The issue of voltage stability has emerged as a prominent concern in ensuring the safety and reliability of large-scale power grids in recent years, necessitating an immediate enhancement of dynamic reactive power reserves. On one hand, the issue of “Strong DC and weak AC” in power grids becomes prominent with the advancement and implementation of DC transmission technology (RAO et al., 2015; LI et al., 2018; RAO et al., 2019). Moreover, due to the dense DC drop point and relatively close electrical distance of the load center at the receiving end, a single AC failure often leads to simultaneous multiple DC commutation failures (LIU et al., 2015; XU et al., 2019). The construction of large-scale new energy bases, primarily wind power and photovoltaic, in remote areas has led to prominent issues such as a weak DC power grid and insufficient short-circuit capacity. Additionally, the decline in dynamic reactive power reserve and voltage stability has become more significant (ZHANG et al., 2011; CHEN et al., 2016). On the contrary, the rapid increase in load on the asynchronous motor at the load center will result in a decline in bus voltage and further depletion of reactive power within the system, ultimately leading to voltage collapse and large-scale power outage incidents. It is imperative to ensure an adequate supply of dynamic reactive power on a large scale to address this insufficiency in voltage support.

The integration of large-scale new energy into the power grid has prompted experts and scholars, both domestic and foreign, to propose the concept of power grid strength as a measure of the interaction between traditional power grids and connected new energy power electronic equipment. This concept is characterized by two dimensions: frequency support strength and voltage support strength (Australian Energy Market Operator, 2020b; Australian Energy Market Operator, 2020a). The voltage support strength refers to the system’s ability to maintain voltage stability during significant disturbances, and it is commonly assessed using the short circuit ratio index (XIN et al., 2020; ZHU et al., 2021; WU et al., 2022). In order to assess the strength of new energy grid-connected systems, scholars have proposed the concept of critical short-circuit ratio (WANG et al., 2019; CAO et al., 2023) as an index for evaluating the short-circuit ratio. The voltage support capability of the system is assessed by comparing the real-time short-circuit ratio with the critical short-circuit ratio. Due to the existence of various short-circuit ratio index construction methods for different Scale of new energy grid connection and new energy access forms (Lin et al., 2022a), there is a challenge in ensuring consistency between the calculation of critical short-circuit ratio index and the boundary value used for system strength assessment (Wu et al., 2018; Lin et al., 2022b; YANG et al., 2023). The short-circuit ratio and critical short-circuit ratio calculation methods proposed in literature (Lin et al., 2022b) can be employed for the evaluation of each new energy station’s short-circuit ratio within the new energy multi-feed system. These methods enable identification of weak links in the new energy cluster system and contribute to formulating strategies for strengthening the weak power grid system. Therefore, this paper primarily adopts these literature-based approaches to assess the strength of the system power grid.

The existing literature on voltage control through reactive power compensation can be categorized into two distinct groups. The first type refers to in situ reactive power compensation (Kirn and Lee, 2005; Abdel-Rahman et al., 2006; Park and Baek, 2012; LIU et al., 2019; Fan et al., 2020; Wen and ZHOU, 2020). The literature (Fan et al., 2020) suggests connecting the fast reactive power compensation device to the heavy-load distribution station and coordinating it with the capacitor bank and on-load regulating transformer. The literature (Wen and ZHOU, 2020) proposed a voltage control technology for distribution stations based on the coordination between SVC and on-load regulator transformers. The literature (Abdel-Rahman et al., 2006) presents a SVC access strategy incorporating two-stage slope regulation and two-stage voltage regulation control. The literature (Park and Baek, 2012) proposes a voltage coordination control method based on local reactive power compensation, which is developed through extensive research on shunt capacitors, reactors, and on-load regulating transformers. The literature (Kirn and Lee, 2005) presents a reactive power compensation approach utilizing artificial neural network. The literature (LIU et al., 2019) presents a centralized/local adaptive reactive voltage coordination control approach for active distribution networks. The limitation of the above method is obvious, the local compensation does not take into account the tidal flow of the power system, and it has little improvement on the economic and stable operation of the power system. The second approach involves coordinating reactive power compensation devices in the power system to achieve system voltage control (FU et al., 2018; NI and LI, 2020; WANG et al., 2020; XIAO et al., 2020; TANG et al., 2021; ZHU and ZHAO, 2021). The literature (XIAO et al., 2020) presents a reactive power optimization model that is based on sensitivity analysis. The characteristics of voltage control in load center SVC were analyzed in reference (WANG et al., 2020). The literature (ZHU and ZHAO, 2021) proposes a voltage control method that utilizes static synchronous compensation and coordinated optimization of on-load regulator transformers for enhanced performance. The literature (TANG et al., 2021) presents a regional voltage control approach that can be applied to the penetration of FACTS devices (NI and LI, 2020). A voltage adaptive coordination control method is proposed in reference (FU et al., 2018). To summarize, the existing studies solely focus on voltage stability as the single objective during stable system operation, neglecting the requirements for system strength in weak grid scenarios of new energy clusters. Hence, this paper aims to enhance the power grid strength of new energy clusters by proposing a multi-type reactive power compensation optimization method that considers constraints related to short-circuit ratios. This approach enables simultaneous consideration of both voltage stability objectives and system power grid strength requirements.

