia in Spain and Hamburg in Germany-were selected based on several criteria. Both are close to or in a large European urban area in one of the five most populated countries. Located in a different climate zone according to the Köppen-Geiger classification, Hamburg has a temperate oceanic climate (Cfb), and Murcia a hot semi-arid climate (BSh). Both locations have salt formations suitable for underground hydrogen gas storage. One location has a high level of solar radiation (Murcia), while the other has a low level (Hamburg).
The two designed smart city areas have the climate characteristics of Hamburg and Murcia; the dimensions are based on, respectively, German and Spanish statistical data. The smart city areas consist of 185,000 m 2 floor area of residential sector buildings, and for Hamburg and Murcia, respectively, 93,000 m 2 and 38,000 m 2 floor area services sector buildings. Hamburg and Murcia have, respectively, a total of 2500 and 2250 fuel cell electric vehicles (FCEVs), of which 2360 and 1850 are passenger cars that can be used for producing electricity via vehicle-to-grid (V2G), so-called fuel cell electric vehicle to grid (FCEV2G). Two thousand households with a total of approximately 4300-5000 inhabitants are the minimum viable economic size for dimensioning the smart city area, as statistically, there is one gas station and one retail food shop per 2000 households. Smaller capacity fueling stations are relatively costlier.
The designed smart city area system is 100% renewable. All electricity and hydrogen can be supplied by solar and wind to fulfill the energy demand for power, heat, and transportation. The transportation and energy sectors are fully integrated, and their final energy use is all-electric. Electricity is generated by solar modules on all roofs. Surplus solar electricity is converted via waterelectrolysis with rainwater into pure hydrogen. The hydrogen is compressed and transported by tube trailer modules to the nearby hydrogen fueling station (HFS) or underground seasonal hydrogen storage (SHS). At the HFS, the hydrogen is further compressed to fuel all types of FCEVs, from passenger cars, vans, buses to trucks, and tractor-trailers. In the case of a temporary shortage of solar electricity, the fuel cells in the parked and grid-connected passenger cars provide the necessary electricity (FCEV2G) by converting hydrogen from the on-board hydrogen storage tanks. At parking places at home, the office, or the local shopping area, vehicle-to-grid points connect the cars to the smart city electrical grid. The SHS is filled with hydrogen from surplus solar electricity and via a very short pipeline with hydrogen produced from wind electricity from an onshore wind turbine park. The stored hydrogen in the SHS is transported via tube trailers to the hydrogen fueling station.
The designed smart city area system is reliable at all times and independent of other energy systems and grid connections. The energy balance is simulated on an hourly basis for an entire year for a Near Future and a Mid Century scenario. Five years are simulated using climate data input from the years 2012-2016, although no significant differences in the energy balance are observed.
Reliable electricity supply can be realized at all times, as extreme peaks in the FCEV2G electricity supply never exceed 50% of the car fleet. The designed smart city area system is affordable, and further cost reductions are possible. It is very difficult to make a fair cost comparison between today's energy system and the one proposed in this study in the Mid Century scenario. Nevertheless, a very simple energy cost comparison for an average household shows that the cost of energy for households (without taxes and levies) in the Mid Century scenario can be 65% lower compared to the present situation-namely 770 and 520 €/year per household for, respectively, Hamburg and Murcia.
The developed system and the technologies used are very different from today's fossil-based energy and transportation system. The designed and analyzed integrated transportation and energy system is an extreme hypothetical scenario because: 1. The city area is not connected to any national grid; it is self-sufficient and stand-alone. 2. Only the residential, services, and road transportation sectors have been taken into account as energy consumers. 3. Space heating and hot water production are all-electric. 4. It uses a single set of technologies for road transportation, transportation fuel, energy storage, and balancing; hydrogen, hydrogen production, and fuel cells (FCEVs and no batteries or BEVs). 5. The city area is relatively small, based on approximately 5000 persons.
