Abstract: The reliability of DC power supplies in power engineering plays a crucial role in ensuring the normal operation and fault handling of power systems. This article points out the problems and solutions related to DC power supply reliability from the perspective of design and selection, including battery selection, charging and discharging equipment selection, monitoring device setup, system wiring, and operation and protection equipment selection. Keywords: DC power supply; reliability; design and selection 0 Overview Power engineering includes various types of power plants and 500kV and below substations, as well as 10kV substations in industrial enterprises and buildings. A reliable DC power supply is indispensable for their normal operation and fault handling. It provides control, protection, and signal power to power equipment in normal operation, and allows high-voltage circuit breakers to operate normally. Especially when the AC power supply fails during a power system fault, a safe and reliable DC power supply is needed. In addition to supplying power to the aforementioned loads, it must also supply power to DC motors, emergency lighting, and UPS loads to ensure the fault handling and power restoration of the power system. For many years, extensive theoretical and practical research has been conducted on the reliability of DC power supplies. This has led to the elimination of outdated equipment and components, improved system wiring, enhanced automation levels, and the adoption of advanced technical specifications, along with principles of long lifespan and minimal maintenance, resulting in significantly improved reliability. Currently, power systems widely utilize valve-regulated sealed lead-acid batteries, high-frequency switching rectifiers or microprocessor-based thyristor rectifiers, DC circuit breakers, and DC power supply monitoring devices. However, some issues affecting the reliability of DC power supplies still exist in areas such as battery selection, charging and discharging equipment selection, monitoring device setup, system wiring, and the selection of operational protection equipment. Besides issues related to equipment technical quality, this paper will address some problems and solutions related to DC power supply reliability from a design and selection perspective. 1. Batteries In the past decade, valve-regulated sealed lead-acid batteries have been widely used. They offer advantages such as no need to add acid, no leakage, no acid mist, low self-discharge current, low internal resistance, long lifespan, and convenient installation with minimal maintenance. However, they are highly sensitive to temperature, thus requiring strict control over the charging power supply and preventing severe overcharging or undercharging. Therefore, the following issues should be considered in the design and selection of battery packs. 1.1 Selection of the Number of Battery Packs The "DC Design Code" has clearly stipulated various types of power engineering projects. Its principle is based on the reliability of DC load power supply. Power plants should have independent DC power supply systems for unit generators and power/control loads, with independent network control systems and separate auxiliary workshops far from the main plant. This aims to minimize the power supply range of each battery DC system and ensure functional independence. Important 110kV substations and 220kV and above substations are required to have two battery packs, based on their importance and the need to meet the dual power supply requirements of relay protection and circuit breaker tripping mechanisms. Therefore, the number of battery packs should be determined based on the needs and reliability of the power supply load, minimizing the power supply range and considering the importance of the project. 1.2 Selection of the Number of Batteries DC systems without terminal batteries and without step-down devices simplify the wiring of the DC system, avoid the sulfation of terminal batteries and the troublesome issues of silicon step-down devices, thus improving reliability. However, the battery bank must operate in accordance with the following requirements: the bus voltage during normal operation must be 105% of the nominal voltage; the bus voltage during online equalization charging should not exceed 110% of the nominal voltage; and the bus voltage at the end of an emergency discharge should be 85% of the nominal voltage. In other words, the bus voltage of a 220V DC system is allowed to fluctuate between 187V and 242V. This results in a float charge voltage of 2.23V, an equalization charge voltage between 2.28V and 2.33V, and a voltage above 1.8V at the end of an emergency discharge, fully satisfying the requirement that the DC bus voltage fluctuate within the permissible range. Calculations show that for a 220V battery bank, the number of individual 2V batteries can only be 10³ or 10⁴. However, most small-scale power projects using 220V DC systems employ batteries with a capacity of 200Ah or less, primarily 12V or 6V battery packs. For 12V packs, 18 cells are often used, equivalent to 108 individual 2V batteries. This results in a high DC bus voltage during normal operation, and lowering the float charge voltage impacts battery life. Since the DC bus voltage is even higher during equalization charging, silicon step-down voltage regulators are commonly used, increasing complexity and reducing reliability. When the DC load is small and battery capacity is sufficient, the voltage at the end of an emergency discharge can be increased to greater than 1.83V, allowing for the selection of 102 individual 2V batteries, or 17 12V packs, or 34 6V packs. Currently, some battery manufacturers can produce battery packs with a dummy battery, i.e., 10V or 4V packs. Choosing 14×12V+4×10V or 34×6V+1×4V also equates to 104 individual 2V battery packs. In summary, the number of battery banks should be strictly controlled to simplify the DC system wiring. 