Enhancing Renewable Energy Use in Residential Communities: Analyzing Storage, Trading, and Combinations

Abstract

Renewable energy resources, especially rooftop solar PV, have gained momentum during the past few years. However, the local consumption of PV power is limited due to the negative correlation between peak PV power and residential loads. Therefore, this study analyzes various cases to maximize the consumption of renewables in communities encompassing dwellings both with and without PV installations. The three cases considered in this study are local energy storage, community energy storage, and internal trading. A total of six cases are analyzed by evaluating these cases individually and in combinations. To achieve this, first, a generalized optimization model with specific constraints for each case is developed. Subsequently, different indices are devised to quantitatively measure trading with the grid and the consumption of renewables under varying cases. The performance of these different cases is analyzed for a community comprising five dwellings over a summer week. Furthermore, the performance of each case is evaluated for various seasons throughout the year. Additionally, a sensitivity analysis of different storage capacities (both local and community) is conducted. Simulation results indicate that community storage results in the highest renewable consumption if only one case is considered. However, the overall combination of internal trading and community storage results in the highest cost reduction, lowest dependence on the grid, and the highest consumption of renewables. Finally, a techno-economic analysis is performed on four widely used battery technologies, taking into account diverse cost and technical considerations.

1. Introduction

The penetration of distributed energy resources (DERs) is increasing across the globe to combat climate change by decarbonizing the power generation sector. A sharp increase in DERs has been recorded in recent years due to incentives from governments and enhancements in DER-related technologies [1]. According to the International Energy Agency (IEA), the renewable energy capacity worldwide will exceed 4500 gigawatts in 2024, roughly on par with fossil fuels for the first time [2]. Solar photovoltaic energy is one of the major DERs which has shown significant growth both on a small scale (residential level) and large scale (commercial/utility level). Solar PV additions are expected to account for two-thirds of the increase in renewable power capacity in 2023 and are projected to continue growing in 2024 [3]. It is forecasted by the Energy Information Administration (EIA) in the US that the United States will generate 14% more electricity from solar energy than from hydroelectric facilities in 2024 [4]. Rooftop solar PV installations are expanding across various countries and are anticipated to experience substantial growth in the next few years. A study [3] shows that 86% of eligible homeowners without solar PV systems were open to considering them, with most planning to install them within one to three years.

With the increasing penetration of DERs, such as solar panels and wind turbines, several challenges arise, demanding careful consideration to maximize their utilization and benefits. For instance, dwellings equipped with solar PV systems often sell their excess power to the grid at a lower price through net metering or net billing [5], while adjacent residences without DERs purchase power from the grid at higher prices. Direct trading with the grid is disadvantageous for both types of dwellings (with and without DERs) [6]. Therefore, several studies have considered power trading among community members via internal markets. This internal trading system not only leads to energy bill savings but also enhances service reliability during power outages by allowing a shared distribution of resources among participants [7]. The internal pricing signal is set between the grid’s buying and selling prices, which benefits both participating dwellings, encouraging mutual exchange within the community [8].

Peer-to-peer (P2P) systems have been extensively discussed in numerous studies for energy trading among community members. For example, a P2P system proposed in [9] aims to optimize energy sharing, reduce bills, and enhance user comfort through flexible demand response. The results from this study have demonstrated significant savings and increased profits for participants, particularly those with PV systems. Similarly, another study focused on electricity trading in communities, as outlined in [10], examining trading dynamics and the factors influencing price and quantity. Additionally, a study introduced a multi-agent framework [11] to facilitate profitable P2P power trading among self-interested residential prosumers within a local electricity market. A bi-level optimization framework for P2P energy trading in microgrids is proposed in [12], focusing on network constraints and uncertainties. Several studies have focused on P2P renewable-based systems. For example, achieving the optimal sizing of PV systems and BESSs is carried out in [13] for prosumers to participate in P2P energy trading. A P2P energy trading system for a net-zero community is proposed in [14], implementing individualized pricing and time-of-use strategies for enhancing consumption of renewables. A P2P energy transaction model for distribution networks, using renewable energy prediction intervals and Nash bargaining, is proposed in [15] to manage uncertainty and optimize demand-side flexibility.

