A Distribution Static Synchronous Compensator Application to Mitigate Voltage Variation for Distribution Feeders

Abstract

With the growing penetration of distributed energy resources (DER), the accompanying challenges have led utilities to limit the hosting capacities of DER installations on distribution feeders. A distribution static synchronous compensator (DSTATCOM) is a power electronic device to provide dynamic injections and absorption of reactive power into the distribution grid with more flexible and reliable voltage control and power quality improvement. A distributed energy resources management system (DERMS) is developed to provide more effective control of a DSTATCOM that can help substantially increase hosting capacity and mitigate overvoltage problems with the existing feeder. A Taiwan power company’s (Taipower) feeder is selected for computer simulation, and the DSTATCOM is employed in the test feeder to demonstrate the effectiveness of the DSTATCOM in improving the overvoltage problems. The voltage/reactive power (Volt/VAR) control of the DSTATCOM helps reduce overvoltage/voltage fluctuations as the DER output increases.

1. Introduction

By implementing the Four-year Wind Power Promotion Plan and Two-year Solar PV Promotion Plan, Taiwan is projected to produce 20 percent of its electricity from renewable energy sources by 2025. Of that 20 percent, nearly 66.3% will be solar power, including 3 GW of rooftop solar PV and 17 GW of utility-scale PV systems. By the end of 2022, Taiwan’s accumulated PV capacity reached 9.7 GWp, while the total solar power generation amounted to 10,675 GWh [1].

Most PV systems in Taiwan are connected directly to the medium voltage (11.4 kV) and low voltage (220/380) levels of distribution networks, and the installation capacities of PV systems in some high-penetration feeders reach the limit of 5 MWp. Reverse power flow on a distribution system may occur during light load and high PV generation times. In the peak solar period, if the penetration of PV systems increases, or a significant number of PV systems are installed on a rural distribution feeder with higher impedance, there is a possibility of exceeding the operational limit for voltage. During planned maintenance or service interruptions, load transfer is executed between interconnected feeders, extending feeder length and increasing total PV capacity. The voltage level of the system could reach 1.1 pu, which may cause a severe issue related to safety.

Recently, some researchers have been dedicated to addressing the voltage rise issue. Two distributed techniques utilizing nodal sensitivities to regulate PV inverters are introduced and evaluated compared to the distributed approach adhering to IEEE-1547 but employing arbitrarily selected droops and a centralized optimal power flow-based method [2]. Smart PV inverters are used for system voltage regulation by providing compensatory reactive power. Active power curtailment methods have been suggested to limit the active power generated by PV systems connected to distribution feeders during periods of intense solar irradiation [3,4,5]. The impact of grid-connected smart inverters with Volt/VAR, Volt/Watt, and dynamic reactive power control capabilities on the voltage of distribution feeders was analyzed in reference [6]. The technique of sequential quadratic programming (SQP) was employed in [7] to achieve the optimal settings of Q(V) and P(V) in a multi-phase network. The optimal selection of power factor droop parameters in PV inverters aims to coordinate step voltage regulators and provide reactive power support. This coordination effectively mitigates voltage issues and reduces excessive tap operations of step voltage regulators [8]. In Taipower’s distribution systems, reference [9] presents an economic evaluation of a hybrid voltage control scheme with active power curtailment to enhance the penetration of PV energy.

Although solar inverters provide various control modes to mitigate the overvoltage problem of high-penetration renewable feeders according to the above research results, the following aspects are considered when assessing solar inverters for utility voltage control and power quality management.

• Most PV inverters and non-utility-owned equipment, cannot provide the utility demand of operational reliability;
• The customer-owned inverters could not establish communication or coordination with the utility control system;
• The solar inverters lacked adequate capacity (inverter capacity/solar panel capacity < 1.0) to resolve the power quality problems fully;
• The effectiveness of mitigating issues on the feeder is limited due to the fixed placement of inverters within the solar plant;
• The utility could not ensure accurate inverter configurations. In practical applications, solar inverters exhibit significant inconsistencies, ranging from variations in sizing, settings, and the quality of implementation.

