Method of determining petroleum hydrocarbon fractions in a sample

Abstract

The present invention relates to a method of determining petroleum hydrocarbon fractions (Cn) in a sample, the method including: inputting the sample into a chamber; emitting infrared light from an optical light source into the chamber with the sample; detecting at a detector a detected infrared light from the chamber; transforming the detected infrared light to a Fourier Transform Infrared (FTIR) spectrum of the sample at a processor, wherein the FTIR spectrum has wavenumbers between 4000 and 400 cm1; processing the FTIR spectrum to identify sub-bands each having at least one doublet of sub-band peaks at respective wavenumbers in a second derivative curve of the FTIR spectrum using a second derivation algorithm implemented by the processor; comparing the at least one doublet of sub-band peaks to data indicative of known doublets of sub-band peaks at known wavenumbers for petroleum hydrocarbon fractions in the FTIR spectrum to classify the petroleum hydrocarbon fractions in the sample; and determining a dominant petroleum hydrocarbon fraction of the petroleum hydrocarbon fractions in the sample based on a ratio of intensities of the sub-band peaks of the at least one doublet for each of the petroleum hydrocarbon fractions in the sample.

Claims

1. A method of determining petroleum hydrocarbon fractions (C.sub.n) in a sample, the method including: inputting the sample into a chamber; emitting infrared light from an optical light source into the chamber with the sample; detecting at a detector a detected infrared light from the chamber; transforming the detected infrared light to a Fourier Transform Infrared (FTIR) spectrum of the sample at a processor, wherein the FTIR spectrum has wavenumbers between 4000 and 400 cm.sup.1; processing the FTIR spectrum to identify sub-bands each having at least one doublet of sub-band peaks at respective wavenumbers in a second derivative curve of the FTIR spectrum using a second derivation algorithm implemented by the processor; comparing the at least one doublet of sub-band peaks to data indicative of known doublets of sub-band peaks at known wavenumbers for petroleum hydrocarbon fractions in the FTIR spectrum to classify the petroleum hydrocarbon fractions in the sample; and determining a dominant petroleum hydrocarbon fraction of the petroleum hydrocarbon fractions in the sample based on a ratio of intensities of the sub-band peaks of the at least one doublet for each of the petroleum hydrocarbon fractions in the sample; wherein the at least one doublet includes a first doublet of sub-band peaks at wavenumbers between 3000 and 2800 cm.sup.1.

2. A method of claim 1, wherein the first doublet of sub-band peaks is at wavenumbers 2954 and 2827 cm.sup.1.

3. A method of claim 2, wherein the at least one doublet further includes a second doublet of sub-band peaks at wavenumbers between 1500 and 1400 cm.sup.1.

4. A method of claim 3, wherein the second doublet of sub-band peaks is at wavenumbers between 1480 and 1450 cm.sup.1.

5. A method of claim 3, wherein the at least one doublet of sub-band peaks further includes a third doublet of sub-band peaks at wavenumbers between 750 and 700 cm.sup.1.

6. A method of claim 5, wherein the third doublet of sub-band peaks is at wavenumbers between 750 and 730 cm.sup.1.

7. A method of claim 5, wherein the sample is soil and the second doublet and the third doublet of sub-band peaks are obscured in the FTIR spectrum by sub-bands for components of the soil.

8. A method of claim 1 further including performing baseline correction of the FTIR spectrum using a baseline correction algorithm implemented by the processor to: locate points on the FTIR spectrum representing wavenumbers with low absorbance values corresponding to valleys in the FTIR spectrum; recursively drawing a new baseline from both sides of the valleys to sides of the FTIR spectrum using straight lines; disregard ones of the points on the FTIR spectrum with absorbance values lower than the straight lines; and generate a baseline corrected FTIR spectrum by connecting remaining ones of the points on the FTIR spectrum.

9. A method of claim 8, further including processing the baseline corrected FTIR spectrum to identify the sub-bands.

10. A method of claim 9, further including filtering the baseline corrected FTIR spectrum using a Gaussian filter algorithm implemented by the processor to remove ones of the sub-bands having sub-band valleys higher than a threshold value in the second derivative curve, wherein the Gaussian filter algorithm is a Gaussian low pass filter algorithm with a standard deviation of 1.5.