Taking into account the aforementioned considerations, the relationship between the evaluation index of power grid strength and the voltage response performance of the system is examined, and the influence mechanism of various reactive power compensation devices on the power grid strength in a new energy cluster system is analyzed. Subsequently, a model for optimizing reactive power operation is proposed to maximize system grid strength while minimizing voltage deviation. A hybrid strategy combining genetic algorithm and CPLEX solver is employed to solve this problem. Finally, a typical simulation example is presented to validate the effectiveness of the proposed method.

In the power system with high penetration rate of new energy, system strength is a common concern, the system “strong” and “weak” is a relative concept, not simply based on the capacity of the new energy equipment connected to the AC grid to evaluate, that is, can not be simply judged as a new energy grid connected to the small capacity of the system, its grid strength is higher than the new energy grid connected to the large capacity of the power system. The short-circuit ratio is currently widely employed to characterize the system strength of new energy clusters (Lin et al., 2022b), as shown in Eq. 1: ξ SCR = S ac P N = U N 2 P N ⋅ 1 Z = 1 Z pu (1) Where, S ac is the short-circuit capacity of the AC system; P N is the rated capacity of new energy equipment; U N is the rated voltage of grid-connected bus; Z is the system impedance; Z pu is the per unit value of system impedance.

The above traditional short-circuit ratio indicators have two limitations for the new energy multi-feed system. Firstly, it is challenging to establish a standardized benchmark capacity due to the varying installed capacities of new energy in each network. Secondly, when considering the impact of junction voltage, it is necessary to take into account not only the directly connected new energy but also other new energy feeds. The proposed short-circuit ratio index ξ SCR − S based on capacity calculation utilizes the ratio between the AC system’s short-circuit capacity and the equivalent grid-connected capacity of new energy. As shown in Eq. 2. ξ SCR − S i = S ac , i S eq , i = S ac , i S i + ∑ j ≠ i Z i j * Z i i * U i U j S j (2) Where, ξ SCR − S i is the voltage support strength at the i node of the system; S eq , i is the equivalent grid-connected capacity of new energy at the system node i; S j , S i respectively are new energy installed capacity of system nodes i, j; U i , U j respectively are the system i, j node and voltage; Z i i * , Z i j * respectively are self-impedance and mutual impedance of the system; * represents the conjugation operation.

The expression of the relationship between short circuit ratio and junction voltage can be further derived as per the literature derivation (Lin et al., 2022b). K SCR − U , i = U N E ˙ eq , i Δ U ˙ i U ˙ i (3) Where, K SCR − U , i is the voltage support strength at grid-connected point i; Δ U ˙ i is the voltage change at grid-connected point i.

The short-circuit ratio index, as indicated by formula (3), exhibits an inverse relationship with the voltage change at the junction point. A higher short-circuit ratio corresponds to a smaller voltage change at the junction point, indicating enhanced system anti-disturbance capability and increased system strength.

In order to facilitate the analysis of the coupling mechanism between the short-circuit ratio index and the reactive power at the system node, the short-circuit ratio shown in Eq. 2 is calculated and simplified as follows: K SCR , i = S ac , i P i + j Q i + ∑ j ≠ i r ˙ j i P j + j Q j (4) r ˙ j i = Z ˙ j i Z ˙ i i (5) In Eq. 4, K SCR , i is the voltage support strength at the junction point i; P _i and P _j are the active power output values of new energy at connecting points i and j respectively; Q _i and Q _j are the reactive power output values of merging points i and j respectively. R _ji is the influence factor of voltage interaction between j and i.

The focus of this paper is primarily on evaluating the operational strength of the short-circuit ratio system, whereby the actual output of new energy at parallel points is utilized for calculating the equivalent short-circuit ratio, rather than relying on the installed capacity of new energy.

In order to enhance the voltage performance of new energy grid-connected systems and strengthen weak grid systems of new energy clusters, common methods for reactive power compensation include the installation of static var compensators (SVC) and static var generators (SVG), as well as utilizing the reactive power emitted by the new energy equipment itself for reactive power compensation. The various methods of reactive power compensation, however, employ distinct mechanisms to enhance the strength of the system grid. The subsequent section will elucidate the mechanism employed by each method during the calculation process for short-circuit ratio.

