Simply put, volatility describes how easily a substance will vaporize. Certainly we all recognize that as the temperature of a liquid rises, some of it becomes a gas or vapor creating a pressure above the liquid. The higher the temperature, the greater the vapor pressure until the boiling point is reached and all of the liquid is vaporized.

In the case of base oils, its even more complicated since base oil is not a single molecule (like water) but a mix of literally thousands of different but related molecules. Generally, the larger the molecule, the higher its viscosity and the lower its volatility. Compare a 4 centiStoke oil (100N) to a 12 cSt. oil (600N) to see the relationship: The heavier weight oil will be more viscous, less volatile.

From the beginnings of engine oil marketing, viscosity rather than volatility was the key factor in deciding what oil to use. The earliest engine oils were all very heavy; theyd be an SAE 30 or SAE 40 in todays version of SAE J300, the Engine Oil Viscosity Classification standard.

Early Days

When the precursor to J300 appeared in 1911, oil volatility didnt play a part in engine operations. Instead, the first volatility measurement was flash point. The concept was simple: Heat an oil in an open cup at a prescribed rate of increase and pass a flame over it every 5 degrees. When there is sufficient vapor over the surface of the oil, it will ignite when the flame is introduced. A very practical test but not scientifically elegant.

The flash point test was primarily a safety check for handling petroleum products. It also was a way to check for the presence of used oils (fuel dilution) in the finished product. Flash point was included in the earliest versions of SAE J300, but was eliminated from the viscosity classification system in 1926.

Engines improved over the next 40 years. However, volatility of the engine oil still was not given much consideration. If one needed to gauge an oils volatility, it was approximated from either the flash point or an actual distillation. Since distillation is a time-consuming process and required some pretty extensive apparatus, the much simpler flash point was preferred.

The first recognition of the need to control volatility per se came in the 1960s and was related to oil consumption. Don Smolenski, formerly with General Motors Research and now the North American OEM liaison manager at Evonik Oil Additives, notes that engines in the 1960s generally consumed less oil than their predecessors and had much longer oil drain intervals – 3,000 miles. Seeing general driver inattention to checking oil levels, automakers began to have significant concerns about evaporative losses and low oil levels in their engines.

Smolenskis remarks are echoed by Dick Kabel, former manager of the Engine Oil Group at GM Research. Both note that GM (and presumably other OEMs) increasingly wanted to reduce customer requirements for maintenance. An additional issue was the introduction of emissions systems with catalytic converters, designed to reduce air pollution. Eventually, OEMs also would identify oil volatility as a contributor to catalyst deactivation, along with some additive components.

Lighter Oil = More Vapor

Historically, the oil consumption problem perhaps was manageable when consumers were staying close to the 3,000-mile or three-month oil change interval recommendation. However, consumers became less diligent about oil changes as time went on. Meanwhile, the effort to improve fuel economy focused in large part on oil viscosity. Lower viscosity engine oils created less viscous drag in the engine and resulted in lower fuel consumption. The question was, how to maintain healthy engine oil viscosity when vapor losses kept eroding the oil in the sump?

By the 1980s, industry was seeing a general reduction in engine oil weights and base oil viscosities, and a corresponding increase in oil volatility. Flash point no longer was enough to confirm satisfactory volatility of the oil. Instead, new methods were developed (and borrowed) to get a better handle on oil volatility.

One of these methods, the Noack volatility test (ASTM D5800, or CEC L-40-A-93 in Europe) dates back to the 1930s but was not commonly used for lubricants until the 1980s. Then it was adopted by OEMs and included in the American Petroleum Institutes engine oil category requirements.

The Noack test begins by placing a precise sample of oil in a cup with a screw-on lid. The lid has a small hole in it to allow oil to volatilize freely. The cup is weighed both with and without its oil contents, and then placed in a closed furnace with a steady airstream blowing over it. The temperature is raised to 250 degrees C over the course of an hour. The cup is then weighed again and the difference in oil weight noted as the percent volatility loss.

Another test used to measure volatility of oils, ASTM D6417, uses GC, shorthand for gas chromatography. Here, the oil sample is put into an open tube or capillary, and heated at a specified rate to separate the hydrocarbons and establish their boiling points. The volatility of the oil sample then is calculated from the percentage of the oils simulated boiling range which falls below a specified temperature (371 degrees C in APIs specification).

