Gas Dilution in Compressor Lubricants: How Dissolved Gas Reduces Operating Viscosity
The lubricant inside a running compressor differs from the lubricant poured in. Process gas or refrigerant dissolves into it, proportional to partial pressure. This dissolved gas—typically hundreds to thousands of times less viscous than the oil—reduces the working viscosity. Heat further contributes to this reduction. As a result, a machine charged with an ISO viscosity grade (ISO VG) 220 lubricant might operate its bearings on a fraction of that figure, yet this isn’t reflected on the data sheet or in a standard oil sample. This page explains the mechanism; the calculation that quantifies it for a specific machine is covered separately.
The viscosity of the lubricant as it exists inside the machine—heated to operating temperature and diluted by dissolved gas—is what protects a compressor. No data sheet shows that number, and a depressurized oil sample understates it.
Dilution is driven by the partial pressure of the dissolving component, oil temperature, and lubricant chemistry—not by total pressure alone. The oil in a flooded-screw separator sits at discharge pressure, which is why discharge normally governs.
Heavier hydrocarbons dissolve into a lubricant far more readily than lighter ones. Methane causes limited viscosity loss; pentane and heavier hydrocarbons present a severe case. CO₂ readily dissolves into polyalkylene glycol (PAG) and significantly contributes to dilution.
Chemistry dictates the extent of dilution. Mineral oils and polyalphaolefin (PAO) are non-polar, like hydrocarbon gas, and mix freely with it. PAG is polar and dissolves considerably less. Polyethylene glycol (PEG) has very low hydrocarbon solubility and maintains operating viscosity closest to its neat value.
In refrigeration the same mechanism works both ways: the dissolved refrigerant that threatens viscosity at the discharge end is what makes oil return physically possible at the cold end.
In a measured NEXT example, a fresh ISO VG 220 lubricant delivers 16.7 cSt at discharge conditions. The gap between the rated grade and the working viscosity is the reason dilution is calculated rather than assumed.
What dissolved gas does to a lubricant
Gas dissolves into a lubricant proportional to its partial pressure—the total absolute pressure multiplied by that component’s mole fraction. Once dissolved, it thins the mixture because the gas itself is typically hundreds to thousands of times less viscous than the oil it enters. Enough dissolved gas can cut a lubricant’s viscosity by as much as an order of magnitude.
Dilution is not a contaminant in the conventional sense. It produces no wear particles, does not change the color of the oil, and leaves no residue. What it changes is the lubricant’s working condition — affecting oil film thickness at journal bearings and rotor contacts, sealing of rotor clearances, separator behavior, oil consumption, and foaming tendency when pressure changes and gas comes out of solution.
The effect is also fast. Laboratory work scaled to real compressor cylinder geometry indicates that dilution reaches equilibrium in roughly 10 to 21 seconds — and probably faster still in a running machine, given turbulence, pressure fluctuation, and thermal convection. There is no meaningful period during which the lubricant is less diluted than the equilibrium value. The equilibrium viscosity is the working viscosity.
One critical distinction must be made. Dissolved gas thins the entire oil charge evenly and reversibly. Liquid refrigerant or condensate arriving in the machine is a different failure: it washes the lubricating film off surfaces, and no chemistry choice prevents it. Dissolved gas is a lubricant selection problem. Condensed liquid is a process problem — requiring scrubbing, superheat margin, and dew point control.
Dissolved gas is not condensed liquid
One critical distinction must be made. Dissolved gas thins the entire oil charge evenly and reversibly. Liquid refrigerant or condensate arriving in the machine is a different failure: it washes the lubricating film off surfaces, and no chemistry choice prevents it. Dissolved gas is a lubricant selection problem. Condensed liquid is a process problem — requiring scrubbing, superheat margin, and dew point control.
What drives gas dilution
The four variables
Four variables set the operating viscosity inside the compressor, and every dilution assessment needs all four.
| Variable | What It Does | The Detail That Gets Missed |
|---|---|---|
| Gas or refrigerant composition | Solubility is calculated component by component and summed. Heavier hydrocarbons dissolve far more than methane; CO2 dissolves readily into PAG. | Lubricants preferentially absorb heavier components, so a stream reading 92% methane does not produce dissolved gas that is 92% methane. Truncating an analysis at butane removes what matters most. |
| Operating pressure | Solubility rises with pressure, so partial pressures—and dilution—peak at discharge. | In a flooded screw, the oil separator sits at discharge pressure. Assessing the machine at suction pressure badly understates the duty. |
| Oil temperature | Hotter oil holds less gas, but hotter oil is also thinner—the two effects overlap at exactly the point where the compressor carries its highest loads. | Viscosity is assessed where the oil is delivered. A 20 °C (36 °F) error in oil temperature moves the answer more than most composition errors. |
| Lubricant chemistry | Often the largest variable of all: different base stocks dissolve the same gas to very different extents under identical conditions. | A heavier grade of the same chemistry dissolves the same proportion of gas by weight. Grade changes the starting point; chemistry changes the loss. |
Polarity: why chemistry sets the spread
The mechanism behind the chemistry variable is polarity, and the working rule is the one every chemist learns first: like dissolves like.
