A Dynamic Diagnostic Water-In-Oil Test

Moisture in oil is one of the most serious contaminants and its presence must be carefully monitored in lubricating oils. From the simple nonquantitative crackle tests to the sophisticated Karl Fischer method, hundreds of water-in-oil tests are performed annually.

Most of the methods used today are performed in the laboratory by trained analytical chemists and technicians. Until now, quick, easy and quantitative results have not been practical and the turnaround time for most analysis can be as long as two weeks. However, a new test kit by Dexsil, called HydroScout, makes real-time water monitoring both easy to run and cost-effective.

HydroScout was originally designed for quantifying percent levels of water in used oil destined for recycling. By increasing the sample size and modifying the chemistry, this new test kit has a detection limit as low as 50 ppm for in-service monitoring of lubricating oils. The benefits of this quantitative method are its field portability for real-time results, ease of use with relatively no set-up time, precalibrated instrumentation, and prestandardized and environmentally safe reagents.

The kit’s method is based on the standard reaction of water with calcium hydride to produce one mole of hydrogen for every mole of water (a mole is simply a measure used by chemists to define Avogadro’s number of molecules).

The easy-to-use kit utilizes a disposable plastic reaction tube that houses a reagent-fill ampule into which the sample oil is poured. By squeezing the reaction tube, the ampule is broken, releasing the reagents to mix with the oil. After shaking the reaction tube, followed by a brief reaction period, the tube is inserted into the HydroScout meter.

A hypodermic needle within the meter punctures a rubber septum that is fitted on the reaction tube’s cap. The meter measures pressure buildup in the reaction tube due to the release of hydrogen. Using the ideal gas law, the internal pressure of the reaction tube is converted into the amount of water in the sample. Depending on the program chosen, stored constants are used to calculate the water content in the oil.

To verify the effectiveness of this method in reacting all of the water in an oil sample, six different turbine oils were spiked with water at various levels, reacted with calcium hydride and the pressure measured. As a control, the actual water content of the oils was determined using a Karl Fischer method using azeotropic co-distillation.

Analyzing these oils using the standard oil analysis test procedure allowed Dexsil to evaluate the accuracy of the HydroScout in describing the water-in-oil content based on the measured pressure (Figure 1). The solid line in Figure 1 is the theoretical result predicted from the Karl Fischer data. Figure 2 shows the same plot for motor oils and hydraulic fluid: note that the R2 from the regression analysis was 0.98, indicating very good correlation between the two test methods.

Figure 1. Comparison Data HydroScout vs. Karl Fischer on Turbine Oils

To achieve complete reaction, all of the water present must come into contact with the calcium hydride. However, because the water is suspended in a generally nonpolar matrix (the oil), it might be expected that the type of oil tested will have an effect on the ability of the calcium hydride to come in contact with all of the water and hence on the accuracy of the HydroScout result.

Among the matrix parameters expected to affect the results is the viscosity of the base stock and the polarity of the additives used to formulate the oil. The viscosity of the oil tends to physically prevent the oil from mixing with the calcium hydride whereas the additives, being generally more polar than the base stock, tend to segregate the water from the reactants by chemical attraction, holding the water in solution in the oil.

To investigate these effects, various oils and hydrocarbon-based fluids (nondetergent and detergent motor oils, single and multiviscosity motor oils, gear oil, brake fluid and hydraulic oils) were spiked at different levels, measured using the new procedure and the results compared to Karl Fischer.

The results from the oils and fluids tested tended to fall into three groups, based on the affinity of the fluid for water. Lighter turbine oils and fuel oils, as well as the more viscous gear oils, produced nearly theoretical pressure readings indicating that the water was reacted completely.

A second group was identified with approximately an 85 percent to 90 percent recovery, comprised of the single viscosity, nondetergent motor oils, brake fluid and mineral oil dielectric fluids. A third group was also evident, with approximately a 60 to 65 percent recovery. This group includes the multiviscosity motor oils, detergent motor oils and hydraulic oils (Figure 2).

Other fluids tested such as brake fluid, gear oils and mineral oil dielectric fluids fall on one of the preprogrammed response curves allowing the instrument to calculate the water content for different fluids based on the oil type and the measured pressure.

Figure 2. HydroScout Results for Motor Oils and Hydraulic Fluids

Programming the meter with three different conversion programs allows for the accurate estimation of low levels of water in all three groups of oils. For the specific oil/fluid-types tested here, the correct response factors have been determined. For other types, a single comparison point can be used to determine the correct program to use.

Through spiking experiments with compounds other than water, it has been determined that this method has no interference from alcohols, ketones, propylene glycol, glycol esters, polyglycols and various metal oxides. However, interferences from ethylene glycol and some organic and inorganic acids have been observed.

The HydroScout has proven to be an accurate method for determining water in a wide variety of lubricating oils. The reagents are premeasured, sealed in glass ampules and can be disposed of in normal laboratory waste making this method an easy-to-use field-screening tool for water-in-oil measurements.

Some cross-reactivity was observed and the user must choose the program that best fits his type of oil. Nevertheless, HydroScout can be an effective diagnostic test on a variety of oils comparing favorably with the Karl Fischer method.

Subscribe to Machinery Lubrication