4.2.1. Techno-Economic Assessment

It is important to analyse each component on an individual basis to fully understand their contribution to the economic and emissions performance within the system. This would not only help to confirm the results seen in the pareto optimality, but also from a practical perspective assist a potential system designer to identify the most important assets, any particularly sensitive parameters, and assess the risks associated with each.

The solar array has the largest capital and operational costs compared with the wind power alternative due to the higher unit costs per kW. Despite this, the PV solar was able to achieve a greater capacity factor, which is the measure of energy output as a ratio of the total potential output of the same period. PV solar is naturally limited by the hours of solar available, while wind power is limited by the average wind speed and distribution. The higher capacity is the main mechanism which produced a lower LCOE for the PV solar of 0.07 compared with wind power's 0.09 EUR/kWh, despite the higher CAPEX and OPEX costs. This result also implies that although wind power is possible on with the REC, it may be beneficial from a to study a PV solar generation only option due to the potential impracticalities of local wind turbines. The inverse was then observed for the environmental impact, in that the PV solar has considerably higher embedded emissions of 40.7 gCO2*e*/kWh compared with the 15.9 gCO2*e*/kWh expected from equivalent wind energy. These results show a good agreement with the reported embedded emissions from the IPCC AR5 report [70] of 45 gCO2*e*/kWh and 13 gCO2*e*/kWh for solar and wind, respectively.

At EUR 106k, the RHFC CAPEX was a factor of twenty higher than the battery. This trend carries over into the LCOE results, where an approximate doubling of the levelized cost was observed for the hydrogen system compared with lithium batteries. These outcomes are in line with similar hydrogen system results found in the literature [55,71]. It is widely known that hydrogen technology is a less financially viable alternative for many applications, so this result was somewhat expected. This could change in the near future as the costs of hydrogen technology reduce.

The emissions output from the battery per kWh delivery was far higher than the hydrogen solution at 72.4 and 18.6 gCO2*e*/kWh, respectively. The trend is also supported by the population variables in Figure 9b, in which it is noted that as the hydrogen assets increased in capacity, the emissions result improved, while the net savings deteriorated.

Figure 11 contains the present value curve of the grid-only case, that is when electricity cost is paid to the utility company over the project period. The curve starts at zero as there is no capital cost associated with grid usage, but the operational cost per year is high. By contrast, the modelled REC requires an initial investment of EUR 380k. However, the lower year-on-year cost means that the system can pay off the investment cost, described as the payback term, in 9 years. The project ends with a final total savings of EUR 178k when the inflation and discounts rates of 2% and 5%, respectively, are considered. The result produces an IRR of 10.9%, year's returns of 9.6%, and an average system LCOE of 0.16 EUR/kWh. This assessment was based on the cost of equipment and installation since 2020.

**Figure 11.** The present value over the project period. The system was primarily compared with a 'business-as-usage' grid-only scenario.

Considering the historical and currently observed trends in renewable generation and storage equipment cost, it is projected that by 2030 and beyond there will continue to be a substantial decrease in the financial requirements for this type of system. The results shown here are therefore towards the upper bounds in terms of uncertainty about the future cost of an REC implementation.

#### 4.2.2. REC Members' Net Savings and Environmental Impacts

The model not only provides a global view of the potential impact of an REC configuration but is also able to analyse the reduction in cost and emissions on a per load basis. There were seven discretised loads within the model, with each being able to mutually accept and trade energy with the decentralised assets. Table 7 below shows the average LCOE and emissions intensity for each building and the percentage decrease in emissions. The REC provided a considerable degree of self-consumption, ranging from 91.1% for the largest load to over 98% for the smallest. In terms of the impact on the energy cost, the new LCOE ranged between 0.16 and 0.17 EUR/kWh compared to 0.30 EUR/kWh for grid-only. The decarbonisation of energy usage was also seen to be in the range of 75–77% in the first year of installation.

**Table 7.** Quantity of energy delivered to the REC compared with the quantity of energy delivered from the grid.


#### *4.3. Best Case and Extremes Comparison*

During the study, it was vital to understand not only the characteristics of the system at the 'best' pareto result, but also the performance at the extremes of the multi-objective optimisation. The result gives an indication as to how sensitive the result was to changing parameters. Table 8 contains the results of the three chosen REC configurations in terms of hybrid generation and storage capacities.

