Performance Optimization of Renewable Energy Communities: The Case Study of Nearby Towns Around Parma, Italy

Over the past few years, energy community initiatives have gained traction in Europe to achieve energy transition goals, allowing citizens to be involved in energy production, consumption and distribution. The rising importance of renewable energy communities (RECs) has been highlighted by Ahmed et al. [1], who presented a comprehensive review of REC.

The energy shared within a REC, which is defined as the minimum on an hourly basis between the electricity produced by the renewable energy sources and the electricity supplied from the grid to members of a REC, enables the establishment of virtual self-consumption that brings both environmental and economic benefits to REC members.

The participants in the energy community can take on these roles: Producers, consumers, and prosumers, who are both producers and consumers.

An important challenge in the design and operation of energy communities is the effective alignment of energy demand and production [2, 3]. The community must be designed to maximise the self-consumption of renewable energy.

However, due to the intermittent nature of renewable sources and the absence of storage systems, the alignment between generation and demand largely depends on the consumption patterns of its members. To increase energy sharing, members with diverse load profiles must be selected.

Lazzari et al. [4] proposed an optimization model based on genetic algorithms that maximize self-consumption, minimize excess solar energy, and achieve the shortest payback period. The optimization model was applied to a case study in Spain.

Volpato et al. [5] presented a novel procedure based on stochastic forecasts that supports decisions regarding the optimal design of multi-energy systems supplying RECs under uncertainties in solar irradiance and users’ energy demand.

Cosic et al. [6] developed an optimization model for REC based on mixed-integer linear programming, which accounts for distributed photovoltaic systems, energy storage systems, different electricity tariff scenarios, and market signals. The optimisation model was applied to three case studies in Austria.

Laurini et al. [7] also presented an optimization model based on mixed-integer linear programming, which enables defining the optimal configuration of a REC, identifying the most suitable technologies, and selecting members. The optimisation model was applied to a case study in Italy.

The present study investigates the energy performance optimisation of a REC in Northern Italy. The case study was selected by considering that in a small town close to Parma (i.e., the municipality of Sorbolo-Mezzani), there is a large industrial facility that generates a large surplus of electricity (i.e., Cosmoproject S.p.A.). This important facility serves the global skin care and cosmetics, pharmaceutical and international luxury brands.

To optimize energy sharing in the REC, different scenarios were explored by considering different energy user demands (i.e., different consumers and prosumers). The REC members were selected based on a map of the electrical primary substation (according to Italian legislation) and the load profile of the industrial facility. More specifically, industrial, commercial, school, and residential buildings located in the municipalities of Colorno and Sorbolo-Mezzani were selected. A simulation model developed in MATLAB^® was employed to assess each configuration, accounting for load profiles and PV generation.

The REC members were selected by considering the following constraints: i) the members might be under the same primary cabin according to Italian legislation, and ii) the REC must include members with different load profiles.

The analysis of the buildings connected to the same primary substation, along with the load profiles of the Cosmoproject S.p.A., indicated that commercial, school, and residential buildings could be incorporated into the REC.

Commercial, school, and residential buildings were included due to their extended operational hours, particularly during weekends when the industrial facility was inactive. Moreover, the integration of another industrial building within the REC was considered due to its high electrical consumption and extended operational hours (i.e., this facility also works on Saturdays).

To optimize the energy shared within the REC, four scenarios were explored, as summarized in Table 4. In particular, the first scenario considered only the industrial building Ind₁ and the commercial buildings Com₁ and Com_2,3. The second scenario considered the integration of school buildings within the REC, while the third scenario also considered the integration of residential buildings within the REC. Finally, in the last scenario, the integration of another industrial building Ind₂ was considered.

Table 4. Simulated scenarios

The ouctomes revealed that in the first scenario (i.e., only Ind₁ and commercial buildings were included within the REC), approximately 186 MWh/a of energy was shared, resulting in a collective SSR of 3.7%. A total of 3,670 MWh/a was drawn from the grid.