Other voltage regulation equipment such as substation on-load tap changers (OLTCs), line regulators, and capacitor banks can also mitigate overvoltage issues due to high distributed generation (DG) integration. The OLTC control could have been under-compensated, and low voltages could result in the feeder circuits due to reverse power situations. The OLTC can only operate the number of times required daily with a loss of life and increased maintenance cycles. The reactive power has to be injected and removed at times measured in cycles to regulators, and switched capacitors are too slow. For example, a temporal voltage drop could occur on the distribution networks when a passing cloud drives a sudden drop in PV generation. Regulators and switched capacitors may take seconds to minutes to respond to the dramatic voltage drop.

A distribution static compensator (DSTATCOM) performs distribution-class voltage regulation. The DSTATCOM provides a continuous and rapid response to correct for voltage issues by injecting or absorbing reactive power. The DSTATCOM, a power electronics device, can adjust its output thousands of times per second. Unlike other voltage regulation equipment, it is not restricted by the number of operations and can be easily controlled by the utility. A case study of the Awada industrial zone in Ethiopia is considered, to show the practical applicability of the proposed Stockwell transform support vector machine [10]. In [11], the authors proposed an operational planning approach to determine the optimal allocation of WT units, PV systems, and hybrid energy storage systems (HESS) in smart grids. The power quality improvement assessed by positioning DSTATCOM was published in [12]. Integration of fixed-step capacitor banks and DSTATCOMs in radial and meshed distribution networks was presented in [13]. The authors developed a stochastic mixed-integer convex (SMIC) model to solve the optimal placement and sizing of DSTATCOMs in electrical distribution networks [14]. The authors of [15] apply the generalized standard distribution optimizer to perform optimal reactive power compensation via DSTATCOMs in electrical distribution systems. The problem of controlling a three-phase three-wire PV-type DSTATCOM is investigated by the authors in [16]. The authors of [17] used adaptive particle swarm optimization (APSO) and hybrid grey wolf particle swarm optimization (GWO-PSO) to execute the concurrent planning of multiple distributed generations (DGs), consisting of solar DG and DSTATCOM with reconfiguration in the radial distribution network. The authors demonstrated the reduced sensor-based operation of PV-DSTATCOM in [18]. In the study, the DSTATCOM is employed in a distribution system to eliminate the connection bottlenecks of larger distribution solar plants and to increase the hosting capacity for DG with existing feeders.

In conventional impact analysis, the operational state of distribution feeders is established through a load aggregation process that relies on the statistical characterization of their loads. This process incorporates monthly customer energy consumption, classification, and the typical load profile associated with each customer class [19,20]. Real-time monitoring of the distribution system operating state is essential for developing a comprehensive real-time distribution network model and determining the appropriate control strategy for DSTATCOMs during load transfer between two feeders with high levels of renewable energy sources. The study’s novelty is that it develops a distributed energy resources management system (DERMS) of a distribution system that collects the bus voltage and the power generation of PV systems from renewable energy terminal units (RETUs). The DERMS integrates the existing distribution management systems, including a distribution mapping management system (DMMS) and a distribution automation system (DAS), to derive the control strategy of DSTATCOMs to support the various application functions of renewable energy management. The Volt/VAR control mode of the DSTATCOM is derived by sensitivity analysis and downloaded to PV systems prior to the implementation of load transfer.

This paper makes the following contributions:

(1)

Integration of the DMMS and DAS databases with the real-time data from RETUs to improve the impact analysis on distribution feeders with high penetration of PV systems;

(2)

Automatic generation of the distribution feeder configuration by retrieving Taipower’s DMMS to support the impact analysis of PV integration;

(3)

Derivation of the Volt/VAR control mode of a DSTATCOM before executing load transfer between the interconnected feeders.