Mineral oils, PAO, and hydrocarbon gas are all non-polar, so they mix freely—which means heavy dilution in hydrocarbon service. PAG is polar: roughly every third atom along its backbone is oxygen, and that polarity makes hydrocarbons considerably less soluble in it. Published comparisons put methane solubility in PAG at roughly half that in PAO or mineral oil, with PAO and mineral behaving similarly to one another. At the far end of the spread, PEG—pure ethylene oxide chemistry—has very low hydrocarbon solubility, and its operating viscosity stays close to the neat viscosity at temperature.
That spread is the entire basis of the chemistry ladder used across NEXT’s application articles: hydrotreated mineral and PAO where dilution is manageable, water-insoluble PAG where viscosity retention starts to govern, water-soluble PAG (an ethylene oxide/propylene oxide copolymer) for heavy hydrocarbon duty, and PEG where dilution must be near-eliminated.
Two things the ladder is not. It is not a quality ranking—each step trades miscibility and cost for dilution resistance, and a step above the duty buys a property the machine cannot use. And it is not one-directional: the same polarity that rejects hydrocarbons makes PAG absorb CO₂ readily, so on a CO₂-bearing stream the PAG advantage narrows and the calculation has to include both contributions.
The scale also reads in a second direction, and refrigeration is where that matters. In gas compression, the lubricant is chosen far from the gas on the polarity scale, because distance means low solubility and retained viscosity. In refrigeration, the deciding property is miscibility, and the scale is read by distance in whichever direction the oil-return design requires. Where the system needs immiscibility—ammonia, with its separation-based oil management—the lubricant comes from the opposite end of the scale: ammonia is polar, so non-polar mineral oil and PAO stay separate from it and drain from defined collection points. Where the system needs miscibility—CO₂, where the refrigerant carries the oil back from the evaporator—the lubricant is chosen close to the refrigerant on the scale, which is why POE is the standard choice there. One scale, two selection rules: distance for dilution resistance, and proximity or distance for the miscibility the circuit demands.
Gas Dilution Measured: Same Conditions, Different Chemistries
Two NEXT calculation sets follow — one on a hydrocarbon stream, one on CO₂. Both use illustrative inputs constructed by NEXT. The calculations are extrapolated from measured solubility data on NEXT products and apply to NEXT products specifically.
Four chemistries on one propane stream
The stream is a manufactured commercial propane composition: propane 96.500, isobutane 2.000, n-butane 0.800, ethane 0.700 mol %. The evaluation point is the oil at 80 °C (176 °F) under 20 bar (290 psi) absolute — specifically, the oil separator of a system at high condensing pressure, representing the most demanding propane duty. All four lubricants are ISO VG 150, allowing for a comparison based solely on chemistry. A minimum viscosity of 10 cSt is required at the lubrication point.
| Gas Component | Mol % | Operating Condition | Value |
|---|---|---|---|
| Propane | 96.500 | Oil temperature | 80 °C (176 °F) |
| Isobutane | 2.000 | Oil separator pressure | 20 bar (290 psi) absolute |
| n-Butane | 0.800 | Evaluation point | Oil separator |
| Ethane | 0.700 | Lubricant viscosity grade | ISO VG 150 |
| Total | 100.000 | Minimum required viscosity | 10 cSt at the lubrication point |
Find below the results
| Lubricant | Chemistry | Dissolved Propane | Operating Viscosity | Against 10 cSt |
|---|---|---|---|---|
| NEXT GPL PAO-150 | Polyalphaolefin | 22.34 wt% | 2.12 cSt | Fails — below the film-failure region |
| NEXT GPL PAG-150 | Water-insoluble PAG (PO) | 13.87 wt% | 7.74 cSt | Fails at this condition |
| NEXT GPL PAG-WS-150 | Water-soluble PAG (EO/PO) | 8.67 wt% | 15.6 cSt | Passes |
| NEXT GPL PAG-EO-150 | PEG | 3.14 wt% | 22.7 cSt | Passes with the most headroom |
With the same gas, operating conditions, and ISO grade, the four chemistries exhibit a tenfold difference in delivered viscosity. This ordering directly reflects the polarity ladder: each step up in polarity leads to less dissolved propane and better viscosity retention.