Increasing the system size and combining several different sectors would create more integration opportunities and could reduce costs. Environmental and health savings and welfare creation (e.g., jobs) compared to the present fossil system are difficult to express in costs for this specific and small-scale system. In the future, multiple energy carriers, storage methods, and energy technologies could work in parallel. The system levelized costs in the Mid Century scenario are 71-104 €/MWh for electricity and 2.6-3.0 €/kg for hydrogen. These results compare favorably with other studies describing fully renewable power, heat, and transportation systems. This gives reason to explore whether variations in system designs and balancing methods and technologies can further reduce total system costs.  Acknowledgments: We thank the Agencia Estatal de Meteorología (AEMET) for providing detailed weather data for multiple years and locations in Spain, which was essential for running the simulations.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results. Table A1 shows some key figures characterizing Hamburg in Germany and Murcia in Spain and their climates. Respectively, 1.8 and 1.5 million inhabitants live in urban areas [231][232][233]. Hamburg has a temperate oceanic (Cfb), and Murcia a hot semi-arid (BSh) climate, according to the Köppen-Geiger climate classification [44,[234][235][236][237]. Local weather station data from 2012 to 2016 [238][239][240] are used to calculate the five-year average (μ) and annual coefficient of variation (CV, also known as the relative standard variation) of the average annual wind speed, solar global horizontal irradiation, precipitation, air temperature, daily maximum and minimum air temperatures, heating degree days (HDD) [127,241], and cooling degree days (CDD) [50,51].  [44,[234][235][236][237] temperate oceanic (Cfb) hot semi-arid (BSh) Weather station data Weather station height above sea-level (m) 1 [238,239] 11 61 Weather station location 1 [238,239] 53°38′ N, 9°59′ E 38°0′ N, 1°10′ W Weather data 2012-2016 means and standard deviations μ (CV) μ (CV) Wind speed at 10 m above ground (m/s) 1 [238,242] 4.1 (4.3%) 3.9 (4.3%) Solar global horizontal irradiation (kWh/m 2 /year) [238,240] 1020 (4%) 1855 (1.8%) Precipitation (l/m 2 /year) [238,240] 735 (4.9%) 255 (24%) Air temperature (°C) [238,240] 9.9 (5.9%) 19.1 (2.8%) Daily maximum air temperature (°C) [238,240] 13.4 (5.1%) 25.5 (2.2%) Daily minimum air temperature (°C) [238,240] 6.3 (8.7%) 13.7 (4.4%) Heating Degree Days (°C•day/year) 2 [238,240] 3066 (6.5%) 854 (16%) Cooling Degree Days (°C•day/year) 2 [238,240] 101 (24%) 1245 (6.9%) 1 Wind speeds measured at the nearby Almeria Airport weather station are used [242,243] because, at the Murcia weather station [239,240], wind speeds are economically less favorable for wind turbines. The non-wind weather data of the Murcia weather station is more complete than that of the Almeria Airport weather station. The Almeria Airport weather station is 21 m above sea-level and has the following coordinates: 36°50′ N, 2°21′ W. The Murcia weather station five-year average (2012-2016) and coefficient of variation of the average annual wind speed at 10 m above ground are 2.4 m/s and 2.8%, respectively. 2 Calculated with a base temperature of 18 °C as in [50,51,127]. Table A2 lists the specific energy consumption and production (SEC and SEP) (kWhe/kg H2) in the Near Future and Mid Century scenarios for the different energy conversion processes, from rainwater collection to hydrogen production, fueling, and reconversion to electricity. Alkaline water electrolysis technology is chosen as, at the moment, it is cheaper than Proton Exchange Membrane (PEM) electrolysis technology [136,244]. The electrolyzer is coupled directly to the direct current renewable energy source, so there is no AC/DC conversion needed at the electrolyzer [245]. Current state-of-the-art alkaline electrolysis technology can respond sufficiently fast [136,246,247] to shortterm solar and wind power fluctuations [248][249][250][251][252][253][254]. The alkaline water electrolysis part-load efficiency curve is used from [255], and its maximum efficiency point (higher heating value (HHV)-based) is 88.8% [256][257][258] for the Near Future and 92.6% for the Mid Century scenario [258,259]. SEC for hydrogen purification is, respectively, 1.3 and 1.1 kWhe/kg H2 in the Near Future and the Mid Century scenario [260,261]. After purification, drying, and oxygen removal, the hydrogen meets the