The rest of this paper is organized as follows. Section 2 provides a concise overview of the distributed energy resources management system. Section 3 introduces the control algorithm for DSTATCOM. Section 4 demonstrates the practical implementation of the assessment approach by conducting computer simulations of DSATACOM operation on an actual Taipower distribution feeder. The effectiveness of DSTATCOM in addressing voltage fluctuations caused by PV power generation has been demonstrated. A conclusion is given in Section 5.

2. Distributed Energy Resources Management System

This study develops a DERMS to provide more effective control of a DSTATCOM that can help substantially increase hosting capacity and mitigate overvoltage problems with existing feeders. Figure 1 shows the configuration diagram of DERMS. The common information model (CIM) provides a standard definition for information exchange between DERMS and these systems. RETUs are installed at the solar plants to collect operation information such as voltage, current, and PV power generation, and upload the data to the upstream data concentrator unit (DCU) using LoRa communication. The DCU then transmits the data back to the DERMS via 4G communication to perform an impact analysis of the PV integration.

2.1. Feeder Network Topology

To increase the hosting capacity of renewable energy, implementation of the AM/FM/GIS database is utilized to streamline the input data for the three-phase load flow analysis. This involves creating distribution network models and executing the customer-to-transformer mapping process. The Oracle database within Taipower’s AM/FM system offers the functionality to combine graphical representations of components with spatial relationships and information management. The database schema incorporates multiple hierarchical levels to accurately represent the distribution system objects, assigning appropriate data types to attributes based on the data requirements of the application functions it supports. Topology processing involves analyzing the attributes of a network connectivity model and the dynamic switch statuses in an AM/FM database to determine the network configuration of distribution feeders. The system network configuration is determined and updated based on the status of line switches by examining the connectivity table FROM and TO fields. These fields indicate the upstream and downstream devices for each component. By tracing this information, the network configuration is adjusted accordingly.

The components of a distribution system are categorized as either branches or nodes. A branch represents an arc that connects two nodes, and a node acts as a pivotal point where multiple branches converge, allowing for the attachment and connection of various branches. A branch refers to any device with two terminals: line sections, switches, transformers, and more. On the other hand, a node is an electrical point that serves as a connection between various branch devices. To perform topology tracing, begin by selecting a specific node or branch. Proceed by designating the other node connected to the branch as the new focal point. The process continues by considering all branches linked to the new node as the updated branches to be explored. Topology processing concludes upon reaching an open tie switch or tracing all devices.

Following the completion of topology processing, the next step involves applying node reduction to eliminate redundant elements like power fuses, poles, high voltage mold joints, and other unnecessary facilities for the hosting capacity analysis and the decision-making of the DSTATCOM control. Data validation is employed to confirm the radial network configuration of distribution feeders, while the impedances of line segments are determined by considering the conductor types and branch lengths.

2.2. Renewable Energy Terminal Units

This study has accomplished the development of a renewable energy terminal unit (RETU). Figure 2 illustrates the functional block diagram of the RETU, comprising an analog processing unit based on a microprocessor and an embedded system dedicated to data processing. The microprocessor retrieves analog signals of voltage and current from the energy metering IC to perform data conversion and calculate power parameters. The microprocessor utilizes an analog/digital converter and a temperature sampling circuit to derive the temperature of the distribution transformer connected to a PV plant. The embedded system utilizes a universal asynchronous receiver/transmitter to transmit the power parameters and temperature data to the data processing unit. Power generation from a solar plant and the voltages at PCC are subsequently recorded in the connected storage medium. They are then transmitted by the RETU to the data concentration unit (DCU) using LoRa RF communication on a minute-by-minute basis. The DCU gathers all of the collected voltages and sends them to the DERMS through public 4G communication for the impact analysis and the voltage regulation of the DSTATCOM.

