A data center indirect evaporative cooling unit anti-freezing early warning system and protection method

Abstract

The application discloses an indirect evaporative cooling unit anti-freezing early warning system and protection method for a data center, which comprises a micro-environment sensor module, a data acquisition gateway, a monitoring platform and a freezing prediction module. The micro-environment sensor module is connected with the monitoring platform through the data acquisition gateway to collect and store the environment data of the IDEC. The freezing prediction module is connected with the monitoring platform. The freezing prediction module calculates the differential pressure deviation rate through a differential pressure determination model and determines the state of the heat exchange core based on the differential pressure deviation rate. The freezing probability is calculated through an LSTM freezing probability model. The risk is divided based on the freezing probability P through a risk grading component and early warning is performed. The micro-environment time series data is analyzed through the LSTM freezing probability model, the freezing probability at a future time can be predicted in advance, time for artificial preventive maintenance is saved, the early warning capability is strong, the differential pressure determination model is used for comprehensive judgment, false alarms and missed alarms caused by a single threshold are reduced, and the judgment accuracy is high.

Description

Technical Field

[0001] This invention relates to the field of data center operation and maintenance technology, specifically to a data center indirect evaporative cooling unit anti-freezing early warning system and protection method. Background Technology

[0002] IDEC units, or indirect evaporative cooling units, are commonly used in data centers, communication equipment rooms, and large public buildings for natural cooling. These units achieve cooling through heat and moisture exchange between outdoor fresh air, indoor exhaust air, and the heat exchange core, offering advantages such as significant energy savings and low operating costs. However, during winter or in low-temperature, high-humidity regions, the cold outdoor fresh air entering the heat exchange core can easily condense and even freeze with the humid air on the exhaust side. This can lead to blockage of the core channels, increased ventilation resistance, and decreased heat exchange efficiency. In severe cases, it can cause fan surge, core cracking, insufficient cooling capacity in the equipment room, and equipment shutdown.

[0003] Existing IDEC units typically employ fixed temperature thresholds, manual inspections, or simple interlocking methods for freeze protection control. For example, when the outdoor temperature falls below a set value, the spray pump is shut down, the fresh air volume is reduced, or the electric heater is activated; when the differential pressure switch activates, an alarm is triggered and the unit shuts down. While these methods can prevent equipment damage to some extent, the fixed temperature thresholds often only reflect the current state and cannot predict future trends in freezing risk. Actions are often triggered only after icing has already occurred or is imminent, resulting in limited early warning effectiveness. Furthermore, freezing is related to outdoor dry-bulb temperature, wet-bulb temperature, relative humidity, inlet and outlet pressure, airflow, valve opening, spray status, and the duration of low temperatures; a single temperature or differential pressure threshold is insufficient to accurately describe the actual freezing risk.

[0004] In view of the above-mentioned defects, the inventors of this invention have finally obtained this invention after a long period of research and practice. Summary of the Invention

[0005] To address the aforementioned technical deficiencies, the present invention provides a data center indirect evaporative cooling unit anti-freezing early warning system, comprising a microenvironment sensor module, a data acquisition gateway, a monitoring platform, and a freezing prediction module. The microenvironment sensor module is connected to the monitoring platform through the data acquisition gateway to collect and store environmental data of the IDEC (Indirect Evaporative Cooling Unit). The freezing prediction module is connected to the monitoring platform to predict the freezing probability based on the collected environmental data.

[0006] Preferably, the microenvironment sensor module includes a dry-bulb and wet-bulb temperature sensor installed at the outdoor shaded area of ​​the IDEC, an air inlet temperature sensor, an air inlet humidity sensor, and an air inlet pressure sensor installed at the outdoor fresh air inlet of the IDEC, and an air outlet temperature sensor, an air outlet humidity sensor, and an air outlet pressure sensor installed at the outdoor exhaust air outlet of the IDEC.

[0007] A protection method using the aforementioned data center indirect evaporative cooling unit anti-freeze early warning system includes the following steps: S1, data is captured through the micro-environment sensor module, and the environmental and device data of IDEC are provided to the monitoring platform through the data acquisition gateway; S2, the freezing prediction module calculates the pressure difference deviation rate through the pressure difference judgment model. And based on the pressure difference deviation rate Determine the state of the heat exchanger core and calculate the freezing probability using an LSTM freezing probability model. ; S3, the freeze prediction module derives the freeze probability P, and the risk classification component classifies risks based on the freeze probability P and issues an early warning.

