Immersion Fluid Comparison: Single vs Dual Phase

Immersion cooling submerges ASICs in a dielectric fluid that carries heat away far faster than air, enabling higher hashrate, quieter operation, and longer hardware life. But the fluid is not an afterthought — it is the most consequential choice in the build. An immersion fluid comparison comes down to single-phase versus dual-phase systems and, within single-phase, mineral oil versus synthetic and engineered dielectrics, each with different costs, performance, and maintenance burdens. This guide compares the options with a trade-off table and explains how to match a fluid to an operation. It assumes familiarity with the basics covered in the immersion-cooling explainer linked below.

Single-phase versus dual-phase: the core split

The first fork is whether the fluid changes state. Single-phase immersion keeps the fluid liquid throughout; a pump circulates it past the miners to a heat exchanger. Dual-phase immersion uses a fluid that boils at a low temperature on the hot chips, rising as vapor and condensing on a coil above the tank — heat transfer by phase change, no pump needed for the primary loop.

Dual-phase moves heat extremely efficiently and runs near-silent, but the engineered fluids that boil at the right temperature are expensive and have faced supply and environmental scrutiny. Single-phase is simpler, cheaper, and the dominant choice for ASIC immersion today. The mechanics of both are introduced in the immersion cooling explainer; this guide focuses on choosing the fluid.

The fluid options compared

Within single-phase, the practical choices range from inexpensive mineral oils to purpose-engineered synthetic dielectrics. The table below summarizes the trade-offs. Figures are general ranges, not vendor quotes — fluid pricing and specifications vary by supplier and should be confirmed against a manufacturer data sheet before purchase.

Why viscosity matters

Viscosity drives pump energy and heat-transfer efficiency in single-phase systems. Thicker mineral oil needs more pumping power and transfers heat slightly less readily than a low-viscosity synthetic. Over a large tank, the pumping-energy difference is a real line on the power bill, which can erode the cooling efficiency gain immersion is meant to deliver. Lower-viscosity synthetics cost more upfront but recover some of that in operating efficiency.

Viscosity also changes with temperature, which complicates the picture. A fluid that pumps easily when warm can be sluggish at start-up in a cold room, straining the pump until the loop heats. In colder climates this argues for a fluid with a flatter viscosity-versus-temperature curve, or for managing the room so the tank does not sit cold. The fluid data sheet states the viscosity at reference temperatures; reading it at the actual operating range matters more than the single headline figure.

Dielectric strength and water ingress

The whole premise of immersion is that the fluid does not conduct electricity. Dielectric strength is the measure of that, and it degrades if water contaminates the fluid. Humid air, condensation, and any leak can introduce moisture, and once dielectric strength falls far enough the fluid stops protecting the hardware it surrounds. This is why fluid testing is not optional maintenance — it is the check that the core safety property still holds. A fluid that tests with falling dielectric strength is a warning to find and stop the water source before it causes a fault.

Material compatibility and hardware prep

Not every miner component tolerates every fluid. Some plastics, thermal pads, fan bearings, and gaskets degrade in certain oils, and stock thermal interfaces are designed for air. Operators typically remove fans, replace incompatible thermal pads, and confirm gasket and label compatibility before submerging a unit. Skipping this prep is a common cause of early failures attributed wrongly to the fluid.

Manufacturers increasingly sell immersion-ready or hydro variants engineered for liquid cooling — Bitmain’s hydro line and Fluminer’s liquid-oriented Scrypt units are examples. Buying purpose-built hardware sidesteps much of the prep risk. The catalog distinction between air, hydro, and immersion-ready models matters here; current options are listed in the Coin Web Mining catalog, including the Scrypt mining hardware suited to liquid setups.

Maintenance burden by fluid type

Fluid is not maintenance-free. Single-phase systems need filtration to remove particulates, periodic testing of dielectric strength and acidity, and top-ups as fluid is lost to handling. Mineral oil degrades faster and may need replacement sooner; synthetics hold their properties longer. Dual-phase systems trade pump maintenance for vapor-seal integrity and condenser upkeep.