2.3. Hosting Capacity Analysis of Renewable Energy

The hosting capacity can be described as the maximum amount of PV that can be integrated without exceeding voltage variation or reliability limitations and without necessitating any upgrades to existing infrastructure or equipment. The hosting capacity relies on factors such as the length of the feeder, PV installations, and specific utility requirements that need to be fulfilled. In this paper, the hosting capacity of distribution feeders is analyzed by collected data, including feeder topology, loadings of service zones, and the power generation of PV systems. According to Taipower’s interconnection guideline of renewable resources, the voltage variation at the point of common coupling (PCC) is determined by calculating the difference between the condition with PV and the condition without PV using the following equation:

$$\Delta V=\frac{V_{\mathrm{wt}\mathrm{PV}}-V_{\mathrm{wo}\mathrm{PV}}}{V_{\mathrm{wo}\mathrm{PV}}}\times 100\%$$

(1)

where $\Delta V$ is the voltage variation (<3%), $V_{\mathrm{wt}\mathrm{PV}}$ is the voltage of PCC with interconnecting all PV systems, and $V_{\mathrm{wo}\mathrm{PV}}$ is the voltage of PCC without interconnecting all PV systems.

With increasing penetration levels of solar PV, the need for voltage regulation technology to provide grid support has become increasingly critical. The DTSTCOM model and Volt/VAR control function are established and simulated for determining the PV hosting capacity of distribution feeders. This study aims to evaluate the impact of DSTATCOM function capability on feeder hosting capacity for solar PV. The flowchart of hosting capacity analysis with and without DSTATCOM is illustrated in Figure 3.

2.4. Coordinated Control of the DERMS and Distribution Automation System

The network configuration of a distribution feeder frequently changes due to the operations of line switches. These switches are used to transfer loads between adjacent feeders, enabling service restoration following a fault and executing uninterrupted load transfers during planned maintenance. As the integration of PV systems increases significantly in distribution feeders, there is a growing concern about potential system overvoltage, which is primarily attributed to the surplus power generated by PV installations and the extended length of the feeder after load transfer. Hence, the DAS transmits the status of the line switches to execute load transfer to the DERAMS system. This allows for the reconfiguration of the feeder network and updates the input data file for analyzing the impact of integrating the PV system. The three-phase load flow analysis determines the bus voltages for the new distribution network configuration.

The optimal Volt/VARs setting of the DSTATCOM is derived before executing load transfer. The DSTATCOM maintains the feeder voltage within a specific range at the regulated point by injecting capacitive and inductive reactive power to prevent overvoltage that would otherwise be caused by the switching operation for load transfer. The DNP3.0 protocol is utilized within the public 4G communication system to facilitate two-way communication between the DERMS and the DSTATCOM. After the DERMS issues the control command for Volt/VAR setting and downloads it to the DSTATCOM, the DAS performs switching operations to complete the load transfer.

3. Control Algorithms of DSTATCOM

Increasing integration of solar PV in Taipower’s feeders challenges the management of the distribution grids. It requires an extensive infrastructure upgrade to avoid voltage limit violations. Power-electronics-based DSTATCOMs provide dynamic injections and absorption of reactive power into the distribution network to fully resolve the power quality problems and to achieve the utility-required operational certainty. Three control algorithms, Volt/VAR mode, power factor mode, and reactive power mode, are embedded in the DSTATCOM to provide continuous and step-free reactive power. In Taipower’s application, the DSTATCOM operates autonomously in the Volt/VAR control mode. The updated voltage set points are derived by Volt/VAR control strategy for the DSTATCOM via SCADA before non-interrupt load transfer is executed between two feeders.

3.1. Volt/VAR Mode

The objective of Volt/VAR is to maintain the feeder voltage within a certain range at the regulated point by injecting capacitive and inductive reactive powers. The DSTATCOM system continuously monitors the feeder voltage and injects leading or lagging currents to regulate the voltage dynamically. The regulation is achieved using a piece-wise schedule of reactive power injection as a function of droop control centered about a user-selected set point, as shown in Figure 4. Configurable parameters include a dead band function (near the set point) and saturation function (near the output limits). In this mode, no external sensors are required. The default control utilizes sensors that are incorporated within the DSTATCOM system. The voltage set point can be updated as needed via the SCADA connection.