1.3 Selection of Test Discharge Equipment DL/T 5044—2004 "Technical Specification for Design of DC Systems in Power Engineering" stipulates that "test discharge devices should preferably use electrothermal devices or active inverter discharge devices." DL/T 724—2000 "Technical Specification for Operation and Maintenance of DC Power Supply Devices for Power Systems Using Batteries" also specifies the verification discharge method and discharge cycle for batteries. The reliability of relying solely on general capacity testing methods for determining the accident discharge capacity of batteries operating in float charging mode year-round is low. The battery terminal voltage is not a reliable indicator of capacity. The only method is to periodically perform verification charge-discharge cycles to activate the batteries and verify their capacity, ensuring that the batteries always operate at over 90% of their capacity. This meets the DC load requirements during AC power outages, power plant shutdowns, and substation emergency handling. This is also a crucial aspect of DC power supply reliability. Due to the backwardness of the discharge equipment level, the selection of discharge equipment is relatively difficult and there is no good selection, which brings a lot of difficulties to the operation and maintenance of the battery. 2 Rectifier The rectifier is an important device of DC power supply. Its quality directly affects the long-term reliable operation of the battery. Therefore, its main technical characteristics should meet the charging and float charging requirements of the battery, long-term continuous operation, and have voltage stabilization, current stabilization and current limiting performance. The technical parameters should meet the requirements of relevant standards. Various functions can be switched automatically or manually. The operation is safe and flexible. Source: Power Transmission and Distribution Equipment Network The following issues should be noted in the design and selection: (1) Float charging voltage For valve-regulated sealed lead-acid batteries, 2.23V should be selected. This is an important issue for the long-term reliable operation of DC system and the life of battery. The correct selection of the float charging voltage of the battery is a relatively complex issue. The float charging voltage should meet the needs of compensating for the battery self-discharge current and maintaining oxygen cycle. In fact, the battery structure, the corrosion rate of the positive plate grid, the gas emission in the battery, and the requirement of the DC system bus voltage of 105%UN should also be considered. If the float charge voltage is too low, the float charge current cannot maintain the oxygen cycle of the battery and compensate for the self-discharge of the battery, resulting in a deviation in the battery terminal voltage. If the float charge voltage is too high, it will aggravate the corrosion rate of the positive plate, gas discharge, and water loss. For valve-regulated batteries, the requirements for overcharging and undercharging are high, so the appropriate float charge voltage should be selected according to the characteristics of the battery. (2) Voltage stabilization, current stabilization and current limiting characteristics. In order to ensure that the battery can operate in the best state and to use the two-stage constant current and constant voltage charging method. In order to ensure the DC bus operating voltage and prevent the generation of lagging batteries, the voltage stabilization characteristics during float charging are very important. The current stabilization characteristics during charging are also very important. When the supply voltage gradually increases, it can maintain a stable current, ensure the normal electrochemical reaction of the battery, and smoothly enter the constant voltage equalization charging stage, so as to improve the battery characteristic parameters or solve the problem of capacity recovery of individual lagging batteries. The current limiting characteristics can prevent the rectifier from easily tripping due to the phenomenon of "load grabbing" and "overshooting" when the load suddenly increases. The voltage stabilization, current stabilization and current limiting characteristic parameters should meet the requirements of Table 1. Table 1. Voltage Stabilization, Current Stabilization, and Current Limiting Characteristic Parameters 3. DC Power Supply Monitoring Device Due to the adoption of computer monitoring technology in power plants, substations, and power dispatching departments, the DC power supply system is a component of the electrical system in power engineering. It plays a crucial role in ensuring the application and reliability of power system automation devices. Its technical conditions, basic parameters, basic functions, safety performance, and structural processes should all meet the requirements of the power industry standard DL/T 856—2003 "DC Power Supply Monitoring Device for Power Use". The industry standard stipulates that the main monitoring contents include automatic adjustment of charging voltage and current for stable operation, automatic switching from float charging to equalization charging or vice versa according to the operating mode, operating status and fault alarms of major DC circuit breakers, normal display and abnormal alarms of DC bus voltage, monitoring of DC system insulation status, online battery detection, and automatic adjustment of inverter discharge. Currently, further efforts are needed in the automatic adjustment of charging voltage and current with temperature changes, automatic switching of charging mode during operation, inverter discharge, automatic tripping in case of severe grounding, and the reliability and intelligence of online battery detection. 4. DC Power Distribution System The wiring, network design, and selection of operating and protective electrical appliances are the main issues affecting the reliability of DC power supplies. 