1. Introduction

Nowadays, renewable energy sources (RES) are developing strongly in the entire world ^{[1][2][3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31][32][33][34][35][36][37][38][39][40][41][42][43][44][45][46][47][48][49]}[1–3]. The total installed capacity of renewable power reached nearly 300 GW by 2021, accounting for 29% of total global electricity production ^[1]. This growth was achieved by the rise of solar and hydropower, with the majority of installed capacity coming from solar photovoltaic (PV). Renewable energy investments continue to grow, with solar energy accounting for nearly half of the total global investment in renewables. Financial incentives for renewable power generation continue to be enacted by governments despite the impact of the COVID-19 pandemic and the Ukraine crisis. Moreover, policies are also designed to ensure the successful integration of RES into energy systems, focusing on the flexibility, operational security, and resilience of networks.

Power stations using RES play an increasingly important role in electricity supply, especially in rural and remote areas ^[1][2][3]. Small stand-alone grids based on these energy sources are often one of the most efficient ways to power off-grid areas, especially solar-powered grids. For this reason, during the pandemic, many renewable-energy-based grid projects slowed down, but new mini-solar grids were still put into use. In addition, during the current Ukraine crisis, the competitiveness of RES with fossil fuels is being improved due to high fossil fuel prices. The deployment of RES also aims to reduce the dependence on imported fuels from abroad, thereby strengthening domestic energy security.

Renewable electricity could be supplied for domestic demand both from centralized power plants via the transmission lines and distributed grids, depending on local conditions ^[1][2]. Currently, distributed PV applications in residential and commercial properties are encouraged in order to reduce electricity consumption by final consumers. Moreover, the integration of distributed generation (DG) in networks is the result of the privatization of the electricity market, environmental protection from emissions, and technological advances. Furthermore, these distributed networks also enable households in developing countries to access clean electrical energy. By 2020, tens of millions of people have access to basic electrical services via the use of solar PV systems at home, and about 87% of mini-networks are operated by RES, mainly solar PV ^[1].

However, the presence of DG units in distribution networks also poses serious problems and challenges for them ^[4][5][6]. The issues of voltage violation, overvoltage, and undervoltage are often present in radial distribution networks with DG. This problem was reported to be caused by the DG units being incompatible with existing voltage regulation methods because of the bidirectional power generation. In fact, this is unavoidable in modern distribution networks where the issues of voltage drop and power loss are very important, and the variable nature of RES could negatively impact the voltage profile and thus degrade the quality of power supplied to final consumers. Nevertheless, the presence of DG is also believed to support and improve voltage profiles in these networks.

The findings in documents ^{[7][8][9][10][11]} show that DG units could minimize power losses (real and reactive power) in distribution networks, and the location and size of these DG units play an essential role in eliminating these losses. For this reason, the optimal structure of DG units in these networks should be determined to minimize losses. Several methods to determine the optimal location and size for DG units are introduced in ^{[7][9][10][12][13][14][15]} to improve voltage quality and reduce power losses in distribution networks. The proposed methods show that this could be achieved by choosing the proper size and position for the DG units.

In Vietnam, DG units using RES in distribution networks have been in place since the 2000s ^{[16][17][18][19]}, and the latest policies also clearly demonstrate the Government’s determination in promoting the development of renewable electricity on islands across the country ^[20][21]. In parallel with the issuance of these mechanisms, many renewable power projects are also implemented, including isolated grid projects with RES-based DG units for off-grid areas ^[22], and one of these is the current isolated network on An-Binh Island (Quang Ngai Province) ^[23].

The islands are now considered to have a strategic role in the socioeconomic development and defense-security assurance of Vietnam ^[24], and therefore, it is essential to establish a solid infrastructure for sustainable development on them, especially for energy security. Due to geographical distance and other natural conditions, most of the networks on these islands are not connected to the national electricity system ^[18][25]. The power supply on these islands is completely dependent on diesel power and RES on site. Integrating DG units close to final consumers could reduce losses to distribution networks, but the structure of the DG units needs to be adapted to local demand as the reliability of these networks could be affected by the high integration of RES power supply, which may lead to the overloading of existing elements and other voltage violations. Moreover, vulnerability is also one of the main challenges for island networks, which manifests itself in their ability to overcome disturbances (transient or severe) in actual operation ^[26].

However, in Vietnam, studies on isolated networks are currently very limited. The recent investigations in ^[27] on isolated networks for Con-Dao Island ^[27](Vietnam) were also carried out based on static models for planning and design purposes. Another study of an islanded microgrid in Vietnam ^[28] investigated the frequency response of the system with the change in load and solar power but was built on the basis of simple hypothetical data. According to these studies, these networks also seem to face the same challenges as other RES-based distribution networks, including losses and disturbances in static and dynamic models.