purity required for FCEVs [262] at a pressure of 15 bar and 30 bar in, respectively, the Near Future and the Mid Century scenario [136,256,257,259]. Compression SEC for all compressors (S6, W6, HFS1, and SHS2) in the Near Future scenario is in the range of 0.5-3.1 kWhe/kg H2, and in the Mid Century scenario 0.4-2.5 kWhe/kg H2 [115,[263][264][265][266]. The reduction in SEC in the Mid Century scenario is due to the increase in isentropic efficiency from 60% to 80% [263] and, for compressors S6 and W6, also the higher hydrogen output pressure of the electrolyzers. The compressor at the wind park (W6) has an outlet pressure of 180 bar, the maximum SHS operating pressure [104], which is assumed to be constant in this study. The operating pressure range of hydrogen tube trailers in both scenarios is 30-500 bar with an effective storage capacity of 1014 kg H2 [136,[267][268][269]. By applying a smart consolidation strategy for emptying and filling tube trailers at the SHS [263,264,270,271], the net electricity consumption is simplified as compressing hydrogen from 180 bar to 500 bar (SHS2). The electricity consumption of compressor S8 is modeled as the compression of hydrogen from the hydrogen purification output pressure of 30 bar to 500 bar. The combined compressor capacity at the HFS (HFS1) is the largest of all compressors and is modeled with a variable inlet pressure of 30-500 bar (emptying the tube trailers) and fixed outlet pressure of 875 bar of the storage (HFS2). Hydrogen cooling SEC for fueling at 700 bar is 0.20 and 0.15 kWhe/kg H2 in, respectively, the Near Future and the Mid Century scenario [177,178], and is assumed to be constant over the entire operating range of the chiller. In this study, no reduction in specific electricity consumption is foreseen in the Mid Century scenario for reverse osmosis [272][273][274]. The specific energy production SEP by the FCEV is, respectively, 20.3 and 23.6 kWhe/kg H2 in the Near Future and the Mid Century scenario. These values correspond to a tank-to-grid efficiency, ηTTG, [37] (analogous to tank-to-wheel, ηTTW, efficiency when driving) of, respectively, 51.5% and 60% (HHV) [275,276]. Furthermore, it is assumed that the SEC and SEP of the conversion processes are location independent. Table A2. Specific energy consumption SEC or production SEP (kWhe/kg H2) of the energy conversion processes in the smart city area for both scenarios and locations. Figure 2 and Table 2  Pure-water tank [121][122][123][124] 120 €/m 3 0.33% 50 120 €/m 3 0.33% 50 1 The size of the electrolyzer at the solar system is smaller than 10 MW and, therefore, still has a higher IC in the Near Future scenario than the wind connected electrolyzer. 2 The OM includes stack replacement [36], which is calculated based on an average of 13 and 24 operating hours per day for, respectively, solar and wind-powered electrolyzers. 3 System LT is 20 and 30 years, and stack economic operating lifetime is 80,000 and 90,000 h in, respectively, the Near Future and the Mid Century scenario. 4 Figures for, respectively, Hamburg/Murcia. The compressor IC costs are dependent on, for example, mass flow capacity, discharge pressure, and pressure ratio [36]. 5 Figures for, respectively, Hamburg/Murcia. The OM is defined by the amount of hydrogen transported [36] and differs per location and scenario. 6 Lifetime of 8 years based on approximately 75,000 km/year; for the cost calculation [36], only the required driven kilometers are included, i.e., there are no "idling" costs. 7 Lifetime = economic lifetime in driving operating hours. The economic lifetime calculation when using the FCEV for both driving and V2G is explained in [36].

B.1. Electricity Consumption and Production
The yellow rectangle in Figure 3 includes the electricity consumption of the services and residential buildings (EB), the HFS compressor and chiller (EHFS), the SHS compressor (ESHS), and the solar electricity production (ES). EHFS and ESHS are calculated by multiplying the hourly hydrogen throughputs by the specific energy consumption component values SEC (kWhe/kg H2) from Table  A2.
The electricity consumption of the services and residential buildings (EB) in the Near Future and the Mid Century scenario is based on the energy consumption at present, called the Present Situation. Therefore, first, the Present annual specific energy consumption of the residential and services buildings SECB (kWh/m 2 /year) is defined for each location. The method described is applicable to any location within Europe.
