Moisture Analysis of Ethanol-Hydrocarbon Blends by Vis-NIR Spectroscopy

While conventional lab methods were generally used to analyze the water content in ethanol-hydrocarbon blends, visible near-infrared spectroscopy (Vis-NIRS) can be used as a faster alternative. Vis-NIRS fast-tracks the inspection of raw material, monitoring the process and the final product control.

Introduction

The presence of water in the fuel can damage the fuel tanks and engine by causing erosion, water etching, cavitation, spalling. In addition, the presence of moisture could also result in microbial growth and fuel oxidation. While rust and corrosion caused by moisture has always been a problem in fuel systems, modern high-pressure fuel systems are more sensitive to the conventional low-pressure systems.

Therefore, manufacturers of fuel systems are specifying that there should be no free water reaching the engine [1]. To avoid the presence of free water fallout and ensure that it stays dissolved in the fuel, it must be kept below its saturation point which vary from 0.005% to 0.18% depending on the temperature and on the ratio of petro diesel / biodiesel. It is therefore important to monitor the amount of free water in the fuel regularly to ensure the longevity of the engine.

Traditional Methods Used for Free Water Detection in Fuels

The standard method used for the determination of free water in lubricating oils, petroleum products, and additives is ASTM standard D6304. The ASTM standard D6304 is a method based on the Karl Fischer (KF) Titration, and is widely used in manufacturing, purchase, sale, or transfer of petroleum products to analyze the water content of petroleum products. This conventional method not only requires use of toxic solvents, but it also has high costs of routine analysis, mainly to utilize highly trained operators to obtain reproducible results.

Vis-NIRS is an Efficient Alternative to Routine KF Titration

To develop alternatives to the traditional KF titration, fifty-one fuel (ethanol-hydrocarbon blend) samples with varying moisture values obtained from coulometric KF-titration, and the correlation between changes in spectral data and moisture values were utilized to develop a prediction model. Of the 51 samples, 47 were used to develop a prediction model and four were used for external validation. The moisture content in the samples obtained were between 0.0026–1.2016% moisture. To predict the validity of using Vis-NIRS as an alternative to KF titration,

NIRS XDS RapidLiquid Analyzer was used to analyze the samples by collecting the spectra over the full wavelength range (400–2500 nm) at room temperature using quartz glass cuvettes with 2 mm path length. Vision Air 2.0 Complete software was used for acquiring and management of data as well as for the development of the quantification method.

Table 1. Used equipment and software.

Figure 1. The NIRS XDS RapidLiquid Analyzer was used for spectral data acquisition over the wavelength range from 400 nm to 2500 nm.

The results from the spectra of samples demonstrated by the correlation between increasing moisture content and increasing absorbance and the absorption bands at 1380–1670 nm and 1875–1990 nm of water were used to build a quantitative prediction model. Not only did the Partial Least Squares (PLS) model using 6 factors showed a high correlation between the provided reference values and the calculated values (R² = 0.999), it also showed low Standard Errors (SEC = 0.0097%, SECV = 0.0116% and SEP = 0.0120%). Taken together, Vis-NIRS can be used as a reliable and fast alternative to routine KF titration thereby saving both time and money.

Figure 2. Non pre-treated spectra of 47 ethanol / hydrocarbon blends samples in the region of 1300–2200 nm. The typical water bands around 1400 nm and 1900 nm are highlighted.

Figure 3. Correlation plot of the predicted moisture content Vis-NIRS versus the KF titration values. Displayed are the calibration (blue) and the validation data (turquoise).

Table 2. Results of the quantitative method development for moisture content.

Table 3. Comparison of moisture content of external validation samples determined by KF titration with Vis-NIR prediction.

References and Further Reading

1. http://www.mycleandiesel.com/
2. https://www.astm.org/Standards/D6304.htm

This information has been sourced, reviewed and adapted from materials provided by Metrohm AG.

For more information on this source, please visit Metrohm AG.