[0008] Preferably, in step S1, the environmental data includes the inlet temperature provided by the inlet temperature sensor. The inlet relative humidity provided by the air inlet humidity sensor The inlet pressure provided by the air inlet pressure sensor The outlet temperature sensor provides the outlet temperature. The outlet relative humidity provided by the air outlet humidity sensor The outlet pressure sensor provides the outlet pressure. The outdoor shading dry-bulb temperature provided by the dry-bulb and wet-bulb temperature sensor outdoor shaded wet-bulb temperature The equipment data includes fan frequency, damper opening, spray status, and unit start / stop status.

[0009] Preferably, in step S2, when Greater than the set threshold, and or If the temperature drops below the antifreeze threshold and persists for more than the set time, it is determined that the heat exchange core is experiencing freezing or a freezing tendency; if If the temperature and humidity conditions rise but do not meet the freezing requirements, the equipment is identified as having a non-freezing malfunction and a maintenance prompt will be issued.

[0010] Preferably, the pressure difference deviation rate The calculation formula is:

[0011] in, The reference pressure difference is the normal operating condition. A represents the valve opening degree, and S represents the fan frequency or air volume status. The expression for normalizing the real-time differential pressure according to the reference air density and air volume is as follows:

[0012] in, For reference air density, Current air density; For reference air volume, This represents the current actual air volume; The formula for calculating the pressure difference ΔP is:

[0013] in, Due to import pressure, This is due to export pressure.

[0014] Preferably, in step S2, the LSTM freeze probability model is established based on real-time data and historical window data read by the monitoring platform, using continuous time series as input and employing the LSTM algorithm. The LSTM freeze probability model captures the characteristics of the data provided by the monitoring platform changing over time and outputs the probability that each IDEC unit will freeze within a set time window in the future. .

[0015] Preferably, the set time window is set to 60min, 120min, 180min, 240min, 300min, and 360min or more according to the site requirements.

[0016] Preferably, the formula for calculating the freezing probability P is:

[0017] in, These are the weighting coefficients; Probability of physical rules; The output probability is for LSTM.

[0018] Better Calculated based on wet-bulb temperature, inlet air temperature, relative humidity, duration of low temperature, and pressure differential deviation rate:

[0019] in, Outdoor wet-bulb temperature The weighting coefficients, For the inlet temperature The weighting coefficients, relative humidity of the import The weighting coefficients, Duration of low temperature The weighting coefficients, Pressure difference deviation rate The weighting coefficients satisfy: ; Outdoor wet-bulb temperature The normalization function, For the inlet temperature The normalization function, relative humidity of the import The normalization function, Duration of low temperature The normalization function, Pressure difference deviation rate The normalized function, and , , , and The output range is 0 to 1.

[0020] Preferably, in step S3, the risk classification component classifies each IDEC unit into low-risk, medium-risk, and high-risk categories based on the freeze probability P calculated by the freeze prediction module: wherein, When 0%≤P≤60%, the risk is low, and the data center indirect evaporative cooling unit anti-freeze early warning system will continue to operate normally and monitor continuously. When 60% < P ≤ 80%, it is considered a medium risk. The data center indirect evaporative cooling unit anti-freeze early warning system issues an early warning, generates a preventive maintenance work order, and prompts maintenance personnel to check the drainage, spray, filter, air valve, sensor and core surface condition. When 80% < P ≤ 100%, it is considered a high risk. The data center indirect evaporative cooling unit antifreeze early warning system determines that there is a high probability of icing and activates antifreeze protection measures automatically or after manual confirmation.

[0021] Compared with the prior art, the beneficial effects of the present invention are as follows: The present invention analyzes micro-environment time series data through the LSTM freezing probability model, which can predict the freezing probability of future moments in advance, buy time for manual preventive maintenance, and has a strong early warning capability. In addition, it combines the differential pressure judgment model to make a comprehensive judgment, reducing false alarms and missed alarms caused by a single threshold, and has high judgment accuracy. Attached Figure Description

[0022] Figure 1 This is a schematic diagram of the anti-freezing early warning system for the indirect evaporative cooling unit in the data center. Figure 2 This is a flowchart illustrating the protection method using the aforementioned data center indirect evaporative cooling unit anti-freeze early warning system. Detailed Implementation