Beginning in 1992 with the introduction of ILSAC GF-1 and API Service SH engine oils, volatility became a key part of engine oil category requirements. Each successive category became more restrictive on volatility through GF-3, but since then have remained constant, limited to 15 percent Noack losses after one hour at 250 C. (General Motors, it should be noted, went even further, and set the Noack limit for its proprietary Dexos1 engine oils at 13 percent.) The table on page 28 shows this progressive tighening of engine oil volatility limits for SAE 5W-30 oils.

Simmering Issues

Evoniks Smolenski reminds us that volatility is still very much on the table. He points to the fact that low-temperature oil viscosity can suffer when vehicles operate under high-temperature conditions (which volatilize the light ends of the oil and lead to thickening). In the GF-5 specification, the Sequence IIIGA and ROBO tests help address these volatility concerns, as well as measuring oxidation and nitration.

Two looming issues surround todays debate over volatility limits, and where they should be set for the next generation of engine oils, GF-6. The issues are Low Speed Pre-Ignition, which is a violent type of abnormal combustion, and the introduction of even lower viscosity grades (e.g. SAE 0W-16).

In the case of LSPI, there is some evidence that oil particles in the combustion chamber of very small (1-liter), highly turbocharged and direct-injected engines might cause problems and that volatility could be a factor. The implication is that less volatile oils would be a plus.

The Pre-Ignition Prevention Program, a two-year-old consortium sponsored by Southwest Research Institute in San Antonio, Texas, is working on the LSPI issue, developing background on the causes and potential fixes. If it determines that lubricants indeed affect LSPI, its expected that a separate consortium will go on to develop new lubricants and lubricant testing methods. There are indications that oil volatility is a potential issue but no results have been reported to date.

In a recent SAE paper titled Investigation of Engine Oil Effect on Abnormal Combustion in Turbocharged Direct Injection-Spark Ignition Engines, Toyota Motor Co.s Kazuo Takeuchi and his co-authors report that LSPI may restrict low speed torque improvements in turbocharged Direct Injection-Spark Ignition engines. Recent investigations have reported that the auto-ignition of an engine oil droplet from the piston crevice in the combustion chamber may cause unexpected and random LSPI, the Toyota scientists say, adding that engine oil formulations have significant effects on LSPI. The authors found that the spontaneous ignition temperature of engine oil, as determined using high-pressure Differential Scanning Calorimetry, correlates with LSPI frequency in a prototype turbocharged DI-SI engine.

Based on these findings, we believe that the oxidation reaction of the oil is a very important factor to the LSPI, added Takeuchi et al. Our test data, using a prototype engine, shows both preventative and contributory effects of base oil and metal-based engine oil additives. From this one could conclude that volatile materials in the lubricant are likely to contribute to LSPI.

Another paper, by Toyotas Satoshi Hirano, reiterated some of those findings, and went on to note that certain engine designs – particularly turbocharged DI-SI engines – seem vulnerable to the abnormal combustion phenomenon. He suggested that engine oil improvements can provide the LSPI prevention performance, which would allow for more extensive use of the benefits of turbocharged DI-SI engines while maintaining engine durability and reliability.

Refiners Dilemma

As for new, energy-saving viscosity grades, like the recently minted SAE 0W-16, these will require lower viscosity base stocks – and the relationship between viscosity and volatility is a given. The question now is what can be done to improve volatility in these lighter-weight engine oils.

One approach would be to take narrower cuts of base stocks. Its a great idea; however the economics begin to look unattractive. For every 1 percent of volatility reduction in the lower end of the boiling range, a similar amount has to be taken from the upper end in order to maintain viscosity at the same level. Unfortunately, compressing the boiling range will result in less base stock produced, less yield. As long as the world is long on base oil supply that might be a possibility, but its a costly one.

A second approach could be to harness gas-to-liquids (GTL) processes to manufacture specific base oil viscosities. It might be possible to tailor a specific stream to meet future performance needs. While this doesnt affect overall base oil capacity it does cost. The question will then be whether or not the cost-benefit is right.

A third approach will be to add or increase the volume of Group IV (polyalphaolefins) or Group V (esters) in the blend of base stocks used. Again the benefit to performance could be there but at a cost that might not be justified, at least at first. As more engines come into the marketplace with very low viscosity, low volatility oil requirements, the technical demand would seem to be there.

Certainly the emphasis on volatility will continue, as engine oil recommendations move towards lower viscosity grades and newer engine oil categories are made available.