Another alternative to enhance the utilization of renewables is the installation of battery energy storage systems (BESSs) alongside PV systems. Several studies have delved into various aspects of BESS and PV integration in residential communities. For instance, [16] explores policies that encourage the combination of residential solar PV systems with BESSs. It focuses on their profitability, emphasizing different aspects such as self-consumption, incentives for energy storage, and dynamic pricing mechanisms. Additionally, [17] focuses on optimizing residential PV–battery microgrid designs to ensure cost efficiency while meeting technical constraints and maximizing component lifetimes. Furthermore, location-specific analyses have been conducted by researchers. For instance, the analysis of distributed PV–battery systems in Ulaanbaatar, Mongolia [18], aimed to identify profitable configurations and potential environmental benefits unique to that area. The analysis of electricity trading practices in a Japanese community is conducted in [19], emphasize policy insights aimed at maximizing renewable energy use and increasing autonomy through effective battery utilization. A P2P system for 21 solar and wind-rich islands in France is analyzed in [20]. The results demonstrate the feasibility of achieving economic efficiency and a substantial annual reduction in CO₂ emissions.

However, individual BESSs for residential homes often face limitations regarding scalability and cost-effectiveness. They can be expensive for individual consumers and may have limitations in efficiently managing excess energy within a community. Therefore, the concept of community BESSs has recently been considered in several studies [21]. The benefits of community storage in enabling demand response for a P2P energy pool are discussed in [22]. Additionally, [23] proposes an energy storage sharing model among multiple buildings, demonstrating fairness, computational efficiency, and substantial economic benefits over individual BESSs. A transactive energy framework is proposed in [24] utilizing auction theory and blockchain for P2P energy trading in a community microgrid. Simulation results have demonstrated substantial monetary benefits and improved self-sufficiency.

This literature review indicates that each of the cases—internal trading, local BESSs, and community BESSs—offers unique opportunities to enhance the utilization of DERs. Combining these technologies can further augment their effectiveness. However, existing literature lacks an in-depth analysis of each individual case and their various combinations. Additionally, there is a lack of focus on analyzing these cases in mixed communities where some residences possess DERs while others do not. Moreover, the absence of generalized indices for evaluating green energy consumption in dwellings with and without PV systems poses challenges for direct comparison. These indices are crucial not only for comparing the performance of different cases and combinations but also for benchmarking them against alternative methodologies. Therefore, there is a critical need to address these gaps in the literature through comprehensive studies that encompass detailed analyses of individual cases, their combinations, and their application in mixed communities. Developing standardized indices will be fundamental in evaluating and comparing the efficacy of various approaches toward leveraging DERs.

This study aims to offer a comprehensive overview of strategies for optimizing renewable energy consumption in residential communities by simulating different cases. These cases are formulated by considering two storage cases—community energy storage and local energy storage—as well as internal trading, both individually and in various combinations. The major contributions of this study are as follows:

• The development of a generalized optimization model tailored to six cases, including variations in storage options (community and local energy storage) and internal trading strategies.
• The introduction of two quantitative indices to measure renewable energy consumption and grid trading across the entire residential community.
• An exploration of the impact of different seasons and varied storage case sizes through impact analysis, providing insights into the model’s adaptability.
• Techno-economic analysis to evaluate the suitability of four commonly used battery technologies—Li-ion, lead–acid, Ni-Cd, and flow batteries—considering factors such as lifespan, efficiency, energy density, power density, and capital costs.

Simulation results indicate that community energy storage exhibits the most significant impact in maximizing renewable energy consumption, reducing grid dependence, and lowering the operational costs of the community. Moreover, the combination of internal trading with the community energy storage system further enhances the benefits.

2. Energy Storage and Trading for Renewable Energy Consumption

2.1. Renewable Energy Consumption

The peak generation period of solar PV energy often does not align with the peak load demand in the residential sector, resulting in limited local consumption of PV-generated electricity. Therefore, prosumers are compelled to trade electricity with the external grid. Moreover, community members without solar PV energy lack access to renewable power. To address these challenges, two cases—energy storage and internal trading—are considered to maximize renewable energy consumption in different communities. It is important to note that the installation of BESSs in each home might not significantly enhance renewable energy consumption in homes without PV systems. However, community-level BESSs can function as a buffer, benefiting all community members by effectively managing renewable energy resources.