The variables used for control depend on the installed feeder configuration. For a three-phase grounded, Y-connected DSTATCOM system, this Volt/VAR curve applies to each measured phase voltage and phase reactive power output. Volt/VAR curve points are checked against the DSTATCOM system and distribution grid limits to ensure stable operation. For example, the slopes of the capacitive and inductive reactive power injection curves must have the polarity as shown in Figure 4. If the reactive power output is below the minimum capacitive or reactive power outputs of the DSTATCOM (VoltVar_Min_Cap_Vars or VoltVar_Min_Ind_Vars), the DSTATCOM will transition to a controlled stop to minimize the power losses. In order to prevent oscillation around the minimum reactive power inception points for capacitive and inductive reactive power output, the DSTATCOM implements a voltage hysteresis band around the turn-on threshold.

3.2. Feeder Power Factor Mode

The objective of the feeder power factor mode of the DSTATCOM is to supply reactive power to maintain a constant power factor according to the power factor set point. This mode requires an external current transformer (CT) sensor to measure the power factor to the feeder, as shown in Figure 5. With the measured power, the DSTATCOM closed loop regulates the net reactive powers (as the sum of the reactive power of load and DSTATCOM) to attain the desired feeder power factor reference. Suppose the DSTATCOM’s contribution to meet the reference of power factor control mode (PowerFactor_Ref) target is below the minimum reactive power output of DSTATCOM (PowerFactor_Min_Vars). In that case, the DSTATCOM will transition to a controlled stop to minimize the power losses.

3.3. Feeder Reactive Power Mode

The feeder reactive power mode aims to supply the net feeder reactive power according to the reactive power reference. This mode requires an external CT sensor as the control sensor to measure the reactive power to the feeder, as shown in Figure 6. In this mode, the DSTATCOM closed loop regulates the net reactive power (as the sum of the load reactive power and DSTATCOM reactive power) to attain the desired reactive power reference. Suppose the DSTATCOM contribution to meet the feeder reactive control mode reference (FeederVarMode_Ref) target is below the minimum reactive power output of the DSTATCOM (FeederVarMode_Min_Vars). In that case, the DSTATCOM will transition to a controlled stop to minimize power losses.

4. Case Studies

To demonstrate the operational effectiveness of the DSTATCOM in addressing over-voltage issues, we have chosen the test feeder XD45, as depicted in Figure 7, for computer simulation purposes. The 11.4 V overhead feeder with a total length of 7.3 km is to supply power to residential and agricultural customers. The Feeder XD45 has a combined installed capacity of 1.2 MWp for PV systems. Figure 8 shows daily profiles of PV power generation and net power loading of the test feeder. The solar irradiation dictated the fluctuation in PV power generation, culminating in a maximum of 630 kW. During the period from 9 a.m. to 3 p.m., the PV systems inject significant power to reduce the net loading of the test feeder, which causes a reverse power flow.

4.1. Hosting Capacity Analysis

This study section evaluates the DSTATCOM’s impact on the hosting capacity of the test feeder under a high penetration level scenario. This section provides the characteristics of the test feeder and the hosting capacity analysis results of the baseline, in which the DSTATCOM does not have the Volt/VAR function. The three candidate locations for installing the DSTATCOM are decided by collecting bus voltages from the RETUs in the field and executing the three-phase load flow. A three-phase DSTATCOM of ±1 MVAR is directly connected on the upstream from (Bus 6, 2.5 km from substation, Scenario 2), middle of (Bus 15, 5.4 km from substation, Scenario 3), or downstream (Bus 31, 7.2 km from substation, Scenario 4) from the test feeder to evaluate the hosting capacity. Figure 9 illustrates that the hosting capacities are compared with that of the Baseline (without DSTATCOM, Scenario 1) when a DSTATCOM is installed at the three candidate locations of the test feeder.

