Green Power Supply Practice of High-Frequency UPS in Cloud Computing Centers
Green Power Supply Practice of High-Frequency UPS in Cloud Computing Centers
Cloud computing centers, as the core infrastructure of the digital economy, are facing severe challenges of high energy consumption and carbon emissions. With the deep advancement of the "double carbon" goal, building green and low-carbon cloud computing centers has become an industry consensus. High-frequency UPS (Uninterruptible Power Supply), as a key device ensuring continuous and stable power supply for computing equipment, plays a pivotal role in the green transformation of power supply systems. Unlike traditional UPS applications that only focus on reliability, green power supply practice of high-frequency UPS in cloud computing centers integrates advanced technologies such as high-efficiency power conversion, renewable energy integration, energy storage coordination, and intelligent energy management. It constructs a "green power + energy storage + high-frequency UPS + intelligent regulation" integrated power supply system, which effectively reduces PUE (Power Usage Effectiveness) and carbon emissions while ensuring power supply reliability. This article elaborates on the system architecture, core technologies, practical paths, and application cases of this green practice, providing a reference for the low-carbon transformation of cloud computing centers.
1. System Architecture of Green Power Supply for High-Frequency UPS
The green power supply system of high-frequency UPS in cloud computing centers breaks through the single backup power supply mode of traditional UPS. It integrates grid power, renewable energy (solar energy, wind energy), energy storage systems, high-frequency UPS, and intelligent management platforms to form a multi-energy complementary and coordinated operation system. The architecture is composed of four core modules, which work together to achieve the goals of high efficiency, low carbon, and stability.
1.1 High-Frequency UPS Mainframe (Core Conversion Unit)
As the core of power conversion and switching, the high-frequency UPS for cloud computing centers adopts a modular design with titanium-level efficiency, and its conversion efficiency can reach more than 99% under full load conditions, which is 4%-6% higher than that of traditional low-frequency UPS. It supports hot-swappable modules, which not only ensures non-stop maintenance but also facilitates capacity expansion according to the growth of computing load, avoiding resource waste caused by over-configuration. Compared with UPS for civil scenarios, it has stronger harmonic suppression capabilities, with total harmonic distortion (THD) less than 2%, which can effectively reduce the impact on the power grid and improve the quality of input and output power.
In addition, the high-frequency UPS is equipped with a DC-coupled interface, which can directly connect to the energy storage system and renewable energy generation equipment, reducing energy loss caused by multiple conversions. It also integrates intelligent load adjustment functions, which can coordinate with the energy management platform to optimize power distribution and lay a foundation for PUE reduction.
1.2 Renewable Energy Integration Module
Renewable energy integration is the key to achieving green power supply. Cloud computing centers usually adopt distributed photovoltaic (PV) power generation as the main green energy source, and can be supplemented by wind power according to regional conditions. For large-scale cloud computing parks, a centralized PV cluster is deployed in the idle area of the park, while the roof and exterior walls of the computer room are equipped with distributed PV panels to maximize the utilization of space resources. The PV modules select high-efficiency monocrystalline silicon products with photoelectric conversion efficiency of over 24%, and are equipped with intelligent cleaning systems to ensure stable power generation efficiency.
The module is connected to the UPS system through a high-efficiency MPPT (Maximum Power Point Tracking) controller, which dynamically adjusts the working parameters according to the changes of light intensity and temperature, improving the energy utilization rate of the PV system by 15%-20% compared with traditional controllers. For regions with unstable renewable energy generation, the system can also access green power from the grid through the green power trading platform to ensure the proportion of green power consumption.
1.3 Energy Storage Coordination Module
The energy storage module plays a role in peak shaving, valley filling, and emergency backup, solving the problem of intermittent and unstable renewable energy generation. Lithium iron phosphate batteries with long cycle life (over 6000 cycles) and high safety are preferred, and a battery energy storage system (BESS) is formed by modular grouping. The capacity configuration is determined according to the peak-valley difference of the cloud computing center's load and the power generation of renewable energy, generally ensuring that it can provide 2-4 hours of backup power for critical loads and realize peak load shifting of the power grid.
In addition to lithium batteries, some high-standard cloud computing centers are also equipped with flywheel energy storage systems as auxiliary backup. Flywheel energy storage has the advantages of fast response speed (millisecond level) and long service life, which can make up for the short-term power gap when the UPS switches working modes and further improve the stability of the power supply system.
1.4 Intelligent Energy Management Platform
The intelligent management platform is the "brain" of the entire green power supply system, integrating digital twin, AI algorithm, and multi-source data monitoring capabilities. It can realize real-time monitoring of high-frequency UPS operating parameters, renewable energy generation, energy storage SOC (State of Charge), and computing center load. Through AI intelligent scheduling algorithms, it dynamically adjusts the power supply ratio of grid power, renewable energy, and energy storage, ensuring that green power is prioritized for consumption and the operating efficiency of the entire system is optimized.
