Sensor-Network Cooling Control for Data Center Energy Waste
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current data center cooling systems are inefficient due to provisioning based on peak load scenarios and inaccurate temperature detection, leading to excessive cooling and energy wastage, as they often operate at a fraction of their capacity.
Innovation Solution
A sensor network is used to control temperature by commissioning sensors and selecting appropriate control schemes for primary actuators, such as CRAC units, to optimize energy utilization and maintain predetermined temperature ranges, thereby reducing energy consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If cooling is provisioned based on peak load scenarios and nameplate power ratings, then reliability is improved, but energy consumption increases excessively
Solution Approach 1:
The system implements feedback control by continuously monitoring actual temperatures at sensor locations near computer systems and using this information to dynamically adjust cooling provision. Temperature sensors detect real thermal conditions, and this feedback drives actuator adjustments to match actual cooling needs rather than relying on static nameplate ratings or peak load assumptions.
Solution Approach 2:
The cooling system transitions from static provisioning based on peak load scenarios to dynamic adjustment based on real-time temperature measurements. Actuators are controlled to vary cooling output according to changing thermal conditions detected by sensors, allowing the system to adapt to actual operational loads rather than maintaining fixed peak-capacity operation.
2Reliability
If cooling is provisioned for worst-case scenarios, then reliability is improved, but efficiency deteriorates due to operating at fraction of capacity
Solution Approach 1:
Real-time temperature feedback from sensor locations enables the system to maintain reliability through continuous monitoring while improving efficiency by adjusting cooling output to match actual thermal demands. The feedback loop prevents both over-provisioning and under-provisioning, keeping the system operating at optimal capacity levels.
Solution Approach 2:
The system uses distributed temperature sensors near computer systems to autonomously detect thermal conditions and trigger appropriate cooling responses. This self-service approach allows the cooling system to respond directly to actual thermal needs without relying on conservative worst-case provisioning, thereby improving overall efficiency.
3Device complexity
If temperatures are detected at air conditioning unit inlets, then measurement simplicity is improved, but measurement precision deteriorates
Solution Approach 1:
The temperature monitoring system is segmented into multiple distributed sensor locations positioned near individual computer systems or racks rather than using a single measurement point at air conditioning unit inlets. This segmentation provides localized temperature data that accurately reflects actual thermal conditions at heat-generating equipment.
Solution Approach 2:
Temperature sensors are positioned as intermediaries between the heat-generating computer systems and the cooling system. These sensors directly measure temperatures near the equipment being cooled and transmit this information to the control system, providing accurate thermal feedback without requiring direct measurement at air conditioning unit inlets.
Data Source
AI summary
In a method for controlling temperature using a sensor network, the sensors of the sensor network are commissioned and one of a plurality of control schemes for operating a primary actuator configured to vary temperatures of the sensors based upon energy utilization requirements of the plurality of control schemes is selected. In addition, the selected one of the plurality of control schemes is implemented to vary the temperatures detected by the sensors.