Abstract: This paper focuses on introducing new products, technologies, and development trends of intelligent low-voltage reactive power compensation devices in China in recent years, and discusses several technical issues in the design scheme, such as the selection of compensation methods, main wiring schemes, thyristor switching circuits, and intelligent automatic controllers. Keywords: Intelligent low-voltage reactive power compensation technology scheme 1 Introduction In recent years, during the implementation of urban and rural power grid transformation, the design scheme of low-voltage parallel capacitor reactive power compensation devices has undergone significant improvements and breakthroughs, achieving satisfactory operational results. It plays a positive role in improving power supply voltage quality, tapping the potential of power supply equipment, reducing line losses, and saving energy. This paper discusses several technical issues and development directions in the development of intelligent low-voltage reactive power compensation devices for reference. 2 Improvement of Low-Voltage Compensation The traditional modes of low-voltage reactive power compensation mainly include the following three types: ① single-unit local compensation installed on low-voltage motors; ② compensation boxes installed on the low-voltage side of distribution transformers; ③ automatic compensation cabinets (such as PGJ cabinets) installed in enterprise power distribution rooms or workshops and high-rise building floor power distribution rooms. Due to space limitations, the issue of single-unit compensation will not be discussed in this article. The technical improvements and new technologies applied to low-voltage compensation boxes and cabinets can be summarized as follows: (1) From three-phase common compensation to phase-by-phase compensation in order to achieve a more ideal compensation effect; (2) From single reactive power compensation to compensation devices that simultaneously have filtering and harmonic suppression functions; (3) From using AC contactors for switching to using thyristor switching circuits for switching, and developing into the optimal switching mode of equal voltage switching and zero current switching; (4) Combining intelligent automatic compensation controllers with the operation recorders of distribution transformers; (5) Incorporating the function of low-voltage compensation into the low-voltage section of box-type substations or American-style box-type substations; (6) Using stainless steel or aviation aluminum plate enclosures, which have the characteristics of cold protection, sun protection, sealing, moisture protection, and rust prevention; (7) Using dry-type or SF6-filled self-healing parallel capacitors to improve operational reliability and extend service life. 3. Wiring Combining Δ-Y Common Compensation and Individual Compensation 3.1 Wiring for Three-Phase Common Compensation Traditional low-voltage compensation uses three-phase common compensation. Based on unified sampling by the controller, each phase receives the same compensation capacity. The wiring for this compensation method is shown in Figure 1. It is suitable for networks with basically balanced three-phase loads and similar cosφ values ​​for each phase. Why do domestic and international manufacturers choose Δ wiring for three-phase common compensation capacitors? Mainly because a 400V self-healing capacitor is much cheaper than a 230V capacitor of the same capacity. This is due to the price of raw materials and the higher inter-electrode electric field strength of the 400V capacitor. Taking a 400V capacitor as an example, with an 8μm thick metallized film, the operating electric field strength is 50MV/m; with a 7μm thick metallized film, the operating electric field strength is 57.14MV/m. For a 230V capacitor, to maintain a similar operating electric field strength, a thinner metallized film must be used. However, 4-5μm films are much more expensive than 7-8μm films. Therefore, 230V capacitors are generally designed with a reduced operating electric field strength. According to domestic pricing, a 230V capacitor of the same capacity costs more than twice as much as a 400V capacitor. 3.2 Three-phase split compensation wiring: Three-phase split compensation involves sampling each phase separately and applying different compensation capacities to each phase. This is suitable for situations where the loads of each phase differ significantly, and their cosφ values ​​also vary considerably. The wiring diagram is shown in Figure 2. The differences from three-phase common compensation are: ① The rated voltage of a single parallel capacitor is 230V, Y-connected; ② The controller operates on separate phases without affecting each other. Of course, its price is higher than that of a three-phase common compensation device, generally 20% to 30% more expensive. 