In fact, studying dynamic security for isolated networks does not seem to be popular due to their simplicity. So far, no real-world studies have been conducted to investigate the issues of both static and dynamic security in isolated networks in Vietnam. Even the research results in ^{[4][5][6][7][8][9][10][11][12][13][14][15]} only stop at determining the losses in static security, while the disturbances in dynamic security have not been mentioned. For these reasons, the problems of operational security in these networks could not be detected and addressed appropriately and, as a result, lead to inefficient deployment of energy services.

To fill this gap, this paper intends to survey the operational status of the current An-Binh grid and analyze the optimal operating scenarios based on the local solar potential. Furthermore, the importance of the efficient use and transition of energy on An-Binh Island is also considered.

2. Application

An-Binh Island belongs to the Quang Ngai Province with a total natural area of about 0.69 km² (Figure 1) ^[29][30]. The island is located in the tropical monsoon climate zone with two distinct seasons in a year, a rainy season and a dry season ^[31]. The rainy season starts from August to December and the dry season is from January to July. The solar potential here is quite good, about 4.7 kWh/m² a day with nearly 2430 h of sunshine a year^[32][33].

The population here is quite small with about 500 people with over 100 households ^[29][30]. The economy is mainly agro-fishery, 60% of which involves fishery, 34% agriculture (growing onions and garlic), and 6% others. This industry is not currently robust, whereas tourism is emerging in recent years.

The An-Binh power grid is an isolated distribution network at a voltage of 0.4 kV and has a radial configuration with a unidirectional power flow ^[34] (Figure 2). This power grid has been built since 2016 and is expected to achieve success in projects to transition energy and improve energy efficiency for people on the island ^[23][30].

The power supply on this island comes from diesel and solar ^[34]. Currently, the power generation mix consists of two diesel generators with a total capacity of 176 kVA (2 × 88 kVA) and ground-mounted PV systems on-site with a total capacity of 96 kWp. The lines of the An-Binh grid are mostly overhead with a total length of nearly 2700 km.

The power demand between the two seasons is markedly different, high in the dry season and low in the rainy season ^[34]. The monthly demand in the dry season is between 25 and 40 kW with the highest being 55.7 kW in May, while in the rainy season, the monthly demand is from 10 to 25 kW, with the lowest at 8 kW in December. Besides that, the peak hours for daily consumption also differ between the two seasons, from 9 a.m. to 4 p.m. in the dry season and from 6 p.m. to 10 p.m. in the rainy season.

Figure 1. An-Binh Island map [29,30].

Figure 2. An-Binh power grid in NEPLAN software ^[34].

3. Proposed Methods

Power systems are often characterized by uncertainty in the processes of generation, transmission, and consumption ^[35][36][37], especially in the context of the high penetration of RES today ^[1][38]. The findings in ^{[35][36][39][40][41][42]} suggest several models for determining the electricity security of a network. Among them, the models of static and dynamic power systems detail the components of a network with endogenous factors/variables toward ensuring the operational security of the infrastructure.

As mentioned at the beginning of this study, given the characteristics of these models and our approach to infrastructure operational security in the short term, static and dynamic models are interesting options for the An-Binh power grid (Figure 3). These models could show the ability of a network to overcome disturbances and return to the normal operating state, as well as represent problems related to the reliability of the system such as power losses or voltage drops.

In this study, a static model was designed based on hourly data, and a dynamic model was embedded in this static model and surveyed with a time frame of several tens of seconds. The electrical value chains examined in the static model were power flows on lines and voltage profiles at nodes, while those in the dynamic model were the responses of frequency and voltage of the system in the face of disturbances. The results obtained were used to evaluate the stability and security of the existing system. Furthermore, these results were also analyzed and compared to choose the most optimal structure for the power mix of the system.

Figure 3. Flowchart of the method proposed.

In order to evaluate the reliability and vulnerability of the An-Binh grid to extreme conditions, 24 h of a peak-load day in May and a bottom-load day in December were selected for simulation in this study. Besides that, two different sizes and locations of the generation mix (PV systems and diesel generators) were identified and simulated on both of the dates mentioned above. For the selected locations, location (1) was sited at the current site of the generation mix, while location (2) was another option for this mix and was located at the point of direct connection to other main branches of the system. For the selected sizes, the scale of 96 kWp solar and 74 kW diesel is the current size of the generation mix, and the other scales were assumed for both solar and diesel in order to determine the operational security of the system at a lower cost. The scenario without PV systems, scenario S1, was added as an important option to better define the impact of these systems on other components in the grid.

Table 1 and Figure 4 show the scenarios designed for these simulations.

Figure 4. Diagram of the An-Binh grid in NEPLAN software.

Table 1. Summary of scenarios for the An-Binh grid.