2.2. Different Combinations of Storage and Trading

In this study, two storage cases are explored—a local storage system (LSS) and a community storage system (CSS)—alongside internal trading, assessing a total of six distinct cases. In the case of the LSS, each dwelling is equipped with a BESS, while the CSS employs a centralized BESS utilized by all community members (dwellings). In all cases, community members have the capability to trade power with the utility grid. An overview of the six cases is outlined in Figure 1 and details for each case are as follows:

• Base: This serves as the base case where each dwelling directly trades power with the utility grid to meet their energy demands. For instance, dwellings equipped with solar PV systems can both purchase and sell power to/from the grid.
• Local storage (LS) only: Each dwelling in this case is equipped with an LSS. Here, dwellings can only store excess renewable power in their LSS and utilize it during high-load periods or in the absence of PV energy generation. Notably, there is no power trading among different dwellings in this case.
• Community storage (CS) only: In this case, a CSS is collectively utilized by all community members. Dwellings with excess PV power can charge surplus power into the CSS, while all dwellings have the capability to discharge power from the CSS.
• Internal trading (IT) only: In this case, community members can share power among themselves. Dwellings with surplus PV power can sell their excess power to other dwellings with higher energy demands or no PV power at a lower price compared to the grid.
• IT + LS: Each dwelling possesses an LSS and can share power with other dwellings. Dwellings without PV systems can purchase excess power from other dwellings with surplus power and store it in their LSS for later use.
• IT + CS: In this case, a CSS is used alongside power sharing among community members. Dwellings with excess power can share their surplus power with other dwellings and charge it to the CSS. All dwellings have access to discharge power from the CSS to meet their energy demands.

3. Problem Formulation

In this section, an optimization problem is formulated to maximize the consumption of renewables, as discussed in the previous section. A generalized objective function is developed, incorporating common constraints applicable to all cases. Furthermore, specific constraints are defined for each individual case. It is worth mentioning that a linear problem is formulated in this study due to its merits, such as convexity and lesser computation time. Due to these advantages, several similar studies [14,17,25] have also used linear models. In addition, the developed model aims to minimize the cost of the community by minimizing the operation cost of individual households. Different combinations of storage and trading are used to achieve this objective.

3.1. Objective Function

The objective function is designed to minimize the overall cost incurred by the entire community when purchasing power from the grid. However, it is important to note that the latter part of the objective function incentivizes selling excess power to the grid. While this selling mechanism may encourage certain dwellings to sell surplus power, it may simultaneously result in other dwellings needing to purchase power from the grid. Hence, the priority remains on fulfilling the local energy demand of the entire community before considering the case of selling power to the grid. It can be modeled as follows:

$$\min {\displaystyle \sum _{t\in T}\left({\displaystyle \sum _{d\in D}\left(P{R}_t^b\cdot P_{t,d}^b-P{R}_t^s\cdot P_{t,d}^s\right)}\right)},$$

(1)

where $P{R}_t^b \mathrm{and} P{R}_t^s$ are, respectively, the per-unit buying and selling price of power from/to the grid. Similarly, $P_{t,d}^b \mathrm{and} P_{t,d}^s$ are, respectively, the power bought from the grid and sold to the grid by dwelling d at time t.

3.2. Constraints

The objective function remains consistent across all cases, while the constraints vary based on the presence or absence of different components such as internal trading, LSS, and CSS. The identifier d signifies that the constraints are applicable for all dwellings (households) in the community. Similarly, the time interval t signifies that these constraints are applicable for each time interval of the scheduling horizon (T). An overview outlining the distinct constraints necessary for each case is provided in Algorithm 1 and details of these constraints are presented as following.

Algorithm 1 Constraint selection for different cases.

1: Get hourly data: load and PV 2: Initialize all cases (all 6cases) 3: if (Base)

3.2.1. Power Balance

Ensuring power balance within each dwelling is essential to guarantee service reliability. However, the methods available for ensuring power balance vary across different cases. For example, in the base case, each dwelling can only trade power with the grid or use available PV power. Therefore, the power balance for the base case can be expressed by Equation (2), where $P_{t,d}^{pv}$ denotes PV power and $P_{t,d}^l$ signifies the load of dwelling d.

