News

Coordinated Power Supply Scheme of UPS and DC Operating Power Supply in Data Centers

Coordinated Power Supply Scheme of UPS and DC Operating Power Supply in Data Centers

As the core hub of digital infrastructure, data centers carry critical businesses such as data storage, computing, and transmission, requiring 24/7 uninterrupted, stable, and high-reliable power support. In data center power supply systems, UPS (Uninterruptible Power Supply) and DC operating power supply undertake different core tasks: UPS focuses on ensuring the stable operation of AC loads such as servers, storage devices, and network switches; DC operating power supply is mainly responsible for powering core DC loads including communication modules, relay protection devices, and intelligent monitoring systems. The coordinated power supply scheme integrates the advantages of both systems, constructs a "dual-core mutual backup" power supply architecture, and solves the problems of single power supply risk, load mismatch, and low energy efficiency in traditional independent operation modes. Based on national standards such as GB/T 3797—2005 and DL/T 781—2010, as well as data center power supply specifications such as GB 50174—2017, this article elaborates on the coordinated architecture, working modes, core design points, and application cases of UPS and DC operating power supply in data centers, providing a standardized solution for optimizing data center power supply system reliability and efficiency.
Note: The following scheme is applicable to medium and large data centers (including Internet, financial, and enterprise-level data centers), supporting AC/DC hybrid load scenarios with 110V/220V DC operating power supply and 220V/380V UPS output. The scheme design must comply with data center power distribution, energy storage safety, and equipment operation technical requirements, and be implemented by professional electrical engineers.

1. Overall Coordinated Architecture of UPS and DC Operating Power Supply

The coordinated power supply system of UPS and DC operating power supply in data centers adopts a "three-layer integrated" architecture (equipment layer, control layer, and distribution layer), with the DC busbar as the core link, realizing seamless connection, intelligent scheduling, and mutual backup between the two systems. The architecture effectively integrates AC and DC power supply resources, ensuring stable power supply for all loads under various working conditions such as normal mains, power outage, and equipment failure.

1.1 Equipment Layer: Core Component Configuration

This layer is the hardware foundation of the coordinated system, consisting of UPS, DC operating power supply, energy storage unit, and auxiliary protection equipment, with reasonable configuration to meet the high reliability requirements of data centers.
  • UPS System Configuration: Select high-frequency online modular UPS with dual-conversion function, adopting N+1 or 2N redundancy configuration to ensure no single point of failure. The UPS input power factor is ≥0.99, output voltage accuracy is ±1%, ripple coefficient ≤0.5%, and full-load conversion efficiency ≥99%. The modular design supports on-demand expansion, and the total rated power is 1.25~1.5 times the total power of AC loads to cope with load surges and harmonic impacts. The UPS is connected to the DC busbar of the DC operating power supply through a bidirectional inverter interface, realizing energy mutual supply.

  • DC Operating Power Supply System: Composed of high-efficiency rectifier modules, 110V/220V DC distribution screens, lithium iron phosphate battery packs, and battery management systems (BMS). The rectifier module adopts three-phase full-bridge rectification technology, with conversion efficiency ≥98.5% and output current stability ±5%. The DC distribution screen serves as the core hub of the DC loop, connecting rectifier modules, battery packs, and DC loads, with each loop equipped with DC circuit breakers for overcurrent and short-circuit protection. The battery pack capacity is configured to ensure 8~24 hours of continuous power supply for critical DC loads, with a cycle life ≥6000 times.

  • Auxiliary Equipment: Including static switches (zero-switching time ≤10ms), insulation monitoring devices, and surge protectors (SPD). The static switch realizes seamless switching between UPS and DC operating power supply; the insulation monitoring device real-time monitors the insulation resistance between the DC busbar and ground; the SPD (protection level ≥20kA) is installed at the input end of the system to prevent lightning and surge damage to equipment.

1.2 Control Layer: Intelligent Coordination Mechanism

The control layer adopts DSP digital control technology and Modbus TCP communication protocol to build a unified intelligent scheduling platform, realizing centralized management and dynamic adjustment of the entire coordinated system.
  • Data Synchronization: Real-time collection of operating parameters such as UPS output voltage/current, DC busbar voltage, battery SOC (State of Charge), load rate, and equipment fault status, ensuring data consistency and synchronization between the two systems.

  • Automatic Scheduling: Based on mains status, load changes, and battery SOC, dynamically adjust the power supply ratio of UPS and DC operating power supply, optimizing energy efficiency and ensuring load stability.

  • Fault Linkage: When either system fails, the control platform immediately sends a fault signal to the other system, triggers automatic switching and backup mechanisms, and isolates the faulty module to avoid affecting the overall power supply.