[0023] The above-mentioned and other technical features and advantages of the present invention will be described in more detail below with reference to the accompanying drawings. Example 1

[0024] like Figure 1 As shown, Figure 1 This is a schematic diagram of the anti-freezing early warning system for the indirect evaporative cooling unit in the data center. The anti-freezing early warning system for the indirect evaporative cooling unit in the data center of the present invention includes a micro-environment sensor module, a data acquisition gateway, a monitoring platform, and a freezing prediction module. The micro-environment sensor module is connected to the monitoring platform through the data acquisition gateway to collect and store environmental data of the IDEC. The freezing prediction module is connected to the monitoring platform to predict the freezing probability based on the collected environmental data.

[0025] Specifically, the microenvironment sensor module includes a dry-bulb and wet-bulb temperature sensor installed at the outdoor shaded area of ​​the IDEC, an inlet temperature sensor, an inlet humidity sensor, and an inlet pressure sensor installed at the outdoor fresh air inlet of the IDEC, and an outlet temperature sensor, an outlet humidity sensor, and an outlet pressure sensor installed at the outdoor exhaust outlet of the IDEC. The inlet temperature sensor, the inlet humidity sensor, and the inlet pressure sensor are used to collect the microenvironmental status of the fresh air before it enters the unit. The outlet temperature sensor, the outlet humidity sensor, and the outlet pressure sensor are used to collect the thermal and humidity status on the exhaust side. The dry-bulb and wet-bulb temperature sensor is used to obtain the outdoor dry-bulb and wet-bulb temperatures that are not directly affected by solar radiation. The inlet pressure sensor and the outlet pressure sensor can be arranged on the front and rear sides of the heat exchange core according to the actual structure of the core, and are used to calculate the pressure difference across the heat exchange core. .

[0026] Preferably, the microenvironment sensor module further includes an air volume sensor, a fan frequency collector, a sprinkler pump status collector, a louver vent opening collector, an exhaust valve opening collector, and a bypass ventilation valve opening collector, which collect real-time data such as air volume, fan frequency, sprinkler pump status, louver vent opening, exhaust valve opening, and bypass ventilation valve opening to improve the accuracy of judgment.

[0027] The data acquisition gateway collects environmental data and device data from the micro-environment sensor module. The gateway preprocesses the environmental and device data, including timestamp marking, outlier removal, missing value imputation, unit standardization, and sampling period alignment. The processed data is uploaded to the monitoring platform. After archiving and preliminary calculations, the monitoring platform provides external data interfaces, such as OPCServer, OPCUA, and ModbusTCPServer interfaces and communication protocols, for real-time reading by the prediction algorithm.

[0028] This invention analyzes micro-environment time-series data using an LSTM freeze probability model, which can predict the freeze probability in advance, giving time for manual preventative maintenance and providing strong early warning capabilities. Combined with a differential pressure judgment model, it reduces false alarms and missed alarms caused by a single threshold, resulting in high judgment accuracy. Example 2

[0029] like Figure 2 As shown, Figure 2 This is a flowchart illustrating the protection method using the aforementioned data center indirect evaporative cooling unit anti-freeze early warning system; the protection method of the present invention using the aforementioned data center indirect evaporative cooling unit anti-freeze early warning system includes the following steps: S1, data is captured through the micro-environment sensor module, and the environmental and device data of IDEC are provided to the monitoring platform through the data acquisition gateway; S2, the freezing prediction module calculates the pressure difference deviation rate through the pressure difference judgment model. And based on the pressure difference deviation rate Determine the state of the heat exchanger core and calculate the freezing probability using an LSTM freezing probability model. ; S3, the freeze prediction module derives the freeze probability P, and the risk classification component classifies risks based on the freeze probability P and issues an early warning.

[0030] Specifically, in step S1, the environmental data includes the inlet temperature provided by the inlet temperature sensor. The inlet relative humidity provided by the air inlet humidity sensor The inlet pressure provided by the air inlet pressure sensor The outlet temperature sensor provides the outlet temperature. The outlet relative humidity provided by the air outlet humidity sensor The outlet pressure sensor provides the outlet pressure. The outdoor shading dry-bulb temperature provided by the dry-bulb and wet-bulb temperature sensor outdoor shaded wet-bulb temperature The equipment data includes fan frequency, damper opening, spray status, and unit start / stop status.

