Seven solutions for Energy-Saving Transformation of data centers

Seven solutions for Energy-Saving Transformation of data centers

With the rapid growth of computer network business volume and the increase of the number of servers, the area and scale of the data center are also constantly expanding, and the energy consumption cost of the data center is also rapidly increasing, and there is a tendency to catch up with the hardware cost. According to a survey by authoritative institutions, the annual total expenditure on server power supply and cooling in China exceeds 2 billion USD. For data centers that have been built and put into operation, how to achieve effective energy-saving transformation needs to be considered in many aspects. This article introduces several main solutions for Energy-Saving Transformation of data center infrastructure.

1. design an overall data center Monitoring System

 

Figure load variation curve of IT devices

According to the data center load and Air-Conditioning cold volume configuration, the data center is generally designed to run in a safe mode of N + X. However, the entire data center's IT equipment layout shows that the data center has a large area, considering the circulating air volume and the reduction of partial room temperature dead corners, and the dynamic changes in the power consumption of load equipment, precision air-conditioning groups cannot be manually controlled to run as needed, and the changes in the heat value of data centers can be irregular, all indoor fans run at full speed, and each air conditioner controls the compressor according to their own needs. This operation mode is not economical and energy-saving in terms of operating costs. If redundant devices in the air-conditioning group are forced to shut down, it is difficult to grasp the air flow organization and temperature change rules of the data room, and manual control of the air-conditioning system is also difficult to ensure the safety of the data room, therefore, local temperature may be uncontrollable. Therefore, it is of vital significance to save energy in the precision air-conditioning group control equipment room.

The idea of energy-saving group control for dedicated air conditioners in the data room is to reduce the energy consumption of the entire data room by reducing the running time of redundant air conditioners while ensuring the normal operation of equipment in the data room, the backup/rotation function and cascade function of the air-conditioning system are implemented to avoid competition and delay self-start function.

Combined with the sugon adaptive cluster power and energy consumption monitoring system, the precision air-conditioning cluster is centrally managed. The management software calculates the server power consumption in each row based on the CPU usage, the operation status of the air conditioner at the corresponding position is precisely controlled based on the Load Distribution and precision air conditioning distribution, including the start and stop of the compressor and the start and stop of the fan of the air conditioner Indoor Unit. Precision air conditioners communicate with protocol converters through RS485 networks, and Protocol converters work collaboratively with management nodes through Ethernet, so that precision air conditioners of different manufacturers (with monitoring function) two wireless temperature and humidity probes can be installed in each server cabinet through Shuguang standard protocol interface and data interaction with management nodes. Each rack is a network and a temperature and humidity monitoring network is formed through the networking of Iot, data is uploaded to the management node over Ethernet to form a dual-layer monitoring network. When the server is monitored to the temperature, enable the air conditioner in the data center to ensure device security while saving energy.

2. Establish data center heat dissipation and airflow organization model

Professional heat dissipation engineers use computational fluid dynamics (CFD) technology to analyze the numerical analysis and Model Experiment of Air Flow Characteristics in the central data center, in-depth analysis of the distribution of the air speed field and temperature field inside the data center, and a reasonable cold Volume Allocation design scheme is obtained on this basis to obtain the optimal air return status to meet the heat dissipation needs of the equipment, at the same time, the air conditioning load is reduced to get the best cold configuration effect. Currently, it is implemented through airpak or fluent simulation software.

 

Figure data center CFD Air Flow Simulation diagram

3. Add blind boards for cabinets

In some cabinets of the IDC room, due to the small number of devices, after the cold air enters the Cabinet from the lower part of the Cabinet, a large part of the cold air directly enters the IDC without passing through the equipment, in this way, the air supply of the air conditioner is short-circuited, and the cooling efficiency of the air conditioner is greatly reduced.

The air duct is blocked to reduce the short circuit of the air and improve the cooling efficiency of the air conditioner. The specific method is to install blind boards in the space where no equipment is installed in the cabinet, so that the cold air can only be sent to the IDC through the equipment.

 

Figure adding blind-panel airflow chart to the cabinet

4. install Active air supply components for cabinets-ADU

Due to the long air supply distance and low air supply efficiency, the traditional underfloor air supply air conditioners may cause the air outlet to be cold and insufficient air volume, which cannot meet the cooling requirements of equipment.

In this case, the active air outlet device ADU is installed in the air outlet, and the ADU is equipped with two or four EC fans. The fan speed can be adjusted according to the load of the Cabinet. This type of active air supply device can improve the air supply efficiency of the air conditioner, so that the number of air conditioners can be reduced to meet the refrigeration and air volume needs, to achieve energy-saving effect.

 

Figure ADU air flow chart for adding a floor in the cabinet

 

Figure ADU module Effect

5. Cold channel sealing for cabinets

The traditional underfloor air supply is a dedicated air conditioner in the IDC. This form may easily cause short-circuit air flows and insufficient air supply, resulting in low cooling efficiency of air conditioners.

Now we can use a closed cold channel to isolate the cold and hot air streams to prevent short-circuit air streams and mixed cold and hot air streams. This will greatly improve the cooling efficiency of the air conditioners, when the normal heat dissipation of equipment in the data center is met, some air conditioners are switched off to achieve energy saving.

6. Add a fresh air system to the IDC

Currently, many large data centers use fresh air systems to implement energy-saving solutions. Due to the closed space in the equipment room, the heat generated by the equipment is large. Precision air conditioners require uninterrupted operation 24 hours a year, resulting in high power consumption during refrigeration and humidification. By introducing fresh air and the control technology of entropy value, outdoor low-temperature air can be introduced in spring, autumn and winter to turn off precision air conditioners, while ensuring the temperature, humidity and cleanliness of the data center, to save energy.

7. Rebuild precision air conditioners in the IDC room so that they can have a natural cooling function-using a fluorine pump energy-saving system

Traditional air conditioners in Fan rooms under the floor still need cooling when the outdoor environment is low in winter. They do not make full use of the outdoor natural cold source, and the energy saving effect is poor. In this case, the air conditioning system can be transformed and upgraded, and the fluorine pump unit can be added to the original air conditioning system to turn off the compressor when the external environment is low in winter through the control system, the fluorine pump system is enabled for refrigeration. Because the fluorine pump consumes less energy than the air-conditioning compressor, the central data center is used as an example. After transformation, the air-conditioning system can save about 20% of power consumption every year.