Ammonia as hydrogen carrier for transport application

. As the interest in hydrogen to help the decarbonization of the transport sector is growing fast, the interest in new methods for its storage is a key point to improve its diffusion in many contexts, investigating innovative methods. Ammonia is a promising solution, as its hydrogen content per volume unit is higher than hydrogen stored in liquid form; furthermore, ammonia does not require cryogenic temperature nor high amounts of energy for liquefaction. In this study, two different plant layouts have been investigated, considering as a case study an ammonia-to-hydrogen conversion plant to feed a bus station composed of ten hydrogen buses (106 kg H2/day). In the end, a techno-economic analysis is performed to investigate the Levelized Cost of Hydrogen production from ammonia for the two cases and evaluate the most feasible solution. For both the plant layouts, the following results are obtained: (i) the optimal size of the main components; (ii) the global energy efficiency; (iii) the purity of H 2 obtained; (iv) the H2 production cost. Finally, the size effect is investigated to evaluate the economic feasibility of the best plant solution for large-scale hydrogen refuelling stations (2000 kg H 2 /day), which are a more representative case for future implementations.


Introduction
The transport sector and electricity producers are responsible for the more than 22 Gtons of CO2 emission on a total of 36 Gtons in 2021 [1].Then, making decisions for decarbonization is essential, and hydrogen (H2) is a promising alternative being a carbon free fuel [2][3].One of the issues related to H2 utilization is its transport and storage [4][5].The three main storage methods available today are i) high-pressure as compressed gas, ii) cryogenic method for liquid H2 and iii) solid-state storage in metal hydrides [6][7][8].Further options are based on the chemical storage of H2 by using compounds, such as methanol [9][10][11] and ammonia (NH3), which need less harsh storage conditions (temperature and pressure) [12].NH3 is a strong candidate as H2 carrier because it provides a higher H2 volumetric density (120 kgH2/m 3 ) than other methods (e.g.compressed H2 at 700 bar is 38 kgH2/m 3 ) and compounds (e.g. for CH3OH is 99 kgH2/m 3 ) [12].Moreover, the risk management associated with its high toxicity is based on worldwide well-consolidated know-how in production, transport and storage, being NH3 one of the top products of the basic chemical industry.Then, such deep knowledge about NH3 represents a strength for designing optimized processes for on-demand H2 production based on ammonia.The final use of H2, obtained by NH3 decomposition, strongly affects the whole process configuration, which must satisfy the purity grade required by the application [13] [14].Ttwo different plant layouts have been considered for H2 production from NH3, for transport application based on a polymer electrolyte membrane fuel cell (PEMFC) system.

Case study
Two different plant configurations have been proposed for H2 production from NH3 for a fuel station dedicated to the public transport.The main difference between the two case studies is the technology employed for H2 separation after the cracking reactor, which affects the H2 purity, the energy/material loops recovery, and the energy efficiency.For both cases, the operating conditions of all the equipment have been set, as well as the compositions of all the fluxes involved have been calculated to evaluate energy and mass process balances.Dataset used for evaluating the H2 daily production for the aforementioned application are based on projects about green mobility [15].Each bus has an average H2 consumption of 8.7 kgH2/100 km (0.11 kgH2/km), a daily driving time of 5.25 hours/day, and an average speed of 18.4 km/h.Then, the overall daily consumption for one bus is 10.6 kgH2/day.In the first case (CASE 1-Fig.1), liquid NH3 (-33 °C, 1 bar) is compressed and heated in several consecutive steps to reach the conditions for the endothermic decomposition reaction (450 °C, 10 bar), allowing an NH3 conversion of 96% as confirmed by simulation, and in accordance with the literature [13].The outlet reactor stream is rich in H2 and cooled up to 25 °C by a heat exchanger and a cooler to be fed to the bottom of the countercurrent absorption column while the water is fed from the top.H2 is separated from the unreacted NH3 using a two steps method.Firstly, the NH3 is recovered in an absorption column by water, followed by a stripping column with air for water regeneration.The absorption stage produces a gas phase mainly composed of H2 (74%) and NH3, and a liquid phase made of water and NH3.The first is compressed to 350 bar, while the latter is treated in a stripping column with air.This last step allows the recovering the water for absorption while the NH3-rich stream is fed to a chiller.Here, the temperature is set to -50°C to condensate NH3 and clean the air stream.The unreacted NH3 is then mixed with the fresh NH3 stream.In the second plant layout (CASE 2 -Fig.2), the first steps of the process are similar to those reported for CASE 1, to heat and compress NH3 from storage (-33 °C, 1 bar) to decomposition conditions (450 °C, 10 bar).The dashed line, which feeds the decomposition reactor, represents the heat supplied by the combustor.The output H2-rich gas stream from cracking reactor is cooled to 25 °C, and the pressure is decreased by an expansion valve to 7 bar, the value required by the pressure swing adsorption (PSA).Here, the separation of H2 is carried out by four absorbent beds made of 5A zeolite, with an H2 recovery of 71.4% and a final H2 purity of 99.51%.This stream is compressed to 350 bar, while the waste stream from PSA is mainly composed of H2 and NH3, then is fed to a combustor producing the heat required by the cracking reaction.The working temperature of the combustor is set at 800 °C; this value has been calculated to guarantee the heating required by the cracking reactor.

