Using NIRS to Analyze Petroleum Products (e.g. TAN, Aromatics, Cetane Index, Sulfur)

Fractional distillation or the refining process of crude oil is integral to many different products, which play an important role in everyday life. For instance, refinery products are typically used as raw materials for regular items or they are utilized as a source of energy. Fuels like diesel, kerosene and fuel oil are a few products that are obtained from crude oil.

Different parameters of fuels such as bromine number, the sulfur contents and aromatics, the cetane index, and the total acid number (TAN) are subjected to rigorous control to ensure that motors operate at their optimum level and the overall product quality is also enhanced. With the aid of near infrared spectroscopy (NIRS), users can easily and quickly determine the critical parameters during the production process and can achieve instant results. In this article, the determination of aromatics, cetane index, sulfur, TAN, and bromine number in various refinery products is described in detail.

Experimental Procedure

In this experiment, a NIRS XDS RapidLiquid analyzer equipped with 8mm-diameter disposable glass vials was used for testing 71 samples of petrochemical products (Table 1, Figure 1).

Figure 1. NIRS XDS RapidLiquid Analyzer (RLA)

Table 1. Used equipment

Table 2. shows five different parameters that were examined in the 71 samples. The samples were collected as raw products such as gas, oil, coker-kerosene, EM, diesel, coker- diesel, and kerosene from the production process, or as final products like HEL and JA. Every week, the samples were collected for a total of 34 days, and transmission measurements were carried out at 35°C temperature in the NIRS XDS system with the equilibration time set at 30s. The customer's reference lab provided the reference values. The analyzed sample parameters are summarized in Table 2.

Table 2. Overview of analyzed sample parameters

Method Development and Results

A quantitative model was developed for each parameter.

Cetane index: For the cetane index, reference values were given for 65 of the 71 samples. These values remained in the range of 39.7 to 57.2, and were spread across the range. The parameters employed for the method development are shown in Table 3.

Table 3. Parameters of the calibration of the cetane index

Except for the VIS range, the entire NIR region was leveraged for calibration development. A 7-factor PLS model was based on the second derivative data of the 65 samples. The NIR data against the reference data is depicted in Figure 2.

Figure 2. Calibration set for the cetane Index -NIR data vs. reference data.

In the case of PLS method development for the cetane index, a link can be found between the reference values and the NIR, and this was defined by R² = 0.9883 and a SEC = 0.56. A SECV = 0.65 was used to cross-validate the calibration.

Bromine number: For the bromine number, reference values were given for 31 of the 71 samples, and these values were in the range of 0.29 to 47.2g/100 g. The parameters used for the method development is shown in Table 4.

Table 4. Parameters of the calibration of the bromine number

Excluding the VIS range, the entire NIR region was leveraged for calibration development to determine the bromine number. A 5-factor PLS model was based on the second derivative spectral data of the 31 samples. The NIR data against reference data is shown in Figure 3.

Figure 3. Calibration set for bromine number -NIR data vs. reference data.

In the case of the PLS method development for the bromine number, a link can be seen between the reference values and the NIR data, and this was defined by R² = 0.9831 and a SEC = 2.1. A SECV = 2.7 was used to cross-validate the calibration. According to the type of product, the bromine number of different samples was divided into four major groups. The coker samples, in particular, have relatively high bromine numbers in comparison to samples of gasoil, diesel and kerosene.

TAN: TAN reference values were given for 40 of the 71 samples, and these values remained in the range of 0.01 to 2.79mg KOH/g. The parameters used for the method development is shown in Table 5.

Table 5. Parameters of the calibration of the TAN

Excluding the VIS range, the entire NIR region was utilized for calibration development to determine the TAN. A 6 factor-PLS model was based on the second derivative data of the 40 samples. The NIR data against the reference data is illustrated in Figure 4.

Figure 4. Calibration set for the TAN - NIR data vs. reference data.

In the case of the PLS method development for the TAN, an association was seen between the reference values and the NIR data, and this was defined by R² = 0.9004 and a SEC = 0.248. A SECV = 0.308 was used to cross-validate the calibration.

The TAN values of different samples were divided into two major groups. To achieve better calibration and an improved robust technique, more number of samples of TAN values greater than 1.5mg KOH/g should be included in the method development.

Aromatics content: For the aromatics content, reference values were provided for 26 of the 71 samples, and these value were in the range of 15.9 to 31.3 wt%. The parameters used for the method development is shown in Table 6.

Table 6. Parameters of the calibration of the aromatics content

Excluding the VIS range, the entire NIR region was utilized for calibration development to determine the aromatics content. A 4-PLS model was based on the second derivative data of the 26 samples. The NIR data against the reference data is shown in Figure 5.

Figure 5. Calibration set for aromatics content -NIR vs. reference data.

In the case of the PLS method development for the aromatics content, a link was found between the reference values and the NIR data and this was defined by R² = 0.9633 and a SEC = 0.849. A SECV = 0.990 was used to cross-validate the calibration.

Sulfur content: For the sulfur content, reference values were given for 47 of the 71 samples. Since the end products have greater than 0.05 wt% of sulfur content, which is less than the detection limit of NIRS, just 35 samples were employed for the method development. The parameters of the method development are shown in Table 7.

Table 7. Parameter of method development for the calibration for the sulfur content

Except for the VIS range, calibration of the sulfur content was developed using the entire NIR region. A 5-factor PLS model was based on the second derivative data of the 35 samples. NIR data against the reference data is shown in Figure 6.

Figure 6. Calibration set for sulfur content - NIR vs. reference data.

For the sulfur content, the PLS method development showed a link between the reference values and NIR, and this was defined by R² = 0.9719 and a SEC = 0.035. A SECV = 0.040 was used to cross-validate the calibration.

Conclusion

From the results, it is evident that different critical parameters of refinery products such as the cetane index, bromine number, aromatics, TAN and sulfur content can be quantitatively analyzed using the advanced NIR XDS RLA analyzer.

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

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