11. A method of claim 1, further including optimising identification of the sub-bands in the second derivative curve using an optimisation algorithm implemented by the processor to minimise a difference between a smoothed second derivative curve and the second derivative curve having the identified sub-bands, wherein the optimisation algorithm is a Monte Carlo algorithm.

12. An apparatus for determining petroleum hydrocarbon fractions (C.sub.n) in a sample, the apparatus including: a housing; a chamber disposed in the housing for inputting the sample therein; an optical light source disposed in the housing for emitting infrared light into the chamber with the sample; a detector for detecting a detected infrared light from the chamber; and a controller disposed in the housing having a processor and a memory in data communication with the processor, the controller being configured to: transform the detected infrared light to a Fourier Transform Infrared (FTIR) spectrum of the sample at a processor, wherein the FTIR spectrum has wavenumbers between 4000 and 400 cm.sup.1; process the FTIR spectrum to identify sub-bands each having at least one doublet of sub-band peaks at respective wavenumbers in a second derivative curve of the FTIR spectrum using a second derivation algorithm implemented by the processor; compare the at least one doublet of sub-band peaks to data indicative of known doublets of sub-band peaks at known wavenumbers for petroleum hydrocarbon fractions in the FTIR spectrum to classify the petroleum hydrocarbon fractions in the sample; and determine a dominant petroleum hydrocarbon fraction of the petroleum hydrocarbon fractions in the sample based on a ratio of intensities of the sub-band peaks of the at least one doublet for each of the petroleum hydrocarbon fractions in the sample; wherein the at least one doublet includes a first doublet of sub-band peaks at wavenumbers between 3000 and 2800 cm.sup.1.

13. An apparatus of claim 12, wherein the first doublet of sub-band peaks is at wavenumbers 2954 and 2827 cm.sup.1.

14. An apparatus of claim 13, wherein the at least one doublet further includes a second doublet of sub-band peaks at wavenumbers between 1500 and 1400 cm.sup.1.

15. An apparatus of claim 14, wherein the second doublet of sub-band peaks is at wavenumbers between 1480 and 1450 cm.sup.1.

16. An apparatus of claim 13, wherein the at least one doublet of sub-band peaks further includes a third doublet of sub-band peaks at wavenumbers between 750 and 700 cm.sup.1.

17. An apparatus of claim 16, wherein the third doublet of sub-band peaks is at wavenumbers between 750 and 730 cm.sup.1.

18. An apparatus of claim 16, wherein the sample is soil and the second doublet and the third doublet of sub-band peaks are obscured in the FTIR spectrum by sub-bands for components of the soil.

19. An apparatus of claim 18, wherein the apparatus is located near the sample.

Description

BRIEF DESCRIPTION OF DRAWINGS

(1) In order that the invention can be more clearly understood, examples of embodiments will now be described with reference to the accompanying drawings, in which:

(2) FIG. 1 shows a representation of an apparatus for determining petroleum hydrocarbon fractions (C.sub.n) in a sample in a sample, according to an embodiment of the invention;

(3) FIG. 2 shows a block diagram of a method of determining petroleum hydrocarbon fractions (C.sub.n) in a sample, according to an embodiment of the invention;

(4) FIGS. 3a-3d shows the FTIR spectrum for alkanes in a sample after baseline correction, obtained according to an embodiment of the invention;

(5) FIGS. 4a-4d shows the FTIR spectrum for selected alkanes, obtained according to an embodiment of the invention;

(6) FIG. 5 shows the absorbance area value related to concentrations of each of the selected alkanes of FIGS. 4a-4d;

(7) FIGS. 6a-6d shows the FTIR spectrum for a number of samples obtained according to an embodiment of the invention;

(8) FIG. 7 shows the baseline corrected FTIR spectra data for the samples of FIGS. 6a-6d; and

(9) FIG. 8 shows the FTIR spectra data for the samples of FIGS. 6a-6d at wavenumbers between 3000 and 2800 cm.sup.1.

