A Faster and Better FluidScan with Version 5
Measure water contamination in turbine oils down to 300 ppm
The latest version of FluidScan software can detect total water presence as low as 300 ppm in turbine oils. The method provides an alternative to laboratory tests such as Karl Fischer titration when fast, simple, and reagent-free analysis is needed. In addition to measuring water contamination, the FluidScan analyzer provides immediate, reagentless measurement of acid number (AN), base number (BN), oxidation, glycol, and other parameters.
Water contamination in turbine and other industrial oils can be a very serious issue and water testing is always a part of any lubricant condition monitoring program. Turbine oils typically are formulated to have high thermal stability, oxidation resistance, and excellent water separation. Lubricants available specifically for gas turbines or steam turbines are designed with specific additive formulations, but there are also many oils that can work with all different types of turbines. Gas turbines have the tendency to build up sludge and varnish whereas steam turbines may experience oxidation, foaming, and sludge. However, a concern of all turbine systems is water contamination. Severe water contamination can cause changes in the oil’s viscosity, accelerated oxidation, additive depletion, and decreased bearing life. Turbine manufacturers typically recommend a warning alarm limit of >1000 ppm.
The most widely accepted method for detecting water in oil is by Karl Fischer (KF) coulometric titration (ASTM D6304). This titration method is somewhat cumbersome, as it requires hazardous reagents, careful sample preparation, expensive equipment, and at least several minutes per analysis. However, Karl Fischer analysis for water can yield highly accurate and repeatable results when executed by a skilled operator and is the comparative method for other analytical techniques for water determination. Also, the water does not have to be fully dissolved in the oil.
The FluidScan can detect the light scattering of water droplets present in oil by a lift in the baseline of the infrared absorbance spectrum. Figure 1 shows several FluidScan spectra of used turbine oil samples with high levels of water contamination.
Figure 1. FluidScan spectra of used turbine oil heavily contaminated with water used to monitor a vacuum dehydration process at a power generation plant.
The degree of light scattering caused by a water-in-oil mixture indeed depends on the concentration of water present, but it also is strongly influenced by how the water is physically dispersed in the oil: the number and size of discrete water droplets present in the oil.
For this reason, it is important to have representative, homogeneous sampling. A portable instrument such as the FluidScan can be used at the sampling site for immediate results where the oil and water will be homogeneous due to the turbulent motion inside the instrument. If the samples are left to settle, perhaps during transit to a designated oil analysis site or laboratory, the water will eventually separate. After the water has completely separated from the oil, it is difficult to get accurate measurement of the water content.
Method
A new water calibration which measures light scattering due to the presence of water droplets is available on the FluidScan for the Industrial Library. The method was developed with water-contaminated samples of several popular brands of turbine and gear/bearing oils for a robust universal calibration of industrial fluids ranging from 1,000 ppm up to 65,000 ppm water. An important component of the method is the use of a homogenizer. The samples were homogenized with a commercially available mechanical homogenizer and allowed to sit at room temperature for two minutes (no more than 30 minutes) prior to measurement on the FluidScan.
Results
Sixteen samples between the range of 500 ppm and 10,000 ppm water contamination were used to test the Total Water FluidScan measurement against Karl Fischer D6304. Each sample was prepared by homogenizing them for 30 seconds on high prior to analysis. They were measured simultaneously on three FluidScans and by Karl Fischer to minimize the effects of sampling errors. The results can be seen below.
To read the full application note click here.
Customer comments
No comments were found for A Faster and Better FluidScan with Version 5. Be the first to comment!