Immersion cooling is an IT cooling practice by which IT components and other electronics, including complete servers, are submerged in a dielectric (electrically non-conductive) fluid that has significantly higher thermal conductivity than air (typically >1,000x the heat capacity by volume). Heat is removed from the system by circulating liquid into direct contact with hot components, then through water cooled heat exchangers.
Immersion Cooling is not new. The first reference to the specific use of dielectric fluids being used to cool “computers” is in 1966 by IBM and in 1982, Cray computing registered a patent for high density immersion cooled technology.
GRC was founded in 2009 and has had commercial deployments of its rack based Immersion Cooling systems for over 10 years. The company deployed the world’s first turnkey immersion cooled (modular) datacenter, the ICEtank, in 2013.
Asperitas was founded in 2014 and launched their first Immersed Computing® solution in 2017, after an extensive R&D phase with an ecosystem of cutting edge partners. Asperitas is leading the Open Compute sub project on Immersion Cooling for standardisation development.
Single-phase Immersion Cooling, as the name suggests, is where the coolant stays in a single (liquid) phase, and does not evaporate. The coolant captures the heat from the immersed components and is circulated through a heat exchanger that transfers the heat to a water loop, which in turn can be cooled with an evaporative/adiabatic cooling tower, dry-cooler (radiator), or an existing chiller plant (not required, but compatible).
Two-phase Immersion Cooling is where the coolant is designed to evaporate (change phase from liquid to gas) at lower temperatures, boiling off when it comes in contact with hot components. The evaporated coolant vapor is then condensed back to the liquid state through the use of condenser coils, typically located at the top of a sealed rack.
There are primarily two kinds of coolants used for Immersion Cooling: Fluorocarbon-based coolants which are typically used for two-phase applications, and Hydrocarbon-based fluids which are solely used for single-phase Immersion Cooling.
Single-phase immersion requires circulation of the dielectric liquids by pumps or by natural convection flow through integrated or external heat exchangers.
Hydrocarbon fluids used for single-phase Immersion Cooling are typically synthetic fluids, which should not be confused with mineral oil. Mineral oils can often have imperfections, impurities, and limit material compatibility. Synthetic fluids, on the other hand, are manufactured at a molecular level which results in an inherently stable product that offers superior performance and material compatibility.
They are clear, odorless, non-toxic fluids that are readily used in domestic products such as cosmetics and other household products, which have well-established material handling practices, minimal regulations, and no direct GWP (Global Warming Potential). The latter can be a challenge with Fluorocarbons.
Both GRC and Asperitas exclusively work with superior synthetic immersion fluids which are well developed by reputable partners and fully tested and optimized for Immersion Cooling.
There are primarily two kinds of coolants used for Immersion Cooling: Fluorocarbon-based coolants which are typically used for two-phase applications, and Hydrocarbon-based fluids which are solely used for single-phase Immersion Cooling.
Fluorocarbon-based coolants may be used in single-phase applications, but typically offer little benefit at a significantly higher cost and added complexity.
One of the advantages of using single-phase Immersion Cooling technology is that the fluids being used are not influencing the business case strongly, while, at the same time, offering the most reliable environment for electronic components.
Unlike fluorocarbon fluids, the hydrocarbon fluids used by GRC and Asperitas do not evaporate under normal operating conditions and do not need to be replaced through the life of the datacenter (15+ years). Regular lab tests are performed on the coolants to test for any variance in critical properties over time. Commercial deployments dating back to as early as 2010 have been running efficiently and reliably, with no degradation in the coolant’s performance.