1.3 Distribution Layer: Unified AC/DC Power Distribution Network

Construct a mixed power distribution network adapted to data center loads, with clear load division and safe and standardized wiring:
  • DC operating power supply directly supplies power to core DC loads (communication modules, relay protection devices, monitoring sensors) through the DC distribution screen, with insulated busbars (brown for positive pole, blue for negative pole) to ensure safe operation.

  • UPS supplies power to AC loads (servers, storage devices, air conditioners) through the AC distribution unit, with each load loop equipped with independent protection devices to avoid mutual interference.

  • The distribution network adopts top-in and top-out wiring mode, with the maximum voltage drop of the distribution screen under full load ≤1V, and the distance between AC and DC cables ≥30cm to prevent electromagnetic interference.

2. Core Working Modes of Coordinated Operation

The coordinated system of UPS and DC operating power supply has four typical working modes, which can be automatically switched according to mains status, equipment operation, and load changes, ensuring continuous and stable power supply for data center loads.

2.1 Normal Mains Mode (Parallel Operation with Load Division)

When the mains is stable, UPS and DC operating power supply operate in parallel, each responsible for their respective loads: the rectifier module of the DC operating power supply converts the mains into stable 110V/220V DC power, supplying power to DC loads and charging the battery pack (maintaining SOC 30%~80%); the UPS rectifies the mains into DC power, inverts it into standard 220V/380V AC power to supply power to AC loads, and its built-in battery is in a floating charge state. The intelligent platform monitors the load rate of both systems in real time, adjusting power distribution to keep the load rate of each system between 50%~80% to optimize energy efficiency.

2.2 Mains Interruption Mode (Battery Energy Sharing)

When the mains is interrupted, the system immediately switches to battery backup mode, and UPS and DC operating power supply realize coordinated power supply through the DC busbar: the DC battery pack supplies power to DC loads through the distribution screen, and at the same time supplies power to the UPS inverter through the DC busbar; the UPS inverts the DC power into AC power to continue supplying power to AC loads. This mode realizes the sharing of battery energy, avoids the waste of independent battery configurations, and extends the overall backup time by 30%~50% compared with independent operation.

2.3 Single System Fault Mode (Mutual Backup Switching)

  • DC Operating Power Supply Fault: When the DC operating power supply fails (such as rectifier module damage, battery undervoltage), the UPS immediately switches the inverter to the extended supply mode, supplying stable DC power to critical DC loads through the DC busbar while maintaining AC load power supply. The fault recovery time is ≤10ms, ensuring no impact on core data center businesses.

  • UPS Fault: When the UPS fails (such as inverter failure, overload), the static switch automatically switches AC loads to the bypass power supply (mains normal) or the DC operating power supply inverted bypass (mains interrupted). The DC operating power supply maintains stable operation of DC loads, and the AC loads are seamlessly connected to the backup power source, avoiding business interruption.

2.4 Mains Recovery Mode (Smooth Switching and Charging)

When the mains recovers, the system smoothly switches back to normal mains mode under the control of the intelligent platform: the rectifier module of the DC operating power supply first resumes operation, charges the battery pack with constant current and constant voltage, and supplies power to DC loads; the UPS rectifier resumes working, switches from battery mode to mains mode, and charges its built-in battery. The switching process avoids voltage fluctuations and inrush current, ensuring stable operation of all loads.

3. Core Design Points of the Coordinated Scheme

To ensure the reliability, safety, and efficiency of the coordinated system in data centers, the following core design points must be strictly implemented in accordance with national standards and industry specifications.

3.1 Capacity Matching and Redundancy Design

  • Capacity Matching: The total rated power of the UPS shall be 1.25~1.5 times the total power of AC loads, considering harmonic current and load surge impacts. The total capacity of the DC operating power supply shall be ≥1.2 times the total power of DC loads, and the DC busbar current-carrying capacity shall meet the system total capacity requirements. The energy storage capacity shall be calculated according to the maximum continuous mains outage time (generally 8~24 hours) to ensure load supply.

  • Redundancy Configuration: UPS modules and rectifier modules of DC operating power supply adopt N+1 redundancy; for data centers with level-4 disaster recovery requirements, 2N redundancy configuration is adopted. The battery pack adopts dual-group parallel design to ensure that a single battery group failure does not affect system operation.

3.2 Switching Control and Protection Technology

  • Zero-Switching Control: The static switch between UPS and DC operating power supply shall adopt zero-voltage zero-current switching technology, with a switching time ≤10ms, avoiding power interruption during switching. The UPS bypass shall have voltage stabilization and frequency stabilization functions to ensure stable power supply for AC loads during switching.

  • Multi-Level Protection: The system integrates overvoltage, undervoltage, overcurrent, short-circuit, and overtemperature protection functions. The DC loop adopts positive and negative dual protection devices, and the same brand and series of products are selected to ensure selective matching of protection functions. The insulation resistance between the DC busbar and ground shall be ≥2MΩ, and an alarm shall be issued immediately when the resistance is lower than the standard.