3.3 Connection of Δ-Y Common Compensation and Three-Phase Compensation Combination From an economic point of view, a capacitor Δ-Y connection can also be used, that is, a connection scheme combining three-phase common compensation and three-phase three-phase three-phase compensation, as shown in Figure 3. The capacitors in the three-phase common compensation part are Δ connected, and the capacitors in the three-phase three-phase compensation part are Y connected. For example, the single capacitors of a certain manufacturer's Δ-connected capacitor bank are 400V, 10, 15, 20, and 30kvar. The single capacitors of the Y-connected capacitor bank are 230V, 3, 4, 5, 6, 8, and 10kvar. The compensation device with this connection method is flexible in operation, and its complete set price is lower than the connection scheme in Figure 2. Some manufacturers still use 400V capacitors for the Y-connected capacitor bank, and the nameplate capacity of each capacitor is the same as that of the Δ-connected capacitor bank. However, the actual output capacity of the capacitors in the Y-connected part is only 1/3 of the nameplate capacity. The purpose of this is that 400V products are relatively inexpensive. Even if the actual capacity is smaller than that of brand-name products, they have a longer lifespan due to the lower operating field strength, and the entire device uses only one type of capacitor, making them highly interchangeable. 4. Switching Switch for Parallel Capacitors 4.1 AC Contactors The PGJ compensation cabinets widely used in the 1970s all used AC contactors as switching switches for parallel capacitors, and this is still in use today. Its disadvantages are: ① When connecting capacitors, a high inrush current is generated, which can easily cause sparks at the contactor contacts, burning the contacts; ② When disconnecting capacitors, the contacts can easily stick together, making them difficult to pull apart; ③ Excessive inrush current is harmful to the capacitor itself and affects its service life. The measures adopted at that time were: (1) Selecting a contactor with a larger rated capacity, such as using a contactor with a rated current of 40A to switch a 15kvar three-phase capacitor (IC = 21.7A); (2) Using a dedicated contactor, the models of which are CJ16, CJ19, CJ20C, B25C~B75C, CJ41 and other series; (3) Adding a series small reactor to each capacitor to suppress inrush current. 4.2 Bidirectional thyristor switching circuit The connection of the bidirectional thyristor contactless switching circuit (also known as solid-state relay) to replace the AC contactor for switching capacitors is shown in Figure 4(a). Its advantages are zero-crossing triggering, no arcing, short action time, and can greatly limit the inrush current of capacitor closing, which is especially suitable for frequent switching occasions. However, it also has the following disadvantages: ① The manufacturing cost of using bidirectional thyristors is high. The price of compensation cabinets using thyristor switching circuits is about 70% to 80% higher than that using contactors; ② Thyristor switching circuits have a large voltage drop during operation, and the power loss and heat generation during operation cannot be ignored. Taking the BZMJ0.4-15-3 parallel capacitor as an example, its rated current is 21.7A. If the voltage drop of the thyristor switch is 1V, the power loss when three thyristor switching circuits are running is: P = 3 × 1 × 21.7 = 65.1W. If the reactive power of the compensation cabinet is 90kvar, then when all are in operation, the power loss of the thyristors is 65.1 × 6 = 390.6W. Assuming an average of 10 hours per day, the daily power consumption reaches 3.906 kWh. The annual power consumption is about 1426 kW·h. The heat generated by the active power consumption will also increase the temperature rise of the entire compensation device, and corresponding heat dissipation and cooling measures need to be adopted. For example, if a contactor is used, it will basically not consume active power; (3) The thyristor circuit itself is also a harmonic source. Its extensive application is not conducive to the waveform of the low voltage power grid. Therefore, in addition to improving the thyristor switching circuit, it should also be made to exit after the opening and closing operation is completed, and the capacitor should still be maintained by the contactor connected in parallel with it. 4.3 Switching circuit with anti-parallel connection of thyristor and diode The wiring scheme of anti-parallel connection of one thyristor and one diode is shown in Figure 4(b). Compared with the wiring scheme in Figure 4(a), since the price of diode of the same capacity is lower than that of thyristor, the manufacturing cost of the contactless switching circuit with anti-parallel connection of one thyristor and one diode is lower, while the technical performance is similar, but the response time is longer. When the capacitor is cut off, from the output of the cut-off command to the completion of the work task, it can be completed within half a cycle (i.e., time t≤10ms). If the scheme in Figure 4(b) is adopted, due to the