DETAILED DESCRIPTION

(10) According to an embodiment of the present invention there is provided an apparatus 10 for determining petroleum hydrocarbon (PHC) fractions (C.sub.n) in a sample, as shown in FIG. 1. The apparatus 10 includes a housing 12, a chamber 14 disposed in the housing 12 for inputting the sample therein at a sample inlet 20. The sample inlet 20 can also be configured to remove the sample from the chamber 20. The housing 12 includes an optical light source 16 disposed in the housing 12 for emitting infrared light into the chamber 14 with the sample and a detector 18 for detecting a detected infrared light from the chamber (not shown is a controller disposed in the housing having a processor and a memory in data communication with the processor).

(11) The controller is configured to perform the following steps to determine petroleum hydrocarbon fractions (C.sub.n) in a sample, the steps including: transforming the detected infrared light to a Fourier Transform Infrared (FTIR) spectrum of the sample at a processor, wherein the FTIR spectrum has wavenumbers between 4000 and 400 cm.sup.1; processing the FTIR spectrum to identify sub-bands each having at least one doublet of sub-band peaks at respective wavenumbers in a second derivative curve of the FTIR spectrum using a second derivation algorithm implemented by the processor; comparing the at least one doublet of sub-band peaks to data indicative of known doublets of sub-band peaks at known wavenumbers for petroleum hydrocarbon fractions in the FTIR spectrum to classify the petroleum hydrocarbon fractions in the sample; and determining a dominant petroleum hydrocarbon fraction of the petroleum hydrocarbon fractions in the sample based on a ratio of intensities of the sub-band peaks of the at least one doublet for each of the petroleum hydrocarbon fractions in the sample.

(12) As mentioned, the hydrocarbon fractions (C.sub.n) are preferably alkanes and the sample could be soil. The apparatus 10 is preferably a handheld FTIR device, and inspects the infrared spectrum wavelength between 4000 and 400 cm.sup.1 using FTIR. As demonstrated in FIG. 3, which shows the output of apparatus 10 in the region of 4000 and 400 cm.sup.1, and in the regions where doublets are shown. That is, in the region of a first doublet of sub-band peaks at wavenumbers between 3000 and 2800 cm.sup.1, a region of a second doublet of sub-band peaks at wavenumbers between 1500 and 1400 cm.sup.1, and a region of a third doublet of sub-band peaks at wavenumbers between 750 and 700 cm.sup.1.

(13) The apparatus 10 was used, in the following examples, to demonstrate a rapid PHC fraction determination method for in-situ PHC assessment of a sample.

(14) These examples further illustrate aspects of the method 100 of FIG. 2 and the apparatus 10 of FIG. 1.

(15) In one example, pure chemicals of long chain alkanes: icosane (C.sub.20), hexacosane (C.sub.26), octacosane (C.sub.28), dotriacontane (C.sub.32) and heptatriacontane (C.sub.37) were obtained. The pure alkane chemicals were added into hexane (95%) to create 20 mL of 10 g/L standard solutions for C.sub.20 to C.sub.32, and 5 g/L for C.sub.37, respectively. It was observed that the alkanes with higher carbon chain numbers have lower a dissolution rate in hexane. 5 g/L C.sub.37 solutions needed to be prepared with the assistance of ultrasonic vibrations. The individual alkanes were spiked into potassium bromide (KBr) for characteristic sub-band identification.

(16) All the samples were measured in triplicate using a handheld FTIR, in the form of apparatus 10, with an 8 cm.sup.1 resolution, 32 sample scans, 64 background scans were co-added in the infrared 4000 to 600 cm.sup.1 region, at a scanning velocity of 2.5 kHz, and 255 beam energy. All measurements for the spiked soil samples were made in diffuse reflectance infrared spectroscopy (DRIFTS) mode, sample non-destructively. It should be mentioned that there are several levels of spectral resolution: from 8 to 2 cm.sup.1. The lower the number present, the higher resolutions, and the detection limits and spectral features for a given compound can be improved through higher resolutions which need high-resolution interferometers.