Why a heavier grade isn't always the solution
The intuitive response to a viscosity shortfall is to use a heavier grade. However, the underlying mechanism explains why this approach has limitations.
Dissolved gas enters the oil as a proportion of its weight, determined by the lubricant’s chemistry, partial pressures, and temperature — not by its fresh viscosity. A heavier grade of the same chemistry will absorb essentially the same percentage of gas, meaning each grade step faces the same dilution penalty and provides diminishing returns in additional working viscosity. In severe hydrocarbon service, published NEXT comparisons demonstrate that a chemistry with very low solubility can meet a viscosity requirement with a light grade, while a high-solubility chemistry cannot achieve the same with its heaviest available grade. Furthermore, a grade heavy enough to compensate for high solubility might be too viscous for cold starts or proper oil return.
The grade selects the starting viscosity, but the chemistry dictates the viscosity loss. When viscosity loss is the problem, chemistry is the key leverage point.
Working viscosity is calculated, not sampled
The working viscosity is determined by calculation, not sampling. This calculation uses measured solubility data for the specific lubricant, an equation of state fitted to that data, and a viscosity mixing rule. NEXT performs this calculation based on a completed Process and Mixed Gas Lubricant Recommendation Form. The results are specific to NEXT products because the underlying model relies on proprietary NEXT product data. The method, its inputs, and its boundaries are fully detailed in the calculation article.
What is gas dilution in a compressor lubricant?
Gas dilution is process gas or refrigerant dissolving into the lubricant under pressure, in proportion to each component’s partial pressure. The dissolved gas reduces the lubricant’s working viscosity inside the machine — evenly, reversibly, and without producing wear particles or changing the oil’s appearance — which is why it is assessed by calculation rather than by inspection.
How much viscosity can dilution remove?
Up to an order of magnitude in severe service. In a measured NEXT comparison, an ISO VG 150 PAO delivered 2.12 cSt on a near-pure propane stream — below the region where an oil film carries load — while a PEG of the same grade delivered 22.7 cSt at the identical condition. The size of the effect depends on the gas composition, the pressures, the oil temperature and above all the lubricant chemistry, which is why it is calculated per operating point.
Does low suction pressure cause dilution?
No — this is the most persistent misconception in the subject. Solubility rises with pressure, so the near-atmospheric suction side is the least diluting condition in the machine. Dilution peaks at discharge and in the oil separator, where the oil sits at discharge pressure in contact with fully compressed gas. Low suction pressure matters indirectly: it produces high compression ratios, and it signals a rich stream that came off a liquid surface.
Why doesn't standard oil analysis show dilution?
Because the sample is depressurized before testing. The dissolved gas leaves the oil between the sample point and the laboratory, so the measured viscosity describes the degassed fluid, not the working fluid. Analysis still earns its place through trends — a sustained viscosity fall usually reports the gas changing — but the working viscosity comes from calculation.
Does charging a heavier grade fix dilution?
No. A heavier grade of the same chemistry dissolves essentially the same proportion of gas by weight, so every grade step starts from the same penalty and buys less than the one before — and a grade heavy enough to compensate for severe dilution fails at cold start and oil return instead. Where dilution is the problem, the lever is a lower-solubility chemistry, not a higher fresh number.
Which chemistry resists gas dilution best?
For hydrocarbon service, PEG by a wide margin: its very low hydrocarbon solubility keeps operating viscosity close to the neat value. Water-soluble PAG is next, then water-insoluble PAG, then PAO and mineral, which behave similarly to one another. The ranking is a severity ladder, not a quality ranking — the most resistant chemistry also places the greatest demand on oil return, and in refrigeration the correct chemistry depends on the refrigerant, not on dilution resistance alone.
How is dilution calculated for a specific machine?
From a completed Process and Mixed Gas Lubricant Recommendation Form: full gas composition in mol % including CO₂, water vapor and H₂S, per-stage suction and discharge conditions, oil injection and sump temperature, and the OEM’s minimum viscosity requirement. The output is the dissolved-gas percentage and operating viscosity at each condition, with a chart plotting the pure lubricant against the diluted mixture. The results apply to NEXT products specifically, because the model rests on measured NEXT product data.