uncontrollability of the diode, its cut-off time is usually between 0.5 and 1 Hz, that is, the cut-off time t ≤ 20 ms. 4.4 Novel Contactless Switching Circuit with Equal Voltage Zero Current Cut-off The wiring of the novel contactless switching circuit with equal voltage zero current cut-off is shown in Figure 5, where J is the contact of the AC contactor. Its operation sequence is explained as follows: When the capacitor is connected, the microcomputer controller first sends a signal to the switching circuit to connect the capacitor at equal voltage. The microcomputer controller then sends a signal to the contactor to close its contacts, short-circuiting the thyristor switching circuit. Since the contact resistance of the contactor J after closing is much smaller than the resistance when the switching circuit is conducting, energy saving and extension of the service life of the switching circuit are achieved. When it is necessary to cut off the capacitor, the controller first sends a signal to the contactor to open the contactor contact J. At this time, the switching circuit is in the conducting state, and the switching circuit cuts off the capacitor when the current crosses zero. The advantages of this scheme are: low power consumption, small inrush current, small harmonic influence, low manufacturing cost, and long service life of the switching circuit and contactor. 4.5 Three-phase delta-connected capacitor bank switched by a two-phase two-tube switching circuit. The wiring diagram for switching a three-phase delta-connected capacitor bank by a two-phase two-tube switching circuit is shown in Figure 6. This switching principle is a patented technology of Beijing Shoudian Technology Co., Ltd., and has been widely used in China's low-voltage distribution network with satisfactory results. 5 Intelligent Automatic Controller 5.1 Detected Quantity and Control Target The detected quantities are mainly cosφ, reactive power Q, and reactive current Iq. In the mid-1980s, controllers using cosφ as the detected quantity were mostly selected. The execution method was to switch capacitors, and the ultimate goal of compensation was to reduce the reactive power entering and leaving the grid. The main disadvantage of this scheme is that it is easy to generate switching oscillations under light loads, and it is not easy to achieve sufficient compensation under heavy loads. Therefore, new controllers no longer use cosφ as the detected quantity. The working principle of a controller with Q as the detected quantity is to send the voltage and current signals to a Hall element or a phase-sensitive amplifier or other device with multiplication function to measure Q = UIsinφ. Since the detected quantity and the control target are the same physical quantity, it is technically reasonable, but the detection difficulty is greater. The controller, whose detected quantity is Iq, utilizes the principle that the instant the phase voltage u crosses zero from positive to negative is precisely the maximum reactive current Iqmax of phase A. It uses the negative zero-crossing signal of the phase voltage u for control, employing a switch and a simple holding circuit to achieve real-time detection of Iq. The advantages of this scheme are: simple detection method, no oscillation, and compensation effect independent of grid voltage fluctuations. 5.2 There are two options for setting the detection point: ① The detection point for the controller's input voltage and current signals is set at the front end of the compensation equipment, as shown at point A in Figure 7; ② The detection point is set at the rear end of the compensation equipment, as shown at point B in Figure 7. Detection point A cannot directly detect the reactive power of the load, making it difficult to achieve rapid switching of multiple capacitor banks at once. It typically uses a gradual switching method, reaching the required compensation value more slowly. Therefore, it is only suitable for situations where the load operation is relatively stable, there are no large-capacity impact loads, and rapid dynamic compensation is not required. If connected to detection point B, its advantage is that it determines the number of capacitor banks to be switched on based solely on the measured values ​​of load Q and Iq. This is a control method that only controls switching without controlling the actual effect after compensation. Its advantages are simple control and the ability to quickly switch multiple capacitor banks on and off at once. Its disadvantage is poor accuracy in static compensation. Some experts have also proposed a closed-loop control method that combines the advantages of both methods. In this method, the detection point is set at point A, and the reactive power ΔQ after compensation is detected. The total reactive power Q of the load is then obtained through ΔQ, which is the reactive power of all capacitors required for complete compensation. This idea of ​​summing the variables can be implemented by a microcomputer. Furthermore, because all capacitors that should be switched on can be switched on at once, it achieves rapid dynamic compensation characteristics and is currently recognized as a more ideal solution. 