The integration of the schools of the municipality of Colorno into the REC (i.e., the second scenario), around 192 MWh/a of energy was shared, achieving a collective self-sufficiency of 3.7%, with 3,818 MWh/a drawn from the grid.

tu_pian_3.png

Figure 3. Energy flows for a sample week in February (Scenario 1)

tu_pian_4.png

Figure 4. Energy flows for a sample week in February (Scenario 2)

tu_pian_5.png

Figure 5. Energy flows for a sample week in February (Scenario 3)

tu_pian_6.png

Figure 6. Energy flows for a sample week in February (Scenario 4)

The integration of residential buildings into the REC (i.e., scenario 3) increased the amount of shared energy to approximately 206 MWh/a, maintaining a self-sufficiency level of 3.8%, while 4,076 MWh/a was sold to the grid.

Finally, by including the nearby industrial building within the REC, the shared energy increased to approximately 230 MWh/a; however, the collective self-sufficiency dropped to 2.3%, and the energy drawn from the grid increased significantly to 7,697 MWh/a.

The simulation results in terms of energy flows within the REC for a sample week in winter are presented in Figures 3-6.

tu_pian_7.png

Figure 7. Energy flows for a sample week in June (Scenario 1)

tu_pian_8.png

Figure 8. Energy flows for a sample week in June (Scenario 2)

tu_pian_9.png

Figure 9. Energy flows for a sample week in June (Scenario 3)

tu_pian_10.png

Figure 10. Energy flows for a sample week in June (Scenario 4)

As expected, in winter, the level of energy sharing within the REC was very low due to the high electrical consumption of Ind₁, which exceeded the energy production of the PV system, with the exception of the weekends in which the building was closed.

Figures 7-10 present the energy flows within the energy community for a sample week in summer (i.e., in June).

Although an increase in energy sharing within the REC was observed, the overall level remained low. This was primarily due to excess photovoltaic generation, which consistently surpassed the community’s energy demand—particularly on weekends, when Cosmoproject S.p.A. was closed.

To gain deeper insights into the factors contributing to the low levels of energy sharing within the REC across all scenarios considered in this study, an analysis of energy flows during summer holidays was carried out.

It was found that in August, the amount of energy fed into the grid was very high, as shown in Figures 11-14, where the energy flows in August are presented for all the investigated scenarios.

It can be observed that although the commercial buildings are opened during the summer holidays, except for public holidays, their electrical consumption is very low with respect to the energy produced by the three PV systems.

However, the amount of energy shared within the REC was high during this period. More specifically, the shared energy in August is around 38 MWh, 40 MWh, 44 MWh, and 45 MWh for scenarios 1, 2, 3, and 4, respectively.

The simulation results in terms of KPIs are detailed in Table 5. The ESP decreases across the scenarios, as it reflects the ratio of total energy produced within the REC to total energy consumed—while production remains constant, overall energy consumption increases from the first to the last scenario.

The SCR and SSR fluctuate due to variations in energy sharing within the REC and in total electrical energy demand, which are influenced by hourly electricity load changes.

tu_pian_11.png

Figure 11. Energy flows for a sample week in August (Scenario 1)

tu_pian_12.png

Figure 12. Energy flows for a sample week in August (Scenario 2)

tu_pian_13.png

Figure 13. Energy flows for a sample week in August (Scenario 3)

tu_pian_14.png

Figure 14. Energy flows for a sample week in August (Scenario 4)

Table 5. Results of scenario analysis in terms of KPIs

tu_pian_15.png

Figure 15. Monthly values of the ratio between the shared energy and the excess energy generated by the industrial facility (Scenario 3)

The scenario analysis revealed that the REC in which commercial, school and residential buildings presented the best energy performance, being characterized by the highest value of the SCR. In this scenario, a detailed analysis was conducted by examining the monthly ratio of shared energy to the excess energy generated by the PV systems. All excess energy generated by the PV systems during winter was shared within the REC, whereas in summer, only approximately 65% of the excess energy was distributed within the community, as shown in Figure 15.

This research is funded by the Emilia-Romagna Regional Development Fund PR-FESR 2021-2027 Program under the project SACER – Sviluppo e integrazione di Accumuli innovativi nelle Comunità Energetiche Rinnovabili, CUP J47G22000760003.

The authors gratefully acknowledge the support of Cosmoproject S.p.A.