For LS and CS cases, two additional terms are added to the power balance equation due to the presence of an energy storage system in these two cases. In Equation (3), $P_{t,d}^{-} \mathrm{and} P_{t,d}^{+}$ represent the power discharged and charged by the energy storage system in dwelling d. It is worth noting that for the CSS, the total charged/discharged power will be, respectively, the sum of individual dwelling outputs as discussed in Section 3.2.2.

The power balance equation for IT is expressed in (4). It is noticeable that two new terms are introduced in (5) to account for internal power transfer among dwellings. In Equation (14), $P_{t,d}^{re} \mathrm{and} P_{t,d}^{sd}$, respectively, represent the amount of power received and sent by dwelling d at time t, as discussed in Section 3.2.3.

Finally, IT + LS and IT + CS cases encompass both energy storage and internal trading cases. Consequently, Equation (5) is employed as the power balance equation for these two cases. Equation (5) incorporates both energy storage-related terms ($P_{t,d}^{-} \mathrm{and} P_{t,d}^{+}$) and internal trading-related terms ($P_{t,d}^{re} \mathrm{and} P_{t,d}^{sd}$), in addition to grid trading ($P_{t,d}^b \mathrm{and} P_{t,d}^s$) and local PV power ($P_{t,d}^{pv}$).

$$P_{t,d}^{pv}+P_{t,d}^b-P_{t,d}^s=P_{t,d}^l$$

(2)

$$P_{t,d}^{pv}+P_{t,d}^b-P_{t,d}^s+P_{t,d}^{-}-P_{t,d}^{+}=P_{t,d}^l$$

(3)

$$P_{t,d}^{pv}+P_{t,d}^b-P_{t,d}^s+P_{t,d}^{re}-P_{t,d}^{sd}=P_{t,d}^l$$

(4)

$$P_{t,d}^{pv}+P_{t,d}^b-P_{t,d}^s+P_{t,d}^{re}-P_{t,d}^{sd}+P_{t,d}^{+}-P_{t,d}^{-}=P_{t,d}^l$$

(5)

3.2.2. Local Energy Storage

In the case of LS, each dwelling (d) can independently charge and discharge their storage system. The charging and discharging constraints for each time interval (t) for any dwelling (d) are, respectively, represented by (6) and (7). $P_d^{ini}$ is the initial power present in the LSS at the beginning of the day. $P_{t,d}^{+} \mathrm{and} P_{t,d}^{-}$ are, respectively, the amount of power charged and discharged during interval t. ${\eta }_d^{+}$ and ${\eta }_d^{-}$ are, respectively, the charging and discharging efficiencies, while $E_d^{}$ is the energy capacity of the LSS in dwelling d. Finally, $SO{C}_d^{max}$ and $SO{C}_d^{\min }$ are, respectively, the upper and lower bounds of battery state of charge (SOC). The LSS is also subjected to the rating of the power converter as shown in (8), where $P_d^{r+}$ and $P_d^{r-}$ are, respectively, the charging and discharging rates of the converter.

$$P_d^{ini}+{\displaystyle \sum _{\tau \le t}\left({\eta }_d^{+}\cdot P_{t,d}^{+}-P_{t,d}^{-}/{\eta }_d^{-}\right)}\le E_d^{}\cdot SO{C}_d^{max}/100$$

(6)

$$E_d^{}\cdot SO{C}_d^{min}/100\le P_d^{ini}+{\displaystyle \sum _{\tau \le t}\left({\eta }_d^{+}\cdot P_{t,d}^{+}-P_{t,d}^{-}/{\eta }_d^{-}\right)}$$

(7)

$$0\le P_{t,d}^{+}\le P_d^{r+}, 0\le P_{t,d}^{-}\le P_d^{r-}$$

(8)