In Taipower’s grid codes for PV integration, the maximum hosting capacity of an 11.4 overhead feeder is 5000 kWp, and the voltage variations for all buses should be less than ±3%. In this study, the hosting capacity analysis of the test feeder is based on Taiwan’s PV integration codes. In Figure 9, the green bar represents the range of PV penetration where all the PV deployments are acceptable regardless of PV locations. The red bar represents the range of PV penetration where any PV deployments are unacceptable. The yellow bar in-between represents the range of PV penetration that may or may not cause issues depending on the specific PV deployment (size and location). As seen in Figure 9, the minimum hosting capacities for the four scenarios are 1500 kW, 3800 kW, 4400 kW, and 2200 kW (green bar), and the maximum hosting capacities are 2400 kW, 5000 kW, 5000 kW, and 4600 kW (yellow bar), respectively. More hosting capacity in Scenario 3 is achieved than those in Scenarios 1, 2, and 4 because the DSTATCOM is located near the large-scale solar facility point (Bus 15) of interconnection.

4.2. Voltage Regulation Simulation

Addressing voltage violations caused by the cumulative impact of numerous solar installations is frequently encountered in practice. Figure 10 shows the real-time waveforms captured from Bus 15 of the test feeder in Figure 7. The excess voltage rise (solid line) or overvoltage (>1.05 pu) can occur during maximum solar production and large voltage deviations can occur during partly cloudy days when solar output becomes intermittent. The DSTATCOM is directly connected to the 11.4 kV test feeder (Bus 15). The overvoltage is eliminated (dashed line) when a DSTATCOM is operated at up to 1000 kVAR inductively. The maximum voltage reduces from 1.061 pu to 1.049 pu, and the average voltage reduces from 1.051 pu to 1.044 pu.

5. Discussion

The DSTATCOM is a high-performance voltage regulation solution in distribution systems. In the study, the DSTATCOM is employed at the distribution primary feeder to increase the PV systems’ penetration levels, solve the power quality problem, and eliminate PV interconnection bottlenecks. The hosting capacities increase by 2300 kWp, 2900 kWp, and 700 kWp when the DSTATCOM is employed upstream, in the middle, and downstream of the test feeder. The results show that the DSTATCOM is recommended to be installed near the large-scale solar facility point of interconnection.

Figure 10 shows a dataset of the voltages at the middle of the test feeder (corresponding to Bus 15 in Figure 7). The injection of inductive reactive power by the 1000 kVAR DSTATCOM, utilizing the Volt/VAR control algorithm, ensures that the feeder voltage at the regulated point remains within 1.05 pu.

6. Conclusions and Future Work

As distributed energy resource (DER) penetration increases, the problems caused by DERs are also increasing, leading utilities to occasionally restrict the hosting capacities of distribution grids. These issues comprise voltage flickering caused by fluctuating wind speeds or cloud cover and the challenge of regulating voltage due to rapidly changing generation levels and reverse power flows. The DSTATCOM is an advanced power electronic device designed to effectively tackle steady-state voltage regulation and transient voltage problems in distribution feeders. Numerical results in the test feeder with a DSTATCOM demonstrated that:

(1)

This study develops a DERMS to provide more effective control of a DSTATCOM that can help substantially increase hosting capacity and mitigate overvoltage problems with existing feeders;

(2)

A Taipower feeder is selected for computer simulation. The DSTATCOM is employed at the test feeder to demonstrate the effectiveness of the DSTATCOM in improving the overvoltage problems. This result shows the high-performance voltage regulation capability of the DSTATCOM solution. The high voltage and voltage deviation issues are clearly resolved.

Future works derived from this research could include the following: (i) DSTATCOM placement with optimization algorithms such as a genetic algorithm, ant colony optimization, particle swarm optimization, etc.; (ii) the coordinated control of on-load tap changers and DSTATCOMs for voltage regulation; and (iii) the optimal parameter settings of the Volt/Var control mode of the DSTATCOM for a distribution network with radial and meshed configurations.