Modular Liquid Cooling System for Data Centers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current liquid cooling systems for data centers are inflexible, costly to modify, and inefficiently utilized, especially in partially populated facilities, leading to difficulties in maintaining uptime and reliability due to the need for custom designs and high costs of acquisition and ownership.
Innovation Solution
A modular liquid cooling system comprising discrete units that can be assembled and tested pre-deployment, featuring a horizontal and vertical assembly with a distribution module, allowing for redundancy and flexible cooling configurations, and enabling easy expansion and maintenance with separate fluid distribution from IT equipment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If custom-designed liquid cooling facilities are built for each data center, then cooling performance is optimized for specific requirements, but system complexity and cost of acquisition increase significantly
Solution Approach 1:
The cooling system is divided into standardized modular units that can be independently deployed and configured. Each module contains integrated components (chillers, pumps, manifolds, piping) that can be assembled in different quantities and arrangements to meet specific data center cooling requirements without requiring custom design of the entire system.
Solution Approach 2:
The modular cooling units are designed with universal interfaces and standardized configurations that allow them to serve multiple data center layouts and cooling requirements. The same module type can be deployed in various quantities and arrangements to accommodate different IT equipment densities and spatial configurations.
2Reliability
If liquid cooling facilities are built to supply cooling to the entire data center, then full cooling capacity is available, but utilization efficiency decreases when data center is not fully populated
Solution Approach 1:
The cooling system is designed to dynamically scale its operational capacity by adding or removing modular units based on actual IT equipment population and cooling demand. Modules can be independently activated or deactivated, allowing the system to match cooling output with actual load requirements and avoid energy waste from operating excess cooling capacity.
Solution Approach 2:
By segmenting the cooling system into independent modular units, each data center can deploy only the number of modules needed for current requirements. Additional modules can be added later as the data center population grows, ensuring that cooling capacity always matches actual demand without over-provisioning.
3Reliability
If liquid cooling facilities are built out, then cooling capacity is established, but modification and upgrade difficulty increases significantly
Solution Approach 1:
The system is composed of discrete modular units with standardized interfaces, allowing individual modules to be removed, replaced, or upgraded without affecting the rest of the system. This segmentation enables flexible modifications and technology upgrades while maintaining overall system functionality.
Solution Approach 2:
The modular architecture enables the system to adapt dynamically to changing requirements through easy addition or removal of modules. Upgrades can be implemented by replacing individual modules with newer versions, allowing the system to evolve without complete redesign or reconstruction.
4Reliability
If traditional liquid cooling facilities are implemented, then cooling is provided, but deployment time and site assembly complexity increase
Solution Approach 1:
Each modular cooling unit is pre-assembled, pre-configured, and pre-tested at the manufacturing facility with all components (chillers, pumps, manifolds, piping, controls) integrated and verified. This preliminary preparation eliminates time-consuming on-site assembly and commissioning activities, allowing rapid deployment when modules are delivered to the data center site.
Solution Approach 2:
By dividing the cooling system into self-contained modular units, the deployment process is simplified to delivering and connecting discrete modules rather than installing complex custom-built systems. Each module arrives ready-to-install with standardized interfaces, significantly reducing site assembly complexity and deployment time.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The modular system enhances reliability, reduces costs, and allows for efficient deployment and maintenance by enabling plug-and-play assembly, flexible scaling, and redundancy, ensuring high performance and uptime while minimizing the need for extensive customization and construction.
Implementation Method 1
Heat removal from processors can be improved by the use of liquid cooling
Implementation Method 2
Chilled water pipes then deliver chilled water from the facilities room into the server room and is connected to the various manifolds to distribute the liquid to the cooling loops
Data Source
AI summary
A two-dimensional liquid cooling system for data centers. A horizontal assembly is positioned above the IT cluster and serves as the main cooling unit. A vertical unit is positioned besides several IT clusters and is connected to several horizontal units and can be used as a backup or for enhanced cooling performance for any of the clusters. The horizontal assembly includes a cooling module that has a cooling unit, a coolant module that receives chilled liquid from the cooling unit and provides cooling liquid to a distribution layer which includes a supply loop and a return loop. The vertical unit is coupled to a plurality of supply and return loops so as to provide cooling liquid to several clusters. The cooling system and entire data center are modularized in different configurations.