(17) However, the relative cost to incorporate such a system for field monitoring would be exorbitant. Further, it would take a considerable amount of time for a scan and obtain point data for computational analysis. The sensitivity level chosen for this experiment is appropriate for reduced downtime and rapid screening for field related applications. With setting the resolution at 8 cm.sup.1, one measurement can be completed within thirty seconds with 32 scans.

(18) Baseline Correction

(19) In this example, a baseline correction algorithm, using computational recursion, was developed and applied to automate baseline correction for MIR (4000 to 400 cm.sup.1) spectra. The algorithm developed will run through all the spectral details and locate the lowest valley; then the baseline can be drawn by connecting the valleys to each side of the spectrum using straight lines. If any part of the spectrum is intersected by the straight lines after baseline correction, the spectrum will create negative absorbance values. In this case, the algorithm will be recursively run for the spectral regions where the intersecting straight lines connected. This simple and rapid baseline correction algorithm can be applied to any IR spectral region, without any intervention. Furthermore, since all the PHCs have similar spectral features, the adopted baseline correction algorithm can give unified baseline corrected IR data for all of the PHCs of interest.

(20) Band Decomposition

(21) An automatic band decomposition algorithm was applied in this example. The sub-bands were established using a second derivation curve (SDC). The band number can be controlled by eliminating the number of SDC valleys. The dominant bands from the original spectrum were presented as the lower valleys in the SDC. The small valleys, representing secondary bands, could be eliminated using a Gaussian low pass filter, with a standard deviation 1.5. The spectrum band can be properly decomposed using Gaussian curves, which the amplitude, width, and location were optimized using Monte Carlo algorithm (MCA), which is a heuristic algorithm based on randomness and statistics to get an optimisation result.

(22) Thus, in the example, the infrared spectral data is first processed by automatically baseline correcting the IR spectra data and band decomposing the IR spectra without visual inspection. Quantitative analysis for each alkane was investigated using the decomposed sub-bands. Orthogonal experimental design (OED) was applied to generate the alkane mixtures with designed heterogeneous concentrations.

(23) Results

(24) The measurements for alkane classification in the example was conducted using four different alkanes, C.sub.20, C.sub.26, C.sub.32 and C.sub.37 with the concentration of 5000 ppm for each alkane, respectively. To prepare these standards, the pure alkane chemicals were diluted with potassium bromide (KBr). The four alkanes' spectra after background subtraction and baseline correction are shown in FIG. 3.a. According to the similarity of the IR bands in the entire region indicates the alkanes have identical molecular structures. The doublet observed between 3000 and 2800 cm.sup.1, which is the evidence of carbon hydrogen bonding for long-chain alkanes due to \\C\\H stretching vibrations. From the FIG. 3.b, it is found there were two bands coherent to the doublet at location 2954 and 2872 cm.sup.1, respectively.

(25) From C.sub.20 to C.sub.37, by the increase of the C.sub.n, the intensities of the two bands were reduced as demonstrated. This phenomenon matched similar results obtained from the Raman spectrum data of C.sub.8 and C.sub.20. In the Raman spectrum data, both the spectra of C.sub.8 and C.sub.20 contained all the similar bands, including the bands at location 2954 and 2872 cm.sup.1. It is observed that the band intensities of the C.sub.20 were less than C.sub.8 at these two locations. On the opposite, there were another two doublets existed in the region from 1480 to 1450 cm.sup.1 and the region at 750 to 730 cm.sup.1. It was observed that the intensity of one coherent band at each of these regions was increased following the increase on the C.sub.n. As shown in FIGS. 3.c and d, the band centre at 1462 and 730 was increased from C.sub.20 to C.sub.37. As mentioned, the band near 730 cm.sup.1 can be seen only in long chain alkanes. The phenomena of the coherent bands in the three regions can be applied for C.sub.n classification.