  • Lightning Protection Design: The entire system adopts a unified grounding system, with grounding resistance ≤1Ω (meeting data center high-reliability requirements). Install SPD at the input end of the mains, UPS, and DC operating power supply to prevent lightning and surge damage to equipment.

3.3 Wiring and Cabling Standards

  • Cables for UPS and DC operating power supply shall adopt flame-retardant shielded cables (complying with YD/T 1173—2001), with cross-sectional area selected according to current-carrying capacity and voltage drop ≤3%. The wiring of the two systems shall be separated by ≥30cm, and cross-connection shall be avoided; if unavoidable, cross at 90° to reduce electromagnetic interference.

  • Exposed busbars in the distribution screen shall be protected by insulating layers, and clear danger warning signs shall be attached. Cables between equipment shall be laid through cable trenches or protective pipes, with connection points wrapped with waterproof insulation tape to prevent water ingress and oxidation.

  • The DC distribution screen and UPS cabinet shall be installed in a dry, well-ventilated area, with a distance of ≥50cm from surrounding objects and heat sources to facilitate heat dissipation and maintenance.

3.4 Intelligent Monitoring and Management

The unified monitoring platform shall realize real-time monitoring of the operating status of UPS and DC operating power supply, including voltage, current, temperature, battery SOC, and fault information. The platform supports remote parameter adjustment, fault alarm, and data statistics, and can transmit operating data to the data center's upper-level energy management system (EMS) to realize global energy optimization. The BMS shall realize balanced management of battery modules and predictive warning of thermal runaway risks, extending battery life.

4. Typical Application Case

4.1 Large Internet Data Center UPS-DC Operating Power Supply Coordinated System

A large Internet data center in East China, with a total load of 2000kW (AC load 1600kW, DC load 400kW), adopted the coordinated power supply scheme of UPS and DC operating power supply, with the following configuration and application effects:
  • System Configuration: 2000kVA modular UPS (2N redundancy, 40 sets of 50kVA modules), 500kVA DC operating power supply system (110V/220V dual voltage output, 1MWh lithium iron phosphate battery pack), and intelligent scheduling platform. The UPS supplies power to servers and storage devices, while the DC operating power supply supplies power to communication modules and monitoring systems. The two systems are connected through a DC busbar to realize energy sharing.

  • Application Effects: The system power supply reliability reached 99.9999%, successfully withstanding multiple mains fluctuations and equipment maintenance tests without business interruption. During a mains outage lasting 8 hours, the coordinated system shared battery energy, ensuring continuous operation of all loads. The overall energy efficiency of the power supply system was improved by 12% compared with the original independent operation mode, and the annual energy saving was about 200,000 kWh. The intelligent monitoring platform reduced on-site maintenance workload by 60%, realizing remote fault diagnosis and handling.

5. Operation and Maintenance and Optimization Suggestions

5.1 Daily Operation and Maintenance

  • Conduct quarterly inspections of the coordinated system, including testing insulation resistance, grounding resistance, and switching function. Clean the UPS cooling system and DC distribution screen regularly to avoid overheating caused by dust accumulation.

  • Monitor the battery SOC, temperature, and internal resistance through the BMS monthly, and perform balanced charging when the voltage difference between battery cells exceeds 0.1V. Replace aging battery modules when the capacity decays to 80% of the rated capacity.

  • Perform annual full-link joint debugging, simulate mains interruption, single system fault, and other scenarios to verify the reliability of the coordination mechanism and the sensitivity of protection functions.

5.2 Optimization Suggestions

  • Energy Efficiency Optimization: Select high-efficiency UPS and rectifier modules, and adjust the load rate of the system to the optimal range through intelligent scheduling. Integrate the system with the data center's cooling system to realize coordinated optimization of power supply and cooling, further reducing PUE.

  • Intelligence Upgrade: Introduce AI algorithms to predict load demand and equipment fault trends, optimize scheduling strategies, and realize predictive maintenance of the system, reducing unplanned downtime.

  • New Energy Integration: Connect the coordinated system to the data center's distributed PV and energy storage system, realizing surplus new energy utilization and peak shaving and valley filling, reducing operating costs and carbon emissions.

Conclusion

The coordinated power supply scheme of UPS and DC operating power supply in data centers realizes the organic integration of AC and DC power supply systems, breaking through the limitations of traditional independent operation modes. Through scientific coordinated architecture, flexible working modes, and strict technical standards, the scheme not only improves the reliability and stability of the data center power supply system but also optimizes energy efficiency and reduces operation and maintenance costs. With the continuous improvement of data center load density and reliability requirements, the coordinated scheme will be further upgraded in terms of intelligent scheduling, new energy integration, and safety protection, providing a more solid power guarantee for the high-quality development of data centers.


Share This Article
Hotline
Email
Message