The platform also supports remote monitoring, fault early warning, and automatic troubleshooting, which can reduce the operation and maintenance cost of the power supply system. At the same time, it can count and analyze carbon emissions, providing data support for the carbon accounting and certification of cloud computing centers.
2. Core Green Technologies and Operating Modes
The green power supply practice of high-frequency UPS relies on advanced technical support and flexible operating modes to balance reliability, efficiency, and low carbon. The following focuses on the core technologies and three typical operating modes.
2.1 Core Green Technologies
Cross-Layer Energy Efficiency Optimization Technology: Integrate high-frequency UPS with intelligent micro-modules, closed hot-aisle cooling, and other technologies to form an L1&L2 cross-layer coordinated optimization system. By adjusting the operating state of UPS and the cooling load in real time, the overall PUE of the data center is reduced to below 1.2, reaching the international advanced level.
DC Microgrid Linkage Technology: Construct a data center-level DC microgrid, realizing direct connection and coordinated operation of PV, energy storage, and UPS. Reduce energy loss caused by AC-DC conversion, and improve the utilization rate of green energy.
AI-Predicted Scheduling Technology: Based on historical data of load, photovoltaic power generation, and grid power price, predict the power supply demand in the next 24 hours through AI algorithms. Formulate optimal scheduling strategies to minimize carbon emissions and power costs.
2.2 Typical Operating Modes
2.2.1 Grid-Tied Green Power Priority Mode (Normal Grid Condition)
When the grid power is stable, the system gives priority to using renewable energy (such as PV power) to supply power to the computing load through the high-frequency UPS. The surplus green power charges the energy storage system. When the green power generation is insufficient, the grid power supplements the gap. When the energy storage system is fully charged, the excess green power can be fed back to the grid (subject to local policy approval), realizing green power sharing. This mode can maximize the consumption of green energy, reduce the reliance on grid power, and lower carbon emissions.
2.2.2 Off-Grid Energy Storage UPS Mode (Grid Interruption)
When the grid power is interrupted due to faults or extreme weather, the high-frequency UPS switches to off-grid mode within milliseconds, and the energy storage system and residual renewable energy jointly supply power to the critical load of the cloud computing center. The AI scheduling system automatically adjusts the load according to the SOC of the energy storage, shutting down non-critical computing tasks to extend the backup time. This mode ensures the continuous operation of core business and avoids economic losses caused by power outages.
2.2.3 Peak Shaving and Valley Filling Mode (Grid Peak-Valley Periods)
During the grid peak load period (usually 9:00-22:00), the energy storage system discharges to supply power to the load, reducing the power purchased from the grid and alleviating the grid load pressure. During the valley load period (usually 22:00 to the next day 7:00), the grid power with lower price charges the energy storage system. This mode not only reduces the power cost of the cloud computing center but also responds to the grid's peak shaving and valley filling requirements, realizing win-win cooperation between the data center and the power grid.
3. Key Design Points for Green Practice in Cloud Computing Centers
Cloud computing centers have the characteristics of high load density, strict reliability requirements, and pursuit of low PUE. The green power supply system of high-frequency UPS needs to be optimized for these characteristics to ensure practicality and economy.
3.1 PUE Lean Optimization Design
PUE is the core indicator of green data centers. The design should focus on reducing the energy consumption of the power supply system and the cooling system. The high-frequency UPS should be installed in the same micro-module as the computing equipment to shorten the cable length and reduce line loss. At the same time, the UPS's own heat dissipation system should be coordinated with the computer room's cooling system to avoid redundant heat dissipation energy consumption. By integrating cross-layer energy efficiency optimization technology, the PUE can be further reduced, and the goal of near-zero carbon operation can be achieved.
3.2 Green Power Full-Consumption Design
According to the local solar radiation, wind energy resources, and green power policy, reasonably configure the capacity of renewable energy and energy storage. For areas with sufficient light resources, the PV installation capacity can be configured according to 30%-50% of the total load of the data center. Equip with a high-performance MPPT controller and energy storage coordination system to solve the problem of intermittent renewable energy generation and ensure that the green power consumption rate reaches more than 90%.
3.3 High-Density and Modular Adaptation Design
With the development of 5G, AI, and big data, the load density of cloud computing centers is increasing. The high-frequency UPS should adopt a high-density modular design, with a single module power of 10-20kW, which can be flexibly expanded according to the load growth. The system should support parallel operation of multiple modules, ensuring that the power supply capacity can be adjusted in real time with the load change, avoiding energy waste caused by low-load operation of the UPS.