3.2.3. Community Energy Storage

The CSS shares technical similarities with the LSS except for its ability to absorb power from all dwellings and supply power to all dwellings when needed. Therefore, Equations (9) and (10) are similar to those of the LSS, except for the absence of a dwelling identifier (d). Equation (11) implies that the power charged to the CSS at time t ($P_t^{+}$) is the sum of the total power charged by each dwelling ($P_{t,d}^{+}$) during that interval. Similarly, the total power discharged from the CSS at time t ($P_t^{-}$) is the sum of the power discharged by each dwelling ($P_{t,d}^{-}$). Similar to the LSS, the CSS is also constrained by the power ratings of the converter, as modeled in (12).

$$P_{}^{ini}+{\displaystyle \sum _{\tau \le t}\left({\eta }_{}^{+}\cdot P_t^{+}-P_t^{-}/{\eta }_{}^{-}\right)}\le E_{}^{}\cdot SO{C}_{}^{max}/100$$

(9)

$$E_{}^{}\cdot SO{C}_{}^{min}/100\le P_{}^{ini}+{\displaystyle \sum _{\tau \le t}\left({\eta }_{}^{+}\cdot P_t^{+}-P_t^{-}/{\eta }_{}^{-}\right)}$$

(10)

$$P_t^{+}={\displaystyle \sum _{d\in D}{P}_{t,d}^{+}}, P_t^{-}={\displaystyle \sum _{d\in D}{P}_{t,d}^{-}}$$

(11)

$$0\le P_t^{+}\le P_{}^{r+}, 0\le P_t^{-}\le P_{}^{r-}$$

(12)

The models of energy storage components (LSS and CSS) are discussed in the previous section. This section deals with additional constraints required to ensure the smooth operation of the community. To avoid selling power by discharging power from the CSS (13), an inequality constraint (13) is introduced, ensuring that discharged power ($P_{t,d}^{-}$) is less than the load ($P_{t,d}^l$). Similarly, to charge only excess power to the BESS, another inequality constraint (14) is introduced.

$$P_{t,d}^{-}\le P_{t,d}^l$$

(13)

$$P_{t,d}^{+}\le \max \left\{P_{t,d}^{pv}-P_{t,d}^l,0\right\}$$

(14)

3.2.4. Internal Trading

The second major component of the different cases analyzed in this study is internal trading (trading among different dwellings in the community). During any interval t, a dwelling d can send power to several other dwellings (d′). The power sent by dwelling d ($P_{t,d}^s$) can be obtained by summing the power sent to all other dwellings d′ ($P_{t,d,d^{\prime }}^s$) and can be represented by (15). Similarly, the power received by dwelling d ($P_{t,d}^r$) during interval t from all other dwellings d’ ($P_{t,d^{\prime },d}^r$) can be represented by (16). Finally, to balance the power of the community, the total internal power trading needs to be balanced, i.e., total sending must equal total receiving. It can be mathematically modeled by (17).

$$P_{t,d}^s={\displaystyle \sum _{\begin{array}{l}{d}^{\prime }\le D\\ d^{\prime }\ne d\end{array}}{P}_{t,d,d^{\prime }}^s}$$

(15)

$$P_{t,d}^r={\displaystyle \sum _{\begin{array}{l}{d}^{\prime }\le D\\ d^{\prime }\ne d\end{array}}{P}_{t,d^{\prime },d}^r}$$

(16)

$${\displaystyle \sum _{d\in D}{P}_{t,d}^s}={\displaystyle \sum _{d\in D}{P}_{t,d}^r}$$

(17)

In addition to the generalized constraints, additional constraints are introduced, aimed at ensuring smooth community operation. To prevent selling power while receiving power from other dwellings, Equation (18) is introduced, ensuring that the sum of received power ($P_{t,d}^{re}$) and bought power ($P_{t,d}^b$) must be less than or equal to the energy deficit ($P_{t,d}^l-P_{t,d}^{pv}$). Similarly, to prioritize self-consumption of renewables, Equation (19) is introduced, implying that dwellings can only share excess power. Lastly, to prevent double usage of excess power for both sending and charging it to the BESS, Equation (20) is introduced, implying that the sum of power charged ($P_{t,d}^{+}$) and sent to other dwellings ($P_{t,d}^{sd}$) must not exceed the excess power in that dwelling.