Building energy consumption is divided into six energy consumption categories: 1. Space heating 2. Space cooling 3. Water heating 4. Cooking 5. Lighting 6. Electrical appliances Space heating and cooling depend on the ambient temperature, which is reflected in the number of annual heating and cooling degree days (HDD and CDD; see Table A1). Hamburg and Murcia differ greatly in this regard. Also, within the respective countries (Spain and Germany), locally, the number of HDDs and CDDs [50,51,308] can differ greatly from the national weighted average [118]. Due to a lack of recent and complete studies on building energy demand relations with respect to climatic parameters, a similar method is developed as in [50,51]. For space heating for both the residential and the services sector, a relation between the specific thermal heating demand and the number of HDDs per country is established using [54,55,118]. The specific thermal heating demand is derived from the used fuel mix, useful thermal energy per fuel type, fuel demand, and floor space [54,55,118,127]. The value used as specific thermal demand for space heating in Murcia and Hamburg is taken from countries with a similar number of HDDs as Murcia and Hamburg. For Murcia, the specific thermal demand is based on the specific thermal heating demand of Cyprus, Malta, and Portugal of the available years 2010-2015. For Hamburg, it is based on the specific thermal demand of Germany, Sweden, Finland, and Lithuania.
For space cooling, the relations between CDDs, specific thermal cooling demand, and specific electricity demand from [50,51] are used.
For water heating, cooking, lighting, and electrical appliances, it is assumed the local consumption in the Murcia and Hamburg regions does not differ from the national average values [54,55]. Table A4 shows the specific annual energy consumption for buildings SECB (kWh/m 2 /year) per energy consumption category, and annual electricity consumption in buildings EB  (Table A5) to the Present SECB. Space heating SECB in the Near Future scenario for both residential and services buildings is a conversion of the Present Situation SEC with its fuel mix [54,55] and corresponding useful thermal energy fractions [127] and a heat pump Coefficient of Performance (COP) of 3.5 [309][310][311][312] into the electrical equivalent energy. For the Mid Century scenario, savings for, respectively, the residential and the services sector of 95% and 85% are achieved, based on [313]. Space cooling SECB in the Near Future remains equivalent to the Present SEC [50,51], whereas, in the Mid Century scenario, savings of 70% are realized for both sectors [309,310,313]. In the Near Future scenario, water heating and cooking SEC is realized by electrification of the Present SEC fuel mix [54,55], with the useful thermal energy fractions [127] into the electrical equivalent energy [127]. Only for water heating, savings of 50% are used in the Mid Century [313], due to the combined application of electrification, heat pump usage, solar thermal heating, and other heat recuperation techniques. By extensive use of LED technology for lighting and LED efficiency increase, savings of 20% and 80% are assumed for, respectively, the Near Future and the Mid Century scenario in both sectors [314][315][316]. A total of 0% of net savings are assumed for the SECB of electrical appliances. Although energy savings will be significant, the net savings will be zero due to an increased number and use of electrical appliances, home automation, and IT services [317][318][319].
Hourly profiles for an entire year are constructed for each energy consumption category, type of building sector, modeled location, and scenario by multiplying the SECB values from Table A4 by  normalized profiles: 1. Space heating SECB is multiplied by the normalized hourly profile of aggregated natural gas consumption profiles for space heating, only in the residential [321] and the services sector [322], and the daily HDD profile with base temperature 18 °C [323]. The natural gas consumption profiles for space heating only are made by subtracting the natural gas consumption for water heating from the total natural gas consumption profiles. 2. Space cooling SECB is multiplied by the hourly CDD profile with base temperature 21 °C [241,323]. 3. Water heating SECB is multiplied by the normalized hourly profile of the aggregated gas consumption profiles for water heating only. The natural gas consumption for water heating is extracted from the total aggregated natural gas consumption profiles during the period of 3 summer weeks (day 205 of the year onwards) with ambient temperatures above 18 °C, where it is assumed no space heating is taking place [321,322]. As the profiles are based on aggregated values, it is assumed that holiday effects are excluded. 4. Cooking, lighting, and electrical appliances SECB values are multiplied by the normalized aggregated electricity consumption profiles for residential [324] and services sector buildings [325].