4. Critical Contributions

It is self-evident that the ability to match demand reflects the generation adequacy of the An-Binh grid at the reference time ^[43]. The results show that the An-Binh grid completely meets the current demand of the island community, even with excess capacity up to double the needs during peak hours of the summer peak day. The power loss on lines and the voltage profile at nodes are still within the allowable range of the Vietnamese grid code but not the best solution for the current system.

Solar PV systems integrated into this grid allowed for a reduction in power loss and improvement in the voltage profile at nodes, but not all locations or sizes could achieve that. The central location (2) seems to be the most adequate one and also the weakest point in the structure and operation of this system ^[44][45][46]. The structural weakness of this location refers to the potential for damage to the system when it is disconnected from other components in the grid, whereas its operational weakness focuses on the cascading faults of overload and overvoltage.

The findings in this study also suggest that location (2) is a good choice for the generation mix in improving the voltage profile at nodes but not for the power loss problem of the entire grid. In all the scenarios with PV systems linked at location (2), the voltage profile at nodes was improved compared with the PV systems connected at location (1), but this improvement was not significant between the simulated PV system sizes.

For the power loss, there was no difference between the two locations selected with the same size of PV systems. Even the presence of PV systems at location (2) posed a loss challenge for the An-Binh grid, even much higher than in other scenarios. This problem is attributed to the incompatibility due to bidirectional power generation from DG units, diesel generators, and PV systems set at two different locations [4–6]. However, at the same location, the smaller the PV system size, the lower the total power loss. This also shows that the An-Binh grid could operate well, with a total generating capacity lower than the current capacity.

Besides the point indicated about the static security of the An-Binh grid, dynamic analysis on this grid once again confirmed the challenges of integrating RES into modern networks [44–46], specifically solar power here. The integration of PV systems increased An-Binh’s vulnerability to both transient and severe disturbances, and depending on the extent of PV production in the generation mix, the degree of this vulnerability was different.

In our experiments for the transient disturbances on the An-Binh grid, the presence of PV systems contributed to enhancing the resilience of the system from faults, but they could also exacerbate the impact of these events on the system. Moreover, the implementation of PV systems is considered to have a certain impact on the recovery time of networks [26]. These impacts were positive or negative depending on their size and location in this study in the 3 s fault clearance experiment. This also shows that smaller size and location (2) would be good options for PV systems.

In our experiments considering the severe disturbances on the An-Binh grid, most of the scenarios with PV systems before fault clearance did not meet the frequency stability within the range specified in the Vietnamese grid code ^[47][48], while the voltage recovered better than in the fault-clearing experiments. The active power imbalance in this case seems to have caused the frequency of the grid to deviate completely from the rated value and lead to voltage instability due to fluctuations in the reactive power [26]. Scenario S7 was the worst for both the frequency and voltage stability of the system, whereas scenario S5 was the best in this study for ensuring dynamic security. The frequency and voltage responses after troubleshooting in scenarios S4 and S6 were also quite good but not in the remaining scenarios. These results once again show that choosing the right location and size for the power mix are very important to improve the resilience of the existing An-Binh grid.

Based on the experiments and analysis above, the best option for the current An-Binh grid would be scenario S5, in which the power mix of 48 kWp solar and 74 kW diesel is located at the grid center point. In addition, protection systems should be established to clear faults within a critical time of 1 s for ensuring the operational security of the system. Moreover, compared with the current An-Binh grid (scenario S0), scenario S6 with 48 kWp solar and 50 kW diesel is also a good option, as it could contribute to saving up to half the total investment cost for the power mix and reducing up to a third of emissions to the local environment ^[49].

5. Conclusions

This study contributes to the existing knowledge of the impact of solar PV systems on the operational security of isolated networks, here the An-Binh Island grid in Vietnam. The results show that the variable nature of solar power can exacerbate the problems of operational security on islanded networks, but these problems can be ameliorated by choosing the right location and size for PV systems.

The studies on static and dynamic models of the current An-Binh power grid are of benefit to stakeholders who may consider and apply these models for other isolated networks in practice. The static and dynamic security analyses in this study also contribute to the assessment of the reliability and vulnerability of these systems.

The evidence from this study suggests that the location and size of the current power mix are not the best options for the An-Binh grid. The adjustment of the size and location of the supply system is really necessary to achieve optimal and sustainable development for this system. However, this study is limited by the use of diesel power to ensure operational security for the current An-Binh power grid. Besides this limitation, the use of 100% renewable electricity with storage was not considered in this work.

For these reasons, these findings provide insights for the next steps in studies on the networks using 100% renewable electricity on Vietnam islands, including emission reduction and environmental protection. Moreover, further research regarding the impact of renewable power in isolated networks would be of great benefit in determining their potential for growth and expansion in off-grid areas in Vietnam.