[0031] In step S2, when Greater than the set threshold, and or If the temperature drops below the antifreeze threshold and persists for more than the set time, it is determined that the heat exchange core is experiencing freezing or a freezing tendency; if If the temperature and humidity conditions rise but do not meet the freezing requirements, the system will prioritize identifying non-freezing abnormalities such as dust accumulation, filter blockage, or airflow issues, and will generate a maintenance prompt instead of directly entering antifreeze protection.

[0032] Preferably, the pressure difference deviation rate The calculation formula is:

[0033] in, This is the baseline pressure difference under normal operating conditions. Since the pressure difference is affected by air volume, valve opening, and air density, A represents the valve opening, and S represents the fan frequency or air volume status. The expression for normalizing the real-time differential pressure according to the reference air density and air volume is as follows:

[0034] in, For reference air density, the standard air density or the air density during system calibration can be used; for current air density; For reference air volume, the rated air volume or the calibration air volume when establishing the benchmark model can be used; This represents the current actual air volume; The formula for calculating the pressure difference ΔP is:

[0035] in, Due to import pressure, This is due to export pressure.

[0036] In step S2, the LSTM freeze probability model is built based on real-time data and historical window data read by the monitoring platform, using continuous time series as input and employing the LSTM algorithm. The LSTM freeze probability model captures the characteristics of the data provided by the monitoring platform changing over time through memory gates, input gates, and output gates, and outputs the probability that each IDEC unit will freeze within a set time window in the future. Generally, historical operational data is used as training samples, with manual inspection records, pressure difference anomaly records, core icing records, and protection action records as labels to train the LSTM freezing probability model.

[0037] The features input to the LSTM freezing probability model include inlet temperature. Imported relative humidity Import pressure outlet temperature Export relative humidity Export pressure Outdoor shade dry bulb temperature Outdoor shaded wet-bulb temperature The parameters include fan frequency, valve opening degree, spray status, duration of low temperature, rate of temperature drop, wet-bulb approach, air density, and differential pressure deviation rate.

[0038] The set time window can be set to 60min, 120min, 180min, 240min, 300min, 360min or more according to on-site requirements.

[0039] To improve engineering reliability, the formula for calculating the final freezing probability P in this invention is as follows:

[0040] in, This is a weighting coefficient, which can be set based on historical data verification results; Output probability for LSTM; The probability is based on physical rules, calculated according to wet-bulb temperature, inlet air temperature, relative humidity, duration of low temperature, and pressure difference deviation rate:

[0041] in, Outdoor wet-bulb temperature The weighting coefficients, For the inlet temperature The weighting coefficients, relative humidity of the import The weighting coefficients, Duration of low temperature The weighting coefficients, Pressure difference deviation rate The weighting coefficients satisfy: ; Outdoor wet-bulb temperature The normalization function, For the inlet temperature The normalization function, relative humidity of the import The normalization function, Duration of low temperature The normalization function, Pressure difference deviation rate The normalized function, and , , , and The output range is 0 to 1.

[0042] When sensor data is incomplete or model confidence is insufficient, the system automatically increases the weight of physical rules to ensure the safety of antifreeze judgment.

[0043] In step S3, the risk grading component classifies each IDEC unit into low-risk, medium-risk, and high-risk categories based on the freeze probability P calculated by the freeze prediction module: where, When 0%≤P≤60%, the risk is low, and the system continues to operate normally and be continuously monitored. When 60% < P ≤ 80%, it is considered a medium risk. The system issues an early warning, generates a preventive maintenance work order, and prompts maintenance personnel to check the condition of drainage, sprinklers, filters, air valves, sensors, and core surface. If the risk decreases after manual maintenance, the system records the result of the action and uses it as a sample for subsequent model calibration. When 80% < P ≤ 100%, it is considered a high risk. The system determines that there is a high probability of freezing and activates antifreeze protection measures automatically or after manual confirmation.

[0044] Through the above-mentioned tiered approach, maintenance personnel can intervene manually in advance before a freeze occurs, such as checking whether the drainage is unobstructed, turning off or reducing the spray water volume, cleaning the water film on the core surface, and confirming the operation status of louvers and air valves, thereby avoiding unplanned downtime caused by sudden freezing.