The solar electricity production (ES) is calculated [326,327] using the hourly global horizontal irradiation values from both Murcia and Hamburg [238][239][240]. The irradiation values are assumed to be equal in both scenarios. With the given fixed roof area available for solar electric modules (Table  2) and the solar electricity system performance ratio and efficiency of, respectively, 0.80 and 0.20 kWp/m 2 in the Near Future scenario and 0.90 and 0.33 kWp/m 2 in the Mid Century scenario [134,[328][329][330][331][332], 11.20 and 18.67 MWp of solar power is installed in the Near Future and Mid Century for Hamburg and Murcia. The solar system inclination is 34° and 39° for, respectively, Murcia and Hamburg [333], both with an azimuth of 0°.

B.2. Road Transportation Hydrogen Demand
Annual hydrogen consumption for road transportation Hroad (kg H2/year) (blue rectangle in Figure 3) of the passenger cars, vans, trucks, tractor-trailers, and buses is calculated in Table A6 using the German and Spanish national average annual distance driven d [57,119,120] and the estimated vehicle fuel economy, specific energy consumption SECT (kg H2/100 km), in the Near Future and the Mid Century scenario [36]. For Hamburg, this results in Hroad of 479,909 kg H2/year in the Near Future scenario and decreases to 316,129 kg H2/year in the Mid Century scenario. For Murcia, this results in Hroad of 545,192 kg H2/year in the Near Future scenario and decreases to 381,732 kg H2/year in the Mid Century scenario. Hroad is then multiplied by a normalized repeating weekly fueling profile [260].
In the Near Future and the Mid Century scenario, the average annual distance driven is assumed to remain constant. The number of tube trailer tractors for hydrogen transportation and their driven kilometers are assumed to be included in the number of road tractors and the number of kilometers they are driven each year. The road vehicles are owned by either the residential or the services sector, and the road transportation energy is consumed in or between smart city areas. The final energy consumption for motorcycles is not included as it currently represents only about 1% of the total road transportation final energy consumption.
At an average annual speed of 45 km/h for passenger cars [129] in Europe and the average annual distance driven d (Table A6), there are only about 305 and 280 driving hours per year per car for, respectively, Hamburg and Murcia, mostly occurring during daylight hours [129]. For most of the non-driving time, passenger cars are mostly parked at home, the office, or close to a services sector building like a supermarket or hospital [129,334].  [57]; therefore, the same relation between the average annual distance driven of cars and vans as in Germany is used. b Including commercial vehicles 3.5-6.0 tons. c Based on a higher heating value (HHV) of 39.41 kWh/kg H2.

B.3. Electricity and Hydrogen Hourly Balance
The red rectangle, in Figure 3, includes both the electricity and the hourly balance. First, the electricity consumption of the services and residential buildings (EB), the HFS compressor and chiller (EHFS), SHS compressor (ESHS) is subtracted from the solar electricity production (ES). Any surplus solar electricity (Esurp) is converted via electrolysis and water (H2OS) into "solar" hydrogen (HS). If there is a shortage of electricity, this is compensated for by electricity from the FCEV2Gs (EV2G) by converting hydrogen (HV2G). The amount of hydrogen consumed for V2G (HV2G) is added the next day to the hydrogen fueling profile for road transportation (Hroad) and follows the same hourly pattern. HV2G and Hroad combined make up the total hydrogen dispensed at the HFS (HHFS).

B.4. Hydrogen Tube Trailer and Tractor Fleet
The grey rectangle, in Figure 3, shows the hydrogen tube trailer transportation. Once a tube trailer (TT1) is filled with "solar" hydrogen (HS), tube trailer tractors (TT2) transport the tube trailers to the HFS and unload them if the high-pressure storage tank (HFS2) is not full. If HFS2 is full, the tube trailer is emptied at the seasonal hydrogen storage (SHS1). If there is insufficient HS, and HFS2 is not full, tube trailers are filled at the SHS with the compressor (SHS2) and transported to the HFS.
Transportation of the tube trailers is modeled as one hour of unavailability; tube trailer tractor averages driving speed of 50 km/h with a single trip distance of 50 km. With a 2-h loading and unloading time [190] and 8 working hours per shift, one tractor can make two roundtrips per shift. The number of tube trailers at the three locations (solar hydrogen production, HFS, and SHS), together with the maximum number of tube trailers transported at the same time, defines the total number of tube trailers needed.