3.4 Safety and Compliance Design
The power supply system of cloud computing centers has strict requirements for safety and compliance. The high-frequency UPS should integrate multiple protection functions such as overvoltage, undervoltage, overcurrent, short circuit, and overheating. The energy storage system should be equipped with a BMS (Battery Management System) to monitor the battery state in real time and prevent safety accidents such as overcharging and thermal runaway. The entire system should comply with national standards such as GB/T 32910 (Green Data Center Evaluation Standard) and meet the requirements of carbon neutrality certification.
4. Operation and Maintenance Strategies for Green Power Supply System
Scientific operation and maintenance are the key to ensuring the long-term stable and efficient operation of the green power supply system. Combined with the characteristics of cloud computing centers, the following operation and maintenance strategies are formulated.
4.1 Intelligent Predictive Maintenance
Make full use of the intelligent energy management platform to monitor the operating parameters of high-frequency UPS, PV modules, and energy storage systems in real time. Through AI algorithms, predict potential faults (such as UPS module aging, battery performance degradation, PV panel contamination) and send early warning signals to maintenance personnel. Conduct targeted maintenance before faults occur, reducing unplanned downtime and maintenance costs.
4.2 Regular Energy Efficiency Inspection and Optimization
Conduct a comprehensive energy efficiency inspection of the power supply system every quarter, including testing the conversion efficiency of UPS, the power generation efficiency of PV system, and the charge-discharge efficiency of energy storage system. Analyze the energy consumption data and adjust the scheduling strategy and operation parameters in time. For example, clean the PV panels regularly to remove dust and dirt, ensuring the photoelectric conversion efficiency; optimize the UPS operating mode to improve the load rate and energy efficiency.
4.3 Energy Storage System Lifecycle Management
Establish a full lifecycle management system for the energy storage battery. Monitor the battery's SOC, cycle life, and internal resistance in real time through BMS. Formulate a reasonable charge-discharge strategy to avoid overcharging and deep discharging, extending the battery's service life. Replace aging battery modules in a timely manner when the battery capacity decays to 80% of the rated capacity to ensure the backup capacity and safety of the energy storage system.
5. Practical Case and Effect Analysis
The green power supply practice of high-frequency UPS has been widely applied in domestic and foreign cloud computing centers, achieving remarkable economic and environmental benefits. The following takes the Tangxia Computing Center project of Guangzhou Power Supply Bureau as an example to elaborate on its application effect.
The Tangxia Computing Center is a key project for Guangzhou Power Supply Bureau to promote the zero-carbon strategy, with a total construction area of nearly 10,000 square meters, carrying key infrastructure such as the Southern Power Grid Cloud Platform and artificial intelligence platform. To achieve the goal of near-zero carbon operation, the project cooperated with Huawei to build a green power supply system integrating high-frequency modular UPS, PV cluster, energy storage system, and intelligent energy management platform <superscript>1.
In terms of hardware configuration, the project deployed Huawei high-frequency modular UPS with a total capacity of 200kW, with a conversion efficiency of up to 99.2% under full load. It is equipped with a 500kW distributed PV cluster and a 1MWh lithium iron phosphate energy storage system. The intelligent energy management platform realizes cross-layer coordination of UPS, cooling system, and energy storage, and optimizes the PUE of the data center to 1.18 through AI scheduling <superscript>1.
After commissioning, the system realizes 100% consumption of PV power, with an annual green power generation of about 600,000 kWh, reducing carbon emissions by more than 400 tons per year. By adopting the peak shaving and valley filling mode, the annual power cost of the data center is reduced by more than 300,000 yuan. The project was successfully selected as a 2025 "Zero Carbon China" standard pilot, providing a replicable model for the green transformation of cloud computing centers <superscript>1.
Conclusion
The green power supply practice of high-frequency UPS in cloud computing centers is an important path to achieve the low-carbon transformation of the digital economy. By integrating high-efficiency power conversion, renewable energy, energy storage, and intelligent management technologies, it breaks through the bottleneck of high energy consumption of traditional UPS power supply systems, realizing the dual goals of reliable power supply and low carbon emission reduction. Its core lies in taking high-frequency UPS as the core, building a multi-energy complementary system, and realizing refined energy management through intelligent technology.
With the continuous advancement of the "double carbon" goal and the rapid development of renewable energy technology, the green power supply practice of high-frequency UPS will be further optimized in terms of efficiency, cost, and intelligence. Future development will focus on the integration of new energy storage technologies, the popularization of DC microgrids, and the deep application of AI scheduling algorithms, helping cloud computing centers move towards "near-zero carbon" and even "zero carbon" operation, and making greater contributions to the construction of a green digital economy.
Share This Article