[0045] For example, if a unit experiences continuous low temperature and high humidity weather, with a fresh air inlet temperature of -6℃, an outdoor wet-bulb temperature of -7℃, an exhaust air temperature of 18℃, and a core pressure differential deviation rate of 22%, the LSTM model outputs a freezing probability P=86% in the next 60 minutes. The system marks this unit as high-risk.

[0046] This invention creates an operational record for each early warning, differential pressure confirmation, manual intervention, protective action, and recovery result. New data is periodically used to retrain or calibrate the LSTM model, allowing it to gradually adapt to the operational characteristics of different data centers, different brands of IDEC generators, and different climate zones. Example 3

[0047] The data center indirect evaporative cooling unit anti-freeze early warning system of the present invention further includes an anti-freeze protection execution module. This module is equipped with a three-way pipe that simultaneously connects the indoor exhaust duct, the outdoor exhaust duct, and the fresh air intake duct to form a controllable mixed air path. The three-way pipe is equipped with an exhaust valve, a bypass valve, and a check valve. The exhaust valve corresponds to the outdoor exhaust duct, the bypass valve corresponds to the fresh air intake duct, and the check valve corresponds to the indoor exhaust duct, ensuring directional airflow from the indoor exhaust duct to the outdoor exhaust duct. When the LSTM freezing probability model determines a high risk or the differential pressure determination model confirms core icing, the control logic is as follows: The louvered air vents corresponding to the fresh air intake duct are fully closed or gradually closed to reduce the direct entry of low-temperature outdoor fresh air into the equipment; the opening of the exhaust air valve is adjusted to allow high-temperature exhaust air to enter the three-way duct at a set ratio; the bypass ventilation valve is fully open to allow the high-temperature exhaust air to fully mix with the low-temperature outdoor fresh air before entering the fresh air intake duct, increasing the temperature of the fresh air entering the equipment to resemble an internal circulation state. If the safe operation of the computer room requires maintaining a minimum fresh air volume, the louvered air vents are kept at their minimum opening, and the mixed air temperature is maintained through the exhaust air valve and the bypass ventilation valve.

[0048] During high-risk protection periods, this invention can also implement the following measures in conjunction: reduce or stop the operation of the spray pump, open the drain valve to empty residual water, reduce the frequency of the fresh air fan, keep the exhaust fan running to provide a heat source, and activate the backup mechanical refrigeration or local heating device. After the freezing probability drops to a low-risk range and the differential pressure deviation rate returns to the normal range and remains so for a set period of time, this invention gradually restores the louvered air vents, bypass ventilation valves, and exhaust air valves to normal operating conditions.

[0049] This invention utilizes the heat from the IDEC exhaust air to mix with the low-temperature outdoor fresh air, thereby increasing the temperature of the incoming fresh air and reducing the need for electric heating or direct shutdown. The protection measures are energy-efficient and reliable. Furthermore, the sensors, data acquisition gateways, OPCServer or Modbus TCPServer interfaces, and damper actuators are all commonly used components in engineering sites, making it easy to implement on existing IDEC units. The prediction algorithm can simultaneously read data from multiple data centers and multiple IDEC units, outputting equipment-level freeze probabilities and risk rankings, allowing maintenance personnel to prioritize the handling of high-risk equipment.

[0050] The above description is merely a preferred embodiment of the present invention and is illustrative rather than restrictive. Those skilled in the art will understand that many changes, modifications, and even equivalents can be made within the spirit and scope defined by the claims of the present invention, all of which will fall within the protection scope of the present invention.

Claims

1. A freeze protection early warning system for indirect evaporative cooling units in data centers, characterized in that, The system includes a microenvironment sensor module, a data acquisition gateway, a monitoring platform, and a freeze prediction module. The microenvironment sensor module is connected to the monitoring platform via the data acquisition gateway to collect and store environmental data from the IDEC. The freeze prediction module is also connected to the monitoring platform and calculates the differential pressure deviation rate using a differential pressure determination model. And based on the pressure difference deviation rate Determine the state of the heat exchanger core and calculate the freezing probability using an LSTM freezing probability model. The risk classification component is used to classify risks based on the freezing probability P and issue early warnings.

2. The data center indirect evaporative cooling unit anti-freeze early warning system as described in claim 1, characterized in that, The microenvironment sensor module includes a dry-bulb and wet-bulb temperature sensor installed at the outdoor shaded area of ​​IDEC, an air inlet temperature sensor, an air inlet humidity sensor, and an air inlet pressure sensor installed at the outdoor fresh air inlet of IDEC, and an air outlet temperature sensor, an air outlet humidity sensor, and an air outlet pressure sensor installed at the outdoor exhaust air outlet of IDEC.