$$P_{t,d}^{re}+P_{t,d}^b\le \max \left\{P_{t,d}^l-P_{t,d}^{pv},0\right\}$$

(18)

$$P_{t,d}^{sd}\le \max \left\{P_{t,d}^{pv}-P_{t,d}^l,0\right\}$$

(19)

$$P_{t,d}^{+}+P_{t,d}^{sd}\le \max \left\{P_{t,d}^{pv}-P_{t,d}^l,0\right\}$$

(20)

4. Performance Evaluation

In this section, the performance of the proposed method is evaluated for a community comprising five dwellings. The developed linear programming model is implemented in Python with the integration of the optimization tool CPLEX [26]. A scheduling horizon of one week (T = 168) with a sample period of an hour (t = 1) is considered. It is important to note that this study considers PV power due to the widespread adoption of rooftop solar PV systems in households. Nevertheless, the developed model is designed in a generalized manner and can be applied to other renewable resources, such as wind energy.

4.1. Input Data

The load consumption data of five actual dwellings in Vancouver, Canada, as reported in [27], have been utilized. Solar irradiance information from the same locality has been used to estimate the solar PV power. The hourly data for a selected summer week of the five dwellings are depicted in Figure 2. It is noticeable that D1 and D2 have lower loads compared to D3–D5. Moreover, the hourly PV power of the dwellings is illustrated in Figure 3, indicating that only dwellings 4 and 5 (PV4 and PV5) possess solar PV systems, while D1-D3 do not have any solar PV installations. Additionally, due to the similar physical locality, the PV profiles of D4 and D5 exhibit similarity, although their magnitudes differ; the solar PV capacity in D4 is 3 kW, whereas in D5, it is 5 kW.

4.2. Comparative Analysis

In this section, a community consisting of five prosumers is examined, wherein two prosumers possess solar PV systems. The analysis focuses on evaluating the performance of six cases with the aim of maximizing the self-consumption of renewables for a specific summer week, as previously discussed. These six cases are base, IT only, LSS only, CSS only, IT + LS, and IT + CS. Different indices are proposed and used to measure renewable consumption, grid dependence, and community cost, as detailed in subsequent sections. Furthermore, an analysis of different seasons is conducted in the following section.

4.2.1. Performance Indices

Two indices are devised for assessing the grid trading and renewable consumption during each case. The objectives of these indices are to quantify the amount of power traded with the grid and the amount of renewable power consumed locally for each case. The grid trading index ($G{T}_d^{}$) is computed as follows:

$$G{T}_d^{}=100\cdot \left(\frac{P_d^b+P_d^s}{P_d^l+P_d^{pv}}\right)$$

(21)

It computes the total trading amount (during the entire scheduling horizon) of dwelling d, which includes buying ($P_d^b$) and selling ($P_d^s$) with respect to its load ($P_d^l$) and available PV power ($P_d^{pv}$). It should be noted that all terms used in this index are accumulated values for the entire scheduling horizon (1 week in this case). Similarly, the renewable consumption ($R{C}_d^{}$) index is computed as follows:

$$R{C}_d^{}=100\cdot \left(\frac{P_d^l-P_d^b}{P_d^l}\right)$$

(22)

where $P_d^l \mathrm{and} P_d^b$ are, respectively, the load and buying power from the grid by dwelling d during the entire scheduling horizon.

The results of the external trading index for each case are shown in Figure 4, while the grid dependence results are illustrated in Figure 5. It is evident that the highest amount of trading occurs during the base case due to the absence of a BESS and the inability to share. Consequently, the lowest amount of renewable consumption is observed for the base case, as shown in Figure 5. In contrast, the lowest amount of external trading and the highest amount renewable consumption are observed for the IT + CS case. This is attributed to the capability of direct sharing among dwellings and the utilization of CS as a communal buffer for all dwellings. Remarkably, the CSS-only case outperforms the other two cases (IT-only and LS-only) when considering a single technology. Furthermore, the CS-only case even performs better than the IT + LS case. This is because dwellings without PV systems cannot utilize the PV power outside the available PV hours during the IT-only, LSS-only, and IT + LS cases. Conversely, in the case of CS and CS + IT, excess power can be stored during surplus intervals and utilized during PV absence intervals by all dwellings.