Main results and discussion
A mass and thermal balance for both plant layouts have been investigated.The most relevant results are summarized in Table 1.CASE 1 requires a lower amount of NH3 (36.52 kmol/d) than CASE 2 (51.42 kmol/day) for providing the same daily H2 production.Nevertheless, CASE 2 provides higher purity hydrogen (99.51% vs 74%), and the flue gas stream from the combustor is mainly composed of N2 (molar fraction, XN2=0.71),water (XH2O=0.22)and H2 (XH2=0.07).The molar fractions of nitrogen compounds such as NH3, NO, and NO2 are evaluated in an order of magnitude of 10 -6 , 10 -13 , 10 -23 , respectively, which are compatible with the emission in the air as well as its final temperature (56 °C).CASE 2 has also lower electrical/thermal energy consumption, but the presence of PSA leads to worse energy efficiency.Indeed, PSA is less performing in H2 separation than the absorption column used in CASE 1 (separation efficiency PSA = 71.4%, absorption column = 90%).This drawback is counterbalanced by the higher purity of the obtained H2 as reported above.Another factor to take into account is the complete process layout.From Fig. 1 and Fig. 2 is well evident that the second solution is simpler with a reduced number of equipment.• 350 operative days per year.To evaluate the Purchased Equipment Cost (PEC), reference value for a similar plant layout used as reference with a capacity of 500 kg H2/day has been considered.In this case, the PECref was equal to 15 M€, as reported in recent literature [18].Thus, considering the plant size scale-down to 106 kg H2/day, the correct PEC can be calculated as follows: Where x is the capacity factor, set equal to 0.65 as reported in literature for industrial plants.
The PEC for the lower scale plant has been determined in 5.48 M€.Assuming a TCI/PEC ratio equal to 2 [10], the TCI is estimated in 10.96 M€.Annual costs distribution for this kind of plant is reported in Fig. 3.It is worth noting that the largest contribute is due to annual fixed costs for this kind of plants, in particular for the small size analysed for this application.However, the plant complexity is high and the hydrogen production is quite limited (106 kg/day), based on the hydrogen demand for ten buses operations.Therefore, the hydrogen production cost, calculated as reported in Eq. (1), results 19.86 €/kg, which is a high value compared to market ones.A further analysis is carried out considering the influence of plant size.As the target is a refuelling station for vehicles, higher production rates, up to 4,000 kg/day, are analysed.Main economic results are shown in Figs. 4 and 5. Fig. 4 shows costs' breakdown: as the size increases, specific CAPEX gets lower and OPEX influence gets higher, in particular the ones related to ammonia consumption.Electrical and thermal energy amounts required in the process are limited thus their economic impact is low too.
Fig. 4 Annual costs' breakdown for different plant sizes Fig. 5 COH trend vs size Fig. 5 shows COH trend according to plant size variation.As the size gets higher, the economic feasibility is improved: for 2,000 kg/day COH is 10 €/kg, which is quite close to green H2 production prices in several present scenarios [21][22].In this configuration, H2 production plant is located close to the refuelling station, thus costs for H2 transport are saved.

Conclusions
This study considers ammonia as candidate for hydrogen storage for the mobility sector, evaluating the impact of process design on hydrogen purity, mass and thermal balances, as well as energy efficiency.Two different plant layouts have been presented, and the most relevant difference is related to the H2 separation system after the cracking reactor.The results show that the H2 separation by absorption and stripping has a higher energy efficiency and a lower ammonia daily consumption than the solution based on PSA.Nevertheless, this second approach improves the final hydrogen purity and reduces the kWh/kgH2 required by the process, and it also has a simpler process layout, making CASE 2 more suitable for the E3S Web of Conferences 414, 02005 (2023) SUPEHR23 https://doi.org/10.1051/e3sconf/202341402005application.Starting from the energy results, a techno-economic analysis is carried out to evaluate the hydrogen production cost COH).For the considered application, as the size is very small, investment costs are too high to achieve economic feasibility: COH is about 20 €/kg.However, considering higher sizes, comparable to the ones for a typical refuelling station (I.e.2000 kg H2/day), costs would be significantly lower (10 €/kg).Furthermore, it must be noted that this configuration would avoid costs related to hydrogen transportation.

Fig. 1
Fig. 1 Plant layout considered in case study 1

Fig. 2
Fig. 2 Plant layout considered in case study 2

Table 1 : energy comparison between the two plant layouts
E3S Web of Conferences 414, 02005 (2023) SUPEHR23 https://doi.org/10.1051/e3sconf/202341402005Annual Fixed Costs (AFC) are evaluated starting from Total Capital Investment (TCI), considering the plant lifetime in years n and the interest rate r: