Lubricant mix up: Hydraulic fluid traces identified in turbine engine oil

Turbine oil contamination

The oils and lubricants for military and civilian aircraft are designed in line with NATO and military specifications to maintain the high performance capabilities of the individual components and contribute to their prolonged existence. These specifications are intended to improve performance rather than stipulate the composition, although some ground rules on content are applied.

For example, turbine oils usually contain a synthetic polyol as base stock. In contrast, hydraulic fluids designated H-537 and H-515 that meet US military specifications contain a synthetic hydrocarbon such as a hydrogenated poly-α-olefin or a mineral oil as base stock, respectively, with various additives. The similar base components make it difficult to identify cross-contamination between hydraulic fluids and turbine oils, which can occur via operator error or from leaks.

A team of scientists at the Defence Science and Technology Organisation, Fishermans Bend, Victoria, Australia, have been looking at the contamination problem in detail. Renée Webster, David Evans, Paul Rawson recognised the failure of bulk testing methods to identify contamination. For instance, FTIR spectroscopy cannot distinguish the C-H absorptions of the base components.

Other methods that rely on identifying the presence of wear additives consisting of phosphorus-containing compounds are unreliable due to the variable low content and composition of these additives, which might also vary from batch to batch. Reliance on the additives is risky because suppliers can change the type and amount of additives without notice.

So, they turned to GC/MS. The electron impact fragmentation patterns of the polyester and polyolefin components, which coelute, are too similar for confident and reliable identification of contamination, as they both have alkane skeletons. In its place, the researchers used positive-ion chemical ionisation, which is a more gentle process than EI and gives different fragmentation patterns for the base components.

Fluid argument resolved

Three hydraulic fluids and four turbine oils meeting US military specifications were studied. The fluids within each group originated from different manufacturers. For validation, small amounts of one hydraulic fluid were added to one turbine oil after dilution with heptane and the samples were injected directly to the GC/MS system.

A conventional 5% diphenyl-95% dimethylpolysiloxane column was used to separate the components and chemical ionisation was carried out with methane. Full-scan and selected ion monitoring (SIM) data were collected at the same time.

The GC/MS chromatograms show that the H-515 hydraulic fluid elutes earlier than the turbine oil so can be identified. Conversely, H-537 overlaps the oil completely. It was impossible to separate the signals of clean turbine oil from those of a mixture of turbine oil containing 5% H-537, even using integration software.

However, the SIM data had the ability to differentiate. Ions in the mass spectra at m/z 309, 335, 421 and 424 belonged uniquely to hydraulic fluids. They were totally absent from the turbine oil. The identities of these ions could not be ascertained but their exclusivity to the hydraulic fluids and their consistent occurrence within a particular retention time window gave the researchers confidence in their assignment to the hydraulic fluids. This conclusion was supported by the absence of the ions in artificially aged oils.

The SIM chromatograms of all four ions from the three hydraulic fluids were remarkably similar, even though they originated from different manufacturers. The areas under the curves were used to estimate their concentrations with detection and quantification limits of 0.05 and 0.5% (w/w), respectively. The recoveries from oils spiked with the fluids were higher than 90%.

The procedure was used to estimate the degree of contamination of turbine oils from in-service aircraft of the Royal Australian Navy Fleet Air Arm that had been spiked with hydraulic fluid at concentrations of 0.8-3.0% (w/w). The recoveries were excellent, ranging from 98.4 to 108.6%, showing that the method can be applied to oils from working aircraft to check their composition and ensure that no contamination has occurred that will affect the performance and efficiency of the turbine engines.