The ongoing routine is its own discipline. Testing the fluid, watching for water ingress, and managing particulate load keep a tank healthy over years rather than months. That routine is detailed in the immersion fluid maintenance guide, and the broader water-and-loop upkeep for hydro hardware is in the water loop maintenance guide. Budgeting for this upkeep is part of the true cost comparison, not an extra.

Total cost: fluid is a multi-year line item

The headline fluid price is only the start. Total cost over a deployment includes the initial fill, top-ups, filtration consumables, periodic testing, and eventual partial or full replacement. A cheaper mineral oil with a shorter service life can cost more over five years than a pricier synthetic that lasts. Frame the comparison per year of service, not per liter at purchase.

Against that, immersion’s payoff is real: higher sustained hashrate per unit, dramatically lower noise, dust elimination, and extended hardware life from stable, lower temperatures. Those benefits interact with tuning — a well-cooled immersion unit can sit higher on its power curve, a point developed in the power curve optimization guide. The fluid choice is what makes that headroom available.

The headroom is the quiet economic argument for immersion beyond noise and dust. An air-cooled unit is capped by how much heat its fans can move; an immersion unit’s ceiling is set by the fluid and heat exchanger, which can be far higher. That lets the same chip run at a more productive point without thermal-throttling, recovering output an air setup leaves on the table in a warm room. Whether that recovered output justifies the fluid and engineering cost depends on the operator’s power rate and ambient conditions, so the comparison is operation-specific rather than universal. Frame the decision as a multi-year total-cost question — fluid, pumps, maintenance, and disposal against the output and longevity gains — not a single up-front number.

Environmental and regulatory considerations

Fluid choice carries an environmental dimension that affects both ethics and compliance. Some engineered two-phase fluids have faced scrutiny over persistence and global-warming potential, and regulation in several jurisdictions has tightened around certain fluorinated compounds. An operator selecting a fluid for a long deployment should check the current regulatory status in their jurisdiction rather than assume availability, because a fluid that is freely sold today may face restriction over a tank’s multi-year life.

Disposal is the other end of the lifecycle. Used dielectric fluid is not something to pour down a drain — it requires proper handling and disposal, and mineral oils, synthetics, and engineered fluids differ in how they must be treated. Factor end-of-life handling into the choice, and treat any disposal question as one for the relevant environmental rules rather than a casual decision. The maintenance routine that keeps fluid healthy also extends its life, deferring that disposal cost; the cadence is in the immersion fluid maintenance guide.

Choosing a fluid for the operation

For a first single-tank build or a budget-constrained home setup, a quality single-phase mineral oil or entry synthetic is the pragmatic start. For a farm prioritizing long service intervals and pumping efficiency at scale, a low-viscosity synthetic or engineered dielectric earns its premium. Dual-phase suits operations that need its peak heat-transfer and silence and can absorb the cost and engineering — a smaller niche given current fluid pricing and scrutiny.

Match the fluid to immersion-ready hardware rather than retrofitting air-cooled units where possible, and confirm every specification against the supplier’s data sheet before committing a tank’s worth of fluid. For sourcing hydro and immersion-suited models to build around, the shop carries current-generation options, with bulk terms available via a quote request for multi-unit tank builds.

Whatever fluid is chosen, treat the first tank as a learning deployment. Fill one tank, run it through a full season, and measure the real pumping energy, the rate of fluid loss, the testing results, and the actual hardware temperatures before scaling to multiple tanks. The data sheet states what a fluid should do; a single tank’s experience reveals what it does in the specific room, climate, and hardware mix at hand. That measured baseline turns the second and third tanks from guesses into informed repeats, and it surfaces any compatibility or maintenance surprise while it is still small and cheap to correct rather than multiplied across a whole immersion farm.

References

• Hydro and immersion-ready Antminer specifications — Bitmain
• Immersion and hydro model efficiency data — ASIC Miner Value
• Liquid-cooling hardware reference — MicroBT
• Immersion deployment trends and economics — Hashrate Index

Planning an immersion build and weighing hardware against fluid? The Coin Web Mining catalog carries current hydro and immersion-suited models with published efficiency figures, and a bulk quote covers multi-unit tank orders with escrow on first purchases.