4.2.2. Component Utilization

In this section, analysis of the energy utilization from the two technologies (BESS and IT) is discussed. It is important to note that the BESS encompasses both LS and CS, depending upon the case being examined. The BESS utilization for each case is depicted in Figure 6, while the utilization of IT is presented in Figure 7. Notably, no BESS is utilized in the first two cases (base and IT only) due to the absence of energy storage in these cases. The maximum BESS utilization is observed in the CS-only case. This is primarily due to the restriction on IT in this case, resulting in the full utilization of the BESS. Further details regarding the charging/discharging and energy status of the LSS and CSS during the last four cases are provided in Appendix A (Figure A1, Figure A2, Figure A3 and Figure A4). The four cases include LS only, CS only, IT + LS, and IT + CS. In the case of storage-only scenarios (LS and CS), energy and power profiles are shown. However, in the combined cases (IT + LS and IT + CS), in addition to the power and energy profiles of storage elements, power trading profiles are also presented.

Similarly, no IT is observed for the base, LS-only, and CS-only cases due to the absence of IT in these cases. The largest amount of IT is observed in the IT-only case, owing to the lack of any storage system in this case. Regarding total energy consumption (internal trading and BESS), the highest amount of energy is utilized in the IT + CS case. This aligns with the findings from the previous section, where the IT + CS case demonstrated the highest amount of renewable consumption. Further details regarding the power shared among different dwellings under these three cases (IT-only, IT + LS, and IT + CS) are provided in Appendix B (Figure A5, Figure A6 and Figure A7).

4.2.3. Community Cost Analysis

In this section, the weekly electricity cost of the entire community is analyzed for all six cases, and the results are depicted in Figure 8. The highest community cost is observed for the base case, primarily due to the incapability to share power among dwellings. Conversely, the lowest cost is observed for the IT + CS case. This is attributed to the capability of both direct power sharing and storing energy in a central BESS, subsequently utilized by all dwellings. A cost reduction of over 60% is observed for the IT + CS case in comparison to the base case.

4.3. Performance during Different Seasons

In this section, a representative week from all four seasons of the year is considered. The proposed indices, external trading, and renewable consumption are analyzed for each case across these seasons. The results of external trading are presented in Figure 9, while those of renewable consumption are displayed in Figure 10. Additionally, the computed cost for the entire community during the selected week is depicted in Figure 11.

It is evident from Figure 9 that the IT + CS case outperforms all other cases across most seasons, except for winter. During winter, there is limited excess PV power available, and households with PV systems tend to self-consume the generated power. Consequently, no power sharing occurs during winter, resulting in consistent external trading values across all cases. This trend can also be observed in renewable consumption, as demonstrated in Figure 10, where the IT + CS case outperforms others across most seasons, except for winter.

The analysis of cost reduction involves comparing the cost of each case with that of the base case. As depicted in Figure 11, the community experiences the greatest cost reduction during the summer season due to the abundance of PV power. Conversely, the lowest reduction is observed during the winter season, attributed to lower solar irradiance and fewer daylight hours. The cost reduction during autumn and spring falls between these extremes due to moderate load and PV power availability

5. Discussion and Analysis

In this section, the impact of various parameters is analyzed to ascertain the most suitable case for maximizing the consumption of renewables. This includes the analysis of different storage cases (LSS and CSS) and techno-economic analysis of different BESS types.

5.1. Impact of Energy Storage Capacity

5.1.1. Local Energy Storage

In this section, the LSS size is varied from 5 kWh to 20 kWh, with a step size of 5 kWh, and four cases are simulated. The results are presented for dwelling D4 and D5 only, as other dwellings do not utilize their LSSs due to the absence of PV power. The state-of-charge (SOC) profiles for D4 under different cases are depicted in Figure 12, and those for D5 are shown in Figure 13. The operational range of the SOC for each LSS is set to {20, 80}%.

Observing the first case (5 kWh), it is evident that the SOC nearly reaches the upper boundary (80%) each day due to the smaller LSS capacity. However, for the last two cases (15 kWh and 20 kWh), both dwellings maintain their LSS below the 80% limit on most days. This suggests that increasing the LSS size beyond 10 kWh does not notably impact renewable consumption. This observation is corroborated by the community cost, as illustrated in Figure 14. The initial two cases exhibit a steeper reduction in the community’s cost. Nonetheless, this reduction becomes less pronounced in the last two cases due to a smaller increase in renewable consumption with the enlargement of the LSS size.