**Table 8.** Comparison REC configurations for extreme cases for net savings and decarbonisation potential compared with the chosen nominal case.


#### *4.4. Pareto Front Comparison of Energy Storage System Technologies*

The hybrid ESS comprised of a lithium battery and RHFC system produces differing performance outcomes based on the relative capacities of the technologies. Therefore, a comparison of the multi-objective optimisation for the same REC set-up with an additional battery-only ESS and RHFC-only ESS are required to ensure that the hybrid design was able to provide the best performance in terms of environmental impact and cost savings for the REC members. Figure 12 compares these pareto fronts, from which the combination of the technologies was able to produce a significant improvement over the technologies working independently. This was likely due to the fact that the battery is better at providing a shortterm response but suffers from increased degradation if used frequently for charge and

discharge, and similarly the hydrogen system requires a high capital cost and is best suited to the long-term storage of grid energy. The battery alone also has increased embedded carbon, which limited its ability to decarbonise. The hydrogen-only system suffered from the limitation that the electrolyser only runs at rated power, limiting flexibility.

**Figure 12.** The present value over project period compared with the 'business-as-usual' grid-only scenario.

Table 9 below summarises a range of LCOE results from the literature with similar smart grid and renewable energy architectures, incorporating hydrogen storage where available. It should be noted that specific parameters such as renewable resource availability, local technology costs, and system sizing may induce uncertainty in the resulting total system LCOE. For example, locations with a higher solar potential are naturally able to achieve a lower PV solar LCOE, and conversely, remote areas with high delivery and installation costs would experience higher project costs. The resulting average is 0.13 EUR/kWh, which is in good agreement with the central economic and environmental trade-off case produced in this work.

**Table 9.** Comparison of the levelized cost of electricity results of similar hybrid hydrogen smart grid systems within the literature.


#### **5. Conclusions**

This work presents a novel decentralised hybrid generation and ESS implementing both battery and hydrogen technology for use in a geographically isolated rural renewable energy community. A review of the existing state of the art was presented and highlighted the gaps in knowledge for such a system, particularly when considering both economic feasibility and a dynamic calculation of the environmental impact. This discussion comes at an interesting time for Europe and around the world as policymakers work to facilitate the potential benefits of aggregating decentralised renewables, one such method being the REC. The steady cost reduction in PV solar and wind power as seen over the past years has also accelerated growth in the decentralised energy sector. The rise of cost-effective hydrogen technology is set to make a considerable impact on how energy is stored and transported as a vector.

The results from this study show that there is an inherent trade-off relationship between cost reduction and the ability to decarbonise the energy system. By using a model built in Python, several different economic and environmental scenarios can be assessed. The implementation of a multi-objective algorithm gives potential system designers and policymakers a range of possible solutions. In this case, the optimal design results in an LCOE of 0.15W EUR/kWh, a project IRR of 10.8%, and an ROI of 9 years. Greenhouse gas emissions were reduced by 72% in the first year of installation to 69 gCO2*e*/kWh.

In further studies, emphasis should be placed on performing a sensitivity analysis and understanding where uncertainty may arise in the energy model. In particular, varying the component input assumptions such as capital cost and embedded emissions in line with reported ranges from the literature would further account for future uncertainty in performance. Additionally, further improvements to the system sizing optimisation method, including additional objective functions such as the loss of load probability from the literature, quantifying social impacts, and applying local space constraints would capture other potential strengths and weaknesses of the hybrid storage technology. Finally, research into the implementation of REC architectures in other locations around the world beyond the case study in this work would not only provide additional validation but also give a global perspective of how effective an REC could be in keeping energy costs stable and curbing the impacts of climate change.

**Author Contributions:** R.G.: Conceptualization, methodology, software, validation, formal analysis, investigation, resources, data curation, writing—original draft preparation, writing—review and editing, visualization. Z.D.: Supervision, project administration, funding acquisition, writing—review and editing. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research is sponsored by the EU Horizon 2020 research and innovation program under the grant agreement No 957852: Virtual Power Plant for Interoperable and Smart isLANDS—VPP4ISLANDS. More information is available at https://cordis.europa.eu/project/id/957852 (accessed on 30 August 2023).

**Data Availability Statement:** Data will be made available on request.

**Acknowledgments:** Special thanks are extended to the Consel de Formentera for allowing access to energy consumption data, from which the energy community model was built.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Abbreviations**

The following abbreviations are used in this manuscript:



#### **References**


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