The calculation process of the short-circuit ratio index is illustrated in Figure 1. Initially, the admittance matrix of the system is established based on its structure (including lines, transformers, etc.), incorporating the subtransient reactance of conventional units into the admittance matrix while disregarding the impedance of the inverter interface for new energy equipment, then the system impedance matrix is derived by taking the inverse of the admittance matrix. According to the principle of symmetrical short-circuit calculation, the per unit value of node short-circuit current is obtained by taking the reciprocal of the self-impedance of each system node. Consequently, the short-circuit capacity for each node in the system is determined. Finally, formula (4) and 5 are utilized to measure the system strength of each new energy node and obtain their respective short-circuit ratios.

The short-circuit ratio calculation process reveals that the inclusion of various reactive power compensation methods impacts the calculation of each node’s short-circuit ratio, subsequently influencing the determination of new energy output limits for each node and ultimately affecting the absorption capacity of each new energy station. Among them, the input of SVC and synchronous compensator primarily impact the short-circuit capacity of each node in the system by modifying the admittance matrix. However, direct access to tuning the synchronous compensator affects the self-admittance matrix of the access node location. SVC is considered as an electrical admittance matrix parallel to the bus, which explains why synchronous compensator has a better compensation effect compared to SVC. The reactive power of Static Var Generator (SVG) and renewable energy sources mainly increases the reactive power of the system node power supply. Table 1 summarizes the impact mechanism of each reactive power compensation mode on short-circuit ratio calculation.

In order to quantitatively analyze the impact of the aforementioned reactive power compensation methods on the system’s power grid strength, a comparative analysis is conducted using a new energy gathering and transmission terminal system as an example to assess the effectiveness of reactive power compensation methods on K _SCR. The calculation results are presented in Table 2, where WT/PV indicates that the reactive power of the new energy station is used to compensate the reactive power. The calculation results demonstrate that the implementation of any of the four reactive power compensation methods can effectively enhance the short-circuit ratio at the grid-connected node of new energy. The increase in K _SCR for SVC is the smallest among them, with a growth rate ranging from 1% to 8%, and the majority of which are below 5%. The two modes of SVG compensation and new energy reactive power compensation exhibit a similar magnitude of improvement for K _SCR, ranging from 6% to 11%. The addition of the synchronous compensator has a significantly noticeable impact on K _SCR, resulting in an evident improvement. The installation of the distributed synchronous compensator at new energy power stations under the gathering station leads to an enhancement level exceeding 40% for K _SCR, with the highest improvement reaching 83%. Therefore, the utilization of synchronous compensator for reactive power compensation and enhancement of power grid strength has become a prominent focus within the industry.

During conventional operations for optimizing reactive power in a power grid, once the placement of each reactive power compensation device is determined, adjustments are made to their outputs (e.g., SVGs and synchronous compensators) in order to comply with system constraints. This ensures economic and secure operation of the grid while achieving minimal voltage deviations at every node. The analysis of the impact of various types of reactive power compensation devices on the power grid strength in the new energy cluster system, as described in Part 2, reveals that adjusting the output of these devices can effectively enhance the short-circuit ratio and overall system strength. Therefore, the paper comprehensively examines the regulatory impact of reactive power compensation devices on system voltage and power grid strength It proposes an optimized operational mode for reactive power optimization that considers both power grid strength requirements and voltage fluctuations.

The genetic algorithm (GA) is an adaptive global optimization probability search algorithm that simulates the genetic and evolutionary processes of organisms in their natural environment. The variables that need to be solved are encoded into a chromosome for genetic modification, while the objective function corresponding to a set of variable values is transformed into an individual fitness. The initial value is randomly assigned, and after the process of natural selection, individuals that meet the criteria are selected. Genetic operations such as replication, crossover, and mutation are then performed on these individuals to obtain the succeeding generation. The process is iterated until an optimal solution is obtained or the genetic algorithm reaches a predetermined threshold.

When applying genetic algorithms to power system optimization, it is important to recognize that the power flow solution within the system is subject to various constraints. The effectiveness of these solutions is evaluated based on an objective function, whereby those with lower evaluation values are discarded while only those with higher evaluation values have the opportunity to iterate their characteristics into subsequent rounds of solutions. Ultimately, this iterative process aims towards achieving optimization.

The paper analyzes the influence mechanism of multi-type reactive power compensation devices on the power grid strength of new energy cluster systems, and proposes a reactive power operation optimization model that aims to maximize the power grid strength while minimizing voltage deviation. The specific conclusions are as follows.

(1) The influence of various reactive power compensation methods on short-circuit ratio calculation is analyzed based on the calculation process, and the impact of multi-type reactive power compensation methods on enhancing the power grid strength is quantified.

The next research direction of this paper is to establish a comprehensive reactive power resource optimization planning model, encompassing various types of reactive power compensation measures, in order to enhance the new energy consumption capacity of weak power grids while considering economic factors.