(26) For the quantification studies, three different concentration levels of each selected alkane were mixed with KBr, to generate the calibration standards of 500, 5000 and 10,000 ppm for C.sub.20, C.sub.26, C.sub.32 and 500, 1000, 2500 and 5000 ppm for alkane C.sub.37. All the calibration spectra of each selected alkane, including the baseline corrected and band decomposed spectra data, are demonstrated in FIG. 4. It is noted that the increasing concentration did significantly affect the absorbance magnitudes. Accordingly, based on the measurements, it was difficult to identify the characteristic bands when the alkanes present were less than 500 ppm for each alkane. In an ideal scenario when only a single alkane presented, the calibration for each alkane can be applied using the bands at this IR region after baseline correction. However, this ideal scenario does not fit into the petroleum products. For example, crude oil can consist of hundreds of individual petroleum hydrocarbon compounds. For the mixtures, the calibration should vary according to the characteristic bands of the sub-bands, after the decomposition, as shown in FIG. 4 for each alkane.

(27) The calibration results of the four alkanes are illustrated in FIG. 5. Comparatively, all alkanes provided reasonable predictions when linking the absorbance area values to the spiked concentrations. FIG. 5 provides evidence that the concentrations were able to be predicted using the calibration curves generated with the single alkane standards.

(28) In order to study the mixture scenario, orthogonal experimental design (OED) was applied to generate the alkane mixtures with designed heterogeneous concentrations. In an example, an orthogonal design table (ODT) was generated and contained nine synthetic mixtures of the selected alkanes, with three concentration levels mentioned before.

(29) The mixtures were also prepared using an artificial soil containing 80% of quartz, 5% of humic acid and 15% of kaolinite. The details are listed in Table 1 and the spectral data for these nine mixtures are shown in FIG. 6. From FIG. 6.b, it is observed the minor band at 2873 cm.sup.1 which indicates the amount of C.sub.20. For example, the band exists in the spectrum of OED 8 rather than OED 9, demonstrated with black continuous line and dash line, respectively. After band decomposition, the length carbon chains, EC.sub.n can be indicated with the ratio of band 2850 cm.sup.1 versus band 2956 cm.sup.1. Additionally, EC.sub.n can also be indicated by the ratio of the other two doublets at the regions 1489 to 1450 cm.sup.1 (FIG. 6.c) and region 750 to 700 cm.sup.1 (FIG. 6.d).

(30) For example, the mixture OED 2 and 8, the proportion of shorter carbon chain (C.sub.20 and C.sub.26) were higher than the longer carbon chain (C.sub.32 and C.sub.37). It is observed the bands of 1470, and 715 cm.sup.1 were suppressed of 1460 and 730 cm.sup.1, respectively, as shown in FIG. 6.c and FIG. 6.d. These suppress phenomena are not shown in other spectral data from more homogeneous mixtures. The total concentration of the alkane mixture can be determined using absorbance area of a region below 3000 to 2800 cm.sup.1, as the EPA method 8440. In a mixture scenario, it is difficult to identify the specified C.sub.n from C.sub.20 to C.sub.37. Fortunately, the proportion of the shorter carbon chain (C.sub.20 and C.sub.26) and the longer carbon chain (C.sub.32 and C.sub.37) can be identified using the ratios as mentioned above.

(31) To further validate the determination concept, 100 mg of four different petroleum products, were spiked separately into 1 g of the artificial soil. The details of the products are listed in Table 2.

(32) The products include petrol, kerosene, diesel, motor/lubricating oils, and grease wax. The density of each product was measured by weighing samples in 50 mL volume containers. From Table 2, as expected, the densities increased with the carbon chain numbers. All samples were prepared fresh daily, and measurements were carried out at room temperature (22 C.) in triplicate, and the average values are presented. As mentioned the artificial soil containing 80% of quartz, 5% of humic acid and 15% of kaolinite. It is observed the characteristic band at 804.5 cm.sup.1, indicating the Si\\O bending vibrations for quartz identification. Kaolinite can be identified with the strong O\\H stretching vibration presenting as the doublets at 3690 to 3620 cm.sup.1. Humic acid can be identified as the small bands like noise covered around 3500 cm.sup.1 and water content band central at around 3400 cm.sup.1. According to the spectrum in FIG. 7, the signal in below 2000 cm.sup.1 was dominated by quartz, and two of the hydrocarbon identical band regions at 740 to 710 cm.sup.1 and 1480 to 1450 cm.sup.1 were submerged in the dominated bands. Therefore, to avoid the interference from soil chemical components, the bands at 3000 to 2800 cm.sup.1 can be applied to the alkane studies.