3. A protection method using the data center indirect evaporative cooling unit anti-freeze early warning system as described in claim 2, characterized in that, Including the following steps: S1, data is captured through the micro-environment sensor module, and the environmental and device data of IDEC are provided to the monitoring platform through the data acquisition gateway; S2, the freezing prediction module calculates the pressure difference deviation rate using the pressure difference judgment model. And based on the pressure difference deviation rate Determine the state of the heat exchanger core and calculate the freezing probability using an LSTM freezing probability model. ; S3, the freeze prediction module derives the freeze probability P, and the risk classification component classifies risks based on the freeze probability P and issues an early warning.

4. The protection method as described in claim 3, characterized in that, In step S1, the environmental data includes the inlet temperature provided by the inlet temperature sensor. The inlet relative humidity provided by the air inlet humidity sensor The inlet pressure provided by the air inlet pressure sensor The outlet temperature sensor provides the outlet temperature. The outlet relative humidity provided by the air outlet humidity sensor The outlet pressure sensor provides the outlet pressure. The outdoor shading dry-bulb temperature provided by the dry-bulb and wet-bulb temperature sensor outdoor shaded wet-bulb temperature The equipment data includes fan frequency, damper opening, spray status, and unit start / stop status.

5. The protection method as described in claim 4, characterized in that, In step S2, when Greater than the set threshold, and or If the temperature drops below the antifreeze threshold and persists for more than the set time, it is determined that the heat exchange core is experiencing freezing or a freezing tendency; if If the temperature and humidity conditions rise but do not meet the freezing requirements, the equipment is identified as having a non-freezing malfunction and a maintenance prompt will be issued.

6. The protection method as described in claim 5, characterized in that, The pressure difference deviation rate The calculation formula is: in, The reference pressure difference is the normal operating condition. A represents the valve opening degree, and S represents the fan frequency or air volume status. The expression for normalizing the real-time differential pressure according to the reference air density and air volume is as follows: in, For reference air density, Current air density; For reference air volume, This represents the current actual air volume; The formula for calculating the pressure difference ΔP is: in, Due to import pressure, This is due to export pressure.

7. The protection method as described in claim 3, characterized in that, In step S2, the LSTM freeze probability model is built based on real-time data and historical window data read by the monitoring platform, using continuous time series as input and employing the LSTM algorithm. The LSTM freeze probability model captures the characteristics of the data provided by the monitoring platform changing over time and outputs the probability that each IDEC unit will freeze within a set time window in the future. .

8. The protection method as described in claim 7, characterized in that, The formula for calculating the freezing probability P is: in, These are the weighting coefficients; Probability of physical rules; The output probability is for LSTM.

9. The protection method as described in claim 8, characterized in that, Calculated based on wet-bulb temperature, inlet air temperature, relative humidity, duration of low temperature, and pressure differential deviation rate: in, Outdoor wet-bulb temperature The weighting coefficients, For the inlet temperature The weighting coefficients, relative humidity of the import The weighting coefficients, Duration of low temperature The weighting coefficients, Pressure difference deviation rate The weighting coefficients satisfy: ; Outdoor wet-bulb temperature The normalization function, For the inlet temperature The normalization function, relative humidity of the import The normalization function, Duration of low temperature The normalization function, Pressure difference deviation rate The normalized function, and , , , and The output range is 0 to 1.

10. The protection method as described in claim 8, characterized in that, In step S3, the risk grading component classifies each IDEC unit into low-risk, medium-risk, and high-risk categories based on the freeze probability P calculated by the freeze prediction module: where, When 0%≤P≤60%, the risk is low, and the data center indirect evaporative cooling unit anti-freeze early warning system will continue to operate normally and monitor continuously. When 60% < P ≤ 80%, it is considered a medium risk. The data center indirect evaporative cooling unit anti-freeze early warning system issues an early warning, generates a preventive maintenance work order, and prompts maintenance personnel to check the drainage, spray, filter, air valve, sensor and core surface condition. When 80% < P ≤ 100%, it is considered a high risk. The data center indirect evaporative cooling unit antifreeze early warning system determines that there is a high probability of icing and activates antifreeze protection measures automatically or after manual confirmation.

Images