Data Center Cooling Optimization via CFD Simulation
Solution Overview
Problem
Current data center cooling systems rely on sensor-based methods that are inefficient, as they often fail to provide accurate temperature estimates and airflow evaluations, leading to incorrect readings and inefficiencies in cooling management, particularly due to aberrant sensor data and lack of real-time adaptation.
Innovation Solution
A simulation-based optimization method that combines airflow and temperature numerical simulations with an energy model and optimization solver to determine optimal cooling set points, reducing energy consumption while maintaining desired inlet temperatures, and includes techniques for variance reduction, automatic calibration, and regression modeling for real-time control.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If sensor-based cooling control systems are used, then cooling adjustments can be made based on temperature readings, but the systems fail to provide accurate temperature estimates in areas without sensors and incorrectly incorporate aberrant sensor readings
Solution Approach 1:
The patent introduces computational fluid dynamics (CFD) simulation as an intermediary between physical sensors and cooling control decisions. The CFD model creates a virtual representation of airflow and temperature distribution, allowing the system to estimate temperatures in areas without physical sensors and to filter/validate sensor readings by comparing them against simulation predictions, thereby resolving the contradiction between measurement precision and reliability
Solution Approach 2:
The patent creates a digital twin (virtual copy) of the data center environment through CFD simulation. This virtual model replicates the physical system's airflow and temperature characteristics, enabling accurate temperature estimation anywhere in the data center without requiring physical sensors at every location, thus improving measurement precision while maintaining reliability through model-based validation
2Adaptability or versatility
If traditional sensor-based cooling control is implemented, then cooling equipment can be adjusted, but the systems fail to provide sophisticated evaluation and classification of sensor data, leading to incorrect incorporation of aberrant readings
Solution Approach 1:
The patent implements a feedback mechanism where CFD simulation results continuously inform and refine cooling control decisions. The system compares simulated airflow patterns and temperature distributions with actual sensor readings, uses this feedback to detect anomalies such as air blockages or missing blanking panels, and adjusts cooling equipment accordingly, thereby improving adaptability while preventing loss of critical airflow information
Solution Approach 2:
The patent replaces traditional mechanical sensor-based detection systems with a computational simulation approach. Instead of relying solely on physical sensors to detect airflow problems, the system uses CFD simulation to model and predict airflow patterns, automatically identifying anomalies through virtual measurements, thus enhancing adaptability while preventing information loss about airflow issues
3Measurement precision
If numerical simulation is used for cooling optimization, then accurate temperature and airflow prediction is achieved, but computational time and complexity increase
Solution Approach 1:
The patent performs preliminary CFD simulations during the design and commissioning phases to establish baseline airflow patterns and temperature distributions. These pre-computed results are stored and used to rapidly assess current system performance without requiring real-time full-scale simulations, thereby maintaining prediction accuracy while significantly reducing optimization computation time
Solution Approach 2:
The patent implements a hybrid approach where full CFD simulations are performed only when necessary (e.g., during commissioning or major changes), while routine optimization uses simplified models or pre-computed data. This partial application of comprehensive simulation maintains accuracy for critical decisions while reducing time loss for routine adjustments
Data Source
AI summary
Aspects of this disclosure are directed to a simulation-based cooling optimization method that provides real-time cooling set points in a data center. The method combines airflow and temperature simulation, energy modeling, and an optimization solver to determine optimal cooling set point values for a data center. Other aspects are also directed to estimating power consumption and cost and energy savings.