(33) FIG. 8 demonstrated the bands of the four products in the region. As mentioned the band at 2950 cm.sup.1 increases adversely with the C.sub.n from C.sub.20 to C.sub.37. By comparison, it is clear to observe that a higher intensive band at 2950 cm.sup.1 existed in IR spectrum of diesel and kerosene, rather than in the lubricating oil and grease wax. The study has shown that it is possible to determine the alkanes with different C.sub.n using their most presented characteristic bands. The phenomena of the three regions, 2954 to 2872 cm.sup.1, 1480 to 1450 cm.sup.1 and 750 to 730 cm.sup.1 can be applied for C.sub.n identification. However, unlike applying pure alkane chemicals to a KBr background. When a petroleum product is applied to soil, the soil IR spectrum will submerge the characteristic bands below 2000 cm.sup.1, due to the quartz and other soil components. Hence, for a soil sample, the first IR region mentioned above should be applied majorly to the alkane studies. Also, using this method, it is possible to determine the dominant fraction of C.sub.n in soil. However, the method may not be able to determinate each alkane in heterogeneous mixtures, especially in a complex background, such as a neutral soil.

(34) Referring back to FIG. 2, there is a shown a flow chart of a method 100 of determining petroleum hydrocarbon fractions (C.sub.n) in a sample, the method 10 including the steps of inputting 102 the sample into a chamber, emitting infrared light from an optical light source into the chamber with the sample, and detecting at a detector a detected infrared light from the chamber; transforming 104 the detected infrared light to a Fourier Transform Infrared (FTIR) spectrum of the sample at a processor, wherein the FTIR spectrum has wavenumbers between 4000 and 400 cm.sup.1; processing 106 the FTIR spectrum to identify sub-bands each having at least one doublet of sub-band peaks at respective wavenumbers in a second derivative curve of the FTIR spectrum using a second derivation algorithm implemented by the processor; comparing 108 the at least one doublet of sub-band peaks to data indicative of known doublets of sub-band peaks at known wavenumbers for petroleum hydrocarbon fractions in the FTIR spectrum to classify the petroleum hydrocarbon fractions in the sample; and determining 110 a dominant petroleum hydrocarbon fraction of the petroleum hydrocarbon fractions in the sample based on a ratio of intensities of the sub-band peaks of the at least one doublet for each of the petroleum hydrocarbon fractions in the sample.

(35) It will be appreciated by those persons skilled in the art that further aspects of the method 100 will be apparent from the above description of the apparatus 10 and the examples. Further, the person skilled in the art will also appreciate that at least part of the method could be embodied in program code that implemented by a processor of the apparatus 10. The program code could be supplied in a number of ways, for example on a memory.

(36) Those skilled in the art will also appreciate that the invention described herein is susceptible to variations and modifications other than those specifically described. It is to be understood that the invention includes all such variations and modifications.

(37) TABLE-US-00001 TABLE 1 Orthogonal experimental design table (Unit: ppm or mg/kg). OED C.sub.20 C.sub.26 C.sub.32 C.sub.37 Total 1 10,000 5,000 10,000 500 25,500 2 10,000 10,000 500 2,500 23,000 3 5,000 500 10,000 2,500 18,000 4 5,000 10,000 5,000 500 20,500 5 5,000 5,000 500 5,000 15,500 6 500 10,000 10,000 5,000 25,500 7 500 500 500 500 2,000 8 10,000 500 5,000 5,000 20,500 9 500 5,000 5,000 2,500 13,000

(38) TABLE-US-00002 TABLE 2 Selected petroleum products Carbon Product type Product Name chains Density Kerosene Diggers C.sub.10 to C.sub.18 0.76 Diesel Caltex C.sub.19 to EC.sub.30 0.81 Lubricating Oil Gear Oil, 85W-140, Penrite >EC.sub.30 0.90 Grease Wax Grease Valpex, Valvoline >EC.sub.30 0.97