4.1 DC System Wiring DC system wiring should strive for simplicity, safety, reliability, and ease of maintenance and operation. A single battery group can be connected via a single busbar with sectionalized sections or a single busbar. Two battery groups are connected via two busbar sections, with a connecting appliance between the two sections, typically a disconnecting switch; protective appliances can be installed if necessary. In short, the DC busbar connects one battery group and the corresponding charging equipment, while the branch loads are supplied by the busbar feeder lines. The busbar is only segmented when there is a dual DC load or one battery group is paired with two sets of charging equipment. Currently, a few power plants and substations still have dual busbar systems with terminal batteries or control and closing busbar systems with step-down devices. Although dual busbar systems with terminal batteries can fully utilize battery capacity, their complex wiring, difficult terminal battery maintenance, and inflexible operation have led to their near obsolescence. For systems using nickel-cadmium (NiCd) batteries, a separate control bus and closing bus configuration is essential because a single NiCd battery cell has a voltage of 1.2V. A 220V system using approximately 180 batteries has a float charge voltage of 1.36–1.39V, an equalization charge voltage of 1.47–1.48V, and a final discharge voltage of 1.10V. This results in a DC bus voltage fluctuating between 266 and 198V, which is insufficient to meet the load control requirements. Therefore, even small-capacity batteries requiring high-current closing mechanisms should have a closing bus. However, this is unnecessary for DC systems using valve-regulated lead-acid (VRC) batteries. 4.2 Network Design: A radial power supply method is recommended for DC power supply networks. For small-capacity (below 200Ah) battery DC systems, due to the limited power supply range, a two-stage network system can be used where the batteries are connected to the DC bus and directly supply power to the loads separately. For medium-to-large capacity (200Ah and above) battery DC systems, due to the large power supply range, DC distribution cabinets can be installed at load concentration points to supply power to the loads. Even with the redistribution of large loads, only a 3-4 level network system is formed. Radial power supply networks provide a single power supply to each load, thus avoiding interference. The distribution cabinet method also saves on cables and simplifies grounding location and protection equipment selection. 4.3 Selection of Operating and Protective Electrical Appliances: DC circuit breakers integrate operating and protection functions, are easy to install, flexible in operation, highly stable, and offer comprehensive protection. Generally, two-stage protection DC circuit breakers have a long-delay thermal tripping function for overload and an instantaneous electromagnetic tripping function for short circuits, making them an ideal choice. However, the rated current of the DC circuit breaker is determined based on the calculated load current. If the rated current is too high, the thermal tripping time under overload (I²t) will be prolonged due to the small load current. If the rated current is too low, the thermal tripping may malfunction due to the large load current and high ambient temperature during prolonged operation. Once the rated current of the circuit breaker is determined, in addition to the overload long-delay thermal trip protection characteristic, the short-circuit instantaneous electromagnetic trip characteristic is also established, generally operating at 10IN±20%. However, the short-circuit current at the circuit breaker installation location determines the sensitivity of the instantaneous short-circuit trip, which must be calculated and verified. The formulas for calculating the short-circuit current and sensitivity at the DC circuit breaker installation location are as follows: Idk=nU0/n(r0+rl)+Σrj+Σrk Kl=Idk/Idz Where, Idk is the short-circuit current at the circuit breaker installation location, A; U0 is the open-circuit voltage of the battery, V; rb is the internal resistance of the battery, Ω; rl is the resistance of the connecting strip or conductor between batteries, Ω; Σrj is the sum of the resistances of the connecting cable or conductor from the battery pack to the circuit breaker installation location, Ω; Σrk is the sum of the resistances of the relevant circuit breaker contacts, Ω; Kl is the sensitivity coefficient, which should not be less than 1.25; Idz is the instantaneous protection (tripper) operating current of the circuit breaker, A. Due to the complexity of parameters, it is impossible for design institutes, equipment manufacturers, or operating units to accurately calculate the short-circuit current, thus making sensitivity verification impossible. Because of the uncertainty of the short-circuit current, even when the rated current is selected based on the load current and the coordination between upstream and downstream protection stages is considered, the instantaneous short-circuit protection cannot guarantee this coordination. A large short-circuit current will inevitably lead to over-tripping and expand the scope of the fault; this is a problem that must be solved. Some units have tried to eliminate the instantaneous trip unit or replace it with a fuse at the battery outlet, but calculations and tests have shown that over-tripping and equipment damage still occur. Before the advent of intelligent DC circuit breakers, using a three-stage (overload long delay + short-circuit instantaneous + short-circuit short delay) DC circuit breaker, with progressively increasing time limits from the load side to the power supply side, eliminates the need for precise short-circuit current calculations. This method can achieve rapid fault clearing while meeting the requirements of coordination, preventing failure to trip, false tripping, and, most importantly, over-tripping. 5. Conclusion The fundamental points for the reliability of DC power supplies in power engineering are: selecting valve-regulated sealed lead-acid batteries; ensuring each battery bank has an independent power supply range; selecting the number of battery banks to meet the DC bus voltage requirements of various operating projects; and considering discharge equipment for the batteries. Rectifiers should be selected as high-frequency switching or thyristor type, with redundancy or backup; the main technical specifications should meet the battery life requirements. DC power supply monitoring devices must first ensure the needs of the charging rectifier; more comprehensive monitoring devices still need to be developed. DC distribution systems should simplify wiring, provide radial power supply, and select DC circuit breakers for protection. While meeting the reliability requirements for overload protection, they must also ensure rapid disconnection during short-circuit protection and possess reliable differential coordination.