B.5. Wind Hydrogen Production and Seasonal Hydrogen Storage Balance
Wind hydrogen production and the seasonal hydrogen storage balance is shown in the green rectangle in Figure 3. As the amount of solar electricity consumption variation is limited due to the limited amount of suitable roof area, the amount of installed wind capacity, together with energy storage, closes both the hourly and the annual energy balance. The large-scale wind turbine park shared with other smart urban areas produces electricity (EW) and is directly connected to a water electrolysis and compression system (W2-W6) and has no connection with any other electricity grid. The wind energy production is sized such that the net amount of consumed hydrogen from the seasonal hydrogen storage in underground salt caverns is zero on a yearly basis. There is no curtailment of wind electricity (EW), and all electricity produced is used for the production and compression of "wind" hydrogen (HW) from water (H2OW).
The wind turbine park performance is based on the 4.2 MW land-based Enercon E-141 EP4 [335] for the Near Future scenario, and, for the Mid Century, it includes future power curve improvements based on [336]. In both scenarios, the hub height is 159 m, and the rotor diameter 141 m. The wind electricity production (EW) is calculated using the hourly wind speed values from both Almeria (see Table A1, footnote 1) and Hamburg [179][180][181][182]. The wind speeds are assumed to be equal in both scenarios and are scaled [337] to the aforementioned hub height with a roughness factor z0 of 0.13 m [338].

C.1. Smart City Area Total System Cost of Energy
The TSCoESCA in euros per year is the sum of the total annual capital and operation and maintenance costs TCi (€/year) of the total number of components (n) in the smart city area: The CCi (€/year) of a component is calculated using the annuity factor AFi (%), installed component capacity Qi (component-specific capacity), and investment cost ICi (€ per componentspecific capacity): Where the annuity factor AFi [339,340] is based on the weighted average cost of capital WACC (%) and the economic lifetime of a component LTi (years): The annual operation and maintenance costs OMCi (€/year) are expressed as an annual percentage OMi (%) of the Qi and ICi:

C.2. System Levelized Cost of Energy
The system levelized cost of energy, for either electricity SLCoEe (€/kWh) or hydrogen SLCoEH (€/kg H2), is calculated by allocating a share of the TSCoESCA related to either electricity TSCoESCA,e or hydrogen consumption TSCoESCA,H. These shares are then divided by either the annual electricity consumption ECe (kWh/year) or the annual hydrogen consumption ECH (kg H2/year), resulting in, respectively, the SLCoEe or the SLCoEH: CoE € year =CoE +CoE , The cost of energy for transportation energy CoEhh,T (€/hh/year) is calculated by multiplying the SLCoEH by the average annual distance driven by passenger cars dpassenger car (km/year/vehicle), the number of passenger cars per household nhh,passenger cars (#/hh), divided by the SECT,passenger cars (kg H2/100 km).

Appendix F Background Figures Cost of Energy for a Household
Results from previous sections serve as input for the cost of energy for a single household CoEhh (€/hh/year) in Table 5 in Section 4.4-the SECT and average annual distance driven d from Table A6, the specific energy consumption in buildings (SECB) from Table A4, the number of passenger cars  and households from Table 2, the average household floor area (Shh) from Section 2.2.2, and the SLCoEe and SLCoEH from Table 4.
For the Present scenario, additional parameters are used as given in the previous sections. An average gasoline fuel consumption of a passenger car in the European Union is approximately 5.6 L/100 km [341]. Gasoline prices without taxes and levies in Germany and Spain in 2017 were 0.500 €/L and 0.544 €/L [342]. For this comparison, it is assumed that natural gas is used for space heating, water heating, cooking, and electricity for space cooling, lighting, and appliances. Average electricity prices without taxes and levies for households in Germany and Spain in 2017 were 164 €/MWh (1000-2500 kWh annual consumption) and 150 €/MWh (5000-15,000 kWh annual consumption), and natural gas prices without taxes and levies were 45 €/MWh (20-200 GJ annual consumption) and 80 €/MWh (<20 GJ annual consumption) [343,344].