5.3 Requirements for the performance and quality of automatic controllers Currently, there are approximately 200 companies producing reactive power compensation automatic controllers nationwide. Most of these are small family workshops with low technical levels, lacking complete testing equipment, producing small batches, and making it difficult to guarantee quality. To improve the technical performance and reliability of automatic controllers, the power industry standard DL/T597-1996, "Technical Conditions for Ordering Low-Voltage Reactive Power Compensators," stipulates the following requirements for the basic functions of controllers: ① They should have the function of setting capacitor switching thresholds, delay settings, and overvoltage protection settings; ② For controllers that can switch according to a set program, they should have a switching program setting function; ③ The operation of the panel function keys should have fault tolerance; ④ The panel settings should have hardware or software interlocking functions. To improve the reliability of controller operation, the following points should also be noted: ① The controller should have measures to prevent switching oscillations under low load conditions; ② The circuit design of the controller should be reasonably simplified; the more components and the more complex the circuit, the higher the failure rate of the controller; ③ Unnecessary additional functions of the controller itself should be reduced, such as automatic and manual switching, and capacitor fault alarm functions, which should be transferred from the controller to the reactive power compensation panel; ④ High-quality microcontrollers and electronic components should be selected; ⑤ Attention should be paid to the electromagnetic interference resistance of the automatic controller; ⑥ Detection quantities and control schemes should be reasonably selected. 5.4 Integration of Distribution Integrated Measurement and Control Instrument and Reactive Power Compensation Automatic Controller The integration of the reactive power compensation automatic controller and the distribution integrated measurement and control instrument is a distribution network automation issue raised by urban power grid transformation. Operating units often require the simultaneous installation of low-voltage capacitors for reactive power compensation and the distribution integrated measurement and control instrument on the low-voltage side of the distribution transformer. Taking the SDPD-2000 distribution integrated measurement and control instrument developed by Beijing Shoudian Technology as an example, it combines two major functions: data acquisition, display, and recording of distribution transformer operating parameters, as well as intelligent control and protection of reactive power compensation. The data acquisition scope includes: voltage, current, power factor, active and reactive power, active and reactive energy, harmonic voltage, harmonic current, daily maximum and minimum values ​​of voltage and load current, power outage time, power restoration time and cumulative power outage time, overvoltage, undervoltage, and phase loss time for each phase, etc., with a data storage period of 2 months. It also has an RS232/485 communication interface, allowing for on-site or remote data acquisition. For display, an LCD screen is used, providing a clear and intuitive display of relevant parameters of the distribution transformer operation in full Chinese. The physical quantity sampled in terms of intelligent control of reactive power compensation is the reactive power Q of the load; it can adjust any combination of Δ-Y capacitor banks; it prevents reactive power switching oscillation and compensation dead zone; when overvoltage, undervoltage, phase loss, harmonic or zero-sequence current exceedance and capacitor temperature rise exceedance occurs in the power grid, the compensation capacitor is quickly disconnected. 6 Application Examples Taking a 400kVA distribution transformer in Huxi Village, Xiamen City as an example, the operation record on July 8, 2000 is shown in Table 1. Note: The calculation formula for the reduction of line loss (ΔPs%) after compensation: References [1] Subhasis Nandi, Pannalal Biswas, et al, Two novel schemes suitable for static switching of three-phase delta-connected capacitor banks with minimum surge current. IEEE Transactions on Industry Applications, 1997, 33(5) [2] Dai Chaobo, Lei Linxu, Lin Haixue. Research on the delta connection scheme for reactive power compensation of thyristor-switched capacitors [J]. Journal of Electrical Engineering, 2001(3) [3] Chen Yunping et al. Phase-by-phase compensation of fundamental reactive power in low-voltage distribution network controlled by microcomputer thyristor switching [J]. Power System Technology, 1999(4)