5.1.2. Community Energy Storage

In this section, the CSS size was varied from 20 to 100 kWh in increments of 20 kWh, and five cases were simulated. Figure 15 illustrates the SOC of the CSS for each case, considering a summer week for analysis. Similar to the previous case, the SOC range for the CSS is set to {20, 80}%.

Observing the initial three cases (20, 40, and 60 kWh), it is evident that the SOC frequently reaches the upper limit (80%) across most days, indicating an excess of renewable power within the community. However, for the latter two cases (80 and 100 kWh), the SOC remains notably below the 80% threshold for most days. This suggests that an 80 kWh CSS is sufficient for this community during the summer season. Considering other seasons, there would likely be less excess power, aligning with the previous section’s discussions. The community cost analysis, presented in Figure 16, demonstrates a sharp reduction in cost for the first three cases. Interestingly, the community cost remains consistent for the last two cases. This reinforces the conclusion that an 80 kWh CSS is adequate for the community, with further CSS size increases showing no significant impact on renewable consumption or community cost.

5.2. Techno-Economic Analysis of Different BESSs

This section analyzes the techno-economic suitability of four commonly used battery technologies in energy arbitrage applications [28]. These technologies include Li-ion, lead–acid, Ni-Cd, and flow batteries. The technical parameters of all four battery technologies are shown in

Table 1. Technical parameters of different BESS types [25].

Battery Type	Life	Energy Density	Roundtrip Efficiency	Power Density
	(Cycles)	(Wh/L)	(%)	(W/L)
Li-ion	1000–20,000	94–700	75–100	56–10,000
Lead–acid	200–2500	25–100	70–90	10–700
Ni-Cd	1000–3500	15–150	60–90	0–33.4
Flow batteries	>10,000	15–70	65–85	15–700

. It can be observed that flow batteries have the longest lifespan but the lowest energy density. Li-ion has the highest efficiency, energy density, and power density. The cost components of different BESS technologies are summarized in

Table 2. Cost components of different BESS types [25].

Battery Type	Power	Energy	Operation and Maintenance
	(EUR/kW)	(EUR/kWh)	(EUR/kW/y)
Lithium-ion	490–4000	170–3800	0–7
Lead–acid	260–1900	90–450	3.4–50
Ni-Cd	450–1500	700–2400	11–20
Flow batteries	550–2500	130–6000	8–70

. The operation and maintenance cost of these technologies are lower compared to capital investment costs. Li-ion batteries have the highest capital cost (accumulated energy and power costs), while lead–acid batteries have the lowest cost. However, the lifespan of lead–acid batteries is the shortest of all (Table 1). In addition, environmental factors also need to be considered when selecting a specific technology.

6. Conclusions

This study conducted a comparative analysis of various cases, including energy storage (local and community-based), internal power trading, and their combinations, to assess their impact on a community’s renewable power consumption. Six cases, including the base case, were simulated. An optimization model, incorporating specific constraints for each case, was developed to evaluate performance under diverse settings. The simulations revealed that combining community energy storage with internal trading yielded the highest consumption of renewables and the lowest community cost. Notably, there was a cost reduction of over 60% observed during the summer season compared to the base case (without storage and trading mechanisms). Seasonality analysis demonstrated that community storage with internal trading significantly reduced grid dependence, with the smallest cost difference observed during winter seasons, approximately 2%, due to reduced PV power and shorter daylight hours. Sensitivity analysis of different storage cases indicated that for community storage, increasing the battery size beyond 80 kWh did not notably impact the community in terms of operational cost and renewable consumption. Similarly, for local storage, a battery size exceeding 15 kWh showed no significant effect on the community cost. Techno-economic analysis showed that Li-ion batteries offer high efficiency but come with higher costs, while lead–acid batteries, although more economical, have a shorter lifespan, therefore emphasizing the need for a balanced consideration of performance and cost in selecting a battery technology.